STM32F405/415, STM32F407/417, STM32F427/437 And STM32F429/439 Advanced Arm® Based 32 Bit MCUs Reference Manual Cortex

User Manual:

Open the PDF directly: View PDF .
Page Count: 1747

Download
Open PDF In Browser	View PDF

RM0090
Reference manual
STM32F405/415, STM32F407/417, STM32F427/437 and
STM32F429/439 advanced Arm®-based 32-bit MCUs
Introduction
This reference manual targets application developers. It provides complete information on
how to use the STM32F405xx/07xx, STM32F415xx/17xx, STM32F42xxx and
STM32F43xxx microcontroller memory and peripherals.
The STM32F405xx/07xx, STM32F415xx/17xx, STM32F42xxx and STM32F43xxx constitute
a family of microcontrollers with different memory sizes, packages and peripherals.
For ordering information, mechanical and electrical device characteristics please refer to the
datasheets.
For information on the Arm® Cortex®-M4 with FPU core, please refer to the Cortex®-M4 with
FPU Technical Reference Manual.

Related documents
Available from STMicroelectronics web site (http://www.st.com):
• STM32F40x and STM32F41x datasheets
• STM32F42x and STM32F43x datasheets
• For information on the Arm® Cortex®-M4 with FPU, refer to the STM32F3xx/F4xxx
Cortex®-M4 with FPU programming manual (PM0214).

June 2018

RM0090 Rev 17

1/1747
www.st.com

1

Contents

RM0090

Contents
1

2

Documentation conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
1.1

List of abbreviations for registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

1.2

Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

1.3

Peripheral availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

Memory and bus architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
2.1

2.1.1

I-bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

2.1.2

D-bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

2.1.3

S-bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

2.1.4

DMA memory bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

2.1.5

DMA peripheral bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

2.1.6

Ethernet DMA bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

2.1.7

USB OTG HS DMA bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

2.1.8

LCD-TFT controller DMA bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

2.1.9

DMA2D bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

2.1.10

BusMatrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

2.1.11

AHB/APB bridges (APB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

2.2

Memory organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

2.3

Memory map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

2.4

3

System architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

2.3.1

Embedded SRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

2.3.2

Flash memory overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

2.3.3

Bit banding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

Boot configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

Embedded Flash memory interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

3.2

Main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

3.3

Embedded Flash memory in
STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . . . . . . . . . . 74

3.4

Embedded Flash memory in STM32F42xxx and STM32F43xxx . . . . . . . 76

3.5

Read interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
3.5.1

2/1747

Relation between CPU clock frequency and Flash memory read time . 80

RM0090 Rev 17

RM0090

Contents
3.5.2

3.6

3.7

4

Adaptive real-time memory accelerator (ART Accelerator™) . . . . . . . . 82

Erase and program operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
3.6.1

Unlocking the Flash control register . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

3.6.2

Program/erase parallelism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

3.6.3

Erase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

3.6.4

Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

3.6.5

Read-while-write (RWW) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

3.6.6

Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

Option bytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
3.7.1

Description of user option bytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

3.7.2

Programming user option bytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92

3.7.3

Read protection (RDP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

3.7.4

Write protections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

3.7.5

Proprietary code readout protection (PCROP) . . . . . . . . . . . . . . . . . . . 95

3.8

One-time programmable bytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

3.9

Flash interface registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
3.9.1

Flash access control register (FLASH_ACR)
for STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . . . . . . 98

3.9.2

Flash access control register (FLASH_ACR)
for STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . . 99

3.9.3

Flash key register (FLASH_KEYR) . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

3.9.4

Flash option key register (FLASH_OPTKEYR) . . . . . . . . . . . . . . . . . . 100

3.9.5

Flash status register (FLASH_SR) for
STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . . . . . . . . 101

3.9.6

Flash status register (FLASH_SR) for
STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

3.9.7

Flash control register (FLASH_CR) for
STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . . . . . . . . 103

3.9.8

Flash control register (FLASH_CR) for
STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

3.9.9

Flash option control register (FLASH_OPTCR) for
STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . . . . . . . . 106

3.9.10

Flash option control register (FLASH_OPTCR)
for STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . 108

3.9.11

Flash option control register (FLASH_OPTCR1)
for STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . 110

3.9.12

Flash interface register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

CRC calculation unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113

RM0090 Rev 17

3/1747
39

Contents

5

RM0090

4.1

CRC introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113

4.2

CRC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113

4.3

CRC functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .114

4.4

CRC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .114
Data register (CRC_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

4.4.2

Independent data register (CRC_IDR) . . . . . . . . . . . . . . . . . . . . . . . . 114

4.4.3

Control register (CRC_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

4.4.4

CRC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

Power controller (PWR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
5.1

5.2

5.3

5.4

5.5

4/1747

4.4.1

Power supplies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .116
5.1.1

Independent A/D converter supply and reference voltage . . . . . . . . . . 117

5.1.2

Battery backup domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

5.1.3

Voltage regulator for STM32F405xx/07xx and STM32F415xx/17xx . . 120

5.1.4

Voltage regulator for STM32F42xxx and STM32F43xxx . . . . . . . . . . . 121

Power supply supervisor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
5.2.1

Power-on reset (POR)/power-down reset (PDR) . . . . . . . . . . . . . . . . . 124

5.2.2

Brownout reset (BOR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

5.2.3

Programmable voltage detector (PVD) . . . . . . . . . . . . . . . . . . . . . . . . 125

Low-power modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
5.3.1

Slowing down system clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128

5.3.2

Peripheral clock gating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128

5.3.3

Sleep mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

5.3.4

Stop mode (STM32F405xx/07xx and STM32F415xx/17xx) . . . . . . . . 130

5.3.5

Stop mode (STM32F42xxx and STM32F43xxx) . . . . . . . . . . . . . . . . . 133

5.3.6

Standby mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136

5.3.7

Programming the RTC alternate functions to wake up the device from
the Stop and Standby modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138

Power control registers (STM32F405xx/07xx and STM32F415xx/17xx) 141
5.4.1

PWR power control register (PWR_CR)
for STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . . . . . 141

5.4.2

PWR power control/status register (PWR_CSR)
for STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . . . . . 142

Power control registers (STM32F42xxx and STM32F43xxx) . . . . . . . . . 144
5.5.1

PWR power control register (PWR_CR)
for STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . 144

5.5.2

PWR power control/status register (PWR_CSR)
for STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . 147
RM0090 Rev 17

RM0090

Contents

5.6

6

PWR register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

Reset and clock control for
STM32F42xxx and STM32F43xxx (RCC) . . . . . . . . . . . . . . . . . . . . . . . 150
6.1

6.2

6.3

Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
6.1.1

System reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150

6.1.2

Power reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150

6.1.3

Backup domain reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
6.2.1

HSE clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

6.2.2

HSI clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

6.2.3

PLL configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

6.2.4

LSE clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156

6.2.5

LSI clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156

6.2.6

System clock (SYSCLK) selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156

6.2.7

Clock security system (CSS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

6.2.8

RTC/AWU clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

6.2.9

Watchdog clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158

6.2.10

Clock-out capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158

6.2.11

Internal/external clock measurement using TIM5/TIM11 . . . . . . . . . . . 158

RCC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
6.3.1

RCC clock control register (RCC_CR) . . . . . . . . . . . . . . . . . . . . . . . . . 161

6.3.2

RCC PLL configuration register (RCC_PLLCFGR) . . . . . . . . . . . . . . . 163

6.3.3

RCC clock configuration register (RCC_CFGR) . . . . . . . . . . . . . . . . . 165

6.3.4

RCC clock interrupt register (RCC_CIR) . . . . . . . . . . . . . . . . . . . . . . . 167

6.3.5

RCC AHB1 peripheral reset register (RCC_AHB1RSTR) . . . . . . . . . . 170

6.3.6

RCC AHB2 peripheral reset register (RCC_AHB2RSTR) . . . . . . . . . . 173

6.3.7

RCC AHB3 peripheral reset register (RCC_AHB3RSTR) . . . . . . . . . . 174

6.3.8

RCC APB1 peripheral reset register (RCC_APB1RSTR) . . . . . . . . . . 174

6.3.9

RCC APB2 peripheral reset register (RCC_APB2RSTR) . . . . . . . . . . 178

6.3.10

RCC AHB1 peripheral clock register (RCC_AHB1ENR) . . . . . . . . . . . 180

6.3.11

RCC AHB2 peripheral clock enable register (RCC_AHB2ENR) . . . . . 182

6.3.12

RCC AHB3 peripheral clock enable register (RCC_AHB3ENR) . . . . . 183

6.3.13

RCC APB1 peripheral clock enable register (RCC_APB1ENR) . . . . . 183

6.3.14

RCC APB2 peripheral clock enable register (RCC_APB2ENR) . . . . . 187

6.3.15

RCC AHB1 peripheral clock enable in low power mode register
(RCC_AHB1LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189

RM0090 Rev 17

5/1747
39

Contents

7

RM0090
RCC AHB2 peripheral clock enable in low power mode register
(RCC_AHB2LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192

6.3.17

RCC AHB3 peripheral clock enable in low power mode register
(RCC_AHB3LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193

6.3.18

RCC APB1 peripheral clock enable in low power mode register
(RCC_APB1LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193

6.3.19

RCC APB2 peripheral clock enabled in low power mode register
(RCC_APB2LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

6.3.20

RCC Backup domain control register (RCC_BDCR) . . . . . . . . . . . . . . 199

6.3.21

RCC clock control & status register (RCC_CSR) . . . . . . . . . . . . . . . . 200

6.3.22

RCC spread spectrum clock generation register (RCC_SSCGR) . . . . 202

6.3.23

RCC PLLI2S configuration register (RCC_PLLI2SCFGR) . . . . . . . . . 203

6.3.24

RCC PLL configuration register (RCC_PLLSAICFGR) . . . . . . . . . . . . 206

6.3.25

RCC Dedicated Clock Configuration Register (RCC_DCKCFGR) . . . 207

6.3.26

RCC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210

Reset and clock control for
STM32F405xx/07xx and STM32F415xx/17xx(RCC) . . . . . . . . . . . . . . 213
7.1

7.2

7.3

6/1747

6.3.16

Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
7.1.1

System reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213

7.1.2

Power reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214

7.1.3

Backup domain reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214

Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
7.2.1

HSE clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217

7.2.2

HSI clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218

7.2.3

PLL configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219

7.2.4

LSE clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219

7.2.5

LSI clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220

7.2.6

System clock (SYSCLK) selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220

7.2.7

Clock security system (CSS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220

7.2.8

RTC/AWU clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221

7.2.9

Watchdog clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221

7.2.10

Clock-out capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222

7.2.11

Internal/external clock measurement using TIM5/TIM11 . . . . . . . . . . . 222

RCC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
7.3.1

RCC clock control register (RCC_CR) . . . . . . . . . . . . . . . . . . . . . . . . . 224

7.3.2

RCC PLL configuration register (RCC_PLLCFGR) . . . . . . . . . . . . . . . 226

7.3.3

RCC clock configuration register (RCC_CFGR) . . . . . . . . . . . . . . . . . 228

RM0090 Rev 17

RM0090

8

Contents
7.3.4

RCC clock interrupt register (RCC_CIR) . . . . . . . . . . . . . . . . . . . . . . . 230

7.3.5

RCC AHB1 peripheral reset register (RCC_AHB1RSTR) . . . . . . . . . . 233

7.3.6

RCC AHB2 peripheral reset register (RCC_AHB2RSTR) . . . . . . . . . . 236

7.3.7

RCC AHB3 peripheral reset register (RCC_AHB3RSTR) . . . . . . . . . . 237

7.3.8

RCC APB1 peripheral reset register (RCC_APB1RSTR) . . . . . . . . . . 237

7.3.9

RCC APB2 peripheral reset register (RCC_APB2RSTR) . . . . . . . . . . 240

7.3.10

RCC AHB1 peripheral clock enable register (RCC_AHB1ENR) . . . . . 242

7.3.11

RCC AHB2 peripheral clock enable register (RCC_AHB2ENR) . . . . . 244

7.3.12

RCC AHB3 peripheral clock enable register (RCC_AHB3ENR) . . . . . 245

7.3.13

RCC APB1 peripheral clock enable register
(RCC_APB1ENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245

7.3.14

RCC APB2 peripheral clock enable register (RCC_APB2ENR) . . . . . 248

7.3.15

RCC AHB1 peripheral clock enable in low power mode register
(RCC_AHB1LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250

7.3.16

RCC AHB2 peripheral clock enable in low power mode register
(RCC_AHB2LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252

7.3.17

RCC AHB3 peripheral clock enable in low power mode register
(RCC_AHB3LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253

7.3.18

RCC APB1 peripheral clock enable in low power mode register
(RCC_APB1LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254

7.3.19

RCC APB2 peripheral clock enabled in low power mode
register (RCC_APB2LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257

7.3.20

RCC Backup domain control register (RCC_BDCR) . . . . . . . . . . . . . . 259

7.3.21

RCC clock control & status register (RCC_CSR) . . . . . . . . . . . . . . . . 260

7.3.22

RCC spread spectrum clock generation register (RCC_SSCGR) . . . . 262

7.3.23

RCC PLLI2S configuration register (RCC_PLLI2SCFGR) . . . . . . . . . 263

7.3.24

RCC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265

General-purpose I/Os (GPIO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
8.1

GPIO introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267

8.2

GPIO main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267

8.3

GPIO functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
8.3.1

General-purpose I/O (GPIO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269

8.3.2

I/O pin multiplexer and mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270

8.3.3

I/O port control registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274

8.3.4

I/O port data registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274

8.3.5

I/O data bitwise handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274

8.3.6

GPIO locking mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274

8.3.7

I/O alternate function input/output . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
RM0090 Rev 17

7/1747
39

Contents

RM0090

8.4

9

8/1747

8.3.8

External interrupt/wakeup lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275

8.3.9

Input configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275

8.3.10

Output configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276

8.3.11

Alternate function configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277

8.3.12

Analog configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278

8.3.13

Using the OSC32_IN/OSC32_OUT pins as GPIO PC14/PC15
port pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278

8.3.14

Using the OSC_IN/OSC_OUT pins as GPIO PH0/PH1 port pins . . . . 278

8.3.15

Selection of RTC_AF1 and RTC_AF2 alternate functions . . . . . . . . . . 279

GPIO registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
8.4.1

GPIO port mode register (GPIOx_MODER) (x = A..I/J/K) . . . . . . . . . . 281

8.4.2

GPIO port output type register (GPIOx_OTYPER)
(x = A..I/J/K) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281

8.4.3

GPIO port output speed register (GPIOx_OSPEEDR)
(x = A..I/J/K) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282

8.4.4

GPIO port pull-up/pull-down register (GPIOx_PUPDR)
(x = A..I/J/K) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282

8.4.5

GPIO port input data register (GPIOx_IDR) (x = A..I/J/K) . . . . . . . . . . 283

8.4.6

GPIO port output data register (GPIOx_ODR) (x = A..I/J/K) . . . . . . . . 283

8.4.7

GPIO port bit set/reset register (GPIOx_BSRR) (x = A..I/J/K) . . . . . . . 284

8.4.8

GPIO port configuration lock register (GPIOx_LCKR)
(x = A..I/J/K) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284

8.4.9

GPIO alternate function low register (GPIOx_AFRL) (x = A..I/J/K) . . . 285

8.4.10

GPIO alternate function high register (GPIOx_AFRH)
(x = A..I/J) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286

8.4.11

GPIO register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287

System configuration controller (SYSCFG) . . . . . . . . . . . . . . . . . . . . 289
9.1

I/O compensation cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289

9.2

SYSCFG registers for STM32F405xx/07xx and STM32F415xx/17xx . . 289
9.2.1

SYSCFG memory remap register (SYSCFG_MEMRMP) . . . . . . . . . . 289

9.2.2

SYSCFG peripheral mode configuration register (SYSCFG_PMC) . . 290

9.2.3

SYSCFG external interrupt configuration register 1
(SYSCFG_EXTICR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291

9.2.4

SYSCFG external interrupt configuration register 2
(SYSCFG_EXTICR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291

9.2.5

SYSCFG external interrupt configuration register 3
(SYSCFG_EXTICR3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292

9.2.6

SYSCFG external interrupt configuration register 4
(SYSCFG_EXTICR4) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293

RM0090 Rev 17

RM0090

Contents

9.3

10

9.2.7

Compensation cell control register (SYSCFG_CMPCR) . . . . . . . . . . . 293

9.2.8

SYSCFG register maps for
STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . . . . . . . . 294

SYSCFG registers for STM32F42xxx and STM32F43xxx . . . . . . . . . . . 294
9.3.1

SYSCFG memory remap register (SYSCFG_MEMRMP) . . . . . . . . . . 294

9.3.2

SYSCFG peripheral mode configuration register (SYSCFG_PMC) . . 296

9.3.3

SYSCFG external interrupt configuration register 1
(SYSCFG_EXTICR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297

9.3.4

SYSCFG external interrupt configuration register 2
(SYSCFG_EXTICR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298

9.3.5

SYSCFG external interrupt configuration register 3
(SYSCFG_EXTICR3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298

9.3.6

SYSCFG external interrupt configuration register 4
(SYSCFG_EXTICR4) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299

9.3.7

Compensation cell control register (SYSCFG_CMPCR) . . . . . . . . . . . 300

9.3.8

SYSCFG register maps for STM32F42xxx and STM32F43xxx . . . . . . 301

DMA controller (DMA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
10.1

DMA introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302

10.2

DMA main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302

10.3

DMA functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
10.3.1

General description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304

10.3.2

DMA transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306

10.3.3

Channel selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307

10.3.4

Arbiter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308

10.3.5

DMA streams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309

10.3.6

Source, destination and transfer modes . . . . . . . . . . . . . . . . . . . . . . . 309

10.3.7

Pointer incrementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312

10.3.8

Circular mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313

10.3.9

Double buffer mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313

10.3.10 Programmable data width, packing/unpacking, endianess . . . . . . . . . 314
10.3.11 Single and burst transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
10.3.12 FIFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
10.3.13 DMA transfer completion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
10.3.14 DMA transfer suspension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
10.3.15 Flow controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
10.3.16 Summary of the possible DMA configurations . . . . . . . . . . . . . . . . . . . 322
10.3.17 Stream configuration procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322

RM0090 Rev 17

9/1747
39

Contents

RM0090
10.3.18 Error management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323

10.4

DMA interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324

10.5

DMA registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
10.5.1

DMA low interrupt status register (DMA_LISR) . . . . . . . . . . . . . . . . . . 325

10.5.2

DMA high interrupt status register (DMA_HISR) . . . . . . . . . . . . . . . . . 326

10.5.3

DMA low interrupt flag clear register (DMA_LIFCR) . . . . . . . . . . . . . . 327

10.5.4

DMA high interrupt flag clear register (DMA_HIFCR) . . . . . . . . . . . . . 327

10.5.5

DMA stream x configuration register (DMA_SxCR) (x = 0..7) . . . . . . . 328

10.5.6

DMA stream x number of data register (DMA_SxNDTR) (x = 0..7) . . . 331

10.5.7

DMA stream x peripheral address register (DMA_SxPAR) (x = 0..7) . 332

10.5.8

DMA stream x memory 0 address register (DMA_SxM0AR) (x = 0..7) 332

10.5.9

DMA stream x memory 1 address register (DMA_SxM1AR) (x = 0..7) 332

10.5.10 DMA stream x FIFO control register (DMA_SxFCR) (x = 0..7) . . . . . . 333
10.5.11 DMA register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335

11

Chrom-Art Accelerator™ controller (DMA2D) . . . . . . . . . . . . . . . . . . 339
11.1

DMA2D introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339

11.2

DMA2D main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340

11.3

DMA2D functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
11.3.1

General description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340

11.3.2

DMA2D control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341

11.3.3

DMA2D foreground and background FIFOs . . . . . . . . . . . . . . . . . . . . 341

11.3.4

DMA2D foreground and background pixel format converter (PFC) . . . 342

11.3.5

DMA2D foreground and background CLUT interface . . . . . . . . . . . . . 344

11.3.6

DMA2D blender . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345

11.3.7

DMA2D output PFC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345

11.3.8

DMA2D output FIFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346

11.3.9

DMA2D AHB master port timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346

11.3.10 DMA2D transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
11.3.11

DMA2D configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347

11.3.12 DMA2D transfer control (start, suspend, abort and completion) . . . . . 350
11.3.13 Watermark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
11.3.14 Error management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
11.3.15 AHB dead time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351

10/1747

11.4

DMA2D interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351

11.5

DMA2D registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352

RM0090 Rev 17

RM0090

Contents
11.5.1

DMA2D control register (DMA2D_CR) . . . . . . . . . . . . . . . . . . . . . . . . 352

11.5.2

DMA2D Interrupt Status Register (DMA2D_ISR) . . . . . . . . . . . . . . . . 354

11.5.3

DMA2D interrupt flag clear register (DMA2D_IFCR) . . . . . . . . . . . . . . 355

11.5.4

DMA2D foreground memory address register (DMA2D_FGMAR) . . . 356

11.5.5

DMA2D foreground offset register (DMA2D_FGOR) . . . . . . . . . . . . . . 356

11.5.6

DMA2D background memory address register (DMA2D_BGMAR) . . 357

11.5.7

DMA2D background offset register (DMA2D_BGOR) . . . . . . . . . . . . . 357

11.5.8

DMA2D foreground PFC control register (DMA2D_FGPFCCR) . . . . . 358

11.5.9

DMA2D foreground color register (DMA2D_FGCOLR) . . . . . . . . . . . . 360

11.5.10 DMA2D background PFC control register (DMA2D_BGPFCCR) . . . . 361
11.5.11

DMA2D background color register (DMA2D_BGCOLR) . . . . . . . . . . . 363

11.5.12 DMA2D foreground CLUT memory address register
(DMA2D_FGCMAR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
11.5.13 DMA2D background CLUT memory address register
(DMA2D_BGCMAR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364
11.5.14 DMA2D output PFC control register (DMA2D_OPFCCR) . . . . . . . . . . 364
11.5.15 DMA2D output color register (DMA2D_OCOLR) . . . . . . . . . . . . . . . . . 365
11.5.16 DMA2D output memory address register (DMA2D_OMAR) . . . . . . . . 366
11.5.17 DMA2D output offset register (DMA2D_OOR) . . . . . . . . . . . . . . . . . . 367
11.5.18 DMA2D number of line register (DMA2D_NLR) . . . . . . . . . . . . . . . . . 367
11.5.19 DMA2D line watermark register (DMA2D_LWR) . . . . . . . . . . . . . . . . . 368
11.5.20 DMA2D AHB master timer configuration register (DMA2D_AMTCR) . 368
11.5.21 DMA2D register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369

12

Interrupts and events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371
12.1

12.2

12.3

Nested vectored interrupt controller (NVIC) . . . . . . . . . . . . . . . . . . . . . . 371
12.1.1

NVIC features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371

12.1.2

SysTick calibration value register . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371

12.1.3

Interrupt and exception vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371

External interrupt/event controller (EXTI) . . . . . . . . . . . . . . . . . . . . . . . . 371
12.2.1

EXTI main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379

12.2.2

EXTI block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380

12.2.3

Wakeup event management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380

12.2.4

Functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380

12.2.5

External interrupt/event line mapping . . . . . . . . . . . . . . . . . . . . . . . . . 382

EXTI registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
12.3.1

Interrupt mask register (EXTI_IMR) . . . . . . . . . . . . . . . . . . . . . . . . . . . 384

RM0090 Rev 17

11/1747
39

Contents

13

RM0090
12.3.2

Event mask register (EXTI_EMR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384

12.3.3

Rising trigger selection register (EXTI_RTSR) . . . . . . . . . . . . . . . . . . 385

12.3.4

Falling trigger selection register (EXTI_FTSR) . . . . . . . . . . . . . . . . . . 385

12.3.5

Software interrupt event register (EXTI_SWIER) . . . . . . . . . . . . . . . . 386

12.3.6

Pending register (EXTI_PR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386

12.3.7

EXTI register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387

Analog-to-digital converter (ADC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388
13.1

ADC introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388

13.2

ADC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388

13.3

ADC functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
13.3.1

ADC on-off control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390

13.3.2

ADC clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390

13.3.3

Channel selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390

13.3.4

Single conversion mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391

13.3.5

Continuous conversion mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392

13.3.6

Timing diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392

13.3.7

Analog watchdog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392

13.3.8

Scan mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393

13.3.9

Injected channel management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394

13.3.10 Discontinuous mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395

13.4

Data alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396

13.5

Channel-wise programmable sampling time . . . . . . . . . . . . . . . . . . . . . 397

13.6

Conversion on external trigger and trigger polarity . . . . . . . . . . . . . . . . 397

13.7

Fast conversion mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399

13.8

Data management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400

13.9

12/1747

13.8.1

Using the DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400

13.8.2

Managing a sequence of conversions without using the DMA . . . . . . 400

13.8.3

Conversions without DMA and without overrun detection . . . . . . . . . . 401

Multi ADC mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
13.9.1

Injected simultaneous mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404

13.9.2

Regular simultaneous mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405

13.9.3

Interleaved mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407

13.9.4

Alternate trigger mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408

13.9.5

Combined regular/injected simultaneous mode . . . . . . . . . . . . . . . . . . 410

13.9.6

Combined regular simultaneous + alternate trigger mode . . . . . . . . . . 411

RM0090 Rev 17

RM0090

Contents

13.10 Temperature sensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
13.11 Battery charge monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
13.12 ADC interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
13.13 ADC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
13.13.1 ADC status register (ADC_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
13.13.2 ADC control register 1 (ADC_CR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
13.13.3 ADC control register 2 (ADC_CR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . 418
13.13.4 ADC sample time register 1 (ADC_SMPR1) . . . . . . . . . . . . . . . . . . . . 420
13.13.5 ADC sample time register 2 (ADC_SMPR2) . . . . . . . . . . . . . . . . . . . . 420
13.13.6 ADC injected channel data offset register x (ADC_JOFRx) (x=1..4) . . 421
13.13.7 ADC watchdog higher threshold register (ADC_HTR) . . . . . . . . . . . . . 421
13.13.8 ADC watchdog lower threshold register (ADC_LTR) . . . . . . . . . . . . . . 422
13.13.9 ADC regular sequence register 1 (ADC_SQR1) . . . . . . . . . . . . . . . . . 422
13.13.10 ADC regular sequence register 2 (ADC_SQR2) . . . . . . . . . . . . . . . . . 423
13.13.11 ADC regular sequence register 3 (ADC_SQR3) . . . . . . . . . . . . . . . . . 423
13.13.12 ADC injected sequence register (ADC_JSQR) . . . . . . . . . . . . . . . . . . 424
13.13.13 ADC injected data register x (ADC_JDRx) (x= 1..4) . . . . . . . . . . . . . . 425
13.13.14 ADC regular data register (ADC_DR) . . . . . . . . . . . . . . . . . . . . . . . . . 425
13.13.15 ADC Common status register (ADC_CSR) . . . . . . . . . . . . . . . . . . . . . 426
13.13.16 ADC common control register (ADC_CCR) . . . . . . . . . . . . . . . . . . . . . 427
13.13.17 ADC common regular data register for dual and triple modes
(ADC_CDR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
13.13.18 ADC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430

14

Digital-to-analog converter (DAC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
14.1

DAC introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433

14.2

DAC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433

14.3

DAC functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
14.3.1

DAC channel enable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

14.3.2

DAC output buffer enable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

14.3.3

DAC data format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

14.3.4

DAC conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436

14.3.5

DAC output voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437

14.3.6

DAC trigger selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437

14.3.7

DMA request . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438

14.3.8

Noise generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438

14.3.9

Triangle-wave generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
RM0090 Rev 17

13/1747
39

Contents

RM0090

14.4

Dual DAC channel conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
14.4.1

Independent trigger without wave generation . . . . . . . . . . . . . . . . . . . 441

14.4.2

Independent trigger with single LFSR generation . . . . . . . . . . . . . . . . 441

14.4.3

Independent trigger with different LFSR generation . . . . . . . . . . . . . . 441

14.4.4

Independent trigger with single triangle generation . . . . . . . . . . . . . . . 442

14.4.5

Independent trigger with different triangle generation . . . . . . . . . . . . . 442

14.4.6

Simultaneous software start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442

14.4.7

Simultaneous trigger without wave generation . . . . . . . . . . . . . . . . . . 443

14.4.8

Simultaneous trigger with single LFSR generation . . . . . . . . . . . . . . . 443

14.4.9

Simultaneous trigger with different LFSR generation . . . . . . . . . . . . . 443

14.4.10 Simultaneous trigger with single triangle generation . . . . . . . . . . . . . . 444
14.4.11 Simultaneous trigger with different triangle generation . . . . . . . . . . . . 444

14.5

DAC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
14.5.1

DAC control register (DAC_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445

14.5.2

DAC software trigger register (DAC_SWTRIGR) . . . . . . . . . . . . . . . . . 448

14.5.3

DAC channel1 12-bit right-aligned data holding register
(DAC_DHR12R1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448

14.5.4

DAC channel1 12-bit left aligned data holding register
(DAC_DHR12L1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449

14.5.5

DAC channel1 8-bit right aligned data holding register
(DAC_DHR8R1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449

14.5.6

DAC channel2 12-bit right aligned data holding register
(DAC_DHR12R2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450

14.5.7

DAC channel2 12-bit left aligned data holding register
(DAC_DHR12L2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450

14.5.8

DAC channel2 8-bit right-aligned data holding register
(DAC_DHR8R2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450

14.5.9

Dual DAC 12-bit right-aligned data holding register
(DAC_DHR12RD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451

14.5.10 DUAL DAC 12-bit left aligned data holding register
(DAC_DHR12LD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
14.5.11 DUAL DAC 8-bit right aligned data holding register
(DAC_DHR8RD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452
14.5.12 DAC channel1 data output register (DAC_DOR1) . . . . . . . . . . . . . . . . 452
14.5.13 DAC channel2 data output register (DAC_DOR2) . . . . . . . . . . . . . . . . 452
14.5.14 DAC status register (DAC_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
14.5.15 DAC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453

15

14/1747

Digital camera interface (DCMI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455

RM0090 Rev 17

RM0090

Contents

15.1

DCMI introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455

15.2

DCMI main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455

15.3

DCMI pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455

15.4

DCMI clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455

15.5

DCMI functional overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456

15.6

15.5.1

DMA interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457

15.5.2

DCMI physical interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457

15.5.3

Synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459

15.5.4

Capture modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462

15.5.5

Crop feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463

15.5.6

JPEG format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465

15.5.7

FIFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465

Data format description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
15.6.1

Data formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465

15.6.2

Monochrome format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466

15.6.3

RGB format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466

15.6.4

YCbCr format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466

15.7

DCMI interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467

15.8

DCMI register description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
15.8.1

DCMI control register 1 (DCMI_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . 467

15.8.2

DCMI status register (DCMI_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470

15.8.3

DCMI raw interrupt status register (DCMI_RIS) . . . . . . . . . . . . . . . . . . 471

15.8.4

DCMI interrupt enable register (DCMI_IER) . . . . . . . . . . . . . . . . . . . . 472

15.8.5

DCMI masked interrupt status register (DCMI_MIS) . . . . . . . . . . . . . . 473

15.8.6

DCMI interrupt clear register (DCMI_ICR) . . . . . . . . . . . . . . . . . . . . . . 474

15.8.7

DCMI embedded synchronization code register (DCMI_ESCR) . . . . . 475

15.8.8

DCMI embedded synchronization unmask register (DCMI_ESUR) . . 476

15.8.9

DCMI crop window start (DCMI_CWSTRT) . . . . . . . . . . . . . . . . . . . . . 477

15.8.10 DCMI crop window size (DCMI_CWSIZE) . . . . . . . . . . . . . . . . . . . . . . 477
15.8.11 DCMI data register (DCMI_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
15.8.12 DCMI register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478

16

LCD-TFT controller (LTDC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480
16.1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480

16.2

LTDC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480

16.3

LTDC functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481

RM0090 Rev 17

15/1747
39

Contents

RM0090

16.4

16.3.1

LTDC block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481

16.3.2

LTDC reset and clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481

16.3.3

LCD-TFT pins and signal interface . . . . . . . . . . . . . . . . . . . . . . . . . . . 483

LTDC programmable parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
16.4.1

LTDC Global configuration parameters . . . . . . . . . . . . . . . . . . . . . . . . 483

16.4.2

Layer programmable parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 486

16.5

LTDC interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490

16.6

LTDC programming procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 492

16.7

LTDC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493
16.7.1

LTDC Synchronization Size Configuration Register (LTDC_SSCR) . . 493

16.7.2

LTDC Back Porch Configuration Register (LTDC_BPCR) . . . . . . . . . . 493

16.7.3

LTDC Active Width Configuration Register (LTDC_AWCR) . . . . . . . . . 494

16.7.4

LTDC Total Width Configuration Register (LTDC_TWCR) . . . . . . . . . . 495

16.7.5

LTDC Global Control Register (LTDC_GCR) . . . . . . . . . . . . . . . . . . . . 495

16.7.6

LTDC Shadow Reload Configuration Register (LTDC_SRCR) . . . . . . 497

16.7.7

LTDC Background Color Configuration Register (LTDC_BCCR) . . . . 497

16.7.8

LTDC Interrupt Enable Register (LTDC_IER) . . . . . . . . . . . . . . . . . . . 498

16.7.9

LTDC Interrupt Status Register (LTDC_ISR) . . . . . . . . . . . . . . . . . . . . 499

16.7.10 LTDC Interrupt Clear Register (LTDC_ICR) . . . . . . . . . . . . . . . . . . . . . 499
16.7.11 LTDC Line Interrupt Position Configuration Register (LTDC_LIPCR) . 500
16.7.12 LTDC Current Position Status Register (LTDC_CPSR) . . . . . . . . . . . . 500
16.7.13 LTDC Current Display Status Register (LTDC_CDSR) . . . . . . . . . . . . 501
16.7.14 LTDC Layerx Control Register (LTDC_LxCR) (where x=1..2) . . . . . . . 502
16.7.15 LTDC Layerx Window Horizontal Position Configuration Register
(LTDC_LxWHPCR) (where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . 503
16.7.16 LTDC Layerx Window Vertical Position Configuration Register
(LTDC_LxWVPCR) (where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . 504
16.7.17 LTDC Layerx Color Keying Configuration Register (LTDC_LxCKCR)
(where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
16.7.18 LTDC Layerx Pixel Format Configuration Register (LTDC_LxPFCR)
(where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
16.7.19 LTDC Layerx Constant Alpha Configuration Register (LTDC_LxCACR)
(where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506
16.7.20 LTDC Layerx Default Color Configuration Register (LTDC_LxDCCR)
(where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506
16.7.21 LTDC Layerx Blending Factors Configuration Register (LTDC_LxBFCR)
(where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 508
16.7.22 LTDC Layerx Color Frame Buffer Address Register (LTDC_LxCFBAR)
(where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509

16/1747

RM0090 Rev 17

RM0090

Contents
16.7.23 LTDC Layerx Color Frame Buffer Length Register (LTDC_LxCFBLR)
(where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509
16.7.24 LTDC Layerx ColorFrame Buffer Line Number Register
(LTDC_LxCFBLNR) (where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . 510
16.7.25 LTDC Layerx CLUT Write Register (LTDC_LxCLUTWR)
(where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511
16.7.26 LTDC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 512

17

Advanced-control timers (TIM1 and TIM8) . . . . . . . . . . . . . . . . . . . . . 515
17.1

TIM1 and TIM8 introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515

17.2

TIM1 and TIM8 main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516

17.3

TIM1 and TIM8 functional description . . . . . . . . . . . . . . . . . . . . . . . . . . 518
17.3.1

Time-base unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 518

17.3.2

Counter modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520

17.3.3

Repetition counter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 529

17.3.4

Clock selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 532

17.3.5

Capture/compare channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535

17.3.6

Input capture mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538

17.3.7

PWM input mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539

17.3.8

Forced output mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539

17.3.9

Output compare mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540

17.3.10 PWM mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
17.3.11 Complementary outputs and dead-time insertion . . . . . . . . . . . . . . . . 544
17.3.12 Using the break function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 546
17.3.13 Clearing the OCxREF signal on an external event . . . . . . . . . . . . . . . 549
17.3.14 6-step PWM generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 550
17.3.15 One-pulse mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
17.3.16 Encoder interface mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552
17.3.17 Timer input XOR function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555
17.3.18 Interfacing with Hall sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555
17.3.19 TIMx and external trigger synchronization . . . . . . . . . . . . . . . . . . . . . . 557
17.3.20 Timer synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 560
17.3.21 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 560

17.4

TIM1 and TIM8 registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561
17.4.1

TIM1 and TIM8 control register 1 (TIMx_CR1) . . . . . . . . . . . . . . . . . . 561

17.4.2

TIM1 and TIM8 control register 2 (TIMx_CR2) . . . . . . . . . . . . . . . . . . 562

17.4.3

TIM1 and TIM8 slave mode control register (TIMx_SMCR) . . . . . . . . 565

17.4.4

TIM1 and TIM8 DMA/interrupt enable register (TIMx_DIER) . . . . . . . 567
RM0090 Rev 17

17/1747
39

Contents

RM0090
17.4.5

TIM1 and TIM8 status register (TIMx_SR) . . . . . . . . . . . . . . . . . . . . . 569

17.4.6

TIM1 and TIM8 event generation register (TIMx_EGR) . . . . . . . . . . . 570

17.4.7

TIM1 and TIM8 capture/compare mode register 1 (TIMx_CCMR1) . . 572

17.4.8

TIM1 and TIM8 capture/compare mode register 2 (TIMx_CCMR2) . . 575

17.4.9

TIM1 and TIM8 capture/compare enable register (TIMx_CCER) . . . . 576

17.4.10 TIM1 and TIM8 counter (TIMx_CNT) . . . . . . . . . . . . . . . . . . . . . . . . . . 580
17.4.11 TIM1 and TIM8 prescaler (TIMx_PSC) . . . . . . . . . . . . . . . . . . . . . . . . 580
17.4.12 TIM1 and TIM8 auto-reload register (TIMx_ARR) . . . . . . . . . . . . . . . . 580
17.4.13 TIM1 and TIM8 repetition counter register (TIMx_RCR) . . . . . . . . . . . 581
17.4.14 TIM1 and TIM8 capture/compare register 1 (TIMx_CCR1) . . . . . . . . . 581
17.4.15 TIM1 and TIM8 capture/compare register 2 (TIMx_CCR2) . . . . . . . . . 582
17.4.16 TIM1 and TIM8 capture/compare register 3 (TIMx_CCR3) . . . . . . . . . 582
17.4.17 TIM1 and TIM8 capture/compare register 4 (TIMx_CCR4) . . . . . . . . . 583
17.4.18 TIM1 and TIM8 break and dead-time register (TIMx_BDTR) . . . . . . . 583
17.4.19 TIM1 and TIM8 DMA control register (TIMx_DCR) . . . . . . . . . . . . . . . 585
17.4.20 TIM1 and TIM8 DMA address for full transfer (TIMx_DMAR) . . . . . . . 586
17.4.21 TIM1 and TIM8 register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587

18

General-purpose timers (TIM2 to TIM5) . . . . . . . . . . . . . . . . . . . . . . . . 589
18.1

TIM2 to TIM5 introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589

18.2

TIM2 to TIM5 main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589

18.3

TIM2 to TIM5 functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
18.3.1

Time-base unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591

18.3.2

Counter modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592

18.3.3

Clock selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602

18.3.4

Capture/compare channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605

18.3.5

Input capture mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607

18.3.6

PWM input mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608

18.3.7

Forced output mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609

18.3.8

Output compare mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609

18.3.9

PWM mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610

18.3.10 One-pulse mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613
18.3.11 Clearing the OCxREF signal on an external event . . . . . . . . . . . . . . . 614
18.3.12 Encoder interface mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
18.3.13 Timer input XOR function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618
18.3.14 Timers and external trigger synchronization . . . . . . . . . . . . . . . . . . . . 618
18.3.15 Timer synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621

18/1747

RM0090 Rev 17

RM0090

Contents
18.3.16 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626

18.4

TIM2 to TIM5 registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 627
18.4.1

TIMx control register 1 (TIMx_CR1) . . . . . . . . . . . . . . . . . . . . . . . . . . 627

18.4.2

TIMx control register 2 (TIMx_CR2) . . . . . . . . . . . . . . . . . . . . . . . . . . 629

18.4.3

TIMx slave mode control register (TIMx_SMCR) . . . . . . . . . . . . . . . . . 630

18.4.4

TIMx DMA/Interrupt enable register (TIMx_DIER) . . . . . . . . . . . . . . . . 632

18.4.5

TIMx status register (TIMx_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 633

18.4.6

TIMx event generation register (TIMx_EGR) . . . . . . . . . . . . . . . . . . . . 635

18.4.7

TIMx capture/compare mode register 1 (TIMx_CCMR1) . . . . . . . . . . . 636

18.4.8

TIMx capture/compare mode register 2 (TIMx_CCMR2) . . . . . . . . . . . 639

18.4.9

TIMx capture/compare enable register (TIMx_CCER) . . . . . . . . . . . . . 640

18.4.10 TIMx counter (TIMx_CNT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642
18.4.11 TIMx prescaler (TIMx_PSC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642
18.4.12 TIMx auto-reload register (TIMx_ARR) . . . . . . . . . . . . . . . . . . . . . . . . 642
18.4.13 TIMx capture/compare register 1 (TIMx_CCR1) . . . . . . . . . . . . . . . . . 643
18.4.14 TIMx capture/compare register 2 (TIMx_CCR2) . . . . . . . . . . . . . . . . . 643
18.4.15 TIMx capture/compare register 3 (TIMx_CCR3) . . . . . . . . . . . . . . . . . 644
18.4.16 TIMx capture/compare register 4 (TIMx_CCR4) . . . . . . . . . . . . . . . . . 644
18.4.17 TIMx DMA control register (TIMx_DCR) . . . . . . . . . . . . . . . . . . . . . . . 645
18.4.18 TIMx DMA address for full transfer (TIMx_DMAR) . . . . . . . . . . . . . . . 646
18.4.19 TIM2 option register (TIM2_OR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647
18.4.20 TIM5 option register (TIM5_OR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647
18.4.21 TIMx register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 648

19

General-purpose timers (TIM9 to TIM14) . . . . . . . . . . . . . . . . . . . . . . . 650
19.1

TIM9 to TIM14 introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650

19.2

TIM9 to TIM14 main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650

19.3

19.2.1

TIM9/TIM12 main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650

19.2.2

TIM10/TIM11 and TIM13/TIM14 main features . . . . . . . . . . . . . . . . . . 651

TIM9 to TIM14 functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . 653
19.3.1

Time-base unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653

19.3.2

Counter modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655

19.3.3

Clock selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658

19.3.4

Capture/compare channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660

19.3.5

Input capture mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 661

19.3.6

PWM input mode (only for TIM9/12) . . . . . . . . . . . . . . . . . . . . . . . . . . 663

19.3.7

Forced output mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664
RM0090 Rev 17

19/1747
39

Contents

RM0090
19.3.8

Output compare mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664

19.3.9

PWM mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665

19.3.10 One-pulse mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666
19.3.11 TIM9/12 external trigger synchronization . . . . . . . . . . . . . . . . . . . . . . . 668
19.3.12 Timer synchronization (TIM9/12) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671
19.3.13 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671

19.4

TIM9 and TIM12 registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 672
19.4.1

TIM9/12 control register 1 (TIMx_CR1) . . . . . . . . . . . . . . . . . . . . . . . . 672

19.4.2

TIM9/12 slave mode control register (TIMx_SMCR) . . . . . . . . . . . . . . 673

19.4.3

TIM9/12 Interrupt enable register (TIMx_DIER) . . . . . . . . . . . . . . . . . 674

19.4.4

TIM9/12 status register (TIMx_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . 676

19.4.5

TIM9/12 event generation register (TIMx_EGR) . . . . . . . . . . . . . . . . . 677

19.4.6

TIM9/12 capture/compare mode register 1 (TIMx_CCMR1) . . . . . . . . 678

19.4.7

TIM9/12 capture/compare enable register (TIMx_CCER) . . . . . . . . . . 681

19.4.8

TIM9/12 counter (TIMx_CNT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 682

19.4.9

TIM9/12 prescaler (TIMx_PSC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 682

19.4.10 TIM9/12 auto-reload register (TIMx_ARR) . . . . . . . . . . . . . . . . . . . . . 682
19.4.11 TIM9/12 capture/compare register 1 (TIMx_CCR1) . . . . . . . . . . . . . . . 683
19.4.12 TIM9/12 capture/compare register 2 (TIMx_CCR2) . . . . . . . . . . . . . . . 683
19.4.13 TIM9/12 register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684

19.5

TIM10/11/13/14 registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 686
19.5.1

TIM10/11/13/14 control register 1 (TIMx_CR1) . . . . . . . . . . . . . . . . . . 686

19.5.2

TIM10/11/13/14 Interrupt enable register (TIMx_DIER) . . . . . . . . . . . . 687

19.5.3

TIM10/11/13/14 status register (TIMx_SR) . . . . . . . . . . . . . . . . . . . . . 687

19.5.4

TIM10/11/13/14 event generation register (TIMx_EGR) . . . . . . . . . . . 688

19.5.5

TIM10/11/13/14 capture/compare mode register 1 (TIMx_CCMR1) . . 688

19.5.6

TIM10/11/13/14 capture/compare enable register (TIMx_CCER) . . . . 691

19.5.7

TIM10/11/13/14 counter (TIMx_CNT) . . . . . . . . . . . . . . . . . . . . . . . . . 692

19.5.8

TIM10/11/13/14 prescaler (TIMx_PSC) . . . . . . . . . . . . . . . . . . . . . . . . 692

19.5.9

TIM10/11/13/14 auto-reload register (TIMx_ARR) . . . . . . . . . . . . . . . . 692

19.5.10 TIM10/11/13/14 capture/compare register 1 (TIMx_CCR1) . . . . . . . . . 693
19.5.11 TIM11 option register 1 (TIM11_OR) . . . . . . . . . . . . . . . . . . . . . . . . . . 693
19.5.12 TIM10/11/13/14 register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 694

20

20/1747

Basic timers (TIM6 and TIM7) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 696
20.1

TIM6 and TIM7 introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 696

20.2

TIM6 and TIM7 main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 696
RM0090 Rev 17

RM0090

Contents

20.3

20.4

21

20.3.1

Time-base unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 697

20.3.2

Counting mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699

20.3.3

Clock source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701

20.3.4

Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702

TIM6 and TIM7 registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702
20.4.1

TIM6 and TIM7 control register 1 (TIMx_CR1) . . . . . . . . . . . . . . . . . . 702

20.4.2

TIM6 and TIM7 control register 2 (TIMx_CR2) . . . . . . . . . . . . . . . . . . 704

20.4.3

TIM6 and TIM7 DMA/Interrupt enable register (TIMx_DIER) . . . . . . . 704

20.4.4

TIM6 and TIM7 status register (TIMx_SR) . . . . . . . . . . . . . . . . . . . . . 705

20.4.5

TIM6 and TIM7 event generation register (TIMx_EGR) . . . . . . . . . . . 705

20.4.6

TIM6 and TIM7 counter (TIMx_CNT) . . . . . . . . . . . . . . . . . . . . . . . . . . 705

20.4.7

TIM6 and TIM7 prescaler (TIMx_PSC) . . . . . . . . . . . . . . . . . . . . . . . . 706

20.4.8

TIM6 and TIM7 auto-reload register (TIMx_ARR) . . . . . . . . . . . . . . . . 706

20.4.9

TIM6 and TIM7 register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707

Independent watchdog (IWDG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708
21.1

IWDG introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708

21.2

IWDG main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708

21.3

IWDG functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708

21.4

22

TIM6 and TIM7 functional description . . . . . . . . . . . . . . . . . . . . . . . . . . 697

21.3.1

Hardware watchdog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708

21.3.2

Register access protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708

21.3.3

Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 709

IWDG registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 710
21.4.1

Key register (IWDG_KR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 710

21.4.2

Prescaler register (IWDG_PR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 710

21.4.3

Reload register (IWDG_RLR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 711

21.4.4

Status register (IWDG_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 711

21.4.5

IWDG register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 712

Window watchdog (WWDG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713
22.1

WWDG introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713

22.2

WWDG main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713

22.3

WWDG functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713

22.4

How to program the watchdog timeout . . . . . . . . . . . . . . . . . . . . . . . . . . 715

22.5

Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 716

RM0090 Rev 17

21/1747
39

Contents

RM0090

22.6

23

WWDG registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 717
22.6.1

Control register (WWDG_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 717

22.6.2

Configuration register (WWDG_CFR) . . . . . . . . . . . . . . . . . . . . . . . . . 718

22.6.3

Status register (WWDG_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 718

22.6.4

WWDG register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 719

Cryptographic processor (CRYP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720
23.1

CRYP introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720

23.2

CRYP main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720

23.3

CRYP functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 722
23.3.1

DES/TDES cryptographic core . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723

23.3.2

AES cryptographic core . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 728

23.3.3

Data type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739

23.3.4

Initialization vectors - CRYP_IV0...1(L/R) . . . . . . . . . . . . . . . . . . . . . . 742

23.3.5

CRYP busy state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 743

23.3.6

Procedure to perform an encryption or a decryption . . . . . . . . . . . . . . 744

23.3.7

Context swapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745

23.4

CRYP interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747

23.5

CRYP DMA interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748

23.6

CRYP registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748
23.6.1

CRYP control register (CRYP_CR) for STM32F415/417xx . . . . . . . . . 748

23.6.2

CRYP control register (CRYP_CR) for STM32F415/417xx . . . . . . . . . 750

23.6.3

CRYP status register (CRYP_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 753

23.6.4

CRYP data input register (CRYP_DIN) . . . . . . . . . . . . . . . . . . . . . . . . 754

23.6.5

CRYP data output register (CRYP_DOUT) . . . . . . . . . . . . . . . . . . . . . 755

23.6.6

CRYP DMA control register (CRYP_DMACR) . . . . . . . . . . . . . . . . . . . 756

23.6.7

CRYP interrupt mask set/clear register (CRYP_IMSCR) . . . . . . . . . . . 756

23.6.8

CRYP raw interrupt status register (CRYP_RISR) . . . . . . . . . . . . . . . 757

23.6.9

CRYP masked interrupt status register (CRYP_MISR) . . . . . . . . . . . . 757

23.6.10 CRYP key registers (CRYP_K0...3(L/R)R) . . . . . . . . . . . . . . . . . . . . . 758
23.6.11 CRYP initialization vector registers (CRYP_IV0...1(L/R)R) . . . . . . . . . 760
23.6.12 CRYP context swap registers (CRYP_CSGCMCCM0..7R and
CRYP_CSGCM0..7R) for STM32F42xxx and STM32F43xxx . . . . . . . 762
23.6.13 CRYP register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 763

24

Random number generator (RNG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767
24.1

22/1747

RNG introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767
RM0090 Rev 17

RM0090

Contents

24.2

RNG main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767

24.3

RNG functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767

24.4

25

Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 768

24.3.2

Error management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 768

RNG registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 768
24.4.1

RNG control register (RNG_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769

24.4.2

RNG status register (RNG_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769

24.4.3

RNG data register (RNG_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 770

24.4.4

RNG register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771

Hash processor (HASH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772
25.1

HASH introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772

25.2

HASH main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772

25.3

HASH functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773

25.4

26

24.3.1

25.3.1

Duration of the processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 775

25.3.2

Data type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 775

25.3.3

Message digest computing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 777

25.3.4

Message padding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 778

25.3.5

Hash operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779

25.3.6

HMAC operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779

25.3.7

Context swapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 780

25.3.8

HASH interrupt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 782

HASH registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 782
25.4.1

HASH control register (HASH_CR) for STM32F415/417xx . . . . . . . . . 782

25.4.2

HASH control register (HASH_CR) for STM32F43xxx . . . . . . . . . . . . 785

25.4.3

HASH data input register (HASH_DIN) . . . . . . . . . . . . . . . . . . . . . . . . 788

25.4.4

HASH start register (HASH_STR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789

25.4.5

HASH digest registers (HASH_HR0..4/5/6/7) . . . . . . . . . . . . . . . . . . . 790

25.4.6

HASH interrupt enable register (HASH_IMR) . . . . . . . . . . . . . . . . . . . 792

25.4.7

HASH status register (HASH_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793

25.4.8

HASH context swap registers (HASH_CSRx) . . . . . . . . . . . . . . . . . . . 794

25.4.9

HASH register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 795

Real-time clock (RTC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798
26.1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798

26.2

RTC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 799

RM0090 Rev 17

23/1747
39

Contents

RM0090

26.3

RTC functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 801
26.3.1

Clock and prescalers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 801

26.3.2

Real-time clock and calendar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 802

26.3.3

Programmable alarms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 802

26.3.4

Periodic auto-wakeup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 802

26.3.5

RTC initialization and configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . 803

26.3.6

Reading the calendar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 805

26.3.7

Resetting the RTC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 806

26.3.8

RTC synchronization

26.3.9

RTC reference clock detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 807

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 806

26.3.10 RTC coarse digital calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 808
26.3.11 RTC smooth digital calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 809
26.3.12 Timestamp function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 811
26.3.13 Tamper detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 811
26.3.14 Calibration clock output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 813
26.3.15 Alarm output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 813

26.4

RTC and low-power modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814

26.5

RTC interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814

26.6

RTC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 816
26.6.1

RTC time register (RTC_TR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 816

26.6.2

RTC date register (RTC_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 817

26.6.3

RTC control register (RTC_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 818

26.6.4

RTC initialization and status register (RTC_ISR) . . . . . . . . . . . . . . . . . 820

26.6.5

RTC prescaler register (RTC_PRER) . . . . . . . . . . . . . . . . . . . . . . . . . 823

26.6.6

RTC wakeup timer register (RTC_WUTR) . . . . . . . . . . . . . . . . . . . . . . 823

26.6.7

RTC calibration register (RTC_CALIBR) . . . . . . . . . . . . . . . . . . . . . . . 824

26.6.8

RTC alarm A register (RTC_ALRMAR) . . . . . . . . . . . . . . . . . . . . . . . . 825

26.6.9

RTC alarm B register (RTC_ALRMBR) . . . . . . . . . . . . . . . . . . . . . . . . 826

26.6.10 RTC write protection register (RTC_WPR) . . . . . . . . . . . . . . . . . . . . . 827
26.6.11 RTC sub second register (RTC_SSR) . . . . . . . . . . . . . . . . . . . . . . . . . 827
26.6.12 RTC shift control register (RTC_SHIFTR) . . . . . . . . . . . . . . . . . . . . . . 828
26.6.13 RTC time stamp time register (RTC_TSTR) . . . . . . . . . . . . . . . . . . . . 828
26.6.14 RTC time stamp date register (RTC_TSDR) . . . . . . . . . . . . . . . . . . . . 829
26.6.15 RTC timestamp sub second register (RTC_TSSSR) . . . . . . . . . . . . . . 830
26.6.16 RTC calibration register (RTC_CALR) . . . . . . . . . . . . . . . . . . . . . . . . . 830
26.6.17 RTC tamper and alternate function configuration register
(RTC_TAFCR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 832

24/1747

RM0090 Rev 17

RM0090

Contents
26.6.18 RTC alarm A sub second register (RTC_ALRMASSR) . . . . . . . . . . . . 834
26.6.19 RTC alarm B sub second register (RTC_ALRMBSSR) . . . . . . . . . . . . 835
26.6.20 RTC backup registers (RTC_BKPxR) . . . . . . . . . . . . . . . . . . . . . . . . . 836
26.6.21 RTC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 836

27

Inter-integrated circuit (I2C) interface . . . . . . . . . . . . . . . . . . . . . . . . . 839
27.1

I2C introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 839

27.2

I2C main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 839

27.3

I2C functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 840
27.3.1

Mode selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 840

27.3.2

I2C slave mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 843

27.3.3

I2C master mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 846

27.3.4

Error conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 851

27.3.5

Programmable noise filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 852

27.3.6

SDA/SCL line control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 853

27.3.7

SMBus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 853

27.3.8

DMA requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 856

27.3.9

Packet error checking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 857

27.4

I2C

27.5

I2C debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 860

27.6

I2C registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 860

interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 858

27.6.1

I2C Control register 1 (I2C_CR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 860

27.6.2

I2C Control register 2 (I2C_CR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 862

27.6.3

I2C Own address register 1 (I2C_OAR1) . . . . . . . . . . . . . . . . . . . . . . . 864

27.6.4

I2C Own address register 2 (I2C_OAR2) . . . . . . . . . . . . . . . . . . . . . . . 864

27.6.5

I2C Data register (I2C_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865

27.6.6

I2C Status register 1 (I2C_SR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865

27.6.7

I2C Status register 2 (I2C_SR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 868

27.6.8

I2C Clock control register (I2C_CCR) . . . . . . . . . . . . . . . . . . . . . . . . . 870

27.6.9

I2C TRISE register (I2C_TRISE) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 871

27.6.10 I2C FLTR register (I2C_FLTR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 871
27.6.11 I2C register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 872

28

Serial peripheral interface (SPI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 873
28.1

SPI introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 873

28.2

SPI and I2S main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 874

RM0090 Rev 17

25/1747
39

Contents

RM0090

28.3

28.2.1

SPI features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 874

28.2.2

I2S features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 875

SPI functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 876
28.3.1

General description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 876

28.3.2

Configuring the SPI in slave mode . . . . . . . . . . . . . . . . . . . . . . . . . . . 879

28.3.3

Configuring the SPI in master mode . . . . . . . . . . . . . . . . . . . . . . . . . . 882

28.3.4

Configuring the SPI for half-duplex communication . . . . . . . . . . . . . . . 884

28.3.5

Data transmission and reception procedures . . . . . . . . . . . . . . . . . . . 885

28.3.6

CRC calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 891

28.3.7

Status flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 893

28.3.8

Disabling the SPI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 894

28.3.9

SPI communication using DMA (direct memory addressing) . . . . . . . 895

28.3.10 Error flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 897
28.3.11 SPI interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 898

28.4

I2S functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 899
28.4.1

I2S general description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 899

28.4.2

I2S full duplex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 900

28.4.3

Supported audio protocols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901

28.4.4

Clock generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 907

28.4.5

I2S master mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 909

28.4.6

I2S slave mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 911

28.4.7

Status flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 913

28.4.8

Error flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 914

28.4.9

I2S interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 915

28.4.10 DMA features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 915

28.5

SPI and I2S registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 916
28.5.1

SPI control register 1 (SPI_CR1) (not used in I2S mode) . . . . . . . . . . 916

28.5.2

SPI control register 2 (SPI_CR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 918

28.5.3

SPI status register (SPI_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 919

28.5.4

SPI data register (SPI_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 920

28.5.5

SPI CRC polynomial register (SPI_CRCPR) (not used in I2S
mode) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 920

28.5.6

SPI RX CRC register (SPI_RXCRCR) (not used in I2S mode) . . . . . . 921

28.5.7

SPI TX CRC register (SPI_TXCRCR) (not used in I2S mode) . . . . . . 921

28.5.8

SPI_I2S configuration register (SPI_I2SCFGR) . . . . . . . . . . . . . . . . . . 922

28.5.9

SPI_I2S prescaler register (SPI_I2SPR) . . . . . . . . . . . . . . . . . . . . . . . 923

28.5.10 SPI register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 925

26/1747

RM0090 Rev 17

RM0090

29

Contents

Serial audio interface (SAI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 926
29.1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 926

29.2

Main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 927

29.3

Functional block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 928

29.4

Main SAI modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 929

29.5

SAI synchronization mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 930

29.6

Audio data size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 930

29.7

Frame synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 930
29.7.1

Frame length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 931

29.7.2

Frame synchronization polarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 931

29.7.3

Frame synchronization active level length . . . . . . . . . . . . . . . . . . . . . . 932

29.7.4

Frame synchronization offset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 932

29.7.5

FS signal role . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 932

29.8

Slot configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 933

29.9

SAI clock generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 935

29.10 Internal FIFOs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 936
29.11 AC’97 link controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 939
29.12 Specific features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 940
29.12.1 Mute mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 940
29.12.2 MONO/STEREO function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 940
29.12.3 Companding mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 941
29.12.4 Output data line management on an inactive slot . . . . . . . . . . . . . . . . 942

29.13 Error flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 944
29.13.1 FIFO overrun/underrun (OVRUDR) . . . . . . . . . . . . . . . . . . . . . . . . . . . 944
29.13.2 Anticipated frame synchronisation detection (AFSDET) . . . . . . . . . . . 946
29.13.3 Late frame synchronization detection . . . . . . . . . . . . . . . . . . . . . . . . . 946
29.13.4 Codec not ready (CNRDY AC’97) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 947
29.13.5 Wrong clock configuration in master mode (with NODIV = 0) . . . . . . . 947

29.14 Interrupt sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948
29.15 Disabling the SAI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948
29.16 SAI DMA interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 949
29.17 SAI registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 950
29.17.1 SAI xConfiguration register 1 (SAI_xCR1) where x is A or B . . . . . . . 950
29.17.2 SAI xConfiguration register 2 (SAI_xCR2) where x is A or B . . . . . . . 953
29.17.3 SAI xFrame configuration register (SAI_XFRCR) where x is A or B . . 955

RM0090 Rev 17

27/1747
39

Contents

RM0090
29.17.4 SAI xSlot register (SAI_xSLOTR) where x is A or B . . . . . . . . . . . . . . 957
29.17.5 SAI xInterrupt mask register (SAI_xIM) where x is A or B . . . . . . . . . . 958
29.17.6 SAI xStatus register (SAI_xSR) where x is A or B . . . . . . . . . . . . . . . . 960
29.17.7 SAI xClear flag register (SAI_xCLRFR) where X is A or B . . . . . . . . . 962
29.17.8 SAI xData register (SAI_xDR) where x is A or B . . . . . . . . . . . . . . . . . 963
29.17.9 SAI register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 963

30

Universal synchronous asynchronous receiver
transmitter (USART) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 965
30.1

USART introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 965

30.2

USART main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 965

30.3

USART functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 966
30.3.1

USART character description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 969

30.3.2

Transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 970

30.3.3

Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 973

30.3.4

Fractional baud rate generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 978

30.3.5

USART receiver tolerance to clock deviation . . . . . . . . . . . . . . . . . . . . 988

30.3.6

Multiprocessor communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 989

30.3.7

Parity control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 991

30.3.8

LIN (local interconnection network) mode . . . . . . . . . . . . . . . . . . . . . . 992

30.3.9

USART synchronous mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 994

30.3.10 Single-wire half-duplex communication . . . . . . . . . . . . . . . . . . . . . . . . 996
30.3.11 Smartcard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 997
30.3.12 IrDA SIR ENDEC block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 999
30.3.13 Continuous communication using DMA . . . . . . . . . . . . . . . . . . . . . . . 1001
30.3.14 Hardware flow control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1003

28/1747

30.4

USART interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1006

30.5

USART mode configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1007

30.6

USART registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1007
30.6.1

Status register (USART_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1007

30.6.2

Data register (USART_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1010

30.6.3

Baud rate register (USART_BRR) . . . . . . . . . . . . . . . . . . . . . . . . . . . 1010

30.6.4

Control register 1 (USART_CR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1010

30.6.5

Control register 2 (USART_CR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1013

30.6.6

Control register 3 (USART_CR3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1014

30.6.7

Guard time and prescaler register (USART_GTPR) . . . . . . . . . . . . . 1017

30.6.8

USART register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1018
RM0090 Rev 17

RM0090

31

Contents

Secure digital input/output interface (SDIO) . . . . . . . . . . . . . . . . . . . 1019
31.1

SDIO main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1019

31.2

SDIO bus topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1020

31.3

SDIO functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1022

31.4

31.3.1

SDIO adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1023

31.3.2

SDIO APB2 interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1033

Card functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1034
31.4.1

Card identification mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1034

31.4.2

Card reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1034

31.4.3

Operating voltage range validation . . . . . . . . . . . . . . . . . . . . . . . . . . 1034

31.4.4

Card identification process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1035

31.4.5

Block write . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1036

31.4.6

Block read . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1037

31.4.7

Stream access, stream write and stream read
(MultiMediaCard only) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1037

31.4.8

Erase: group erase and sector erase . . . . . . . . . . . . . . . . . . . . . . . . 1039

31.4.9

Wide bus selection or deselection . . . . . . . . . . . . . . . . . . . . . . . . . . . 1039

31.4.10 Protection management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1039
31.4.11 Card status register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1042
31.4.12 SD status register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1045
31.4.13 SD I/O mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1049
31.4.14 Commands and responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1050

31.5

31.6

Response formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1054
31.5.1

R1 (normal response command) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1054

31.5.2

R1b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1054

31.5.3

R2 (CID, CSD register) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1054

31.5.4

R3 (OCR register) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1055

31.5.5

R4 (Fast I/O) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1055

31.5.6

R4b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1056

31.5.7

R5 (interrupt request) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1056

31.5.8

R6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1057

SDIO I/O card-specific operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1057
31.6.1

SDIO I/O read wait operation by SDIO_D2 signaling . . . . . . . . . . . . 1058

31.6.2

SDIO read wait operation by stopping SDIO_CK . . . . . . . . . . . . . . . 1058

31.6.3

SDIO suspend/resume operation . . . . . . . . . . . . . . . . . . . . . . . . . . . 1058

31.6.4

SDIO interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1058

RM0090 Rev 17

29/1747
39

Contents

RM0090

31.7

CE-ATA specific operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1059
31.7.1

Command completion signal disable . . . . . . . . . . . . . . . . . . . . . . . . . 1059

31.7.2

Command completion signal enable . . . . . . . . . . . . . . . . . . . . . . . . . 1059

31.7.3

CE-ATA interrupt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1059

31.7.4

Aborting CMD61 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1059

31.8

HW flow control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1060

31.9

SDIO registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1060
31.9.1

SDIO power control register (SDIO_POWER) . . . . . . . . . . . . . . . . . . 1060

31.9.2

SDI clock control register (SDIO_CLKCR) . . . . . . . . . . . . . . . . . . . . 1061

31.9.3

SDIO argument register (SDIO_ARG) . . . . . . . . . . . . . . . . . . . . . . . . 1062

31.9.4

SDIO command register (SDIO_CMD) . . . . . . . . . . . . . . . . . . . . . . . 1062

31.9.5

SDIO command response register (SDIO_RESPCMD) . . . . . . . . . . 1063

31.9.6

SDIO response 1..4 register (SDIO_RESPx) . . . . . . . . . . . . . . . . . . 1064

31.9.7

SDIO data timer register (SDIO_DTIMER) . . . . . . . . . . . . . . . . . . . . 1064

31.9.8

SDIO data length register (SDIO_DLEN) . . . . . . . . . . . . . . . . . . . . . 1065

31.9.9

SDIO data control register (SDIO_DCTRL) . . . . . . . . . . . . . . . . . . . . 1066

31.9.10 SDIO data counter register (SDIO_DCOUNT) . . . . . . . . . . . . . . . . . . 1067
31.9.11 SDIO status register (SDIO_STA) . . . . . . . . . . . . . . . . . . . . . . . . . . . 1068
31.9.12 SDIO interrupt clear register (SDIO_ICR) . . . . . . . . . . . . . . . . . . . . . 1069
31.9.13 SDIO mask register (SDIO_MASK) . . . . . . . . . . . . . . . . . . . . . . . . . . 1071
31.9.14 SDIO FIFO counter register (SDIO_FIFOCNT) . . . . . . . . . . . . . . . . . 1073
31.9.15 SDIO data FIFO register (SDIO_FIFO) . . . . . . . . . . . . . . . . . . . . . . . 1074
31.9.16 SDIO register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1074

32

Controller area network (bxCAN) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1076
32.1

bxCAN introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1076

32.2

bxCAN main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1076

32.3

bxCAN general description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1077

32.4

30/1747

32.3.1

CAN 2.0B active core . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1077

32.3.2

Control, status and configuration registers . . . . . . . . . . . . . . . . . . . . 1078

32.3.3

Tx mailboxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1078

32.3.4

Acceptance filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1078

bxCAN operating modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1079
32.4.1

Initialization mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1080

32.4.2

Normal mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1080

32.4.3

Sleep mode (low power) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1080

RM0090 Rev 17

RM0090

Contents

32.5

33

Test mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1081
32.5.1

Silent mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1081

32.5.2

Loop back mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1082

32.5.3

Loop back combined with silent mode . . . . . . . . . . . . . . . . . . . . . . . . 1082

32.6

Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1083

32.7

bxCAN functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1083
32.7.1

Transmission handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1083

32.7.2

Time triggered communication mode . . . . . . . . . . . . . . . . . . . . . . . . . 1085

32.7.3

Reception handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1085

32.7.4

Identifier filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1086

32.7.5

Message storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1090

32.7.6

Error management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1092

32.7.7

Bit timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1092

32.8

bxCAN interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1094

32.9

CAN registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1096
32.9.1

Register access protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1096

32.9.2

CAN control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1096

32.9.3

CAN mailbox registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1106

32.9.4

CAN filter registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1113

32.9.5

bxCAN register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1117

Ethernet (ETH): media access control (MAC) with
DMA controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1121
33.1

Ethernet introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1121

33.2

Ethernet main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1121
33.2.1

MAC core features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1122

33.2.2

DMA features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1123

33.2.3

PTP features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1123

33.3

Ethernet pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1124

33.4

Ethernet functional description: SMI, MII and RMII . . . . . . . . . . . . . . . .1125

33.5

33.4.1

Station management interface: SMI . . . . . . . . . . . . . . . . . . . . . . . . . . 1125

33.4.2

Media-independent interface: MII . . . . . . . . . . . . . . . . . . . . . . . . . . . 1128

33.4.3

Reduced media-independent interface: RMII . . . . . . . . . . . . . . . . . . 1131

33.4.4

MII/RMII selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1132

Ethernet functional description: MAC 802.3 . . . . . . . . . . . . . . . . . . . . . .1133
33.5.1

MAC 802.3 frame format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1133

RM0090 Rev 17

31/1747
39

Contents

RM0090

33.6

34

MAC frame transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1137

33.5.3

MAC frame reception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1144

33.5.4

MAC interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1149

33.5.5

MAC filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1150

33.5.6

MAC loopback mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1153

33.5.7

MAC management counters: MMC . . . . . . . . . . . . . . . . . . . . . . . . . . 1153

33.5.8

Power management: PMT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1154

33.5.9

Precision time protocol (IEEE1588 PTP) . . . . . . . . . . . . . . . . . . . . . . 1157

Ethernet functional description: DMA controller operation . . . . . . . . . . .1163
33.6.1

Initialization of a transfer using DMA . . . . . . . . . . . . . . . . . . . . . . . . . 1164

33.6.2

Host bus burst access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1164

33.6.3

Host data buffer alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1165

33.6.4

Buffer size calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1165

33.6.5

DMA arbiter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1166

33.6.6

Error response to DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1166

33.6.7

Tx DMA configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1166

33.6.8

Rx DMA configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1178

33.6.9

DMA interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1189

33.7

Ethernet interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1190

33.8

Ethernet register descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1191
33.8.1

MAC register description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1191

33.8.2

MMC register description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1210

33.8.3

IEEE 1588 time stamp registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1215

33.8.4

DMA register description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1223

33.8.5

Ethernet register maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1237

USB on-the-go full-speed (OTG_FS) . . . . . . . . . . . . . . . . . . . . . . . . . 1241
34.1

OTG_FS introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1241

34.2

OTG_FS main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1242

34.3

32/1747

33.5.2

34.2.1

General features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1242

34.2.2

Host-mode features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1243

34.2.3

Peripheral-mode features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1243

OTG_FS functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1244
34.3.1

OTG pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1244

34.3.2

OTG full-speed core . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1244

34.3.3

Full-speed OTG PHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1245

RM0090 Rev 17

RM0090

Contents

34.4

34.5

34.6

34.7

OTG dual role device (DRD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246
34.4.1

ID line detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246

34.4.2

HNP dual role device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246

34.4.3

SRP dual role device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1247

USB peripheral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1247
34.5.1

SRP-capable peripheral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1248

34.5.2

Peripheral states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1248

34.5.3

Peripheral endpoints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1249

USB host . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1251
34.6.1

SRP-capable host . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1252

34.6.2

USB host states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1252

34.6.3

Host channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1254

34.6.4

Host scheduler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1255

SOF trigger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1256
34.7.1

Host SOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1256

34.7.2

Peripheral SOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1257

34.8

OTG low-power modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1257

34.9

Dynamic update of the OTG_FS_HFIR register . . . . . . . . . . . . . . . . . . 1258

34.10 USB data FIFOs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1259
34.11 Peripheral FIFO architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1260
34.11.1 Peripheral Rx FIFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1260
34.11.2 Peripheral Tx FIFOs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1261

34.12 Host FIFO architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1261
34.12.1 Host Rx FIFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1261
34.12.2 Host Tx FIFOs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1262

34.13 FIFO RAM allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1262
34.13.1 Device mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1262
34.13.2 Host mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1263

34.14 USB system performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1263
34.15 OTG_FS interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1264
34.16 OTG_FS control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . . 1266
34.16.1 CSR memory map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1267
34.16.2 OTG_FS global registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1272
34.16.3 Host-mode registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1293
34.16.4 Device-mode registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1303

RM0090 Rev 17

33/1747
39

Contents

RM0090
34.16.5 OTG_FS power and clock gating control register
(OTG_FS_PCGCCTL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1326
34.16.6 OTG_FS register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1327

34.17 OTG_FS programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1336
34.17.1 Core initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1336
34.17.2 Host initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1337
34.17.3 Device initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1337
34.17.4 Host programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1338
34.17.5 Device programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1354
34.17.6 Operational model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1356
34.17.7 Worst case response time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1372
34.17.8 OTG programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1374

35

USB on-the-go high-speed (OTG_HS) . . . . . . . . . . . . . . . . . . . . . . . . 1380
35.1

OTG_HS introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1380

35.2

OTG_HS main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1380

35.3

35.4

35.5

35.6

34/1747

35.2.1

General features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1381

35.2.2

Host-mode features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1382

35.2.3

Peripheral-mode features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1382

OTG_HS functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1382
35.3.1

OTG pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1383

35.3.2

High-speed OTG PHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1383

35.3.3

Embedded Full-speed OTG PHY . . . . . . . . . . . . . . . . . . . . . . . . . . . 1384

OTG dual-role device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1384
35.4.1

ID line detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1384

35.4.2

HNP dual role device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1384

35.4.3

SRP dual-role device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1385

USB functional description in peripheral mode . . . . . . . . . . . . . . . . . . 1385
35.5.1

SRP-capable peripheral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1385

35.5.2

Peripheral states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1386

35.5.3

Peripheral endpoints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1387

USB functional description on host mode . . . . . . . . . . . . . . . . . . . . . . 1390
35.6.1

SRP-capable host . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1390

35.6.2

USB host states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1391

35.6.3

Host channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1392

35.6.4

Host scheduler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1394

RM0090 Rev 17

RM0090

Contents

35.7

SOF trigger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1395
35.7.1

Host SOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1395

35.7.2

Peripheral SOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1395

35.8

OTG_HS low-power modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1396

35.9

Dynamic update of the OTG_HS_HFIR register . . . . . . . . . . . . . . . . . 1397

35.10 FIFO RAM allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1398
35.10.1 Peripheral mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1398
35.10.2 Host mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1398

35.11 OTG_HS interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1399
35.12 OTG_HS control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . 1401
35.12.1 CSR memory map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1401
35.12.2 OTG_HS global registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1406
35.12.3 Host-mode registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1429
35.12.4 Device-mode registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1442
35.12.5 OTG_HS power and clock gating control register
(OTG_HS_PCGCCTL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1471
35.12.6 OTG_HS register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1471

35.13 OTG_HS programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1486
35.13.1 Core initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1486
35.13.2 Host initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1487
35.13.3 Device initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1488
35.13.4 DMA mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1488
35.13.5 Host programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1488
35.13.6 Device programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1514
35.13.7 Operational model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1516
35.13.8 Worst case response time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1534
35.13.9 OTG programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1536

36

Flexible static memory controller (FSMC) . . . . . . . . . . . . . . . . . . . . 1542
36.1

FSMC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1542

36.2

Block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1543

36.3

AHB interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1543
36.3.1

36.4

Supported memories and transactions . . . . . . . . . . . . . . . . . . . . . . . 1544

External device address mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1545
36.4.1

NOR/PSRAM address mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1545

36.4.2

NAND/PC Card address mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . 1546

RM0090 Rev 17

35/1747
39

Contents

RM0090

36.5

36.6

37

NOR Flash/PSRAM controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1547
36.5.1

External memory interface signals . . . . . . . . . . . . . . . . . . . . . . . . . . . 1548

36.5.2

Supported memories and transactions . . . . . . . . . . . . . . . . . . . . . . . 1550

36.5.3

General timing rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1551

36.5.4

NOR Flash/PSRAM controller asynchronous transactions . . . . . . . . 1552

36.5.5

Synchronous transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1569

36.5.6

NOR/PSRAM control registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1575

NAND Flash/PC Card controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1582
36.6.1

External memory interface signals . . . . . . . . . . . . . . . . . . . . . . . . . . . 1583

36.6.2

NAND Flash / PC Card supported memories and transactions . . . . . 1585

36.6.3

Timing diagrams for NAND and PC Card . . . . . . . . . . . . . . . . . . . . . 1585

36.6.4

NAND Flash operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1586

36.6.5

NAND Flash prewait functionality . . . . . . . . . . . . . . . . . . . . . . . . . . . 1587

36.6.6

Computation of the error correction code (ECC)
in NAND Flash memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1588

36.6.7

PC Card/CompactFlash operations . . . . . . . . . . . . . . . . . . . . . . . . . . 1589

36.6.8

NAND Flash/PC Card control registers . . . . . . . . . . . . . . . . . . . . . . . 1591

36.6.9

FSMC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1598

Flexible memory controller (FMC) . . . . . . . . . . . . . . . . . . . . . . . . . . . 1600
37.1

FMC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1600

37.2

Block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1601

37.3

AHB interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1602
37.3.1

37.4

37.5

37.6

External device address mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1604
37.4.1

NOR/PSRAM address mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1605

37.4.2

NAND Flash memory/PC Card address mapping . . . . . . . . . . . . . . . 1606

37.4.3

SDRAM address mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1607

NOR Flash/PSRAM controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1610
37.5.1

External memory interface signals . . . . . . . . . . . . . . . . . . . . . . . . . . . 1611

37.5.2

Supported memories and transactions . . . . . . . . . . . . . . . . . . . . . . . 1613

37.5.3

General timing rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1615

37.5.4

NOR Flash/PSRAM controller asynchronous transactions . . . . . . . . 1615

37.5.5

Synchronous transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1632

37.5.6

NOR/PSRAM controller registers . . . . . . . . . . . . . . . . . . . . . . . . . . . 1638

NAND Flash/PC Card controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1645
37.6.1

36/1747

Supported memories and transactions . . . . . . . . . . . . . . . . . . . . . . . 1603

External memory interface signals . . . . . . . . . . . . . . . . . . . . . . . . . . . 1646
RM0090 Rev 17

RM0090

Contents

37.7

37.8

38

37.6.2

NAND Flash / PC Card supported memories and transactions . . . . . 1648

37.6.3

Timing diagrams for NAND Flash memory and PC Card . . . . . . . . . 1648

37.6.4

NAND Flash operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1649

37.6.5

NAND Flash prewait functionality . . . . . . . . . . . . . . . . . . . . . . . . . . . 1650

37.6.6

Computation of the error correction code (ECC)
in NAND Flash memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1651

37.6.7

PC Card/CompactFlash operations . . . . . . . . . . . . . . . . . . . . . . . . . . 1652

37.6.8

NAND Flash/PC Card controller registers . . . . . . . . . . . . . . . . . . . . . 1654

SDRAM controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1661
37.7.1

SDRAM controller main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1661

37.7.2

SDRAM External memory interface signals . . . . . . . . . . . . . . . . . . . . 1661

37.7.3

SDRAM controller functional description . . . . . . . . . . . . . . . . . . . . . . 1662

37.7.4

Low power modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1669

37.7.5

SDRAM controller registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1672

FMC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1678

Debug support (DBG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1681
38.1

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1681

38.2

Reference Arm® documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1682

38.3

SWJ debug port (serial wire and JTAG) . . . . . . . . . . . . . . . . . . . . . . . . 1682
38.3.1

38.4

Mechanism to select the JTAG-DP or the SW-DP . . . . . . . . . . . . . . . 1683

Pinout and debug port pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1683
38.4.1

SWJ debug port pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1684

38.4.2

Flexible SWJ-DP pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . 1684

38.4.3

Internal pull-up and pull-down on JTAG pins . . . . . . . . . . . . . . . . . . . 1685

38.4.4

Using serial wire and releasing the unused debug pins as GPIOs . . 1686

38.5

STM32F4xx JTAG TAP connection . . . . . . . . . . . . . . . . . . . . . . . . . . . 1686

38.6

ID codes and locking mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1688
38.6.1

MCU device ID code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1688

38.6.2

Boundary scan TAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1689

38.6.3

Cortex®-M4 with FPU TAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1689

38.6.4

Cortex®-M4 with FPU JEDEC-106 ID code . . . . . . . . . . . . . . . . . . . . 1689

38.7

JTAG debug port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1689

38.8

SW debug port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1691
38.8.1

SW protocol introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1691

38.8.2

SW protocol sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1691

RM0090 Rev 17

37/1747
39

Contents

RM0090

38.9

38.8.3

SW-DP state machine (reset, idle states, ID code) . . . . . . . . . . . . . . 1692

38.8.4

DP and AP read/write accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1692

38.8.5

SW-DP registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1693

38.8.6

SW-AP registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1694

AHB-AP (AHB access port) - valid for both JTAG-DP
and SW-DP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1694

38.10 Core debug . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1695
38.11 Capability of the debugger host to connect under system reset . . . . . 1696
38.12 FPB (Flash patch breakpoint) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1696
38.13 DWT (data watchpoint trigger) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1697
38.14 ITM (instrumentation trace macrocell) . . . . . . . . . . . . . . . . . . . . . . . . . 1697
38.14.1 General description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1697
38.14.2 Time stamp packets, synchronization and overflow packets . . . . . . . 1697

38.15 ETM (Embedded trace macrocell) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1699
38.15.1 ETM general description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1699
38.15.2 ETM signal protocol and packet types . . . . . . . . . . . . . . . . . . . . . . . . 1699
38.15.3 Main ETM registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1700
38.15.4 ETM configuration example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1700

38.16 MCU debug component (DBGMCU) . . . . . . . . . . . . . . . . . . . . . . . . . . 1700
38.16.1 Debug support for low-power modes . . . . . . . . . . . . . . . . . . . . . . . . . 1700
38.16.2 Debug support for timers, watchdog, bxCAN and I2C . . . . . . . . . . . . 1701
38.16.3 Debug MCU configuration register . . . . . . . . . . . . . . . . . . . . . . . . . . 1701
38.16.4 Debug MCU APB1 freeze register (DBGMCU_APB1_FZ) . . . . . . . . 1703
38.16.5 Debug MCU APB2 Freeze register (DBGMCU_APB2_FZ) . . . . . . . . 1704

38.17 TPIU (trace port interface unit) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1705
38.17.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1705
38.17.2 TRACE pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1706
38.17.3 TPUI formatter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1707
38.17.4 TPUI frame synchronization packets . . . . . . . . . . . . . . . . . . . . . . . . . 1708
38.17.5 Transmission of the synchronization frame packet . . . . . . . . . . . . . . 1708
38.17.6 Synchronous mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1708
38.17.7 Asynchronous mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1709
38.17.8 TRACECLKIN connection inside the STM32F4xx . . . . . . . . . . . . . . . 1709
38.17.9 TPIU registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1709
38.17.10 Example of configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1710

38.18 DBG register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1711

38/1747

RM0090 Rev 17

RM0090

39

40

Contents

Device electronic signature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1712
39.1

Unique device ID register (96 bits) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1712

39.2

Flash size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1713

Revision history . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1714

RM0090 Rev 17

39/1747
39

List of tables

RM0090

List of tables
Table 1.
Table 2.
Table 3.
Table 4.
Table 5.
Table 6.
Table 7.
Table 8.
Table 9.
Table 10.
Table 11.
Table 12.
Table 13.
Table 14.
Table 15.
Table 16.
Table 17.
Table 18.
Table 19.
Table 20.
Table 21.
Table 22.
Table 23.
Table 24.
Table 25.
Table 26.
Table 27.
Table 28.
Table 29.
Table 30.
Table 31.
Table 32.
Table 33.
Table 34.
Table 35.
Table 36.
Table 37.
Table 38.
Table 39.

40/1747

STM32F4xx register boundary addresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Boot modes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Memory mapping vs. Boot mode/physical remap
in STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Memory mapping vs. Boot mode/physical remap
in STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Flash module organization (STM32F40x and STM32F41x) . . . . . . . . . . . . . . . . . . . . . . . . 75
Flash module - 2 Mbyte dual bank organization (STM32F42xxx and STM32F43xxx) . . . . 77
1 Mbyte Flash memory single bank vs dual bank organization
(STM32F42xxx and STM32F43xxx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
1 Mbyte single bank Flash memory organization
(STM32F42xxx and STM32F43xxx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
1 Mbyte dual bank Flash memory organization (STM32F42xxx and STM32F43xxx) . . . . 79
Number of wait states according to CPU clock (HCLK) frequency
(STM32F405xx/07xx and STM32F415xx/17xx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Number of wait states according to CPU clock (HCLK) frequency
(STM32F42xxx and STM32F43xxx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Program/erase parallelism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Flash interrupt request . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Option byte organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Description of the option bytes (STM32F405xx/07xx and STM32F415xx/17xx) . . . . . . . . 89
Description of the option bytes
(STM32F42xxx and STM32F43xxx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Access versus read protection level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
OTP area organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Flash register map and reset values
(STM32F405xx/07xx and STM32F415xx/17xx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Flash register map and reset values (STM32F42xxx and STM32F43xxx) . . . . . . . . . . . . 111
CRC calculation unit register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Voltage regulator configuration mode versus device operating mode . . . . . . . . . . . . . . . 122
Low-power mode summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Sleep-now entry and exit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Sleep-on-exit entry and exit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Stop operating modes
(STM32F405xx/07xx and STM32F415xx/17xx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Stop mode entry and exit (for STM32F405xx/07xx and STM32F415xx/17xx) . . . . . . . . . 132
Stop operating modes (STM32F42xxx and STM32F43xxx) . . . . . . . . . . . . . . . . . . . . . . . 134
Stop mode entry and exit (STM32F42xxx and STM32F43xxx) . . . . . . . . . . . . . . . . . . . . 136
Standby mode entry and exit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
PWR - register map and reset values for STM32F405xx/07xx and STM32F415xx/17xx. 149
PWR - register map and reset values for STM32F42xxx and STM32F43xxx . . . . . . . . . 149
RCC register map and reset values for STM32F42xxx and STM32F43xxx . . . . . . . . . . . 210
RCC register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Port bit configuration table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Flexible SWJ-DP pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
RTC_AF1 pin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
RTC_AF2 pin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
GPIO register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287

RM0090 Rev 17

RM0090
Table 40.
Table 41.
Table 42.
Table 43.
Table 44.
Table 45.
Table 46.
Table 47.
Table 48.
Table 49.
Table 50.
Table 51.
Table 52.
Table 53.
Table 54.
Table 55.
Table 56.
Table 57.
Table 58.
Table 59.
Table 60.
Table 61.
Table 62.
Table 63.
Table 64.
Table 65.
Table 66.
Table 67.
Table 68.
Table 69.
Table 70.
Table 71.
Table 72.
Table 73.
Table 74.
Table 75.
Table 76.
Table 77.
Table 78.
Table 79.
Table 80.
Table 81.
Table 82.
Table 83.
Table 84.
Table 85.
Table 86.
Table 87.
Table 88.
Table 89.
Table 90.
Table 91.

List of tables
SYSCFG register map and reset values (STM32F405xx/07xx and STM32F415xx/17xx) 294
SYSCFG register map and reset values (STM32F42xxx and STM32F43xxx) . . . . . . . . . 301
DMA1 request mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
DMA2 request mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
Source and destination address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Source and destination address registers in Double buffer mode (DBM=1). . . . . . . . . . . 314
Packing/unpacking & endian behavior (bit PINC = MINC = 1) . . . . . . . . . . . . . . . . . . . . . 315
Restriction on NDT versus PSIZE and MSIZE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
FIFO threshold configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 318
Possible DMA configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
DMA interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
DMA register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
Supported color mode in input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342
Data order in memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
Alpha mode configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
Supported CLUT color mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
CLUT data order in memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
Supported color mode in output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
Data order in memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
DMA2D interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
DMA2D register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
Vector table for STM32F405xx/07xx and STM32F415xx/17xx. . . . . . . . . . . . . . . . . . . . . 372
Vector table for STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
External interrupt/event controller register map and reset values. . . . . . . . . . . . . . . . . . . 387
External interrupt/event controller register map and reset values. . . . . . . . . . . . . . . . . . . 387
ADC pins. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
Analog watchdog channel selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
Configuring the trigger polarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
External trigger for regular channels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
External trigger for injected channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
ADC interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
ADC global register map. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
ADC register map and reset values for each ADC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
ADC register map and reset values (common ADC registers) . . . . . . . . . . . . . . . . . . . . . 432
DAC pins. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434
External triggers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
DAC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
DCMI pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
DCMI signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
Positioning of captured data bytes in 32-bit words (8-bit width) . . . . . . . . . . . . . . . . . . . . 458
Positioning of captured data bytes in 32-bit words (10-bit width) . . . . . . . . . . . . . . . . . . . 458
Positioning of captured data bytes in 32-bit words (12-bit width) . . . . . . . . . . . . . . . . . . . 459
Positioning of captured data bytes in 32-bit words (14-bit width) . . . . . . . . . . . . . . . . . . . 459
Data storage in monochrome progressive video format . . . . . . . . . . . . . . . . . . . . . . . . . . 466
Data storage in RGB progressive video format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
Data storage in YCbCr progressive video format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
DCMI interrupts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
DCMI register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
LTDC registers versus clock domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482
LCD-TFT pins and signal interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
Pixel Data mapping versus Color Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
LTDC interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491

RM0090 Rev 17

41/1747
46

List of tables
Table 92.
Table 93.
Table 94.
Table 95.
Table 96.
Table 97.
Table 98.
Table 99.
Table 100.
Table 101.
Table 102.
Table 103.
Table 104.
Table 105.
Table 106.
Table 107.
Table 108.
Table 109.
Table 110.
Table 111.
Table 112.
Table 113.
Table 114.
Table 115.
Table 116.
Table 117.
Table 118.
Table 119.
Table 120.
Table 121.
Table 122.
Table 123.
Table 124.
Table 125.
Table 126.
Table 127.
Table 128.
Table 129.
Table 130.
Table 131.
Table 132.
Table 133.
Table 134.
Table 135.
Table 136.
Table 137.

42/1747

RM0090

LTDC register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 512
Counting direction versus encoder signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 553
TIMx Internal trigger connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
Output control bits for complementary OCx and OCxN channels with
break feature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
TIM1 and TIM8 register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
Counting direction versus encoder signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 616
TIMx internal trigger connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 632
Output control bit for standard OCx channels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 641
TIM2 to TIM5 register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 648
TIMx internal trigger connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 674
Output control bit for standard OCx channels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 682
TIM9/12 register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684
Output control bit for standard OCx channels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 691
TIM10/11/13/14 register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 694
TIM6 and TIM7 register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707
Min/max IWDG timeout period (in ms) at 32 kHz (LSI). . . . . . . . . . . . . . . . . . . . . . . . . . . 709
IWDG register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 712
Minimum and maximum timeout values at 30 MHz (fPCLK1). . . . . . . . . . . . . . . . . . . . . . . 716
WWDG register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 719
Number of cycles required to process each 128-bit block
(STM32F415/417xx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720
Number of cycles required to process each 128-bit block
(STM32F43xxx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720
Data types. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 740
CRYP register map and reset values for STM32F415/417xx . . . . . . . . . . . . . . . . . . . . . . 763
CRYP register map and reset values for STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . 764
RNG register map and reset map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771
HASH register map and reset values on STM32F415/417xx . . . . . . . . . . . . . . . . . . . . . . 795
HASH register map and reset values on STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . 796
Effect of low-power modes on RTC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814
Interrupt control bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 815
RTC register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 836
Maximum DNF[3:0] value to be compliant with Thd:dat(max) . . . . . . . . . . . . . . . . . . . . . 852
SMBus vs. I2C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 854
I2C Interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 858
I2C register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 872
SPI interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 898
Audio frequency precision (for PLLM VCO = 1 MHz or 2 MHz) . . . . . . . . . . . . . . . . . . . . 909
I2S interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 915
SPI register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 925
Example of possible audio frequency sampling range . . . . . . . . . . . . . . . . . . . . . . . . . . . 936
Interrupt sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948
SAI register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 963
Noise detection from sampled data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 977
Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 12 MHz,
oversampling by 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 980
Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK =12 MHz,
oversampling by 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 981
Error calculation for programmed baud rates at fPCLK = 16 MHz or fPCLK = 24 MHz,
oversampling by 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 981
Error calculation for programmed baud rates at fPCLK = 16 MHz or fPCLK = 24 MHz,

RM0090 Rev 17

RM0090

List of tables

oversampling by 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 982
Table 138. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 16 MHz,
oversampling by 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 983
Table 139. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 16 MHz,
oversampling by 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 983
Table 140. Error calculation for programmed baud rates at fPCLK = 30 MHz or fPCLK = 60 MHz,
oversampling by 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 984
Table 141. Error calculation for programmed baud rates at fPCLK = 30 MHz or fPCLK = 60 MHz,
oversampling by 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 985
Table 142. Error calculation for programmed baud rates at fPCLK = 42 MHz or fPCLK = 84 Hz,
oversampling by 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 986
Table 143. Error calculation for programmed baud rates at fPCLK = 42 MHz or fPCLK = 84 MHz,
oversampling by 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 987
Table 144. USART receiver’s tolerance when DIV fraction is 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 988
Table 145. USART receiver tolerance when DIV_Fraction is different from 0 . . . . . . . . . . . . . . . . . . 989
Table 146. Frame formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 991
Table 147. USART interrupt requests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1006
Table 148. USART mode configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1007
Table 149. USART register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1018
Table 150. SDIO I/O definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1023
Table 151. Command format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1027
Table 152. Short response format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1028
Table 153. Long response format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1028
Table 154. Command path status flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1028
Table 155. Data token format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1031
Table 156. Transmit FIFO status flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1032
Table 157. Receive FIFO status flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1033
Table 158. Card status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1043
Table 159. SD status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1046
Table 160. Speed class code field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1047
Table 161. Performance move field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1047
Table 162. AU_SIZE field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1048
Table 163. Maximum AU size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1048
Table 164. Erase size field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1048
Table 165. Erase timeout field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1049
Table 166. Erase offset field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1049
Table 167. Block-oriented write commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1051
Table 168. Block-oriented write protection commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1052
Table 169. Erase commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1052
Table 170. I/O mode commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1053
Table 171. Lock card . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1053
Table 172. Application-specific commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1053
Table 173. R1 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1054
Table 174. R2 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1055
Table 175. R3 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1055
Table 176. R4 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1055
Table 177. R4b response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1056
Table 178. R5 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1056
Table 179. R6 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1057
Table 180. Response type and SDIO_RESPx registers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1064
Table 181. SDIO register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1074
Table 182. Transmit mailbox mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1091

RM0090 Rev 17

43/1747
46

List of tables
Table 183.
Table 184.
Table 185.
Table 186.
Table 187.
Table 188.
Table 189.
Table 190.
Table 191.
Table 192.
Table 193.
Table 194.
Table 195.
Table 196.
Table 197.
Table 198.
Table 199.
Table 200.
Table 201.
Table 202.
Table 203.
Table 204.
Table 205.
Table 206.
Table 207.
Table 208.
Table 209.
Table 210.
Table 211.
Table 212.
Table 213.
Table 214.
Table 215.
Table 216.
Table 217.
Table 218.
Table 219.
Table 220.
Table 221.
Table 222.
Table 223.
Table 224.
Table 225.
Table 226.
Table 227.
Table 228.
Table 229.
Table 230.
Table 231.
Table 232.
Table 233.

44/1747

RM0090

Receive mailbox mapping. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1091
bxCAN register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1117
Alternate function mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1124
Management frame format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1126
Clock range. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1128
TX interface signal encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1130
RX interface signal encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1130
Frame statuses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1146
Destination address filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1152
Source address filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1153
Receive descriptor 0 - encoding for bits 7, 5 and 0 (normal descriptor
format only, EDFE=0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1184
Time stamp snapshot dependency on registers bits . . . . . . . . . . . . . . . . . . . . . . . . . . . 1217
Ethernet register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1237
OTG_FS input/output pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1244
Compatibility of STM32 low power modes with the OTG . . . . . . . . . . . . . . . . . . . . . . . . 1257
Core global control and status registers (CSRs). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1267
Host-mode control and status registers (CSRs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1268
Device-mode control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1269
Data FIFO (DFIFO) access register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1270
Power and clock gating control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . . . 1271
TRDT values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1277
Minimum duration for soft disconnect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1305
OTG_FS register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1327
OTG_HS input/output pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1383
Compatibility of STM32 low power modes with the OTG . . . . . . . . . . . . . . . . . . . . . . . . 1396
Core global control and status registers (CSRs). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1402
Host-mode control and status registers (CSRs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1403
Device-mode control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1404
Data FIFO (DFIFO) access register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1406
Power and clock gating control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . . . 1406
TRDT values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1413
Minimum duration for soft disconnect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1445
OTG_HS register map and reset values. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1471
NOR/PSRAM bank selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1546
External memory address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1546
Memory mapping and timing registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1546
NAND bank selections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1547
Programmable NOR/PSRAM access parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1548
Nonmultiplexed I/O NOR Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1548
Multiplexed I/O NOR Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1549
Nonmultiplexed I/Os PSRAM/SRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1549
Multiplexed I/O PSRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1550
NOR Flash/PSRAM controller: example of supported memories and transactions . . . . 1550
FSMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1553
FSMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1554
FSMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1555
FSMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1556
FSMC_BWTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1556
FSMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1558
FSMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1559
FSMC_BWTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1559

RM0090 Rev 17

RM0090
Table 234.
Table 235.
Table 236.
Table 237.
Table 238.
Table 239.
Table 240.
Table 241.
Table 242.
Table 243.
Table 244.
Table 245.
Table 246.
Table 247.
Table 248.
Table 249.
Table 250.
Table 251.
Table 252.
Table 253.
Table 254.
Table 255.
Table 256.
Table 257.
Table 258.
Table 259.
Table 260.
Table 261.
Table 262.
Table 263.
Table 264.
Table 265.
Table 266.
Table 267.
Table 268.
Table 269.
Table 270.
Table 271.
Table 272.
Table 273.
Table 274.
Table 275.
Table 276.
Table 277.
Table 278.
Table 279.
Table 280.
Table 281.
Table 282.
Table 283.
Table 284.
Table 285.

List of tables
FSMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1561
FSMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1561
FSMC_BWTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1562
FSMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1563
FSMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1564
FSMC_BWTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1564
FSMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1566
FSMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1566
FSMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1571
FSMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1572
FSMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1573
FSMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1574
Programmable NAND/PC Card access parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1583
8-bit NAND Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1583
16-bit NAND Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1584
16-bit PC Card . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1584
Supported memories and transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1585
16-bit PC-Card signals and access type. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1590
ECC result relevant bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1597
FSMC register map. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1598
NOR/PSRAM bank selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1605
NOR/PSRAM External memory address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1606
NAND/PC Card memory mapping and timing registers . . . . . . . . . . . . . . . . . . . . . . . . . 1606
NAND bank selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1607
SDRAM bank selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1607
SDRAM address mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1607
SDRAM address mapping with 8-bit data bus width. . . . . . . . . . . . . . . . . . . . . . . . . . . . 1608
SDRAM address mapping with 16-bit data bus width. . . . . . . . . . . . . . . . . . . . . . . . . . . 1609
SDRAM address mapping with 32-bit data bus width. . . . . . . . . . . . . . . . . . . . . . . . . . . 1609
Programmable NOR/PSRAM access parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1611
Non-multiplexed I/O NOR Flash memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1612
16-bit multiplexed I/O NOR Flash memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1612
Non-multiplexed I/Os PSRAM/SRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1612
16-Bit multiplexed I/O PSRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1613
NOR Flash/PSRAM: Example of supported memories and transactions . . . . . . . . . . . . 1614
FMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1617
FMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1617
FMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1619
FMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1620
FMC_BWTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1620
FMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1622
FMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1623
FMC_BWTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1623
FMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1625
FMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1625
FMC_BWTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1626
FMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1627
FMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1628
FMC_BWTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1628
FMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1630
FMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1630
FMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1635

RM0090 Rev 17

45/1747
46

List of tables
Table 286.
Table 287.
Table 288.
Table 289.
Table 290.
Table 291.
Table 292.
Table 293.
Table 294.
Table 295.
Table 296.
Table 297.
Table 298.
Table 299.
Table 300.
Table 301.
Table 302.
Table 303.
Table 304.
Table 305.
Table 306.
Table 307.
Table 308.
Table 309.
Table 310.
Table 311.
Table 312.
Table 313.
Table 314.
Table 315.

46/1747

RM0090

FMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1635
FMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1636
FMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1637
Programmable NAND Flash/PC Card access parameters . . . . . . . . . . . . . . . . . . . . . . . 1646
8-bit NAND Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1646
16-bit NAND Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1647
16-bit PC Card . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1647
Supported memories and transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1648
16-bit PC-Card signals and access type. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1652
ECC result relevant bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1660
SDRAM signals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1661
FMC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1678
SWJ debug port pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1684
Flexible SWJ-DP pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1684
JTAG debug port data registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1689
32-bit debug port registers addressed through the shifted value A[3:2] . . . . . . . . . . . . 1690
Packet request (8-bits) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1691
ACK response (3 bits). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1692
DATA transfer (33 bits) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1692
SW-DP registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1693
Cortex®-M4 with FPU AHB-AP registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1694
Core debug registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1695
Main ITM registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1698
Main ETM registers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1700
Asynchronous TRACE pin assignment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1706
Synchronous TRACE pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1706
Flexible TRACE pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1707
Important TPIU registers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1709
DBG register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1711
Document revision history . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1714

RM0090 Rev 17

RM0090

List of figures

List of figures
Figure 1.
Figure 2.
Figure 3.
Figure 4.
Figure 5.
Figure 6.
Figure 7.
Figure 8.
Figure 9.
Figure 10.
Figure 11.
Figure 12.
Figure 13.
Figure 14.
Figure 15.
Figure 16.
Figure 17.
Figure 18.
Figure 19.
Figure 20.
Figure 21.
Figure 22.
Figure 23.
Figure 24.
Figure 25.
Figure 26.
Figure 27.
Figure 28.
Figure 29.
Figure 30.
Figure 31.
Figure 32.
Figure 33.
Figure 34.
Figure 35.
Figure 36.
Figure 37.
Figure 38.
Figure 39.
Figure 40.
Figure 41.
Figure 42.
Figure 43.
Figure 44.
Figure 45.

System architecture for STM32F405xx/07xx and STM32F415xx/17xx devices. . . . . . . . . 60
System architecture for STM32F42xxx and STM32F43xxx devices . . . . . . . . . . . . . . . . . 62
Flash memory interface connection inside system architecture
(STM32F405xx/07xx and STM32F415xx/17xx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Flash memory interface connection inside system architecture
(STM32F42xxx and STM32F43xxx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Sequential 32-bit instruction execution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
RDP levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
PCROP levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
CRC calculation unit block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Power supply overview for STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . 116
Power supply overview for STM32F42xxx and STM32F43xxx. . . . . . . . . . . . . . . . . . . . . 117
Backup domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Power-on reset/power-down reset waveform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
BOR thresholds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
PVD thresholds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
Simplified diagram of the reset circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Clock tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
HSE/ LSE clock sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
Frequency measurement with TIM5 in Input capture mode . . . . . . . . . . . . . . . . . . . . . . . 159
Frequency measurement with TIM11 in Input capture mode . . . . . . . . . . . . . . . . . . . . . . 160
Simplified diagram of the reset circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
Clock tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
HSE/ LSE clock sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Frequency measurement with TIM5 in Input capture mode . . . . . . . . . . . . . . . . . . . . . . . 223
Frequency measurement with TIM11 in Input capture mode . . . . . . . . . . . . . . . . . . . . . . 223
Basic structure of a five-volt tolerant I/O port bit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Selecting an alternate function on STM32F405xx/07xx and STM32F415xx/17xx . . . . 272
Selecting an alternate function on STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . 273
Input floating/pull up/pull down configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
Output configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Alternate function configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
High impedance-analog configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
DMA block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
System implementation of the two DMA controllers
(STM32F405xx/07xx and STM32F415xx/17xx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
System implementation of the two DMA controllers (STM32F42xxx and STM32F43xxx) 306
Channel selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
Peripheral-to-memory mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Memory-to-peripheral mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Memory-to-memory mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
FIFO structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
DMA2D block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
External interrupt/event controller block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
External interrupt/event GPIO mapping (STM32F405xx/07xx and STM32F415xx/17xx). 382
External interrupt/event GPIO mapping (STM32F42xxx and STM32F43xxx) . . . . . . . . . 383
Single ADC block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
Timing diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392

RM0090 Rev 17

47/1747
56

List of figures
Figure 46.
Figure 47.
Figure 48.
Figure 49.
Figure 50.
Figure 51.
Figure 52.
Figure 53.
Figure 54.
Figure 55.
Figure 56.
Figure 57.
Figure 58.
Figure 59.
Figure 60.
Figure 61.
Figure 62.
Figure 63.
Figure 64.
Figure 65.
Figure 66.
Figure 67.
Figure 68.
Figure 69.
Figure 70.
Figure 71.
Figure 72.
Figure 73.
Figure 74.
Figure 75.
Figure 76.
Figure 77.
Figure 78.
Figure 79.
Figure 80.
Figure 81.
Figure 82.
Figure 83.
Figure 84.
Figure 85.
Figure 86.
Figure 87.
Figure 88.
Figure 89.
Figure 90.
Figure 91.
Figure 92.
Figure 93.
Figure 94.
Figure 95.
Figure 96.
Figure 97.

48/1747

RM0090

Analog watchdog’s guarded area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
Injected conversion latency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
Right alignment of 12-bit data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
Left alignment of 12-bit data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
Left alignment of 6-bit data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
Multi ADC block diagram(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
Injected simultaneous mode on 4 channels: dual ADC mode . . . . . . . . . . . . . . . . . . . . . 405
Injected simultaneous mode on 4 channels: triple ADC mode . . . . . . . . . . . . . . . . . . . . . 405
Regular simultaneous mode on 16 channels: dual ADC mode . . . . . . . . . . . . . . . . . . . . 406
Regular simultaneous mode on 16 channels: triple ADC mode . . . . . . . . . . . . . . . . . . . . 406
Interleaved mode on 1 channel in continuous conversion mode: dual ADC mode. . . . . . 407
Interleaved mode on 1 channel in continuous conversion mode: triple ADC mode . . . . . 408
Alternate trigger: injected group of each ADC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
Alternate trigger: 4 injected channels (each ADC) in discontinuous mode . . . . . . . . . . . . 410
Alternate trigger: injected group of each ADC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410
Alternate + regular simultaneous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
Case of trigger occurring during injected conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
Temperature sensor and VREFINT channel block diagram . . . . . . . . . . . . . . . . . . . . . . 413
DAC channel block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434
Data registers in single DAC channel mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
Data registers in dual DAC channel mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
Timing diagram for conversion with trigger disabled TEN = 0 . . . . . . . . . . . . . . . . . . . . . 437
DAC LFSR register calculation algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
DAC conversion (SW trigger enabled) with LFSR wave generation. . . . . . . . . . . . . . . . . 439
DAC triangle wave generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
DAC conversion (SW trigger enabled) with triangle wave generation . . . . . . . . . . . . . . . 440
DCMI block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456
Top-level block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
DCMI signal waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
Timing diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460
Frame capture waveforms in Snapshot mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462
Frame capture waveforms in continuous grab mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463
Coordinates and size of the window after cropping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464
Data capture waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464
Pixel raster scan order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
LTDC block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481
LCD-TFT Synchronous timings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 484
Layer window programmable parameters: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
Blending two layers with background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490
Interrupt events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491
Advanced-control timer block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517
Counter timing diagram with prescaler division change from 1 to 2 . . . . . . . . . . . . . . . . . 519
Counter timing diagram with prescaler division change from 1 to 4 . . . . . . . . . . . . . . . . . 519
Counter timing diagram, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520
Counter timing diagram, internal clock divided by 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521
Counter timing diagram, internal clock divided by 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521
Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521
Counter timing diagram, update event when ARPE=0 (TIMx_ARR not preloaded) . . . . . 522
Counter timing diagram, update event when ARPE=1 (TIMx_ARR preloaded) . . . . . . . . 522
Counter timing diagram, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524
Counter timing diagram, internal clock divided by 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524
Counter timing diagram, internal clock divided by 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525

RM0090 Rev 17

RM0090
Figure 98.
Figure 99.
Figure 100.
Figure 101.
Figure 102.
Figure 103.
Figure 104.
Figure 105.
Figure 106.
Figure 107.
Figure 108.
Figure 109.
Figure 110.
Figure 111.
Figure 112.
Figure 113.
Figure 114.
Figure 115.
Figure 116.
Figure 117.
Figure 118.
Figure 119.
Figure 120.
Figure 121.
Figure 122.
Figure 123.
Figure 124.
Figure 125.
Figure 126.
Figure 127.
Figure 128.
Figure 129.
Figure 130.
Figure 131.
Figure 132.
Figure 133.
Figure 134.
Figure 135.
Figure 136.
Figure 137.
Figure 138.
Figure 139.
Figure 140.
Figure 141.
Figure 142.
Figure 143.
Figure 144.
Figure 145.
Figure 146.
Figure 147.
Figure 148.
Figure 149.

List of figures
Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525
Counter timing diagram, update event when repetition counter is not used . . . . . . . . . . . 526
Counter timing diagram, internal clock divided by 1, TIMx_ARR = 0x6 . . . . . . . . . . . . . . 527
Counter timing diagram, internal clock divided by 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527
Counter timing diagram, internal clock divided by 4, TIMx_ARR=0x36 . . . . . . . . . . . . . . 528
Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528
Counter timing diagram, update event with ARPE=1 (counter underflow) . . . . . . . . . . . . 529
Counter timing diagram, Update event with ARPE=1 (counter overflow) . . . . . . . . . . . . . 529
Update rate examples depending on mode and TIMx_RCR register settings . . . . . . . . . 531
Control circuit in normal mode, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . 532
TI2 external clock connection example. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533
Control circuit in external clock mode 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 534
External trigger input block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 534
Control circuit in external clock mode 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535
Capture/compare channel (example: channel 1 input stage) . . . . . . . . . . . . . . . . . . . . . . 536
Capture/compare channel 1 main circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 536
Output stage of capture/compare channel (channel 1 to 3) . . . . . . . . . . . . . . . . . . . . . . . 537
Output stage of capture/compare channel (channel 4). . . . . . . . . . . . . . . . . . . . . . . . . . . 537
PWM input mode timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539
Output compare mode, toggle on OC1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
Edge-aligned PWM waveforms (ARR=8) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
Center-aligned PWM waveforms (ARR=8) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
Complementary output with dead-time insertion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
Dead-time waveforms with delay greater than the negative pulse. . . . . . . . . . . . . . . . . . 545
Dead-time waveforms with delay greater than the positive pulse. . . . . . . . . . . . . . . . . . . 545
Output behavior in response to a break.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 548
Clearing TIMx OCxREF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549
6-step generation, COM example (OSSR=1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 550
Example of one pulse mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
Example of counter operation in encoder interface mode. . . . . . . . . . . . . . . . . . . . . . . . . 554
Example of encoder interface mode with TI1FP1 polarity inverted. . . . . . . . . . . . . . . . . . 554
Example of Hall sensor interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556
Control circuit in reset mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 557
Control circuit in gated mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 558
Control circuit in trigger mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 559
Control circuit in external clock mode 2 + trigger mode . . . . . . . . . . . . . . . . . . . . . . . . . . 560
General-purpose timer block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 590
Counter timing diagram with prescaler division change from 1 to 2 . . . . . . . . . . . . . . . . . 592
Counter timing diagram with prescaler division change from 1 to 4 . . . . . . . . . . . . . . . . . 592
Counter timing diagram, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
Counter timing diagram, internal clock divided by 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
Counter timing diagram, internal clock divided by 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594
Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594
Counter timing diagram, Update event when ARPE=0 (TIMx_ARR not preloaded). . . . . 595
Counter timing diagram, Update event when ARPE=1 (TIMx_ARR preloaded). . . . . . . . 595
Counter timing diagram, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
Counter timing diagram, internal clock divided by 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597
Counter timing diagram, internal clock divided by 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597
Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597
Counter timing diagram, Update event . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 598
Counter timing diagram, internal clock divided by 1, TIMx_ARR=0x6 . . . . . . . . . . . . . . . 599
Counter timing diagram, internal clock divided by 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599

RM0090 Rev 17

49/1747
56

List of figures
Figure 150.
Figure 151.
Figure 152.
Figure 153.
Figure 154.
Figure 155.
Figure 156.
Figure 157.
Figure 158.
Figure 159.
Figure 160.
Figure 161.
Figure 162.
Figure 163.
Figure 164.
Figure 165.
Figure 166.
Figure 167.
Figure 168.
Figure 169.
Figure 170.
Figure 171.
Figure 172.
Figure 173.
Figure 174.
Figure 175.
Figure 176.
Figure 177.
Figure 178.
Figure 179.
Figure 180.
Figure 181.
Figure 182.
Figure 183.
Figure 184.
Figure 185.
Figure 186.
Figure 187.
Figure 188.
Figure 189.
Figure 190.
Figure 191.
Figure 192.
Figure 193.
Figure 194.
Figure 195.
Figure 196.
Figure 197.
Figure 198.
Figure 199.
Figure 200.
Figure 201.

50/1747

RM0090

Counter timing diagram, internal clock divided by 4, TIMx_ARR=0x36 . . . . . . . . . . . . . . 600
Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600
Counter timing diagram, Update event with ARPE=1 (counter underflow). . . . . . . . . . . . 601
Counter timing diagram, Update event with ARPE=1 (counter overflow) . . . . . . . . . . . . . 601
Control circuit in normal mode, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . 602
TI2 external clock connection example. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 603
Control circuit in external clock mode 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604
External trigger input block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604
Control circuit in external clock mode 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605
Capture/compare channel (example: channel 1 input stage) . . . . . . . . . . . . . . . . . . . . . . 605
Capture/compare channel 1 main circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606
Output stage of capture/compare channel (channel 1). . . . . . . . . . . . . . . . . . . . . . . . . . . 606
PWM input mode timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608
Output compare mode, toggle on OC1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610
Edge-aligned PWM waveforms (ARR=8) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 611
Center-aligned PWM waveforms (ARR=8) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 612
Example of one-pulse mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613
Clearing TIMx OCxREF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
Example of counter operation in encoder interface mode . . . . . . . . . . . . . . . . . . . . . . . . 617
Example of encoder interface mode with TI1FP1 polarity inverted . . . . . . . . . . . . . . . . . 617
Control circuit in reset mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618
Control circuit in gated mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 619
Control circuit in trigger mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 620
Control circuit in external clock mode 2 + trigger mode . . . . . . . . . . . . . . . . . . . . . . . . . . 621
Master/Slave timer example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621
Gating timer 2 with OC1REF of timer 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 622
Gating timer 2 with Enable of timer 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 623
Triggering timer 2 with update of timer 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624
Triggering timer 2 with Enable of timer 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625
Triggering timer 1 and 2 with timer 1 TI1 input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626
General-purpose timer block diagram (TIM9 and TIM12) . . . . . . . . . . . . . . . . . . . . . . . . 651
General-purpose timer block diagram (TIM10/11/13/14) . . . . . . . . . . . . . . . . . . . . . . . . 652
Counter timing diagram with prescaler division change from 1 to 2 . . . . . . . . . . . . . . . . . 654
Counter timing diagram with prescaler division change from 1 to 4 . . . . . . . . . . . . . . . . . 654
Counter timing diagram, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655
Counter timing diagram, internal clock divided by 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656
Counter timing diagram, internal clock divided by 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656
Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656
Counter timing diagram, update event when ARPE=0 (TIMx_ARR not preloaded) . . . . . 657
Counter timing diagram, update event when ARPE=1 (TIMx_ARR preloaded) . . . . . . . . 657
Control circuit in normal mode, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . 658
TI2 external clock connection example. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
Control circuit in external clock mode 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
Capture/compare channel (example: channel 1 input stage) . . . . . . . . . . . . . . . . . . . . . . 660
Capture/compare channel 1 main circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 661
Output stage of capture/compare channel (channel 1). . . . . . . . . . . . . . . . . . . . . . . . . . . 661
PWM input mode timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663
Output compare mode, toggle on OC1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
Edge-aligned PWM waveforms (ARR=8) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666
Example of one pulse mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667
Control circuit in reset mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669
Control circuit in gated mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670

RM0090 Rev 17

RM0090
Figure 202.
Figure 203.
Figure 204.
Figure 205.
Figure 206.
Figure 207.
Figure 208.
Figure 209.
Figure 210.
Figure 211.
Figure 212.
Figure 213.
Figure 214.
Figure 215.
Figure 216.
Figure 217.
Figure 218.
Figure 219.
Figure 220.
Figure 221.
Figure 222.
Figure 223.
Figure 224.
Figure 225.
Figure 226.
Figure 227.
Figure 228.
Figure 229.
Figure 230.
Figure 231.
Figure 232.
Figure 233.
Figure 234.
Figure 235.
Figure 236.
Figure 237.
Figure 238.
Figure 239.
Figure 240.
Figure 241.
Figure 242.
Figure 243.
Figure 244.
Figure 245.
Figure 246.
Figure 247.
Figure 248.
Figure 249.
Figure 250.
Figure 251.

List of figures
Control circuit in trigger mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670
Basic timer block diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 696
Counter timing diagram with prescaler division change from 1 to 2 . . . . . . . . . . . . . . . . . 698
Counter timing diagram with prescaler division change from 1 to 4 . . . . . . . . . . . . . . . . . 698
Counter timing diagram, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699
Counter timing diagram, internal clock divided by 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 700
Counter timing diagram, internal clock divided by 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 700
Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 700
Counter timing diagram, update event when ARPE=0 (TIMx_ARR not
preloaded). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701
Counter timing diagram, update event when ARPE=1 (TIMx_ARR
preloaded). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701
Control circuit in normal mode, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . 702
Independent watchdog block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 709
Watchdog block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 714
Window watchdog timing diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715
Block diagram (STM32F415/417xx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 722
Block diagram (STM32F43xxx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723
DES/TDES-ECB mode encryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
DES/TDES-ECB mode decryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
DES/TDES-CBC mode encryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 727
DES/TDES-CBC mode decryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 728
AES-ECB mode encryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 729
AES-ECB mode decryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 730
AES-CBC mode encryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 731
AES-CBC mode decryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 732
AES-CTR mode encryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 733
AES-CTR mode decryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 734
Initial counter block structure for the Counter mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 734
64-bit block construction according to DATATYPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 741
Initialization vectors use in the TDES-CBC encryption . . . . . . . . . . . . . . . . . . . . . . . . . . . 743
CRYP interrupt mapping diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748
Block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767
Block diagram for STM32F415/417xx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773
Block diagram for STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774
Bit, byte and half-word swapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 776
HASH interrupt mapping diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 782
RTC block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 800
I2C bus protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 841
I2C block diagram for STM32F40x/41x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 842
I2C block diagram for STM32F42x/43x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 843
Transfer sequence diagram for slave transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 845
Transfer sequence diagram for slave receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 846
Transfer sequence diagram for master transmitter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 849
Transfer sequence diagram for master receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 850
I2C interrupt mapping diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 859
SPI block diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 876
Single master/ single slave application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 877
Data clock timing diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 879
TI mode - Slave mode, single transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881
TI mode - Slave mode, continuous transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 882
TI mode - master mode, single transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 883

RM0090 Rev 17

51/1747
56

List of figures

RM0090

Figure 252. TI mode - master mode, continuous transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 884
Figure 253. TXE/RXNE/BSY behavior in Master / full-duplex mode (BIDIMODE=0 and
RXONLY=0) in case of continuous transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 887
Figure 254. TXE/RXNE/BSY behavior in Slave / full-duplex mode (BIDIMODE=0,
RXONLY=0) in case of continuous transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 888
Figure 255. TXE/BSY behavior in Master transmit-only mode (BIDIMODE=0 and RXONLY=0)
in case of continuous transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 889
Figure 256. TXE/BSY in Slave transmit-only mode (BIDIMODE=0 and RXONLY=0) in case of
continuous transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 889
Figure 257. RXNE behavior in receive-only mode (BIDIRMODE=0 and RXONLY=1)
in case of continuous transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 890
Figure 258. TXE/BSY behavior when transmitting (BIDIRMODE=0 and RXONLY=0)
in case of discontinuous transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 891
Figure 259. Transmission using DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 896
Figure 260. Reception using DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 896
Figure 261. TI mode frame format error detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 898
Figure 262. I2S block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 899
Figure 263. I2S full duplex block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 900
Figure 264. I2S Philips protocol waveforms (16/32-bit full accuracy, CPOL = 0). . . . . . . . . . . . . . . . . 902
Figure 265. I2S Philips standard waveforms (24-bit frame with CPOL = 0) . . . . . . . . . . . . . . . . . . . . . 902
Figure 266. Transmitting 0x8EAA33 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902
Figure 267. Receiving 0x8EAA33 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 903
Figure 268. I2S Philips standard (16-bit extended to 32-bit packet frame with CPOL = 0) . . . . . . . . . 903
Figure 269. Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 903
Figure 270. MSB justified 16-bit or 32-bit full-accuracy length with CPOL = 0 . . . . . . . . . . . . . . . . . . 904
Figure 271. MSB justified 24-bit frame length with CPOL = 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 904
Figure 272. MSB justified 16-bit extended to 32-bit packet frame with CPOL = 0 . . . . . . . . . . . . . . . . 904
Figure 273. LSB justified 16-bit or 32-bit full-accuracy with CPOL = 0 . . . . . . . . . . . . . . . . . . . . . . . . 905
Figure 274. LSB justified 24-bit frame length with CPOL = 0. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 905
Figure 275. Operations required to transmit 0x3478AE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 905
Figure 276. Operations required to receive 0x3478AE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 906
Figure 277. LSB justified 16-bit extended to 32-bit packet frame with CPOL = 0 . . . . . . . . . . . . . . . . 906
Figure 278. Example of LSB justified 16-bit extended to 32-bit packet frame . . . . . . . . . . . . . . . . . . . 906
Figure 279. PCM standard waveforms (16-bit) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 907
Figure 280. PCM standard waveforms (16-bit extended to 32-bit packet frame). . . . . . . . . . . . . . . . . 907
Figure 281. Audio sampling frequency definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 908
Figure 282. I2S clock generator architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 908
Figure 283. Functional block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 928
Figure 284. Audio frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 930
Figure 285. FS role is start of frame + channel side identification (FSDEF = TRIS = 1) . . . . . . . . . . . 933
Figure 286. FS role is start of frame (FSDEF = 0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 933
Figure 287. Slot size configuration with FBOFF = 0 in SAI_xSLOTR . . . . . . . . . . . . . . . . . . . . . . . . . 934
Figure 288. First bit offset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 934
Figure 289. Audio block clock generator overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 935
Figure 290. AC’97 audio frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 939
Figure 291. Data companding hardware in an audio block in the SAI . . . . . . . . . . . . . . . . . . . . . . . . . 941
Figure 292. Tristate strategy on SD output line on an inactive slot . . . . . . . . . . . . . . . . . . . . . . . . . . . 943
Figure 293. Tristate on output data line in a protocol like I2S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 944
Figure 294. Overrun detection error. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 945
Figure 295. FIFO underrun event . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 946
Figure 296. USART block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 968
Figure 297. Word length programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 969

52/1747

RM0090 Rev 17

RM0090
Figure 298.
Figure 299.
Figure 300.
Figure 301.
Figure 302.
Figure 303.
Figure 304.
Figure 305.
Figure 306.
Figure 307.
Figure 308.
Figure 309.
Figure 310.
Figure 311.
Figure 312.
Figure 313.
Figure 314.
Figure 315.
Figure 316.
Figure 317.
Figure 318.
Figure 319.
Figure 320.
Figure 321.
Figure 322.
Figure 323.
Figure 324.
Figure 325.
Figure 326.
Figure 327.
Figure 328.
Figure 329.
Figure 330.
Figure 331.
Figure 332.
Figure 333.
Figure 334.
Figure 335.
Figure 336.
Figure 337.
Figure 338.
Figure 339.
Figure 340.
Figure 341.
Figure 342.
Figure 343.
Figure 344.
Figure 345.
Figure 346.
Figure 347.
Figure 348.
Figure 349.

List of figures
Configurable stop bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 971
TC/TXE behavior when transmitting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 972
Start bit detection when oversampling by 16 or 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 973
Data sampling when oversampling by 16 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 976
Data sampling when oversampling by 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 977
Mute mode using Idle line detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 990
Mute mode using address mark detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 990
Break detection in LIN mode (11-bit break length - LBDL bit is set) . . . . . . . . . . . . . . . . . 993
Break detection in LIN mode vs. Framing error detection. . . . . . . . . . . . . . . . . . . . . . . . . 994
USART example of synchronous transmission. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 995
USART data clock timing diagram (M=0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 995
USART data clock timing diagram (M=1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 996
RX data setup/hold time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 996
ISO 7816-3 asynchronous protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 997
Parity error detection using the 1.5 stop bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 998
IrDA SIR ENDEC- block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1000
IrDA data modulation (3/16) -Normal mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1000
Transmission using DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1002
Reception using DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1003
Hardware flow control between 2 USARTs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1003
RTS flow control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1004
CTS flow control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1005
USART interrupt mapping diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1006
SDIO “no response” and “no data” operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1020
SDIO (multiple) block read operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1020
SDIO (multiple) block write operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1021
SDIO sequential read operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1021
SDIO sequential write operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1021
SDIO block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1022
SDIO adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1023
Control unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1024
SDIO adapter command path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1025
Command path state machine (CPSM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1026
SDIO command transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1027
Data path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1029
Data path state machine (DPSM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1030
CAN network topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1077
Dual CAN block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1079
bxCAN operating modes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1081
bxCAN in silent mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1082
bxCAN in loop back mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1082
bxCAN in combined mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1083
Transmit mailbox states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1084
Receive FIFO states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1085
Filter bank scale configuration - register organization . . . . . . . . . . . . . . . . . . . . . . . . . . 1088
Example of filter numbering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1089
Filtering mechanism - example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1090
CAN error state diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1091
Bit timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1093
CAN frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1094
Event flags and interrupt generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1095
RX and TX mailboxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1106

RM0090 Rev 17

53/1747
56

List of figures
Figure 350.
Figure 351.
Figure 352.
Figure 353.
Figure 354.
Figure 355.
Figure 356.
Figure 357.
Figure 358.
Figure 359.
Figure 360.
Figure 361.
Figure 362.
Figure 363.
Figure 364.
Figure 365.
Figure 366.
Figure 367.
Figure 368.
Figure 369.
Figure 370.
Figure 371.
Figure 372.
Figure 373.
Figure 374.
Figure 375.
Figure 376.
Figure 377.
Figure 378.
Figure 379.
Figure 380.
Figure 381.
Figure 382.
Figure 383.
Figure 384.
Figure 385.
Figure 386.
Figure 387.
Figure 388.
Figure 389.
Figure 390.
Figure 391.
Figure 392.
Figure 393.
Figure 394.
Figure 395.
Figure 396.
Figure 397.
Figure 398.
Figure 399.
Figure 400.
Figure 401.

54/1747

RM0090

ETH block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1125
SMI interface signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1126
MDIO timing and frame structure - Write cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1127
MDIO timing and frame structure - Read cycle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1128
Media independent interface signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1129
MII clock sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1131
Reduced media-independent interface signals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1131
RMII clock sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1132
Clock scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1132
Address field format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1134
MAC frame format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1136
Tagged MAC frame format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1136
Transmission bit order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1143
Transmission with no collision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1143
Transmission with collision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1144
Frame transmission in MMI and RMII modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1144
Receive bit order. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1148
Reception with no error. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1149
Reception with errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1149
Reception with false carrier indication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1149
MAC core interrupt masking scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1150
Wakeup frame filter register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1155
Networked time synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1158
System time update using the Fine correction method. . . . . . . . . . . . . . . . . . . . . . . . . . 1160
PTP trigger output to TIM2 ITR1 connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1162
PPS output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1163
Descriptor ring and chain structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1164
TxDMA operation in Default mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1168
TxDMA operation in OSF mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1170
Normal transmit descriptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1171
Enhanced transmit descriptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1177
Receive DMA operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1179
Normal Rx DMA descriptor structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1181
Enhanced receive descriptor field format with IEEE1588 time stamp enabled. . . . . . . . 1187
Interrupt scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1190
Ethernet MAC remote wakeup frame filter register (ETH_MACRWUFFR). . . . . . . . . . . 1200
OTG full-speed block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1244
OTG A-B device connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246
USB peripheral-only connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1248
USB host-only connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1252
SOF connectivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1256
Updating OTG_FS_HFIR dynamically . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1259
Device-mode FIFO address mapping and AHB FIFO access mapping . . . . . . . . . . . . . 1260
Host-mode FIFO address mapping and AHB FIFO access mapping . . . . . . . . . . . . . . . 1261
Interrupt hierarchy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1265
CSR memory map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1267
Transmit FIFO write task . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1339
Receive FIFO read task . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1340
Normal bulk/control OUT/SETUP and bulk/control IN transactions . . . . . . . . . . . . . . . . 1341
Bulk/control IN transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1344
Normal interrupt OUT/IN transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1346
Normal isochronous OUT/IN transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1351

RM0090 Rev 17

RM0090
Figure 402.
Figure 403.
Figure 404.
Figure 405.
Figure 406.
Figure 407.
Figure 408.
Figure 409.
Figure 410.
Figure 411.
Figure 412.
Figure 413.
Figure 414.
Figure 415.
Figure 416.
Figure 417.
Figure 418.
Figure 419.
Figure 420.
Figure 421.
Figure 422.
Figure 423.
Figure 424.
Figure 425.
Figure 426.
Figure 427.
Figure 428.
Figure 429.
Figure 430.
Figure 431.
Figure 432.
Figure 433.
Figure 434.
Figure 435.
Figure 436.
Figure 437.
Figure 438.
Figure 439.
Figure 440.
Figure 441.
Figure 442.
Figure 443.
Figure 444.
Figure 445.
Figure 446.
Figure 447.
Figure 448.
Figure 449.
Figure 450.
Figure 451.

List of figures
Receive FIFO packet read . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1357
Processing a SETUP packet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1359
Bulk OUT transaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1365
TRDT max timing case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1373
A-device SRP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1374
B-device SRP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1375
A-device HNP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1376
B-device HNP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1378
USB OTG interface block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1383
USB host-only connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1390
SOF trigger output to TIM2 ITR1 connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1395
Updating OTG_HS_HFIR dynamically . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1397
Interrupt hierarchy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1400
CSR memory map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1402
Transmit FIFO write task . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1491
Receive FIFO read task . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1492
Normal bulk/control OUT/SETUP and bulk/control IN transactions - DMA
mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1493
Normal bulk/control OUT/SETUP and bulk/control IN transactions - Slave
mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1494
Bulk/control IN transactions - DMA mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1497
Bulk/control IN transactions - Slave mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1498
Normal interrupt OUT/IN transactions - DMA mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1500
Normal interrupt OUT/IN transactions - Slave mode . . . . . . . . . . . . . . . . . . . . . . . . . . . 1501
Normal isochronous OUT/IN transactions - DMA mode . . . . . . . . . . . . . . . . . . . . . . . . . 1506
Normal isochronous OUT/IN transactions - Slave mode . . . . . . . . . . . . . . . . . . . . . . . . 1507
Receive FIFO packet read in slave mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1518
Processing a SETUP packet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1520
Slave mode bulk OUT transaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1526
TRDT max timing case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1535
A-device SRP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1536
B-device SRP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1537
A-device HNP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1538
B-device HNP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1540
FSMC block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1543
FSMC memory banks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1545
Mode1 read accesses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1552
Mode1 write accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1553
ModeA read accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1554
ModeA write accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1555
Mode2 and mode B read accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1557
Mode2 write accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1557
Mode B write accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1558
Mode C read accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1560
Mode C write accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1560
Mode D read accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1562
Mode D write accesses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1563
Multiplexed read accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1565
Multiplexed write accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1565
Asynchronous wait during a read access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1568
Asynchronous wait during a write access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1568
Wait configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1570

RM0090 Rev 17

55/1747
56

List of figures
Figure 452.
Figure 453.
Figure 454.
Figure 455.
Figure 456.
Figure 457.
Figure 458.
Figure 459.
Figure 460.
Figure 461.
Figure 462.
Figure 463.
Figure 464.
Figure 465.
Figure 466.
Figure 467.
Figure 468.
Figure 469.
Figure 470.
Figure 471.
Figure 472.
Figure 473.
Figure 474.
Figure 475.
Figure 476.
Figure 477.
Figure 478.
Figure 479.
Figure 480.
Figure 481.
Figure 482.
Figure 483.
Figure 484.
Figure 485.
Figure 486.
Figure 487.
Figure 488.

56/1747

RM0090

Synchronous multiplexed read mode - NOR, PSRAM (CRAM) . . . . . . . . . . . . . . . . . . . 1571
Synchronous multiplexed write mode - PSRAM (CRAM) . . . . . . . . . . . . . . . . . . . . . . . . 1573
NAND/PC Card controller timing for common memory access . . . . . . . . . . . . . . . . . . . 1586
Access to non ‘CE don’t care’ NAND-Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1587
FMC block diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1602
FMC memory banks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1605
Mode1 read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1616
Mode1 write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1616
ModeA read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1618
ModeA write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1619
Mode2 and mode B read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1621
Mode2 write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1621
ModeB write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1622
ModeC read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1624
ModeC write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1624
ModeD read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1626
ModeD write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1627
Muxed read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1629
Muxed write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1629
Asynchronous wait during a read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . 1631
Asynchronous wait during a write access waveforms. . . . . . . . . . . . . . . . . . . . . . . . . . . 1632
Wait configuration waveforms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1634
Synchronous multiplexed read mode waveforms - NOR, PSRAM (CRAM) . . . . . . . . . . 1634
Synchronous multiplexed write mode waveforms - PSRAM (CRAM). . . . . . . . . . . . . . . 1636
NAND Flash/PC Card controller waveforms for common memory access. . . . . . . . . . . 1649
Access to non ‘CE don’t care’ NAND-Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1650
Burst write SDRAM access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1663
Burst read SDRAM access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1664
Logic diagram of Read access with RBURST bit set (CAS=2, RPIPE=0) . . . . . . . . . . . 1665
Read access crossing row boundary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1667
Write access crossing row boundary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1667
Self-refresh mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1670
Power-down mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1671
Block diagram of STM32 MCU and Cortex®-M4 with FPU-level debug support . . . . . . 1681
SWJ debug port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1683
JTAG TAP connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1687
TPIU block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1705

RM0090 Rev 17

RM0090

1

Documentation conventions

Documentation conventions
The STM32F401xx devices have an Arm®(a) Cortex®-M4 with FPU and an Arm® Cortex®A7 with FPU cores.

1.1

List of abbreviations for registers
The following abbreviations are used in register descriptions:
read/write (rw)

Software can read and write to these bits.

read-only (r)

Software can only read these bits.

write-only (w)

Software can only write to this bit. Reading the bit returns the reset
value.

read/clear (rc_w1) Software can read as well as clear this bit by writing 1. Writing ‘0’ has
no effect on the bit value.
read/clear (rc_w0) Software can read as well as clear this bit by writing 0. Writing ‘1’ has
no effect on the bit value.
read/clear by read Software can read this bit. Reading this bit automatically clears it to ‘0’.
(rc_r)
Writing ‘0’ has no effect on the bit value.
read/set (rs)

Software can read as well as set this bit. Writing ‘0’ has no effect on the
bit value.

read-only write
trigger (rt_w)

Software can read this bit. Writing ‘0’ or ‘1’ triggers an event but has no
effect on the bit value.

toggle (t)

Software can only toggle this bit by writing ‘1’. Writing ‘0’ has no effect.

Reserved (Res.)

Reserved bit, must be kept at reset value.

a. Arm is a registered trademark of Arm Limited (or its subsidiaries) in the US and/or elsewhere.

RM0090 Rev 17

57/1747
112

Documentation conventions

1.2

RM0090

Glossary
This section gives a brief definition of acronyms and abbreviations used in this document:
•

The CPU core integrates two debug ports:
–

JTAG debug port (JTAG-DP) provides a 5-pin standard interface based on the
Joint Test Action Group (JTAG) protocol.

–

SWD debug port (SWD-DP) provides a 2-pin (clock and data) interface based on
the Serial Wire Debug (SWD) protocol.
For both the JTAG and SWD protocols, please refer to the Cortex®-M4 with FPU
Technical Reference Manual

1.3

•

Word: data/instruction of 32-bit length.

•

Half word: data/instruction of 16-bit length.

•

Byte: data of 8-bit length.

•

Double word: data of 64-bit length.

•

IAP (in-application programming): IAP is the ability to reprogram the Flash memory of a
microcontroller while the user program is running.

•

ICP (in-circuit programming): ICP is the ability to program the Flash memory of a
microcontroller using the JTAG protocol, the SWD protocol or the bootloader while the
device is mounted on the user application board.

•

I-Code: this bus connects the Instruction bus of the CPU core to the Flash instruction
interface. Prefetch is performed on this bus.

•

D-Code: this bus connects the D-Code bus (literal load and debug access) of the CPU
to the Flash data interface.

•

Option bytes: product configuration bits stored in the Flash memory.

•

OBL: option byte loader.

•

AHB: advanced high-performance bus.

•

CPU: refers to the Cortex®-M4 with FPU core.

Peripheral availability
For peripheral availability and number across all STM32F405xx/07xx and
STM32F415xx/17xx sales types, please refer to the STM32F405xx/07xx and
STM32F415xx/17xx datasheets.
For peripheral availability and number across all STM32F42xxx and STM32F43xxx sales
types, please refer to the STM32F42xxx and STM32F43xxx datasheets.

58/1747

RM0090 Rev 17

RM0090

Memory and bus architecture

2

Memory and bus architecture

2.1

System architecture
In STM32F405xx/07xx and STM32F415xx/17xx, the main system consists of 32-bit
multilayer AHB bus matrix that interconnects:
The main system consists of 32-bit multilayer AHB bus matrix that interconnects:
•

•

Eight masters:
–

Cortex®-M4 with FPU core I-bus, D-bus and S-bus

–

DMA1 memory bus

–

DMA2 memory bus

–

DMA2 peripheral bus

–

Ethernet DMA bus

–

USB OTG HS DMA bus

Seven slaves:
–

Internal Flash memory ICode bus

–

Internal Flash memory DCode bus

–

Main internal SRAM1 (112 KB)

–

Auxiliary internal SRAM2 (16 KB)

–

AHB1 peripherals including AHB to APB bridges and APB peripherals

–

AHB2 peripherals

–

FSMC

The bus matrix provides access from a master to a slave, enabling concurrent access and
efficient operation even when several high-speed peripherals work simultaneously. The 64Kbyte CCM (core coupled memory) data RAM is not part of the bus matrix and can be
accessed only through the CPU. This architecture is shown in Figure 1.

RM0090 Rev 17

59/1747
112

Memory and bus architecture

RM0090

Figure 1. System architecture for STM32F405xx/07xx and STM32F415xx/17xx devices

)#/$%
$#/$%

!##%,

53"?(3?-

-!#
53" /4'
%THERNET
(3
%4(%2.%4?-

$-!?0

'0
$-!
$-!?-%-

$-!?-%-

'0
$-!
$-!?0)

3 BUS

) BUS

$ BUS

!2#ORTEX -

 +BYTE
##- DATA 2!-

&LASH
MEMORY
32!-
 +BYTE
32!-
 +BYTE
!("
PERIPHERALS
!("
PERIPHERALS
&3-#
3TATIC -EM#TL

!0"
!0"

"US MATRIX 3
AID

60/1747

RM0090 Rev 17

RM0090

Memory and bus architecture
In the STM32F42xx and STM32F43xx devices, the main system consists of 32-bit multilayer
AHB bus matrix that interconnects:
•

•

Ten masters:
–

Cortex®-M4 with FPU core I-bus, D-bus and S-bus

–

DMA1 memory bus

–

DMA2 memory bus

–

DMA2 peripheral bus

–

Ethernet DMA bus

–

USB OTG HS DMA bus

–

LCD Controller DMA-bus

–

DMA2D (Chrom-Art Accelerator™) memory bus

Eight slaves:
–

Internal Flash memory ICode bus

–

Internal Flash memory DCode bus

–

Main internal SRAM1 (112 KB)

–

Auxiliary internal SRAM2 (16 KB)

–

Auxiliary internal SRAM3 (64 KB)

–

AHB1peripherals including AHB to APB bridges and APB peripherals

–

AHB2 peripherals

–

FMC

The bus matrix provides access from a master to a slave, enabling concurrent access and
efficient operation even when several high-speed peripherals work simultaneously. The 64Kbyte CCM (core coupled memory) data RAM is not part of the bus matrix and can be
accessed only through the CPU. This architecture is shown in Figure 2.

RM0090 Rev 17

61/1747
112

Memory and bus architecture

RM0090

Figure 2. System architecture for STM32F42xxx and STM32F43xxx devices
&KURP$57$FFHOHUDWRU
'0$'

,&2'(
'&2'(

$&&(/

'0$'

/&'7)7B0

86%B+6B0

0$&
86%27* /&'7)7
(WKHUQHW
+6
(7+(51(7B0

'0$B3

*3
'0$
'0$B0(0

'0$B0(0

*3
'0$
'0$B3,

6EXV

'EXV

$50
&RUWH[0
,EXV

.E\WH
&&0GDWD5$0

)ODVK
PHPRU\
65$0
.E\WH
65$0
.E\WH
65$0
.E\WH
$+%
SHULSKHUDOV

$3%

$+%
SHULSKHUDOV
)0&H[WHUQDO
0HP&WO

$3%

%XVPDWUL[6
-36

2.1.1

I-bus
This bus connects the Instruction bus of the Cortex®-M4 with FPU core to the BusMatrix.
This bus is used by the core to fetch instructions. The target of this bus is a memory
containing code (internal Flash memory/SRAM or external memories through the
FSMC/FMC).

2.1.2

D-bus
This bus connects the databus of the Cortex®-M4 with FPU to the 64-Kbyte CCM data RAM
to the BusMatrix. This bus is used by the core for literal load and debug access. The target
of this bus is a memory containing code or data (internal Flash memory or external
memories through the FSMC/FMC).

2.1.3

S-bus
This bus connects the system bus of the Cortex®-M4 with FPU core to a BusMatrix. This
bus is used to access data located in a peripheral or in SRAM. Instructions may also be
fetched on this bus (less efficient than ICode). The targets of this bus are the internal
SRAM1, SRAM2 and SRAM3, the AHB1 peripherals including the APB peripherals, the
AHB2 peripherals and the external memories through the FSMC/FMC.

62/1747

RM0090 Rev 17

RM0090

2.1.4

Memory and bus architecture

DMA memory bus
This bus connects the DMA memory bus master interface to the BusMatrix. It is used by the
DMA to perform transfer to/from memories. The targets of this bus are data memories:
internal SRAMs (SRAM1, SRAM2 and SRAM3) and external memories through the
FSMC/FMC.

2.1.5

DMA peripheral bus
This bus connects the DMA peripheral master bus interface to the BusMatrix. This bus is
used by the DMA to access AHB peripherals or to perform memory-to-memory transfers.
The targets of this bus are the AHB and APB peripherals plus data memories: internal
SRAMs (SRAM1, SRAM2 and SRAM3) and external memories through the FSMC/FMC.

2.1.6

Ethernet DMA bus
This bus connects the Ethernet DMA master interface to the BusMatrix. This bus is used by
the Ethernet DMA to load/store data to a memory. The targets of this bus are data
memories: internal SRAMs (SRAM1, SRAM2, SRAM3), internal Flash memory, and
external memories through the FSMC/FMC.

2.1.7

USB OTG HS DMA bus
This bus connects the USB OTG HS DMA master interface to the BusMatrix. This bus is
used by the USB OTG DMA to load/store data to a memory. The targets of this bus are data
memories: internal SRAMs (SRAM1, SRAM2, SRAM3), internal Flash memory, and
external memories through the FSMC/FMC.

2.1.8

LCD-TFT controller DMA bus
This bus connects the LCD controller DMA master interface to the BusMatrix. It is used by
the LCD-TFT DMA to load/store data to a memory. The targets of this bus are data
memories: internal SRAMs (SRAM1, SRAM2, SRAM3), external memories through FMC,
and internal Flash memory.

2.1.9

DMA2D bus
This bus connect the DMA2D master interface to the BusMatrix. This bus is used by the
DMA2D graphic Accelerator to load/store data to a memory. The targets of this bus are data
memories: internal SRAMs (SRAM1, SRAM2, SRAM3), external memories through FMC,
and internal Flash memory.

2.1.10

BusMatrix
The BusMatrix manages the access arbitration between masters. The arbitration uses a
round-robin algorithm.

RM0090 Rev 17

63/1747
112

Memory and bus architecture

2.1.11

RM0090

AHB/APB bridges (APB)
The two AHB/APB bridges, APB1 and APB2, provide full synchronous connections between
the AHB and the two APB buses, allowing flexible selection of the peripheral frequency.
Refer to the device datasheets for more details on APB1 and APB2 maximum frequencies,
and to Table 1 for the address mapping of AHB and APB peripherals.
After each device reset, all peripheral clocks are disabled (except for the SRAM and Flash
memory interface). Before using a peripheral you have to enable its clock in the
RCC_AHBxENR or RCC_APBxENR register.

Note:

When a 16- or an 8-bit access is performed on an APB register, the access is transformed
into a 32-bit access: the bridge duplicates the 16- or 8-bit data to feed the 32-bit vector.

2.2

Memory organization
Program memory, data memory, registers and I/O ports are organized within the same linear
4 Gbyte address space.
The bytes are coded in memory in little endian format. The lowest numbered byte in a word
is considered the word’s least significant byte and the highest numbered byte, the word’s
most significant.
For the detailed mapping of peripheral registers, please refer to the related chapters.
The addressable memory space is divided into 8 main blocks, each of 512 MB.
All the memory areas that are not allocated to on-chip memories and peripherals are
considered “Reserved”). Refer to the memory map figure in the product datasheet.

2.3

Memory map
See the datasheet corresponding to your device for a comprehensive diagram of the
memory map. Table 1 gives the boundary addresses of the peripherals available in all
STM32F4xx devices.
Table 1. STM32F4xx register boundary addresses
Boundary address

Peripheral

Bus

FSMC control register
(STM32F405xx/07xx
and
0xA000 0000 - 0xA000 0FFF STM32F415xx/17xx)/ AHB3
FMC control register
(STM32F42xxx and
STM32F43xxx)

64/1747

Register map

Section 36.6.9: FSMC register map on page 1598
Section 37.8: FMC register map on page 1678

RM0090 Rev 17

RM0090

Memory and bus architecture
Table 1. STM32F4xx register boundary addresses (continued)

Boundary address

Peripheral

Bus

0x5006 0800 - 0x5006 0BFF

RNG

Section 24.4.4: RNG register map on page 771

0x5006 0400 - 0x5006 07FF

HASH

Section 25.4.9: HASH register map on page 795

0x5006 0000 - 0x5006 03FF

CRYP

0x5005 0000 - 0x5005 03FF

DCMI

0x5000 0000 - 0x5003 FFFF

USB OTG FS

Section 34.16.6: OTG_FS register map on
page 1327

0x4004 0000 - 0x4007 FFFF

USB OTG HS

Section 35.12.6: OTG_HS register map on
page 1471

0x4002 B000 - 0x4002 BBFF

DMA2D

0x4002 8000 - 0x4002 93FF

ETHERNET MAC

0x4002 6400 - 0x4002 67FF

DMA2

0x4002 6000 - 0x4002 63FF

DMA1

0x4002 4000 - 0x4002 4FFF

BKPSRAM

0x4002 3C00 - 0x4002 3FFF

Flash interface
register

0x4002 3800 - 0x4002 3BFF

RCC

0x4002 3000 - 0x4002 33FF

CRC

0x4002 2800 - 0x4002 2BFF

GPIOK

0x4002 2400 - 0x4002 27FF

GPIOJ

0x4002 2000 - 0x4002 23FF

GPIOI

0x4002 1C00 - 0x4002 1FFF

GPIOH

0x4002 1800 - 0x4002 1BFF

GPIOG

0x4002 1400 - 0x4002 17FF

GPIOF

0x4002 1000 - 0x4002 13FF

GPIOE

0x4002 0C00 - 0x4002 0FFF

GPIOD

0x4002 0800 - 0x4002 0BFF

GPIOC

0x4002 0400 - 0x4002 07FF

GPIOB

0x4002 0000 - 0x4002 03FF

GPIOA

0x4001 6800 - 0x4001 6BFF

LCD-TFT

0x4001 5800 - 0x4001 5BFF

SAI1

0x4001 5400 - 0x4001 57FF

SPI6

0x4001 5000 - 0x4001 53FF

SPI5

AHB2

Register map

Section 23.6.13: CRYP register map on page 763
Section 15.8.12: DCMI register map on page 478

Section 11.5: DMA2D registers on page 352
Section 33.8.5: Ethernet register maps on
page 1237
Section 10.5.11: DMA register map on page 335

Section 3.9: Flash interface registers
Section 7.3.24: RCC register map on page 265
AHB1

Section 4.4.4: CRC register map on page 115
Section 8.4.11: GPIO register map on page 287

Section 8.4.11: GPIO register map on page 287

APB2

APB2

Section 16.7.26: LTDC register map on page 512
Section 29.17.9: SAI register map on page 963
Section 28.5.10: SPI register map on page 925

RM0090 Rev 17

65/1747
112

Memory and bus architecture

RM0090

Table 1. STM32F4xx register boundary addresses (continued)
Boundary address

Peripheral

0x4001 4800 - 0x4001 4BFF

TIM11

0x4001 4400 - 0x4001 47FF

TIM10

0x4001 4000 - 0x4001 43FF

TIM9

0x4001 3C00 - 0x4001 3FFF

EXTI

Bus

Register map
Section 19.5.12: TIM10/11/13/14 register map on
page 694
Section 19.4.13: TIM9/12 register map on
page 684

APB2 Section 12.3.7: EXTI register map on page 387
Section 9.2.8: SYSCFG register maps for
STM32F405xx/07xx and STM32F415xx/17xx on
page 294 and Section 9.3.8: SYSCFG register
maps for STM32F42xxx and STM32F43xxx on
page 301

0x4001 3800 - 0x4001 3BFF

SYSCFG

0x4001 3400 - 0x4001 37FF

SPI4

APB2 Section 28.5.10: SPI register map on page 925

0x4001 3000 - 0x4001 33FF

SPI1

Section 28.5.10: SPI register map on page 925

0x4001 2C00 - 0x4001 2FFF

SDIO

Section 31.9.16: SDIO register map on page 1074

0x4001 2000 - 0x4001 23FF

ADC1 - ADC2 - ADC3

Section 13.13.18: ADC register map on page 430

0x4001 1400 - 0x4001 17FF

USART6

0x4001 1000 - 0x4001 13FF

USART1

0x4001 0400 - 0x4001 07FF

TIM8

0x4001 0000 - 0x4001 03FF

TIM1

0x4000 7C00 - 0x4000 7FFF

UART8

0x4000 7800 - 0x4000 7BFF

UART7

66/1747

APB2

Section 30.6.8: USART register map on page 1018
Section 17.4.21: TIM1 and TIM8 register map on
page 587

APB1 Section 30.6.8: USART register map on page 1018

RM0090 Rev 17

RM0090

Memory and bus architecture
Table 1. STM32F4xx register boundary addresses (continued)

Boundary address

Peripheral

Bus

Register map

0x4000 7400 - 0x4000 77FF

DAC

Section 14.5.15: DAC register map on page 453

0x4000 7000 - 0x4000 73FF

PWR

Section 5.6: PWR register map on page 149

0x4000 6800 - 0x4000 6BFF

CAN2

0x4000 6400 - 0x4000 67FF

CAN1

0x4000 5C00 - 0x4000 5FFF

I2C3

0x4000 5800 - 0x4000 5BFF

I2C2

0x4000 5400 - 0x4000 57FF

I2C1

0x4000 5000 - 0x4000 53FF

UART5

0x4000 4C00 - 0x4000 4FFF

UART4

0x4000 4800 - 0x4000 4BFF

USART3

0x4000 4400 - 0x4000 47FF

USART2

0x4000 4000 - 0x4000 43FF

I2S3ext

0x4000 3C00 - 0x4000 3FFF

SPI3 / I2S3

0x4000 3800 - 0x4000 3BFF

SPI2 / I2S2

0x4000 3400 - 0x4000 37FF

I2S2ext

0x4000 3000 - 0x4000 33FF

IWDG

0x4000 2C00 - 0x4000 2FFF

WWDG

0x4000 2800 - 0x4000 2BFF

RTC & BKP Registers

Section 26.6.21: RTC register map on page 836

0x4000 2000 - 0x4000 23FF

TIM14

0x4000 1C00 - 0x4000 1FFF

TIM13

Section 19.5.12: TIM10/11/13/14 register map on
page 694

0x4000 1800 - 0x4000 1BFF

TIM12

0x4000 1400 - 0x4000 17FF

TIM7

0x4000 1000 - 0x4000 13FF

TIM6

0x4000 0C00 - 0x4000 0FFF

TIM5

0x4000 0800 - 0x4000 0BFF

TIM4

0x4000 0400 - 0x4000 07FF

TIM3

0x4000 0000 - 0x4000 03FF

TIM2

Section 32.9.5: bxCAN register map on page 1117

Section 27.6.11: I2C register map on page 872

Section 30.6.8: USART register map on page 1018

Section 28.5.10: SPI register map on page 925
APB1
Section 21.4.5: IWDG register map on page 712
Section 22.6.4: WWDG register map on page 719

Section 19.4.13: TIM9/12 register map on
page 684
Section 20.4.9: TIM6 and TIM7 register map on
page 707

Section 18.4.21: TIMx register map on page 648

RM0090 Rev 17

67/1747
112

Memory and bus architecture

2.3.1

RM0090

Embedded SRAM
The STM32F405xx/07xx and STM32F415xx/17xx feature 4 Kbytes of backup SRAM (see
Section 5.1.2: Battery backup domain) plus 192 Kbytes of system SRAM.
The STM32F42xxx and STM32F43xxx feature 4 Kbytes of backup SRAM (see
Section 5.1.2: Battery backup domain) plus 256 Kbytes of system SRAM.
The embedded SRAM can be accessed as bytes, half-words (16 bits) or full words (32 bits).
Read and write operations are performed at CPU speed with 0 wait state. The embedded
SRAM is divided into up to three blocks:
•

SRAM1 and SRAM2 mapped at address 0x2000 0000 and accessible by all AHB
masters.

•

SRAM3 (available on STM32F42xxx and STM32F43xxx) mapped at address
0x2002 0000 and accessible by all AHB masters.

•

CCM (core coupled memory) mapped at address 0x1000 0000 and accessible only by
the CPU through the D-bus.

The AHB masters support concurrent SRAM accesses (from the Ethernet or the USB OTG
HS): for instance, the Ethernet MAC can read/write from/to SRAM2 while the CPU is
reading/writing from/to SRAM1 or SRAM3.
The CPU can access the SRAM1, SRAM2, and SRAM3 through the System Bus or through
the I-Code/D-Code buses when boot from SRAM is selected or when physical remap is
selected (Section 9.2.1: SYSCFG memory remap register (SYSCFG_MEMRMP) in the
SYSCFG controller). To get the max performance on SRAM execution, physical remap
should be selected (boot or software selection).

2.3.2

Flash memory overview
The Flash memory interface manages CPU AHB I-Code and D-Code accesses to the Flash
memory. It implements the erase and program Flash memory operations and the read and
write protection mechanisms. It accelerates code execution with a system of instruction
prefetch and cache lines.
The Flash memory is organized as follows:
•

A main memory block divided into sectors.

•

System memory from which the device boots in System memory boot mode

•

512 OTP (one-time programmable) bytes for user data.

•

Option bytes to configure read and write protection, BOR level, watchdog
software/hardware and reset when the device is in Standby or Stop mode.

Refer to Section 3: Embedded Flash memory interface for more details.

2.3.3

Bit banding
The Cortex®-M4 with FPU memory map includes two bit-band regions. These regions map
each word in an alias region of memory to a bit in a bit-band region of memory. Writing to a
word in the alias region has the same effect as a read-modify-write operation on the
targeted bit in the bit-band region.
In the STM32F4xx devices both the peripheral registers and the SRAM are mapped to a bitband region, so that single bit-band write and read operations are allowed. The operations

68/1747

RM0090 Rev 17

RM0090

Memory and bus architecture
are only available for Cortex®-M4 with FPU accesses, and not from other bus masters (e.g.
DMA).
A mapping formula shows how to reference each word in the alias region to a corresponding
bit in the bit-band region. The mapping formula is:
bit_word_addr = bit_band_base + (byte_offset x 32) + (bit_number × 4)
where:
–

bit_word_addr is the address of the word in the alias memory region that maps to
the targeted bit

–

bit_band_base is the starting address of the alias region

–

byte_offset is the number of the byte in the bit-band region that contains the
targeted bit

–

bit_number is the bit position (0-7) of the targeted bit

Example
The following example shows how to map bit 2 of the byte located at SRAM address
0x20000300 to the alias region:
0x22006008 = 0x22000000 + (0x300*32) + (2*4)
Writing to address 0x22006008 has the same effect as a read-modify-write operation on bit
2 of the byte at SRAM address 0x20000300.
Reading address 0x22006008 returns the value (0x01 or 0x00) of bit 2 of the byte at SRAM
address 0x20000300 (0x01: bit set; 0x00: bit reset).
For more information on bit-banding, please refer to the Cortex®-M4 with FPU programming
manual (see Related documents on page 1).

2.4

Boot configuration
Due to its fixed memory map, the code area starts from address 0x0000 0000 (accessed
through the ICode/DCode buses) while the data area (SRAM) starts from address
0x2000 0000 (accessed through the system bus). The Cortex®-M4 with FPU CPU always
fetches the reset vector on the ICode bus, which implies to have the boot space available
only in the code area (typically, Flash memory). STM32F4xx microcontrollers implement a
special mechanism to be able to boot from other memories (like the internal SRAM).
In the STM32F4xx, three different boot modes can be selected through the BOOT[1:0] pins
as shown in Table 2.
Table 2. Boot modes
Boot mode selection pins
Boot mode

Aliasing

BOOT1

BOOT0

x

0

Main Flash memory Main Flash memory is selected as the boot space

0

1

System memory

System memory is selected as the boot space

1

1

Embedded SRAM

Embedded SRAM is selected as the boot space

RM0090 Rev 17

69/1747
112

Memory and bus architecture

RM0090

The values on the BOOT pins are latched on the 4th rising edge of SYSCLK after a reset. It
is up to the user to set the BOOT1 and BOOT0 pins after reset to select the required boot
mode.
BOOT0 is a dedicated pin while BOOT1 is shared with a GPIO pin. Once BOOT1 has been
sampled, the corresponding GPIO pin is free and can be used for other purposes.
The BOOT pins are also resampled when the device exits the Standby mode. Consequently,
they must be kept in the required Boot mode configuration when the device is in the Standby
mode. After this startup delay is over, the CPU fetches the top-of-stack value from address
0x0000 0000, then starts code execution from the boot memory starting from 0x0000 0004.
Note:

When the device boots from SRAM, in the application initialization code, you have to
relocate the vector table in SRAM using the NVIC exception table and the offset register.
In STM32F42xxx and STM32F43xxx devices, when booting from the main Flash memory,
the application software can either boot from bank 1 or from bank 2. By default, boot from
bank 1 is selected.
To select boot from Flash memory bank 2, set the BFB2 bit in the user option bytes. When
this bit is set and the boot pins are in the boot from main Flash memory configuration, the
device boots from system memory, and the boot loader jumps to execute the user
application programmed in Flash memory bank 2. For further details, please refer to
AN2606.

Embedded bootloader
The embedded bootloader mode is used to reprogram the Flash memory using one of the
following serial interfaces:
•

USART1 (PA9/PA10)

•

USART3 (PB10/11 and PC10/11)

•

CAN2 (PB5/13)

•

USB OTG FS (PA11/12) in Device mode (DFU: device firmware upgrade).

The USART peripherals operate at the internal 16 MHz oscillator (HSI) frequency, while the
CAN and USB OTG FS require an external clock (HSE) multiple of 1 MHz (ranging from 4 to
26 MHz).
The embedded bootloader code is located in system memory. It is programmed by ST
during production. For additional information, refer to application note AN2606.

Physical remap in STM32F405xx/07xx and STM32F415xx/17xx
Once the boot pins are selected, the application software can modify the memory
accessible in the code area (in this way the code can be executed through the ICode bus in
place of the System bus). This modification is performed by programming the Section 9.2.1:
SYSCFG memory remap register (SYSCFG_MEMRMP) in the SYSCFG controller.
The following memories can thus be remapped:

70/1747

•

Main Flash memory

•

System memory

•

Embedded SRAM1 (112 KB)

•

FSMC bank 1 (NOR/PSRAM 1 and 2)

RM0090 Rev 17

RM0090

Memory and bus architecture
Table 3. Memory mapping vs. Boot mode/physical remap
in STM32F405xx/07xx and STM32F415xx/17xx
Boot/Remap in
Boot/Remap in
Boot/Remap in
main Flash memory embedded SRAM System memory

Addresses

Remap in FSMC

0x2001 C000 - 0x2001 FFFF

SRAM2 (16 KB)

SRAM2 (16 KB)

SRAM2 (16 KB)

SRAM2 (16 KB)

0x2000 0000 - 0x2001 BFFF

SRAM1 (112 KB)

SRAM1 (112 KB)

SRAM1 (112 KB)

SRAM1 (112 KB)

0x1FFF 0000 - 0x1FFF 77FF

System memory

System memory

System memory

System memory

0x0810 0000 - 0x0FFF FFFF

Reserved

Reserved

Reserved

Reserved

0x0800 0000 - 0x080F FFFF

Flash memory

Flash memory

Flash memory

Flash memory

0x0400 0000 - 0x07FF FFFF

Reserved

Reserved

Reserved

FSMC bank 1
NOR/PSRAM 2
(128 MB Aliased)

Flash (1 MB) Aliased

SRAM1 (112 KB)
Aliased

System memory
(30 KB) Aliased

FSMC bank 1
NOR/PSRAM 1
(128 MB Aliased)

0x0000 0000 0x000F FFFF(1)(2)

1. When the FSMC is remapped at address 0x0000 0000, only the first two regions of bank 1 memory controller (bank 1
NOR/PSRAM 1 and NOR/PSRAM 2) can be remapped. In remap mode, the CPU can access the external memory via
ICode bus instead of System bus which boosts up the performance.
2. Even when aliased in the boot memory space, the related memory is still accessible at its original memory space.

Physical remap in STM32F42xxx and STM32F43xxx
Once the boot pins are selected, the application software can modify the memory
accessible in the code area (in this way the code can be executed through the ICode bus in
place of the System bus). This modification is performed by programming the Section 9.2.1:
SYSCFG memory remap register (SYSCFG_MEMRMP) in the SYSCFG controller.
The following memories can thus be remapped:
•

Main Flash memory

•

System memory

•

Embedded SRAM1 (112 KB)

•

FMC bank 1 (NOR/PSRAM 1 and 2)

•

FMC SDRAM bank 1
Table 4. Memory mapping vs. Boot mode/physical remap
in STM32F42xxx and STM32F43xxx

Addresses

Boot/Remap in
Boot/Remap in
Boot/Remap in
main Flash memory embedded SRAM System memory

Remap in FMC

0x2002 0000 - 0x2002 FFFF

SRAM3 (64 KB)

SRAM3 (64 KB)

SRAM3 (64 KB)

SRAM3 (64 KB)

0x2001 C000 - 0x2001 FFFF

SRAM2 (16 KB)

SRAM2 (16 KB)

SRAM2 (16 KB)

SRAM2 (16 KB)

0x2000 0000 - 0x2001 BFFF

SRAM1 (112 KB)

SRAM1 (112 KB)

SRAM1 (112 KB)

SRAM1 (112 KB)

0x1FFF 0000 - 0x1FFF 77FF

System memory

System memory

System memory

System memory

0x0810 0000 - 0x0FFF FFFF

Reserved

Reserved

Reserved

Reserved

0x0800 0000 - 0x081F FFFF

Flash memory

Flash memory

Flash memory

Flash memory

RM0090 Rev 17

71/1747
112

Memory and bus architecture

RM0090

Table 4. Memory mapping vs. Boot mode/physical remap
in STM32F42xxx and STM32F43xxx (continued)
Addresses

0x0400 0000 - 0x07FF FFFF

0x0000 0000 0x001F FFFF(1)(2)

Boot/Remap in
Boot/Remap in
Boot/Remap in
main Flash memory embedded SRAM System memory
Reserved

Reserved

Flash (2 MB) Aliased

SRAM1 (112 KB)
Aliased

Remap in FMC

Reserved

FMC bank 1
NOR/PSRAM 2
(128 MB Aliased)

System memory
(30 KB) Aliased

FMC bank 1
NOR/PSRAM 1
(128 MB Aliased)
or FMC SDRAM
bank 1 (128 MB
Aliased)

1. When the FMC is remapped at address 0x0000 0000, only the first two regions of bank 1 memory controller (bank 1
NOR/PSRAM 1 and NOR/PSRAM 2) or SDRAM bank 1 can be remapped. In remap mode, the CPU can access the
external memory via ICode bus instead of System bus which boosts up the performance.
2. Even when aliased in the boot memory space, the related memory is still accessible at its original memory space.

72/1747

RM0090 Rev 17

RM0090

Embedded Flash memory interface

3

Embedded Flash memory interface

3.1

Introduction
The Flash memory interface manages CPU AHB I-Code and D-Code accesses to the Flash
memory. It implements the erase and program Flash memory operations and the read and
write protection mechanisms.
The Flash memory interface accelerates code execution with a system of instruction
prefetch and cache lines.

3.2

Main features
•

Flash memory read operations

•

Flash memory program/erase operations

•

Read / write protections

•

Prefetch on I-Code

•

64 cache lines of 128 bits on I-Code

•

8 cache lines of 128 bits on D-Code

Figure 3 shows the Flash memory interface connection inside the system architecture.
Figure 3. Flash memory interface connection inside system architecture
(STM32F405xx/07xx and STM32F415xx/17xx)
#ORTEX - WITH &05
) #ODE BUS

) #ODE

!("
 BIT
INSTRUCTION &LASH INTERFACE
BUS

&LASH
MEMORY

$ #ODE

#ORTEX
CORE

3 BUS
$ CODE BUS

##- DATA
2!-

&LASH
MEMORY
BUS
 BITS

!("
 BIT
DATA BUS

&,)4& REGISTERS

!("
 BIT
SYSTEM BUS

!("
PERIPH

$-!

32!- AND
%XTERNAL
MEMORIES

53" (3

!("
PERIPH

$-!

%THERNET
!CCESS TO INSTRUCTION IN &LASH MEMORY
!CCESS TO DATA AND LITERAL POOL IN &LASH MEMORY
&,)4& REGISTER ACCESS
-36

RM0090 Rev 17

73/1747
112

Embedded Flash memory interface

RM0090

Figure 4. Flash memory interface connection inside system architecture
(STM32F42xxx and STM32F43xxx)
#ORTEX - WITH &05
) #ODE BUS

) #ODE

!("
 BIT
INSTRUCTION &LASH INTERFACE
BUS

&LASH
MEMORY

$ #ODE

#ORTEX
CORE

3 BUS
$ CODE BUS

##- DATA
2!-

!("
 BIT
DATA BUS

&,)4& REGISTERS

!("
 BIT
SYSTEM BUS

$-!

&LASH
MEMORY
BUS
 BITS

$-!

!("
PERIPH
32!- AND
EXTERNAL
MEMORIES
!("
PERIPH

$-!$

,#$ 4&4
53" (3
%THERNET
!CCESS TO INSTRUCTION IN &LASH MEMORY
!CCESS TO DATA AND LITERAL POOL IN &LASH MEMORY
&,)4& REGISTER ACCESS

3.3

-36

Embedded Flash memory in
STM32F405xx/07xx and STM32F415xx/17xx
The Flash memory has the following main features:
•

Capacity up to 1 Mbyte

•

128 bits wide data read

•

Byte, half-word, word and double word write

•

Sector and mass erase

•

Memory organization
The Flash memory is organized as follows:
–

A main memory block divided into 4 sectors of 16 Kbytes, 1 sector of 64 Kbytes,
and 7 sectors of 128 Kbytes

–

System memory from which the device boots in System memory boot mode

–

512 OTP (one-time programmable) bytes for user data
The OTP area contains 16 additional bytes used to lock the corresponding OTP
data block.

–
•

74/1747

Option bytes to configure read and write protection, BOR level, watchdog
software/hardware and reset when the device is in Standby or Stop mode.

Low-power modes (for details refer to the Power control (PWR) section of the reference
manual)

RM0090 Rev 17

RM0090

Embedded Flash memory interface
Table 5. Flash module organization (STM32F40x and STM32F41x)
Block

Name

Block base addresses

Size

Sector 0

0x0800 0000 - 0x0800 3FFF

16 Kbytes

Sector 1

0x0800 4000 - 0x0800 7FFF

16 Kbytes

Sector 2

0x0800 8000 - 0x0800 BFFF

16 Kbytes

Sector 3

0x0800 C000 - 0x0800 FFFF

16 Kbytes

Sector 4

0x0801 0000 - 0x0801 FFFF

64 Kbytes

Sector 5

0x0802 0000 - 0x0803 FFFF

128 Kbytes

Sector 6

0x0804 0000 - 0x0805 FFFF

128 Kbytes

.
.
.

.
.
.

.
.
.

Sector 11

0x080E 0000 - 0x080F FFFF

128 Kbytes

System memory

0x1FFF 0000 - 0x1FFF 77FF

30 Kbytes

OTP area

0x1FFF 7800 - 0x1FFF 7A0F

528 bytes

Option bytes

0x1FFF C000 - 0x1FFF C00F

16 bytes

Main memory

RM0090 Rev 17

75/1747
112

Embedded Flash memory interface

3.4

RM0090

Embedded Flash memory in STM32F42xxx and
STM32F43xxx
The Flash memory has the following main features:
•

Capacity up to 2 Mbyte with dual bank architecture supporting read-while-write
capability (RWW)

•

128 bits wide data read

•

Byte, half-word, word and double word write

•

Sector, bank, and mass erase (both banks)

•

Dual bank memory organization
The Flash memory is organized as follows:
–

For each bank, a main memory block (1 Mbyte) divided into 4 sectors of 16
Kbytes, 1 sector of 64 Kbytes, and 7 sectors of 128 Kbytes

–

System memory from which the device boots in System memory boot mode

–

512 OTP (one-time programmable) bytes for user data
The OTP area contains 16 additional bytes used to lock the corresponding OTP
data block.

–

•

Option bytes to configure read and write protection, BOR level, watchdog, dual
bank boot mode, dual bank feature, software/hardware and reset when the device
is in Standby or Stop mode.

Dual bank organization on 1 Mbyte devices
The dual bank feature on 1 Mbyte devices is enabled by setting the DB1M option bit.
To obtain a dual bank Flash memory, the last 512 Kbytes of the single bank (sectors
[8:11]) are re-structured in the same way as the first 512 Kbytes.
The sector numbering of dual bank memory organization is different from the single
bank: the single bank memory contains 12 sectors whereas the dual bank memory
contains 16 sectors (see Table 7: 1 Mbyte Flash memory single bank vs dual bank
organization (STM32F42xxx and STM32F43xxx)).
For erase operation, the right sector numbering must be considered according the
DB1M option bit.
–

When the DB1M bit is reset, the erase operation must be performed on the default
sector number.

–

When the DB1M bit is set, to perform an erase operation on bank 2, the sector
number must be programmed (sector number from 12 to 19). Refer to FLASH_CR
register for SNB (Sector number) configuration.

Refer to Table 8: 1 Mbyte single bank Flash memory organization (STM32F42xxx and
STM32F43xxx) and Table 9: 1 Mbyte dual bank Flash memory organization (STM32F42xxx
and STM32F43xxx) for details on 1 Mbyte single bank and 1 Mbyte dual bank
organizations.

76/1747

RM0090 Rev 17

RM0090

Embedded Flash memory interface

Table 6. Flash module - 2 Mbyte dual bank organization (STM32F42xxx and STM32F43xxx)
Block

Bank

Name

Block base addresses

Size

Sector 0

0x0800 0000 - 0x0800 3FFF

16 Kbytes

Sector 1

0x0800 4000 - 0x0800 7FFF

16 Kbytes

Sector 2

0x0800 8000 - 0x0800 BFFF

16 Kbytes

Sector 3

0x0800 C000 - 0x0800 FFFF

16 Kbyte

Sector 4

0x0801 0000 - 0x0801 FFFF

64 Kbytes

Sector 5

0x0802 0000 - 0x0803 FFFF

128 Kbytes

Sector 6

0x0804 0000 - 0x0805 FFFF

128 Kbytes

-

-

-

-

-

-

-

-

-

Sector 11

0x080E 0000 - 0x080F FFFF

128 Kbytes

Sector 12

0x0810 0000 - 0x0810 3FFF

16 Kbytes

Sector 13

0x0810 4000 - 0x0810 7FFF

16 Kbytes

Sector 14

0x0810 8000 - 0x0810 BFFF

16 Kbytes

Sector 15

0x0810 C000 - 0x0810 FFFF

16 Kbytes

Sector 16

0x0811 0000 - 0x0811 FFFF

64 Kbytes

Sector 17

0x0812 0000 - 0x0813 FFFF

128 Kbytes

Sector 18

0x0814 0000 - 0x0815 FFFF

128 Kbytes

-

-

-

-

-

-

0x081E 0000 - 0x081F FFFF

128 Kbytes

System memory

0x1FFF 0000 - 0x1FFF 77FF

30 Kbytes

OTP

0x1FFF 7800 - 0x1FFF 7A0F

528 bytes

Bank 1

0x1FFF C000 - 0x1FFF C00F

16 bytes

Bank 2

0x1FFE C000 - 0x1FFE C00F

16 bytes

Bank 1

Main memory

Bank 2

Sector 23

Option bytes

RM0090 Rev 17

77/1747
112

Embedded Flash memory interface

RM0090

Table 7. 1 Mbyte Flash memory single bank vs dual bank organization
(STM32F42xxx and STM32F43xxx)
1 Mbyte single bank Flash memory (default)

1 Mbyte dual bank Flash memory

DB1M=0

DB1M=1

Main memory

1MB

Sector number

Sector size

Sector 0

Main memory

Sector number

Sector size

16 Kbytes

Sector 0

16 Kbytes

Sector 1

16 Kbytes

Sector 1

16 Kbytes

Sector 2

16 Kbytes

Sector 2

16 Kbytes

Sector 3

16 Kbytes

Sector 3

16 Kbytes

Sector 4

64 Kbytes

Sector 4

64 Kbytes

Sector 5

128 Kbytes

Sector 5

128 Kbytes

Sector 6

128 Kbytes

Sector 6

128 Kbytes

Sector 7

128 Kbytes

Sector 7

128 Kbytes

Sector 8

128 Kbytes

Sector 12

16 Kbytes

Sector 9

128 Kbytes

Sector 13

16 Kbytes

Sector 10

128 Kbytes

Sector 14

16 Kbytes

Sector 11

128 Kbytes

Sector 15

16 Kbytes

-

-

Sector 16

64 Kbytes

-

-

Sector 17

128 Kbytes

-

-

Sector 18

128 Kbytes

-

-

Sector 19

128 Kbytes

Bank 1
512KB

Bank 2
512KB

Table 8. 1 Mbyte single bank Flash memory organization
(STM32F42xxx and STM32F43xxx)
Block

Main memory

78/1747

Bank

Single bank

Name

Block base addresses

Size

Sector 0

0x0800 0000 - 0x0800 3FFF

16 Kbytes

Sector 1

0x0800 4000 - 0x0800 7FFF

16 Kbytes

Sector 2

0x0800 8000 - 0x0800 BFFF

16 Kbytes

Sector 3

0x0800 C000 - 0x0800 FFFF

16 Kbytes

Sector 4

0x0801 0000 - 0x0801 FFFF

64 Kbytes

Sector 5

0x0802 0000 - 0x0803 FFFF

128 Kbytes

Sector 6

0x0804 0000 - 0x0805 FFFF

128 Kbytes

Sector 7

0x0806 0000 - 0x0807 FFFF

128 Kbytes

Sector 8

0x0808 0000 - 0x0809 FFFF

128 Kbytes

Sector 9

0x080A 0000 - 0x080B FFFF

128 Kbytes

Sector 10

0x080C 0000 - 0x080D FFFF

128 Kbytes

Sector 11

0x080E 0000 - 0x080F FFFF

128 Kbytes

RM0090 Rev 17

RM0090

Embedded Flash memory interface
Table 8. 1 Mbyte single bank Flash memory organization
(STM32F42xxx and STM32F43xxx) (continued)

Block

Bank

Name

Block base addresses

Size

System memory

0x1FFF 0000 - 0x1FFFF 77FF

30 Kbytes

OTP

0x1FFF 7800 - 0x1FFF 7A0F

528 bytes

0x1FFF C000 - 0x1FFF C00F

16 bytes

0x1FFE C000 - 0x1FFE C00F

16 bytes

Option bytes

Table 9. 1 Mbyte dual bank Flash memory organization (STM32F42xxx and STM32F43xxx)
Block

Name

Block base addresses

Size

Sector 0

0x0800 0000 - 0x0800 3FFF

16 Kbytes

Sector 1

0x0800 4000 - 0x0800 7FFF

16 Kbytes

Sector 2

0x0800 8000 - 0x0800 BFFF

16 Kbytes

Sector 3

0x0800 C000 - 0x0800 FFFF

16 Kbyte

Sector 4

0x0801 0000 - 0x0801 FFFF

64 Kbytes

Sector 5

0x0802 0000 - 0x0803 FFFF

128 Kbytes

Sector 6

0x0804 0000 - 0x0805 FFFF

128 Kbytes

Sector 7

0x0806 0000 - 0x0807 FFFF

128 Kbytes

Sector 12

0x0808 0000 - 0x0808 3FFF

16 Kbytes

Sector 13

0x0808 4000 - 0x0808 7FFF

16 Kbytes

Sector 14

0x0808 0000 - 0x0808 BFFF

16 Kbytes

Sector 15

0x0808 C000 - 0x0808 FFFF

16 Kbytes

Sector 16

0x0809 0000 - 0x0809 FFFF

64 Kbytes

Sector 17

0x080A 0000 - 0x080B FFFF

128 Kbytes

Sector 18

0x080C 0000 - 0x080D FFFF

128 Kbytes

Sector 19

0x080E 0000 - 0x080F FFFF

128 Kbytes

System memory

0x1FFF 0000 - 0x1FFF 77FF

30 Kbytes

OTP

0x1FFF 7800 - 0x1FFF 7A0F

528 bytes

Bank 1

0x1FFF C000 - 0x1FFF C00F

16 bytes

Bank 2

0x1FFE C000 - 0x1FFE C00F

16 bytes

Bank 1

Main memory

Bank 2

Option bytes

RM0090 Rev 17

79/1747
112

Embedded Flash memory interface

RM0090

3.5

Read interface

3.5.1

Relation between CPU clock frequency and Flash memory read time
To correctly read data from Flash memory, the number of wait states (LATENCY) must be
correctly programmed in the Flash access control register (FLASH_ACR) according to the
frequency of the CPU clock (HCLK) and the supply voltage of the device.
The prefetch buffer must be disabled when the supply voltage is below 2.1 V. The
correspondence between wait states and CPU clock frequency is given in Table 10 and
Table 11.

Note:

On STM32F405xx/07xx and STM32F415xx/17xx devices:
- when VOS = '0', the maximum value of fHCLK = 144 MHz.
- when VOS = '1', the maximum value of fHCLK = 168 MHz.
On STM32F42xxx and STM32F43xxx devices:
- when VOS[1:0] = '0x01', the maximum value of fHCLK is 120 MHz.
- when VOS[1:0] = '0x10', the maximum value of fHCLK is 144 MHz. It can be extended to
168 MHz by activating the over-drive mode.
- when VOS[1:0] = '0x11, the maximum value of fHCLK is 168 MHz. It can be extended to
180 MHz by activating the over-drive mode.
- The over-drive mode is not available when VDD ranges from 1.8 to 2.1 V.
Refer to Section 5.1.4: Voltage regulator for STM32F42xxx and STM32F43xxx for details on
how to activate the over-drive mode.
Table 10. Number of wait states according to CPU clock (HCLK) frequency
(STM32F405xx/07xx and STM32F415xx/17xx)
HCLK (MHz)

Wait states (WS)

Voltage range

Voltage range

Voltage range

Voltage range

2.7 V - 3.6 V

2.4 V - 2.7 V

2.1 V - 2.4 V

0 WS (1 CPU cycle)

0 < HCLK ≤30

0 < HCLK ≤24

0 < HCLK ≤22

0 < HCLK ≤20

1 WS (2 CPU cycles)

30 < HCLK ≤60

24 < HCLK ≤48

22 < HCLK ≤44

20 1, 1->2, 0->2)
there is no mass erase.
•

Level 2: debug/chip read protection disabled
The read protection Level 2 is activated by writing 0xCC to the RDP option byte. When
the read protection Level 2 is set:
–

All protections provided by Level 1 are active.

–

Booting from RAM or system memory bootloader is no more allowed.

–

JTAG, SWV (single-wire viewer), ETM, and boundary scan are disabled.

–

User option bytes can no longer be changed.

–

When booting from Flash memory, accesses (read, erase and program) to Flash
memory and backup SRAM from user code are allowed.

Memory read protection Level 2 is an irreversible operation. When Level 2 is activated,
the level of protection cannot be decreased to Level 0 or Level 1.
Note:

The JTAG port is permanently disabled when Level 2 is active (acting as a JTAG fuse). As a
consequence, boundary scan cannot be performed. STMicroelectronics is not able to
perform analysis on defective parts on which the Level 2 protection has been set.

RM0090 Rev 17

93/1747
112

Embedded Flash memory interface
--

Memory area

RM0090

Table 17. Access versus read protection level
Protection
Level

Debug features, Boot from RAM or
from System memory bootloader
Read

Main Flash Memory
and Backup SRAM
Option Bytes

OTP

Level 1

Write

Booting from Flash memory

Erase

Read

Write

NO(1)

NO

Erase

YES

Level 2

NO

YES

Level 1

YES

YES

Level 2

NO

NO

Level 1

NO

NA

YES

NA

Level 2

NO

NA

YES

NA

1. The main Flash memory and backup SRAM are only erased when the RDP changes from level 1 to 0. The OTP area
remains unchanged.

Figure 6 shows how to go from one RDP level to another.
Figure 6. RDP levels
5'3 $$K  &&K
2WKHUVRSWLRQVPRGLILHG

:ULWHRSWLRQV
LQFOXGLQJ
5'3 &&K

/HYHO
5'3 &&K

/HYHO
5'3 $$K
5'3 &&K
GHIDXOW
:ULWHRSWLRQVLQFOXGLQJ
5'3 &&K  $$K

:ULWHRSWLRQV
LQFOXGLQJ
5'3 &&K

2SWLRQVZULWH 5'3OHYHOLQFUHDVH LQFOXGHV
2SWLRQVHUDVH
1HZRSWLRQVSURJUDP
2SWLRQVZULWH 5'3OHYHOGHFUHDVH LQFOXGHV
0DVVHUDVH
2SWLRQVHUDVH
1HZRSWLRQVSURJUDP

3.7.4

:ULWHRSWLRQV
LQFOXGLQJ
5'3 $$K

/HYHO
5'3 $$K

5'3 $$K
2WKHUVRSWLRQ V PRGLILHG

2SWLRQVZULWH 5'3OHYHOLGHQWLFDO LQFOXGHV
2SWLRQVHUDVH
1HZRSWLRQVSURJUDP
069

Write protections
Up to 24 user sectors in Flash memory can be protected against unwanted write operations
due to loss of program counter contexts. When the non-write protection nWRPi bit (0 ≤i ≤11)
in the FLASH_OPTCR or FLASH_OPTCR1 registers is low, the corresponding sector

94/1747

RM0090 Rev 17

RM0090

Embedded Flash memory interface
cannot be erased or programmed. Consequently, a mass erase cannot be performed if one
of the sectors is write-protected.
If an erase/program operation to a write-protected part of the Flash memory is attempted
(sector protected by write protection bit, OTP part locked or part of the Flash memory that
can never be written like the ICP), the write protection error flag (WRPERR) is set in the
FLASH_SR register.
On STM32F42xxx and STM32F43xxx devices, when the PCROP mode is set, the active
level of nWRPi is high, and the corresponding sector i is write protected when nWRPi is
high. A PCROP sector is automatically write protected.

Note:

When the memory read protection level is selected (RDP level = 1), it is not possible to
program or erase Flash memory sector i if the CPU debug features are connected (JTAG or
single wire) or boot code is being executed from RAM, even if nWRPi = 1.

Write protection error flag
If an erase/program operation to a write protected area of the Flash memory is performed,
the Write Protection Error flag (WRPERR) is set in the FLASH_SR register.
If an erase operation is requested, the WRPERR bit is set when:
•

Mass, bank, sector erase are configured (MER or MER/MER1 and SER = 1)

•

A sector erase is requested and the Sector Number SNB field is not valid

•

A mass erase is requested while at least one of the user sector is write protected by
option bit (MER or MER/MER1 = 1 and nWRPi = 0 with 0 ≤i ≤11 bits in the
FLASH_OPTCRx register

•

A sector erase is requested on a write protected sector. (SER = 1, SNB = i and
nWRPi = 0 with 0 ≤i ≤11 bits in the FLASH_OPTCRx register)

•

The Flash memory is readout protected and an intrusion is detected.

If a program operation is requested, the WRPERR bit is set when:

3.7.5

•

A write operation is performed on system memory or on the reserved part of the user
specific sector.

•

A write operation is performed to the user configuration sector

•

A write operation is performed on a sector write protected by option bit.

•

A write operation is requested on an OTP area which is already locked

•

The Flash memory is read protected and an intrusion is detected.

Proprietary code readout protection (PCROP)
The proprietary readout protection (PCROP) is available only on STM32F42xxx and
STM32F43xxx devices.
Flash memory user sectors (0 to 23) can be protected against D-bus read accesses by
using the proprietary readout protection (PCROP).
The PCROP protection is selected as follows, through the SPRMOD option bit in the
FLASH_OPTCR register:
•

SPRMOD = 0: nWRPi control the write protection of respective user sectors

•

SPRMOD = 1: nWRPi control the read and write protection (PCROP) of respective
user sectors.

RM0090 Rev 17

95/1747
112

Embedded Flash memory interface

RM0090

When a sector is readout protected (PCROP mode activated), it can only be accessed for
code fetch through ICODE Bus on Flash interface:
•

Any read access performed through the D-bus triggers a RDERR flag error.

•

Any program/erase operation on a PCROPed sector triggers a WRPERR flag error.
Figure 7. PCROP levels
7RITE OPTIONS
30-/$  ACTIVE
AND VALID N720I

,EVEL 
2$0  X!!
2$0  X##
DEFAULT

7RITE OPTIONS
30-/$  ACTIVE
AND VALID N720I

7RITE OPTIONS
30-/$  ACTIVE
AND VALID N720I

.O RESTRICTION ON
7RITE OPTIONS

,EVEL 

,EVEL 

2$0  X##

2$0  X!!

7RITE OPTIONS
30-/$  ACTIVE
AND VALID N720I

5SER OPTION SECTOR ERASE
0ROGRAM NEW OPTIONS

7RITE OPTIONS
30-/$  ACTIVE
AND VALID N720I

'LOBAL MASS ERASE
5SER OPTION SECTION ERASE
0ROGRAM NEW OPTIONS
6ALID N720I MEANS THAT NONE OF THE N720 BITS SET CAN BE RESET TRANSITION FROM  TO 

-36

The deactivation of the SPRMOD and/or the unprotection of PCROPed user sectors can
only occur when, at the same time, the RDP level changes from 1 to 0. If this condition is not
respected, the user option byte modification is cancelled and the write error WRPERR flag
is set. The modification of the users option bytes (BOR_LEV, RST_STDBY, ..) is allowed
since none of the active nWRPi bits is reset and SPRMOD is kept active.
Note:

The active value of nWRPi bits is inverted when PCROP mode is active (SPRMOD =1).
If SPRMOD = 1 and nWRPi =1, then user sector i of bank 1, respectively bank 2 is
read/write protected (PCROP).

96/1747

RM0090 Rev 17

RM0090

3.8

Embedded Flash memory interface

One-time programmable bytes
Table 18 shows the organization of the one-time programmable (OTP) part of the OTP area.
Table 18. OTP area organization
Block
0

1

[128:96]

[95:64]

[63:32]

[31:0]

Address byte 0

OTP0

OTP0

OTP0

OTP0

0x1FFF 7800

OTP0

OTP0

OTP0

OTP0

0x1FFF 7810

OTP1

OTP1

OTP1

OTP1

0x1FFF 7820

OTP1

OTP1

OTP1

OTP1

0x1FFF 7830

.
.
.
15
Lock block

.
.
.

.
.
.

OTP15

OTP15

OTP15

OTP15

0x1FFF 79E0

OTP15

OTP15

OTP15

OTP15

0x1FFF 79F0

LOCKB15 ...
LOCKB12

LOCKB11 ...
LOCKB8

LOCKB7 ...
LOCKB4

LOCKB3 ...
LOCKB0

0x1FFF 7A00

The OTP area is divided into 16 OTP data blocks of 32 bytes and one lock OTP block of 16
bytes. The OTP data and lock blocks cannot be erased. The lock block contains 16 bytes
LOCKBi (0 ≤i ≤15) to lock the corresponding OTP data block (blocks 0 to 15). Each OTP
data block can be programmed until the value 0x00 is programmed in the corresponding
OTP lock byte. The lock bytes must only contain 0x00 and 0xFF values, otherwise the OTP
bytes might not be taken into account correctly.

RM0090 Rev 17

97/1747
112

Embedded Flash memory interface

RM0090

3.9

Flash interface registers

3.9.1

Flash access control register (FLASH_ACR)
for STM32F405xx/07xx and STM32F415xx/17xx
The Flash access control register is used to enable/disable the acceleration features and
control the Flash memory access time according to CPU frequency.
Address offset: 0x00
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access

31

30

29

28

27

26

25

24

23

22

21

20

19

18

6

5

4

3

2

17

16

1

0

Reserved
15

14

13

Reserved

12

11

DCRST ICRST
rw

w

10

9

8

DCEN

ICEN

PRFTEN

rw

rw

rw

7

Reserved

LATENCY[2:0]
rw

rw

rw

Bits 31:13 Reserved, must be kept cleared.
Bit 12 DCRST: Data cache reset
0: Data cache is not reset
1: Data cache is reset
This bit can be written only when the D cache is disabled.
Bit 11 ICRST: Instruction cache reset
0: Instruction cache is not reset
1: Instruction cache is reset
This bit can be written only when the I cache is disabled.
Bit 10 DCEN: Data cache enable
0: Data cache is disabled
1: Data cache is enabled
Bit 9 ICEN: Instruction cache enable
0: Instruction cache is disabled
1: Instruction cache is enabled
Bit 8 PRFTEN: Prefetch enable
0: Prefetch is disabled
1: Prefetch is enabled
Bits 7:3 Reserved, must be kept cleared.
Bits 2:0 LATENCY[2:0]: Latency
These bits represent the ratio of the CPU clock period to the Flash memory access time.
000: Zero wait state
001: One wait state
010: Two wait states
011: Three wait states
100: Four wait states
101: Five wait states
110: Six wait states
111: Seven wait states

98/1747

RM0090 Rev 17

RM0090

Embedded Flash memory interface

3.9.2

Flash access control register (FLASH_ACR)
for STM32F42xxx and STM32F43xxx
The Flash access control register is used to enable/disable the acceleration features and
control the Flash memory access time according to CPU frequency.
Address offset: 0x00
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access

31

30

29

28

27

26

25

24

23

22

21

20

19

6

5

4

3

18

17

16

2

1

0

Reserved
15

14

13

Reserved

12

11

DCRST ICRST
rw

w

10

9

8

DCEN

ICEN

PRFTEN

rw

rw

rw

7

Reserved

LATENCY[3:0]
rw

rw

rw

rw

Bits 31:11 Reserved, must be kept cleared.
Bit 12 DCRST: Data cache reset
0: Data cache is not reset
1: Data cache is reset
This bit can be written only when the D cache is disabled.
Bit 11 ICRST: Instruction cache reset
0: Instruction cache is not reset
1: Instruction cache is reset
This bit can be written only when the I cache is disabled.
Bit 10 DCEN: Data cache enable
0: Data cache is disabled
1: Data cache is enabled
Bit 9 ICEN: Instruction cache enable
0: Instruction cache is disabled
1: Instruction cache is enabled
Bit 8 PRFTEN: Prefetch enable
0: Prefetch is disabled
1: Prefetch is enabled
Bits 7:4 Reserved, must be kept cleared.
Bits 3:0 LATENCY[3:0]: Latency
These bits represent the ratio of the CPU clock period to the Flash memory access time.
0000: Zero wait state
0001: One wait state
0010: Two wait states
...
1110: Fourteen wait states
1111: Fifteen wait states

RM0090 Rev 17

99/1747
112

Embedded Flash memory interface

3.9.3

RM0090

Flash key register (FLASH_KEYR)
The Flash key register is used to allow access to the Flash control register and so, to allow
program and erase operations.
Address offset: 0x04
Reset value: 0x0000 0000
Access: no wait state, word access

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

KEY[31:16]
w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

w

w

w

w

w

w

w

KEY[15:0]
w

w

w

w

w

w

w

w

w

Bits 31:0 FKEYR[31:0]: FPEC key
The following values must be programmed consecutively to unlock the FLASH_CR register
and allow programming/erasing it:
a) KEY1 = 0x45670123
b) KEY2 = 0xCDEF89AB

3.9.4

Flash option key register (FLASH_OPTKEYR)
The Flash option key register is used to allow program and erase operations in the user
configuration sector.
Address offset: 0x08
Reset value: 0x0000 0000
Access: no wait state, word access

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

OPTKEYR[31:16
w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

w

w

w

w

w

w

w

OPTKEYR[15:0]
w

w

w

w

w

w

w

w

w

Bits 31:0 OPTKEYR[31:0]: Option byte key
The following values must be programmed consecutively to unlock the FLASH_OPTCR
register and allow programming it:
a) OPTKEY1 = 0x08192A3B
b) OPTKEY2 = 0x4C5D6E7F

100/1747

RM0090 Rev 17

RM0090

Embedded Flash memory interface

3.9.5

Flash status register (FLASH_SR) for
STM32F405xx/07xx and STM32F415xx/17xx
The Flash status register gives information on ongoing program and erase operations.
Address offset: 0x0C
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

Reserved
15

14

13

12

11

Reserved

10

9

8

7

r
6

5

4

PGSERR PGPERR PGAERR WRPERR
rc_w1

16
BSY

rc_w1

rc_w1

rc_w1

3

2

Reserved

1

0

OPERR

EOP

rc_w1

rc_w1

Bits 31:17 Reserved, must be kept cleared.
Bit 16 BSY: Busy
This bit indicates that a Flash memory operation is in progress. It is set at the beginning of a
Flash memory operation and cleared when the operation finishes or an error occurs.
0: no Flash memory operation ongoing
1: Flash memory operation ongoing
Bits 15:8 Reserved, must be kept cleared.
Bit 7 PGSERR: Programming sequence error
Set by hardware when a write access to the Flash memory is performed by the code while the
control register has not been correctly configured.
Cleared by writing 1.
Bit 6 PGPERR: Programming parallelism error
Set by hardware when the size of the access (byte, half-word, word, double word) during the
program sequence does not correspond to the parallelism configuration PSIZE (x8, x16, x32,
x64).
Cleared by writing 1.
Bit 5 PGAERR: Programming alignment error
Set by hardware when the data to program cannot be contained in the same 128-bit Flash
memory row.
Cleared by writing 1.
Bit 4 WRPERR: Write protection error
Set by hardware when an address to be erased/programmed belongs to a write-protected part
of the Flash memory.
Cleared by writing 1.

RM0090 Rev 17

101/1747
112

Embedded Flash memory interface

RM0090

Bits 3:2 Reserved, must be kept cleared.
Bit 1 OPERR: Operation error
Set by hardware when a flash operation (programming / erase /read) request is detected and
can not be run because of parallelism, alignment, or write protection error. This bit is set only if
error interrupts are enabled (ERRIE = 1).
Bit 0 EOP: End of operation
Set by hardware when one or more Flash memory operations (program/erase) has/have
completed successfully. It is set only if the end of operation interrupts are enabled (EOPIE = 1).
Cleared by writing a 1.

3.9.6

Flash status register (FLASH_SR) for
STM32F42xxx and STM32F43xxx
The Flash status register gives information on ongoing program and erase operations.
Address offset: 0x0C
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

Reserved
15

14

13

12
Reserved

11

10

9

8

7

r
6

5

4

RDERR PGSERR PGPERR PGAERR WRPERR
rc_w1

rc_w1

16
BSY

rc_w1

rc_w1

rc_w1

3

2

Reserved

1

0

OPERR

EOP

rc_w1

rc_w1

Bits 31:17 Reserved, must be kept cleared.
Bit 16 BSY: Busy
This bit indicates that a Flash memory operation is in progress to/from one bank. It is set at the
beginning of a Flash memory operation and cleared when the operation finishes or an error
occurs.
0: no Flash memory operation ongoing
1: Flash memory operation ongoing
Bits 15:9 Reserved, must be kept cleared.
Bit 8 RDERR: Proprietary readout protection (PCROP) error
Set by hardware when a read access through the D-bus is performed to an address belonging
to a proprietary readout protected Flash sector.
Cleared by writing 1.
Bit 7 PGSERR: Programming sequence error
Set by hardware when a write access to the Flash memory is performed by the code while the
control register has not been correctly configured.
Cleared by writing 1.

102/1747

RM0090 Rev 17

RM0090

Embedded Flash memory interface

Bit 6 PGPERR: Programming parallelism error
Set by hardware when the size of the access (byte, half-word, word, double word) during the
program sequence does not correspond to the parallelism configuration PSIZE (x8, x16, x32,
x64).
Cleared by writing 1.
Bit 5 PGAERR: Programming alignment error
Set by hardware when the data to program cannot be contained in the same 128-bit Flash
memory row.
Cleared by writing 1.
Bit 4 WRPERR: Write protection error
Set by hardware when an address to be erased/programmed belongs to a write-protected part
of the Flash memory.
Cleared by writing 1.
Bits 3:2 Reserved, must be kept cleared.
Bit 1 OPERR: Operation error
Set by hardware when a flash operation (programming/erase/read) request is detected and
can not be run because of parallelism, alignment, write or read (PCROP) protection error. This
bit is set only if error interrupts are enabled (ERRIE = 1).
Bit 0 EOP: End of operation
Set by hardware when one or more Flash memory operations (program/erase) has/have
completed successfully. It is set only if the end of operation interrupts are enabled (EOPIE = 1).
Cleared by writing a 1.

3.9.7

Flash control register (FLASH_CR) for
STM32F405xx/07xx and STM32F415xx/17xx
The Flash control register is used to configure and start Flash memory operations.
Address offset: 0x10
Reset value: 0x8000 0000
Access: no wait state when no Flash memory operation is ongoing, word, half-word and
byte access.

31

30

29

LOCK

27

26

Reserved

rs
15

28

14

13

12

Reserved

11

10

25

24

ERRIE

EOPIE

rw

rw

9

8

PSIZE[1:0]
rw

rw

23

22

21

20

19

18

17

16
STRT

Reserved
7
Res.

RM0090 Rev 17

6

5

4

rs
3

SNB[3:0]
rw

rw

rw

rw

2

1

0

MER

SER

PG

rw

rw

rw

103/1747
112

Embedded Flash memory interface

RM0090

Bit 31 LOCK: Lock
Write to 1 only. When it is set, this bit indicates that the FLASH_CR register is locked. It is
cleared by hardware after detecting the unlock sequence.
In the event of an unsuccessful unlock operation, this bit remains set until the next reset.
Bits 30:26 Reserved, must be kept cleared.
Bit 25 ERRIE: Error interrupt enable
This bit enables the interrupt generation when the OPERR bit in the FLASH_SR register is set
to 1.
0: Error interrupt generation disabled
1: Error interrupt generation enabled
Bit 24 EOPIE: End of operation interrupt enable
This bit enables the interrupt generation when the EOP bit in the FLASH_SR register goes to
1.
0: Interrupt generation disabled
1: Interrupt generation enabled
Bits 23:17 Reserved, must be kept cleared.
Bit 16 STRT: Start
This bit triggers an erase operation when set. It is set only by software and cleared when the
BSY bit is cleared.
Bits 15:10 Reserved, must be kept cleared.
Bits 9:8 PSIZE[1:0]: Program size
These bits select the program parallelism.
00 program x8
01 program x16
10 program x32
11 program x64
Bit 7 Reserved, must be kept cleared.
Bits 6:3 SNB[3:0]: Sector number
These bits select the sector to erase.
0000 sector 0
0001 sector 1
...
1011 sector 11
Others not allowed
Bit 2 MER: Mass Erase
Erase activated for all user sectors.
Bit 1 SER: Sector Erase
Sector Erase activated.
Bit 0 PG: Programming
Flash programming activated.

104/1747

RM0090 Rev 17

RM0090

Embedded Flash memory interface

3.9.8

Flash control register (FLASH_CR) for
STM32F42xxx and STM32F43xxx
The Flash control register is used to configure and start Flash memory operations.
Address offset: 0x10
Reset value: 0x8000 0000
Access: no wait state when no Flash memory operation is ongoing, word, half-word and
byte access.

31

30

29

LOCK

MER1
rw

27

26

Reserved

rs
15

28

14

13

12
Reserved

11

10

25

24

ERRIE

EOPIE

rw

rw

9

8

23

22

rw

20

19

18

17

16
STRT

Reserved
7

6

PSIZE[1:0]
rw

21

5

4

rs
3

SNB[4:0]
rw

rw

rw

rw

rw

2

1

0

MER

SER

PG

rw

rw

rw

Bit 31 LOCK: Lock
Write to 1 only. When it is set, this bit indicates that the FLASH_CR register is locked. It is
cleared by hardware after detecting the unlock sequence.
In the event of an unsuccessful unlock operation, this bit remains set until the next reset.
Bits 30:26 Reserved, must be kept cleared.
Bit 25 ERRIE: Error interrupt enable
This bit enables the interrupt generation when the OPERR bit in the FLASH_SR register is set
to 1.
0: Error interrupt generation disabled
1: Error interrupt generation enabled
Bit 24 EOPIE: End of operation interrupt enable
This bit enables the interrupt generation when the EOP bit in the FLASH_SR register goes to
1.
0: Interrupt generation disabled
1: Interrupt generation enabled
Bits 23:17 Reserved, must be kept cleared.
Bit 16 STRT: Start
This bit triggers an erase operation when set. It is set only by software and cleared when the
BSY bit is cleared.
Bit 15 MER1: Mass Erase of bank 2 sectors
Erase activated for bank 2 user sectors 12 to 23.
Bits 14:10 Reserved, must be kept cleared.
Bits 9:8 PSIZE[1:0]: Program size
These bits select the program parallelism.
00 program x8
01 program x16
10 program x32
11 program x64

RM0090 Rev 17

105/1747
112

Embedded Flash memory interface

RM0090

Bits 7:3 SNB[3:0]: Sector number
These bits select the sector to erase.
0000: sector 0
0001: sector 1
...
01011: sector 11
01100: not allowed
01101: not allowed
01110: not allowed
01111: not allowed
10000: section 12
10001: section 13
...
11011 sector 23
11100: not allowed
11101: not allowed
11110: not allowed
11111: not allowed
Bit 2 MER: Mass Erase of bank 1 sectors
Erase activated of bank 1 sectors.
Bit 1 SER: Sector Erase
Sector Erase activated.
Bit 0 PG: Programming
Flash programming activated.

3.9.9

Flash option control register (FLASH_OPTCR) for
STM32F405xx/07xx and STM32F415xx/17xx
The FLASH_OPTCR register is used to modify the user option bytes.
Address offset: 0x14
Reset value: 0x0FFF AAED. The option bits are loaded with values from Flash memory at
reset release.
Access: no wait state when no Flash memory operation is ongoing, word, half-word and
byte access.

31

30

29

28

14

26

25

24

13

12

rw

rw

rw

rw

11

10

9

8

RDP[7:0]
rw

106/1747

rw

23

22

21

20

19

18

17

16

rw

rw

rw

rw

rw

rw

5

4

3

2

nWRP[11:0]

Reserved
15

27

rw

rw

rw

rw

rw

rw

rw

rw

7

6

nRST_
STDBY

nRST_
STOP

rw

rw

RM0090 Rev 17

WDG_
Reserve
SW
d
rw

BOR_LEV
rw

rw

1

0

OPTST
RT

OPTLO
CK

rs

rs

RM0090

Embedded Flash memory interface

Bits 31:28 Reserved, must be kept cleared.
Bits 27:16 nWRP[11:0]: Not write protect
These bits contain the value of the write-protection option bytes after reset. They can be
written to program a new write protect value into Flash memory.
0: Write protection active on selected sector
1: Write protection inactive on selected sector
Bits 15:8 RDP[7:0]: Read protect
These bits contain the value of the read-protection option level after reset. They can be written
to program a new read protection value into Flash memory.
0xAA: Level 0, read protection not active
0xCC: Level 2, chip read protection active
Others: Level 1, read protection of memories active
Bits 7:5 USER[2:0]: User option bytes
These bits contain the value of the user option byte after reset. They can be written to program
a new user option byte value into Flash memory.
Bit 7: nRST_STDBY
Bit 6: nRST_STOP
Bit 5: WDG_SW
Note: When changing the WDG mode from hardware to software or from software to
hardware, a system reset is required to make the change effective.
Bit 4 Reserved, must be kept cleared. Always read as “0”.
Bits 3:2 BOR_LEV[1:0]: BOR reset Level
These bits contain the supply level threshold that activates/releases the reset. They can be
written to program a new BOR level. By default, BOR is off. When the supply voltage (VDD)
drops below the selected BOR level, a device reset is generated.
00: BOR Level 3 (VBOR3), brownout threshold level 3
01: BOR Level 2 (VBOR2), brownout threshold level 2
10: BOR Level 1 (VBOR1), brownout threshold level 1
11: BOR off, POR/PDR reset threshold level is applied
Note: For full details about BOR characteristics, refer to the “Electrical characteristics” section
in the device datasheet.
Bit 1 OPTSTRT: Option start
This bit triggers a user option operation when set. It is set only by software and cleared when
the BSY bit is cleared.
Bit 0 OPTLOCK: Option lock
Write to 1 only. When this bit is set, it indicates that the FLASH_OPTCR register is locked. This
bit is cleared by hardware after detecting the unlock sequence.
In the event of an unsuccessful unlock operation, this bit remains set until the next reset.

RM0090 Rev 17

107/1747
112

Embedded Flash memory interface

3.9.10

RM0090

Flash option control register (FLASH_OPTCR)
for STM32F42xxx and STM32F43xxx
The FLASH_OPTCR register is used to modify the user option bytes.
Address offset: 0x14
Reset value: 0x0FFF AAED. The option bits are loaded with values from Flash memory at
reset release.
Access: no wait state when no Flash memory operation is ongoing, word, half-word and
byte access.

31

30

SPR
MOD

DB1M

rw

rw

15

14

29

28

27

26

25

24

13

rw

rw

22

21

20

19

18

17

16

nWRP[11:0]

Reserved
12

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

11

10

9

8

7

6

5

4

3

2

1

0

nRST_ nRST_
STDBY STOP

RDP[7:0]
rw

23

rw

rw

rw

rw

rw

rw

rw

WDG_
SW

BFB2

rw

rw

BOR_LEV
rw

rw

OPTST OPTLO
RT
CK
rs

rs

Bit 31 SPRMOD: Selection of protection mode for nWPRi bits
0: PCROP disabled. nWPRi bits used for Write protection on sector i.
1: PCROP enabled. nWPRi bits used for PCROP protection on sector i
Bit 30 DB1M: Dual-bank on 1 Mbyte Flash memory devices
0: 1 Mbyte single bank Flash memory (contiguous addresses in bank1)
1: 1 Mbyte dual bank Flash memory. The Flash memory is organized as two banks of
512 Kbytes each (see Table 7: 1 Mbyte Flash memory single bank vs dual bank organization
(STM32F42xxx and STM32F43xxx) and Table 9: 1 Mbyte dual bank Flash memory
organization (STM32F42xxx and STM32F43xxx)). To perform an erase operation, the right
sector must be programmed (see Table 7 for information on the sector numbering scheme).
Note: If DB1M is set and an erase operation is performed on Bank 2 while the default sector
number is selected (as an example, sector 8 is configured instead of sector 12), the
erase operation on Bank 2 sector is not performed.
Bits 29:28 Reserved, must be kept cleared.
Bits 27:16 nWRP[11:0]: Not write protect
These bits contain the value of the write-protection and read-protection (PCROP) option bytes
for sectors 0 to 11 after reset. They can be written to program a new write-protect or PCROP
value into Flash memory.
If SPRMOD is reset:
0: Write protection active on sector i
1: Write protection not active on sector i
If SPRMOD is set:
0: PCROP protection not active on sector i
1: PCROP protection active on sector i
Bits 15:8 RDP[7:0]: Read protect
These bits contain the value of the read-protection option level after reset. They can be written
to program a new read protection value into Flash memory.
0xAA: Level 0, read protection not active
0xCC: Level 2, chip read protection active
Others: Level 1, read protection of memories active

108/1747

RM0090 Rev 17

RM0090

Embedded Flash memory interface

Bits 7:5 USER: User option bytes
These bits contain the value of the user option byte after reset. They can be written to program
a new user option byte value into Flash memory.
Bit 7: nRST_STDBY
Bit 6: nRST_STOP
Bit 5: WDG_SW
Note: When changing the WDG mode from hardware to software or from software to
hardware, a system reset is required to make the change effective.
Bit 4 BFB2: Dual-bank Boot option byte
0: Dual-bank boot disabled. Boot can be performed either from Flash memory bank 1 or from
system memory depending on boot pin state (default)
1: Dual-bank boot enabled. Boot is always performed from system memory.
Note: For STM32F42xx and STM32F43xx 1MB part numbers, this option bit must be kept
cleared when DB1M=0.
Bits 3:2 BOR_LEV: BOR reset Level
These bits contain the supply level threshold that activates/releases the reset. They can be
written to program a new BOR level. By default, BOR is off. When the supply voltage (VDD)
drops below the selected BOR level, a device reset is generated.
00: BOR Level 3 (VBOR3), brownout threshold level 3
01: BOR Level 2 (VBOR2), brownout threshold level 2
10: BOR Level 1 (VBOR1), brownout threshold level 1
11: BOR off, POR/PDR reset threshold level is applied
Note: For full details on BOR characteristics, refer to the “Electrical characteristics” section of
the product datasheet.
Bit 1 OPTSTRT: Option start
This bit triggers a user option operation when set. It is set only by software and cleared when
the BSY bit is cleared.
Bit 0 OPTLOCK: Option lock
Write to 1 only. When this bit is set, it indicates that the FLASH_OPTCR register is locked. This
bit is cleared by hardware after detecting the unlock sequence.
In the event of an unsuccessful unlock operation, this bit remains set until the next reset.

RM0090 Rev 17

109/1747
112

Embedded Flash memory interface

3.9.11

RM0090

Flash option control register (FLASH_OPTCR1)
for STM32F42xxx and STM32F43xxx
This register is available only on STM32F42xxx and STM32F43xxx.
The FLASH_OPTCR1 register is used to modify the user option bytes for bank 2.
Address offset: 0x18
Reset value: 0x0FFF 0000. The option bits are loaded with values from Flash memory at
reset release.
Access: no wait state when no Flash memory operation is ongoing, word, half-word and
byte access.

31

30

29

28

14

13

26

25

24

23

22

21

20

19

18

17

16

nWRP[11:0]

Reserved
15

27

12

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

11

10

9

8

7

6

5

4

3

2

1

0

Reserved

Bits 31:28 Reserved, must be kept cleared.
Bits 27:16 nWRP[11:0]: Not write protect
These bits contain the value of the write-protection and read-protection (PCROP) option bytes
for sectors 0 to 11 after reset. They can be written to program a new write-protect or PCROP
value into Flash memory.
If SPRMOD is reset (default value):
0: Write protection active on sector i.
1: Write protection not active on sector i.
If SPRMOD is set:
0: PCROP protection not active on sector i.
1: PCROP protection active on sector i.
Bits 15:0 Reserved, must be kept cleared.

110/1747

RM0090 Rev 17

RM0090

3.9.12

Embedded Flash memory interface

Flash interface register map

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Reserved

Reserved

0

Reserved

0

nWRP[11:0]

Reserved

Reset value

1

1

1

1

1

1

0

0

0

0

0

0

0

0

1

1

0

0

1

0

RDP[7:0]

1

1

1

1

1

0

1

0

1

0

0

0

0

0

0

0

0

0

Reserved

1

STRT

Reset value

EOPIE

LOCK

0

Reserved

0

OPTKEYR[15:0]

Reserved

FLASH_CR

0

SNB[3:0]

0

0

1

1

1

0

0

1

1

EOP

0

0

OPERR

0

0

0

0

PG

0

0

SER

0

0

0

0

OPTSTRT

0

FLASH_SR

FLASH_
OPTCR

0

OPTLOCK

0

OPTKEYR[31:16]
0

0

Reserved

0

0

MER

0

0

BOR_LEV[1:0]

0

PGAERR

0

WRPERR

0

Reserved

0

PGPERR

0

PGSERR

0

WDG_SW

0x14

0
KEY[15:0]

Reset value

0x10

0

nRST_STOP

0x0C

0

FLASH_OPT
KEYR
Reset value

0

LATENCY
[2:0]

Reserved

nRST_STDBY

0x08

0

PSIZE[1:0]

Reset value

0
KEY[31:16]

BSY

0x04

FLASH_
KEYR

ICEN

Reset value

PRFTEN

Reserved

DCEN

FLASH_ACR

ICRST

0x00

Register

DCRST

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 19. Flash register map and reset values
(STM32F405xx/07xx and STM32F415xx/17xx)

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

KEY[15:0]
0

0

0

0

0

0

0

0

0

OPTKEYR[31:16]
0

0

EOP

0

0

0

0

0

0

0

OPTKEYR[15:0]
0

0

0

0

0

Reserved

Reset value

0

RM0090 Rev 17

0

0

0

0

0

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

Reserved

0

0

WRPERR

0

0

OPERR

0

PGAERR

FLASH_SR

0

PGPERR

0x0C

0

KEY[31:16]
0

FLASH_
OPTKEYR
Reset value

0

PGSERR

Reset value

0

LATENCY[3:0]

Reserved

RDERR

0x08

FLASH_KEYR

BSY

0x04

PRFTEN

Reset value

ICEN

Reserved

DCEN

FLASH_ACR

ICRST

0x00

Register

DCRST

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 20. Flash register map and reset values (STM32F42xxx and STM32F43xxx)

111/1747
112

Embedded Flash memory interface

RM0090

0x18

Reset value

112/1747

1

1

1

1

1

1

1

1

1

1

nWRP[11:0]
1

1

1

1

1

1

1

1

1

0

1

0

1

0

0

1

1

1

0

Reserved

RM0090 Rev 17

0

0

0

0

1

0
BOR_LEV[1:0]

1

PG

1

0

SER

1

0

OPTLOCK

1

0

MER

1

RDP[7:0]

0

OPTSTRT

1

0

BFB2

Reserved
1

0

WDG_SW

0

nWRP[11:0]

SNB[4:0]

nRST_STOP

0

FLASH_
OPTCR1

0

Reserved

nRST_STDBY

Reset value

0

PSIZE[1:0]

FLASH_OPTCR

0

Reserved

0x14

Reserved

STRT

1

Reserved

MER1

LOCK

Reset value

0x10

EOPIE

FLASH_CR

DB1M

Register

SPRMOD

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 20. Flash register map and reset values (STM32F42xxx and STM32F43xxx) (continued)

1

1

RM0090

4

CRC calculation unit

CRC calculation unit
This section applies to the whole STM32F4xx family, unless otherwise specified.

4.1

CRC introduction
The CRC (cyclic redundancy check) calculation unit is used to get a CRC code from a 32-bit
data word and a fixed generator polynomial.
Among other applications, CRC-based techniques are used to verify data transmission or
storage integrity. In the scope of the EN/IEC 60335-1 standard, they offer a means of
verifying the Flash memory integrity. The CRC calculation unit helps compute a signature of
the software during runtime, to be compared with a reference signature generated at linktime and stored at a given memory location.

4.2

CRC main features
•

Uses CRC-32 (Ethernet) polynomial: 0x4C11DB7
–

•

X32 + X26 + X23 + X22 + X16 + X12 + X11 + X10 +X8 + X7 + X5 + X4 + X2+ X + 1

Single input/output 32-bit data register

•

CRC computation done in 4 AHB clock cycles (HCLK)

•

General-purpose 8-bit register (can be used for temporary storage)

The block diagram is shown in Figure 8.
Figure 8. CRC calculation unit block diagram
$+%EXV
ELW UHDGDFFHVV
'DWDUHJLVWHU RXWSXW

&5&FRPSXWDWLRQ SRO\QRPLDO[&'%
ELW ZULWHDFFHVV
'DWDUHJLVWHU LQSXW
DL

RM0090 Rev 17

113/1747
115

CRC calculation unit

4.3

RM0090

CRC functional description
The CRC calculation unit mainly consists of a single 32-bit data register, which:
•

is used as an input register to enter new data in the CRC calculator (when writing into
the register)

•

holds the result of the previous CRC calculation (when reading the register)

Each write operation into the data register creates a combination of the previous CRC value
and the new one (CRC computation is done on the whole 32-bit data word, and not byte per
byte).
The write operation is stalled until the end of the CRC computation, thus allowing back-toback write accesses or consecutive write and read accesses.
The CRC calculator can be reset to 0xFFFF FFFF with the RESET control bit in the
CRC_CR register. This operation does not affect the contents of the CRC_IDR register.

4.4

CRC registers
The CRC calculation unit contains two data registers and a control register.The peripheral
The CRC registers have to be accessed by words (32 bits).

4.4.1

Data register (CRC_DR)
Address offset: 0x00
Reset value: 0xFFFF FFFF

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

DR [31:16]
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

DR [15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 Data register bits
Used as an input register when writing new data into the CRC calculator.
Holds the previous CRC calculation result when it is read.

4.4.2

Independent data register (CRC_IDR)
Address offset: 0x04
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

Reserved
15

14

13

12

11

Reserved

114/1747

10

9

8

IDR[7:0]

RM0090 Rev 17

rw

RM0090

CRC calculation unit

Bits 31:8 Reserved, must be kept at reset value.
Bits 7:0 General-purpose 8-bit data register bits
Can be used as a temporary storage location for one byte.
This register is not affected by CRC resets generated by the RESET bit in the CRC_CR
register.

4.4.3

Control register (CRC_CR)
Address offset: 0x08
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

6

5

4

3

2

1

16

Reserved
15

14

13

12

11

10

9

8

7

0
RESET

Reserved

w

Bits 31:1 Reserved, must be kept at reset value.
Bit 0 RESET bit
Resets the CRC calculation unit and sets the data register to 0xFFFF FFFF.
This bit can only be set, it is automatically cleared by hardware.

4.4.4

CRC register map
The following table provides the CRC register map and reset values.
Table 21. CRC calculation unit register map and reset values
Offset

0x00

Register

31-24

23-16

15-8

7

6

Data register

Reset
value

0xFFFF FFFF

Reset
value

Reset
value

3

2

1

0

Independent data register
Reserved

0x00

CRC_CR
0x08

4

CRC_DR

CRC_IDR
0x04

5

RESET
Reserved

RM0090 Rev 17

0

115/1747
115

Power controller (PWR)

5

RM0090

Power controller (PWR)
This section applies to the whole STM32F4xx family, unless otherwise specified.

5.1

Power supplies
The device requires a 1.8 to 3.6 V operating voltage supply (VDD). An embedded linear
voltage regulator is used to supply the internal 1.2 V digital power.
The real-time clock (RTC), the RTC backup registers, and the backup SRAM (BKP SRAM)
can be powered from the VBAT voltage when the main VDD supply is powered off.

Note:

Depending on the operating power supply range, some peripheral may be used with limited
functionality and performance. For more details refer to section “General operating
conditions” in STM32F4xx datasheets.
Figure 9. Power supply overview for STM32F405xx/07xx and STM32F415xx/17xx
9%$7
9%$7
WR9

*3,2V
,1
9&$3B
9&$3B

î—)

9''

966


îQ)
î—)

/HYHOVKLIWHU

287

9''

%DFNXSFLUFXLWU\
26&.57&
:DNHXSORJLF
%DFNXSUHJLVWHUV
EDFNXS5$0

3RZHU
VZLWFK

,2
/RJLF
.HUQHOORJLF
&38GLJLWDO
5$0 

9ROWDJH
UHJXODWRU

)ODVKPHPRU\

%<3$66B5(*
3'5B21
9''

9''$
95()

Q)
—)

5HVHW
FRQWUROOHU

Q)
—)

95()
95()

$'&

$QDORJ
5&V
3//

966$

069

1. VDDA and VSSA must be connected to VDD and VSS, respectively.

116/1747

RM0090 Rev 17

RM0090

Power controller (PWR)
Figure 10. Power supply overview for STM32F42xxx and STM32F43xxx
6"!4

/54

'0 )/S
).

 §  —&

)/
,OGIC
+ERNEL LOGIC
#05
DIGITAL
 2!-

6#!0?
6#!0?

6$$


 §  N&
 §  —&

,EVEL SHIFTER

6"!4 
 TO 6

6$$

"ACKUP CIRCUITRY
/3#+ 24#
7AKEUP LOGIC
"ACKUP REGISTERS
BACKUP 2!-

0O WER SWI TCH

6OLTAGE
REGULATOR

633



&LASH MEMORY

"90!33?2%'
0$2?/.
6$$

6$$!
62%&

 N&
 —&

2ESET
CONTROLLER

 N&
 —&

62%&
62%&

!$#

!NALOG
2#S 0,,


633!

-36

1. VDDA and VSSA must be connected to VDD and VSS, respectively.

5.1.1

Independent A/D converter supply and reference voltage
To improve conversion accuracy, the ADC has an independent power supply which can be
separately filtered and shielded from noise on the PCB.
•

The ADC voltage supply input is available on a separate VDDA pin.

•

An isolated supply ground connection is provided on pin VSSA.

To ensure a better accuracy of low voltage inputs, the user can connect a separate external
reference voltage ADC input on VREF. The voltage on VREF ranges from 1.8 V to VDDA.

RM0090 Rev 17

117/1747
149

Power controller (PWR)

5.1.2

RM0090

Battery backup domain
Backup domain description
To retain the content of the RTC backup registers, backup SRAM, and supply the RTC when
VDD is turned off, VBAT pin can be connected to an optional standby voltage supplied by a
battery or by another source.
To allow the RTC to operate even when the main digital supply (VDD) is turned off, the VBAT
pin powers the following blocks:
•

The RTC

•

The LSE oscillator

•

The backup SRAM when the low-power backup regulator is enabled

•

PC13 to PC15 I/Os, plus PI8 I/O (when available)

The switch to the VBAT supply is controlled by the power-down reset embedded in the Reset
block.

Warning:

During tRSTTEMPO (temporization at VDD startup) or after a PDR
is detected, the power switch between VBAT and VDD remains
connected to VBAT.
During the startup phase, if VDD is established in less than
tRSTTEMPO (Refer to the datasheet for the value of tRSTTEMPO)
and VDD > VBAT + 0.6 V, a current may be injected into VBAT
through an internal diode connected between VDD and the
power switch (VBAT).
If the power supply/battery connected to the VBAT pin cannot
support this current injection, it is strongly recommended to
connect an external low-drop diode between this power
supply and the VBAT pin.

If no external battery is used in the application, it is recommended to connect the VBAT pin to
VDD with a 100 nF external decoupling ceramic capacitor in parallel.
When the backup domain is supplied by VDD (analog switch connected to VDD), the
following functions are available:

Note:

118/1747

•

PC14 and PC15 can be used as either GPIO or LSE pins

•

PC13 can be used as a GPIOas the RTC_AF1 pin (refer to Table 37: RTC_AF1 pin for
more details about this pin configuration)

Due to the fact that the switch only sinks a limited amount of current (3 mA), the use of PI8
and PC13 to PC15 GPIOs in output mode is restricted: the speed has to be limited to 2 MHz
with a maximum load of 30 pF and these I/Os must not be used as a current source (e.g. to
drive an LED).

RM0090 Rev 17

RM0090

Power controller (PWR)
When the backup domain is supplied by VBAT (analog switch connected to VBAT because
VDD is not present), the following functions are available:
•

PC14 and PC15 can be used as LSE pins only

•

PC13 can be used as the RTC_AF1 pin (refer to Table 37: RTC_AF1 pin for more
details about this pin configuration)

•

PI8 can be used as RTC_AF2

Backup domain access
After reset, the backup domain (RTC registers, RTC backup register and backup SRAM) is
protected against possible unwanted write accesses. To enable access to the backup
domain, proceed as follows:
•

Access to the RTC and RTC backup registers

1.

Enable the power interface clock by setting the PWREN bits in the RCC_APB1ENR
register (see Section 7.3.13 and Section 6.3.13)

2.

Set the DBP bit in the Section 5.4.1 and PWR power control register (PWR_CR) for
STM32F42xxx and STM32F43xxx to enable access to the backup domain

3.

Select the RTC clock source: see Section 7.2.8: RTC/AWU clock

4.

Enable the RTC clock by programming the RTCEN [15] bit in the Section 7.3.20: RCC
Backup domain control register (RCC_BDCR)

•

Access to the backup SRAM

1.

Enable the power interface clock by setting the PWREN bits in the RCC_APB1ENR
register (see Section 7.3.13 and Section 6.3.13 for STM32F405xx/07xx and
STM32F415xx/17xx and STM32F42xxx and STM32F43xxx, respectively)

2.

Set the DBP bit in the PWR power control register (PWR_CR) for STM32F405xx/07xx
and STM32F415xx/17xx and PWR power control register (PWR_CR) for
STM32F42xxx and STM32F43xxx to enable access to the backup domain

3.

Enable the backup SRAM clock by setting BKPSRAMEN bit in the RCC AHB1
peripheral clock enable register (RCC_AHB1ENR).

RTC and RTC backup registers
The real-time clock (RTC) is an independent BCD timer/counter. The RTC provides a timeof-day clock/calendar, two programmable alarm interrupts, and a periodic programmable
wakeup flag with interrupt capability. The RTC contains 20 backup data registers (80 bytes)
which are reset when a tamper detection event occurs. For more details refer to Section 26:
Real-time clock (RTC).

Backup SRAM
The backup domain includes 4 Kbytes of backup SRAM addressed in 32-bit, 16-bit or 8-bit
mode. Its content is retained even in Standby or VBAT mode when the low-power backup
regulator is enabled. It can be considered as an internal EEPROM when VBAT is always
present.
When the backup domain is supplied by VDD (analog switch connected to VDD), the backup
SRAM is powered from VDD which replaces the VBAT power supply to save battery life.
When the backup domain is supplied by VBAT (analog switch connected to VBAT because
VDD is not present), the backup SRAM is powered by a dedicated low-power regulator. This
regulator can be ON or OFF depending whether the application needs the backup SRAM
function in Standby and VBAT modes or not. The power-down of this regulator is controlled
RM0090 Rev 17

119/1747
149

Power controller (PWR)

RM0090

by a dedicated bit, the BRE control bit of the PWR_CSR register (see Section 5.4.2: PWR
power control/status register (PWR_CSR) for STM32F405xx/07xx and
STM32F415xx/17xx).
The backup SRAM is not mass erased by an tamper event. When the Flash memory is read
protected, the backup SRAM is also read protected to prevent confidential data, such as
cryptographic private key, from being accessed. When the protection level change from
level 1 to level 0 is requested, the backup SRAM content is erased.
Figure 11. Backup domain

3RZHUVZLWFK
9ROWDJHUHJXODWRU
!9

/3YROWDJHUHJXODWRU
!9

9GRPDLQ

%DFNXS65$0
LQWHUIDFH

%DFNXS65$0
9
57&

/6(+]

%DFNXSGRPDLQ

069

5.1.3

Voltage regulator for STM32F405xx/07xx and STM32F415xx/17xx
An embedded linear voltage regulator supplies all the digital circuitries except for the backup
domain and the Standby circuitry. The regulator output voltage is around 1.2 V.
This voltage regulator requires one or two external capacitors to be connected to one or two
dedicated pins, VCAP_1 and VCAP_2 available in all packages. Specific pins must be
connected either to VSS or VDD to activate or deactivate the voltage regulator. These pins
depend on the package.
When activated by software, the voltage regulator is always enabled after Reset. It works in
three different modes depending on the application modes.
•

In Run mode, the regulator supplies full power to the 1.2 V domain (core, memories
and digital peripherals). In this mode, the regulator output voltage (around 1.2 V) can
be scaled by software to different voltage values:
Scale 1 or scale 2 can be configured on the fly through VOS (bit 15 of the
PWR_CR register).
The voltage scaling allows optimizing the power consumption when the device is
clocked below the maximum system frequency.

•

120/1747

In Stop mode, the main regulator or the low-power regulator supplies to the 1.2 V
domain, thus preserving the content of registers and internal SRAM. The voltage

RM0090 Rev 17

RM0090

Power controller (PWR)
regulator can be put either in main regulator mode (MR) or in low-power mode (LPR).
The programmed voltage scale remains the same during Stop mode:
The programmed voltage scale remains the same during Stop mode (see
Section 5.4.1: PWR power control register (PWR_CR) for STM32F405xx/07xx
and STM32F415xx/17xx).
•

In Standby mode, the regulator is powered down. The content of the registers and
SRAM are lost except for the Standby circuitry and the backup domain.

Note:

For more details, refer to the voltage regulator section in the STM32F405xx/07xx and
STM32F415xx/17xx datasheets.

5.1.4

Voltage regulator for STM32F42xxx and STM32F43xxx
An embedded linear voltage regulator supplies all the digital circuitries except for the backup
domain and the Standby circuitry. The regulator output voltage is around 1.2 V.
This voltage regulator requires two external capacitors to be connected to two dedicated
pins, VCAP_1 and VCAP_2 available in all packages. Specific pins must be connected either to
VSS or VDD to activate or deactivate the voltage regulator. These pins depend on the
package.
When activated by software, the voltage regulator is always enabled after Reset. It works in
three different modes depending on the application modes (Run, Stop, or Standby mode).
•

In Run mode, the main regulator supplies full power to the 1.2 V domain (core,
memories and digital peripherals). In this mode, the regulator output voltage (around
1.2 V) can be scaled by software to different voltage values (scale 1, scale 2, and scale
3 can be configured through VOS[1:0] bits of the PWR_CR register). The scale can be
modified only when the PLL is OFF and the HSI or HSE clock source is selected as
system clock source. The new value programmed is active only when the PLL is ON.
When the PLL is OFF, the voltage scale 3 is automatically selected.
The voltage scaling allows optimizing the power consumption when the device is
clocked below the maximum system frequency. After exit from Stop mode, the voltage
scale 3 is automatically selected.(see Section 5.4.1: PWR power control register
(PWR_CR) for STM32F405xx/07xx and STM32F415xx/17xx.
2 operating modes are available:

•

–

Normal mode: The CPU and core logic operate at maximum frequency at a given
voltage scaling (scale 1, scale 2 or scale 3)

–

Over-drive mode: This mode allows the CPU and the core logic to operate at a
higher frequency than the normal mode for the voltage scaling scale 1 and scale
2.

In Stop mode: the main regulator or low-power regulator supplies a low-power voltage
to the 1.2V domain, thus preserving the content of registers and internal SRAM.
The voltage regulator can be put either in main regulator mode (MR) or in low-power
mode (LPR). Both modes can be configured by software as follows:
–

Normal mode: the 1.2 V domain is preserved in nominal leakage mode. It is the
default mode when the main regulator (MR) or the low-power regulator (LPR) is
enabled.

–

Under-drive mode: the 1.2 V domain is preserved in reduced leakage mode. This
mode is only available with the main regulator or in low-power regulator mode

RM0090 Rev 17

121/1747
149

Power controller (PWR)

RM0090

(see Table 22).
•
Note:

In Standby mode: the regulator is powered down. The content of the registers and
SRAM are lost except for the Standby circuitry and the backup domain.

Over-drive and under-drive mode are not available when the regulator is bypassed.
For more details, refer to the voltage regulator section in the STM32F42xxx and
STM32F43xxx datasheets.
Table 22. Voltage regulator configuration mode versus device operating mode(1)
Voltage regulator
configuration

Run mode

Sleep mode

Stop mode

Standby mode

Normal mode

MR

MR

MR or LPR

-

Over-drive
mode(2)

MR

MR

-

-

Under-drive mode

-

-

MR or LPR

-

Power-down
mode

-

-

-

Yes

1. ‘-’ means that the corresponding configuration is not available.
2. The over-drive mode is not available when VDD = 1.8 to 2.1 V.

Entering Over-drive mode
It is recommended to enter Over-drive mode when the application is not running critical
tasks and when the system clock source is either HSI or HSE. To optimize the configuration
time, enable the Over-drive mode during the PLL lock phase.
To enter Over-drive mode, follow the sequence below:

Note:

1.

Select HSI or HSE as system clock.

2.

Configure RCC_PLLCFGR register and set PLLON bit of RCC_CR register.

3.

Set ODEN bit of PWR_CR register to enable the Over-drive mode and wait for the
ODRDY flag to be set in the PWR_CSR register.

4.

Set the ODSW bit in the PWR_CR register to switch the voltage regulator from Normal
mode to Over-drive mode. The System will be stalled during the switch but the PLL
clock system will be still running during locking phase.

5.

Wait for the ODSWRDY flag in the PWR_CSR to be set.

6.

Select the required Flash latency as well as AHB and APB prescalers.

7.

Wait for PLL lock.

8.

Switch the system clock to the PLL.

9.

Enable the peripherals that are not generated by the System PLL (I2S clock, LCD-TFT
clock, SAI1 clock, USB_48MHz clock....).

The PLLI2S and PLLSAI can be configured at the same time as the system PLL.
During the Over-drive switch activation, no peripheral clocks should be enabled. The
peripheral clocks must be enabled once the Over-drive mode is activated.
Entering Stop mode disables the Over-drive mode, as well as the PLL. The application
software has to configure again the Over-drive mode and the PLL after exiting from Stop
mode.

122/1747

RM0090 Rev 17

RM0090

Power controller (PWR)

Exiting from Over-drive mode
It is recommended to exit from Over-drive mode when the application is not running critical
tasks and when the system clock source is either HSI or HSE.There are two sequences that
allow exiting from over-drive mode:
•

By resetting simultaneously the ODEN and ODSW bits bit in the PWR_CR register
(sequence 1)

•

By resetting first the ODSW bit to switch the voltage regulator to Normal mode and then
resetting the ODEN bit to disable the Over-drive mode (sequence 2).

Example of sequence 1:
1.

Select HSI or HSE as system clock source.

2.

Disable the peripheral clocks that are not generated by the System PLL (I2S clock,
LCD-TFT clock, SAI1 clock, USB_48MHz clock,....)

3.

Reset simultaneously the ODEN and the ODSW bits in the PWR_CR register to switch
back the voltage regulator to Normal mode and disable the Over-drive mode.

4.

Wait for the ODWRDY flag of PWR_CSR to be reset.

Example of sequence 2:
1.

Select HSI or HSE as system clock source.

2.

Disable the peripheral clocks that are not generated by the System PLL (I2S clock,
LCD-TFT clock, SAI1 clock, USB_48MHz clock,....).

3.

Reset the ODSW bit in the PWR_CR register to switch back the voltage regulator to
Normal mode. The system clock is stalled during voltage switching.

4.

Wait for the ODWRDY flag of PWR_CSR to be reset.

5.

Reset the ODEN bit in the PWR_CR register to disable the Over-drive mode.

Note:

During step 3, the ODEN bit remains set and the Over-drive mode is still enabled but not
active (ODSW bit is reset). If the ODEN bit is reset instead, the Over-drive mode is disabled
and the voltage regulator is switched back to the initial voltage.

5.2

Power supply supervisor

5.2.1

Power-on reset (POR)/power-down reset (PDR)
The device has an integrated POR/PDR circuitry that allows proper operation starting
from 1.8 V.
The device remains in Reset mode when VDD/VDDA is below a specified threshold,
VPOR/PDR, without the need for an external reset circuit. For more details concerning the

RM0090 Rev 17

123/1747
149

Power controller (PWR)

RM0090

power on/power-down reset threshold, refer to the electrical characteristics of the
datasheet.
Figure 12. Power-on reset/power-down reset waveform
6$$6$$!

0$2
 M6
HYSTERESIS

0$2

4EMPORIZATION
T2344%-0/

2ESET
-36

5.2.2

Brownout reset (BOR)
During power on, the Brownout reset (BOR) keeps the device under reset until the supply
voltage reaches the specified VBOR threshold.
VBOR is configured through device option bytes. By default, BOR is off. 3 programmable
VBOR threshold levels can be selected:

Note:

•

BOR Level 3 (VBOR3). Brownout threshold level 3.

•

BOR Level 2 (VBOR2). Brownout threshold level 2.

•

BOR Level 1 (VBOR1). Brownout threshold level 1.

For full details about BOR characteristics, refer to the "Electrical characteristics" section in
the device datasheet.
When the supply voltage (VDD) drops below the selected VBOR threshold, a device reset is
generated.
The BOR can be disabled by programming the device option bytes. In this case, the
power-on and power-down is then monitored by the POR/ PDR (see Section 5.2.1: Poweron reset (POR)/power-down reset (PDR)).
The BOR threshold hysteresis is ~100 mV (between the rising and the falling edge of the
supply voltage).

124/1747

RM0090 Rev 17

RM0090

Power controller (PWR)
Figure 13. BOR thresholds
6$$6$$!

"/2 THRESHOLD

 M6
HYSTERESIS

2ESET

-36

5.2.3

Programmable voltage detector (PVD)
You can use the PVD to monitor the VDD power supply by comparing it to a threshold
selected by the PLS[2:0] bits in the PWR power control register (PWR_CR) for
STM32F405xx/07xx and STM32F415xx/17xx and PWR power control register (PWR_CR)
for STM32F42xxx and STM32F43xxx.
The PVD is enabled by setting the PVDE bit.
A PVDO flag is available, in the PWR power control/status register (PWR_CSR) for
STM32F405xx/07xx and STM32F415xx/17xx, to indicate if VDD is higher or lower than the
PVD threshold. This event is internally connected to the EXTI line16 and can generate an
interrupt if enabled through the EXTI registers. The PVD output interrupt can be generated
when VDD drops below the PVD threshold and/or when VDD rises above the PVD threshold
depending on EXTI line16 rising/falling edge configuration. As an example the service
routine could perform emergency shutdown tasks.

RM0090 Rev 17

125/1747
149

Power controller (PWR)

RM0090
Figure 14. PVD thresholds

9''

39'WKUHVKROG

P9
K\VWHUHVLV

39'RXWSXW

069

5.3

Low-power modes
By default, the microcontroller is in Run mode after a system or a power-on reset. In Run
mode the CPU is clocked by HCLK and the program code is executed. Several low-power
modes are available to save power when the CPU does not need to be kept running, for
example when waiting for an external event. It is up to the user to select the mode that gives
the best compromise between low-power consumption, short startup time and available
wakeup sources.
The devices feature three low-power modes:
•

Sleep mode (Cortex®-M4 with FPU core stopped, peripherals kept running)

•

Stop mode (all clocks are stopped)

•

Standby mode (1.2 V domain powered off)

In addition, the power consumption in Run mode can be reduce by one of the following
means:
•

Slowing down the system clocks

•

Gating the clocks to the APBx and AHBx peripherals when they are unused.

Entering low-power mode
Low-power modes are entered by the MCU by executing the WFI (Wait For Interrupt), or
WFE (Wait for Event) instructions, or when the SLEEPONEXIT bit in the Cortex®-M4 with
FPU System Control register is set on Return from ISR.
Entering Low-power mode through WFI or WFE will be executed only if no interrupt is
pending or no event is pending.

126/1747

RM0090 Rev 17

RM0090

Power controller (PWR)

Exiting low-power mode
The MCU exits from Sleep and Stop modes low-power mode depending on the way the lowpower mode was entered:
•
If the WFI instruction or Return from ISR was used to enter the low-power mode, any
peripheral interrupt acknowledged by the NVIC can wake up the device.
•

If the WFE instruction is used to enter the low-power mode, the MCU exits the lowpower mode as soon as an event occurs. The wakeup event can be generated either
by:
–

NVIC IRQ interrupt:
When SEVONPEND = 0 in the Cortex®-M4 with FPU System Control register: by
enabling an interrupt in the peripheral control register and in the NVIC. When the
MCU resumes from WFE, the peripheral interrupt pending bit and the NVIC
peripheral IRQ channel pending bit (in the NVIC interrupt clear pending register)
have to be cleared. Only NVIC interrupts with sufficient priority will wakeup and
interrupt the MCU.
When SEVONPEND = 1 in the Cortex®-M4 with FPU System Control register: by
enabling an interrupt in the peripheral control register and optionally in the NVIC.
When the MCU resumes from WFE, the peripheral interrupt pending bit and when
enabled the NVIC peripheral IRQ channel pending bit (in the NVIC interrupt clear
pending register) have to be cleared. All NVIC interrupts will wakeup the MCU,
even the disabled ones.Only enabled NVIC interrupts with sufficient priority will
wakeup and interrupt the MCU.

–

Event
This is done by configuring a EXTI line in event mode. When the CPU resumes
from WFE, it is not necessary to clear the EXTI peripheral interrupt pending bit or
the NVIC IRQ channel pending bit as the pending bits corresponding to the event
line is not set. It may be necessary to clear the interrupt flag in the peripheral.

The MCU exits from Standby low-power mode through an external reset (NRST pin), an
IWDG reset, a rising edge on one of the enabled WKUPx pins or a RTC event occurs (see
Figure 237: RTC block diagram).
After waking up from Standby mode, program execution restarts in the same way as after a
Reset (boot pin sampling, option bytes loading, reset vector is fetched, etc.).

RM0090 Rev 17

127/1747
149

Power controller (PWR)

RM0090

Only enabled NVIC interrupts with sufficient priority will wakeup and interrupt the MCU.
Table 23. Low-power mode summary
Mode name

Entry

Wakeup

Sleep
(Sleep now or
Sleep-onexit)

WFI or Return
from ISR

Any interrupt

WFE

Wakeup event

Stop

Effect on 1.2 V
domain clocks

Effect on
VDD
domain
clocks

Voltage regulator

CPU CLK OFF
no effect on other
clocks or analog
clock sources

None

ON

PDDS and LPDS
bits +
Any EXTI line (configured
SLEEPDEEP bit
in the EXTI registers,
+ WFI, Return internal and external lines)
from ISR or WFE
All 1.2 V domain
clocks OFF

Standby

5.3.1

PDDS bit +
SLEEPDEEP bit
+ WFI, Return
from ISR or WFE

WKUP pin rising edge,
RTC alarm (Alarm A or
Alarm B), RTC Wakeup
event, RTC tamper
events, RTC time stamp
event, external reset in
NRST pin, IWDG reset

ON or in low- power
mode (depends on
PWR power control
register (PWR_CR)
for
STM32F405xx/07x
x and
STM32F415xx/17x
x and PWR power
control register
(PWR_CR) for
HSI and STM32F405xx/07x
x and
HSE
STM32F415xx/17x
oscillator
xPWR power
s OFF
control register
(PWR_CR) for
STM32F42xxx and
STM32F43xxx

OFF

Slowing down system clocks
In Run mode the speed of the system clocks (SYSCLK, HCLK, PCLK1, PCLK2) can be
reduced by programming the prescaler registers. These prescalers can also be used to slow
down peripherals before entering Sleep mode.
For more details refer to Section 7.3.3: RCC clock configuration register (RCC_CFGR).

5.3.2

Peripheral clock gating
In Run mode, the HCLKx and PCLKx for individual peripherals and memories can be
stopped at any time to reduce power consumption.
To further reduce power consumption in Sleep mode the peripheral clocks can be disabled
prior to executing the WFI or WFE instructions.

128/1747

RM0090 Rev 17

RM0090

Power controller (PWR)
Peripheral clock gating is controlled by the AHB1 peripheral clock enable register
(RCC_AHB1ENR), AHB2 peripheral clock enable register (RCC_AHB2ENR), AHB3
peripheral clock enable register (RCC_AHB3ENR) (see Section 7.3.10: RCC AHB1
peripheral clock enable register (RCC_AHB1ENR), Section 7.3.11: RCC AHB2 peripheral
clock enable register (RCC_AHB2ENR), Section 7.3.12: RCC AHB3 peripheral clock
enable register (RCC_AHB3ENR) for STM32F405xx/07xx and STM32F415xx/17xx, and
Section 6.3.10: RCC AHB1 peripheral clock register (RCC_AHB1ENR), Section 6.3.11:
RCC AHB2 peripheral clock enable register (RCC_AHB2ENR), and Section 6.3.12: RCC
AHB3 peripheral clock enable register (RCC_AHB3ENR) for STM32F42xxx and
STM32F43xxx).
Disabling the peripherals clocks in Sleep mode can be performed automatically by resetting
the corresponding bit in RCC_AHBxLPENR and RCC_APBxLPENR registers.

5.3.3

Sleep mode
Entering Sleep mode
The Sleep mode is entered according to Section : Entering low-power mode, when the
SLEEPDEEP bit in the Cortex®-M4 with FPU System Control register is cleared.
Refer to Table 24 and Table 25 for details on how to enter Sleep mode.

Note:

All interrupt pending bits must be cleared before the sleep mode entry.

Exiting Sleep mode
The Sleep mode is exited according to Section : Exiting low-power mode.
Refer to Table 24 and Table 25 for more details on how to exit Sleep mode.
Table 24. Sleep-now entry and exit
Sleep-now mode

Description
WFI (Wait for Interrupt) or WFE (Wait for Event) while:
– SLEEPDEEP = 0, and
– No interrupt (for WFI) or event (for WFE) is pending.
Refer to the Cortex®-M4 with FPU System Control register.

Mode entry

On Return from ISR while:
– SLEEPDEEP = 0 and
– SLEEPONEXIT = 1,
– No interrupt is pending.
Refer to the Cortex®-M4 with FPU System Control register.

RM0090 Rev 17

129/1747
149

Power controller (PWR)

RM0090
Table 24. Sleep-now entry and exit (continued)

Sleep-now mode

Description

Mode exit

If WFI or Return from ISR was used for entry:
Interrupt: Refer to Table 61: Vector table for STM32F405xx/07xx and
STM32F415xx/17xx and Table 62: Vector table for STM32F42xxx and
STM32F43xxx
If WFE was used for entry and SEVONPEND = 0
Wakeup event: Refer to Section 12.2.3: Wakeup event management
f WFE was used for entry and SEVONPEND = 1
Interrupt even when disabled in NVIC: refer to Table 61: Vector table for
STM32F405xx/07xx and STM32F415xx/17xx and Table 62: Vector table
for STM32F42xxx and STM32F43xxx or Wakeup event (see
Section 12.2.3: Wakeup event management).

Wakeup latency

None

Table 25. Sleep-on-exit entry and exit
Sleep-on-exit

Description
WFI (Wait for Interrupt) or WFE (Wait for Event) while:
– SLEEPDEEP = 0, and
– No interrupt (for WFI) or event (for WFE) is pending.
Refer to the Cortex®-M4 with FPU System Control register.

Mode entry

5.3.4

On Return from ISR while:
– SLEEPDEEP = 0, and
– SLEEPONEXIT = 1, and
– No interrupt is pending.
Refer to the Cortex®-M4 with FPU System Control register.

Mode exit

Interrupt: refer to Table 61: Vector table for STM32F405xx/07xx and
STM32F415xx/17xx and Table 62: Vector table for STM32F42xxx and
STM32F43xxx

Wakeup latency

None

Stop mode (STM32F405xx/07xx and STM32F415xx/17xx)
The Stop mode is based on the Cortex®-M4 with FPU deepsleep mode combined with
peripheral clock gating. The voltage regulator can be configured either in normal or lowpower mode. In Stop mode, all clocks in the 1.2 V domain are stopped, the PLLs, the HSI
and the HSE RC oscillators are disabled. Internal SRAM and register contents are
preserved.
By setting the FPDS bit in the PWR_CR register, the Flash memory also enters power-down
mode when the device enters Stop mode. When the Flash memory is in power-down mode,
an additional startup delay is incurred when waking up from Stop mode (see Table 26: Stop
operating modes (STM32F405xx/07xx and STM32F415xx/17xx) and Section 5.4.1: PWR
power control register (PWR_CR) for STM32F405xx/07xx and STM32F415xx/17xx).

130/1747

RM0090 Rev 17

RM0090

Power controller (PWR)
Table 26. Stop operating modes
(STM32F405xx/07xx and STM32F415xx/17xx)
Stop mode

LPDS bit

FPDS bit

Wake-up latency

STOP MR
(Main regulator)

0

0

HSI RC startup time

STOP MR-FPD

0

1

HSI RC startup time +
Flash wakeup time from Power Down
mode

STOP LP

1

0

HSI RC startup time +
regulator wakeup time from LP mode

1

HSI RC startup time +
Flash wakeup time from Power Down
mode +
regulator wakeup time from LP mode

STOP LP-FPD

1

Entering Stop mode (for STM32F405xx/07xx and STM32F415xx/17xx)
The Stop mode is entered according to Section : Entering low-power mode, when the
SLEEPDEEP bit in the Cortex®-M4 with FPU System Control register is set.
Refer to Table 27 for details on how to enter the Stop mode.
To further reduce power consumption in Stop mode, the internal voltage regulator can be put
in low-power mode. This is configured by the LPDS bit of the PWR power control register
(PWR_CR) for STM32F405xx/07xx and STM32F415xx/17xx and PWR power control
register (PWR_CR) for STM32F42xxx and STM32F43xxx.
If Flash memory programming is ongoing, the Stop mode entry is delayed until the memory
access is finished.
If an access to the APB domain is ongoing, The Stop mode entry is delayed until the APB
access is finished.
In Stop mode, the following features can be selected by programming individual control bits:
•

Independent watchdog (IWDG): the IWDG is started by writing to its Key register or by
hardware option. Once started it cannot be stopped except by a Reset. See
Section 21.3 in Section 21: Independent watchdog (IWDG).

•

Real-time clock (RTC): this is configured by the RTCEN bit in the Section 7.3.20: RCC
Backup domain control register (RCC_BDCR)

•

Internal RC oscillator (LSI RC): this is configured by the LSION bit in the
Section 7.3.21: RCC clock control & status register (RCC_CSR).

•

External 32.768 kHz oscillator (LSE OSC): this is configured by the LSEON bit in the
Section 7.3.20: RCC Backup domain control register (RCC_BDCR).

The ADC or DAC can also consume power during the Stop mode, unless they are disabled
before entering it. To disable them, the ADON bit in the ADC_CR2 register and the ENx bit
in the DAC_CR register must both be written to 0.

RM0090 Rev 17

131/1747
149

Power controller (PWR)
Note:

RM0090

If the application needs to disable the external clock before entering Stop mode, the HSEON
bit must first be disabled and the system clock switched to HSI.
Otherwise, if the HSEON bit is kept enabled while the external clock (external oscillator) can
be removed before entering stop mode, the clock security system (CSS) feature must be
enabled to detect any external oscillator failure and avoid a malfunction behavior when
entering stop mode.

Exiting Stop mode (for STM32F405xx/07xx and STM32F415xx/17xx)
The Stop mode is exited according to Section : Exiting low-power mode.
Refer to Table 27 for more details on how to exit Stop mode.
When exiting Stop mode by issuing an interrupt or a wakeup event, the HSI RC oscillator is
selected as system clock.
When the voltage regulator operates in low-power mode, an additional startup delay is
incurred when waking up from Stop mode. By keeping the internal regulator ON during Stop
mode, the consumption is higher although the startup time is reduced.
Table 27. Stop mode entry and exit (for STM32F405xx/07xx and STM32F415xx/17xx)
Stop mode

Description
WFI (Wait for Interrupt) or WFE (Wait for Event) while:
– No interrupt (for WFI) or event (for WFE) is pending,
– SLEEPDEEP bit is set in Cortex®-M4 with FPU System Control register,
– PDDS bit is cleared in Power Control register (PWR_CR),
– Select the voltage regulator mode by configuring LPDS bit in PWR_CR.

Mode entry

On Return from ISR:
– No interrupt is pending,
– SLEEPDEEP bit is set in Cortex®-M4 with FPU System Control register,
– SLEEPONEXIT = 1,
– PDDS bit is cleared in Power Control register (PWR_CR).
Note: To enter Stop mode, all EXTI Line pending bits (in Pending register
(EXTI_PR)), all peripheral interrupts pending bits, the RTC Alarm
(Alarm A and Alarm B), RTC wakeup, RTC tamper, and RTC time
stamp flags, must be reset. Otherwise, the Stop mode entry
procedure is ignored and program execution continues.

132/1747

RM0090 Rev 17

RM0090

Power controller (PWR)
Table 27. Stop mode entry and exit (for STM32F405xx/07xx and STM32F415xx/17xx)
Stop mode

5.3.5

Description

Mode exit

If WFI or Return from ISR was used for entry:
Any EXTI lines configured in Interrupt mode (the corresponding EXTI
Interrupt vector must be enabled in the NVIC). The interrupt source can
be external interrupts or peripherals with wakeup capability. Refer to
Table 61: Vector table for STM32F405xx/07xx and STM32F415xx/17xx
on page 372 and Table 62: Vector table for STM32F42xxx and
STM32F43xxx.
If WFE was used for entry and SEVONPEND = 0
Any EXTI lines configured in event mode. Refer to Section 12.2.3:
Wakeup event management on page 380.
If WFE was used for entry and SEVONPEND = 1:
– Any EXTI lines configured in Interrupt mode (even if the corresponding
EXTI Interrupt vector is disabled in the NVIC). The interrupt source can
be an external interrupt or a peripheral with wakeup capability. Refer to
Table 61: Vector table for STM32F405xx/07xx and STM32F415xx/17xx
on page 372 and Table 62: Vector table for STM32F42xxx and
STM32F43xxx.
– Wakeup event: refer to Section 12.2.3: Wakeup event management on
page 380.

Wakeup latency

Table 26: Stop operating modes (STM32F405xx/07xx and
STM32F415xx/17xx)

Stop mode (STM32F42xxx and STM32F43xxx)
The Stop mode is based on the Cortex®-M4 with FPU deepsleep mode combined with
peripheral clock gating. The voltage regulator can be configured either in normal or lowpower mode. In Stop mode, all clocks in the 1.2 V domain are stopped, the PLLs, the HSI
and the HSE RC oscillators are disabled. Internal SRAM and register contents are
preserved.
In Stop mode, the power consumption can be further reduced by using additional settings in
the PWR_CR register. However this will induce an additional startup delay when waking up
from Stop mode (see Table 28).

RM0090 Rev 17

133/1747
149

Power controller (PWR)

RM0090

Table 28. Stop operating modes (STM32F42xxx and STM32F43xxx)
UDEN[1:0]
bits

MRUDS
bit

LPUDS
bit

STOP MR
(Main Regulator)

-

0

-

0

0

HSI RC startup time

STOP MR- FPD

-

0

-

0

1

HSI RC startup time +
Flash wakeup time from powerdown mode

STOP LP

-

0

0

1

0

HSI RC startup time +
regulator wakeup time from LP
mode

1

HSI RC startup time +
Flash wakeup time from powerdown mode +
regulator wakeup time from LP
mode

-

HSI RC startup time +
Flash wakeup time from powerdown mode +
Main regulator wakeup time from
under-drive mode + Core logic to
nominal mode

-

HSI RC startup time +
Flash wakeup time from powerdown mode +
regulator wakeup time from LP
under-drive mode + Core logic to
nominal mode

Voltage Regulator Mode

Normal
mode

STOP LP-FPD

STOP UMRFPD

-

3

-

1

LPDS FPDS
bit
bit

0

-

1

0

Underdrive
Mode
STOP ULP-FPD

3

-

1

1

Wakeup latency

Entering Stop mode (STM32F42xxx and STM32F43xxx)
The Stop mode is entered according to Section : Entering low-power mode, when the
SLEEPDEEP bit in the Cortex®-M4 with FPU System Control register is set.
Refer to Table 29 for details on how to enter the Stop mode.
When the microcontroller enters in Stop mode, the voltage scale 3 is automatically selected.
To further reduce power consumption in Stop mode, the internal voltage regulator can be put
in low voltage mode. This is configured by the LPDS, MRUDS, LPUDS and UDEN bits of
the PWR power control register (PWR_CR) for STM32F405xx/07xx and
STM32F415xx/17xx.
If Flash memory programming is ongoing, the Stop mode entry is delayed until the memory
access is finished.
If an access to the APB domain is ongoing, The Stop mode entry is delayed until the APB
access is finished.
If the Over-drive mode was enabled before entering Stop mode, it is automatically disabled
during when the Stop mode is activated.

134/1747

RM0090 Rev 17

RM0090

Power controller (PWR)
In Stop mode, the following features can be selected by programming individual control bits:
•

Independent watchdog (IWDG): the IWDG is started by writing to its Key register or by
hardware option. Once started it cannot be stopped except by a Reset. See
Section 21.3 in Section 21: Independent watchdog (IWDG).

•

Real-time clock (RTC): this is configured by the RTCEN bit in the Section 7.3.20: RCC
Backup domain control register (RCC_BDCR)

•

Internal RC oscillator (LSI RC): this is configured by the LSION bit in the
Section 7.3.21: RCC clock control & status register (RCC_CSR).

•

External 32.768 kHz oscillator (LSE OSC): this is configured by the LSEON bit in the
RCC Backup domain control register (RCC_BDCR).

The ADC or DAC can also consume power during the Stop mode, unless they are disabled
before entering it. To disable them, the ADON bit in the ADC_CR2 register and the ENx bit
in the DAC_CR register must both be written to 0.
Note:

Before entering Stop mode, it is recommended to enable the clock security system (CSS)
feature to prevent external oscillator (HSE) failure from impacting the internal MCU
behavior.

Exiting Stop mode (STM32F42xxx and STM32F43xxx)
The Stop mode is exited according to Section : Exiting low-power mode.
Refer to Table 29 for more details on how to exit Stop mode.
When exiting Stop mode by issuing an interrupt or a wakeup event, the HSI RC oscillator is
selected as system clock.
If the Under-drive mode was enabled, it is automatically disabled after exiting Stop mode.
When the voltage regulator operates in low voltage mode, an additional startup delay is
incurred when waking up from Stop mode. By keeping the internal regulator ON during Stop
mode, the consumption is higher although the startup time is reduced.
When the voltage regulator operates in Under-drive mode, an additional startup delay is
induced when waking up from Stop mode.

RM0090 Rev 17

135/1747
149

Power controller (PWR)

RM0090

Table 29. Stop mode entry and exit (STM32F42xxx and STM32F43xxx)
Stop mode

Description
WFI (Wait for Interrupt) or WFE (Wait for Event) while:
– No interrupt or event is pending,
– SLEEPDEEP bit is set in Cortex®-M4 with FPU System Control register,
– PDDS bit is cleared in Power Control register (PWR_CR),
– Select the voltage regulator mode by configuring LPDS, MRUDS, LPUDS
and UDEN bits in PWR_CR (see Table 28: Stop operating modes
(STM32F42xxx and STM32F43xxx)).

Mode entry

On Return from ISR while:
– No interrupt is pending,
– SLEEPDEEP bit is set in Cortex®-M4 with FPU System Control register,
and
– SLEEPONEXIT = 1, and
– PDDS is cleared in PWR_CR1.
Note: To enter Stop mode, all EXTI Line pending bits (in Pending register
(EXTI_PR)), all peripheral interrupts pending bits, the RTC Alarm
(Alarm A and Alarm B), RTC wakeup, RTC tamper, and RTC time
stamp flags, must be reset. Otherwise, the Stop mode entry
procedure is ignored and program execution continues.

5.3.6

Mode exit

If WFI or Return from ISR was used for entry:
All EXTI lines configured in Interrupt mode (the corresponding EXTI
Interrupt vector must be enabled in the NVIC). The interrupt source can
be external interrupts or peripherals with wakeup capability. Refer to
Table 61: Vector table for STM32F405xx/07xx and STM32F415xx/17xx
on page 372.
If WFE was used for entry and SEVONPEND = 0:
All EXTI Lines configured in event mode. Refer to Section 12.2.3:
Wakeup event management on page 380
If WFE was used for entry and SEVONPEND = 1:
– Any EXTI lines configured in Interrupt mode (even if the corresponding
EXTI Interrupt vector is disabled in the NVIC). The interrupt source can
be external interrupts or peripherals with wakeup capability. Refer to
Table 61: Vector table for STM32F405xx/07xx and STM32F415xx/17xx
on page 372 and Table 62: Vector table for STM32F42xxx and
STM32F43xxx.
– Wakeup event: refer to Section 12.2.3: Wakeup event management on
page 380.

Wakeup latency

Refer to Table 28: Stop operating modes (STM32F42xxx and
STM32F43xxx)

Standby mode
The Standby mode allows to achieve the lowest power consumption. It is based on the
Cortex®-M4 with FPU deepsleep mode, with the voltage regulator disabled. The 1.2 V
domain is consequently powered off. The PLLs, the HSI oscillator and the HSE oscillator are
also switched off. SRAM and register contents are lost except for registers in the backup
domain (RTC registers, RTC backup register and backup SRAM), and Standby circuitry (see
Figure 9).

136/1747

RM0090 Rev 17

RM0090

Power controller (PWR)

Entering Standby mode
The Standby mode is entered according to Section : Entering low-power mode, when the
SLEEPDEEP bit in the Cortex®-M4 with FPU System Control register is set.
Refer to Table 30 for more details on how to enter Standby mode.
In Standby mode, the following features can be selected by programming individual control
bits:
•

Independent watchdog (IWDG): the IWDG is started by writing to its Key register or by
hardware option. Once started it cannot be stopped except by a reset. See
Section 21.3 in Section 21: Independent watchdog (IWDG).

•

Real-time clock (RTC): this is configured by the RTCEN bit in the backup domain
control register (RCC_BDCR)

•

Internal RC oscillator (LSI RC): this is configured by the LSION bit in the Control/status
register (RCC_CSR).

•

External 32.768 kHz oscillator (LSE OSC): this is configured by the LSEON bit in the
backup domain control register (RCC_BDCR)

Exiting Standby mode
The Standby mode is exited according to Section : Exiting low-power mode. The SBF status
flag in PWR_CR (see Section 5.4.2: PWR power control/status register (PWR_CSR) for
STM32F405xx/07xx and STM32F415xx/17xx) indicates that the MCU was in Standby
mode. All registers are reset after wakeup from Standby except for PWR_CR.
Refer to Table 30 for more details on how to exit Standby mode.
Table 30. Standby mode entry and exit
Standby mode

Description
WFI (Wait for Interrupt) or WFE (Wait for Event) while:
– SLEEPDEEP is set in Cortex®-M4 with FPU System Control register,
– PDDS bit is set in Power Control register (PWR_CR),
– No interrupt (for WFI) or event (for WFE) is pending,
– WUF bit is cleared in Power Control register (PWR_CR),
– the RTC flag corresponding to the chosen wakeup source (RTC Alarm A,
RTC Alarm B, RTC wakeup, Tamper or Timestamp flags) is cleared

Mode entry

On return from ISR while:
– SLEEPDEEP bit is set in Cortex®-M4 with FPU System Control register,
and
– SLEEPONEXIT = 1, and
– PDDS bit is set in Power Control register (PWR_CR), and
– No interrupt is pending,
– WUF bit is cleared in Power Control/Status register (PWR_SR),
– The RTC flag corresponding to the chosen wakeup source (RTC Alarm
A, RTC Alarm B, RTC wakeup, Tamper or Timestamp flags) is cleared.

Mode exit

WKUP pin rising edge, RTC alarm (Alarm A and Alarm B), RTC wakeup,
tamper event, time stamp event, external reset in NRST pin, IWDG reset.

Wakeup latency

Reset phase.

RM0090 Rev 17

137/1747
149

Power controller (PWR)

RM0090

I/O states in Standby mode
In Standby mode, all I/O pins are high impedance except for:
•

Reset pad (still available)

•

RTC_AF1 pin (PC13) if configured for tamper, time stamp, RTC Alarm out, or RTC
clock calibration out

•

WKUP pin (PA0), if enabled

Debug mode
By default, the debug connection is lost if the application puts the MCU in Stop or Standby
mode while the debug features are used. This is due to the fact that the Cortex®-M4 with
FPU core is no longer clocked.
However, by setting some configuration bits in the DBGMCU_CR register, the software can
be debugged even when using the low-power modes extensively. For more details, refer to
Section 38.16.1: Debug support for low-power modes.

5.3.7

Programming the RTC alternate functions to wake up the device from
the Stop and Standby modes
The MCU can be woken up from a low-power mode by an RTC alternate function.
The RTC alternate functions are the RTC alarms (Alarm A and Alarm B), RTC wakeup, RTC
tamper event detection and RTC time stamp event detection.
These RTC alternate functions can wake up the system from the Stop and Standby lowpower modes.
The system can also wake up from low-power modes without depending on an external
interrupt (Auto-wakeup mode), by using the RTC alarm or the RTC wakeup events.
The RTC provides a programmable time base for waking up from the Stop or Standby mode
at regular intervals.
For this purpose, two of the three alternate RTC clock sources can be selected by
programming the RTCSEL[1:0] bits in the Section 7.3.20: RCC Backup domain control
register (RCC_BDCR):

138/1747

•

Low-power 32.768 kHz external crystal oscillator (LSE OSC)
This clock source provides a precise time base with a very low-power consumption
(additional consumption of less than 1 µA under typical conditions)

•

Low-power internal RC oscillator (LSI RC)
This clock source has the advantage of saving the cost of the 32.768 kHz crystal. This
internal RC oscillator is designed to use minimum power.

RM0090 Rev 17

RM0090

Power controller (PWR)

RTC alternate functions to wake up the device from the Stop mode
•

•

•

To wake up the device from the Stop mode with an RTC alarm event, it is necessary to:
a)

Configure the EXTI Line 17 to be sensitive to rising edges (Interrupt or Event
modes)

b)

Enable the RTC Alarm Interrupt in the RTC_CR register

c)

Configure the RTC to generate the RTC alarm

To wake up the device from the Stop mode with an RTC tamper or time stamp event, it
is necessary to:
a)

Configure the EXTI Line 21 to be sensitive to rising edges (Interrupt or Event
modes)

b)

Enable the RTC time stamp Interrupt in the RTC_CR register or the RTC tamper
interrupt in the RTC_TAFCR register

c)

Configure the RTC to detect the tamper or time stamp event

To wake up the device from the Stop mode with an RTC wakeup event, it is necessary
to:
a)

Configure the EXTI Line 22 to be sensitive to rising edges (Interrupt or Event
modes)

b)

Enable the RTC wakeup interrupt in the RTC_CR register

c)

Configure the RTC to generate the RTC Wakeup event

RTC alternate functions to wake up the device from the Standby mode
•

•

•

To wake up the device from the Standby mode with an RTC alarm event, it is necessary
to:
a)

Enable the RTC alarm interrupt in the RTC_CR register

b)

Configure the RTC to generate the RTC alarm

To wake up the device from the Standby mode with an RTC tamper or time stamp
event, it is necessary to:
a)

Enable the RTC time stamp interrupt in the RTC_CR register or the RTC tamper
interrupt in the RTC_TAFCR register

b)

Configure the RTC to detect the tamper or time stamp event

To wake up the device from the Standby mode with an RTC wakeup event, it is
necessary to:
a)

Enable the RTC wakeup interrupt in the RTC_CR register

b)

Configure the RTC to generate the RTC wakeup event

Safe RTC alternate function wakeup flag clearing sequence
If the selected RTC alternate function is set before the PWR wakeup flag (WUTF) is cleared,
it will not be detected on the next event as detection is made once on the rising edge.
To avoid bouncing on the pins onto which the RTC alternate functions are mapped, and exit
correctly from the Stop and Standby modes, it is recommended to follow the sequence
below before entering the Standby mode:
•

When using RTC alarm to wake up the device from the low-power modes:
a)

Disable the RTC alarm interrupt (ALRAIE or ALRBIE bits in the RTC_CR register)

b)

Clear the RTC alarm (ALRAF/ALRBF) flag

RM0090 Rev 17

139/1747
149

Power controller (PWR)

•

•

c)

Clear the PWR Wakeup (WUF) flag

d)

Enable the RTC alarm interrupt

e)

Re-enter the low-power mode

When using RTC wakeup to wake up the device from the low-power modes:
a)

Disable the RTC Wakeup interrupt (WUTIE bit in the RTC_CR register)

b)

Clear the RTC Wakeup (WUTF) flag

c)

Clear the PWR Wakeup (WUF) flag

d)

Enable the RTC Wakeup interrupt

e)

Re-enter the low-power mode

When using RTC tamper to wake up the device from the low-power modes:
a)

•

140/1747

RM0090

Disable the RTC tamper interrupt (TAMPIE bit in the RTC_TAFCR register)

b)

Clear the Tamper (TAMP1F/TSF) flag

c)

Clear the PWR Wakeup (WUF) flag

d)

Enable the RTC tamper interrupt

e)

Re-enter the low-power mode

When using RTC time stamp to wake up the device from the low-power modes:
a)

Disable the RTC time stamp interrupt (TSIE bit in RTC_CR)

b)

Clear the RTC time stamp (TSF) flag

c)

Clear the PWR Wakeup (WUF) flag

d)

Enable the RTC TimeStamp interrupt

e)

Re-enter the low-power mode

RM0090 Rev 17

RM0090

Power controller (PWR)

5.4

Power control registers (STM32F405xx/07xx and
STM32F415xx/17xx)

5.4.1

PWR power control register (PWR_CR)
for STM32F405xx/07xx and STM32F415xx/17xx
Address offset: 0x00
Reset value: 0x0000 4000 (reset by wakeup from Standby mode)

31

30

29

28

27

26

25

24

23

22

21

6

5

20

19

18

17

16

Reserved
15
Res.

14
VOS
rw

13

12

11

Reserved

10

9

8

FPDS

DBP

rw

rw

7

PLS[2:0]
rw

rw

rw

4

3

2

1

0

PVDE

CSBF

CWUF

PDDS

LPDS

rw

w

w

rw

rw

Bits 31:15 Reserved, must be kept at reset value.
Bit 14 VOS: Regulator voltage scaling output selection
This bit controls the main internal voltage regulator output voltage to achieve a trade-off
between performance and power consumption when the device does not operate at the
maximum frequency.
0: Scale 2 mode
1: Scale 1 mode (default value at reset)
Bits 13:10 Reserved, must be kept at reset value.
Bit 9 FPDS: Flash power-down in Stop mode
When set, the Flash memory enters power-down mode when the device enters Stop mode.
This allows to achieve a lower consumption in stop mode but a longer restart time.
0: Flash memory not in power-down when the device is in Stop mode
1: Flash memory in power-down when the device is in Stop mode
Bit 8 DBP: Disable backup domain write protection
In reset state, the RCC_BDCR register, the RTC registers (including the backup registers),
and the BRE bit of the PWR_CSR register, are protected against parasitic write access. This
bit must be set to enable write access to these registers.
0: Access to RTC and RTC Backup registers and backup SRAM disabled
1: Access to RTC and RTC Backup registers and backup SRAM enabled
Bits 7:5 PLS[2:0]: PVD level selection
These bits are written by software to select the voltage threshold detected by the Power
Voltage Detector
000: 2.0 V
001: 2.1 V
010: 2.3 V
011: 2.5 V
100: 2.6 V
101: 2.7 V
110: 2.8 V
111: 2.9 V
Note: Refer to the electrical characteristics of the datasheet for more details.

RM0090 Rev 17

141/1747
149

Power controller (PWR)

RM0090

Bit 4 PVDE: Power voltage detector enable
This bit is set and cleared by software.
0: PVD disabled
1: PVD enabled
Bit 3 CSBF: Clear standby flag
This bit is always read as 0.
0: No effect
1: Clear the SBF Standby Flag (write).
Bit 2 CWUF: Clear wakeup flag
This bit is always read as 0.
0: No effect
1: Clear the WUF Wakeup Flag after 2 System clock cycles
Bit 1 PDDS: Power-down deepsleep
This bit is set and cleared by software. It works together with the LPDS bit.
0: Enter Stop mode when the CPU enters deepsleep. The regulator status depends on the
LPDS bit.
1: Enter Standby mode when the CPU enters deepsleep.
Bit 0 LPDS: Low-power deepsleep
This bit is set and cleared by software. It works together with the PDDS bit.
0: Voltage regulator on during Stop mode
1: Voltage regulator in low-power mode during Stop mode

5.4.2

PWR power control/status register (PWR_CSR)
for STM32F405xx/07xx and STM32F415xx/17xx
Address offset: 0x04
Reset value: 0x0000 0000 (not reset by wakeup from Standby mode)
Additional APB cycles are needed to read this register versus a standard APB read.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

BRR

PVDO

SBF

WUF

r

r

r

r

Reserved
15

14

Res

VOS
RDY
r

13

12

11

Reserved

10

9

8

BRE

EWUP

rw

rw

7

Reserved

Bits 31:15 Reserved, must be kept at reset value.
Bit 14 VOSRDY: Regulator voltage scaling output selection ready bit
0: Not ready
1: Ready

142/1747

RM0090 Rev 17

RM0090

Power controller (PWR)

Bits 13:10 Reserved, must be kept at reset value.
Bit 9 BRE: Backup regulator enable
When set, the Backup regulator (used to maintain backup SRAM content in Standby and
VBAT modes) is enabled. If BRE is reset, the backup regulator is switched off. The backup
SRAM can still be used but its content will be lost in the Standby and VBAT modes. Once set,
the application must wait that the Backup Regulator Ready flag (BRR) is set to indicate that
the data written into the RAM will be maintained in the Standby and VBAT modes.
0: Backup regulator disabled
1: Backup regulator enabled
Note: This bit is not reset when the device wakes up from Standby mode, by a system reset,
or by a power reset.
Bit 8 EWUP: Enable WKUP pin
This bit is set and cleared by software.
0: WKUP pin is used for general purpose I/O. An event on the WKUP pin does not wakeup
the device from Standby mode.
1: WKUP pin is used for wakeup from Standby mode and forced in input pull down
configuration (rising edge on WKUP pin wakes-up the system from Standby mode).
Note: This bit is reset by a system reset.
Bits 7:4 Reserved, must be kept at reset value.
Bit 3 BRR: Backup regulator ready
Set by hardware to indicate that the Backup Regulator is ready.
0: Backup Regulator not ready
1: Backup Regulator ready
Note: This bit is not reset when the device wakes up from Standby mode or by a system reset
or power reset.
Bit 2 PVDO: PVD output
This bit is set and cleared by hardware. It is valid only if PVD is enabled by the PVDE bit.
0: VDD is higher than the PVD threshold selected with the PLS[2:0] bits.
1: VDD is lower than the PVD threshold selected with the PLS[2:0] bits.
Note: The PVD is stopped by Standby mode. For this reason, this bit is equal to 0 after
Standby or reset until the PVDE bit is set.
Bit 1 SBF: Standby flag
This bit is set by hardware and cleared only by a POR/PDR (power-on reset/power-down
reset) or by setting the CSBF bit in the PWR_CR register.
0: Device has not been in Standby mode
1: Device has been in Standby mode
Bit 0 WUF: Wakeup flag
This bit is set by hardware and cleared either by a system reset or by setting the CWUF bit in
the PWR_CR register.
0: No wakeup event occurred
1: A wakeup event was received from the WKUP pin or from the RTC alarm (Alarm A or
Alarm B), RTC Tamper event, RTC TimeStamp event or RTC Wakeup).
Note: An additional wakeup event is detected if the WKUP pin is enabled (by setting the
EWUP bit) when the WKUP pin level is already high.

RM0090 Rev 17

143/1747
149

Power controller (PWR)

RM0090

5.5

Power control registers (STM32F42xxx and STM32F43xxx)

5.5.1

PWR power control register (PWR_CR)
for STM32F42xxx and STM32F43xxx
Address offset: 0x00
Reset value: 0x0000 C000 (reset by wakeup from Standby mode)

31

30

29

28

27

26

25

24

23

22

21

20

14

VOS[1:0]
rw

13
ADCDC1

rw

rw

12
Res.

18

UDEN[1:0]

Reserved
15

19

11

10

9

8

MRUDS

LPUDS

FPDS

DBP

rw

rw

rw

rw

7

6

5

PLS[2:0]
rw

rw

rw

17

16

ODSWE
ODEN
N

rw

rw

rw

rw

3

2

1

0

PVDE

CSBF

CWUF

PDDS

LPDS

rw

rc_w1

rc_w1

rw

rw

4

Bits 31:20 Reserved, must be kept at reset value.
Bits 19:18 UDEN[1:0]: Under-drive enable in stop mode
These bits are set by software. They allow to achieve a lower power consumption in Stop
mode but with a longer wakeup time.
When set, the digital area has less leakage consumption when the device enters Stop mode.
00: Under-drive disable
01: Reserved
10: Reserved
11:Under-drive enable
Bit 17 ODSWEN: Over-drive switching enabled.
This bit is set by software. It is cleared automatically by hardware after exiting from Stop
mode or when the ODEN bit is reset. When set, It is used to switch to Over-drive mode.
To set or reset the ODSWEN bit, the HSI or HSE must be selected as system clock.
The ODSWEN bit must only be set when the ODRDY flag is set to switch to Over-drive
mode.
0: Over-drive switching disabled
1: Over-drive switching enabled
Note: On any over-drive switch (enabled or disabled), the system clock will be stalled during
the internal voltage set up.
Bit 16 ODEN: Over-drive enable
This bit is set by software. It is cleared automatically by hardware after exiting from Stop
mode. It is used to enabled the Over-drive mode in order to reach a higher frequency.
To set or reset the ODEN bit, the HSI or HSE must be selected as system clock. When the
ODEN bit is set, the application must first wait for the Over-drive ready flag (ODRDY) to be
set before setting the ODSWEN bit.
0: Over-drive disabled
1: Over-drive enabled

144/1747

RM0090 Rev 17

RM0090

Power controller (PWR)

Bits 15:14 VOS[1:0]: Regulator voltage scaling output selection
These bits control the main internal voltage regulator output voltage to achieve a trade-off
between performance and power consumption when the device does not operate at the
maximum frequency (refer to the STM32F42xx and STM32F43xx datasheets for more
details).
These bits can be modified only when the PLL is OFF. The new value programmed is active
only when the PLL is ON. When the PLL is OFF, the voltage scale 3 is automatically
selected.
00: Reserved (Scale 3 mode selected)
01: Scale 3 mode
10: Scale 2 mode
11: Scale 1 mode (reset value)
Bit 13 ADCDC1:
0: No effect.
1: Refer to AN4073 for details on how to use this bit.
Note: This bit can only be set when operating at supply voltage range 2.7 to 3.6V and when
the Prefetch is OFF.
Bit 12 Reserved, must be kept at reset value.
Bit 11 MRUDS: Main regulator in deepsleep under-drive mode
This bit is set and cleared by software.
0: Main regulator ON when the device is in Stop mode
1: Main Regulator in under-drive mode and Flash memory in power-down when the device is
in Stop under-drive mode.
Bit 10 LPUDS: Low-power regulator in deepsleep under-drive mode
This bit is set and cleared by software.
0: Low-power regulator ON if LPDS bit is set when the device is in Stop mode
1: Low-power regulator in under-drive mode if LPDS bit is set and Flash memory in powerdown when the device is in Stop under-drive mode.
Bit 9 FPDS: Flash power-down in Stop mode
When set, the Flash memory enters power-down mode when the device enters Stop mode.
This allows to achieve a lower consumption in stop mode but a longer restart time.
0: Flash memory not in power-down when the device is in Stop mode
1: Flash memory in power-down when the device is in Stop mode
Bit 8 DBP: Disable backup domain write protection
In reset state, the RCC_BDCR register, the RTC registers (including the backup registers),
and the BRE bit of the PWR_CSR register, are protected against parasitic write access. This
bit must be set to enable write access to these registers.
0: Access to RTC and RTC Backup registers and backup SRAM disabled
1: Access to RTC and RTC Backup registers and backup SRAM enabled

RM0090 Rev 17

145/1747
149

Power controller (PWR)

RM0090

Bits 7:5 PLS[2:0]: PVD level selection
These bits are written by software to select the voltage threshold detected by the Power
Voltage Detector
000: 2.0 V
001: 2.1 V
010: 2.3 V
011: 2.5 V
100: 2.6 V
101: 2.7 V
110: 2.8 V
111: 2.9 V
Note: Refer to the electrical characteristics of the datasheet for more details.
Bit 4 PVDE: Power voltage detector enable
This bit is set and cleared by software.
0: PVD disabled
1: PVD enabled
Bit 3 CSBF: Clear standby flag
This bit is always read as 0.
0: No effect
1: Clear the SBF Standby Flag (write).
Bit 2 CWUF: Clear wakeup flag
This bit is always read as 0.
0: No effect
1: Clear the WUF Wakeup Flag after 2 System clock cycles
Bit 1 PDDS: Power-down deepsleep
This bit is set and cleared by software. It works together with the LPDS bit.
0: Enter Stop mode when the CPU enters deepsleep. The regulator status depends on the
LPDS bit.
1: Enter Standby mode when the CPU enters deepsleep.
Bit 0 LPDS: Low-power deepsleep
This bit is set and cleared by software. It works together with the PDDS bit.
0:Main voltage regulator ON during Stop mode
1: Low-power voltage regulator ON during Stop mode

146/1747

RM0090 Rev 17

RM0090

Power controller (PWR)

5.5.2

PWR power control/status register (PWR_CSR)
for STM32F42xxx and STM32F43xxx
Address offset: 0x04
Reset value: 0x0000 0000 (not reset by wakeup from Standby mode)
Additional APB cycles are needed to read this register versus a standard APB read.

31

30

29

28

27

26

25

24

23

22

21

20

14

Res

VOS
RDY
r

13

12

11

Reserved

10

18

UDRDY[1:0]

Reserved
15

19

9

8

BRE

EWUP

rw

rw

7

6

5

4

Reserved.

17

16

ODSWRDY ODRDY

rc_w1

rc_w1

r

r

3

2

1

0

BRR

PVDO

SBF

WUF

r

r

r

r

Bits 31:20 Reserved, must be kept at reset value.
Bits 19:18 UDRDY[1:0]: Under-drive ready flag
These bits are set by hardware when MCU entered stop Under-drive mode and exited.
When the under-drive mode is enabled, these bits are not set as long as the MCU has not
entered stop mode yet. They are cleared by programming them to 1.
00: Under-drive is disabled
01: Reserved
10: Reserved
11:Under-drive mode is activated in Stop mode.
Bit 17 ODSWRDY: Over-drive mode switching ready
0: Over-drive mode is not active.
1: Over-drive mode is active on digital area on 1.2 V domain
Bit 16 ODRDY: Over-drive mode ready
0: Over-drive mode not ready.
1: Over-drive mode ready
Bit 14 VOSRDY: Regulator voltage scaling output selection ready bit
0: Not ready
1: Ready
Bits 13:10 Reserved, must be kept at reset value.
Bit 9 BRE: Backup regulator enable
When set, the Backup regulator (used to maintain backup SRAM content in Standby and
VBAT modes) is enabled. If BRE is reset, the backup regulator is switched off. The backup
SRAM can still be used but its content will be lost in the Standby and VBAT modes. Once set,
the application must wait that the Backup Regulator Ready flag (BRR) is set to indicate that
the data written into the RAM will be maintained in the Standby and VBAT modes.
0: Backup regulator disabled
1: Backup regulator enabled
Note: This bit is not reset when the device wakes up from Standby mode, by a system reset,
or by a power reset.

RM0090 Rev 17

147/1747
149

Power controller (PWR)

RM0090

Bit 8 EWUP: Enable WKUP pin
This bit is set and cleared by software.
0: WKUP pin is used for general purpose I/O. An event on the WKUP pin does not wakeup
the device from Standby mode.
1: WKUP pin is used for wakeup from Standby mode and forced in input pull down
configuration (rising edge on WKUP pin wakes-up the system from Standby mode).
Note: This bit is reset by a system reset.
Bits 7:4 Reserved, must be kept at reset value.
Bit 3 BRR: Backup regulator ready
Set by hardware to indicate that the Backup Regulator is ready.
0: Backup Regulator not ready
1: Backup Regulator ready
Note: This bit is not reset when the device wakes up from Standby mode or by a system reset
or power reset.
Bit 2 PVDO: PVD output
This bit is set and cleared by hardware. It is valid only if PVD is enabled by the PVDE bit.
0: VDD is higher than the PVD threshold selected with the PLS[2:0] bits.
1: VDD is lower than the PVD threshold selected with the PLS[2:0] bits.
Note: The PVD is stopped by Standby mode. For this reason, this bit is equal to 0 after
Standby or reset until the PVDE bit is set.
Bit 1 SBF: Standby flag
This bit is set by hardware and cleared only by a POR/PDR (power-on reset/power-down
reset) or by setting the CSBF bit in the PWR power control register (PWR_CR) for
STM32F405xx/07xx and STM32F415xx/17xx
0: Device has not been in Standby mode
1: Device has been in Standby mode
Bit 0 WUF: Wakeup flag
This bit is set by hardware and cleared either by a system reset or by setting the CWUF bit in
the PWR_CR register.
0: No wakeup event occurred
1: A wakeup event was received from the WKUP pin or from the RTC alarm (Alarm A or
Alarm B), RTC Tamper event, RTC TimeStamp event or RTC Wakeup).
Note: An additional wakeup event is detected if the WKUP pin is enabled (by setting the
EWUP bit) when the WKUP pin level is already high.

148/1747

RM0090 Rev 17

RM0090

5.6

Power controller (PWR)

PWR register map
The following table summarizes the PWR registers.

PDDS

LPDS

0

0

0

0

0

SBF

WUF

0

CSBF

0

0

CWUF

0

0

PVDO

Reset value

Reserved

0

PVDE

Reserved

0

PLS[2:0]

BRR

PWR_CSR

DBP

0x004

1
VOSRDY

Reset value

Reserved

FPDS

Reserved

BRE

PWR_CR

EWUP

0x000

Register

VOS

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 31. PWR - register map and reset values for STM32F405xx/07xx and STM32F415xx/17xx

0

0

0

0

0

Reserved

0

LPDS

0

0

PDDS

Reserved

0

0

0

0

0

0

SBF

DBP
0

WUF

FPDS
0

CWUF

LPUDS
1

PVDO

MRUDS
1

CSBF

0

0

PVDE

VOSRDY

1

Reserved

0

ADCDC1

0

VOS[1:0]
1

PLS[2:0]

BRR

0

1

EWUP

0

1

BRE

Reset value

Reserved

ODEN

PWR_CSR

1

Reserved

0x004

1

ODSWEN

Reset value

ODRDY

Reserved

ODSWRDY

PWR_CR

UDEN[1:0]

0x000

Register

UDRDY[1:0]

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 32. PWR - register map and reset values for STM32F42xxx and STM32F43xxx

0

0

0

0

0

Reserved

Refer to Section 2.3: Memory mapfor the register boundary addresses.

RM0090 Rev 17

149/1747
149

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

6

Reset and clock control for
STM32F42xxx and STM32F43xxx (RCC)

6.1

Reset

RM0090

There are three types of reset, defined as system Reset, power Reset and backup domain
Reset.

6.1.1

System reset
A system reset sets all registers to their reset values except the reset flags in the clock
controller CSR register and the registers in the Backup domain (see Figure 4).
A system reset is generated when one of the following events occurs:
1.

A low level on the NRST pin (external reset)

2.

Window watchdog end of count condition (WWDG reset)

3.

Independent watchdog end of count condition (IWDG reset)

4.

A software reset (SW reset) (see Software reset)

5.

Low-power management reset (see Low-power management reset)

Software reset
The reset source can be identified by checking the reset flags in the RCC clock control &
status register (RCC_CSR).
The SYSRESETREQ bit in Cortex®-M4 with FPU Application Interrupt and Reset Control
Register must be set to force a software reset on the device. Refer to the Cortex®-M4 with
FPU technical reference manual for more details.

Low-power management reset
There are two ways of generating a low-power management reset:
1.

Reset generated when entering the Standby mode:
This type of reset is enabled by resetting the nRST_STDBY bit in the user option bytes.
In this case, whenever a Standby mode entry sequence is successfully executed, the
device is reset instead of entering the Standby mode.

2.

Reset when entering the Stop mode:
This type of reset is enabled by resetting the nRST_STOP bit in the user option bytes.
In this case, whenever a Stop mode entry sequence is successfully executed, the
device is reset instead of entering the Stop mode.

6.1.2

Power reset
A power reset is generated when one of the following events occurs:
1.

Power-on/power-down reset (POR/PDR reset) or brownout (BOR) reset

2.

When exiting the Standby mode

A power reset sets all registers to their reset values except the Backup domain (see
Figure 4)

150/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)
These sources act on the NRST pin and it is always kept low during the delay phase. The
RESET service routine vector is fixed at address 0x0000_0004 in the memory map.
The system reset signal provided to the device is output on the NRST pin. The pulse
generator guarantees a minimum reset pulse duration of 20 µs for each internal reset
source. In case of an external reset, the reset pulse is generated while the NRST pin is
asserted low.
Figure 15. Simplified diagram of the reset circuit
9''9''$

538
([WHUQDO
UHVHW

6\VWHPUHVHW

)LOWHU

1567
3XOVH
JHQHUDWRU
PLQ—V

::'*UHVHW
,:'*UHVHW
3RZHUUHVHW
6RIWZDUHUHVHW
/RZSRZHUPDQDJHPHQWUHVHW
DLF

The Backup domain has two specific resets that affect only the Backup domain (see
Figure 4).

6.1.3

Backup domain reset
The backup domain reset sets all RTC registers and the RCC_BDCR register to their reset
values. The BKPSRAM is not affected by this reset. The only way of resetting the
BKPSRAM is through the Flash interface by requesting a protection level change from 1 to
0.
A backup domain reset is generated when one of the following events occurs:

6.2

1.

Software reset, triggered by setting the BDRST bit in the RCC Backup domain control
register (RCC_BDCR).

2.

VDD or VBAT power on, if both supplies have previously been powered off.

Clocks
Three different clock sources can be used to drive the system clock (SYSCLK):
•

HSI oscillator clock

•

HSE oscillator clock

•

Main PLL (PLL) clock

The devices have the two following secondary clock sources:
•

32 kHz low-speed internal RC (LSI RC) which drives the independent watchdog and,
optionally, the RTC used for Auto-wakeup from the Stop/Standby mode.

•

32.768 kHz low-speed external crystal (LSE crystal) which optionally drives the RTC
clock (RTCCLK)

RM0090 Rev 17

151/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

RM0090

Each clock source can be switched on or off independently when it is not used, to optimize
power consumption.
Figure 16. Clock tree
/6,5&
N+]

/6,

:DWFKGRJ
HQDEOH

,:'*&/.

7R,QGHSHQGHQW
ZDWFKGRJ

57&6(/>@
26&B,1
26&B287
0&2

57&&/.

WR57&

6<6&/.
WR
+6(B57&

/6(
0&2

57&
HQDEOH

/6(

/6(26&
N+]

WR

3HULSKHUDO
FORFNHQDEOH

WR

(WKHUQHW
373FORFN
+&/.
WR$+%EXVFRUH
PHPRU\DQG'0$

0+]PD[
0+]
+6,5&

&ORFN
HQDEOH

+6,

+6,
+6(
3//&/.

26&B287
26&B,1

WR&RUWH[6\VWHPWLPHU
)&/.&RUWH[
IUHHUXQQLQJFORFN

$+%
35(6&


6<6&/.
0+]
PD[

+6(

0+]
+6(26&



6:

3HULSKHUDOFORFN
HQDEOH

$3%[
35(6&


3HULSKHUDOFORFN
LI $3%[SUHVF  HQDEOH
[HOVH[

$3%[SHULSKHUDOFORFNV

$3%[WLPHUFORFNV

0
9&2
[1

3//

9&2
[1

3//,6

3

3HULSKHUDO
FORFNHQDEOH

3//&.

4
5

3

,665&

4

([WFORFN
3//6$,&/.

3//6$,

4
5

3///&'&/.

(7+B0,,B7;B&/.B0,,
3+<(WKHUQHW
WR0+]


(7+B0,,B5;B&/.B0,
,

86%3+<
WR0+]

',9

3

3//6$,&/.
[1

3HULSKHUDO
FORFNHQDEOH

3//,6&/.

5

,6B&.,1

9&2

0+]
FORFNV

',9

3HULSKHUDO
FORFNHQDEOH

6$,B$FORFN

3HULSKHUDO
FORFNHQDEOH

6$,B%FORFN

3HULSKHUDO
FORFNHQDEOH

',9
3HULSKHUDO
FORFNHQDEOH
3HULSKHUDO
FORFNHQDEOH
3HULSKHUDO
FORFNHQDEOH

,6FORFNV

/&'7)7
FORFN

0$&7;&/.
0,,B50,,B6(/LQ6<6&)*B30&WR(WKHUQHW0$&
0$&5;&/.
0$&50,,&/.
3HULSKHUDO
FORFNHQDEOH

27*B+6B8/3,B&.

86%+6
8/3,FORFN
069

1. For full details about the internal and external clock source characteristics, refer to the Electrical characteristics section in
the device datasheet.
2. When TIMPRE bit of the RCC_DCKCFGR register is reset, if APBx prescaler is 1, then TIMxCLK = PCLKx, otherwise
TIMxCLK = 2x PCLKx.
3. When TIMPRE bit in the RCC_DCKCFGR register is set, if APBx prescaler is 1,2 or 4, then TIMxCLK = HCLK, otherwise
TIMxCLK = 4x PCLKx.

152/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)
The clock controller provides a high degree of flexibility to the application in the choice of the
external crystal or the oscillator to run the core and peripherals at the highest frequency
and, guarantee the appropriate frequency for peripherals that need a specific clock like
Ethernet, USB OTG FS and HS, I2S, SAI, LTDC, and SDIO.
Several prescalers are used to configure the AHB frequency, the high-speed APB (APB2)
and the low-speed APB (APB1) domains. The maximum frequency of the AHB domain is
180 MHz. The maximum allowed frequency of the high-speed APB2 domain is 90 MHz. The
maximum allowed frequency of the low-speed APB1 domain is 45 MHz
All peripheral clocks are derived from the system clock (SYSCLK) except for:
•

The USB OTG FS clock (48 MHz), the random analog generator (RNG) clock
(≤48 MHz) and the SDIO clock (≤48 MHz) which are coming from a specific output of
PLL (PLL48CLK)

•

I2S clock
To achieve high-quality audio performance, the I2S clock can be derived either from a
specific PLL (PLLI2S) or from an external clock mapped on the I2S_CKIN pin. For
more information about I2S clock frequency and precision, refer to Section 28.4.4:
Clock generator.

•

SAI1 clock
The SAI1 clock is generated from a specific PLL (PLLSAI or PLLI2S) or from an
external clock mapped on the I2S_CKIN pin.
The PLLSAI can be used as clock source for SAI1 peripheral in case the PLLI2S is
programmed to achieve another audio sampling frequency (49.152 MHz or
11.2896 MHz), and the application requires both frequencies at the same time.

•

LTDC clock
The LTDC clock is generated from a specific PLL (PLLSAI).

•

The USB OTG HS (60 MHz) clock which is provided from the external PHY

•

The Ethernet MAC clocks (TX, RX and RMII) which are provided from the external
PHY. For further information on the Ethernet configuration, please refer to
Section 33.4.4: MII/RMII selection in the Ethernet peripheral description. When the
Ethernet is used, the AHB clock frequency must be at least 25 MHz.

The RCC feeds the external clock of the Cortex System Timer (SysTick) with the AHB clock
(HCLK) divided by 8. The SysTick can work either with this clock or with the Cortex clock
(HCLK), configurable in the SysTick control and status register.
The timer clock frequencies for STM32F42xxx and STM32F43xxx are automatically set by
hardware. There are two cases depending on the value of TIMPRE bit in RCC_CFGR
register:
•

If TIMPRE bit in RCC_DKCFGR register is reset:
If the APB prescaler is configured to a division factor of 1, the timer clock frequencies
(TIMxCLK) are set to PCLKx. Otherwise, the timer clock frequencies are twice the
frequency of the APB domain to which the timers are connected: TIMxCLK = 2xPCLKx.

•

If TIMPRE bit in RCC_DKCFGR register is set:
If the APB prescaler is configured to a division factor of 1, 2 or 4, the timer clock
frequencies (TIMxCLK) are set to HCLK. Otherwise, the timer clock frequencies is four
times the frequency of the APB domain to which the timers are connected: TIMxCLK =
4xPCLKx.

RM0090 Rev 17

153/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

RM0090

FCLK acts as Cortex®-M4 with FPU free-running clock. For more details, refer to the
Cortex®-M4 with FPU technical reference manual.

6.2.1

HSE clock
The high speed external clock signal (HSE) can be generated from two possible clock
sources:
•

HSE external crystal/ceramic resonator

•

HSE external user clock

The resonator and the load capacitors have to be placed as close as possible to the
oscillator pins in order to minimize output distortion and startup stabilization time. The
loading capacitance values must be adjusted according to the selected oscillator.
Figure 17. HSE/ LSE clock sources
Hardware configuration

OSC_OUT
External clock

(HI-Z)
External
source

OSC_IN OSC_OUT
Crystal/ceramic
resonators
CL1

Load
capacitors

CL2

External source (HSE bypass)
In this mode, an external clock source must be provided. You select this mode by setting the
HSEBYP and HSEON bits in the RCC clock control register (RCC_CR). The external clock
signal (square, sinus or triangle) with ~50% duty cycle has to drive the OSC_IN pin while the
OSC_OUT pin should be left HI-Z. See Figure 17.

External crystal/ceramic resonator (HSE crystal)
The HSE has the advantage of producing a very accurate rate on the main clock.
The associated hardware configuration is shown in Figure 17. Refer to the electrical
characteristics section of the datasheet for more details.
The HSERDY flag in the RCC clock control register (RCC_CR) indicates if the high-speed
external oscillator is stable or not. At startup, the clock is not released until this bit is set by
hardware. An interrupt can be generated if enabled in the RCC clock interrupt register
(RCC_CIR).

154/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)
The HSE Crystal can be switched on and off using the HSEON bit in the RCC clock control
register (RCC_CR).

6.2.2

HSI clock
The HSI clock signal is generated from an internal 16 MHz RC oscillator and can be used
directly as a system clock, or used as PLL input.
The HSI RC oscillator has the advantage of providing a clock source at low cost (no external
components). It also has a faster startup time than the HSE crystal oscillator however, even
with calibration the frequency is less accurate than an external crystal oscillator or ceramic
resonator.

Calibration
RC oscillator frequencies can vary from one chip to another due to manufacturing process
variations, this is why each device is factory calibrated by ST for 1% accuracy at TA= 25 °C.
After reset, the factory calibration value is loaded in the HSICAL[7:0] bits in the RCC clock
control register (RCC_CR).
If the application is subject to voltage or temperature variations this may affect the RC
oscillator speed. You can trim the HSI frequency in the application using the HSITRIM[4:0]
bits in the RCC clock control register (RCC_CR).
The HSIRDY flag in the RCC clock control register (RCC_CR) indicates if the HSI RC is
stable or not. At startup, the HSI RC output clock is not released until this bit is set by
hardware.
The HSI RC can be switched on and off using the HSION bit in the RCC clock control
register (RCC_CR).
The HSI signal can also be used as a backup source (Auxiliary clock) if the HSE crystal
oscillator fails. Refer to Section 6.2.7: Clock security system (CSS) on page 157.

6.2.3

PLL configuration
The STM32F4xx devices feature three PLLs:
•

•

A main PLL (PLL) clocked by the HSE or HSI oscillator and featuring two different
output clocks:
–

The first output is used to generate the high speed system clock (up to 180 MHz)

–

The second output is used to generate the clock for the USB OTG FS (48 MHz),
the random analog generator (≤48 MHz) and the SDIO (≤48 MHz).

Two dedicated PLLs (PLLI2S and PLLSAI) used to generate an accurate clock to
achieve high-quality audio performance on the I2S and SAI1 interfaces. PLLSAI is also
used for the LCD-TFT clock.

Since the main-PLL configuration parameters cannot be changed once PLL is enabled, it is
recommended to configure PLL before enabling it (selection of the HSI or HSE oscillator as
PLL clock source, and configuration of division factors M, N, P, and Q).
The PLLI2S and PLLSAI use the same input clock as PLL (PLLM[5:0] and PLLSRC bits are
common to both PLLs). However, the PLLI2S and PLLSAI have dedicated enable/disable
and division factors (N and R) configuration bits. Once the PLLI2S and PLLSAI are enabled,
the configuration parameters cannot be changed.

RM0090 Rev 17

155/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

RM0090

The three PLLs are disabled by hardware when entering Stop and Standby modes, or when
an HSE failure occurs when HSE or PLL (clocked by HSE) are used as system clock. RCC
PLL configuration register (RCC_PLLCFGR),RCC clock configuration register
(RCC_CFGR), and RCC Dedicated Clock Configuration Register (RCC_DCKCFGR) can be
used to configure PLL, PLLI2S, and PLLSAI.

6.2.4

LSE clock
The LSE clock is generated from a 32.768 kHz low-speed external crystal or ceramic
resonator. It has the advantage providing a low-power but highly accurate clock source to
the real-time clock peripheral (RTC) for clock/calendar or other timing functions.
The LSE oscillator is switched on and off using the LSEON bit in RCC Backup domain
control register (RCC_BDCR).
The LSERDY flag in the RCC Backup domain control register (RCC_BDCR) indicates if the
LSE crystal is stable or not. At startup, the LSE crystal output clock signal is not released
until this bit is set by hardware. An interrupt can be generated if enabled in the RCC clock
interrupt register (RCC_CIR).

External source (LSE bypass)
In this mode, an external clock source must be provided. It must have a frequency up to
1 MHz. You select this mode by setting the LSEBYP and LSEON bits in the RCC Backup
domain control register (RCC_BDCR). The external clock signal (square, sinus or triangle)
with ~50% duty cycle has to drive the OSC32_IN pin while the OSC32_OUT pin should be
left HI-Z. See Figure 17.

6.2.5

LSI clock
The LSI RC acts as an low-power clock source that can be kept running in Stop and
Standby mode for the independent watchdog (IWDG) and Auto-wakeup unit (AWU). The
clock frequency is around 32 kHz. For more details, refer to the electrical characteristics
section of the datasheets.
The LSI RC can be switched on and off using the LSION bit in the RCC clock control &
status register (RCC_CSR).
The LSIRDY flag in the RCC clock control & status register (RCC_CSR) indicates if the lowspeed internal oscillator is stable or not. At startup, the clock is not released until this bit is
set by hardware. An interrupt can be generated if enabled in the RCC clock interrupt register
(RCC_CIR).

6.2.6

System clock (SYSCLK) selection
After a system reset, the HSI oscillator is selected as the system clock. When a clock source
is used directly or through PLL as the system clock, it is not possible to stop it.
A switch from one clock source to another occurs only if the target clock source is ready
(clock stable after startup delay or PLL locked). If a clock source that is not yet ready is
selected, the switch occurs when the clock source is ready. Status bits in the RCC clock
control register (RCC_CR) indicate which clock(s) is (are) ready and which clock is currently
used as the system clock.

156/1747

RM0090 Rev 17

RM0090

6.2.7

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Clock security system (CSS)
The clock security system can be activated by software. In this case, the clock detector is
enabled after the HSE oscillator startup delay, and disabled when this oscillator is stopped.
If a failure is detected on the HSE clock, this oscillator is automatically disabled, a clock
failure event is sent to the break inputs of advanced-control timers TIM1 and TIM8, and an
interrupt is generated to inform the software about the failure (clock security system
interrupt CSSI), allowing the MCU to perform rescue operations. The CSSI is linked to the
Cortex®-M4 with FPU NMI (non-maskable interrupt) exception vector.

Note:

When the CSS is enabled, if the HSE clock happens to fail, the CSS generates an interrupt,
which causes the automatic generation of an NMI. The NMI is executed indefinitely unless
the CSS interrupt pending bit is cleared. As a consequence, the application has to clear the
CSS interrupt in the NMI ISR by setting the CSSC bit in the Clock interrupt register
(RCC_CIR).
If the HSE oscillator is used directly or indirectly as the system clock (indirectly meaning that
it is directly used as PLL input clock, and that PLL clock is the system clock) and a failure is
detected, then the system clock switches to the HSI oscillator and the HSE oscillator is
disabled.
If the HSE oscillator clock was the clock source of PLL used as the system clock when the
failure occurred, PLL is also disabled. In this case, if the PLLI2S was enabled, it is also
disabled when the HSE fails.

6.2.8

RTC/AWU clock
Once the RTCCLK clock source has been selected, the only possible way of modifying the
selection is to reset the power domain.
The RTCCLK clock source can be either the HSE 1 MHz (HSE divided by a programmable
prescaler), the LSE or the LSI clock. This is selected by programming the RTCSEL[1:0] bits
in the RCC Backup domain control register (RCC_BDCR) and the RTCPRE[4:0] bits in RCC
clock configuration register (RCC_CFGR). This selection cannot be modified without
resetting the Backup domain.
If the LSE is selected as the RTC clock, the RTC operates normally if the backup or the
system supply disappears. If the LSI is selected as the AWU clock, the AWU state is not
guaranteed if the system supply disappears. If the HSE oscillator divided by a value
between 2 and 31 is used as the RTC clock, the RTC state is not guaranteed if the backup
or the system supply disappears.
The LSE clock is in the Backup domain, whereas the HSE and LSI clocks are not. As a
consequence:
•

If LSE is selected as the RTC clock:
–

•

If LSI is selected as the Auto-wakeup unit (AWU) clock:
–

•

The RTC continues to work even if the VDD supply is switched off, provided the
VBAT supply is maintained.
The AWU state is not guaranteed if the VDD supply is powered off. Refer to
Section 6.2.5: LSI clock on page 156 for more details on LSI calibration.

If the HSE clock is used as the RTC clock:
–

The RTC state is not guaranteed if the VDD supply is powered off or if the internal
voltage regulator is powered off (removing power from the 1.2 V domain).

RM0090 Rev 17

157/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

RM0090

Note:

To read the RTC calendar register when the APB1 clock frequency is less than seven times
the RTC clock frequency (fAPB1 < 7xfRTCLCK), the software must read the calendar time and
date registers twice. The data are correct if the second read access to RTC_TR gives the
same result than the first one. Otherwise a third read access must be performed.

6.2.9

Watchdog clock
If the independent watchdog (IWDG) is started by either hardware option or software
access, the LSI oscillator is forced ON and cannot be disabled. After the LSI oscillator
temporization, the clock is provided to the IWDG.

6.2.10

Clock-out capability
Two microcontroller clock output (MCO) pins are available:
•

MCO1
You can output four different clock sources onto the MCO1 pin (PA8) using the
configurable prescaler (from 1 to 5):
–

HSI clock

–

LSE clock

–

HSE clock

–

PLL clock

The desired clock source is selected using the MCO1PRE[2:0] and MCO1[1:0] bits in
the RCC clock configuration register (RCC_CFGR).
•

MCO2
You can output four different clock sources onto the MCO2 pin (PC9) using the
configurable prescaler (from 1 to 5):
–

HSE clock

–

PLL clock

–

System clock (SYSCLK)

–

PLLI2S clock

The desired clock source is selected using the MCO2PRE[2:0] and MCO2 bits in the
RCC clock configuration register (RCC_CFGR).
For the different MCO pins, the corresponding GPIO port has to be programmed in alternate
function mode.
The selected clock to output onto MCO must not exceed 100 MHz (the maximum I/O
speed).

6.2.11

Internal/external clock measurement using TIM5/TIM11
It is possible to indirectly measure the frequencies of all on-board clock source generators
by means of the input capture of TIM5 channel4 and TIM11 channel1 as shown in Figure 18
and Figure 19.

Internal/external clock measurement using TIM5 channel4
TIM5 has an input multiplexer which allows choosing whether the input capture is triggered
by the I/O or by an internal clock. This selection is performed through the TI4_RMP [1:0] bits
in the TIM5_OR register.

158/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)
The primary purpose of having the LSE connected to the channel4 input capture is to be
able to precisely measure the HSI (this requires to have the HSI used as the system clock
source). The number of HSI clock counts between consecutive edges of the LSE signal
provides a measurement of the internal clock period. Taking advantage of the high precision
of LSE crystals (typically a few tens of ppm) we can determine the internal clock frequency
with the same resolution, and trim the source to compensate for manufacturing-process
and/or temperature- and voltage-related frequency deviations.
The HSI oscillator has dedicated, user-accessible calibration bits for this purpose.
The basic concept consists in providing a relative measurement (e.g. HSI/LSE ratio): the
precision is therefore tightly linked to the ratio between the two clock sources. The greater
the ratio, the better the measurement.
It is also possible to measure the LSI frequency: this is useful for applications that do not
have a crystal. The ultralow-power LSI oscillator has a large manufacturing process
deviation: by measuring it versus the HSI clock source, it is possible to determine its
frequency with the precision of the HSI. The measured value can be used to have more
accurate RTC time base timeouts (when LSI is used as the RTC clock source) and/or an
IWDG timeout with an acceptable accuracy.
Use the following procedure to measure the LSI frequency:
1.

Enable the TIM5 timer and configure channel4 in Input capture mode.

2.

This bit is set the TI4_RMP bits in the TIM5_OR register to 0x01 to connect the LSI
clock internally to TIM5 channel4 input capture for calibration purposes.

3.

Measure the LSI clock frequency using the TIM5 capture/compare 4 event or interrupt.

4.

Use the measured LSI frequency to update the prescaler of the RTC depending on the
desired time base and/or to compute the IWDG timeout.
Figure 18. Frequency measurement with TIM5 in Input capture mode
4)-

4)?2-0;=
'0)/
24#?7AKE5P?)4
,3%
,3)

4)

AI6

Internal/external clock measurement using TIM11 channel1
TIM11 has an input multiplexer which allows choosing whether the input capture is triggered
by the I/O or by an internal clock. This selection is performed through TI1_RMP [1:0] bits in
the TIM11_OR register. The HSE_RTC clock (HSE divided by a programmable prescaler) is
connected to channel 1 input capture to have a rough indication of the external crystal
frequency. This requires that the HSI is the system clock source. This can be useful for
instance to ensure compliance with the IEC 60730/IEC 61335 standards which require to be
able to determine harmonic or subharmonic frequencies (–50/+100% deviations).

RM0090 Rev 17

159/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

RM0090

Figure 19. Frequency measurement with TIM11 in Input capture mode
4)-

4)?2-0;=
'0)/
4)
(3%?24# -(Z
AI

160/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

6.3

RCC registers
Refer to Section 1.1: List of abbreviations for registers for a list of abbreviations used in
register descriptions.

6.3.1

RCC clock control register (RCC_CR)
Address offset: 0x00
Reset value: 0x0000 XX83 where X is undefined.
Access: no wait state, word, half-word and byte access

31

30

Reserved
15

14

29

28

27

26

25

PLLSAI PLLSAI PLLI2S PLLI2S PLLRD
RDY
ON
RDY
ON
Y

24

23

PLLON

r

rw

r

rw

r

rw

13

12

11

10

9

8

22

r

r

r

r

20

Reserved
7

6

HSICAL[7:0]
r

21

5

4

19

18

17

16

CSS
ON

HSE
BYP

HSE
RDY

HSE
ON

rw

rw

r

rw

3

2

1

0

Res.

HSI
RDY

HSION

r

rw

HSITRIM[4:0]
r

r

r

rw

rw

rw

rw

rw

Bits 31:28 Reserved, must be kept at reset value.
Bit 29 PLLSAIRDY: PLLSAI clock ready flag
Set by hardware to indicate that the PLLSAI is locked.
0: PLLSAI unlocked
1: PLLSAI locked
Bit 28 PLLSAION: PLLSAI enable
Set and cleared by software to enable PLLSAI.
Cleared by hardware when entering Stop or Standby mode.
0: PLLSAI OFF
1: PLLSAI ON
Bit 27 PLLI2SRDY: PLLI2S clock ready flag
Set by hardware to indicate that the PLLI2S is locked.
0: PLLI2S unlocked
1: PLLI2S locked
Bit 26 PLLI2SON: PLLI2S enable
Set and cleared by software to enable PLLI2S.
Cleared by hardware when entering Stop or Standby mode.
0: PLLI2S OFF
1: PLLI2S ON
Bit 25 PLLRDY: Main PLL (PLL) clock ready flag
Set by hardware to indicate that PLL is locked.
0: PLL unlocked
1: PLL locked
Bit 24 PLLON: Main PLL (PLL) enable
Set and cleared by software to enable PLL.
Cleared by hardware when entering Stop or Standby mode. This bit cannot be reset if PLL
clock is used as the system clock.
0: PLL OFF
1: PLL ON

RM0090 Rev 17

161/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

RM0090

Bits 23:20 Reserved, must be kept at reset value.
Bit 19 CSSON: Clock security system enable
Set and cleared by software to enable the clock security system. When CSSON is set, the
clock detector is enabled by hardware when the HSE oscillator is ready, and disabled by
hardware if an oscillator failure is detected.
0: Clock security system OFF (Clock detector OFF)
1: Clock security system ON (Clock detector ON if HSE oscillator is stable, OFF if not)
Bit 18 HSEBYP: HSE clock bypass
Set and cleared by software to bypass the oscillator with an external clock. The external
clock must be enabled with the HSEON bit, to be used by the device.
The HSEBYP bit can be written only if the HSE oscillator is disabled.
0: HSE oscillator not bypassed
1: HSE oscillator bypassed with an external clock
Bit 17 HSERDY: HSE clock ready flag
Set by hardware to indicate that the HSE oscillator is stable. After the HSEON bit is cleared,
HSERDY goes low after 6 HSE oscillator clock cycles.
0: HSE oscillator not ready
1: HSE oscillator ready
Bit 16 HSEON: HSE clock enable
Set and cleared by software.
Cleared by hardware to stop the HSE oscillator when entering Stop or Standby mode. This
bit cannot be reset if the HSE oscillator is used directly or indirectly as the system clock.
0: HSE oscillator OFF
1: HSE oscillator ON
Bits 15:8 HSICAL[7:0]: Internal high-speed clock calibration
These bits are initialized automatically at startup.
Bits 7:3 HSITRIM[4:0]: Internal high-speed clock trimming
These bits provide an additional user-programmable trimming value that is added to the
HSICAL[7:0] bits. It can be programmed to adjust to variations in voltage and temperature
that influence the frequency of the internal HSI RC.
Bit 2 Reserved, must be kept at reset value.
Bit 1 HSIRDY: Internal high-speed clock ready flag
Set by hardware to indicate that the HSI oscillator is stable. After the HSION bit is cleared,
HSIRDY goes low after 6 HSI clock cycles.
0: HSI oscillator not ready
1: HSI oscillator ready
Bit 0 HSION: Internal high-speed clock enable
Set and cleared by software.
Set by hardware to force the HSI oscillator ON when leaving the Stop or Standby mode or in
case of a failure of the HSE oscillator used directly or indirectly as the system clock. This bit
cannot be cleared if the HSI is used directly or indirectly as the system clock.
0: HSI oscillator OFF
1: HSI oscillator ON

162/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

6.3.2

RCC PLL configuration register (RCC_PLLCFGR)
Address offset: 0x04
Reset value: 0x2400 3010
Access: no wait state, word, half-word and byte access.
This register is used to configure the PLL clock outputs according to the formulas:

31

30

•

f(VCO clock) = f(PLL clock input) × (PLLN / PLLM)

•

f(PLL general clock output) = f(VCO clock) / PLLP

•

f(USB OTG FS, SDIO, RNG clock output) = f(VCO clock) / PLLQ
29

28

27

PLLQ3 PLLQ2

Reserved

rw
15
Reserv
ed

14

13

26

12

11

25

24

PLLQ1

PLLQ0

rw

rw

rw

10

9

8

23

22

7

6

PLLN
rw

rw

rw

rw

rw

21

PLLSR
Reserv
C
ed
rw

20

19

18

Reserved
5

4

3

2

PLLM5 PLLM4 PLLM3 PLLM2
rw

rw

rw

rw

rw

rw

rw

rw

17

16

PLLP1

PLLP0

rw

rw

1

0

PLLM1

PLLM0

rw

rw

Bits 31:28 Reserved, must be kept at reset value.
Bits 27:24 PLLQ: Main PLL (PLL) division factor for USB OTG FS, SDIO and random number generator
clocks
Set and cleared by software to control the frequency of USB OTG FS clock, the random
number generator clock and the SDIO clock. These bits should be written only if PLL is
disabled.
Caution: The USB OTG FS requires a 48 MHz clock to work correctly. The SDIO and the
random number generator need a frequency lower than or equal to 48 MHz to work
correctly.
USB OTG FS clock frequency = VCO frequency / PLLQ with 2 ≤PLLQ ≤15
0000: PLLQ = 0, wrong configuration
0001: PLLQ = 1, wrong configuration
0010: PLLQ = 2
0011: PLLQ = 3
0100: PLLQ = 4
...
1111: PLLQ = 15
Bit 23 Reserved, must be kept at reset value.
Bit 22 PLLSRC: Main PLL(PLL) and audio PLL (PLLI2S) entry clock source
Set and cleared by software to select PLL and PLLI2S clock source. This bit can be written
only when PLL and PLLI2S are disabled.
0: HSI clock selected as PLL and PLLI2S clock entry
1: HSE oscillator clock selected as PLL and PLLI2S clock entry
Bits 21:18 Reserved, must be kept at reset value.

RM0090 Rev 17

163/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

RM0090

Bits 17:16 PLLP: Main PLL (PLL) division factor for main system clock
Set and cleared by software to control the frequency of the general PLL output clock. These
bits can be written only if PLL is disabled.
Caution: The software has to set these bits correctly not to exceed 180 MHz on this domain.
PLL output clock frequency = VCO frequency / PLLP with PLLP = 2, 4, 6, or 8
00: PLLP = 2
01: PLLP = 4
10: PLLP = 6
11: PLLP = 8
Bits 14:6 PLLN: Main PLL (PLL) multiplication factor for VCO
Set and cleared by software to control the multiplication factor of the VCO. These bits can
be written only when PLL is disabled. Only half-word and word accesses are allowed to
write these bits.
Caution: The software has to set these bits correctly to ensure that the VCO output
frequency is between 100 and 432 MHz.
VCO output frequency = VCO input frequency × PLLN with 50 ≤PLLN ≤432
000000000: PLLN = 0, wrong configuration
000000001: PLLN = 1, wrong configuration
...
000110010: PLLN = 50
...
001100011: PLLN = 99
001100100: PLLN = 100
...
110110000: PLLN = 432
110110001: PLLN = 433, wrong configuration
...
111111111: PLLN = 511, wrong configuration
Note: Multiplication factors ranging from 50 and 99 are possible for VCO input frequency
higher than 1 MHz. However care must be taken that the minimum VCO output
frequency respects the value specified above.
Bits 5:0 PLLM: Division factor for the main PLL (PLL) and audio PLL (PLLI2S) input clock
Set and cleared by software to divide the PLL and PLLI2S input clock before the VCO.
These bits can be written only when the PLL and PLLI2S are disabled.
Caution: The software has to set these bits correctly to ensure that the VCO input frequency
ranges from 1 to 2 MHz. It is recommended to select a frequency of 2 MHz to limit
PLL jitter.
VCO input frequency = PLL input clock frequency / PLLM with 2 ≤PLLM ≤63
000000: PLLM = 0, wrong configuration
000001: PLLM = 1, wrong configuration
000010: PLLM = 2
000011: PLLM = 3
000100: PLLM = 4
...
111110: PLLM = 62
111111: PLLM = 63

164/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

6.3.3

RCC clock configuration register (RCC_CFGR)
Address offset: 0x08
Reset value: 0x0000 0000
Access: 0 ≤ wait state ≤ 2, word, half-word and byte access
1 or 2 wait states inserted only if the access occurs during a clock source switch.

31

30

29

MCO2

14

26

rw

25

24

23

22

I2SSC
R

MCO1 PRE[2:0]

rw

rw

rw

rw

rw

rw

rw

rw

12

11

10

9

8

7

6

rw

rw

rw

rw

PPRE1[2:0]
rw

rw

Reserved

21

20

19

MCO1

13

PPRE2[2:0]
rw

27

MCO2 PRE[2:0]

rw
15

28

18

17

16

RTCPRE[4:0]

5

rw

rw

rw

rw

rw

4

3

2

1

0

SWS1

SWS0

SW1

SW0

r

r

rw

rw

HPRE[3:0]
rw

rw

Bits 31:30 MCO2[1:0]: Microcontroller clock output 2
Set and cleared by software. Clock source selection may generate glitches on MCO2. It is
highly recommended to configure these bits only after reset before enabling the external
oscillators and the PLLs.
00: System clock (SYSCLK) selected
01: PLLI2S clock selected
10: HSE oscillator clock selected
11: PLL clock selected
Bits 27:29 MCO2PRE: MCO2 prescaler
Set and cleared by software to configure the prescaler of the MCO2. Modification of this
prescaler may generate glitches on MCO2. It is highly recommended to change this
prescaler only after reset before enabling the external oscillators and the PLLs.
0xx: no division
100: division by 2
101: division by 3
110: division by 4
111: division by 5
Bits 24:26 MCO1PRE: MCO1 prescaler
Set and cleared by software to configure the prescaler of the MCO1. Modification of this
prescaler may generate glitches on MCO1. It is highly recommended to change this
prescaler only after reset before enabling the external oscillators and the PLL.
0xx: no division
100: division by 2
101: division by 3
110: division by 4
111: division by 5
Bit 23 I2SSRC: I2S clock selection
Set and cleared by software. This bit allows to select the I2S clock source between the
PLLI2S clock and the external clock. It is highly recommended to change this bit only after
reset and before enabling the I2S module.
0: PLLI2S clock used as I2S clock source
1: External clock mapped on the I2S_CKIN pin used as I2S clock source

RM0090 Rev 17

165/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

RM0090

Bits 22:21 MCO1: Microcontroller clock output 1
Set and cleared by software. Clock source selection may generate glitches on MCO1. It is
highly recommended to configure these bits only after reset before enabling the external
oscillators and PLL.
00: HSI clock selected
01: LSE oscillator selected
10: HSE oscillator clock selected
11: PLL clock selected
Bits 20:16 RTCPRE: HSE division factor for RTC clock
Set and cleared by software to divide the HSE clock input clock to generate a 1 MHz clock
for RTC.
Caution: The software has to set these bits correctly to ensure that the clock supplied to the
RTC is 1 MHz. These bits must be configured if needed before selecting the RTC
clock source.
00000: no clock
00001: no clock
00010: HSE/2
00011: HSE/3
00100: HSE/4
...
11110: HSE/30
11111: HSE/31
Bits 15:13 PPRE2: APB high-speed prescaler (APB2)
Set and cleared by software to control APB high-speed clock division factor.
Caution: The software has to set these bits correctly not to exceed 90 MHz on this domain.
The clocks are divided with the new prescaler factor from 1 to 16 AHB cycles after
PPRE2 write.
0xx: AHB clock not divided
100: AHB clock divided by 2
101: AHB clock divided by 4
110: AHB clock divided by 8
111: AHB clock divided by 16
Bits 12:10 PPRE1: APB Low speed prescaler (APB1)
Set and cleared by software to control APB low-speed clock division factor.
Caution: The software has to set these bits correctly not to exceed 45 MHz on this domain.
The clocks are divided with the new prescaler factor from 1 to 16 AHB cycles after
PPRE1 write.
0xx: AHB clock not divided
100: AHB clock divided by 2
101: AHB clock divided by 4
110: AHB clock divided by 8
111: AHB clock divided by 16
Bits 9:8 Reserved, must be kept at reset value.

166/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bits 7:4 HPRE: AHB prescaler
Set and cleared by software to control AHB clock division factor.
Caution: The clocks are divided with the new prescaler factor from 1 to 16 AHB cycles after
HPRE write.
Caution: The AHB clock frequency must be at least 25 MHz when the Ethernet is used.
0xxx: system clock not divided
1000: system clock divided by 2
1001: system clock divided by 4
1010: system clock divided by 8
1011: system clock divided by 16
1100: system clock divided by 64
1101: system clock divided by 128
1110: system clock divided by 256
1111: system clock divided by 512
Bits 3:2 SWS: System clock switch status
Set and cleared by hardware to indicate which clock source is used as the system clock.
00: HSI oscillator used as the system clock
01: HSE oscillator used as the system clock
10: PLL used as the system clock
11: not applicable
Bits 1:0 SW: System clock switch
Set and cleared by software to select the system clock source.
Set by hardware to force the HSI selection when leaving the Stop or Standby mode or in
case of failure of the HSE oscillator used directly or indirectly as the system clock.
00: HSI oscillator selected as system clock
01: HSE oscillator selected as system clock
10: PLL selected as system clock
11: not allowed

6.3.4

RCC clock interrupt register (RCC_CIR)
Address offset: 0x0C
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access

31

30

29

28

27

26

25

24

Reserv
ed

14

13

PLLSAI
RDYIE

PLLI2S
RDYIE

rw

rw

12

11

10

PLL
HSE
HSI
RDYIE RDYIE RDYIE
rw

rw

22

21

PLLSAI PLLI2S
CSSC
RDYC RDYC

Reserved
15

23

rw

20

19

18

17

16

PLL
RDYC

HSE
RDYC

HSI
RDYC

LSE
RDYC

LSI
RDYC

w

w

w

w

w

w

w

w

9

8

7

6

5

4

3

2

1

0

LSE
RDYIE

LSI
RDYIE

CSSF

PLL
RDYF

HSE
RDYF

HSI
RDYF

LSE
RDYF

LSI
RDYF

rw

rw

r

r

r

r

r

r

PLLSAI PLLI2S
RDYF
RDYF

RM0090 Rev 17

r

r

167/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

RM0090

Bits 31:24 Reserved, must be kept at reset value.
Bit 23 CSSC: Clock security system interrupt clear
This bit is set by software to clear the CSSF flag.
0: No effect
1: Clear CSSF flag
Bit 22 PLLSAIRDYC: PLLSAI Ready Interrupt Clear
This bit is set by software to clear PLLSAIRDYF flag. It is reset by hardware when the
PLLSAIRDYF is cleared.
0: PLLSAIRDYF not cleared
1: PLLSAIRDYF cleared
Bit 21 PLLI2SRDYC: PLLI2S ready interrupt clear
This bit is set by software to clear the PLLI2SRDYF flag.
0: No effect
1: PLLI2SRDYF cleared
Bit 20 PLLRDYC: Main PLL(PLL) ready interrupt clear
This bit is set by software to clear the PLLRDYF flag.
0: No effect
1: PLLRDYF cleared
Bit 19 HSERDYC: HSE ready interrupt clear
This bit is set by software to clear the HSERDYF flag.
0: No effect
1: HSERDYF cleared
Bit 18 HSIRDYC: HSI ready interrupt clear
This bit is set software to clear the HSIRDYF flag.
0: No effect
1: HSIRDYF cleared
Bit 17 LSERDYC: LSE ready interrupt clear
This bit is set by software to clear the LSERDYF flag.
0: No effect
1: LSERDYF cleared
Bit 16 LSIRDYC: LSI ready interrupt clear
This bit is set by software to clear the LSIRDYF flag.
0: No effect
1: LSIRDYF cleared
Bit 15 Reserved, must be kept at reset value.
Bit 14 PLLSAIRDYIE: PLLSAI Ready Interrupt Enable
This bit is set and reset by software to enable/disable interrupt caused by PLLSAI lock.
0: PLLSAI lock interrupt disabled
1: PLLSAI lock interrupt enabled
Bit 13 PLLI2SRDYIE: PLLI2S ready interrupt enable
This bit is set and cleared by software to enable/disable interrupt caused by PLLI2S lock.
0: PLLI2S lock interrupt disabled
1: PLLI2S lock interrupt enabled

168/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 12 PLLRDYIE: Main PLL (PLL) ready interrupt enable
This bit is set and cleared by software to enable/disable interrupt caused by PLL lock.
0: PLL lock interrupt disabled
1: PLL lock interrupt enabled
Bit 11 HSERDYIE: HSE ready interrupt enable
This bit is set and cleared by software to enable/disable interrupt caused by the HSE
oscillator stabilization.
0: HSE ready interrupt disabled
1: HSE ready interrupt enabled
Bit 10 HSIRDYIE: HSI ready interrupt enable
This bit is set and cleared by software to enable/disable interrupt caused by the HSI
oscillator stabilization.
0: HSI ready interrupt disabled
1: HSI ready interrupt enabled
Bit 9 LSERDYIE: LSE ready interrupt enable
This bit is set and cleared by software to enable/disable interrupt caused by the LSE
oscillator stabilization.
0: LSE ready interrupt disabled
1: LSE ready interrupt enabled
Bit 8 LSIRDYIE: LSI ready interrupt enable
This bit is set and cleared by software to enable/disable interrupt caused by LSI oscillator
stabilization.
0: LSI ready interrupt disabled
1: LSI ready interrupt enabled
Bit 7 CSSF: Clock security system interrupt flag
This bit is set by hardware when a failure is detected in the HSE oscillator.
It is cleared by software by setting the CSSC bit.
0: No clock security interrupt caused by HSE clock failure
1: Clock security interrupt caused by HSE clock failure
Bit 6 PLLSAIRDYF: PLLSAI Ready Interrupt flag
This bit is set by hardware when the PLLSAI is locked and PLLSAIRDYDIE is set.
It is cleared by software by setting the PLLSAIRDYC bit.
0: No clock ready interrupt caused by PLLSAI lock
1: Clock ready interrupt caused by PLLSAI lock
Bit 5 PLLI2SRDYF: PLLI2S ready interrupt flag
This bit is set by hardware when the PLLI2S is locked and PLLI2SRDYDIE is set.
It is cleared by software by setting the PLLRI2SDYC bit.
0: No clock ready interrupt caused by PLLI2S lock
1: Clock ready interrupt caused by PLLI2S lock
Bit 4 PLLRDYF: Main PLL (PLL) ready interrupt flag
This bit is set by hardware when PLL is locked and PLLRDYDIE is set.
It is cleared by software setting the PLLRDYC bit.
0: No clock ready interrupt caused by PLL lock
1: Clock ready interrupt caused by PLL lock

RM0090 Rev 17

169/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

RM0090

Bit 3 HSERDYF: HSE ready interrupt flag
This bit is set by hardware when External High Speed clock becomes stable and
HSERDYDIE is set.
It is cleared by software by setting the HSERDYC bit.
0: No clock ready interrupt caused by the HSE oscillator
1: Clock ready interrupt caused by the HSE oscillator
Bit 2 HSIRDYF: HSI ready interrupt flag
This bit is set by hardware when the Internal High Speed clock becomes stable and
HSIRDYDIE is set.
It is cleared by software by setting the HSIRDYC bit.
0: No clock ready interrupt caused by the HSI oscillator
1: Clock ready interrupt caused by the HSI oscillator
Bit 1 LSERDYF: LSE ready interrupt flag
This bit is set by hardware when the External Low Speed clock becomes stable and
LSERDYDIE is set.
It is cleared by software by setting the LSERDYC bit.
0: No clock ready interrupt caused by the LSE oscillator
1: Clock ready interrupt caused by the LSE oscillator
Bit 0 LSIRDYF: LSI ready interrupt flag
This bit is set by hardware when the internal low speed clock becomes stable and
LSIRDYDIE is set.
It is cleared by software by setting the LSIRDYC bit.
0: No clock ready interrupt caused by the LSI oscillator
1: Clock ready interrupt caused by the LSI oscillator

6.3.5

RCC AHB1 peripheral reset register (RCC_AHB1RSTR)
Address offset: 0x10
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31

30

29

28

OTGH
S
RST

Reserved

27

26

Reserved

rw
15

14
Reserved

13

25

24

23

22

21

ETHMAC
RST

Res.

DMA2D
RST

DMA2
RST

DMA1
RST

rw

rw

rw

7

6

5

rw
12
CRCR
ST
rw

11

10

9

8

Res.

GPIOK
RST

GPIOJ
RST

GPIOI
RST

rw

rw

rw

19

18

17

rw

Bit 29 OTGHSRST: USB OTG HS module reset
This bit is set and cleared by software.
0: does not reset the USB OTG HS module
1: resets the USB OTG HS module
Bits 28:26 Reserved, must be kept at reset value.

RM0090 Rev 17

rw

rw

16

Reserved

4

3

2

1

GPIOH GPIOGG GPIOF GPIOE GPIOD GPIOC GPIOB
RST
RST
RST
RST
RST
RST
RST

Bits 31:30 Reserved, must be kept at reset value.

170/1747

20

rw

rw

rw

rw

0
GPIOA
RST
rw

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 25 ETHMACRST: Ethernet MAC reset
This bit is set and cleared by software.
0: does not reset Ethernet MAC
1: resets Ethernet MAC
Bit 24 Reserved, must be kept at reset value.
Bit 23 DMA2DRST: DMA2D reset
This bit is set and reset by software.
0: does not reset DMA2D
1: resets DMA2D
Bit 22 DMA2RST: DMA2 reset
This bit is set and cleared by software.
0: does not reset DMA2
1: resets DMA2
Bit 21 DMA1RST: DMA2 reset
This bit is set and cleared by software.
0: does not reset DMA2
1: resets DMA2
Bits 20:13 Reserved, must be kept at reset value.
Bit 12 CRCRST: CRC reset
This bit is set and cleared by software.
0: does not reset CRC
1: resets CRC
Bit 11 Reserved, must be kept at reset value.
Bit 10 GPIOKRST: IO port K reset
This bit is set and cleared by software.
0: does not reset IO port K
1: resets IO port K
Bit 9 GPIOJRST: IO port J reset
This bit is set and cleared by software.
0: does not reset IO port J
1: resets IO port J
Bit 8 GPIOIRST: IO port I reset
This bit is set and cleared by software.
0: does not reset IO port I
1: resets IO port I
Bit 7 GPIOHRST: IO port H reset
This bit is set and cleared by software.
0: does not reset IO port H
1: resets IO port H
Bit 6 GPIOGRST: IO port G reset
This bit is set and cleared by software.
0: does not reset IO port G
1: resets IO port G

RM0090 Rev 17

171/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 5 GPIOFRST: IO port F reset
This bit is set and cleared by software.
0: does not reset IO port F
1: resets IO port F
Bit 4 GPIOERST: IO port E reset
This bit is set and cleared by software.
0: does not reset IO port E
1: resets IO port E
Bit 3 GPIODRST: IO port D reset
This bit is set and cleared by software.
0: does not reset IO port D
1: resets IO port D
Bit 2 GPIOCRST: IO port C reset
This bit is set and cleared by software.
0: does not reset IO port C
1: resets IO port C
Bit 1 GPIOBRST: IO port B reset
This bit is set and cleared by software.
0: does not reset IO port B
1:resets IO port B
Bit 0 GPIOARST: IO port A reset
This bit is set and cleared by software.
0: does not reset IO port A
1: resets IO port A

172/1747

RM0090 Rev 17

RM0090

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

6.3.6

RCC AHB2 peripheral reset register (RCC_AHB2RSTR)
Address offset: 0x14
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

3

2

1

16

Reserved
15

14

13

12

11

Reserved

10

9

8

7

6

5

4

OTGFS
RST

RNG
RST

HASH
RST

CRYP
RST

rw

rw

rw

rw

Reserved

0
DCMI
RST
rw

Bits 31:8 Reserved, must be kept at reset value.
Bit 7 OTGFSRST: USB OTG FS module reset
Set and cleared by software.
0: does not reset the USB OTG FS module
1: resets the USB OTG FS module
Bit 6 RNGRST: Random number generator module reset
Set and cleared by software.
0: does not reset the random number generator module
1: resets the random number generator module
Bit 5 HASHRST: Hash module reset
Set and cleared by software.
0: does not reset the HASH module
1: resets the HASH module
Bit 4 CRYPRST: Cryptographic module reset
Set and cleared by software.
0: does not reset the cryptographic module
1: resets the cryptographic module
Bits 3:1 Reserved, must be kept at reset value.
Bit 0 DCMIRST: Camera interface reset
Set and cleared by software.
0: does not reset the Camera interface
1: resets the Camera interface

RM0090 Rev 17

173/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

6.3.7

RM0090

RCC AHB3 peripheral reset register (RCC_AHB3RSTR)
Address offset: 0x18
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

Reserved
15

14

13

12

11

10

9

8

7

FMCRST

Reserved

rw

Bits 31:1 Reserved, must be kept at reset value.
Bit 0 FMCRST: Flexible memory controller module reset
Set and cleared by software.
0: does not reset the FMC module
1: resets the FMC module

6.3.8

RCC APB1 peripheral reset register (RCC_APB1RSTR)
Address offset: 0x20
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31

30

29

28

PWR
UART8R UART7R
DACRST
ST
ST
RST
rw

rw

rw

rw

15

14

13

12

SPI3
RST

SPI2
RST

rw

rw

174/1747

Reserved

27
Reserved
11
WWDG
RST
rw

26

25

CAN2
RST

CAN1
RST
rw

10

9

Reserved

24

23

22

21

I2C3
RST

I2C2
RST

I2C1
RST

rw

rw

rw

8

7

6

5

4

3

2

1

0

TIM14
RST

TIM13
RST

TIM12
RST

TIM7
RST

TIM6
RST

TIM5
RST

TIM4
RST

TIM3
RST

TIM2
RST

rw

rw

rw

rw

rw

rw

rw

rw

rw

Reserved

RM0090 Rev 17

20

19

18

17

16

UART5 UART4 UART3 UART2
ReserRST
RST
RST
RST
ved
rw
rw
rw
rw

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 31 UART8RST: UART8 reset
Set and cleared by software.
0: does not reset UART8
1: resets UART8
Bit 30 UART7RST: UART7 reset
Set and cleared by software.
0: does not reset UART7
1: resets UART7
Bit 29 DACRST: DAC reset
Set and cleared by software.
0: does not reset the DAC interface
1: resets the DAC interface
Bit 28 PWRRST: Power interface reset
Set and cleared by software.
0: does not reset the power interface
1: resets the power interface
Bit 27 Reserved, must be kept at reset value.
Bit 26 CAN2RST: CAN2 reset
Set and cleared by software.
0: does not reset CAN2
1: resets CAN2
Bit 25 CAN1RST: CAN1 reset
Set and cleared by software.
0: does not reset CAN1
1: resets CAN1
Bit 24 Reserved, must be kept at reset value.
Bit 23 I2C3RST: I2C3 reset
Set and cleared by software.
0: does not reset I2C3
1: resets I2C3
Bit 22 I2C2RST: I2C2 reset
Set and cleared by software.
0: does not reset I2C2
1: resets I2C2
Bit 21 I2C1RST: I2C1 reset
Set and cleared by software.
0: does not reset I2C1
1: resets I2C1
Bit 20 UART5RST: UART5 reset
Set and cleared by software.
0: does not reset UART5
1: resets UART5
Bit 19 UART4RST: USART4 reset
Set and cleared by software.
0: does not reset UART4
1: resets UART4

RM0090 Rev 17

175/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 18 USART3RST: USART3 reset
Set and cleared by software.
0: does not reset USART3
1: resets USART3
Bit 17 USART2RST: USART2 reset
Set and cleared by software.
0: does not reset USART2
1: resets USART2
Bit 16 Reserved, must be kept at reset value.
Bit 15 SPI3RST: SPI3 reset
Set and cleared by software.
0: does not reset SPI3
1: resets SPI3
Bit 14 SPI2RST: SPI2 reset
Set and cleared by software.
0: does not reset SPI2
1: resets SPI2
Bits 13:12 Reserved, must be kept at reset value.
Bit 11 WWDGRST: Window watchdog reset
Set and cleared by software.
0: does not reset the window watchdog
1: resets the window watchdog
Bits 10:9 Reserved, must be kept at reset value.
Bit 8 TIM14RST: TIM14 reset
Set and cleared by software.
0: does not reset TIM14
1: resets TIM14
Bit 7 TIM13RST: TIM13 reset
Set and cleared by software.
0: does not reset TIM13
1: resets TIM13
Bit 6 TIM12RST: TIM12 reset
Set and cleared by software.
0: does not reset TIM12
1: resets TIM12
Bit 5 TIM7RST: TIM7 reset
Set and cleared by software.
0: does not reset TIM7
1: resets TIM7
Bit 4 TIM6RST: TIM6 reset
Set and cleared by software.
0: does not reset TIM6
1: resets TIM6

176/1747

RM0090 Rev 17

RM0090

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 3 TIM5RST: TIM5 reset
Set and cleared by software.
0: does not reset TIM5
1: resets TIM5
Bit 2 TIM4RST: TIM4 reset
Set and cleared by software.
0: does not reset TIM4
1: resets TIM4
Bit 1 TIM3RST: TIM3 reset
Set and cleared by software.
0: does not reset TIM3
1: resets TIM3
Bit 0 TIM2RST: TIM2 reset
Set and cleared by software.
0: does not reset TIM2
1: resets TIM2

RM0090 Rev 17

177/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

6.3.9

RM0090

RCC APB2 peripheral reset register (RCC_APB2RSTR)
Address offset: 0x24
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31

30

29

28

27

26

25

LTD
CRST

Reserved

24

23

Reserved

22

21

20

SAI1
RST

SPI6
RST

SPI5
RST

rw

rw

rw

6

5

4

USART
6
RST

USART
1
RST

rw

rw

rw
15
Reserved

14

13

12

11

SYSCF
G RST

SPI4
RST

SPI1
RST

SDIO
RST

rw

rw

rw

rw

10

9

Reserved

8

7

ADC
RST

Reserved

rw

Bits 31:27 Reserved, must be kept at reset value.
Bit 26 LTDCRST: LTDC reset
This bit is set and reset by software.
0: does not reset LCD-TFT
1: resets LCD-TFT
Bits 27:23 Reserved, must be kept at reset value.
Bit 22 SAI1RST: SAI1 reset
This bit is set and reset by software.
0: does not reset SAI1
1: resets SAI1
Bit 21 SPI6RST: SPI6 reset
This bit is set and cleared by software.
0: does not reset SPI6
1: resets SPI6
Bit 20 SPI5RST: SPI5 reset
This bit is set and cleared by software.
0: does not reset SPI5
1: resets SPI5
Bit 19 Reserved, must be kept at reset value.
Bit 18 TIM11RST: TIM11 reset
This bit is set and cleared by software.
0: does not reset TIM11
1: resets TIM14
Bit 17 TIM10RST: TIM10 reset
This bit is set and cleared by software.
0: does not reset TIM10
1: resets TIM10
Bit 16 TIM9RST: TIM9 reset
This bit is set and cleared by software.
0: does not reset TIM9
1: resets TIM9

178/1747

RM0090 Rev 17

19

18

17

16

Res.

TIM11
RST

TIM10
RST

TIM9
RST

rw

rw

rw

2

1

0

TIM8
RST

TIM1
RST

rw

rw

3

Reserved

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 15 Reserved, must be kept at reset value.
Bit 14 SYSCFGRST: System configuration controller reset
This bit is set and cleared by software.
0: does not reset the System configuration controller
1: resets the System configuration controller
Bit 13 SPI4RST: SPI4 reset
This bit is set and cleared by software.
0: does not reset SPI4
1: resets SPI4
Bit 12 SPI1RST: SPI1 reset
This bit is set and cleared by software.
0: does not reset SPI1
1: resets SPI1
Bit 11 SDIORST: SDIO reset
This bit is set and cleared by software.
0: does not reset the SDIO module
1: resets the SDIO module
Bits 10:9 Reserved, must be kept at reset value.
Bit 8 ADCRST: ADC interface reset (common to all ADCs)
This bit is set and cleared by software.
0: does not reset the ADC interface
1: resets the ADC interface
Bits 7:6 Reserved, must be kept at reset value.
Bit 5 USART6RST: USART6 reset
This bit is set and cleared by software.
0: does not reset USART6
1: resets USART6
Bit 4

USART1RST: USART1 reset
This bit is set and cleared by software.
0: does not reset USART1
1: resets USART1

Bits 3:2 Reserved, must be kept at reset value.
Bit 1 TIM8RST: TIM8 reset
This bit is set and cleared by software.
0: does not reset TIM8
1: resets TIM8
Bit 0 TIM1RST: TIM1 reset
This bit is set and cleared by software.
0: does not reset TIM1
1: resets TIM1

RM0090 Rev 17

179/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

6.3.10

RM0090

RCC AHB1 peripheral clock register (RCC_AHB1ENR)
Address offset: 0x30
Reset value: 0x0010 0000
Access: no wait state, word, half-word and byte access.

31

30

OTGH
S
Reser- ULPIE
ved
N

15

29

28

27

26

25

ETHM ETHM ETHM
OTGH
ETHMA
ACPTP ACRXE ACTXE
SEN
CEN
EN
N
N

24

Res.

23

22

DMA2D DMA2E
EN
N

21

20

19

18

DMA1E
N

CCMDAT
ARAMEN

Res.

BKPSR
AMEN

4

3

rw

rw

rw

rw

rw

rw

rw

rw

rw

14

13

12

11

10

9

8

7

6

5

Res.

GPIOK
EN

GPIOJ
EN

GPIOIE
N

GPIOH
EN

rw

rw

rw

rw

Reserved

CRCE
N
rw

16

Reserved

rw
2

1

0

GPIOD GPIOC GPIO GPIO
GPIOG GPIOFE
GPIOEEN
EN
EN
BEN AEN
EN
N
rw

Bit 31 Reserved, must be kept at reset value.
Bit 30 OTGHSULPIEN: USB OTG HSULPI clock enable
This bit is set and cleared by software.
0: USB OTG HS ULPI clock disabled
1: USB OTG HS ULPI clock enabled
Bit 29 OTGHSEN: USB OTG HS clock enable
This bit is set and cleared by software.
0: USB OTG HS clock disabled
1: USB OTG HS clock enabled
Bit 28 ETHMACPTPEN: Ethernet PTP clock enable
This bit is set and cleared by software.
0: Ethernet PTP clock disabled
1: Ethernet PTP clock enabled
Bit 27 ETHMACRXEN: Ethernet Reception clock enable
This bit is set and cleared by software.
0: Ethernet Reception clock disabled
1: Ethernet Reception clock enabled
Bit 26 ETHMACTXEN: Ethernet Transmission clock enable
This bit is set and cleared by software.
0: Ethernet Transmission clock disabled
1: Ethernet Transmission clock enabled
Bit 25 ETHMACEN: Ethernet MAC clock enable
This bit is set and cleared by software.
0: Ethernet MAC clock disabled
1: Ethernet MAC clock enabled
Bit 24 Reserved, must be kept at reset value.
Bit 23 DMA2DEN: DMA2D clock enable
This bit is set and cleared by software.
0: DMA2D clock disabled
1: DMA2D clock enabled

180/1747

17

RM0090 Rev 17

rw

rw

rw

rw

rw

rw

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 22 DMA2EN: DMA2 clock enable
This bit is set and cleared by software.
0: DMA2 clock disabled
1: DMA2 clock enabled
Bit 21 DMA1EN: DMA1 clock enable
This bit is set and cleared by software.
0: DMA1 clock disabled
1: DMA1 clock enabled
Bit 20 CCMDATARAMEN: CCM data RAM clock enable
This bit is set and cleared by software.
0: CCM data RAM clock disabled
1: CCM data RAM clock enabled
Bit 19 Reserved, must be kept at reset value.
Bit 18 BKPSRAMEN: Backup SRAM interface clock enable
This bit is set and cleared by software.
0: Backup SRAM interface clock disabled
1: Backup SRAM interface clock enabled
Bits 17:13 Reserved, must be kept at reset value.
Bit 12 CRCEN: CRC clock enable
This bit is set and cleared by software.
0: CRC clock disabled
1: CRC clock enabled
Bit 11 Reserved, must be kept at reset value.
Bit 10 GPIOKEN: IO port K clock enable
This bit is set and cleared by software.
0: IO port K clock disabled
1: IO port K clock enabled
Bit 9 GPIOJEN: IO port J clock enable
This bit is set and cleared by software.
0: IO port J clock disabled
1: IO port J clock enabled
Bit 8 GPIOIEN: IO port I clock enable
This bit is set and cleared by software.
0: IO port I clock disabled
1: IO port I clock enabled
Bit 7 GPIOHEN: IO port H clock enable
This bit is set and cleared by software.
0: IO port H clock disabled
1: IO port H clock enabled
Bit 6 GPIOGEN: IO port G clock enable
This bit is set and cleared by software.
0: IO port G clock disabled
1: IO port G clock enabled

RM0090 Rev 17

181/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

RM0090

Bit 5 GPIOFEN: IO port F clock enable
This bit is set and cleared by software.
0: IO port F clock disabled
1: IO port F clock enabled
Bit 4 GPIOEEN: IO port E clock enable
This bit is set and cleared by software.
0: IO port E clock disabled
1: IO port E clock enabled
Bit 3 GPIODEN: IO port D clock enable
This bit is set and cleared by software.
0: IO port D clock disabled
1: IO port D clock enabled
Bit 2 GPIOCEN: IO port C clock enable
This bit is set and cleared by software.
0: IO port C clock disabled
1: IO port C clock enabled
Bit 1 GPIOBEN: IO port B clock enable
This bit is set and cleared by software.
0: IO port B clock disabled
1: IO port B clock enabled
Bit 0 GPIOAEN: IO port A clock enable
This bit is set and cleared by software.
0: IO port A clock disabled
1: IO port A clock enabled

6.3.11

RCC AHB2 peripheral clock enable register (RCC_AHB2ENR)
Address offset: 0x34
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

7

6

5

4

3

2

1

0

OTGFS
EN

RNG
EN

HASH
EN

CRYP
EN

rw

rw

rw

rw

Reserved
15

14

13

12

11

Reserved

10

9

8

Bits 31:8 Reserved, must be kept at reset value.
Bit 7 OTGFSEN: USB OTG FS clock enable
This bit is set and cleared by software.
0: USB OTG FS clock disabled
1: USB OTG FS clock enabled
Bit 6 RNGEN: Random number generator clock enable
This bit is set and cleared by software.
0: Random number generator clock disabled
1: Random number generator clock enabled

182/1747

RM0090 Rev 17

Reserved

DCMI
EN
rw

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 5 HASHEN: Hash modules clock enable
This bit is set and cleared by software.
0: Hash modules clock disabled
1: Hash modules clock enabled
Bit 4 CRYPEN: Cryptographic modules clock enable
This bit is set and cleared by software.
0: cryptographic module clock disabled
1: cryptographic module clock enabled
Bits 3:1 Reserved, must be kept at reset value.
Bit 0 DCMIEN: Camera interface enable
This bit is set and cleared by software.
0: Camera interface clock disabled
1: Camera interface clock enabled

6.3.12

RCC AHB3 peripheral clock enable register (RCC_AHB3ENR)
Address offset: 0x38
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

6

5

4

3

2

1

16

Reserved
15

14

13

12

11

10

9

8

7

0
FMCEN

Reserved

rw

Bits 31:1 Reserved, must be kept at reset value.
Bit 0 FMCEN: Flexible memory controller module clock enable
This bit is set and cleared by software.
0: FMC module clock disabled
1: FMC module clock enabled

6.3.13

RCC APB1 peripheral clock enable register (RCC_APB1ENR)
Address offset: 0x40
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31

30

UART8 UART7
EN
EN

29
DAC
EN

28
PWR
EN

rw

rw

rw

rw

15

14

13

12

SPI3
EN

SPI2
EN

rw

rw

Reserved

27
Reserved
11
WWDG
EN
rw

26

25

CAN2
EN

CAN1
EN

rw

rw

10

9

Reserved

24
Reserved

23

22

21

I2C3
EN

I2C2
EN

I2C1
EN

20

19

18

17

16

USART USART
UART5 UART4
3
2
EN
EN
EN
EN

Reserved

rw

rw

rw

rw

rw

rw

8

7

6

5

4

3

2

1

0

TIM14
EN

TIM13
EN

TIM12
EN

TIM7
EN

TIM6
EN

TIM5
EN

TIM4
EN

TIM3
EN

TIM2
EN

rw

rw

rw

rw

rw

rw

rw

rw

rw

RM0090 Rev 17

rw

183/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 31 UART8EN: UART8 clock enable
This bit is set and cleared by software.
0: UART8 clock disabled
1: UART8 clock enabled
Bit 30 UART7EN: UART7 clock enable
This bit is set and cleared by software.
0: UART7 clock disabled
1: UART7 clock enabled
Bit 29 DACEN: DAC interface clock enable
This bit is set and cleared by software.
0: DAC interface clock disabled
1: DAC interface clock enable
Bit 28 PWREN: Power interface clock enable
This bit is set and cleared by software.
0: Power interface clock disabled
1: Power interface clock enable
Bit 27 Reserved, must be kept at reset value.
Bit 26 CAN2EN: CAN 2 clock enable
This bit is set and cleared by software.
0: CAN 2 clock disabled
1: CAN 2 clock enabled
Bit 25 CAN1EN: CAN 1 clock enable
This bit is set and cleared by software.
0: CAN 1 clock disabled
1: CAN 1 clock enabled
Bit 24 Reserved, must be kept at reset value.
Bit 23 I2C3EN: I2C3 clock enable
This bit is set and cleared by software.
0: I2C3 clock disabled
1: I2C3 clock enabled
Bit 22 I2C2EN: I2C2 clock enable
This bit is set and cleared by software.
0: I2C2 clock disabled
1: I2C2 clock enabled
Bit 21 I2C1EN: I2C1 clock enable
This bit is set and cleared by software.
0: I2C1 clock disabled
1: I2C1 clock enabled
Bit 20 UART5EN: UART5 clock enable
This bit is set and cleared by software.
0: UART5 clock disabled
1: UART5 clock enabled
Bit 19 UART4EN: UART4 clock enable
This bit is set and cleared by software.
0: UART4 clock disabled
1: UART4 clock enabled

184/1747

RM0090 Rev 17

RM0090

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 18 USART3EN: USART3 clock enable
This bit is set and cleared by software.
0: USART3 clock disabled
1: USART3 clock enabled
Bit 17 USART2EN: USART2 clock enable
This bit is set and cleared by software.
0: USART2 clock disabled
1: USART2 clock enabled
Bit 16 Reserved, must be kept at reset value.
Bit 15 SPI3EN: SPI3 clock enable
This bit is set and cleared by software.
0: SPI3 clock disabled
1: SPI3 clock enabled
Bit 14 SPI2EN: SPI2 clock enable
This bit is set and cleared by software.
0: SPI2 clock disabled
1: SPI2 clock enabled
Bits 13:12 Reserved, must be kept at reset value.
Bit 11 WWDGEN: Window watchdog clock enable
This bit is set and cleared by software.
0: Window watchdog clock disabled
1: Window watchdog clock enabled
Bit 10:9 Reserved, must be kept at reset value.
Bit 8 TIM14EN: TIM14 clock enable
This bit is set and cleared by software.
0: TIM14 clock disabled
1: TIM14 clock enabled
Bit 7 TIM13EN: TIM13 clock enable
This bit is set and cleared by software.
0: TIM13 clock disabled
1: TIM13 clock enabled
Bit 6 TIM12EN: TIM12 clock enable
This bit is set and cleared by software.
0: TIM12 clock disabled
1: TIM12 clock enabled
Bit 5 TIM7EN: TIM7 clock enable
This bit is set and cleared by software.
0: TIM7 clock disabled
1: TIM7 clock enabled
Bit 4 TIM6EN: TIM6 clock enable
This bit is set and cleared by software.
0: TIM6 clock disabled
1: TIM6 clock enabled

RM0090 Rev 17

185/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 3 TIM5EN: TIM5 clock enable
This bit is set and cleared by software.
0: TIM5 clock disabled
1: TIM5 clock enabled
Bit 2 TIM4EN: TIM4 clock enable
This bit is set and cleared by software.
0: TIM4 clock disabled
1: TIM4 clock enabled
Bit 1 TIM3EN: TIM3 clock enable
This bit is set and cleared by software.
0: TIM3 clock disabled
1: TIM3 clock enabled
Bit 0 TIM2EN: TIM2 clock enable
This bit is set and cleared by software.
0: TIM2 clock disabled
1: TIM2 clock enabled

186/1747

RM0090 Rev 17

RM0090

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

6.3.14

RCC APB2 peripheral clock enable register (RCC_APB2ENR)
Address offset: 0x44
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31

30

29

28

27

26

25

LTDC
EN

Reserved

24

23

22

Reserved

14

13

12

11

10

9

8

SYSCF
G EN

SPI4E
N

SPI1
EN

SDIO
EN

ADC3
EN

ADC2
EN

ADC1
EN

rw

rw

rw

rw

rw

rw

rw

20

SAI1EN SPI6EN SPI5EN

Reserved

rw
15

21

7

rw

rw

rw

6

5

4

Reserved

USART USART
1
6
EN
EN
rw

rw

19

18

17

16

Res.

TIM11
EN

TIM10
EN

TIM9
EN

rw

rw

rw

2

1

0

TIM8
EN

TIM1
EN

rw

rw

3

Reserved

Bits 31:27 Reserved, must be kept at reset value.
Bit 26 LTDCEN: LTDC clock enable
This bit is set and cleared by software.
0: LTDC clock disabled
1: LTDC clock enabled
Bits 27: 23 Reserved, must be kept at reset value.
Bit 22 SAI1EN: SAI1 clock enable
This bit is set and cleared by software.
0: SAI1 clock disabled
1: SAI1 clock enabled
Bit 21 SPI6EN: SPI6 clock enable
This bit is set and cleared by software.
0: SPI6 clock disabled
1: SPI6 clock enabled
Bit 20 SPI5EN: SPI5 clock enable
This bit is set and cleared by software.
0: SPI5 clock disabled
1: SPI5 clock enabled
Bit 18 TIM11EN: TIM11 clock enable
This bit is set and cleared by software.
0: TIM11 clock disabled
1: TIM11 clock enabled
Bit 17 TIM10EN: TIM10 clock enable
This bit is set and cleared by software.
0: TIM10 clock disabled
1: TIM10 clock enabled
Bit 16 TIM9EN: TIM9 clock enable
This bit is set and cleared by software.
0: TIM9 clock disabled
1: TIM9 clock enabled
Bit 15 Reserved, must be kept at reset value.

RM0090 Rev 17

187/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 14 SYSCFGEN: System configuration controller clock enable
This bit is set and cleared by software.
0: System configuration controller clock disabled
1: System configuration controller clock enabled
Bit 13 SPI4EN: SPI4 clock enable
This bit is set and cleared by software.
0: SPI4 clock disabled
1: SPI4 clock enabled
Bit 12 SPI1EN: SPI1 clock enable
This bit is set and cleared by software.
0: SPI1 clock disabled
1: SPI1 clock enabled
Bit 11 SDIOEN: SDIO clock enable
This bit is set and cleared by software.
0: SDIO module clock disabled
1: SDIO module clock enabled
Bit 10 ADC3EN: ADC3 clock enable
This bit is set and cleared by software.
0: ADC3 clock disabled
1: ADC3 clock enabled
Bit 9

ADC2EN: ADC2 clock enable
This bit is set and cleared by software.
0: ADC2 clock disabled
1: ADC2 clock enabled

Bit 8

ADC1EN: ADC1 clock enable
This bit is set and cleared by software.
0: ADC1 clock disabled
1: ADC1 clock enabled

Bits 7:6 Reserved, must be kept at reset value.
Bit 5 USART6EN: USART6 clock enable
This bit is set and cleared by software.
0: USART6 clock disabled
1: USART6 clock enabled
Bit 4 USART1EN: USART1 clock enable
This bit is set and cleared by software.
0: USART1 clock disabled
1: USART1 clock enabled
Bits 3:2 Reserved, must be kept at reset value.
Bit 1 TIM8EN: TIM8 clock enable
This bit is set and cleared by software.
0: TIM8 clock disabled
1: TIM8 clock enabled
Bit 0 TIM1EN: TIM1 clock enable
This bit is set and cleared by software.
0: TIM1 clock disabled
1: TIM1 clock enabled

188/1747

RM0090 Rev 17

RM0090

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

6.3.15

RCC AHB1 peripheral clock enable in low power mode register
(RCC_AHB1LPENR)
Address offset: 0x50
Reset value: 0x7EEF 97FF
Access: no wait state, word, half-word and byte access.

31

30

29

Res.

OTGHS
ULPILPE
N

OTGH
S
LPEN

rw

rw

rw

rw

rw

rw

14

13

12

11

10

9

15
FLITF
LPEN
rw

Reserved

28

27

26

25

24

ETHPT
ETHMA
ETHRX ETHTX
P
C
LPEN
LPEN
LPEN
LPEN

CRC
LPEN
rw

Res.

Res.

23

22

21

DMA2D
LPEN

DMA2
LPEN

DMA1
LPEN

rw

rw

rw

8

7

6

GPIOK
LPEN

GPIOIJ
LPEN

GPIOI
LPEN

GPIOH
LPEN

GPIOG
G
LPEN

rw

rw

rw

rw

rw

5

20

Res.

19

17

16

4

rw

rw

rw

rw

3

2

1

0

GPIO
GPIOE GPIOD
F
LPEN LPEN
LPEN
rw

18

BKPSRA SRAM SRAM
SRAM3
M
2
1
LPEN
LPEN
LPEN LPEN

rw

GPIOC
LPEN

rw

rw

GPIOB GPIOA
LPEN LPEN
rw

rw

Bit 31 Reserved, must be kept at reset value.
Bit 30 OTGHSULPILPEN: USB OTG HS ULPI clock enable during Sleep mode
This bit is set and cleared by software.
0: USB OTG HS ULPI clock disabled during Sleep mode
1: USB OTG HS ULPI clock enabled during Sleep mode
Bit 29 OTGHSLPEN: USB OTG HS clock enable during Sleep mode
This bit is set and cleared by software.
0: USB OTG HS clock disabled during Sleep mode
1: USB OTG HS clock enabled during Sleep mode
Bit 28 ETHMACPTPLPEN: Ethernet PTP clock enable during Sleep mode
This bit is set and cleared by software.
0: Ethernet PTP clock disabled during Sleep mode
1: Ethernet PTP clock enabled during Sleep mode
Bit 27 ETHMACRXLPEN: Ethernet reception clock enable during Sleep mode
This bit is set and cleared by software.
0: Ethernet reception clock disabled during Sleep mode
1: Ethernet reception clock enabled during Sleep mode
Bit 26 ETHMACTXLPEN: Ethernet transmission clock enable during Sleep mode
This bit is set and cleared by software.
0: Ethernet transmission clock disabled during sleep mode
1: Ethernet transmission clock enabled during sleep mode
Bit 25 ETHMACLPEN: Ethernet MAC clock enable during Sleep mode
This bit is set and cleared by software.
0: Ethernet MAC clock disabled during Sleep mode
1: Ethernet MAC clock enabled during Sleep mode
Bit 24 Reserved, must be kept at reset value.
Bit 23 DMA2DLPEN: DMA2D clock enable during Sleep mode
This bit is set and cleared by software.
0: DMA2D clock disabled during Sleep mode
1: DMA2D clock enabled during Sleep mode

RM0090 Rev 17

189/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 22 DMA2LPEN: DMA2 clock enable during Sleep mode
This bit is set and cleared by software.
0: DMA2 clock disabled during Sleep mode
1: DMA2 clock enabled during Sleep mode
Bit 21 DMA1LPEN: DMA1 clock enable during Sleep mode
This bit is set and cleared by software.
0: DMA1 clock disabled during Sleep mode
1: DMA1 clock enabled during Sleep mode
Bit 20 Reserved, must be kept at reset value.
Bit 19 SRAM3LPEN: SRAM3 interface clock enable during Sleep mode
This bit is set and cleared by software.
0: SRAM3 interface clock disabled during Sleep mode
1: SRAM3 interface clock enabled during Sleep mode
Bit 18 BKPSRAMLPEN: Backup SRAM interface clock enable during Sleep mode
This bit is set and cleared by software.
0: Backup SRAM interface clock disabled during Sleep mode
1: Backup SRAM interface clock enabled during Sleep mode
Bit 17 SRAM2LPEN: SRAM2 interface clock enable during Sleep mode
This bit is set and cleared by software.
0: SRAM2 interface clock disabled during Sleep mode
1: SRAM2 interface clock enabled during Sleep mode
Bit 16 SRAM1LPEN: SRAM1 interface clock enable during Sleep mode
This bit is set and cleared by software.
0: SRAM1 interface clock disabled during Sleep mode
1: SRAM1 interface clock enabled during Sleep mode
Bit 15 FLITFLPEN: Flash interface clock enable during Sleep mode
This bit is set and cleared by software.
0: Flash interface clock disabled during Sleep mode
1: Flash interface clock enabled during Sleep mode
Bits 14:13 Reserved, must be kept at reset value.
Bit 12 CRCLPEN: CRC clock enable during Sleep mode
This bit is set and cleared by software.
0: CRC clock disabled during Sleep mode
1: CRC clock enabled during Sleep mode
Bit 11 Reserved, must be kept at reset value.
Bit 10 GPIOKLPEN: IO port K clock enable during Sleep mode
This bit is set and cleared by software.
0: IO port K clock disabled during Sleep mode
1: IO port K clock enabled during Sleep mode
Bit 9 GPIOJLPEN: IO port J clock enable during Sleep mode
This bit is set and cleared by software.
0: IO port J clock disabled during Sleep mode
1: IO port J clock enabled during Sleep mode

190/1747

RM0090 Rev 17

RM0090

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 8 GPIOILPEN: IO port I clock enable during Sleep mode
This bit is set and cleared by software.
0: IO port I clock disabled during Sleep mode
1: IO port I clock enabled during Sleep mode
Bit 7

GPIOHLPEN: IO port H clock enable during Sleep mode
This bit is set and cleared by software.
0: IO port H clock disabled during Sleep mode
1: IO port H clock enabled during Sleep mode

Bits 6 GPIOGLPEN: IO port G clock enable during Sleep mode
This bit is set and cleared by software.
0: IO port G clock disabled during Sleep mode
1: IO port G clock enabled during Sleep mode
Bit 5 GPIOFLPEN: IO port F clock enable during Sleep mode
This bit is set and cleared by software.
0: IO port F clock disabled during Sleep mode
1: IO port F clock enabled during Sleep mode
Bit 4 GPIOELPEN: IO port E clock enable during Sleep mode
Set and cleared by software.
0: IO port E clock disabled during Sleep mode
1: IO port E clock enabled during Sleep mode
Bit 3 GPIODLPEN: IO port D clock enable during Sleep mode
This bit is set and cleared by software.
0: IO port D clock disabled during Sleep mode
1: IO port D clock enabled during Sleep mode
Bit 2 GPIOCLPEN: IO port C clock enable during Sleep mode
This bit is set and cleared by software.
0: IO port C clock disabled during Sleep mode
1: IO port C clock enabled during Sleep mode
Bit 1 GPIOBLPEN: IO port B clock enable during Sleep mode
This bit is set and cleared by software.
0: IO port B clock disabled during Sleep mode
1: IO port B clock enabled during Sleep mode
Bit 0 GPIOALPEN: IO port A clock enable during sleep mode
This bit is set and cleared by software.
0: IO port A clock disabled during Sleep mode
1: IO port A clock enabled during Sleep mode

RM0090 Rev 17

191/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

6.3.16

RM0090

RCC AHB2 peripheral clock enable in low power mode register
(RCC_AHB2LPENR)
Address offset: 0x54
Reset value: 0x0000 00F1
Access: no wait state, word, half-word and byte access.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

3

2

1

16

Reserved
15

14

13

12

11

Reserved

10

9

8

7

6

OTGFS RNG
LPEN LPEN
rw

rw

5

4

HASH
LPEN

CRYP
LPEN

rw

rw

Reserved

Bits 31:8 Reserved, must be kept at reset value.
Bit 7 OTGFSLPEN: USB OTG FS clock enable during Sleep mode
This bit is set and cleared by software.
0: USB OTG FS clock disabled during Sleep mode
1: USB OTG FS clock enabled during Sleep mode
Bit 6 RNGLPEN: Random number generator clock enable during Sleep mode
This bit is set and cleared by software.
0: Random number generator clock disabled during Sleep mode
1: Random number generator clock enabled during Sleep mode
Bit 5 HASHLPEN: Hash modules clock enable during Sleep mode
This bit is set and cleared by software.
0: Hash modules clock disabled during Sleep mode
1: Hash modules clock enabled during Sleep mode
Bit 4 CRYPLPEN: Cryptography modules clock enable during Sleep mode
This bit is set and cleared by software.
0: cryptography modules clock disabled during Sleep mode
1: cryptography modules clock enabled during Sleep mode
Bits 3:1 Reserved, must be kept at reset value.
Bit 0 DCMILPEN: Camera interface enable during Sleep mode
This bit is set and cleared by software.
0: Camera interface clock disabled during Sleep mode
1: Camera interface clock enabled during Sleep mode

192/1747

RM0090 Rev 17

0
DCMI
LPEN
rw

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

6.3.17

RCC AHB3 peripheral clock enable in low power mode register
(RCC_AHB3LPENR)
Address offset: 0x58
Reset value: 0x0000 0001
Access: no wait state, word, half-word and byte access.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

6

5

4

3

2

1

16

Reserved
15

14

13

12

11

10

9

8

7

0
FMC
LPEN

Reserved

rw

Bits 31:1Reserved, must be kept at reset value.
FMCLPEN: Flexible memory controller module clock enable during Sleep mode
This bit is set and cleared by software.
Bit 0
0: FMC module clock disabled during Sleep mode
1: FMC module clock enabled during Sleep mode

6.3.18

RCC APB1 peripheral clock enable in low power mode register
(RCC_APB1LPENR)
Address offset: 0x60
Reset value: 0xF6FE C9FF
Access: no wait state, word, half-word and byte access.

31

30

UART8 UART7
LPEN LPEN

29

28

27

26

25

DAC
LPEN

PWR
LPEN

RESER
VED

CAN2
LPEN

CAN1
LPEN

rw

rw

11

10

9

rw

rw

rw

rw

15

14

13

12

SPI3
LPEN

SPI2
LPEN

rw

rw

Reserved

WWDG
LPEN
rw

Reserved

24
Reserved

23

22

21

I2C3
LPEN

I2C2
LPEN

I2C1
LPEN

20

19

UART5 UART4
LPEN
LPEN

18

17

16

USART USART
2
3
LPEN
LPEN

Reserved

rw

rw

rw

rw

rw

rw

rw

8

7

6

5

4

3

2

1

0

TIM14
LPEN

TIM13
LPEN

TIM12
LPEN

TIM7
LPEN

TIM6
LPEN

TIM5
LPEN

TIM4
LPEN

TIM3
LPEN

TIM2
LPEN

rw

rw

rw

rw

rw

rw

rw

rw

rw

RM0090 Rev 17

193/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 31 UART8LPEN: UART8 clock enable during Sleep mode
This bit is set and cleared by software.
0: UART8 clock disabled during Sleep mode
1: UART8 clock enabled during Sleep mode
Bit 30 UART7LPEN: UART7 clock enable during Sleep mode
This bit is set and cleared by software.
0: UART7 clock disabled during Sleep mode
1: UART7 clock enabled during Sleep mode
Bit 29 DACLPEN: DAC interface clock enable during Sleep mode
This bit is set and cleared by software.
0: DAC interface clock disabled during Sleep mode
1: DAC interface clock enabled during Sleep mode
Bit 28 PWRLPEN: Power interface clock enable during Sleep mode
This bit is set and cleared by software.
0: Power interface clock disabled during Sleep mode
1: Power interface clock enabled during Sleep mode
Bit 27 Reserved, must be kept at reset value.
Bit 26 CAN2LPEN: CAN 2 clock enable during Sleep mode
This bit is set and cleared by software.
0: CAN 2 clock disabled during sleep mode
1: CAN 2 clock enabled during sleep mode
Bit 25 CAN1LPEN: CAN 1 clock enable during Sleep mode
This bit is set and cleared by software.
0: CAN 1 clock disabled during Sleep mode
1: CAN 1 clock enabled during Sleep mode
Bit 24 Reserved, must be kept at reset value.
Bit 23 I2C3LPEN: I2C3 clock enable during Sleep mode
This bit is set and cleared by software.
0: I2C3 clock disabled during Sleep mode
1: I2C3 clock enabled during Sleep mode
Bit 22 I2C2LPEN: I2C2 clock enable during Sleep mode
This bit is set and cleared by software.
0: I2C2 clock disabled during Sleep mode
1: I2C2 clock enabled during Sleep mode
Bit 21 I2C1LPEN: I2C1 clock enable during Sleep mode
This bit is set and cleared by software.
0: I2C1 clock disabled during Sleep mode
1: I2C1 clock enabled during Sleep mode
Bit 20 UART5LPEN: UART5 clock enable during Sleep mode
This bit is set and cleared by software.
0: UART5 clock disabled during Sleep mode
1: UART5 clock enabled during Sleep mode
Bit 19 UART4LPEN: UART4 clock enable during Sleep mode
This bit is set and cleared by software.
0: UART4 clock disabled during Sleep mode
1: UART4 clock enabled during Sleep mode

194/1747

RM0090 Rev 17

RM0090

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 18

USART3LPEN: USART3 clock enable during Sleep mode
This bit is set and cleared by software.
0: USART3 clock disabled during Sleep mode
1: USART3 clock enabled during Sleep mode

Bit 17 USART2LPEN: USART2 clock enable during Sleep mode
This bit is set and cleared by software.
0: USART2 clock disabled during Sleep mode
1: USART2 clock enabled during Sleep mode
Bit 16 Reserved, must be kept at reset value.
Bit 15 SPI3LPEN: SPI3 clock enable during Sleep mode
This bit is set and cleared by software.
0: SPI3 clock disabled during Sleep mode
1: SPI3 clock enabled during Sleep mode
Bit 14 SPI2LPEN: SPI2 clock enable during Sleep mode
This bit is set and cleared by software.
0: SPI2 clock disabled during Sleep mode
1: SPI2 clock enabled during Sleep mode
Bits 13:12 Reserved, must be kept at reset value.
Bit 11 WWDGLPEN: Window watchdog clock enable during Sleep mode
This bit is set and cleared by software.
0: Window watchdog clock disabled during sleep mode
1: Window watchdog clock enabled during sleep mode
Bits 10:9 Reserved, must be kept at reset value.
Bit 8 TIM14LPEN: TIM14 clock enable during Sleep mode
This bit is set and cleared by software.
0: TIM14 clock disabled during Sleep mode
1: TIM14 clock enabled during Sleep mode
Bit 7 TIM13LPEN: TIM13 clock enable during Sleep mode
This bit is set and cleared by software.
0: TIM13 clock disabled during Sleep mode
1: TIM13 clock enabled during Sleep mode
Bit 6 TIM12LPEN: TIM12 clock enable during Sleep mode
This bit is set and cleared by software.
0: TIM12 clock disabled during Sleep mode
1: TIM12 clock enabled during Sleep mode
Bit 5 TIM7LPEN: TIM7 clock enable during Sleep mode
This bit is set and cleared by software.
0: TIM7 clock disabled during Sleep mode
1: TIM7 clock enabled during Sleep mode
Bit 4 TIM6LPEN: TIM6 clock enable during Sleep mode
This bit is set and cleared by software.
0: TIM6 clock disabled during Sleep mode
1: TIM6 clock enabled during Sleep mode

RM0090 Rev 17

195/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 3 TIM5LPEN: TIM5 clock enable during Sleep mode
This bit is set and cleared by software.
0: TIM5 clock disabled during Sleep mode
1: TIM5 clock enabled during Sleep mode
Bit 2 TIM4LPEN: TIM4 clock enable during Sleep mode
This bit is set and cleared by software.
0: TIM4 clock disabled during Sleep mode
1: TIM4 clock enabled during Sleep mode
Bit 1 TIM3LPEN: TIM3 clock enable during Sleep mode
This bit is set and cleared by software.
0: TIM3 clock disabled during Sleep mode
1: TIM3 clock enabled during Sleep mode
Bit 0

196/1747

TIM2LPEN: TIM2 clock enable during Sleep mode
This bit is set and cleared by software.
0: TIM2 clock disabled during Sleep mode
1: TIM2 clock enabled during Sleep mode

RM0090 Rev 17

RM0090

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

6.3.19

RCC APB2 peripheral clock enabled in low power mode register
(RCC_APB2LPENR)
Address offset: 0x64
Reset value: 0x0x0477 7F33
Access: no wait state, word, half-word and byte access.

31

30

29

28

27

26

25

LTDC
LPEN

Reserved

24

23

Reserved

22

21

20

SAI1
LPEN

SPI6
LPEN

SPI5
LPEN

rw

rw

rw

rw
15

14

SYSC
FG
ReserLPEN
ved
rw

13

12

11

10

9

8

SPI4
LPEN

SPI1
LPEN

SDIO
LPEN

ADC3
LPEN

ADC2
LPEN

ADC1
LPEN

rw

rw

rw

rw

rw

rw

7

6

Reserved

5

4

USART
6
LPEN

USART
1
LPEN

rw

rw

19
Reserved
3

18

17

16

TIM11
LPEN

TIM10
LPEN

TIM9
LPEN

rw

rw

rw

2

1

0

TIM8
LPEN

TIM1
LPEN

rw

rw

Reserved

Bits 31:27 Reserved, must be kept at reset value.
Bit 26 LTDCLPEN: LTDC clock enable during Sleep mode
This bit is set and cleared by software.
0: LTDC clock disabled during Sleep mode
1: LTDC clock enabled during Sleep mode
Bits 25:23 Reserved, must be kept at reset value.
Bit 22 SAI1LPEN: SAI1 clock enable during Sleep mode
This bit is set and cleared by software.
0: SAI1 clock disabled during Sleep mode
1: SAI1 clock enabled during Sleep mode
Bit 21 SPI6LPEN: SPI6 clock enable during Sleep mode
This bit is set and cleared by software.
0: SPI6 clock disabled during Sleep mode
1: SPI6 clock enabled during Sleep mode
Bit 20 SPI5LPEN: SPI5 clock enable during Sleep mode
This bit is set and cleared by software.
0: SPI5 clock disabled during Sleep mode
1: SPI5 clock enabled during Sleep mode
Bit 19 Reserved, must be kept at reset value.
Bit 18 TIM11LPEN: TIM11 clock enable during Sleep mode
This bit is set and cleared by software.
0: TIM11 clock disabled during Sleep mode
1: TIM11 clock enabled during Sleep mode
Bit 17 TIM10LPEN: TIM10 clock enable during Sleep mode
This bit is set and cleared by software.
0: TIM10 clock disabled during Sleep mode
1: TIM10 clock enabled during Sleep mode

RM0090 Rev 17

197/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 16 TIM9LPEN: TIM9 clock enable during sleep mode
This bit is set and cleared by software.
0: TIM9 clock disabled during Sleep mode
1: TIM9 clock enabled during Sleep mode
Bit 15 Reserved, must be kept at reset value.
Bit 14 SYSCFGLPEN: System configuration controller clock enable during Sleep mode
This bit is set and cleared by software.
0: System configuration controller clock disabled during Sleep mode
1: System configuration controller clock enabled during Sleep mode
Bit 13 SPI4LPEN: SPI4 clock enable during Sleep mode
This bit is set and cleared by software.
0: SPI4 clock disabled during Sleep mode
1: SPI4 clock enabled during Sleep mode
Bit 12 SPI1LPEN: SPI1 clock enable during Sleep mode
This bit is set and cleared by software.
0: SPI1 clock disabled during Sleep mode
1: SPI1 clock enabled during Sleep mode
Bit 11 SDIOLPEN: SDIO clock enable during Sleep mode
This bit is set and cleared by software.
0: SDIO module clock disabled during Sleep mode
1: SDIO module clock enabled during Sleep mode
Bit 10 ADC3LPEN: ADC 3 clock enable during Sleep mode
This bit is set and cleared by software.
0: ADC 3 clock disabled during Sleep mode
1: ADC 3 clock disabled during Sleep mode
Bit 9 ADC2LPEN: ADC2 clock enable during Sleep mode
This bit is set and cleared by software.
0: ADC2 clock disabled during Sleep mode
1: ADC2 clock disabled during Sleep mode
Bit 8 ADC1LPEN: ADC1 clock enable during Sleep mode
This bit is set and cleared by software.
0: ADC1 clock disabled during Sleep mode
1: ADC1 clock disabled during Sleep mode
Bits 7:6 Reserved, must be kept at reset value.
Bit 5 USART6LPEN: USART6 clock enable during Sleep mode
This bit is set and cleared by software.
0: USART6 clock disabled during Sleep mode
1: USART6 clock enabled during Sleep mode
Bit 4 USART1LPEN: USART1 clock enable during Sleep mode
This bit is set and cleared by software.
0: USART1 clock disabled during Sleep mode
1: USART1 clock enabled during Sleep mode

198/1747

RM0090 Rev 17

RM0090

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bits 3:2 Reserved, must be kept at reset value.
Bit 1 TIM8LPEN: TIM8 clock enable during Sleep mode
This bit is set and cleared by software.
0: TIM8 clock disabled during Sleep mode
1: TIM8 clock enabled during Sleep mode
Bit 0 TIM1LPEN: TIM1 clock enable during Sleep mode
This bit is set and cleared by software.
0: TIM1 clock disabled during Sleep mode
1: TIM1 clock enabled during Sleep mode

6.3.20

RCC Backup domain control register (RCC_BDCR)
Address offset: 0x70
Reset value: 0x0000 0000, reset by Backup domain reset.
Access: 0 ≤ wait state ≤ 3, word, half-word and byte access
Wait states are inserted in case of successive accesses to this register.
The LSEON, LSEBYP, RTCSEL and RTCEN bits in the RCC Backup domain control
register (RCC_BDCR) are in the Backup domain. As a result, after Reset, these bits are
write-protected and the DBP bit in the PWR power control register (PWR_CR) for
STM32F42xxx and STM32F43xxx has to be set before these can be modified. Refer to
Section 6.1.1: System reset on page 150 for further information. These bits are only reset
after a Backup domain Reset (see Section 6.1.3: Backup domain reset). Any internal or
external Reset will not have any effect on these bits.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

Reserved
15
RTCEN
rw

14

13

12
Reserved

11

10

9

8

rw
7

RTCSEL[1:0]
rw

16
BDRST

rw

6

5
Reserved

4

3

2

1

0

LSEBY LSERD
LSEON
P
Y
rw

r

rw

Bits 31:17 Reserved, must be kept at reset value.
Bit 16 BDRST: Backup domain software reset
This bit is set and cleared by software.
0: Reset not activated
1: Resets the entire Backup domain
Note: The BKPSRAM is not affected by this reset, the only way of resetting the BKPSRAM is
through the Flash interface when a protection level change from level 1 to level 0 is
requested.
Bit 15 RTCEN: RTC clock enable
This bit is set and cleared by software.
0: RTC clock disabled
1: RTC clock enabled
Bits 14:10 Reserved, must be kept at reset value.

RM0090 Rev 17

199/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

RM0090

Bits 9:8 RTCSEL[1:0]: RTC clock source selection
These bits are set by software to select the clock source for the RTC. Once the RTC clock
source has been selected, it cannot be changed anymore unless the Backup domain is
reset. The BDRST bit can be used to reset them.
00: No clock
01: LSE oscillator clock used as the RTC clock
10: LSI oscillator clock used as the RTC clock
11: HSE oscillator clock divided by a programmable prescaler (selection through the
RTCPRE[4:0] bits in the RCC clock configuration register (RCC_CFGR)) used as the RTC
clock
Bits 7:3 Reserved, must be kept at reset value.
Bit 2 LSEBYP: External low-speed oscillator bypass
This bit is set and cleared by software to bypass the oscillator. This bit can be written only
when the LSE clock is disabled.
0: LSE oscillator not bypassed
1: LSE oscillator bypassed
Bit 1 LSERDY: External low-speed oscillator ready
This bit is set and cleared by hardware to indicate when the external 32 kHz oscillator is
stable. After the LSEON bit is cleared, LSERDY goes low after 6 external low-speed
oscillator clock cycles.
0: LSE clock not ready
1: LSE clock ready
Bit 0 LSEON: External low-speed oscillator enable
This bit is set and cleared by software.
0: LSE clock OFF
1: LSE clock ON

6.3.21

RCC clock control & status register (RCC_CSR)
Address offset: 0x74
Reset value: 0x0E00 0000, reset by system reset, except reset flags by power reset only.
Access: 0 ≤ wait state ≤ 3, word, half-word and byte access
Wait states are inserted in case of successive accesses to this register.

31

30

LPWR WWDG
RSTF RSTF

29

28

27

26

25

24

IWDG
RSTF

SFT
RSTF

POR
RSTF

PIN
RSTF

BORRS
TF

RMVF

r

r

r

r

r

r

r

rt_w

15

14

13

12

11

10

9

8

23

21

20

19

18

17

16

1

0

Reserved
7

Reserved

200/1747

22

RM0090 Rev 17

6

5

4

3

2

LSIRDY LSION
r

rw

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bit 31 LPWRRSTF: Low-power reset flag
This bit is set by hardware when a Low-power management reset occurs.
Cleared by writing to the RMVF bit.
0: No Low-power management reset occurred
1: Low-power management reset occurred
For further information on Low-power management reset, refer to Low-power management
reset.
Bit 30 WWDGRSTF: Window watchdog reset flag
This bit is set by hardware when a window watchdog reset occurs.
Cleared by writing to the RMVF bit.
0: No window watchdog reset occurred
1: Window watchdog reset occurred
Bit 29 IWDGRSTF: Independent watchdog reset flag
This bit is set by hardware when an independent watchdog reset from VDD domain occurs.
Cleared by writing to the RMVF bit.
0: No watchdog reset occurred
1: Watchdog reset occurred
Bit 28 SFTRSTF: Software reset flag
This bit is set by hardware when a software reset occurs.
Cleared by writing to the RMVF bit.
0: No software reset occurred
1: Software reset occurred
Bit 27 PORRSTF: POR/PDR reset flag
This bit is set by hardware when a POR/PDR reset occurs.
Cleared by writing to the RMVF bit.
0: No POR/PDR reset occurred
1: POR/PDR reset occurred
Bit 26 PINRSTF: PIN reset flag
This bit is set by hardware when a reset from the NRST pin occurs.
Cleared by writing to the RMVF bit.
0: No reset from NRST pin occurred
1: Reset from NRST pin occurred
Bit 25 BORRSTF: BOR reset flag
Cleared by software by writing the RMVF bit.
This bit is set by hardware when a POR/PDR or BOR reset occurs.
0: No POR/PDR or BOR reset occurred
1: POR/PDR or BOR reset occurred
Bit 24 RMVF: Remove reset flag
This bit is set by software to clear the reset flags.
0: No effect
1: Clear the reset flags

RM0090 Rev 17

201/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

RM0090

Bits 23:2 Reserved, must be kept at reset value.
Bit 1 LSIRDY: Internal low-speed oscillator ready
This bit is set and cleared by hardware to indicate when the internal RC 40 kHz oscillator is
stable. After the LSION bit is cleared, LSIRDY goes low after 3 LSI clock cycles.
0: LSI RC oscillator not ready
1: LSI RC oscillator ready
Bit 0 LSION: Internal low-speed oscillator enable
This bit is set and cleared by software.
0: LSI RC oscillator OFF
1: LSI RC oscillator ON

6.3.22

RCC spread spectrum clock generation register (RCC_SSCGR)
Address offset: 0x80
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.
The spread spectrum clock generation is available only for the main PLL.
The RCC_SSCGR register must be written either before the main PLL is enabled or after
the main PLL disabled.

Note:

For full details about PLL spread spectrum clock generation (SSCG) characteristics, refer to
the “Electrical characteristics” section in your device datasheet.

31

30

SSCG
EN

SPR
EAD
SEL

rw

rw

15

14

29

28

27

26

25

24

23

rw

21

20

19

18

17

16

INCSTEP

Reserved
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

INCSTEP
rw

22

MODPER
rw

Bit 31 SSCGEN: Spread spectrum modulation enable
This bit is set and cleared by software.
0: Spread spectrum modulation DISABLE. (To write after clearing CR[24]=PLLON bit)
1: Spread spectrum modulation ENABLE. (To write before setting CR[24]=PLLON bit)
Bit 30 SPREADSEL: Spread Select
This bit is set and cleared by software.
To write before to set CR[24]=PLLON bit.
0: Center spread
1: Down spread

202/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bits 29:28 Reserved, must be kept at reset value.
Bits 27:13 INCSTEP: Incrementation step
These bits are set and cleared by software. To write before setting CR[24]=PLLON bit.
Configuration input for modulation profile amplitude.
Bits 12:0 MODPER: Modulation period
These bits are set and cleared by software. To write before setting CR[24]=PLLON bit.
Configuration input for modulation profile period.

6.3.23

RCC PLLI2S configuration register (RCC_PLLI2SCFGR)
Address offset: 0x84
Reset value: 0x2400 3000
Access: no wait state, word, half-word and byte access.
This register is used to configure the PLLI2S clock outputs according to the formulas:
f(VCO clock) = f(PLLI2S clock input) × (PLLI2SN / PLLM)
f(PLL I2S clock output) = f(VCO clock) / PLLI2SR

31
Reserv
ed
15

30

29

28

PLLI2S
R2

PLLI2S
R1

PLLI2S
R0

27

26

25

24

rw

rw

rw

rw

rw

rw

rw

14

13

12

11

10

9

8

23

22

21

PLLI2SQ

20

19

18

17

16

3

2

1

0

Reserved
7

6

PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN
Reserv
8
7
6
5
4
3
2
1
0
ed
rw
rw
rw
rw
rw
rw
rw
rw
rw

RM0090 Rev 17

5

4

Reserved

203/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

RM0090

Bit 31 Reserved, must be kept at reset value.
Bits 30:28 PLLI2SR: PLLI2S division factor for I2S clocks
These bits are set and cleared by software to control the I2S clock frequency. These bits
should be written only if the PLLI2S is disabled. The factor must be chosen in accordance
with the prescaler values inside the I2S peripherals, to reach 0.3% error when using
standard crystals and 0% error with audio crystals. For more information about I2S clock
frequency and precision, refer to Section 28.4.4: Clock generator in the I2S chapter.
Caution: The I2Ss requires a frequency lower than or equal to 192 MHz to work correctly.
I2S clock frequency = VCO frequency / PLLR with 2 ≤PLLR ≤7
000: PLLR = 0, wrong configuration
001: PLLR = 1, wrong configuration
010: PLLR = 2
...
111: PLLR = 7
Bits 27:24 PLLI2SQ: PLLI2S division factor for SAI1 clock
These bits are set and cleared by software to control the SAI1 clock frequency.
They should be written when the PLLI2S is disabled.
SAI1 clock frequency = VCO frequency / PLLI2SQ with 2 <= PLLI2SIQ <= 15
0000: PLLI2SQ = 0, wrong configuration
0001: PLLI2SQ = 1, wrong configuration
0010: PLLI2SQ = 2
0011: PLLI2SQ = 3
0100: PLLI2SQ = 4
0101: PLLI2SQ = 5
...
1111: PLLI2SQ = 15

204/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bits 23:15 Reserved, must be kept at reset value.
Bits 14:6 PLLI2SN: PLLI2S multiplication factor for VCO
These bits are set and cleared by software to control the multiplication factor of the VCO.
These bits can be written only when PLLI2S is disabled. Only half-word and word accesses
are allowed to write these bits.
Caution: The software has to set these bits correctly to ensure that the VCO output
frequency is between 100 and 432 MHz.
VCO output frequency = VCO input frequency × PLLI2SN with 50 ≤PLLI2SN ≤432
000000000: PLLI2SN = 0, wrong configuration
000000001: PLLI2SN = 1, wrong configuration
...
000110010: PLLI2SN = 50
...
001100011: PLLI2SN = 99
001100100: PLLI2SN = 100
001100101: PLLI2SN = 101
001100110: PLLI2SN = 102
...
110110000: PLLI2SN = 432
110110001: PLLI2SN = 433, wrong configuration
...
111111111: PLLI2SN = 511, wrong configuration
Note: Multiplication factors ranging from 50 and 99 are possible for VCO input frequency
higher than 1 MHz. However care must be taken that the minimum VCO output
frequency respects the value specified above.
Bits 5:0 Reserved, must be kept at reset value.

RM0090 Rev 17

205/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

6.3.24

RM0090

RCC PLL configuration register (RCC_PLLSAICFGR)
Address offset: 0x88
Reset value: 0x2400 3000
Access: no wait state, word, half-word and byte access.
This register is used to configure the PLLSAI clock outputs according to the formulas:
•

31
Reserved
15
Reserved

30

f(VCO clock) = f(PLLSAI clock input) × (PLLSAIN / PLLM)

•

f(PLLSAI1 clock output) = f(VCO clock) / PLLSAIQ

•

f(PLL LCD clock output) = f(VCO clock) / PLLSAIR
29

28

27

PLLSAIR

26

25

24

23

22

rw

rw

rw

rw

14

13

12

11

rw

rw

rw

rw

rw

10

9

8

7

6

rw

rw

rw

rw

rw

20

19

5

4

3

17

16

2

1

0

Reserved

Bit 31 Reserved, must be kept at reset value.
Bits 30:28 PLLSAIR: PLLSAI division factor for LCD clock
Set and reset by software to control the LCD clock frequency.
These bits should be written when the PLLSAI is disabled.
LCD clock frequency = VCO frequency / PLLSAIR with 2 ≤PLLSAIR ≤7
000: PLLSAIR = 0, wrong configuration
001: PLLSAIR = 1, wrong configuration
010: PLLSAIR = 2
...
111: PLLSAIR = 7
Bits 27:24 PLLSAIQ: PLLSAI division factor for SAI1 clock
Set and reset by software to control the frequency of SAI1 clock.
These bits should be written when the PLLSAI is disabled.
SAI1 clock frequency = VCO frequency / PLLSAIQ with 2 ≤PLLSAIQ ≤15
0000: PLLSAIQ = 0, wrong configuration
0001: PLLSAIQ = 1, wrong configuration
...
0010: PLLSAIQ = 2
0011: PLLSAIQ = 3
0100: PLLSAIQ = 4
0101: PLLSAIQ = 5
...
1111: PLLSAIQ = 15

206/1747

18

Reserved

rw

PLLSAIN
rw

21

PLLSAIQ

RM0090 Rev 17

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bits 23:15 Reserved, must be kept at reset value.
Bits 14:6 PLLSAIN: PLLSAI division factor for VCO
These bits are set and cleared by software to control the multiplication factor of the VCO.
These bits can be written only when PLLSAI is disabled. Only half-word and word accesses
are allowed to write these bits.
Caution: The software has to set these bits correctly to ensure that the VCO output frequency
is between 100 and 432 MHz.
VCO output frequency = VCO input frequency × PLLISAIN with 50 ≤PLLISAIN ≤432
000000000: PLLISAIN = 0, wrong configuration
000000001: PLLISAIN = 1, wrong configuration
...
000110010: PLLISAIN = 50
...
001100011: PLLISAIN = 99
001100100: PLLISAIN = 100
001100101: PLLISAIN = 101
001100110: PLLISAIN = 102
...
110110000: PLLISAIN = 432
110110001: PLLISAIN = 433, wrong configuration
...
111111111: PLLISAIN = 511, wrong configuration
Note: Multiplication factors ranging from 50 and 99 are possible for VCO input frequency
higher than 1 MHz. However care must be taken that the minimum VCO output
frequency respects the value specified above.
Bits 5:0 Reserved, must be kept at reset value

6.3.25

RCC Dedicated Clock Configuration Register (RCC_DCKCFGR)
Address offset: 0x8C
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.
The RCC_DCKCFGR register allows to configure the timer clock prescalers and the PLLSAI
and PLLI2S output clock dividers for SAI1 and LTDC peripherals according to the following
formula:
f(PLLSAIDIVQ clock output) = f(PLLSAI_Q) / PLLSAIDIVQ
f(PLLSAIDIVR clock output) = f(PLLSAI_R) / PLLSAIDIVR
f(PLLI2SDIVQ clock output) = f(PLLI2S_Q) / PLLI2SDIVQ

31

30

29

28

27

26

25

TIMPRE

Reserved
15

14
Reserved

13

12

24

11

10

9

rw

rw

rw

22

SAI1BSRC

21

20

SAI1ASRC

rw

rw

rw

rw

rw

8

7

6

5

4

PLLSAIDIVQ
rw

23

rw

Reserved

RM0090 Rev 17

19

18

Reserved
3

2

17

16

PLLSAIDIVR
rw

rw

1

0

rw

rw

PLLS2DIVQ
rw

rw

rw

207/1747
212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

RM0090

Bits 31:25 Reserved, must be kept at reset value.
Bit 24 TIMPRE: Timers clocks prescalers selection
This bit is set and reset by software to control the clock frequency of all the timers connected
to APB1 and APB2 domain.
0: If the APB prescaler (PPRE1, PPRE2 in the RCC_CFGR register) is configured to a
division factor of 1, TIMxCLK = PCLKx. Otherwise, the timer clock frequencies are set to
twice to the frequency of the APB domain to which the timers are connected:
TIMxCLK = 2xPCLKx.
1:If the APB prescaler (PPRE1, PPRE2 in the RCC_CFGR register) is configured to a
division factor of 1, 2 or 4, TIMxCLK = HCLK. Otherwise, the timer clock frequencies are set
to four times to the frequency of the APB domain to which the timers are connected:
TIMxCLK = 4xPCLKx.
Bits 23:22 SAI1BSRC: SAI1-B clock source selection
These bits are set and cleared by software to control the SAI1-B clock frequency.
They should be written when the PLLSAI and PLLI2S are disabled.
00: SAI1-B clock frequency = f(PLLSAI_Q) / PLLSAIDIVQ
01: SAI1-B clock frequency = f(PLLI2S_Q) / PLLI2SDIVQ
10: SAI1-B clock frequency = Alternate function input frequency
11: wrong configuration
Bits 21:20 SAI1ASRC: SAI1-A clock source selection
These bits are set and cleared by software to control the SAI1-A clock frequency.
They should be written when the PLLSAI and PLLI2S are disabled.
00: SAI1-A clock frequency = f(PLLSAI_Q) / PLLSAIDIVQ
01: SAI1-A clock frequency = f(PLLI2S_Q) / PLLI2SDIVQ
10: SAI1-A clock frequency = Alternate function input frequency
11: wrong configuration
Bits 19: 18 Reserved, must be kept at reset value.
Bits 17:16 PLLSAIDIVR: division factor for LCD_CLK
These bits are set and cleared by software to control the frequency of LCD_CLK.
They should be written only if PLLSAI is disabled.
LCD_CLK frequency = f(PLLSAI_R) / PLLSAIDIVR with 2 ≤PLLSAIDIVR ≤16
00: PLLSAIDIVR = /2
01: PLLSAIDIVR = /4
10: PLLSAIDIVR = /8
11: PLLSAIDIVR = /16
Bits 15: 13 Reserved, must be kept at reset value.

208/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Bits 12:8 PLLSAIDIVQ: PLLSAI division factor for SAI1 clock
These bits are set and reset by software to control the SAI1 clock frequency.
They should be written only if PLLSAI is disabled.
SAI1 clock frequency = f(PLLSAI_Q) / PLLSAIDIVQ with 1 ≤PLLSAIDIVQ ≤31
00000: PLLSAIDIVQ = /1
00001: PLLSAIDIVQ = /2
00010: PLLSAIDIVQ = /3
00011: PLLSAIDIVQ = /4
00100: PLLSAIDIVQ = /5
...
11111: PLLSAIDIVQ = /32
Bits 7:5 Reserved, must be kept at reset value.
Bits 4:0 PLLI2SDIVQ: PLLI2S division factor for SAI1 clock
These bits are set and reset by software to control the SAI1 clock frequency.
They should be written only if PLLI2S is disabled.
SAI1 clock frequency = f(PLLI2S_Q) / PLLI2SDIVQ with 1 <= PLLI2SDIVQ <= 31
00000: PLLI2SDIVQ = /1
00001: PLLI2SDIVQ = /2
00010: PLLI2SDIVQ = /3
00011: PLLI2SDIVQ = /4
00100: PLLI2SDIVQ = /5
...
11111: PLLI2SDIVQ = /32

RM0090 Rev 17

209/1747
212

RCC_PLLCFG
R

0x08
RCC_CFGR

0x0C
RCC_CIR

0x10
RCC_AHB1RS
TR

0x14
RCC_AHB2RS
TR

0x18
RCC_AHB3RS
TR

0x1C
Reserved

0x20
RCC_APB1RS
TR

0x24
RCC_APB2RS
TR

0x28
0x2C
Reserved
Reserved

0x30

RCC_
AHB1ENR

0x34

RCC_
AHB2ENR

0x38

RCC_
AHB3ENR

0x3C

Reserved

210/1747

OTGHSEN

RM0090 Rev 17

Reserved

Reserved
GPIOCRST

GPIOAEN

GPIOBEN

TIM3RST
TIM2RST

TIM8RST
TIM1RST

FMCRST DCMIRST

Reserved

TIM4RST

Reserved

GPIODRST

TIM5RST

GPIOCEN

CRYPRST GPIOERST

USART1RST TIM6RST

GPIOEEN

GPIOARST

GPIOBRST

LSIRDYF

LSERDYF

HSIRDYF

HSERDYF

PLLRDYF

GPIOFRST PLLI2SRDYF

GPIODEN

CRYPEN

CSSF

LSIRDYIE

LSERDYIE

HSIRDYIE

HSERDYIE

CSSON

HSEON

HSERDY

HSEBYP

SW 0

PLLM 0

HSION

HSIRDY

Reserved

PLLM 2

SWS 0

PLLM 1

PLLM 3 HSITRIM 0

SWS 1
SW 1

PLLM 4 HSITRIM 1

HPRE 0

PLLM 5 HSITRIM 2

PLLN 0 HSITRIM 3
HPRE 1

HPRE 2

HSICAL 0

HSICAL 1

HSICAL 2

HSICAL 3

HSICAL 4

HSICAL 5

HSICAL 6

PLLN 1 HSITRIM 4

PLLN 2

PLLN 3

PLLN 4

PLLN 5

PLLN 6

PLLN 7

PLLN 8

HPRE 3

Reserved

PPRE1 0

PPRE1 2
PPRE1 1

PLLI2SRDYIE PPRE2 0
PLLRDYIE

PLL ON

PLL RDY

PPRE2 2 Reserved HSICAL 7

RTCPRE 0 PLLP 0

PLLSAIRDYIE PPRE2 1

Reserved

LSIRDYC

RTCPRE 1 PLLP 1

RTCPRE 2

LSERDYC

RTCPRE 3

HSIRDYC

PLLRDYC

PLLSRC

MCO1 1
MCO1 0
RTCPRE 4

Reserved

I2SSRC

MCO1PRE1 PLLQ 1
MCO1PRE0 PLLQ 0

MCO1PRE2 PLLQ 2 PLL I2SON

HSERDYC

GPIOGRST PLLSAIRDYF

RNGRST
HSAHRST

GPIOFEN

HASHEN

Reserved

OTGFSRST GPIOHRST

TIM13RST
TIM12RST

GPIOGEN

RNGEN

GPIOIRST

TIM14RST

GPIOJRST

ADCRST

GPIOKRST

Reserved

CRCRST

Reserved

Reserved

Reserved

Reserved

USART6RST TIM7RST

GPIOHEN

GPIOIEN

GPIOJEN

GPIOKEN

SDIORST WWDGRST

Reserved

Reserved

SPI1RST

CRCEN

OTGFSEN

Reserved
SP45RST

Reserved

FMCEN DCMIEN

SPI3RST

Reserved

Reserved

Reserved

Reserved
Reserved
SYSCFGRST SPI2RST

Reserved

TIM9RST

TIM10RST UART2RST

TIM11RST UART3RST

UART4RST

DMA1RST PLLI2SRDYC

DMA2DRST

CSSC
DMA2RST PLLSAIRDYC

Reserved

ETHMACRST

PLL SAION

PLL SAIRDY

MCO2PRE0 PLLQ 3 PLL I2SRDY

Reserved

Reserved

BKPSRAMEN

UART5RST

SPI5RST

I2C1RST

I2C2RST

Reserved

SPI6RST

SAI1RST

I2C3RST

Reserved

CAN1RST

CAN2RST

Reserved

MCO2PRE1

MCO2PRE2

Reserved

Reserved

CCMDATARAMEN

DMA1EN

DMA2EN

Reserved

LTDCRST

Reserved

OTGHSRST
Reserved

DMA2DEN

Reserved

ETHMACEN

Reserved

ETHMACTXEN

DACRST
PWRRST

MCO2 0

Reserved

ETHMACRXEN

ETHMACPTPEN

UART7RST

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

0x04
Reserved

RCC_CR

MCO2 1

0x00

Reserved

Register
name

UART8RST

Addr.
offset

Reserved

6.3.26

OTGHSULPIEN

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)
RM0090

RCC register map
Table 33 gives the register map and reset values.

Table 33. RCC register map and reset values for STM32F42xxx and STM32F43xxx

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
UART8EN

0x44
RCC_
APB2ENR

0x48
0x4C
Reserved
Reserved

0x50
RCC_AHB1LP
ENR

0x54
RCC_AHB2LP
ENR

0x58
RCC_AHB3LP
ENR

0x5C
Reserved

0x60
RCC_APB1LP
ENR

0x64
RCC_APB2LP
ENR

0x68
0x6C
Reserved
Reserved

0x70
RCC_BDCR

0x74
RCC_CSR

0x78
0x7C

Reserved
Reserved

0x80

RCC_SSCGR

0x84

RCC_PLLI2SC
FGR

Reserved

SPREADSEL

PLLI2SRx
RMVF

BORRSTF
Reserved

PLLI2SQ

Reserved

RM0090 Rev 17
RTCEN

BDRST

Reserved
I2C1LPEN
UART5LPEN
UART4LPEN

SPI6LPEN
SPI5LPEN
Reserved

Reserved
Reserved
SPI3LPEN

Reserved

INCSTEP

PLLI2SNx
Reserved
LSERDY
LSEON

LSIRDY
LSION

Reserved

CRYPLPEN

USART1LPEN TIM6LPEN

TIM3LPEN
TIM2LPEN

TIM1LPEN

TIM4LPEN
TIM8LPEN

Reserved

FMCLPEN DCMILPEN

Reserved

HASHLPEN

USART6LPEN TIM7LPEN

GPIOALPEN

GPIOBLPEN

GPIOCLPEN

GPIODLPEN

GPIOELPEN

GPIOFLPEN

GPIOGLPEN

RNGLPEN

TIM5LPEN

SPI1EN

SPI4EN

TIM1EN

TIM8EN

Reserved

USART1EN

USART6EN

Reserved

ADC1EN

ADC2EN

SDIOEN

TIM12LPEN

SPI3EN

Reserved

TIM2EN

TIM3EN

TIM4EN

TIM5EN

TIM6EN

TIM7EN

TIM12EN

TIM13EN

TIM14EN

Reserved

WWDGEN

Reserved

SYSCFGEN SPI2EN

Reserved

TIM9EN

TIM10EN USART2EN

ADC3EN
GPIOILPEN

Reserved

UART4EN

Reserved

TIM11EN USART3EN

I2C1EN
UART5EN

SPI5EN

I2C2EN

I2C3EN

Reserved

CAN1EN

CAN2EN

Reserved

SPI6EN

Reserved
GPIOJLPEN

DACEN
PWREN

SAI1EN

Reserved

LTDCEN

GPIOKLPEN

CRCLPEN

Reserved

FLITFLPEN

SRAM1LPEN

SRAM2LPEN

BKPSRAMLPEN

SRAM3LPEN

Reserved

DMA1LPEN

DMA2LPEN

DMA2DLPEN

Reserved

ETHMACLPEN

GPIOHLPEN

Reserved
OTGFSLPEN

Reserved

TIM13LPEN

ADC1LPEN

RTCSEL 0

Reserved

WWDGLPEN

Reserved

Reserved

TIM14LPEN

ADC2LPEN

RTCSEL 1

Reserved
ADC3LPEN

SDIOLPEN

SPI1LPEN

SPI4LPEN

SYSCFGLPEN SPI2LPEN

Reserved

TIM9LPEN

TIM10LPEN USART2LPEN

TIM11LPEN USART3LPEN

I2C2LPEN

I2C3LPEN

Reserved

CAN1LPEN

ETHMACTXLPEN

ETHMACRXLPEN

ETHMACPTPLPEN

OTGHSLPEN

OTGHSULPILPEN

Reserved
Reserved

LSEBYP

Reserved
CAN2LPEN

SAI1LPEN

Reserved

LTDCLPEN

PWRLPEN

DACLPEN

UART7LPEN
Reserved

PADRSTF

PORRSTF

SFTRSTF

WDGRSTF

WWDGRSTF

UART7EN

RCC_
APB1ENR

Reserved

0x40

UART8LPEN

Register
name

LPWRRSTF

Addr.
offset

Reserved SSCGEN

RM0090
Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

Table 33. RCC register map and reset values for STM32F42xxx and STM32F43xxx (continued)

Reserved
Reserved

MODPER

Reserved

211/1747

212

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)

RM0090

Register
name

0x88

RCC_PLLSAI
CFGR

0x8C

RCC_DCKCF
GR

Reserved

Reserved

PLLSAIN
PLLSAIDIVR

Reserved

SAI1ASCR

SAI1BSCR

TIMPRE

PLLSAIQ

Reserved

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Addr.
offset

PLLSAIR

Table 33. RCC register map and reset values for STM32F42xxx and STM32F43xxx (continued)

Reserved

PLLSAIDIVQ

Reserved

Refer to Section 2.3: Memory map for the register boundary addresses.

212/1747

RM0090 Rev 17

PLLI2SDIVQ

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

7

Reset and clock control for
STM32F405xx/07xx and STM32F415xx/17xx(RCC)

7.1

Reset
There are three types of reset, defined as system Reset, power Reset and backup domain
Reset.

7.1.1

System reset
A system reset sets all registers to their reset values except the reset flags in the clock
controller CSR register and the registers in the Backup domain (see Figure 4).
A system reset is generated when one of the following events occurs:
1.

A low level on the NRST pin (external reset)

2.

Window watchdog end of count condition (WWDG reset)

3.

Independent watchdog end of count condition (IWDG reset)

4.

A software reset (SW reset) (see Software reset)

5.

Low-power management reset (see Low-power management reset)

Software reset
The reset source can be identified by checking the reset flags in the RCC clock control &
status register (RCC_CSR).
The SYSRESETREQ bit in Cortex®-M4 with FPU Application Interrupt and Reset Control
Register must be set to force a software reset on the device. Refer to the Cortex®-M4 with
FPU technical reference manual for more details.

RM0090 Rev 17

213/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

RM0090

Low-power management reset
There are two ways of generating a low-power management reset:
1.

Reset generated when entering the Standby mode:
This type of reset is enabled by resetting the nRST_STDBY bit in the user option bytes.
In this case, whenever a Standby mode entry sequence is successfully executed, the
device is reset instead of entering the Standby mode.

2.

Reset when entering the Stop mode:
This type of reset is enabled by resetting the nRST_STOP bit in the user option bytes.
In this case, whenever a Stop mode entry sequence is successfully executed, the
device is reset instead of entering the Stop mode.

7.1.2

Power reset
A power reset is generated when one of the following events occurs:
1.

Power-on/power-down reset (POR/PDR reset) or brownout (BOR) reset

2.

When exiting the Standby mode

A power reset sets all registers to their reset values except the Backup domain (see
Figure 4)
These sources act on the NRST pin and it is always kept low during the delay phase. The
RESET service routine vector is fixed at address 0x0000_0004 in the memory map.
The system reset signal provided to the device is output on the NRST pin. The pulse
generator guarantees a minimum reset pulse duration of 20 µs for each internal reset
source. In case of an external reset, the reset pulse is generated while the NRST pin is
asserted low.
Figure 20. Simplified diagram of the reset circuit
9''9''$

538
([WHUQDO
UHVHW

)LOWHU

6\VWHPUHVHW

1567
3XOVH
JHQHUDWRU
PLQ—V

::'*UHVHW
,:'*UHVHW
3RZHUUHVHW
6RIWZDUHUHVHW
/RZSRZHUPDQDJHPHQWUHVHW
DLF

The Backup domain has two specific resets that affect only the Backup domain (see
Figure 4).

7.1.3

Backup domain reset
The backup domain reset sets all RTC registers and the RCC_BDCR register to their reset
values. The BKPSRAM is not affected by this reset. The only way of resetting the
BKPSRAM is through the Flash interface by requesting a protection level change from 1 to
0.

214/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)
A backup domain reset is generated when one of the following events occurs:

7.2

1.

Software reset, triggered by setting the BDRST bit in the RCC Backup domain control
register (RCC_BDCR).

2.

VDD or VBAT power on, if both supplies have previously been powered off.

Clocks
Three different clock sources can be used to drive the system clock (SYSCLK):
•

HSI oscillator clock

•

HSE oscillator clock

•

Main PLL (PLL) clock

The devices have the two following secondary clock sources:
•

32 kHz low-speed internal RC (LSI RC) which drives the independent watchdog and,
optionally, the RTC used for Auto-wakeup from the Stop/Standby mode.

•

32.768 kHz low-speed external crystal (LSE crystal) which optionally drives the RTC
clock (RTCCLK)

Each clock source can be switched on or off independently when it is not used, to optimize
power consumption.

RM0090 Rev 17

215/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

RM0090

Figure 21. Clock tree
tĂƚĐŚĚŽŐ
ĞŶĂďůĞ

>^/

>^/ Z
ϯϮ Ŭ,nj

/t'><

ƚŽ ŝŶĚĞƉĞŶĚĞŶƚ
ǁĂƚĐŚĚŽŐ

Zd ^  >΀ϭ͗Ϭ΁
K^ϯϮͺ/E
K^ϯϮͺKhd

DKϮ

Zd><

^z^><
,^ͺZd

ͬϭƚŽϱ

WĞƌŝƉŚĞƌĂů
ĐůŽĐŬĞŶĂďůĞ

ͬϮ ƚŽ ϯϭ

ϭϲ D,nj
,^ / Z 

ϭϲϴD,njŵĂdž͘
ůŽĐŬ
ŶĂďůĞ
ͬϴ

,^/
,^/
,^

^t

^z^><
ϭϲϴD,nj
ŵĂdž

K^  ͺ/E

Wdž
WZ^
ͬϭ͕Ϯ͕ϰ͕ϴ͕ϭϲ

WĞƌŝƉŚĞƌĂů
ĐůŽĐŬĞŶĂďůĞ
WĞƌŝƉŚĞƌĂů
ĐůŽĐŬĞŶĂďůĞ

ŝĨ;WdžƉƌĞƐĐсϭdžϭ
ĞůƐĞdžϮ

,^

ϰͲϮϲ D,nj
,^ K ^ 

ƚŚĞƌŶĞƚ
WdWĐůŽĐŬ
,><
ƚŽ,ďƵƐ͕ĐŽƌĞ͕
ŵĞŵŽƌLJĂŶĚD
ƚŽŽƌƚĞdž^LJƐƚĞŵ
ƚŝŵĞƌ
&><ŽƌƚĞdž
ĨƌĞĞͲƌƵŶŶŝŶŐĐůŽĐŬ

,
WZ^
ͬϭ͕Ϯ͕͘͘ϱϭϮ

W>>><

K^ ͺKhd

ƚŽZd

ͬϭƚŽϱ

>^
DKϭ

Zd
ĞŶĂďůĞ

>^

>^ K^ 
ϯϮ͘ϳϲϴ Ŭ,nj

Wdž
ƉĞƌŝƉŚĞƌĂů
ĐůŽĐŬƐ
WdžƚŝŵĞƌ
ĐůŽĐŬƐ

ͬD
sK

ͬW
ͬY

W>>ϰϴ<

WĞƌŝƉŚĞƌĂů
ĐůŽĐŬĞŶĂďůĞ

ϰϴD,nj
ĐůŽĐŬƐ

WĞƌŝƉŚĞƌĂů
ĐůŽĐŬĞŶĂďůĞ

/Ϯ^ĐůŽĐŬƐ

džE
ͬZ

W>>

sK

ͬW

/Ϯ^^Z

ͬY
džE

W>>/Ϯ^

ͬZ

W>>/Ϯ^><
džƚ͘ĐůŽĐŬ

/Ϯ^ͺ<ͺD//
W,zƚŚĞƌŶĞƚ

WĞƌŝƉŚĞƌĂů
ĐůŽĐŬĞŶĂďůĞ

Ddy><

ϮϱƚŽϱϬD,nj
ͬϮ͕ϮϬ

d,ͺD//ͺZyͺ
><ͺD//
h^Ϯ͘ϬW,z
ϮϰƚŽϲϬD,nj

D//ͺZD//ͺ^>ŝŶ^z^&'ͺWD
WĞƌŝƉŚĞƌĂů
ĐůŽĐŬĞŶĂďůĞ
WĞƌŝƉŚĞƌĂů
ĐůŽĐŬĞŶĂďůĞ

Kd'ͺ,^ͺh>W/ͺ<

ƚŽƚŚĞƌŶĞƚD
DZy><
DZD//><

WĞƌŝƉŚĞƌĂů
ĐůŽĐŬĞŶĂďůĞ

h^,^
h>W/ĐůŽĐŬ
ĂŝϭϲϬϴϴĚ

1. For full details about the internal and external clock source characteristics, refer to the Electrical characteristics section in
the device datasheet.

216/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)
The clock controller provides a high degree of flexibility to the application in the choice of the
external crystal or the oscillator to run the core and peripherals at the highest frequency
and, guarantee the appropriate frequency for peripherals that need a specific clock like
Ethernet, USB OTG FS and HS, I2S and SDIO.
Several prescalers are used to configure the AHB frequency, the high-speed APB (APB2)
and the low-speed APB (APB1) domains. The maximum frequency of the AHB domain is
168 MHz. The maximum allowed frequency of the high-speed APB2 domain is 84 MHz. The
maximum allowed frequency of the low-speed APB1 domain is 42 MHz
All peripheral clocks are derived from the system clock (SYSCLK) except for:
•

The USB OTG FS clock (48 MHz), the random analog generator (RNG) clock
(≤48 MHz) and the SDIO clock (≤48 MHz) which are coming from a specific output of
PLL (PLL48CLK)

•

The I2S clock
To achieve high-quality audio performance, the I2S clock can be derived either from a
specific PLL (PLLI2S) or from an external clock mapped on the I2S_CKIN pin. For
more information about I2S clock frequency and precision, refer to Section 28.4.4:
Clock generator.

•

The USB OTG HS (60 MHz) clock which is provided from the external PHY

•

The Ethernet MAC clocks (TX, RX and RMII) which are provided from the external
PHY. For further information on the Ethernet configuration, please refer to
Section 33.4.4: MII/RMII selection in the Ethernet peripheral description. When the
Ethernet is used, the AHB clock frequency must be at least 25 MHz.

The RCC feeds the external clock of the Cortex System Timer (SysTick) with the AHB clock
(HCLK) divided by 8. The SysTick can work either with this clock or with the Cortex clock
(HCLK), configurable in the SysTick control and status register.
The timer clock frequencies are automatically set by hardware. There are two cases:
1.

If the APB prescaler is 1, the timer clock frequencies are set to the same frequency as
that of the APB domain to which the timers are connected.

2.

Otherwise, they are set to twice (×2) the frequency of the APB domain to which the
timers are connected.

FCLK acts as Cortex®-M4 with FPU free-running clock. For more details, refer to the
Cortex®-M4 with FPU technical reference manual.

7.2.1

HSE clock
The high speed external clock signal (HSE) can be generated from two possible clock
sources:
•

HSE external crystal/ceramic resonator

•

HSE external user clock

The resonator and the load capacitors have to be placed as close as possible to the
oscillator pins in order to minimize output distortion and startup stabilization time. The
loading capacitance values must be adjusted according to the selected oscillator.

RM0090 Rev 17

217/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

RM0090

Figure 22. HSE/ LSE clock sources
Hardware configuration

OSC_OUT
External clock

(HI-Z)
External
source

OSC_IN OSC_OUT
Crystal/ceramic
resonators
CL1

Load
capacitors

CL2

External source (HSE bypass)
In this mode, an external clock source must be provided. You select this mode by setting the
HSEBYP and HSEON bits in the RCC clock control register (RCC_CR). The external clock
signal (square, sinus or triangle) with ~50% duty cycle has to drive the OSC_IN pin while the
OSC_OUT pin should be left HI-Z. See Figure 22.

External crystal/ceramic resonator (HSE crystal)
The HSE has the advantage of producing a very accurate rate on the main clock.
The associated hardware configuration is shown in Figure 22. Refer to the electrical
characteristics section of the datasheet for more details.
The HSERDY flag in the RCC clock control register (RCC_CR) indicates if the high-speed
external oscillator is stable or not. At startup, the clock is not released until this bit is set by
hardware. An interrupt can be generated if enabled in the RCC clock interrupt register
(RCC_CIR).
The HSE Crystal can be switched on and off using the HSEON bit in the RCC clock control
register (RCC_CR).

7.2.2

HSI clock
The HSI clock signal is generated from an internal 16 MHz RC oscillator and can be used
directly as a system clock, or used as PLL input.
The HSI RC oscillator has the advantage of providing a clock source at low cost (no external
components). It also has a faster startup time than the HSE crystal oscillator however, even
with calibration the frequency is less accurate than an external crystal oscillator or ceramic
resonator.

218/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Calibration
RC oscillator frequencies can vary from one chip to another due to manufacturing process
variations, this is why each device is factory calibrated by ST for 1% accuracy at TA= 25 °C.
After reset, the factory calibration value is loaded in the HSICAL[7:0] bits in the RCC clock
control register (RCC_CR).
If the application is subject to voltage or temperature variations this may affect the RC
oscillator speed. You can trim the HSI frequency in the application using the HSITRIM[4:0]
bits in the RCC clock control register (RCC_CR).
The HSIRDY flag in the RCC clock control register (RCC_CR) indicates if the HSI RC is
stable or not. At startup, the HSI RC output clock is not released until this bit is set by
hardware.
The HSI RC can be switched on and off using the HSION bit in the RCC clock control
register (RCC_CR).
The HSI signal can also be used as a backup source (Auxiliary clock) if the HSE crystal
oscillator fails. Refer to Section 7.2.7: Clock security system (CSS) on page 220.

7.2.3

PLL configuration
The STM32F4xx devices feature two PLLs:
•

•

A main PLL (PLL) clocked by the HSE or HSI oscillator and featuring two different
output clocks:
–

The first output is used to generate the high speed system clock (up to 168 MHz)

–

The second output is used to generate the clock for the USB OTG FS (48 MHz),
the random analog generator (≤48 MHz) and the SDIO (≤48 MHz).

A dedicated PLL (PLLI2S) used to generate an accurate clock to achieve high-quality
audio performance on the I2S interface.

Since the main-PLL configuration parameters cannot be changed once PLL is enabled, it is
recommended to configure PLL before enabling it (selection of the HSI or HSE oscillator as
PLL clock source, and configuration of division factors M, N, P, and Q).
The PLLI2S uses the same input clock as PLL (PLLM[5:0] and PLLSRC bits are common to
both PLLs). However, the PLLI2S has dedicated enable/disable and division factors (N and
R) configuration bits. Once the PLLI2S is enabled, the configuration parameters cannot be
changed.
The two PLLs are disabled by hardware when entering Stop and Standby modes, or when
an HSE failure occurs when HSE or PLL (clocked by HSE) are used as system clock. RCC
PLL configuration register (RCC_PLLCFGR) and RCC clock configuration register
(RCC_CFGR) can be used to configure PLL and PLLI2S, respectively.

7.2.4

LSE clock
The LSE clock is generated from a 32.768 kHz low-speed external crystal or ceramic
resonator. It has the advantage providing a low-power but highly accurate clock source to
the real-time clock peripheral (RTC) for clock/calendar or other timing functions.
The LSE oscillator is switched on and off using the LSEON bit in RCC Backup domain
control register (RCC_BDCR).

RM0090 Rev 17

219/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

RM0090

The LSERDY flag in the RCC Backup domain control register (RCC_BDCR) indicates if the
LSE crystal is stable or not. At startup, the LSE crystal output clock signal is not released
until this bit is set by hardware. An interrupt can be generated if enabled in the RCC clock
interrupt register (RCC_CIR).

External source (LSE bypass)
In this mode, an external clock source must be provided. It must have a frequency up to
1 MHz. You select this mode by setting the LSEBYP and LSEON bits in the RCC Backup
domain control register (RCC_BDCR). The external clock signal (square, sinus or triangle)
with ~50% duty cycle has to drive the OSC32_IN pin while the OSC32_OUT pin should be
left HI-Z. See Figure 22.

7.2.5

LSI clock
The LSI RC acts as an low-power clock source that can be kept running in Stop and
Standby mode for the independent watchdog (IWDG) and Auto-wakeup unit (AWU). The
clock frequency is around 32 kHz. For more details, refer to the electrical characteristics
section of the datasheets.
The LSI RC can be switched on and off using the LSION bit in the RCC clock control &
status register (RCC_CSR).
The LSIRDY flag in the RCC clock control & status register (RCC_CSR) indicates if the lowspeed internal oscillator is stable or not. At startup, the clock is not released until this bit is
set by hardware. An interrupt can be generated if enabled in the RCC clock interrupt register
(RCC_CIR).

7.2.6

System clock (SYSCLK) selection
After a system reset, the HSI oscillator is selected as the system clock. When a clock source
is used directly or through PLL as the system clock, it is not possible to stop it.
A switch from one clock source to another occurs only if the target clock source is ready
(clock stable after startup delay or PLL locked). If a clock source that is not yet ready is
selected, the switch occurs when the clock source is ready. Status bits in the RCC clock
control register (RCC_CR) indicate which clock(s) is (are) ready and which clock is currently
used as the system clock.

7.2.7

Clock security system (CSS)
The clock security system can be activated by software. In this case, the clock detector is
enabled after the HSE oscillator startup delay, and disabled when this oscillator is stopped.
If a failure is detected on the HSE clock, this oscillator is automatically disabled, a clock
failure event is sent to the break inputs of advanced-control timers TIM1 and TIM8, and an
interrupt is generated to inform the software about the failure (clock security system
interrupt CSSI), allowing the MCU to perform rescue operations. The CSSI is linked to the
Cortex®-M4 with FPU NMI (non-maskable interrupt) exception vector.

Note:

220/1747

When the CSS is enabled, if the HSE clock happens to fail, the CSS generates an interrupt,
which causes the automatic generation of an NMI. The NMI is executed indefinitely unless
the CSS interrupt pending bit is cleared. As a consequence, the application has to clear the
CSS interrupt in the NMI ISR by setting the CSSC bit in the Clock interrupt register
(RCC_CIR).

RM0090 Rev 17

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)
If the HSE oscillator is used directly or indirectly as the system clock (indirectly meaning that
it is directly used as PLL input clock, and that PLL clock is the system clock) and a failure is
detected, then the system clock switches to the HSI oscillator and the HSE oscillator is
disabled.
If the HSE oscillator clock was the clock source of PLL used as the system clock when the
failure occurred, PLL is also disabled. In this case, if the PLLI2S was enabled, it is also
disabled when the HSE fails.

7.2.8

RTC/AWU clock
Once the RTCCLK clock source has been selected, the only possible way of modifying the
selection is to reset the power domain.
The RTCCLK clock source can be either the HSE 1 MHz (HSE divided by a programmable
prescaler), the LSE or the LSI clock. This is selected by programming the RTCSEL[1:0] bits
in the RCC Backup domain control register (RCC_BDCR) and the RTCPRE[4:0] bits in RCC
clock configuration register (RCC_CFGR). This selection cannot be modified without
resetting the Backup domain.
If the LSE is selected as the RTC clock, the RTC operates normally if the backup or the
system supply disappears. If the LSI is selected as the AWU clock, the AWU state is not
guaranteed if the system supply disappears. If the HSE oscillator divided by a value
between 2 and 31 is used as the RTC clock, the RTC state is not guaranteed if the backup
or the system supply disappears.
The LSE clock is in the Backup domain, whereas the HSE and LSI clocks are not. As a
consequence:
•

If LSE is selected as the RTC clock:
–

•

If LSI is selected as the Auto-wakeup unit (AWU) clock:
–

•

The RTC continues to work even if the VDD supply is switched off, provided the
VBAT supply is maintained.
The AWU state is not guaranteed if the VDD supply is powered off. Refer to
Section 7.2.5: LSI clock on page 220 for more details on LSI calibration.

If the HSE clock is used as the RTC clock:
–

The RTC state is not guaranteed if the VDD supply is powered off or if the internal
voltage regulator is powered off (removing power from the 1.2 V domain).

Note:

To read the RTC calendar register when the APB1 clock frequency is less than seven times
the RTC clock frequency (fAPB1 < 7xfRTCLCK), the software must read the calendar time and
date registers twice. The data are correct if the second read access to RTC_TR gives the
same result than the first one. Otherwise a third read access must be performed.

7.2.9

Watchdog clock
If the independent watchdog (IWDG) is started by either hardware option or software
access, the LSI oscillator is forced ON and cannot be disabled. After the LSI oscillator
temporization, the clock is provided to the IWDG.

RM0090 Rev 17

221/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

7.2.10

RM0090

Clock-out capability
Two microcontroller clock output (MCO) pins are available:
•

MCO1
You can output four different clock sources onto the MCO1 pin (PA8) using the
configurable prescaler (from 1 to 5):
–

HSI clock

–

LSE clock

–

HSE clock

–

PLL clock

The desired clock source is selected using the MCO1PRE[2:0] and MCO1[1:0] bits in
the RCC clock configuration register (RCC_CFGR).
•

MCO2
You can output four different clock sources onto the MCO2 pin (PC9) using the
configurable prescaler (from 1 to 5):
–

HSE clock

–

PLL clock

–

System clock (SYSCLK)

–

PLLI2S clock

The desired clock source is selected using the MCO2PRE[2:0] and MCO2 bits in the
RCC clock configuration register (RCC_CFGR).
For the different MCO pins, the corresponding GPIO port has to be programmed in alternate
function mode.
The selected clock to output onto MCO must not exceed 100 MHz (the maximum I/O
speed).

7.2.11

Internal/external clock measurement using TIM5/TIM11
It is possible to indirectly measure the frequencies of all on-board clock source generators
by means of the input capture of TIM5 channel4 and TIM11 channel1 as shown in Figure 23
and Figure 24.

Internal/external clock measurement using TIM5 channel4
TIM5 has an input multiplexer which allows choosing whether the input capture is triggered
by the I/O or by an internal clock. This selection is performed through the TI4_RMP [1:0] bits
in the TIM5_OR register.
The primary purpose of having the LSE connected to the channel4 input capture is to be
able to precisely measure the HSI (this requires to have the HSI used as the system clock
source). The number of HSI clock counts between consecutive edges of the LSE signal
provides a measurement of the internal clock period. Taking advantage of the high precision
of LSE crystals (typically a few tens of ppm) we can determine the internal clock frequency
with the same resolution, and trim the source to compensate for manufacturing-process
and/or temperature- and voltage-related frequency deviations.
The HSI oscillator has dedicated, user-accessible calibration bits for this purpose.

222/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)
The basic concept consists in providing a relative measurement (e.g. HSI/LSE ratio): the
precision is therefore tightly linked to the ratio between the two clock sources. The greater
the ratio, the better the measurement.
It is also possible to measure the LSI frequency: this is useful for applications that do not
have a crystal. The ultralow-power LSI oscillator has a large manufacturing process
deviation: by measuring it versus the HSI clock source, it is possible to determine its
frequency with the precision of the HSI. The measured value can be used to have more
accurate RTC time base timeouts (when LSI is used as the RTC clock source) and/or an
IWDG timeout with an acceptable accuracy.
Use the following procedure to measure the LSI frequency:
1.

Enable the TIM5 timer and configure channel4 in Input capture mode.

2.

Set the TI4_RMP bits in the TIM5_OR register to 0x01 to connect the LSI clock
internally to TIM5 channel4 input capture for calibration purposes.

3.

Measure the LSI clock frequency using the TIM5 capture/compare 4 event or interrupt.

4.

Use the measured LSI frequency to update the prescaler of the RTC depending on the
desired time base and/or to compute the IWDG timeout.
Figure 23. Frequency measurement with TIM5 in Input capture mode
4)-

4)?2-0;=
'0)/
24#?7AKE5P?)4
,3%
,3)

4)

AI6

Internal/external clock measurement using TIM11 channel1
TIM11 has an input multiplexer which allows choosing whether the input capture is triggered
by the I/O or by an internal clock. This selection is performed through TI1_RMP [1:0] bits in
the TIM11_OR register. The HSE_RTC clock (HSE divided by a programmable prescaler) is
connected to channel 1 input capture to have a rough indication of the external crystal
frequency. This requires that the HSI is the system clock source. This can be useful for
instance to ensure compliance with the IEC 60730/IEC 61335 standards which require to be
able to determine harmonic or subharmonic frequencies (–50/+100% deviations).
Figure 24. Frequency measurement with TIM11 in Input capture mode
4)-

4)?2-0;=
'0)/
4)
(3%?24# -(Z
AI

RM0090 Rev 17

223/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

7.3

RM0090

RCC registers
Refer to Section 1.1: List of abbreviations for registers for a list of abbreviations used in
register descriptions.

7.3.1

RCC clock control register (RCC_CR)
Address offset: 0x00
Reset value: 0x0000 XX83 where X is undefined.
Access: no wait state, word, half-word and byte access

31

30

29

28

14

13

26

25

PLLI2S PLLI2S PLLRD
RDY
ON
Y

Reserved
15

27

12

24

23

22

r

r

r

Reserved

r

rw

r

rw

11

10

9

8

r

20

PLLON

7

6

HSICAL[7:0]
r

21

5

4

19

18

17

16

CSS
ON

HSE
BYP

HSE
RDY

HSE
ON

rw

rw

r

rw

3

2

1

0
HSION

Res.

HSI
RDY
r

rw

HSITRIM[4:0]
r

r

r

rw

rw

rw

rw

rw

Bits 31:28 Reserved, must be kept at reset value.
Bit 27 PLLI2SRDY: PLLI2S clock ready flag
Set by hardware to indicate that the PLLI2S is locked.
0: PLLI2S unlocked
1: PLLI2S locked
Bit 26 PLLI2SON: PLLI2S enable
Set and cleared by software to enable PLLI2S.
Cleared by hardware when entering Stop or Standby mode.
0: PLLI2S OFF
1: PLLI2S ON
Bit 25 PLLRDY: Main PLL (PLL) clock ready flag
Set by hardware to indicate that PLL is locked.
0: PLL unlocked
1: PLL locked
Bit 24 PLLON: Main PLL (PLL) enable
Set and cleared by software to enable PLL.
Cleared by hardware when entering Stop or Standby mode. This bit cannot be reset if PLL
clock is used as the system clock.
0: PLL OFF
1: PLL ON
Bits 23:20 Reserved, must be kept at reset value.
Bit 19 CSSON: Clock security system enable
Set and cleared by software to enable the clock security system. When CSSON is set, the
clock detector is enabled by hardware when the HSE oscillator is ready, and disabled by
hardware if an oscillator failure is detected.
0: Clock security system OFF (Clock detector OFF)
1: Clock security system ON (Clock detector ON if HSE oscillator is stable, OFF if not)

224/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bit 18 HSEBYP: HSE clock bypass
Set and cleared by software to bypass the oscillator with an external clock. The external
clock must be enabled with the HSEON bit, to be used by the device.
The HSEBYP bit can be written only if the HSE oscillator is disabled.
0: HSE oscillator not bypassed
1: HSE oscillator bypassed with an external clock
Bit 17 HSERDY: HSE clock ready flag
Set by hardware to indicate that the HSE oscillator is stable. After the HSEON bit is cleared,
HSERDY goes low after 6 HSE oscillator clock cycles.
0: HSE oscillator not ready
1: HSE oscillator ready
Bit 16 HSEON: HSE clock enable
Set and cleared by software.
Cleared by hardware to stop the HSE oscillator when entering Stop or Standby mode. This
bit cannot be reset if the HSE oscillator is used directly or indirectly as the system clock.
0: HSE oscillator OFF
1: HSE oscillator ON
Bits 15:8 HSICAL[7:0]: Internal high-speed clock calibration
These bits are initialized automatically at startup.
Bits 7:3 HSITRIM[4:0]: Internal high-speed clock trimming
These bits provide an additional user-programmable trimming value that is added to the
HSICAL[7:0] bits. It can be programmed to adjust to variations in voltage and temperature
that influence the frequency of the internal HSI RC.
Bit 2 Reserved, must be kept at reset value.
Bit 1 HSIRDY: Internal high-speed clock ready flag
Set by hardware to indicate that the HSI oscillator is stable. After the HSION bit is cleared,
HSIRDY goes low after 6 HSI clock cycles.
0: HSI oscillator not ready
1: HSI oscillator ready
Bit 0 HSION: Internal high-speed clock enable
Set and cleared by software.
Set by hardware to force the HSI oscillator ON when leaving the Stop or Standby mode or in
case of a failure of the HSE oscillator used directly or indirectly as the system clock. This bit
cannot be cleared if the HSI is used directly or indirectly as the system clock.
0: HSI oscillator OFF
1: HSI oscillator ON

RM0090 Rev 17

225/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

7.3.2

RM0090

RCC PLL configuration register (RCC_PLLCFGR)
Address offset: 0x04
Reset value: 0x2400 3010
Access: no wait state, word, half-word and byte access.
This register is used to configure the PLL clock outputs according to the formulas:

31

30

•

f(VCO clock) = f(PLL clock input) × (PLLN / PLLM)

•

f(PLL general clock output) = f(VCO clock) / PLLP

•

f(USB OTG FS, SDIO, RNG clock output) = f(VCO clock) / PLLQ
29

28

27

PLLQ3 PLLQ2

Reserved

rw
15
Reserv
ed

14

13

26

12

11

25

24

PLLQ1

PLLQ0

rw

rw

rw

10

9

8

23

22

7

6

PLLN
rw

rw

rw

rw

rw

21

PLLSR
Reserv
C
ed
rw

20

19

18

Reserved
5

4

3

2

PLLM5 PLLM4 PLLM3 PLLM2
rw

rw

rw

rw

rw

rw

rw

rw

17

16

PLLP1

PLLP0

rw

rw

1

0

PLLM1

PLLM0

rw

rw

Bits 31:28 Reserved, must be kept at reset value.
Bits 27:24 PLLQ: Main PLL (PLL) division factor for USB OTG FS, SDIO and random number generator
clocks
Set and cleared by software to control the frequency of USB OTG FS clock, the random
number generator clock and the SDIO clock. These bits should be written only if PLL is
disabled.
Caution: The USB OTG FS requires a 48 MHz clock to work correctly. The SDIO and the
random number generator need a frequency lower than or equal to 48 MHz to work
correctly.
USB OTG FS clock frequency = VCO frequency / PLLQ with 2 ≤PLLQ ≤15
0000: PLLQ = 0, wrong configuration
0001: PLLQ = 1, wrong configuration
0010: PLLQ = 2
0011: PLLQ = 3
0100: PLLQ = 4
...
1111: PLLQ = 15
Bit 23 Reserved, must be kept at reset value.
Bit 22 PLLSRC: Main PLL(PLL) and audio PLL (PLLI2S) entry clock source
Set and cleared by software to select PLL and PLLI2S clock source. This bit can be written
only when PLL and PLLI2S are disabled.
0: HSI clock selected as PLL and PLLI2S clock entry
1: HSE oscillator clock selected as PLL and PLLI2S clock entry
Bits 21:18 Reserved, must be kept at reset value.

226/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bits 17:16 PLLP: Main PLL (PLL) division factor for main system clock
Set and cleared by software to control the frequency of the general PLL output clock. These
bits can be written only if PLL is disabled.
Caution: The software has to set these bits correctly not to exceed 168 MHz on this domain.
PLL output clock frequency = VCO frequency / PLLP with PLLP = 2, 4, 6, or 8
00: PLLP = 2
01: PLLP = 4
10: PLLP = 6
11: PLLP = 8
Bits 14:6 PLLN: Main PLL (PLL) multiplication factor for VCO
Set and cleared by software to control the multiplication factor of the VCO. These bits can
be written only when PLL is disabled. Only half-word and word accesses are allowed to
write these bits.
Caution: The software has to set these bits correctly to ensure that the VCO output
frequency is between 100 and 432 MHz.
VCO output frequency = VCO input frequency × PLLN with 50 ≤PLLN ≤432
000000000: PLLN = 0, wrong configuration
000000001: PLLN = 1, wrong configuration
...
000110010: PLLN = 50
...
001100011: PLLN = 99
001100100: PLLN = 100
...
110110000: PLLN = 432
110110001: PLLN = 433, wrong configuration
...
111111111: PLLN = 511, wrong configuration
Note: Multiplication factors ranging from 50 and 99 are possible for VCO input frequency
higher than 1 MHz. However care must be taken that the minimum VCO output
frequency respects the value specified above.
Bits 5:0 PLLM: Division factor for the main PLL (PLL) and audio PLL (PLLI2S) input clock
Set and cleared by software to divide the PLL and PLLI2S input clock before the VCO.
These bits can be written only when the PLL and PLLI2S are disabled.
Caution: The software has to set these bits correctly to ensure that the VCO input frequency
ranges from 1 to 2 MHz. It is recommended to select a frequency of 2 MHz to limit
PLL jitter.
VCO input frequency = PLL input clock frequency / PLLM with 2 ≤PLLM ≤63
000000: PLLM = 0, wrong configuration
000001: PLLM = 1, wrong configuration
000010: PLLM = 2
000011: PLLM = 3
000100: PLLM = 4
...
111110: PLLM = 62
111111: PLLM = 63

RM0090 Rev 17

227/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

7.3.3

RM0090

RCC clock configuration register (RCC_CFGR)
Address offset: 0x08
Reset value: 0x0000 0000
Access: 0 ≤ wait state ≤ 2, word, half-word and byte access
1 or 2 wait states inserted only if the access occurs during a clock source switch.

31

30

29

MCO2

14

26

rw

25

24

23

22

I2SSC
R

MCO1 PRE[2:0]

rw

rw

rw

rw

rw

rw

rw

rw

12

11

10

9

8

7

6

rw

rw

rw

rw

PPRE1[2:0]
rw

rw

Reserved

21

20

19

MCO1

13

PPRE2[2:0]
rw

27

MCO2 PRE[2:0]

rw
15

28

18

17

16

RTCPRE[4:0]

5

rw

rw

rw

rw

rw

4

3

2

1

0

SWS1

SWS0

SW1

SW0

r

r

rw

rw

HPRE[3:0]
rw

rw

Bits 31:30 MCO2[1:0]: Microcontroller clock output 2
Set and cleared by software. Clock source selection may generate glitches on MCO2. It is
highly recommended to configure these bits only after reset before enabling the external
oscillators and the PLLs.
00: System clock (SYSCLK) selected
01: PLLI2S clock selected
10: HSE oscillator clock selected
11: PLL clock selected
Bits 27:29 MCO2PRE: MCO2 prescaler
Set and cleared by software to configure the prescaler of the MCO2. Modification of this
prescaler may generate glitches on MCO2. It is highly recommended to change this
prescaler only after reset before enabling the external oscillators and the PLLs.
0xx: no division
100: division by 2
101: division by 3
110: division by 4
111: division by 5
Bits 24:26 MCO1PRE: MCO1 prescaler
Set and cleared by software to configure the prescaler of the MCO1. Modification of this
prescaler may generate glitches on MCO1. It is highly recommended to change this
prescaler only after reset before enabling the external oscillators and the PLL.
0xx: no division
100: division by 2
101: division by 3
110: division by 4
111: division by 5
Bit 23 I2SSRC: I2S clock selection
Set and cleared by software. This bit allows to select the I2S clock source between the
PLLI2S clock and the external clock. It is highly recommended to change this bit only after
reset and before enabling the I2S module.
0: PLLI2S clock used as I2S clock source
1: External clock mapped on the I2S_CKIN pin used as I2S clock source

228/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bits 22:21 MCO1: Microcontroller clock output 1
Set and cleared by software. Clock source selection may generate glitches on MCO1. It is
highly recommended to configure these bits only after reset before enabling the external
oscillators and PLL.
00: HSI clock selected
01: LSE oscillator selected
10: HSE oscillator clock selected
11: PLL clock selected
Bits 20:16 RTCPRE: HSE division factor for RTC clock
Set and cleared by software to divide the HSE clock input clock to generate a 1 MHz clock
for RTC.
Caution: The software has to set these bits correctly to ensure that the clock supplied to the
RTC is 1 MHz. These bits must be configured if needed before selecting the RTC
clock source.
00000: no clock
00001: no clock
00010: HSE/2
00011: HSE/3
00100: HSE/4
...
11110: HSE/30
11111: HSE/31
Bits 15:13 PPRE2: APB high-speed prescaler (APB2)
Set and cleared by software to control APB high-speed clock division factor.
Caution: The software has to set these bits correctly not to exceed 84 MHz on this domain.
The clocks are divided with the new prescaler factor from 1 to 16 AHB cycles after
PPRE2 write.
0xx: AHB clock not divided
100: AHB clock divided by 2
101: AHB clock divided by 4
110: AHB clock divided by 8
111: AHB clock divided by 16
Bits 12:10 PPRE1: APB Low speed prescaler (APB1)
Set and cleared by software to control APB low-speed clock division factor.
Caution: The software has to set these bits correctly not to exceed 42 MHz on this domain.
The clocks are divided with the new prescaler factor from 1 to 16 AHB cycles after
PPRE1 write.
0xx: AHB clock not divided
100: AHB clock divided by 2
101: AHB clock divided by 4
110: AHB clock divided by 8
111: AHB clock divided by 16
Bits 9:8 Reserved, must be kept at reset value.

RM0090 Rev 17

229/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

RM0090

Bits 7:4 HPRE: AHB prescaler
Set and cleared by software to control AHB clock division factor.
Caution: The clocks are divided with the new prescaler factor from 1 to 16 AHB cycles after
HPRE write.
Caution: The AHB clock frequency must be at least 25 MHz when the Ethernet is used.
0xxx: system clock not divided
1000: system clock divided by 2
1001: system clock divided by 4
1010: system clock divided by 8
1011: system clock divided by 16
1100: system clock divided by 64
1101: system clock divided by 128
1110: system clock divided by 256
1111: system clock divided by 512
Bits 3:2 SWS: System clock switch status
Set and cleared by hardware to indicate which clock source is used as the system clock.
00: HSI oscillator used as the system clock
01: HSE oscillator used as the system clock
10: PLL used as the system clock
11: not applicable
Bits 1:0 SW: System clock switch
Set and cleared by software to select the system clock source.
Set by hardware to force the HSI selection when leaving the Stop or Standby mode or in
case of failure of the HSE oscillator used directly or indirectly as the system clock.
00: HSI oscillator selected as system clock
01: HSE oscillator selected as system clock
10: PLL selected as system clock
11: not allowed

7.3.4

RCC clock interrupt register (RCC_CIR)
Address offset: 0x0C
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access

31

30

29

28

27

26

25

24

23
CSSC

Reserved

w
15

14

Reserved

13
PLLI2S
RDYIE
rw

230/1747

12

11

10

PLL
HSE
HSI
RDYIE RDYIE RDYIE
rw

rw

rw

22

21

PLLI2S
Reserv RDYC
ed
w

9

8

7

6

LSE
RDYIE

LSI
RDYIE

CSSF

rw

rw

r

Reserv
ed

RM0090 Rev 17

20

19

18

17

16

PLL
RDYC

HSE
RDYC

HSI
RDYC

LSE
RDYC

LSI
RDYC

w

w

w

w

w

5

4

3

2

1

0

PLLI2S
RDYF

PLL
RDYF

HSE
RDYF

HSI
RDYF

LSE
RDYF

LSI
RDYF

r

r

r

r

r

r

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bits 31:24 Reserved, must be kept at reset value.
Bit 23 CSSC: Clock security system interrupt clear
This bit is set by software to clear the CSSF flag.
0: No effect
1: Clear CSSF flag
Bits 22 Reserved, must be kept at reset value.
Bit 21 PLLI2SRDYC: PLLI2S ready interrupt clear
This bit is set by software to clear the PLLI2SRDYF flag.
0: No effect
1: PLLI2SRDYF cleared
Bit 20 PLLRDYC: Main PLL(PLL) ready interrupt clear
This bit is set by software to clear the PLLRDYF flag.
0: No effect
1: PLLRDYF cleared
Bit 19 HSERDYC: HSE ready interrupt clear
This bit is set by software to clear the HSERDYF flag.
0: No effect
1: HSERDYF cleared
Bit 18 HSIRDYC: HSI ready interrupt clear
This bit is set software to clear the HSIRDYF flag.
0: No effect
1: HSIRDYF cleared
Bit 17 LSERDYC: LSE ready interrupt clear
This bit is set by software to clear the LSERDYF flag.
0: No effect
1: LSERDYF cleared
Bit 16 LSIRDYC: LSI ready interrupt clear
This bit is set by software to clear the LSIRDYF flag.
0: No effect
1: LSIRDYF cleared
Bits 15:12 Reserved, must be kept at reset value.
Bit 13 PLLI2SRDYIE: PLLI2S ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by PLLI2S lock.
0: PLLI2S lock interrupt disabled
1: PLLI2S lock interrupt enabled
Bit 12 PLLRDYIE: Main PLL (PLL) ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by PLL lock.
0: PLL lock interrupt disabled
1: PLL lock interrupt enabled
Bit 11 HSERDYIE: HSE ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by the HSE oscillator
stabilization.
0: HSE ready interrupt disabled
1: HSE ready interrupt enabled

RM0090 Rev 17

231/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

RM0090

Bit 10 HSIRDYIE: HSI ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by the HSI oscillator
stabilization.
0: HSI ready interrupt disabled
1: HSI ready interrupt enabled
Bit 9 LSERDYIE: LSE ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by the LSE oscillator
stabilization.
0: LSE ready interrupt disabled
1: LSE ready interrupt enabled
Bit 8 LSIRDYIE: LSI ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by LSI oscillator
stabilization.
0: LSI ready interrupt disabled
1: LSI ready interrupt enabled
Bit 7 CSSF: Clock security system interrupt flag
Set by hardware when a failure is detected in the HSE oscillator.
Cleared by software setting the CSSC bit.
0: No clock security interrupt caused by HSE clock failure
1: Clock security interrupt caused by HSE clock failure
Bits 6 Reserved, must be kept at reset value.
Bit 5 PLLI2SRDYF: PLLI2S ready interrupt flag
Set by hardware when the PLLI2S locks and PLLI2SRDYDIE is set.
Cleared by software setting the PLLRI2SDYC bit.
0: No clock ready interrupt caused by PLLI2S lock
1: Clock ready interrupt caused by PLLI2S lock
Bit 4 PLLRDYF: Main PLL (PLL) ready interrupt flag
Set by hardware when PLL locks and PLLRDYDIE is set.
Cleared by software setting the PLLRDYC bit.
0: No clock ready interrupt caused by PLL lock
1: Clock ready interrupt caused by PLL lock
Bit 3 HSERDYF: HSE ready interrupt flag
Set by hardware when External High Speed clock becomes stable and HSERDYDIE is set.
Cleared by software setting the HSERDYC bit.
0: No clock ready interrupt caused by the HSE oscillator
1: Clock ready interrupt caused by the HSE oscillator

232/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bit 2 HSIRDYF: HSI ready interrupt flag
Set by hardware when the Internal High Speed clock becomes stable and HSIRDYDIE is
set.
Cleared by software setting the HSIRDYC bit.
0: No clock ready interrupt caused by the HSI oscillator
1: Clock ready interrupt caused by the HSI oscillator
Bit 1 LSERDYF: LSE ready interrupt flag
Set by hardware when the External Low Speed clock becomes stable and LSERDYDIE is
set.
Cleared by software setting the LSERDYC bit.
0: No clock ready interrupt caused by the LSE oscillator
1: Clock ready interrupt caused by the LSE oscillator
Bit 0 LSIRDYF: LSI ready interrupt flag
Set by hardware when the internal low speed clock becomes stable and LSIRDYDIE is set.
Cleared by software setting the LSIRDYC bit.
0: No clock ready interrupt caused by the LSI oscillator
1: Clock ready interrupt caused by the LSI oscillator

7.3.5

RCC AHB1 peripheral reset register (RCC_AHB1RSTR)
Address offset: 0x10
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31

30

29

28

OTGH
S
RST

Reserved

27

26

25
ETHMAC
RST

Reserved

rw
15

14
Reserved

13

24

23

Reserved

22

21

DMA2
RST

DMA1
RST

rw

rw

6

5

rw
12
CRCR
ST
rw

11

10
Reserved

9

8
GPIOI
RST
rw

7

20

19

18

17

16

Reserved

4

3

2

1

0

GPIOH GPIOGG GPIOF GPIOE GPIOD GPIOC GPIOB
RST
RST
RST
RST
RST
RST
RST
rw

rw

rw

rw

rw

rw

rw

GPIOA
RST
rw

Bits 31:30 Reserved, must be kept at reset value.
Bit 29 OTGHSRST: USB OTG HS module reset
Set and cleared by software.
0: does not reset the USB OTG HS module
1: resets the USB OTG HS module
Bits 28:26 Reserved, must be kept at reset value.
Bit 25 ETHMACRST: Ethernet MAC reset
Set and cleared by software.
0: does not reset Ethernet MAC
1: resets Ethernet MAC
Bits 24:23 Reserved, must be kept at reset value.

RM0090 Rev 17

233/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bit 22 DMA2RST: DMA2 reset
Set and cleared by software.
0: does not reset DMA2
1: resets DMA2
Bit 21 DMA1RST: DMA1 reset
Set and cleared by software.
0: does not reset DMA1
1: resets DMA1
Bits 20:13 Reserved, must be kept at reset value.
Bit 12 CRCRST: CRC reset
Set and cleared by software.
0: does not reset CRC
1: resets CRC
Bits 11:9 Reserved, must be kept at reset value.
Bit 8 GPIOIRST: IO port I reset
Set and cleared by software.
0: does not reset IO port I
1: resets IO port I
Bit 7 GPIOHRST: IO port H reset
Set and cleared by software.
0: does not reset IO port H
1: resets IO port H
Bits 6 GPIOGRST: IO port G reset
Set and cleared by software.
0: does not reset IO port G
1: resets IO port G
Bit 5 GPIOFRST: IO port F reset
Set and cleared by software.
0: does not reset IO port F
1: resets IO port F
Bit 4 GPIOERST: IO port E reset
Set and cleared by software.
0: does not reset IO port E
1: resets IO port E
Bit 3 GPIODRST: IO port D reset
Set and cleared by software.
0: does not reset IO port D
1: resets IO port D

234/1747

RM0090 Rev 17

RM0090

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bit 2 GPIOCRST: IO port C reset
Set and cleared by software.
0: does not reset IO port C
1: resets IO port C
Bit 1 GPIOBRST: IO port B reset
Set and cleared by software.
0: does not reset IO port B
1:resets IO port B
Bit 0 GPIOARST: IO port A reset
Set and cleared by software.
0: does not reset IO port A
1: resets IO port A

RM0090 Rev 17

235/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

7.3.6

RM0090

RCC AHB2 peripheral reset register (RCC_AHB2RSTR)
Address offset: 0x14
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

3

2

1

16

Reserved
15

14

13

12

11

Reserved

10

9

8

7

6

5

4

OTGFS
RST

RNG
RST

HASH
RST

CRYP
RST

rw

rw

rw

rw

Bits 31:8 Reserved, must be kept at reset value.
Bit 7 OTGFSRST: USB OTG FS module reset
Set and cleared by software.
0: does not reset the USB OTG FS module
1: resets the USB OTG FS module
Bit 6 RNGRST: Random number generator module reset
Set and cleared by software.
0: does not reset the random number generator module
1: resets the random number generator module
Bit 5 HASHRST: Hash module reset
Set and cleared by software.
0: does not reset the HASH module
1: resets the HASH module
Bit 4 CRYPRST: Cryptographic module reset
Set and cleared by software.
0: does not reset the cryptographic module
1: resets the cryptographic module
Bits 3:1 Reserved, must be kept at reset value.
Bit 0 DCMIRST: Camera interface reset
Set and cleared by software.
0: does not reset the Camera interface
1: resets the Camera interface

236/1747

RM0090 Rev 17

Reserved

0
DCMI
RST
rw

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

7.3.7

RCC AHB3 peripheral reset register (RCC_AHB3RSTR)
Address offset: 0x18
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

Reserved
15

14

13

12

11

10

9

8

7

FSMCRST

Reserved

rw

Bits 31:1 Reserved, must be kept at reset value.
Bit 0 FSMCRST: Flexible static memory controller module reset
Set and cleared by software.
0: does not reset the FSMC module
1: resets the FSMC module

7.3.8

RCC APB1 peripheral reset register (RCC_APB1RSTR)
Address offset: 0x20
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31

30

Reserved

15

14

SPI3
RST

SPI2
RST

rw

rw

29

28

PWR
DACRST
RST
rw

rw

13

12

Reserved

27
Reserved
11
WWDG
RST
rw

26

25

CAN2
RST

CAN1
RST
rw

10

9

Reserved

24

23

22

21

I2C3
RST

I2C2
RST

I2C1
RST

rw

rw

rw

8

7

6

5

4

3

2

1

0

TIM14
RST

TIM13
RST

TIM12
RST

TIM7
RST

TIM6
RST

TIM5
RST

TIM4
RST

TIM3
RST

TIM2
RST

rw

rw

rw

rw

rw

rw

rw

rw

rw

Reserved

20

19

18

17

16

UART5 UART4 UART3 UART2
ReserRST
RST
RST
RST
ved
rw
rw
rw
rw

Bits 31:30 Reserved, must be kept at reset value.
Bit 29 DACRST: DAC reset
Set and cleared by software.
0: does not reset the DAC interface
1: resets the DAC interface
Bit 28 PWRRST: Power interface reset
Set and cleared by software.
0: does not reset the power interface
1: resets the power interface
Bit 27 Reserved, must be kept at reset value.

RM0090 Rev 17

237/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bit 26 CAN2RST: CAN2 reset
Set and cleared by software.
0: does not reset CAN2
1: resets CAN2
Bit 25 CAN1RST: CAN1 reset
Set and cleared by software.
0: does not reset CAN1
1: resets CAN1
Bit 24 Reserved, must be kept at reset value.
Bit 23 I2C3RST: I2C3 reset
Set and cleared by software.
0: does not reset I2C3
1: resets I2C3
Bit 22 I2C2RST: I2C2 reset
Set and cleared by software.
0: does not reset I2C2
1: resets I2C2
Bit 21 I2C1RST: I2C1 reset
Set and cleared by software.
0: does not reset I2C1
1: resets I2C1
Bit 20 UART5RST: UART5 reset
Set and cleared by software.
0: does not reset UART5
1: resets UART5
Bit 19 UART4RST: USART4 reset
Set and cleared by software.
0: does not reset UART4
1: resets UART4
Bit 18 USART3RST: USART3 reset
Set and cleared by software.
0: does not reset USART3
1: resets USART3
Bit 17 USART2RST: USART2 reset
Set and cleared by software.
0: does not reset USART2
1: resets USART2
Bit 16 Reserved, must be kept at reset value.
Bit 15 SPI3RST: SPI3 reset
Set and cleared by software.
0: does not reset SPI3
1: resets SPI3
Bit 14 SPI2RST: SPI2 reset
Set and cleared by software.
0: does not reset SPI2
1: resets SPI2

238/1747

RM0090 Rev 17

RM0090

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bits 13:12 Reserved, must be kept at reset value.
Bit 11 WWDGRST: Window watchdog reset
Set and cleared by software.
0: does not reset the window watchdog
1: resets the window watchdog
Bits 10:9 Reserved, must be kept at reset value.
Bit 8 TIM14RST: TIM14 reset
Set and cleared by software.
0: does not reset TIM14
1: resets TIM14
Bit 7 TIM13RST: TIM13 reset
Set and cleared by software.
0: does not reset TIM13
1: resets TIM13
Bit 6 TIM12RST: TIM12 reset
Set and cleared by software.
0: does not reset TIM12
1: resets TIM12
Bit 5 TIM7RST: TIM7 reset
Set and cleared by software.
0: does not reset TIM7
1: resets TIM7
Bit 4 TIM6RST: TIM6 reset
Set and cleared by software.
0: does not reset TIM6
1: resets TIM6
Bit 3 TIM5RST: TIM5 reset
Set and cleared by software.
0: does not reset TIM5
1: resets TIM5
Bit 2 TIM4RST: TIM4 reset
Set and cleared by software.
0: does not reset TIM4
1: resets TIM4
Bit 1 TIM3RST: TIM3 reset
Set and cleared by software.
0: does not reset TIM3
1: resets TIM3
Bit 0 TIM2RST: TIM2 reset
Set and cleared by software.
0: does not reset TIM2
1: resets TIM2

RM0090 Rev 17

239/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

7.3.9

RM0090

RCC APB2 peripheral reset register (RCC_APB2RSTR)
Address offset: 0x24
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31

30

29

28

27

26

25

24

23

22

21

20

19

Reserved
15
Reserved

14

13

SYSCF
G RST Reserved
rw

12

11

SPI1
RST

SDIO
RST

rw

rw

10

9

Reserved

8
ADC
RST

7

6

Reserved

rw

5

4

USART
6
RST

USART
1
RST

rw

rw

Bits 31:19 Reserved, must be kept at reset value.
Bit 18 TIM11RST: TIM11 reset
Set and cleared by software.
0: does not reset TIM11
1: resets TIM14
Bit 17 TIM10RST: TIM10 reset
Set and cleared by software.
0: does not reset TIM10
1: resets TIM10
Bit 16 TIM9RST: TIM9 reset
Set and cleared by software.
0: does not reset TIM9
1: resets TIM9
Bit 15 Reserved, must be kept at reset value.
Bit 14 SYSCFGRST: System configuration controller reset
Set and cleared by software.
0: does not reset the System configuration controller
1: resets the System configuration controller
Bit 13 Reserved, must be kept at reset value.
Bit 12 SPI1RST: SPI1 reset
Set and cleared by software.
0: does not reset SPI1
1: resets SPI1
Bit 11 SDIORST: SDIO reset
Set and cleared by software.
0: does not reset the SDIO module
1: resets the SDIO module
Bits 10:9 Reserved, must be kept at reset value.

240/1747

RM0090 Rev 17

3

18

17

16

TIM11
RST

TIM10
RST

TIM9
RST

rw

rw

rw

2

1

0

TIM8
RST

TIM1
RST

rw

rw

Reserved

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bit 8 ADCRST: ADC interface reset (common to all ADCs)
Set and cleared by software.
0: does not reset the ADC interface
1: resets the ADC interface
Bits 7:6 Reserved, must be kept at reset value.
Bit 5 USART6RST: USART6 reset
Set and cleared by software.
0: does not reset USART6
1: resets USART6
Bit 4

USART1RST: USART1 reset
Set and cleared by software.
0: does not reset USART1
1: resets USART1

Bits 3:2 Reserved, must be kept at reset value.
Bit 1 TIM8RST: TIM8 reset
Set and cleared by software.
0: does not reset TIM8
1: resets TIM8
Bit 0 TIM1RST: TIM1 reset
Set and cleared by software.
0: does not reset TIM1
1: resets TIM1

RM0090 Rev 17

241/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

7.3.10

RM0090

RCC AHB1 peripheral clock enable register (RCC_AHB1ENR)
Address offset: 0x30
Reset value: 0x0010 0000
Access: no wait state, word, half-word and byte access.

31

30

OTGH
S
Reser- ULPIE
ved
N

15

29

28

27

26

25

ETHM ETHM ETHM
OTGH
ETHMA
ACPTP ACRXE ACTXE
SEN
CEN
EN
N
N

rw

rw

rw

rw

rw

rw

14

13

12

11

10

9

Reserved

24

CRCE
N
rw

Reserved

23

Reserved

8

21

20

19

18

DMA2E
N

DMA1E
N

CCMDAT
ARAMEN

Res.

BKPSR
AMEN

rw

rw

6

5

4

3

7

GPIOIE GPIOH
N
EN
rw

22

rw

rw

Bit 30 OTGHSULPIEN: USB OTG HSULPI clock enable
Set and cleared by software.
0: USB OTG HS ULPI clock disabled
1: USB OTG HS ULPI clock enabled
Bit 29 OTGHSEN: USB OTG HS clock enable
Set and cleared by software.
0: USB OTG HS clock disabled
1: USB OTG HS clock enabled
Bit 28 ETHMACPTPEN: Ethernet PTP clock enable
Set and cleared by software.
0: Ethernet PTP clock disabled
1: Ethernet PTP clock enabled
Bit 27 ETHMACRXEN: Ethernet Reception clock enable
Set and cleared by software.
0: Ethernet Reception clock disabled
1: Ethernet Reception clock enabled
Bit 26 ETHMACTXEN: Ethernet Transmission clock enable
Set and cleared by software.
0: Ethernet Transmission clock disabled
1: Ethernet Transmission clock enabled
Bit 25 ETHMACEN: Ethernet MAC clock enable
Set and cleared by software.
0: Ethernet MAC clock disabled
1: Ethernet MAC clock enabled
Bits 24:23 Reserved, must be kept at reset value.
Bit 22 DMA2EN: DMA2 clock enable
Set and cleared by software.
0: DMA2 clock disabled
1: DMA2 clock enabled

RM0090 Rev 17

rw

16

Reserved

rw
2

1

GPIOD GPIOC GPIO
GPIOG GPIOFE
GPIOEEN
EN
EN
BEN
EN
N

Bit 31 Reserved, must be kept at reset value.

242/1747

17

rw

rw

rw

rw

0
GPIO
AEN
rw

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bit 21 DMA1EN: DMA1 clock enable
Set and cleared by software.
0: DMA1 clock disabled
1: DMA1 clock enabled
Bit 20 CCMDATARAMEN: CCM data RAM clock enable
Set and cleared by software.
0: CCM data RAM clock disabled
1: CCM data RAM clock enabled
Bit 19 Reserved, must be kept at reset value.
Bit 18 BKPSRAMEN: Backup SRAM interface clock enable
Set and cleared by software.
0: Backup SRAM interface clock disabled
1: Backup SRAM interface clock enabled
Bits 17:13 Reserved, must be kept at reset value.
Bit 12 CRCEN: CRC clock enable
Set and cleared by software.
0: CRC clock disabled
1: CRC clock enabled
Bits 11:9 Reserved, must be kept at reset value.
Bit 8 GPIOIEN: IO port I clock enable
Set and cleared by software.
0: IO port I clock disabled
1: IO port I clock enabled
Bit 7 GPIOHEN: IO port H clock enable
Set and cleared by software.
0: IO port H clock disabled
1: IO port H clock enabled
Bit 6 GPIOGEN: IO port G clock enable
Set and cleared by software.
0: IO port G clock disabled
1: IO port G clock enabled
Bit 5 GPIOFEN: IO port F clock enable
Set and cleared by software.
0: IO port F clock disabled
1: IO port F clock enabled
Bit 4 GPIOEEN: IO port E clock enable
Set and cleared by software.
0: IO port E clock disabled
1: IO port E clock enabled
Bit 3 GPIODEN: IO port D clock enable
Set and cleared by software.
0: IO port D clock disabled
1: IO port D clock enabled

RM0090 Rev 17

243/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

RM0090

Bit 2 GPIOCEN: IO port C clock enable
Set and cleared by software.
0: IO port C clock disabled
1: IO port C clock enabled
Bit 1 GPIOBEN: IO port B clock enable
Set and cleared by software.
0: IO port B clock disabled
1: IO port B clock enabled
Bit 0 GPIOAEN: IO port A clock enable
Set and cleared by software.
0: IO port A clock disabled
1: IO port A clock enabled

7.3.11

RCC AHB2 peripheral clock enable register (RCC_AHB2ENR)
Address offset: 0x34
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

7

6

5

4

3

2

1

0

OTGFS
EN

RNG
EN

HASH
EN

CRYP
EN

rw

rw

rw

rw

Reserved
15

14

13

12

11

Reserved

10

9

8

Bits 31:8 Reserved, must be kept at reset value.
Bit 7 OTGFSEN: USB OTG FS clock enable
Set and cleared by software.
0: USB OTG FS clock disabled
1: USB OTG FS clock enabled
Bit 6 RNGEN: Random number generator clock enable
Set and cleared by software.
0: Random number generator clock disabled
1: Random number generator clock enabled
Bit 5 HASHEN: Hash modules clock enable
Set and cleared by software.
0: Hash modules clock disabled
1: Hash modules clock enabled

244/1747

RM0090 Rev 17

Reserved

DCMI
EN
rw

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bit 4 CRYPEN: Cryptographic modules clock enable
Set and cleared by software.
0: cryptographic module clock disabled
1: cryptographic module clock enabled
Bits 3:1 Reserved, must be kept at reset value.
Bit 0 DCMIEN: Camera interface enable
Set and cleared by software.
0: Camera interface clock disabled
1: Camera interface clock enabled

7.3.12

RCC AHB3 peripheral clock enable register (RCC_AHB3ENR)
Address offset: 0x38
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

Reserved
15

14

13

12

11

10

9

8

7

FSMCEN

Reserved

rw

Bits 31:1 Reserved, must be kept at reset value.
Bit 0 FSMCEN: Flexible static memory controller module clock enable
Set and cleared by software.
0: FSMC module clock disabled
1: FSMC module clock enabled

7.3.13

RCC APB1 peripheral clock enable register
(RCC_APB1ENR)
Address offset: 0x40
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31

30

Reserved

15

14

SPI3
EN

SPI2
EN

rw

rw

29

28

DAC
EN

PWR
EN

rw

rw

13

12

Reserved

27
Reserved
11
WWDG
EN
rw

26

25

CAN2
EN

CAN1
EN

rw

rw

10

9

Reserved

24
Reserved

23

22

21

I2C3
EN

I2C2
EN

I2C1
EN

20

19

UART5 UART4
EN
EN

18

17

16

USART USART
3
2
EN
EN

Reserved

rw

rw

rw

rw

rw

rw

rw

8

7

6

5

4

3

2

1

0

TIM14
EN

TIM13
EN

TIM12
EN

TIM7
EN

TIM6
EN

TIM5
EN

TIM4
EN

TIM3
EN

TIM2
EN

rw

rw

rw

rw

rw

rw

rw

rw

rw

RM0090 Rev 17

245/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bits 31:30 Reserved, must be kept at reset value.
Bit 29 DACEN: DAC interface clock enable
Set and cleared by software.
0: DAC interface clock disabled
1: DAC interface clock enable
Bit 28 PWREN: Power interface clock enable
Set and cleared by software.
0: Power interface clock disabled
1: Power interface clock enable
Bit 27 Reserved, must be kept at reset value.
Bit 26 CAN2EN: CAN 2 clock enable
Set and cleared by software.
0: CAN 2 clock disabled
1: CAN 2 clock enabled
Bit 25 CAN1EN: CAN 1 clock enable
Set and cleared by software.
0: CAN 1 clock disabled
1: CAN 1 clock enabled
Bit 24 Reserved, must be kept at reset value.
Bit 23 I2C3EN: I2C3 clock enable
Set and cleared by software.
0: I2C3 clock disabled
1: I2C3 clock enabled
Bit 22 I2C2EN: I2C2 clock enable
Set and cleared by software.
0: I2C2 clock disabled
1: I2C2 clock enabled
Bit 21 I2C1EN: I2C1 clock enable
Set and cleared by software.
0: I2C1 clock disabled
1: I2C1 clock enabled
Bit 20 UART5EN: UART5 clock enable
Set and cleared by software.
0: UART5 clock disabled
1: UART5 clock enabled
Bit 19 UART4EN: UART4 clock enable
Set and cleared by software.
0: UART4 clock disabled
1: UART4 clock enabled
Bit 18 USART3EN: USART3 clock enable
Set and cleared by software.
0: USART3 clock disabled
1: USART3 clock enabled

246/1747

RM0090 Rev 17

RM0090

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bit 17 USART2EN: USART2 clock enable
Set and cleared by software.
0: USART2 clock disabled
1: USART2 clock enabled
Bit 16 Reserved, must be kept at reset value.
Bit 15 SPI3EN: SPI3 clock enable
Set and cleared by software.
0: SPI3 clock disabled
1: SPI3 clock enabled
Bit 14 SPI2EN: SPI2 clock enable
Set and cleared by software.
0: SPI2 clock disabled
1: SPI2 clock enabled
Bits 13:12 Reserved, must be kept at reset value.
Bit 11 WWDGEN: Window watchdog clock enable
Set and cleared by software.
0: Window watchdog clock disabled
1: Window watchdog clock enabled
Bits 10:9 Reserved, must be kept at reset value.
Bit 8 TIM14EN: TIM14 clock enable
Set and cleared by software.
0: TIM14 clock disabled
1: TIM14 clock enabled
Bit 7 TIM13EN: TIM13 clock enable
Set and cleared by software.
0: TIM13 clock disabled
1: TIM13 clock enabled
Bit 6 TIM12EN: TIM12 clock enable
Set and cleared by software.
0: TIM12 clock disabled
1: TIM12 clock enabled
Bit 5 TIM7EN: TIM7 clock enable
Set and cleared by software.
0: TIM7 clock disabled
1: TIM7 clock enabled
Bit 4 TIM6EN: TIM6 clock enable
Set and cleared by software.
0: TIM6 clock disabled
1: TIM6 clock enabled
Bit 3 TIM5EN: TIM5 clock enable
Set and cleared by software.
0: TIM5 clock disabled
1: TIM5 clock enabled

RM0090 Rev 17

247/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

RM0090

Bit 2 TIM4EN: TIM4 clock enable
Set and cleared by software.
0: TIM4 clock disabled
1: TIM4 clock enabled
Bit 1 TIM3EN: TIM3 clock enable
Set and cleared by software.
0: TIM3 clock disabled
1: TIM3 clock enabled
Bit 0 TIM2EN: TIM2 clock enable
Set and cleared by software.
0: TIM2 clock disabled
1: TIM2 clock enabled

7.3.14

RCC APB2 peripheral clock enable register (RCC_APB2ENR)
Address offset: 0x44
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31

30

29

28

27

26

25

24

23

22

21

20

19

Reserved
15
Reserved

14

13

SYSCF
G EN Reserved
rw

12

11

10

9

8

SPI1
EN

SDIO
EN

ADC3
EN

ADC2
EN

ADC1
EN

rw

rw

rw

rw

rw

7

Reserved

Bits 31:19 Reserved, must be kept at reset value.
Bit 18 TIM11EN: TIM11 clock enable
Set and cleared by software.
0: TIM11 clock disabled
1: TIM11 clock enabled
Bit 17 TIM10EN: TIM10 clock enable
Set and cleared by software.
0: TIM10 clock disabled
1: TIM10 clock enabled
Bit 16 TIM9EN: TIM9 clock enable
Set and cleared by software.
0: TIM9 clock disabled
1: TIM9 clock enabled
Bit 15 Reserved, must be kept at reset value.

248/1747

6

RM0090 Rev 17

5

4

USART USART
6
1
EN
EN
rw

rw

3

18

17

16

TIM11
EN

TIM10
EN

TIM9
EN

rw

rw

rw

2

1

0

TIM8
EN

TIM1
EN

rw

rw

Reserved

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bit 14 SYSCFGEN: System configuration controller clock enable
Set and cleared by software.
0: System configuration controller clock disabled
1: System configuration controller clock enabled
Bit 13 Reserved, must be kept at reset value.
Bit 12 SPI1EN: SPI1 clock enable
Set and cleared by software.
0: SPI1 clock disabled
1: SPI1 clock enabled
Bit 11 SDIOEN: SDIO clock enable
Set and cleared by software.
0: SDIO module clock disabled
1: SDIO module clock enabled
Bit 10 ADC3EN: ADC3 clock enable
Set and cleared by software.
0: ADC3 clock disabled
1: ADC3 clock disabled
Bit 9

ADC2EN: ADC2 clock enable
Set and cleared by software.
0: ADC2 clock disabled
1: ADC2 clock disabled

Bit 8

ADC1EN: ADC1 clock enable
Set and cleared by software.
0: ADC1 clock disabled
1: ADC1 clock disabled

Bits 7:6 Reserved, must be kept at reset value.
Bit 5 USART6EN: USART6 clock enable
Set and cleared by software.
0: USART6 clock disabled
1: USART6 clock enabled
Bit 4 USART1EN: USART1 clock enable
Set and cleared by software.
0: USART1 clock disabled
1: USART1 clock enabled
Bits 3:2 Reserved, must be kept at reset value.
Bit 1 TIM8EN: TIM8 clock enable
Set and cleared by software.
0: TIM8 clock disabled
1: TIM8 clock enabled
Bit 0 TIM1EN: TIM1 clock enable
Set and cleared by software.
0: TIM1 clock disabled
1: TIM1 clock enabled

RM0090 Rev 17

249/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

7.3.15

RM0090

RCC AHB1 peripheral clock enable in low power mode register
(RCC_AHB1LPENR)
Address offset: 0x50
Reset value: 0x7E67 91FF
Access: no wait state, word, half-word and byte access.

31

30

OTGHS
Reser ULPILPE
N
-ved
15

29
OTGH
S
LPEN

28

27

26

25

rw

rw

rw

rw

rw

rw

14

13

12

11

10

9

FLITF
LPEN
rw

Reserved

24

ETHMA
ETHPT
ETHRX ETHTX
P
C
LPEN
LPEN
LPEN
LPEN

CRC
LPEN
rw

Reserved

23

Reserved

8

7

22

21

DMA2
LPEN

DMA1
LPEN

rw

rw

6

GPIOG
GPIOI GPIOH
G
LPEN LPEN
LPEN
rw

rw

rw

5

20

19

Reserved

4

3

GPIO
GPIOE GPIOD
F
LPEN
LPEN
LPEN
rw

rw

18

rw

Bit 30 OTGHSULPILPEN: USB OTG HS ULPI clock enable during Sleep mode
Set and cleared by software.
0: USB OTG HS ULPI clock disabled during Sleep mode
1: USB OTG HS ULPI clock enabled during Sleep mode
Bit 29 OTGHSLPEN: USB OTG HS clock enable during Sleep mode
Set and cleared by software.
0: USB OTG HS clock disabled during Sleep mode
1: USB OTG HS clock enabled during Sleep mode
Bit 28 ETHMACPTPLPEN: Ethernet PTP clock enable during Sleep mode
Set and cleared by software.
0: Ethernet PTP clock disabled during Sleep mode
1: Ethernet PTP clock enabled during Sleep mode
Bit 27 ETHMACRXLPEN: Ethernet reception clock enable during Sleep mode
Set and cleared by software.
0: Ethernet reception clock disabled during Sleep mode
1: Ethernet reception clock enabled during Sleep mode
Bit 26 ETHMACTXLPEN: Ethernet transmission clock enable during Sleep mode
Set and cleared by software.
0: Ethernet transmission clock disabled during sleep mode
1: Ethernet transmission clock enabled during sleep mode

Bits 24:23 Reserved, must be kept at reset value.
Bit 22 DMA2LPEN: DMA2 clock enable during Sleep mode
Set and cleared by software.
0: DMA2 clock disabled during Sleep mode
1: DMA2 clock enabled during Sleep mode

250/1747

RM0090 Rev 17

16

rw

rw

rw

2

1

0

GPIOC
LPEN

Bit 31 Reserved, must be kept at reset value.

Bit 25 ETHMACLPEN: Ethernet MAC clock enable during Sleep mode
Set and cleared by software.
0: Ethernet MAC clock disabled during Sleep mode
1: Ethernet MAC clock enabled during Sleep mode

17

BKPSRA SRAM SRAM
M
2
1
LPEN
LPEN LPEN

rw

GPIOB GPIOA
LPEN LPEN
rw

rw

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bit 21 DMA1LPEN: DMA1 clock enable during Sleep mode
Set and cleared by software.
0: DMA1 clock disabled during Sleep mode
1: DMA1 clock enabled during Sleep mode
Bits 20:19 Reserved, must be kept at reset value.
Bit 18 BKPSRAMLPEN: Backup SRAM interface clock enable during Sleep mode
Set and cleared by software.
0: Backup SRAM interface clock disabled during Sleep mode
1: Backup SRAM interface clock enabled during Sleep mode
Bit 17 SRAM2LPEN: SRAM 2 interface clock enable during Sleep mode
Set and cleared by software.
0: SRAM 2 interface clock disabled during Sleep mode
1: SRAM 2 interface clock enabled during Sleep mode
Bit 16 SRAM1LPEN: SRAM 1interface clock enable during Sleep mode
Set and cleared by software.
0: SRAM 1 interface clock disabled during Sleep mode
1: SRAM 1 interface clock enabled during Sleep mode
Bit 15 FLITFLPEN: Flash interface clock enable during Sleep mode
Set and cleared by software.
0: Flash interface clock disabled during Sleep mode
1: Flash interface clock enabled during Sleep mode
Bits 14:13 Reserved, must be kept at reset value.
Bit 12 CRCLPEN: CRC clock enable during Sleep mode
Set and cleared by software.
0: CRC clock disabled during Sleep mode
1: CRC clock enabled during Sleep mode
Bits 11:9 Reserved, must be kept at reset value.
Bit 8 GPIOILPEN: IO port I clock enable during Sleep mode
Set and cleared by software.
0: IO port I clock disabled during Sleep mode
1: IO port I clock enabled during Sleep mode
Bit 7

GPIOHLPEN: IO port H clock enable during Sleep mode
Set and cleared by software.
0: IO port H clock disabled during Sleep mode
1: IO port H clock enabled during Sleep mode

Bits 6 GPIOGLPEN: IO port G clock enable during Sleep mode
Set and cleared by software.
0: IO port G clock disabled during Sleep mode
1: IO port G clock enabled during Sleep mode
Bit 5 GPIOFLPEN: IO port F clock enable during Sleep mode
Set and cleared by software.
0: IO port F clock disabled during Sleep mode
1: IO port F clock enabled during Sleep mode

RM0090 Rev 17

251/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

RM0090

Bit 4 GPIOELPEN: IO port E clock enable during Sleep mode
Set and cleared by software.
0: IO port E clock disabled during Sleep mode
1: IO port E clock enabled during Sleep mode
Bit 3 GPIODLPEN: IO port D clock enable during Sleep mode
Set and cleared by software.
0: IO port D clock disabled during Sleep mode
1: IO port D clock enabled during Sleep mode
Bit 2 GPIOCLPEN: IO port C clock enable during Sleep mode
Set and cleared by software.
0: IO port C clock disabled during Sleep mode
1: IO port C clock enabled during Sleep mode
Bit 1 GPIOBLPEN: IO port B clock enable during Sleep mode
Set and cleared by software.
0: IO port B clock disabled during Sleep mode
1: IO port B clock enabled during Sleep mode
Bit 0 GPIOALPEN: IO port A clock enable during sleep mode
Set and cleared by software.
0: IO port A clock disabled during Sleep mode
1: IO port A clock enabled during Sleep mode

7.3.16

RCC AHB2 peripheral clock enable in low power mode register
(RCC_AHB2LPENR)
Address offset: 0x54
Reset value: 0x0000 00F1
Access: no wait state, word, half-word and byte access.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

HASH
LPEN

CRYP
LPEN

rw

rw

Reserved
15

14

13

12

11

Reserved

10

9

8

7

OTGFS RNG
LPEN LPEN
rw

rw

Reserved

Bits 31:8 Reserved, must be kept at reset value.
Bit 7 OTGFSLPEN: USB OTG FS clock enable during Sleep mode
Set and cleared by software.
0: USB OTG FS clock disabled during Sleep mode
1: USB OTG FS clock enabled during Sleep mode
Bit 6 RNGLPEN: Random number generator clock enable during Sleep mode
Set and cleared by software.
0: Random number generator clock disabled during Sleep mode
1: Random number generator clock enabled during Sleep mode

252/1747

RM0090 Rev 17

DCMI
LPEN
rw

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bit 5 HASHLPEN: Hash modules clock enable during Sleep mode
Set and cleared by software.
0: Hash modules clock disabled during Sleep mode
1: Hash modules clock enabled during Sleep mode
Bit 4 CRYPLPEN: Cryptography modules clock enable during Sleep mode
Set and cleared by software.
0: cryptography modules clock disabled during Sleep mode
1: cryptography modules clock enabled during Sleep mode
Bits 3:1 Reserved, must be kept at reset value.
Bit 0 DCMILPEN: Camera interface enable during Sleep mode
Set and cleared by software.
0: Camera interface clock disabled during Sleep mode
1: Camera interface clock enabled during Sleep mode

7.3.17

RCC AHB3 peripheral clock enable in low power mode register
(RCC_AHB3LPENR)
Address offset: 0x58
Reset value: 0x0000 0001
Access: no wait state, word, half-word and byte access.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

6

5

4

3

2

1

16

Reserved
15

14

13

12

11

10

9

8

7

0
FSMC
LPEN

Reserved

rw

Bits 31:1Reserved, must be kept at reset value.
FSMCLPEN: Flexible static memory controller module clock enable during Sleep mode
Set and cleared by software.
Bit 0
0: FSMC module clock disabled during Sleep mode
1: FSMC module clock enabled during Sleep mode

RM0090 Rev 17

253/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

7.3.18

RM0090

RCC APB1 peripheral clock enable in low power mode register
(RCC_APB1LPENR)
Address offset: 0x60
Reset value: 0x36FE C9FF
Access: no wait state, word, half-word and byte access.

31

30

Reserved

15

14

SPI3
LPEN

SPI2
LPEN

rw

rw

29

28

DAC
LPEN

PWR
LPEN

rw

rw

13

12

Reserved

27
RESER
VED

11
WWDG
LPEN
rw

26

25

CAN2
LPEN

CAN1
LPEN

rw

rw

10

9

Reserved

24
Reserved

23

22

21

I2C3
LPEN

I2C2
LPEN

I2C1
LPEN

rw

rw

rw

20

18

17

USART USART
UART5 UART4
2
3
LPEN
LPEN
LPEN
LPEN
rw

rw

rw

16
Reserved

rw

8

7

6

5

4

3

2

1

0

TIM14
LPEN

TIM13
LPEN

TIM12
LPEN

TIM7
LPEN

TIM6
LPEN

TIM5
LPEN

TIM4
LPEN

TIM3
LPEN

TIM2
LPEN

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:30 Reserved, must be kept at reset value.
Bit 29 DACLPEN: DAC interface clock enable during Sleep mode
Set and cleared by software.
0: DAC interface clock disabled during Sleep mode
1: DAC interface clock enabled during Sleep mode
Bit 28 PWRLPEN: Power interface clock enable during Sleep mode
Set and cleared by software.
0: Power interface clock disabled during Sleep mode
1: Power interface clock enabled during Sleep mode
Bit 27 Reserved, must be kept at reset value.
Bit 26 CAN2LPEN: CAN 2 clock enable during Sleep mode
Set and cleared by software.
0: CAN 2 clock disabled during sleep mode
1: CAN 2 clock enabled during sleep mode
Bit 25 CAN1LPEN: CAN 1 clock enable during Sleep mode
Set and cleared by software.
0: CAN 1 clock disabled during Sleep mode
1: CAN 1 clock enabled during Sleep mode
Bit 24 Reserved, must be kept at reset value.
Bit 23 I2C3LPEN: I2C3 clock enable during Sleep mode
Set and cleared by software.
0: I2C3 clock disabled during Sleep mode
1: I2C3 clock enabled during Sleep mode
Bit 22 I2C2LPEN: I2C2 clock enable during Sleep mode
Set and cleared by software.
0: I2C2 clock disabled during Sleep mode
1: I2C2 clock enabled during Sleep mode

254/1747

19

RM0090 Rev 17

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bit 21 I2C1LPEN: I2C1 clock enable during Sleep mode
Set and cleared by software.
0: I2C1 clock disabled during Sleep mode
1: I2C1 clock enabled during Sleep mode
Bit 20 UART5LPEN: UART5 clock enable during Sleep mode
Set and cleared by software.
0: UART5 clock disabled during Sleep mode
1: UART5 clock enabled during Sleep mode
Bit 19 UART4LPEN: UART4 clock enable during Sleep mode
Set and cleared by software.
0: UART4 clock disabled during Sleep mode
1: UART4 clock enabled during Sleep mode
Bit 18

USART3LPEN: USART3 clock enable during Sleep mode
Set and cleared by software.
0: USART3 clock disabled during Sleep mode
1: USART3 clock enabled during Sleep mode

Bit 17 USART2LPEN: USART2 clock enable during Sleep mode
Set and cleared by software.
0: USART2 clock disabled during Sleep mode
1: USART2 clock enabled during Sleep mode
Bit 16 Reserved, must be kept at reset value.
Bit 15 SPI3LPEN: SPI3 clock enable during Sleep mode
Set and cleared by software.
0: SPI3 clock disabled during Sleep mode
1: SPI3 clock enabled during Sleep mode
Bit 14 SPI2LPEN: SPI2 clock enable during Sleep mode
Set and cleared by software.
0: SPI2 clock disabled during Sleep mode
1: SPI2 clock enabled during Sleep mode
Bits 13:12 Reserved, must be kept at reset value.
Bit 11 WWDGLPEN: Window watchdog clock enable during Sleep mode
Set and cleared by software.
0: Window watchdog clock disabled during sleep mode
1: Window watchdog clock enabled during sleep mode
Bits 10:9 Reserved, must be kept at reset value.
Bit 8 TIM14LPEN: TIM14 clock enable during Sleep mode
Set and cleared by software.
0: TIM14 clock disabled during Sleep mode
1: TIM14 clock enabled during Sleep mode
Bit 7 TIM13LPEN: TIM13 clock enable during Sleep mode
Set and cleared by software.
0: TIM13 clock disabled during Sleep mode
1: TIM13 clock enabled during Sleep mode

RM0090 Rev 17

255/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bit 6 TIM12LPEN: TIM12 clock enable during Sleep mode
Set and cleared by software.
0: TIM12 clock disabled during Sleep mode
1: TIM12 clock enabled during Sleep mode
Bit 5 TIM7LPEN: TIM7 clock enable during Sleep mode
Set and cleared by software.
0: TIM7 clock disabled during Sleep mode
1: TIM7 clock enabled during Sleep mode
Bit 4 TIM6LPEN: TIM6 clock enable during Sleep mode
Set and cleared by software.
0: TIM6 clock disabled during Sleep mode
1: TIM6 clock enabled during Sleep mode
Bit 3 TIM5LPEN: TIM5 clock enable during Sleep mode
Set and cleared by software.
0: TIM5 clock disabled during Sleep mode
1: TIM5 clock enabled during Sleep mode
Bit 2 TIM4LPEN: TIM4 clock enable during Sleep mode
Set and cleared by software.
0: TIM4 clock disabled during Sleep mode
1: TIM4 clock enabled during Sleep mode
Bit 1 TIM3LPEN: TIM3 clock enable during Sleep mode
Set and cleared by software.
0: TIM3 clock disabled during Sleep mode
1: TIM3 clock enabled during Sleep mode
Bit 0

256/1747

TIM2LPEN: TIM2 clock enable during Sleep mode
Set and cleared by software.
0: TIM2 clock disabled during Sleep mode
1: TIM2 clock enabled during Sleep mode

RM0090 Rev 17

RM0090

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

7.3.19

RCC APB2 peripheral clock enabled in low power mode
register (RCC_APB2LPENR)
Address offset: 0x64
Reset value: 0x0407 5F33
Access: no wait state, word, half-word and byte access.

31

30

29

28

27

26

25

24

23

22

21

20

19

Reserved
15

14

13

12

11

10

9

8

SYSC
FG
ReserReserLPEN
ved
ved

SPI1
LPEN

SDIO
LPEN

ADC3
LPEN

ADC2
LPEN

ADC1
LPEN

rw

rw

rw

rw

rw

rw

7

6

Reserved

5

4

USART
6
LPEN

USART
1
LPEN

rw

rw

3

18

17

16

TIM11
LPEN

TIM10
LPEN

TIM9
LPEN

rw

rw

rw

2

1

0

TIM8
LPEN

TIM1
LPEN

rw

rw

Reserved

Bits 31:19 Reserved, must be kept at reset value.
Bit 18 TIM11LPEN: TIM11 clock enable during Sleep mode
Set and cleared by software.
0: TIM11 clock disabled during Sleep mode
1: TIM11 clock enabled during Sleep mode
Bit 17 TIM10LPEN: TIM10 clock enable during Sleep mode
Set and cleared by software.
0: TIM10 clock disabled during Sleep mode
1: TIM10 clock enabled during Sleep mode
Bit 16 TIM9LPEN: TIM9 clock enable during sleep mode
Set and cleared by software.
0: TIM9 clock disabled during Sleep mode
1: TIM9 clock enabled during Sleep mode
Bit 15 Reserved, must be kept at reset value.
Bit 14 SYSCFGLPEN: System configuration controller clock enable during Sleep mode
Set and cleared by software.
0: System configuration controller clock disabled during Sleep mode
1: System configuration controller clock enabled during Sleep mode
Bit 13 Reserved, must be kept at reset value.
Bit 12 SPI1LPEN: SPI1 clock enable during Sleep mode
Set and cleared by software.
0: SPI1 clock disabled during Sleep mode
1: SPI1 clock enabled during Sleep mode
Bit 11 SDIOLPEN: SDIO clock enable during Sleep mode
Set and cleared by software.
0: SDIO module clock disabled during Sleep mode
1: SDIO module clock enabled during Sleep mode

RM0090 Rev 17

257/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bit 10 ADC3LPEN: ADC 3 clock enable during Sleep mode
Set and cleared by software.
0: ADC 3 clock disabled during Sleep mode
1: ADC 3 clock disabled during Sleep mode
Bit 9 ADC2LPEN: ADC2 clock enable during Sleep mode
Set and cleared by software.
0: ADC2 clock disabled during Sleep mode
1: ADC2 clock disabled during Sleep mode
Bit 8 ADC1LPEN: ADC1 clock enable during Sleep mode
Set and cleared by software.
0: ADC1 clock disabled during Sleep mode
1: ADC1 clock disabled during Sleep mode
Bits 7:6 Reserved, must be kept at reset value.
Bit 5 USART6LPEN: USART6 clock enable during Sleep mode
Set and cleared by software.
0: USART6 clock disabled during Sleep mode
1: USART6 clock enabled during Sleep mode
Bit 4 USART1LPEN: USART1 clock enable during Sleep mode
Set and cleared by software.
0: USART1 clock disabled during Sleep mode
1: USART1 clock enabled during Sleep mode
Bits 3:2 Reserved, must be kept at reset value.
Bit 1 TIM8LPEN: TIM8 clock enable during Sleep mode
Set and cleared by software.
0: TIM8 clock disabled during Sleep mode
1: TIM8 clock enabled during Sleep mode
Bit 0 TIM1LPEN: TIM1 clock enable during Sleep mode
Set and cleared by software.
0: TIM1 clock disabled during Sleep mode
1: TIM1 clock enabled during Sleep mode

258/1747

RM0090 Rev 17

RM0090

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

7.3.20

RCC Backup domain control register (RCC_BDCR)
Address offset: 0x70
Reset value: 0x0000 0000, reset by Backup domain reset.
Access: 0 ≤ wait state ≤ 3, word, half-word and byte access
Wait states are inserted in case of successive accesses to this register.
The LSEON, LSEBYP, RTCSEL and RTCEN bits in the RCC Backup domain control
register (RCC_BDCR) are in the Backup domain. As a result, after Reset, these bits are
write-protected and the DBP bit in the PWR power control register (PWR_CR) for
STM32F405xx/07xx and STM32F415xx/17xx has to be set before these can be modified.
Refer to Section 7.1.1: System reset on page 213 for further information. These bits are only
reset after a Backup domain Reset (see Section 7.1.3: Backup domain reset). Any internal
or external Reset will not have any effect on these bits.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

Reserved
15
RTCEN
rw

14

13

12
Reserved

11

10

9

8

rw
7

RTCSEL[1:0]
rw

16
BDRST

rw

6

5
Reserved

4

3

2

1

0

LSEBY LSERD
LSEON
P
Y
rw

r

rw

Bits 31:17 Reserved, must be kept at reset value.
Bit 16 BDRST: Backup domain software reset
Set and cleared by software.
0: Reset not activated
1: Resets the entire Backup domain
Note: The BKPSRAM is not affected by this reset, the only way of resetting the BKPSRAM is
through the Flash interface when a protection level change from level 1 to level 0 is
requested.
Bit 15 RTCEN: RTC clock enable
Set and cleared by software.
0: RTC clock disabled
1: RTC clock enabled
Bits 14:10 Reserved, must be kept at reset value.
Bits 9:8 RTCSEL[1:0]: RTC clock source selection
Set by software to select the clock source for the RTC. Once the RTC clock source has been
selected, it cannot be changed anymore unless the Backup domain is reset. The BDRST bit
can be used to reset them.
00: No clock
01: LSE oscillator clock used as the RTC clock
10: LSI oscillator clock used as the RTC clock
11: HSE oscillator clock divided by a programmable prescaler (selection through the
RTCPRE[4:0] bits in the RCC clock configuration register (RCC_CFGR)) used as the RTC
clock
Bits 7:3 Reserved, must be kept at reset value.

RM0090 Rev 17

259/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

RM0090

Bit 2 LSEBYP: External low-speed oscillator bypass
Set and cleared by software to bypass the oscillator. This bit can be written only when the
LSE clock is disabled.
0: LSE oscillator not bypassed
1: LSE oscillator bypassed
Bit 1 LSERDY: External low-speed oscillator ready
Set and cleared by hardware to indicate when the external 32 kHz oscillator is stable. After
the LSEON bit is cleared, LSERDY goes low after 6 external low-speed oscillator clock
cycles.
0: LSE clock not ready
1: LSE clock ready
Bit 0 LSEON: External low-speed oscillator enable
Set and cleared by software.
0: LSE clock OFF
1: LSE clock ON

7.3.21

RCC clock control & status register (RCC_CSR)
Address offset: 0x74
Reset value: 0x0E00 0000, reset by system reset, except reset flags by power reset only.
Access: 0 ≤ wait state ≤ 3, word, half-word and byte access
Wait states are inserted in case of successive accesses to this register.

31

30

LPWR WWDG
RSTF RSTF

29

28

27

26

25

24

IWDG
RSTF

SFT
RSTF

POR
RSTF

PIN
RSTF

BORRS
TF

RMVF

r

r

r

r

r

r

r

rt_w

15

14

13

12

11

10

9

8

23

22

21

20

19

18

17

16

1

0

Reserved
7

6

5

Reserved

4

3

2

LSIRDY LSION
r

rw

Bit 31 LPWRRSTF: Low-power reset flag
Set by hardware when a Low-power management reset occurs.
Cleared by writing to the RMVF bit.
0: No Low-power management reset occurred
1: Low-power management reset occurred
For further information on Low-power management reset, refer to Low-power management
reset.
Bit 30 WWDGRSTF: Window watchdog reset flag
Set by hardware when a window watchdog reset occurs.
Cleared by writing to the RMVF bit.
0: No window watchdog reset occurred
1: Window watchdog reset occurred
Bit 29 IWDGRSTF: Independent watchdog reset flag
Set by hardware when an independent watchdog reset from VDD domain occurs.
Cleared by writing to the RMVF bit.
0: No watchdog reset occurred
1: Watchdog reset occurred

260/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

Bit 28 SFTRSTF: Software reset flag
Set by hardware when a software reset occurs.
Cleared by writing to the RMVF bit.
0: No software reset occurred
1: Software reset occurred
Bit 27 PORRSTF: POR/PDR reset flag
Set by hardware when a POR/PDR reset occurs.
Cleared by writing to the RMVF bit.
0: No POR/PDR reset occurred
1: POR/PDR reset occurred
Bit 26 PINRSTF: PIN reset flag
Set by hardware when a reset from the NRST pin occurs.
Cleared by writing to the RMVF bit.
0: No reset from NRST pin occurred
1: Reset from NRST pin occurred
Bit 25 BORRSTF: BOR reset flag
Cleared by software by writing the RMVF bit.
Set by hardware when a POR/PDR or BOR reset occurs.
0: No POR/PDR or BOR reset occurred
1: POR/PDR or BOR reset occurred
Bit 24 RMVF: Remove reset flag
Set by software to clear the reset flags.
0: No effect
1: Clear the reset flags
Bits 23:2 Reserved, must be kept at reset value.
Bit 1 LSIRDY: Internal low-speed oscillator ready
Set and cleared by hardware to indicate when the internal RC 40 kHz oscillator is stable.
After the LSION bit is cleared, LSIRDY goes low after 3 LSI clock cycles.
0: LSI RC oscillator not ready
1: LSI RC oscillator ready
Bit 0 LSION: Internal low-speed oscillator enable
Set and cleared by software.
0: LSI RC oscillator OFF
1: LSI RC oscillator ON

RM0090 Rev 17

261/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

7.3.22

RM0090

RCC spread spectrum clock generation register (RCC_SSCGR)
Address offset: 0x80
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.
The spread spectrum clock generation is available only for the main PLL.
The RCC_SSCGR register must be written either before the main PLL is enabled or after
the main PLL disabled.

Note:

For full details about PLL spread spectrum clock generation (SSCG) characteristics, refer to
the “Electrical characteristics” section in your device datasheet.

31

30

SSCG
EN

SPR
EAD
SEL

rw

rw

15

14

29

28

27

26

25

24

23

13

12

rw

21

20

19

18

17

16

INCSTEP

Reserved
rw

rw

rw

11

10

9

rw

rw

rw

rw

rw

rw

rw

rw

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

8

INCSTEP
rw

22

MODPER
rw

rw

rw

rw

rw

rw

rw

rw

Bit 31 SSCGEN: Spread spectrum modulation enable
Set and cleared by software.
0: Spread spectrum modulation DISABLE. (To write after clearing CR[24]=PLLON bit)
1: Spread spectrum modulation ENABLE. (To write before setting CR[24]=PLLON bit)
Bit 30 SPREADSEL: Spread Select
Set and cleared by software.
To write before to set CR[24]=PLLON bit.
0: Center spread
1: Down spread
Bits 29:28 Reserved, must be kept at reset value.
Bits 27:13 INCSTEP: Incrementation step
Set and cleared by software. To write before setting CR[24]=PLLON bit.
Configuration input for modulation profile amplitude.
Bits 12:0 MODPER: Modulation period
Set and cleared by software. To write before setting CR[24]=PLLON bit.
Configuration input for modulation profile period.

262/1747

RM0090 Rev 17

RM0090

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

7.3.23

RCC PLLI2S configuration register (RCC_PLLI2SCFGR)
Address offset: 0x84
Reset value: 0x2000 3000
Access: no wait state, word, half-word and byte access.
This register is used to configure the PLLI2S clock outputs according to the formulas:

31
Reserv
ed
15

•

f(VCO clock) = f(PLLI2S clock input) × (PLLI2SN / PLLM)

•

f(PLL I2S clock output) = f(VCO clock) / PLLI2SR

30

29

28

PLLI2S
R2

PLLI2S
R1

PLLI2S
R0

rw

rw

rw

14

13

12

27

26

25

24

23

22

21

20

19

18

17

16

5

4

3

2

1

0

Reserved
11

10

9

8

7

6

PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN
Reserv
8
7
6
5
4
3
2
1
0
ed
rw
rw
rw
rw
rw
rw
rw
rw
rw

Reserved

Bit 31 Reserved, must be kept at reset value.
Bits 30:28 PLLI2SR: PLLI2S division factor for I2S clocks
Set and cleared by software to control the I2S clock frequency. These bits should be written
only if the PLLI2S is disabled. The factor must be chosen in accordance with the prescaler
values inside the I2S peripherals, to reach 0.3% error when using standard crystals and 0%
error with audio crystals. For more information about I2S clock frequency and precision,
refer to Section 28.4.4: Clock generator in the I2S chapter.
Caution: The I2Ss requires a frequency lower than or equal to 192 MHz to work correctly.
I2S clock frequency = VCO frequency / PLLR with 2 ≤PLLR ≤7
000: PLLR = 0, wrong configuration
001: PLLR = 1, wrong configuration
010: PLLR = 2
...
111: PLLR = 7

RM0090 Rev 17

263/1747
266

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

RM0090

Bits 27:15 Reserved, must be kept at reset value.
Bits 14:6 PLLI2SN: PLLI2S multiplication factor for VCO
These bits are set and cleared by software to control the multiplication factor of the VCO.
These bits can be written only when PLLI2S is disabled. Only half-word and word accesses
are allowed to write these bits.
Caution: The software has to set these bits correctly to ensure that the VCO output
frequency is between 100 and 432 MHz.
VCO output frequency = VCO input frequency × PLLI2SN with 50 ≤PLLI2SN ≤432
000000000: PLLI2SN = 0, wrong configuration
000000001: PLLI2SN = 1, wrong configuration
...
000110010: PLLI2SN = 50
...
001100011: PLLI2SN = 99
001100100: PLLI2SN = 100
001100101: PLLI2SN = 101
001100110: PLLI2SN = 102
...
110110000: PLLI2SN = 432
110110001: PLLI2SN = 433, wrong configuration
...
111111111: PLLI2SN = 511, wrong configuration
Note: Multiplication factors ranging from 50 and 99 are possible for VCO input frequency
higher than 1 MHz. However care must be taken that the minimum VCO output
frequency respects the value specified above.
Bits 5:0 Reserved, must be kept at reset value.

264/1747

RM0090 Rev 17

Reserved

0x04
RCC_
PLLCFGR
Reserved

0x08
RCC_CFGR

0x0C
RCC_CIR

0x10
RCC_
AHB1RSTR

0x14
RCC_
AHB2RSTR

0x18
RCC_
AHB3RSTR

0x1C
Reserved

0x20
RCC_
APB1RSTR

0x24
RCC_
APB2RSTR

0x28
0x2C
Reserved
Reserved

0x30

RCC_
AHB1ENR

0x34

RCC_
AHB2ENR

0x38

RCC_
AHB3ENR

0x3C

Reserved

OTGHSEN

RM0090 Rev 17

Reserved

Reserved

GPIOAEN

GPIOBEN

TIM3RST
TIM2RST

TIM1RST

TIM4RST

TIM8RST

Reserved

FSMCRST DCMIRST

Reserved

GPIODRST

TIM5RST

GPIOCEN

CRYPRST GPIOERST

USART1RST TIM6RST

GPIOEEN

GPIODEN

CRYPEN

Reserved

CSSF

LSIRDYIE

LSERDYIE

HSIRDYIE

HSERDYIE

GPIOARST

GPIOBRST

GPIOCRST

PLLRDYF

LSIRDYF

LSERDYF

HSIRDYF

HSERDYF

CSSON

HSEON

HSERDY

HSEBYP

SW 0

SW 1

PLLM 0

PLLM 1

PLLM 2

HSION

HSIRDY

Reserved

PLLM 3 HSITRIM 0

SWS 1
SWS 0

PLLM 4 HSITRIM 1

HPRE 0

PLLM 5 HSITRIM 2

PLLN 0 HSITRIM 3
HPRE 1

HPRE 2

HSICAL 0

HSICAL 1

HSICAL 2

HSICAL 3

PLLN 1 HSITRIM 4

PLLN 2

PLLN 3

PLLN 4

HSICAL 4

HSICAL 5

HSICAL 6

HPRE 3

Reserved

PPRE1 0

PLLN 5

PLLN 6

PLLN 7
PPRE1 2
PPRE1 1

PLLN 8
PPRE2 1

PLLRDYIE

PLL ON

PLL RDY

PPRE2 2 Reserved HSICAL 7

RTCPRE 0 PLLP 0

PLLI2SRDYIE PPRE2 0

Reserved

LSIRDYC

PLLSRC

RTCPRE 1 PLLP 1

RTCPRE 2

LSERDYC

RTCPRE 3

HSIRDYC

PLLRDYC

MCO1 0
RTCPRE 4

MCO1 1

I2SSRC Reserved

MCO1PRE1 PLLQ 1
MCO1PRE0 PLLQ 0

MCO1PRE2 PLLQ 2 PLL I2SON

MCO2PRE0 PLLQ 3 PLL I2SRDY

MCO2PRE1

HSERDYC

GPIOFRST PLLI2SRDYF

GPIOGRST

RNGRST
HSAHRST

GPIOFEN

HASHEN

TIM13RST

Reserved

OTGFSRST GPIOHRST

TIM14RST

ADCRST
TIM12RST

GPIOGEN

RNGEN

GPIOIRST

Reserved

Reserved

CRCRST

Reserved

Reserved

USART6RST TIM7RST

GPIOHEN

Reserved

SDIORST WWDGRST

SPI1RST

Reserved

Reserved

FSMCEN DCMIEN

Reserved

GPIOIEN

SPI3RST

Reserved

OTGFSEN

Reserved

Reserved

Reserved
Reserved
SYSCFGRST SPI2RST

Reserved

Reserved

Reserved

CRCEN

Reserved

TIM9RST

Reserved

CSSC

DMA1RST PLLI2SRDYC

DMA2RST

Reserved

ETHMACRST

Reserved

MCO2PRE2

Reserved

TIM10RST UART2RST

TIM11RST UART3RST

Reserved

BKPSRAMEN

UART4RST

UART5RST

I2C1RST

I2C2RST

I2C3RST

Reserved

CAN1RST

CAN2RST

Reserved

OTGHSRST

Reserved

Reserved

CCMDATARAMEN

DMA1EN

DACRST
PWRRST

Reserved

DMA2EN

Reserved

ETHMACEN

ETHMACTXEN

ETHMACRXEN

ETHMACPTPEN

MCO2 0

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

RCC_CR

MCO2 1

0x00

Reserved

Register
name

Reserved

Addr.
offset

Reserved

7.3.24

OTGHSULPIEN

RM0090
Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

RCC register map
Table 34 gives the register map and reset values.
Table 34. RCC register map and reset values

265/1747

266

RCC_
APB2ENR

0x48
0x4C
Reserved
Reserved

0x50
RCC_AHB1LP
ENR

0x54
RCC_AHB2LP
ENR

0x58
RCC_AHB3LP
ENR

0x5C
Reserved

0x60
RCC_APB1LP
ENR

0x64
RCC_APB2LP
ENR

0x68
0x6C
Reserved
Reserved

0x70
RCC_BDCR

0x74
RCC_CSR

0x78
0x7C

Reserved
Reserved

0x80

RCC_SSCGR

0x84

RCC_PLLI2SC
FGR

266/1747

Reserved

SPREADSEL

PLLI2SRx
RMVF

BORRSTF

PADRSTF

PORRSTF

SFTRSTF

WDGRSTF

WWDGRSTF

Reserved

Reserved

Reserved
Reserved

Reserved

RM0090 Rev 17
SPI3LPEN

Reserved

INCSTEP

PLLI2SNx

Refer to Section 2.3: Memory map for the register boundary addresses.
Reserved
LSERDY
LSEON

LSIRDY
LSION

Reserved

CRYPLPEN

USART1LPEN TIM6LPEN

TIM3LPEN
TIM2LPEN

TIM1LPEN

FSMCLPEN DCMILPEN

Reserved
Reserved

MODPER

Reserved

GPIOALPEN

GPIOBLPEN

GPIOCLPEN

TIM4LPEN
TIM8LPEN

Reserved

GPIODLPEN

GPIOELPEN

GPIOFLPEN
TIM5LPEN
Reserved

HASHLPEN

USART6LPEN TIM7LPEN

GPIOGLPEN

RNGLPEN

TIM12LPEN

Reserved

GPIOILPEN

Reserved

CRCLPEN

Reserved

FLITFLPEN

SRAM1LPEN

SRAM2LPEN

BKPSRAMLPEN

Reserved

DMA1LPEN

DMA2LPEN

Reserved

ETHMACLPEN

GPIOHLPEN

Reserved
OTGFSLPEN

Reserved

TIM13LPEN

ADC1LPEN

RTCSEL 0

Reserved

WWDGLPEN

Reserved

Reserved

TIM14LPEN

ADC2LPEN

RTCSEL 1

Reserved
ADC3LPEN

SDIOLPEN

SPI1LPEN

Reserved

SYSCFGLPEN SPI2LPEN

Reserved

TIM9LPEN

TIM10LPEN USART2LPEN

TIM11LPEN USART3LPEN

UART4LPEN

UART5LPEN

I2C1LPEN

I2C2LPEN

I2C3LPEN

Reserved

CAN1LPEN

ETHMACTXLPEN

ETHMACRXLPEN

DACEN

Reserved

TIM9EN

SPI3EN

Reserved

TIM10EN USART2EN

TIM11EN USART3EN

UART4EN

UART5EN

I2C1EN

I2C2EN

I2C3EN

Reserved

CAN1EN

CAN2EN

Reserved

PWREN

TIM1EN

TIM8EN

Reserved

USART1EN

USART6EN

Reserved

ADC1EN

ADC2EN

ADC3EN

SDIOEN

SPI1EN

Reserved

TIM2EN

TIM3EN

TIM4EN

TIM5EN

TIM6EN

TIM7EN

TIM12EN

TIM13EN

TIM14EN

Reserved

WWDGEN

Reserved

SYSCFGEN SPI2EN

Reserved
Reserved

LSEBYP

Reserved
CAN2LPEN

OTGHSLPEN
ETHMACPTPLPEN

Reserved

RTCEN

BDRST

DACLPEN
PWRLPEN

OTGHSULPILPEN

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

0x44
Reserved

RCC_
APB1ENR

Reserved

0x40

Reserved

Register
name

LPWRRSTF

Addr.
offset

Reserved SSCGEN

Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)
RM0090

Table 34. RCC register map and reset values (continued)

RM0090

8

General-purpose I/Os (GPIO)

General-purpose I/Os (GPIO)
This section applies to the whole STM32F4xx family, unless otherwise specified.

8.1

GPIO introduction
Each general-purpose I/O port has four 32-bit configuration registers (GPIOx_MODER,
GPIOx_OTYPER, GPIOx_OSPEEDR and GPIOx_PUPDR), two 32-bit data registers
(GPIOx_IDR and GPIOx_ODR), a 32-bit set/reset register (GPIOx_BSRR), a 32-bit locking
register (GPIOx_LCKR) and two 32-bit alternate function selection register (GPIOx_AFRH
and GPIOx_AFRL).

8.2

8.3

GPIO main features
•

Up to 16 I/Os under control

•

Output states: push-pull or open drain + pull-up/down

•

Output data from output data register (GPIOx_ODR) or peripheral (alternate function
output)

•

Speed selection for each I/O

•

Input states: floating, pull-up/down, analog

•

Input data to input data register (GPIOx_IDR) or peripheral (alternate function input)

•

Bit set and reset register (GPIOx_BSRR) for bitwise write access to GPIOx_ODR

•

Locking mechanism (GPIOx_LCKR) provided to freeze the I/O configuration

•

Analog function

•

Alternate function input/output selection registers (at most 16 AFs per I/O)

•

Fast toggle capable of changing every two clock cycles

•

Highly flexible pin multiplexing allows the use of I/O pins as GPIOs or as one of several
peripheral functions

GPIO functional description
Subject to the specific hardware characteristics of each I/O port listed in the datasheet, each
port bit of the general-purpose I/O (GPIO) ports can be individually configured by software in
several modes:
•

Input floating

•

Input pull-up

•

Input-pull-down

•

Analog

•

Output open-drain with pull-up or pull-down capability

•

Output push-pull with pull-up or pull-down capability

•

Alternate function push-pull with pull-up or pull-down capability

•

Alternate function open-drain with pull-up or pull-down capability

RM0090 Rev 17

267/1747
288

General-purpose I/Os (GPIO)

RM0090

Each I/O port bit is freely programmable, however the I/O port registers have to be
accessed as 32-bit words, half-words or bytes. The purpose of the GPIOx_BSRR register is
to allow atomic read/modify accesses to any of the GPIO registers. In this way, there is no
risk of an IRQ occurring between the read and the modify access.
Figure 25 shows the basic structure of a 5 V tolerant I/O port bit. Table 39 gives the possible
port bit configurations.
Figure 25. Basic structure of a five-volt tolerant I/O port bit

!NALOG

4O ON CHIP
PERIPHERAL

!LTERNATE FUNCTION INPUT

2EADWRITE
&ROM ON CHIP
PERIPHERAL

)NPUT DATA REGISTER

6$$
44, 3CHMITT
TRIGGER

ONOFF

6$$?&4 

0ROTECTION
DIODE

0ULL
UP

)NPUT DRIVER

/UTPUT DATA REGISTER

"IT SETRESET REGISTERS

2EAD

7RITE

ONOFF

)/ PIN

/UTPUT DRIVER

6$$

ONOFF

0 -/3

633

/UTPUT
CONTROL

0ROTECTION
DIODE

0ULL
DOWN

633

. -/3
633

!LTERNATE FUNCTION OUTPUT

0USH PULL
OPEN DRAIN OR
DISABLED

!NALOG
AIB

1. VDD_FT is a potential specific to five-volt tolerant I/Os and different from VDD.

Table 35. Port bit configuration table(1)
MODER(i)
[1:0]

01

268/1747

OTYPER(i)

OSPEEDR(i)
[B:A]

PUPDR(i)
[1:0]

I/O configuration

0

0

0

GP output

PP

0

0

1

GP output

PP + PU

0

1

0

GP output

PP + PD

1

1

Reserved

0

0

GP output

OD

1

0

1

GP output

OD + PU

1

1

0

GP output

OD + PD

1

1

1

Reserved (GP output OD)

0
1

SPEED
[B:A]

RM0090 Rev 17

RM0090

General-purpose I/Os (GPIO)
Table 35. Port bit configuration table(1) (continued)
MODER(i)
[1:0]

10

00

11

OTYPER(i)

OSPEEDR(i)
[B:A]

PUPDR(i)
[1:0]

I/O configuration

0

0

0

AF

PP

0

0

1

AF

PP + PU

0

1

0

AF

PP + PD

1

1

Reserved

0

0

AF

OD

1

0

1

AF

OD + PU

1

1

0

AF

OD + PD

1

1

1

Reserved

0

SPEED
[B:A]

1

x

x

x

0

0

Input

Floating

x

x

x

0

1

Input

PU

x

x

x

1

0

Input

PD

x

x

x

1

1

Reserved (input floating)

x

x

x

0

0

Input/output

x

x

x

0

1

x

x

x

1

0

x

x

x

1

1

Analog

Reserved

1. GP = general-purpose, PP = push-pull, PU = pull-up, PD = pull-down, OD = open-drain, AF = alternate
function.

8.3.1

General-purpose I/O (GPIO)
During and just after reset, the alternate functions are not active and the I/O ports are
configured in input floating mode.
The debug pins are in AF pull-up/pull-down after reset:
•

PA15: JTDI in pull-up

•

PA14: JTCK/SWCLK in pull-down

•

PA13: JTMS/SWDAT in pull-up

•

PB4: NJTRST in pull-up

•

PB3: JTDO in floating state

When the pin is configured as output, the value written to the output data register
(GPIOx_ODR) is output on the I/O pin. It is possible to use the output driver in push-pull
mode or open-drain mode (only the N-MOS is activated when 0 is output).
The input data register (GPIOx_IDR) captures the data present on the I/O pin at every AHB1
clock cycle.
All GPIO pins have weak internal pull-up and pull-down resistors, which can be activated or
not depending on the value in the GPIOx_PUPDR register.

RM0090 Rev 17

269/1747
288

General-purpose I/Os (GPIO)

8.3.2

RM0090

I/O pin multiplexer and mapping
The microcontroller I/O pins are connected to onboard peripherals/modules through a
multiplexer that allows only one peripheral’s alternate function (AF) connected to an I/O pin
at a time. In this way, there can be no conflict between peripherals sharing the same I/O pin.
Each I/O pin has a multiplexer with sixteen alternate function inputs (AF0 to AF15) that can
be configured through the GPIOx_AFRL (for pin 0 to 7) and GPIOx_AFRH (for pin 8 to 15)
registers:
•

After reset all I/Os are connected to the system’s alternate function 0 (AF0)

•

The peripherals’ alternate functions are mapped from AF1 to AF13

•

Cortex®-M4 with FPU EVENTOUT is mapped on AF15

This structure is shown in Figure 26 below.
In addition to this flexible I/O multiplexing architecture, each peripheral has alternate
functions mapped onto different I/O pins to optimize the number of peripherals available in
smaller packages.
To use an I/O in a given configuration, proceed as follows:
•

System function
Connect the I/O to AF0 and configure it depending on the function used:

Note:

–

JTAG/SWD, after each device reset these pins are assigned as dedicated pins
immediately usable by the debugger host (not controlled by the GPIO controller)

–

RTC_REFIN: this pin should be configured in Input floating mode

–

MCO1 and MCO2: these pins have to be configured in alternate function mode.

The user can disable some or all of the JTAG/SWD pins and so release the associated pins
for GPIO usage (released pins highlighted in gray in the table).
For more details please refer to Section 7.2.10: Clock-out capability and Section 6.2.10:
Clock-out capability.

270/1747

RM0090 Rev 17

RM0090

General-purpose I/Os (GPIO)
Table 36. Flexible SWJ-DP pin assignment
SWJ I/O pin assigned
PA13 /
JTMS/
SWDIO

PA14 /
JTCK/
SWCLK

PA15 /
JTDI

PB3 /
JTDO

PB4/
NJTRST

Full SWJ (JTAG-DP + SW-DP) - Reset state

X

X

X

X

X

Full SWJ (JTAG-DP + SW-DP) but without
NJTRST

X

X

X

X

JTAG-DP Disabled and SW-DP Enabled

X

X

Available debug ports

JTAG-DP Disabled and SW-DP Disabled

•

Released

GPIO
Configure the desired I/O as output or input in the GPIOx_MODER register.

•

Peripheral alternate function
For the ADC and DAC, configure the desired I/O as analog in the GPIOx_MODER
register.
For other peripherals:

•

–

Configure the desired I/O as an alternate function in the GPIOx_MODER register

–

Select the type, pull-up/pull-down and output speed via the GPIOx_OTYPER,
GPIOx_PUPDR and GPIOx_OSPEEDR registers, respectively

–

Connect the I/O to the desired AFx in the GPIOx_AFRL or GPIOx_AFRH register

EVENTOUT
Configure the I/O pin used to output the Cortex®-M4 with FPU EVENTOUT signal by
connecting it to AF15

Note:

EVENTOUT is not mapped onto the following I/O pins: PC13, PC14, PC15, PH0, PH1 and
PI8.
Please refer to the “Alternate function mapping” table in the datasheets for the detailed
mapping of the system and peripherals’ alternate function I/O pins.

RM0090 Rev 17

271/1747
288

General-purpose I/Os (GPIO)

RM0090

Figure 26. Selecting an alternate function on STM32F405xx/07xx and
STM32F415xx/17xx
)RUSLQVWRWKH*3,2[B$)5/>@UHJLVWHUVHOHFWVWKHGHGLFDWHGDOWHUQDWHIXQFWLRQ
$) V\VWHP
$) 7,07,0
$) 7,0
$) 7,0
$) ,&
$) 63,63,
$) 63,
$) 86$57
$) 86$57
$) &$1&$17,0
$) 27*B)627*B+6
$) (7+
$) )60&6',227*B+6 
$) '&0,
$)
$) (9(17287

3LQ[ [ 


$)5/>@
)RUSLQVWRWKH*3,2[B$)5+>@UHJLVWHUVHOHFWVWKHGHGLFDWHGDOWHUQDWHIXQFWLRQ
$) V\VWHP
$) 7,07,0
$) 7,0
$) 7,0
$) ,&
$) 63,63,
$) 63,
$) 86$57
$) 86$57
$) &$1&$17,0
$) 27*B)627*B+6
$) (7+
$) )60&6',227*B+6 
$) '&0,
$)
$) (9(17287

3LQ[ [ 


$)5+>@
1. Configured in FS.

272/1747

RM0090 Rev 17

DL

RM0090

General-purpose I/Os (GPIO)
Figure 27. Selecting an alternate function on STM32F42xxx and STM32F43xxx

)RUSLQVWRWKH*3,2[B$)5/>@UHJLVWHUVHOHFWVWKHGHGLFDWHGDOWHUQDWHIXQFWLRQ
$) V\VWHP
$) 7,07,0
$) 7,0
$) 7,0
$) ,&
$) 63,
$) 63,6$,
$) 86$57
$) 86$57
$) &$1&$1/7'&7,0
$) 27*B)627*B+6
$) (7+
$) )0&6',227*B+6 
$) '&0,
$) /7'&
$) (9(17287

3LQ[ [ 


$)5/>@
)RUSLQVWRWKH*3,2[B$)5+>@UHJLVWHUVHOHFWVWKHGHGLFDWHGDOWHUQDWHIXQFWLRQ

$) V\VWHP
$) 7,07,0
$) 7,0
$) 7,0
$) ,&
$) 63,
$) 63,6$,
$) 86$57
$) 86$57
$) &$1&$17,0
$) 27*B)627*B+6
$) (7+
$) )0&6',227*B+6 
$) '&0,
$) /7'&
$) (9(17287

3LQ[ [ 


$)5+>@
069

1. Configured in FS.

RM0090 Rev 17

273/1747
288

General-purpose I/Os (GPIO)

8.3.3

RM0090

I/O port control registers
Each of the GPIOs has four 32-bit memory-mapped control registers (GPIOx_MODER,
GPIOx_OTYPER, GPIOx_OSPEEDR, GPIOx_PUPDR) to configure up to 16 I/Os.
The GPIOx_MODER register is used to select the I/O direction (input, output, AF, analog).
The GPIOx_OTYPER and GPIOx_OSPEEDR registers are used to select the output type
(push-pull or open-drain) and speed (the I/O speed pins are directly connected to the
corresponding GPIOx_OSPEEDR register bits whatever the I/O direction). The
GPIOx_PUPDR register is used to select the pull-up/pull-down whatever the I/O direction.

8.3.4

I/O port data registers
Each GPIO has two 16-bit memory-mapped data registers: input and output data registers
(GPIOx_IDR and GPIOx_ODR). GPIOx_ODR stores the data to be output, it is read/write
accessible. The data input through the I/O are stored into the input data register
(GPIOx_IDR), a read-only register.
See Section 8.4.5: GPIO port input data register (GPIOx_IDR) (x = A..I/J/K) and
Section 8.4.6: GPIO port output data register (GPIOx_ODR) (x = A..I/J/K) for the register
descriptions.

8.3.5

I/O data bitwise handling
The bit set reset register (GPIOx_BSRR) is a 32-bit register which allows the application to
set and reset each individual bit in the output data register (GPIOx_ODR). The bit set reset
register has twice the size of GPIOx_ODR.
To each bit in GPIOx_ODR, correspond two control bits in GPIOx_BSRR: BSRR(i) and
BSRR(i+SIZE). When written to 1, bit BSRR(i) sets the corresponding ODR(i) bit. When
written to 1, bit BSRR(i+SIZE) resets the ODR(i) corresponding bit.
Writing any bit to 0 in GPIOx_BSRR does not have any effect on the corresponding bit in
GPIOx_ODR. If there is an attempt to both set and reset a bit in GPIOx_BSRR, the set
action takes priority.
Using the GPIOx_BSRR register to change the values of individual bits in GPIOx_ODR is a
“one-shot” effect that does not lock the GPIOx_ODR bits. The GPIOx_ODR bits can always
be accessed directly. The GPIOx_BSRR register provides a way of performing atomic
bitwise handling.
There is no need for the software to disable interrupts when programming the GPIOx_ODR
at bit level: it is possible to modify one or more bits in a single atomic AHB1 write access.

8.3.6

GPIO locking mechanism
It is possible to freeze the GPIO control registers by applying a specific write sequence to
the GPIOx_LCKR register. The frozen registers are GPIOx_MODER, GPIOx_OTYPER,
GPIOx_OSPEEDR, GPIOx_PUPDR, GPIOx_AFRL and GPIOx_AFRH.
To write the GPIOx_LCKR register, a specific write / read sequence has to be applied. When
the right LOCK sequence is applied to bit 16 in this register, the value of LCKR[15:0] is used
to lock the configuration of the I/Os (during the write sequence the LCKR[15:0] value must
be the same). When the LOCK sequence has been applied to a port bit, the value of the port
bit can no longer be modified until the next MCU or peripheral reset. Each GPIOx_LCKR bit

274/1747

RM0090 Rev 17

RM0090

General-purpose I/Os (GPIO)
freezes the corresponding bit in the control registers (GPIOx_MODER, GPIOx_OTYPER,
GPIOx_OSPEEDR, GPIOx_PUPDR, GPIOx_AFRL and GPIOx_AFRH).
The LOCK sequence (refer to Section 8.4.8: GPIO port configuration lock register
(GPIOx_LCKR) (x = A..I/J/K)) can only be performed using a word (32-bit long) access to
the GPIOx_LCKR register due to the fact that GPIOx_LCKR bit 16 has to be set at the same
time as the [15:0] bits.
For more details please refer to LCKR register description in Section 8.4.8: GPIO port
configuration lock register (GPIOx_LCKR) (x = A..I/J/K).

8.3.7

I/O alternate function input/output
Two registers are provided to select one out of the sixteen alternate function inputs/outputs
available for each I/O. With these registers, you can connect an alternate function to some
other pin as required by your application.
This means that a number of possible peripheral functions are multiplexed on each GPIO
using the GPIOx_AFRL and GPIOx_AFRH alternate function registers. The application can
thus select any one of the possible functions for each I/O. The AF selection signal being
common to the alternate function input and alternate function output, a single channel is
selected for the alternate function input/output of one I/O.
To know which functions are multiplexed on each GPIO pin, refer to the datasheets.

Note:

The application is allowed to select one of the possible peripheral functions for each I/O at a
time.

8.3.8

External interrupt/wakeup lines
All ports have external interrupt capability. To use external interrupt lines, the port must be
configured in input mode, refer to Section 12.2: External interrupt/event controller (EXTI)
and Section 12.2.3: Wakeup event management.

8.3.9

Input configuration
When the I/O port is programmed as Input:
•

the output buffer is disabled

•

the Schmitt trigger input is activated

•

the pull-up and pull-down resistors are activated depending on the value in the
GPIOx_PUPDR register

•

The data present on the I/O pin are sampled into the input data register every AHB1
clock cycle

•

A read access to the input data register provides the I/O State

Figure 28 shows the input configuration of the I/O port bit.

RM0090 Rev 17

275/1747
288

General-purpose I/Os (GPIO)

RM0090

)NPUT DATA REGISTER

Figure 28. Input floating/pull up/pull down configurations

2EADWRITE

/UTPUT DATA REGISTER

7RITE

"IT SETRESET REGISTERS

2EAD

ON

44, 3CHMITT
TRIGGER

6$$ 6$$
ONOFF
PULL
UP

INPUT DRIVER

PROTECTION
DIODE
)/ PIN

ONOFF
OUTPUT DRIVER
PULL
DOWN
633

PROTECTION
DIODE

633

AIB

8.3.10

Output configuration
When the I/O port is programmed as output:
•

The output buffer is enabled:
–

Open drain mode: A “0” in the Output register activates the N-MOS whereas a “1”
in the Output register leaves the port in Hi-Z (the P-MOS is never activated)

–

Push-pull mode: A “0” in the Output register activates the N-MOS whereas a “1” in
the Output register activates the P-MOS

•

The Schmitt trigger input is activated

•

The weak pull-up and pull-down resistors are activated or not depending on the value
in the GPIOx_PUPDR register

•

The data present on the I/O pin are sampled into the input data register every AHB1
clock cycle

•

A read access to the input data register gets the I/O state

•

A read access to the output data register gets the last written value

Figure 29 shows the output configuration of the I/O port bit.

276/1747

RM0090 Rev 17

RM0090

General-purpose I/Os (GPIO)

)NPUT DATA REGISTER

Figure 29. Output configuration

7RITE

2EADWRITE

ON

6$$

6$$

44, 3CHMITT
TRIGGER

ONOFF

)NPUT DRIVER

/UTPUT DATA REGISTER

"IT SETRESET REGISTERS

2EAD

PROTECTION
DIODE

PULL
UP

/UTPUT DRIVER

6$$

)/ PIN

ONOFF

0 -/3
/UTPUT
CONTROL

PROTECTION
DIODE

PULL
DOWN

633

633

. -/3
0USH PULL OR
633
/PEN DRAIN

AIB

Alternate function configuration
When the I/O port is programmed as alternate function:
•

The output buffer can be configured as open-drain or push-pull

•

The output buffer is driven by the signal coming from the peripheral (transmitter enable
and data)

•

The Schmitt trigger input is activated

•

The weak pull-up and pull-down resistors are activated or not depending on the value
in the GPIOx_PUPDR register

•

The data present on the I/O pin are sampled into the input data register every AHB1
clock cycle

•

A read access to the input data register gets the I/O state

Figure 30 shows the Alternate function configuration of the I/O port bit.
Figure 30. Alternate function configuration
!LTERNATE FUNCTION INPUT

2EAD

2EADWRITE
&ROM ON CHIP
PERIPHERAL

6$$ 6$$
44, 3CHMITT
TRIGGER

ONOFF
PROTECTION
DIODE

0ULL
UP

)NPUT DRIVER
/UTPUT DATA REGISTER

7RITE

ON

)NPUT DATA REGISTER

4O ON CHIP
PERIPHERAL

"IT SETRESET REGISTERS

8.3.11

)/ PIN
/UTPUT DRIVER

ONOFF

6$$
0 -/3

/UTPUT
CONTROL

PROTECTION
DIODE

0ULL
DOWN

633

633

. -/3
633

PUSH PULL OR
OPEN DRAIN

!LTERNATE FUNCTION OUTPUT
AIB

RM0090 Rev 17

277/1747
288

General-purpose I/Os (GPIO)

8.3.12

RM0090

Analog configuration
When the I/O port is programmed as analog configuration:

Note:

•

The output buffer is disabled

•

The Schmitt trigger input is deactivated, providing zero consumption for every analog
value of the I/O pin. The output of the Schmitt trigger is forced to a constant value (0).

•

The weak pull-up and pull-down resistors are disabled

•

Read access to the input data register gets the value “0”

In the analog configuration, the I/O pins cannot be 5 Volt tolerant.
Figure 31 shows the high-impedance, analog-input configuration of the I/O port bit.
Figure 31. High impedance-analog configuration

)NPUT DATA REGISTER

!NALOG

4O ON CHIP
PERIPHERAL

2EADWRITE

&ROM ON CHIP
PERIPHERAL

8.3.13

/UTPUT DATA REGISTER

7RITE

"IT SETRESET REGISTERS

2EAD

OFF

6$$

44, 3CHMITT
TRIGGER

PROTECTION
DIODE

)NPUT DRIVER

)/ PIN
PROTECTION
DIODE
633

!NALOG
AI

Using the OSC32_IN/OSC32_OUT pins as GPIO PC14/PC15
port pins
The LSE oscillator pins OSC32_IN and OSC32_OUT can be used as general-purpose
PC14 and PC15 I/Os, respectively, when the LSE oscillator is off. The PC14 and PC15 I/Os
are only configured as LSE oscillator pins OSC32_IN and OSC32_OUT when the LSE
oscillator is ON. This is done by setting the LSEON bit in the RCC_BDCR register. The LSE
has priority over the GPIO function.

Note:

The PC14/PC15 GPIO functionality is lost when the 1.2 V domain is powered off (by the
device entering the standby mode) or when the backup domain is supplied by VBAT (VDD no
more supplied). In this case the I/Os are set in analog input mode.

8.3.14

Using the OSC_IN/OSC_OUT pins as GPIO PH0/PH1 port pins
The HSE oscillator pins OSC_IN/OSC_OUT can be used as general-purpose PH0/PH1
I/Os, respectively, when the HSE oscillator is OFF. (after reset, the HSE oscillator is off). The
PH0/PH1 I/Os are only configured as OSC_IN/OSC_OUT HSE oscillator pins when the
HSE oscillator is ON. This is done by setting the HSEON bit in the RCC_CR register. The
HSE has priority over the GPIO function.

278/1747

RM0090 Rev 17

RM0090

8.3.15

General-purpose I/Os (GPIO)

Selection of RTC_AF1 and RTC_AF2 alternate functions
The STM32F4xx feature two GPIO pins RTC_AF1 and RTC_AF2 that can be used for the
detection of a tamper or time stamp event, or RTC_ALARM, or RTC_CALIB RTC outputs.
•

The RTC_AF1 (PC13) can be used for the following purposes:

RTC_ALARM output: this output can be RTC Alarm A, RTC Alarm B or RTC Wakeup
depending on the OSEL[1:0] bits in the RTC_CR register
•

RTC_CALIB output: this feature is enabled by setting the COE[23] in the RTC_CR
register

•

RTC_TAMP1: tamper event detection

•

RTC_TS: time stamp event detection

The RTC_AF2 (PI8) can be used for the following purposes:
•

RTC_TAMP1: tamper event detection

•

RTC_TAMP2: tamper event detection

•

RTC_TS: time stamp event detection

The selection of the corresponding pin is performed through the RTC_TAFCR register as
follows:
•

TAMP1INSEL is used to select which pin is used as the RTC_TAMP1 tamper input

•

TSINSEL is used to select which pin is used as the RTC_TS time stamp input

•

ALARMOUTTYPE is used to select whether the RTC_ALARM is output in push-pull or
open-drain mode

The output mechanism follows the priority order listed in Table 37 and Table 38.
Table 37. RTC_AF1 pin(1)
TSINSEL
Pin
Time TAMP1INSEL
ALARMOUTTYPE
RTC_ALARM RTC_CALIB Tamper
TIMESTAMP
configuration
stamp
TAMPER1
RTC_ALARM
enabled
enabled
enabled
pin
and function
enabled pin selection
configuration
selection
Alarm out
output OD

1

Don’t care

Don’t
care

Don’t
care

Don’t care

Don’t care

0

Alarm out
output PP

1

Don’t care

Don’t
care

Don’t
care

Don’t care

Don’t care

1

Calibration
out output PP

0

1

Don’t
care

Don’t
care

Don’t care

Don’t care

Don’t care

TAMPER1
input floating

0

0

1

0

0

Don’t care

Don’t care

TIMESTAMP
and
TAMPER1
input floating

0

0

1

1

0

0

Don’t care

TIMESTAMP
input floating

0

0

0

1

Don’t care

0

Don’t care

Standard
GPIO

0

0

0

0

Don’t care

Don’t care

Don’t care

1. OD: open drain; PP: push-pull.

RM0090 Rev 17

279/1747
288

General-purpose I/Os (GPIO)

RM0090

Table 38. RTC_AF2 pin
TSINSEL
TAMP1INSEL
ALARMOUTTYPE
TIMESTAMP
TAMPER1
RTC_ALARM
pin
pin selection
configuration
selection

Tamper
enabled

Time
stamp
enabled

TAMPER1 input floating

1

0

1

Don’t care

Don’t care

TIMESTAMP and TAMPER1 input
floating

1

1

1

1

Don’t care

TIMESTAMP input floating

0

1

Don’t care

1

Don’t care

Standard GPIO

0

0

Don’t care

Don’t care

Don’t care

Pin configuration and function

280/1747

RM0090 Rev 17

RM0090

General-purpose I/Os (GPIO)

8.4

GPIO registers
This section gives a detailed description of the GPIO registers.
For a summary of register bits, register address offsets and reset values, refer to Table 39.
The GPIO registers can be accessed by byte (8 bits), half-words (16 bits) or words (32 bits).

8.4.1

GPIO port mode register (GPIOx_MODER) (x = A..I/J/K)
Address offset: 0x00
Reset values:

31

30

MODER15[1:0]

•

0xA800 0000 for port A

•

0x0000 0280 for port B

•

0x0000 0000 for other ports
29

28

MODER14[1:0]

27

26

MODER13[1:0]

25

24

23

MODER12[1:0]

22

MODER11[1:0]

21

20

MODER10[1:0]

19

18

MODER9[1:0]

17

16

MODER8[1:0]

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

MODER7[1:0]
rw

rw

MODER6[1:0]
rw

rw

MODER5[1:0]
rw

rw

MODER4[1:0]
rw

rw

MODER3[1:0]
rw

rw

MODER2[1:0]
rw

rw

MODER1[1:0]
rw

MODER0[1:0]

rw

rw

rw

18

17

16

Bits 2y:2y+1 MODERy[1:0]: Port x configuration bits (y = 0..15)
These bits are written by software to configure the I/O direction mode.
00: Input (reset state)
01: General purpose output mode
10: Alternate function mode
11: Analog mode

8.4.2

GPIO port output type register (GPIOx_OTYPER)
(x = A..I/J/K)
Address offset: 0x04
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

OT15

OT14

OT13

OT12

OT11

OT10

OT9

OT8

OT7

OT6

OT5

OT4

OT3

OT2

OT1

OT0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Reserved

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 OTy: Port x configuration bits (y = 0..15)
These bits are written by software to configure the output type of the I/O port.
0: Output push-pull (reset state)
1: Output open-drain

RM0090 Rev 17

281/1747
288

General-purpose I/Os (GPIO)

8.4.3

RM0090

GPIO port output speed register (GPIOx_OSPEEDR)
(x = A..I/J/K)
Address offset: 0x08
Reset values:

31

•

0x0C00 0000 for port A

•

0x0000 00C0 for port B

•

0x0000 0000 for other ports

30

29

OSPEEDR15
[1:0]

28

27

OSPEEDR14
[1:0]

26

25

OSPEEDR13
[1:0]

24

OSPEEDR12
[1:0]

23

22

OSPEEDR11
[1:0]

21

20

OSPEEDR10
[1:0]

19

18

17

16

OSPEEDR9
[1:0]

OSPEEDR8
[1:0]

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

OSPEEDR7[1:0] OSPEEDR6[1:0] OSPEEDR5[1:0] OSPEEDR4[1:0]
rw

rw

rw

rw

rw

rw

rw

OSPEEDR3[1:0]

rw

rw

OSPEEDR2[1:0]

rw

rw

OSPEEDR1
[1:0]

OSPEEDR0
1:0]

rw

rw

rw

rw

rw

Bits 2y:2y+1 OSPEEDRy[1:0]: Port x configuration bits (y = 0..15)
These bits are written by software to configure the I/O output speed.
00: Low speed
01: Medium speed
10: High speed
11: Very high speed
Note: Refer to the product datasheets for the values of OSPEEDRy bits versus VDD
range and external load.

8.4.4

GPIO port pull-up/pull-down register (GPIOx_PUPDR)
(x = A..I/J/K)
Address offset: 0x0C
Reset values:

31

30

PUPDR15[1:0]

•

0x6400 0000 for port A

•

0x0000 0100 for port B

•

0x0000 0000 for other ports
29

28

PUPDR14[1:0]

27

26

PUPDR13[1:0]

25

24

PUPDR12[1:0]

23

22

PUPDR11[1:0]

21

20

PUPDR10[1:0]

19

18

PUPDR9[1:0]

17

16

PUPDR8[1:0]

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

PUPDR7[1:0]
rw

rw

282/1747

PUPDR6[1:0]
rw

rw

PUPDR5[1:0]
rw

rw

PUPDR4[1:0]
rw

rw

PUPDR3[1:0]
rw

RM0090 Rev 17

rw

PUPDR2[1:0]
rw

rw

PUPDR1[1:0]
rw

rw

PUPDR0[1:0]
rw

rw

RM0090

General-purpose I/Os (GPIO)

Bits 2y:2y+1 PUPDRy[1:0]: Port x configuration bits (y = 0..15)
These bits are written by software to configure the I/O pull-up or pull-down
00: No pull-up, pull-down
01: Pull-up
10: Pull-down
11: Reserved

8.4.5

GPIO port input data register (GPIOx_IDR) (x = A..I/J/K)
Address offset: 0x10
Reset value: 0x0000 XXXX (where X means undefined)

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

Reserved
15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

IDR15

IDR14

IDR13

IDR12

IDR11

IDR10

IDR9

IDR8

IDR7

IDR6

IDR5

IDR4

IDR3

IDR2

IDR1

IDR0

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 IDRy: Port input data (y = 0..15)
These bits are read-only and can be accessed in word mode only. They contain the input
value of the corresponding I/O port.

8.4.6

GPIO port output data register (GPIOx_ODR) (x = A..I/J/K)
Address offset: 0x14
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

ODR9

ODR8

ODR7

ODR6

ODR5

ODR4

ODR3

ODR2

ODR1

ODR0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Reserved

ODR15 ODR14 ODR13 ODR12 ODR11 ODR10
rw

rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 ODRy: Port output data (y = 0..15)
These bits can be read and written by software.
Note: For atomic bit set/reset, the ODR bits can be individually set and reset by writing to the
GPIOx_BSRR register (x = A..I/J/K).

RM0090 Rev 17

283/1747
288

General-purpose I/Os (GPIO)

8.4.7

RM0090

GPIO port bit set/reset register (GPIOx_BSRR) (x = A..I/J/K)
Address offset: 0x18
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

BR15

BR14

BR13

BR12

BR11

BR10

BR9

BR8

BR7

BR6

BR5

BR4

BR3

BR2

BR1

BR0

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

BS15

BS14

BS13

BS12

BS11

BS10

BS9

BS8

BS7

BS6

BS5

BS4

BS3

BS2

BS1

BS0

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

Bits 31:16 BRy: Port x reset bit y (y = 0..15)
These bits are write-only and can be accessed in word, half-word or byte mode. A read to
these bits returns the value 0x0000.
0: No action on the corresponding ODRx bit
1: Resets the corresponding ODRx bit
Note: If both BSx and BRx are set, BSx has priority.
Bits 15:0 BSy: Port x set bit y (y= 0..15)
These bits are write-only and can be accessed in word, half-word or byte mode. A read to
these bits returns the value 0x0000.
0: No action on the corresponding ODRx bit
1: Sets the corresponding ODRx bit

8.4.8

GPIO port configuration lock register (GPIOx_LCKR)
(x = A..I/J/K)
This register is used to lock the configuration of the port bits when a correct write sequence
is applied to bit 16 (LCKK). The value of bits [15:0] is used to lock the configuration of the
GPIO. During the write sequence, the value of LCKR[15:0] must not change. When the
LOCK sequence has been applied on a port bit, the value of this port bit can no longer be
modified until the next MCU or peripheral reset.

Note:

A specific write sequence is used to write to the GPIOx_LCKR register. Only word access
(32-bit long) is allowed during this write sequence.
Each lock bit freezes a specific configuration register (control and alternate function
registers).
Address offset: 0x1C
Reset value: 0x0000 0000
Access: 32-bit word only, read/write register

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16
LCKK

Reserved

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

LCK15

LCK14

LCK13

LCK12

LCK11

LCK10

LCK9

LCK8

LCK7

LCK6

LCK5

LCK4

LCK3

LCK2

LCK1

LCK0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

284/1747

RM0090 Rev 17

RM0090

General-purpose I/Os (GPIO)

Bits 31:17 Reserved, must be kept at reset value.
Bit 16 LCKK[16]: Lock key
This bit can be read any time. It can only be modified using the lock key write sequence.
0: Port configuration lock key not active
1: Port configuration lock key active. The GPIOx_LCKR register is locked until an MCU reset
or a peripheral reset occurs.
LOCK key write sequence:
WR LCKR[16] = ‘1’ + LCKR[15:0]
WR LCKR[16] = ‘0’ + LCKR[15:0]
WR LCKR[16] = ‘1’ + LCKR[15:0]
RD LCKR
RD LCKR[16] = ‘1’ (this read operation is optional but it confirms that the lock is active)
Note: During the LOCK key write sequence, the value of LCK[15:0] must not change.
Any error in the lock sequence aborts the lock.
After the first lock sequence on any bit of the port, any read access on the LCKK bit will
return ‘1’ until the next CPU reset.
Bits 15:0 LCKy: Port x lock bit y (y= 0..15)
These bits are read/write but can only be written when the LCKK bit is ‘0.
0: Port configuration not locked
1: Port configuration locked

8.4.9

GPIO alternate function low register (GPIOx_AFRL) (x = A..I/J/K)
Address offset: 0x20
Reset value: 0x0000 0000

31

30

29

28

27

AFRL7[3:0]

26

25

24

23

22

AFRL6[3:0]

21

20

19

AFRL5[3:0]

18

17

16

AFRL4[3:0]

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

AFRL3[3:0]
rw

rw

rw

AFRL2[3:0]
rw

rw

rw

rw

AFRL1[3:0]
rw

rw

rw

rw

AFRL0[3:0]
rw

rw

rw

rw

rw

Bits 31:0 AFRLy: Alternate function selection for port x bit y (y = 0..7)
These bits are written by software to configure alternate function I/Os
AFRLy selection:
0000: AF0
0001: AF1
0010: AF2
0011: AF3
0100: AF4
0101: AF5
0110: AF6
0111: AF7

1000: AF8
1001: AF9
1010: AF10
1011: AF11
1100: AF12
1101: AF13
1110: AF14
1111: AF15

RM0090 Rev 17

285/1747
288

General-purpose I/Os (GPIO)

8.4.10

RM0090

GPIO alternate function high register (GPIOx_AFRH)
(x = A..I/J)
Address offset: 0x24
Reset value: 0x0000 0000

31

30

29

28

27

AFRH15[3:0]

26

25

24

23

AFRH14[3:0]

22

21

20

19

AFRH13[3:0]

18

17

16

AFRH12[3:0]

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

AFRH11[3:0]
rw

rw

rw

AFRH10[3:0]
rw

rw

rw

rw

AFRH9[3:0]
rw

rw

rw

rw

AFRH8[3:0]
rw

rw

Bits 31:0 AFRHy: Alternate function selection for port x bit y (y = 8..15)
These bits are written by software to configure alternate function I/Os
AFRHy selection:
0000: AF0
0001: AF1
0010: AF2
0011: AF3
0100: AF4
0101: AF5
0110: AF6
0111: AF7

286/1747

1000: AF8
1001: AF9
1010: AF10
1011: AF11
1100: AF12
1101: AF13
1110: AF14
1111: AF15

RM0090 Rev 17

rw

rw

rw

Reset value

0x08

0x0C

0x0C
GPIOB_
OSPEEDR

Reset value

GPIOA_PUPDR

Reset value

GPIOB_PUPDR

Reset value
0

0

0

0
0
0

0
0

1

1

0

0
0
0

0
0

0

0

0

0
0
0

0
0

1

0

0

0
0
0

0
0

0

0

0

0
0
0

0
0

0

0

0

0
0
0

0
0

0

0

0

0
0
0

0
0

0

0

0

0

RM0090 Rev 17

Reserved

Reset value

0
0

0
0

0

0

0

0
0
0

0
0

0
0

0

0

0

0
0
0

0
0

0
0

0

0

0

0
0
0

0
0

0
0

0

0

0

0

0
0

0
0

0
1

0

0

1

0

0
0

0
0

1
0

0

0

0

0

0
0

0
0

0
0

0

0

0

0

OT1
OT0

0

0
0
0
0

0
0

0
0

0

0

0

0

OSPEEDR0[1:0]

0

OSPEEDR0[1:0]

0

MODER1[1:0]

0
MODER1[1:0]

0

PUPDR0[1:0]

OT2

0

PUPDR0[1:0]

OT3

0

OSPEEDR1[1:0]

0

OSPEEDR1[1:0]

0

MODER2[1:0]

0
MODER2[1:0]

0

PUPDR1[1:0]

OT4

0

PUPDR1[1:0]

OT5

0

OSPEEDR2[1:0]

0

OSPEEDR2[1:0]

1

MODER3[1:0]

0
MODER3[1:0]

0

PUPDR2[1:0]

OT6

0

PUPDR2[1:0]

OT7

0

OSPEEDR3[1:0]

0

OSPEEDR3[1:0]

1

MODER4[1:0]

0
MODER4[1:0]

0

PUPDR3[1:0]

OT8

0

PUPDR3[1:0]

OT9

0

OSPEEDR4[1:0]

0

OSPEEDR4[1:0]

0

PUPDR4[1:0]

0
MODER5[1:0]

0

MODER5[1:0]

0

PUPDR4[1:0]

OT11
OT10

0

OSPEEDR5[1:0]

0

OSPEEDR5[1:0]

0

PUPDR5[1:0]

0
MODER6[1:0]

0

MODER6[1:0]

0

PUPDR5[1:0]

OT12

0
OT13

0

OSPEEDR6[1:0]

0

OSPEEDR6[1:0]

0

PUPDR6[1:0]

MODER7[1:0]

0

PUPDR6[1:0]

MODER7[1:0]

0
OT14

0
OT15

0

OSPEEDR7[1:0]

0

OSPEEDR7[1:0]

0

PUPDR7[1:0]

0

PUPDR7[1:0]

0

MODER8[1:0]

0
MODER8[1:0]

0

OSPEEDR8[1:0]

0

OSPEEDR8[1:0]

0

PUPDR8[1:0]

0

PUPDR8[1:0]

0

MODER9[1:0]

0
MODER9[1:0]

0

OSPEEDR9[1:0]

0

OSPEEDR9[1:0]

0

PUPDR9[1:0]

0

PUPDR9[1:0]

0

MODER10[1:0]

0
MODER10[1:0]

0

OSPEEDR10[1:0]

0

OSPEEDR10[1:0]

0

PUPDR10[1:0]

0
MODER11[1:0]

0
0

PUPDR10[1:0]

GPIOx_
OTYPER
(where x =
A..I/J/K)
MODER11[1:0]

0

OSPEEDR11[1:0]

0

OSPEEDR11[1:0]

0

PUPDR11[1:0]

0

PUPDR11[1:0]

0
MODER12[1:0]

0
MODER12[1:0]

0

OSPEEDR12[1:0]

0

OSPEEDR12[1:0]

1

PUPDR12[1:0]

0

PUPDR12[1:0]

0
MODER13[1:0]

0
MODER13[1:0]

MODER14[1:0]

0

OSPEEDR13[1:0]

MODER14[1:0]

0

OSPEEDR13[1:0]

0x08
GPIOx_
OSPEEDR
(where x =
A..I/J/K except
B)
0
1

PUPDR13[1:0]

Reset value
0

PUPDR13[1:0]

0x04
GPIOx_MODER
(where x =
C..I/J/K)
0

OSPEEDR14[1:0]

Reset value

OSPEEDR14[1:0]

0x00
GPIOB_
MODER

PUPDR14[1:0]

0x00
1

PUPDR14[1:0]

Reset value

MODER15[1:0]

0

0
0

0
0
0
0

MODER0[1:0]

0

MODER0[1:0]

MODER0[1:0]

MODER1[1:0]

MODER2[1:0]

MODER3[1:0]

MODER4[1:0]

MODER5[1:0]

MODER6[1:0]

MODER7[1:0]

MODER8[1:0]

MODER9[1:0]

MODER10[1:0]

MODER11[1:0]

MODER12[1:0]

MODER13[1:0]

MODER14[1:0]

MODER15[1:0]

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

GPIOA_
MODER

MODER15[1:0]

0x00

OSPEEDR15[1:0]

Register

OSPEEDR15[1:0]

Offset

PUPDR15[1:0]

8.4.11

PUPDR15[1:0]

RM0090
General-purpose I/Os (GPIO)

GPIO register map

The following table gives the GPIO register map and the reset values.
Table 39. GPIO register map and reset values

0

0

0

0

0

0

287/1747

288

General-purpose I/Os (GPIO)

RM0090

Register

0x0C

GPIOx_PUPDR
(where x =
C..I/J/K)

0x10

GPIOx_IDR
(where x =
A..I/J/K)

0x14

GPIOx_ODR
(where x =
A..I/J/K)

0x18

GPIOx_BSRR
(where x =
A..I/J/K)

BR14

BR13

BR12

Reset value

0

0

0

0

0x1C

GPIOx_LCKR
(where x =
A..I/J/K)

0x20

0x24

GPIOx_AFRH
(where x =
A..I/J)

Reset value

Reset value

PUPDR0[1:0]

PUPDR1[1:0]

x

x

x

x

x

x

x

x

x

x

BS5

BS4

BS3

BS2

BS1

BS0

0

0

0

0

0

0

0

0

0

0

LCK4

LCK3

LCK2

LCK1

LCK0

0

LCK5

0

BS6

0

BS7

0

LCK6

0

LCK7

0

BS8

0

BS9

0

LCK8

0

LCK9

0

LCK11

0

LCK10

0

LCK12

0

LCK13

0

LCK14

0

LCK15

0

LCKK

BS11

0

BS10

0

BS12

0

BS13

0

BS14

0

BS15

0

BR0

0

BR1

0

BR2

0

BR3

0

BR4

0

BR5

0

BR6

0

BR7

0

BR8

0

BR9

0
BR11

ODR0

x

ODR1

x

ODR2

x

ODR3

x

ODR4

x

ODR5

x

ODR6

IDR0

0

IDR1

0

IDR2

0

IDR3

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

AFRL7[3:0]

AFRL6[3:0]

AFRL5[3:0]

AFRL4[3:0]

AFRL3[3:0]

AFRL2[3:0]

AFRL1[3:0]

AFRL0[3:0]

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

AFRH15[3:0] AFRH14[3:0] AFRH13[3:0] AFRH12[3:0] AFRH11[3:0] AFRH10[3:0]

AFRH9[3:0]

AFRH8[3:0]

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Refer to Section 2.3: Memory map for the register boundary addresses.

288/1747

PUPDR2[1:0]
0
IDR4

0
IDR5

0
IDR6

0
IDR7

0
IDR8

0

0

Reserved

PUPDR3[1:0]

PUPDR4[1:0]

PUPDR5[1:0]

PUPDR6[1:0]

0

0

Reset value
GPIOx_AFRL
(where x =
A..I/J/K)

0

BR10

Reset value

0

ODR7

Reserved

0

IDR9

Reset value

0

ODR8

Reserved

0

ODR9

0

IDR11

0

PUPDR7[1:0]

PUPDR8[1:0]
0

IDR10

0

ODR11

0

ODR10

0

PUPDR9[1:0]

PUPDR10[1:0]
0

IDR12

0

IDR13

0

ODR12

0

PUPDR11[1:0]

PUPDR12[1:0]
0

ODR13

0

IDR14

0

IDR15

0

ODR14

0

PUPDR13[1:0]

PUPDR14[1:0]

PUPDR15[1:0]
0

ODR15

Reset value

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Offset

BR15

Table 39. GPIO register map and reset values (continued)

RM0090 Rev 17

0

0

0

0

0

0

RM0090

System configuration controller (SYSCFG)

9

System configuration controller (SYSCFG)
The system configuration controller is mainly used to remap the memory accessible in the
code area, select the Ethernet PHY interface and manage the external interrupt line
connection to the GPIOs.
This section applies to the whole STM32F4xx family, unless otherwise specified.

9.1

I/O compensation cell
By default the I/O compensation cell is not used. However when the I/O output buffer speed
is configured in 50 MHz or 100 MHz mode, it is recommended to use the compensation cell
for slew rate control on I/O tf(IO)out)/tr(IO)out commutation to reduce the I/O noise on power
supply.
When the compensation cell is enabled, a READY flag is set to indicate that the
compensation cell is ready and can be used. The I/O compensation cell can be used only
when the supply voltage ranges from 2.4 to 3.6 V.

9.2

SYSCFG registers for STM32F405xx/07xx and
STM32F415xx/17xx

9.2.1

SYSCFG memory remap register (SYSCFG_MEMRMP)
This register is used for specific configurations on memory remap:
•

Two bits are used to configure the type of memory accessible at address 0x0000 0000.
These bits are used to select the physical remap by software and so, bypass the BOOT
pins.

•

After reset these bits take the value selected by the BOOT pins. When booting from
main Flash memory with BOOT pins set to 10 [(BOOT1,BOOT0) = (1,0)] this register
takes the value 0x00.

When the FSMC is remapped at address 0x0000 0000, only the first two regions of Bank 1
memory controller (Bank1 NOR/PSRAM 1 and NOR/PSRAM 2) can be remapped. In remap
mode, the CPU can access the external memory via ICode bus instead of System bus
which boosts up the performance.
Address offset: 0x00
Reset value: 0x0000 000X (X is the memory mode selected by the BOOT pins)
)

31

30

29

28

27

26

25

24

15

14

13

12

11

10

9

8

23

22

21

20

19

18

6

5

4

3

2

17

16

1

0

Reserved
7

Reserved

RM0090 Rev 17

MEM_MODE
rw

rw

289/1747
301

System configuration controller (SYSCFG)

RM0090

Bits 31:2 Reserved, must be kept at reset value.
Bits 1:0 MEM_MODE: Memory mapping selection
Set and cleared by software. This bit controls the memory internal mapping at
address 0x0000 0000. After reset these bits take the value selected by the Boot
pins (except for FSMC).
00: Main Flash memory mapped at 0x0000 0000
01: System Flash memory mapped at 0x0000 0000
10: FSMC Bank1 (NOR/PSRAM 1 and 2) mapped at 0x0000 0000
11: Embedded SRAM (SRAM1) mapped at 0x0000 0000
Note: Refer to Section 2.3: Memory map for details about the memory mapping at
address 0x0000 0000.

9.2.2

SYSCFG peripheral mode configuration register (SYSCFG_PMC)
Address offset: 0x04
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

MII_RMII
_SEL

Reserved

19

18

17

16

2

1

0

Reserved

rw
15

14

13

12

11

10

9

8

7

6

5

4

3

Reserved

Bits 31:24 Reserved, must be kept at reset value.
Bit 23 MII_RMII_SEL: Ethernet PHY interface selection
Set and Cleared by software.These bits control the PHY interface for the
Ethernet MAC.
0: MII interface is selected
1: RMII PHY interface is selected
Note: This configuration must be done while the MAC is under reset and before
enabling the MAC clocks.
Bits 22:0 Reserved, must be kept at reset value.

290/1747

RM0090 Rev 17

RM0090

System configuration controller (SYSCFG)

9.2.3

SYSCFG external interrupt configuration register 1
(SYSCFG_EXTICR1)
Address offset: 0x08
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

6

5

4

3

18

17

16

2

1

0

Reserved
15

14

13

12

11

EXTI3[3:0]
rw

rw

rw

10

9

8

7

EXTI2[3:0]
rw

rw

rw

rw

EXTI1[3:0]
rw

rw

rw

rw

EXTI0[3:0]
rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 EXTIx[3:0]: EXTI x configuration (x = 0 to 3)
These bits are written by software to select the source input for the EXTIx
external interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin
1000: PI[x] pin

9.2.4

SYSCFG external interrupt configuration register 2
(SYSCFG_EXTICR2)
Address offset: 0x0C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

6

5

4

3

18

17

16

2

1

0

Reserved
15

14

13

12

11

EXTI7[3:0]
rw

rw

rw

10

9

8

7

EXTI6[3:0]
rw

rw

rw

rw

EXTI5[3:0]
rw

rw

RM0090 Rev 17

rw

rw

EXTI4[3:0]
rw

rw

rw

rw

rw

291/1747
301

System configuration controller (SYSCFG)

RM0090

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 EXTIx[3:0]: EXTI x configuration (x = 4 to 7)
These bits are written by software to select the source input for the EXTIx
external interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin
1000: PI[x] pin

9.2.5

SYSCFG external interrupt configuration register 3
(SYSCFG_EXTICR3)
Address offset: 0x10
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

5

4

3

2

1

0

rw

rw

rw

Reserved
15

14

13

12

11

rw

rw

EXTI11[3:0]
rw

rw

rw

10

9

8

7

6

rw

rw

rw

EXTI10[3:0]
rw

rw

EXTI9[3:0]
rw

EXTI8[3:0]
rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 EXTIx[3:0]: EXTI x configuration (x = 8 to 11)
These bits are written by software to select the source input for the EXTIx external
interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin
1000: PI[x] pin

292/1747

RM0090 Rev 17

RM0090

System configuration controller (SYSCFG)

9.2.6

SYSCFG external interrupt configuration register 4
(SYSCFG_EXTICR4)
Address offset: 0x14
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

6

5

4

3

18

17

16

2

1

0

Reserved
15

14

13

12

11

EXTI15[3:0]
rw

rw

rw

10

9

8

7

EXTI14[3:0]
rw

rw

rw

rw

EXTI13[3:0]
rw

rw

rw

rw

EXTI12[3:0]
rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 EXTIx[3:0]: EXTI x configuration (x = 12 to 15)
These bits are written by software to select the source input for the EXTIx external
interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin
Note: PI[15:12] are not used.

9.2.7

Compensation cell control register (SYSCFG_CMPCR)
Address offset: 0x20
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

Reserved
15

14

13

12
Reserved

11

10

9

8

7

READY

r

Reserved

CMP_PD
rw

Bits 31:9 Reserved, must be kept at reset value.
Bit 8 READY: Compensation cell ready flag
0: I/O compensation cell not ready
1: O compensation cell ready
Bits 7:2 Reserved, must be kept at reset value.
Bit 0 CMP_PD: Compensation cell power-down
0: I/O compensation cell power-down mode
1: I/O compensation cell enabled

RM0090 Rev 17

293/1747
301

System configuration controller (SYSCFG)

9.2.8

RM0090

SYSCFG register maps for
STM32F405xx/07xx and STM32F415xx/17xx
The following table gives the SYSCFG register map and the reset values.

0x00

SYSCFG_
MEMRMP

MEM_MODE

Register

Reserved

Reset value
0x08
0x0C
0x10
0x14

0x20

SYSCFG_EXTICR1
Reset value
SYSCFG_EXTICR2
Reset value
SYSCFG_EXTICR3
Reset value
SYSCFG_EXTICR4
Reset value
SYSCFG_CMPCR

Reserved

Reserved

x

Reserved

0
EXTI3[3:0]

Reserved

0

0

0

0

EXTI7[3:0]

Reserved

0

Reserved
Reserved

0

0

0

EXTI2[3:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

EXTI5[3:0]

0

EXTI10[3:0]

0

0

EXTI6[3:0]

EXTI11[3:0]
0

EXTI1[3:0]

0

0

0

0

0

EXTI9[3:0]

0

0

0

0

0

EXTI0[3:0]
0

0

0

0

EXTI4[3:0]
0

0

0

0

EXTI8[3:0]
0

0

0

0

EXTI15[3:0]

EXTI14[3:0]

EXTI13[3:0]

EXTI12[3:0]

0

0

0

0

0

0

0

0

0

0

Reserved

READY

SYSCFG_PMC

0x04

x
MII_RMII_SEL

Reset value

Reset value

0

0

0

0

0

Reserved

0
CMP_PD

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 40. SYSCFG register map and reset values (STM32F405xx/07xx and STM32F415xx/17xx)

0

0

Refer to Section 2.3: Memory map for the register boundary addresses.

9.3

SYSCFG registers for STM32F42xxx and STM32F43xxx

9.3.1

SYSCFG memory remap register (SYSCFG_MEMRMP)
This register is used for specific configurations on memory remap:

294/1747

•

Three bits are used to configure the type of memory accessible at address
0x0000 0000. These bits are used to select the physical remap by software and so,
bypass the BOOT pins.

•

After reset these bits take the value selected by the BOOT pins. When booting from
main Flash memory with BOOT pins set to 10 [(BOOT1,BOOT0) = (1,0)] this register
takes the value 0x00.

•

Other bits are used to swap FMC SDRAM Bank 1/2 with FMC Bank 3/4 and configure
the Flash Bank 1/2 mapping

RM0090 Rev 17

RM0090

System configuration controller (SYSCFG)
There are two possible FMC remap at address 0x0000 0000:
•

FMC Bank 1 (NOR/PSRAM 1 and 2) remap:
Only the first two regions of Bank 1 memory controller (Bank1 NOR/PSRAM 1 and
NOR/PSRAM 2) can be remapped.

•

FMC SDRAM Bank 1 remap.

In remap mode at address 0x0000 0000, the CPU can access the external memory via
ICode bus instead of System bus which boosts up the performance.
Address offset: 0x00
Reset value: 0x0000 000X (X is the memory mode selected by the BOOT pins)
Note:

Booting from NOR Flash memory or SDRAM is not allowed. The regions can only be
mapped at 0x0000 0000 through software remap.
)

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

Reserved
15

14

13

Reserved

12

11

10

SWP_FMC

rw

rw

9

8

Res.

FB_
MODE

7

rw

Reserved

MEM_MODE[2:0]
rw

rw

rw

Bits 31:12 Reserved, must be kept at reset value.
Bits 11:10 SWP_FMC: FMC memory mapping swap
Set and cleared by software. These bits are used to swap the FMC SDRAM
Bank 1/2 and FMC Bank 3/4 (SDRAM Bank 1/2 and NAND Bank 2/PCCARD
Bank) in order to enable the code execution from SDRAM Banks without a
physical remapping at 0x0000 0000 address.
00: No FMC memory mapping swap
01: SDRAM banks and NAND Bank 2/PCCARD mapping are swapped. SDRAM
Bank 1 and 2 are mapped at NAND Bank 2 (0x8000 0000) and PCCARD Bank
(0x9000 0000) address, respectively. NAND Bank 2 and PCCARD Bank are
mapped at 0xC000 0000 and 0xD000 0000, respectively.
10: Reserved
11: Reserved
Bit 9 Reserved, must be kept at reset value.

RM0090 Rev 17

295/1747
301

System configuration controller (SYSCFG)

RM0090

Bit 8 FB_MODE: Flash Bank mode selection
Set and cleared by software. This bit controls the Flash Bank 1/2 mapping.
0: Flash Bank 1 is mapped at 0x0800 0000 (and aliased at 0x0000 0000) and
Flash Bank 2 is mapped at 0x0810 0000 (and aliased at 0x0010 0000)
1: Flash Bank 2 is mapped at 0x0800 0000 (and aliased at 0x0000 0000) and
Flash Bank 1 is mapped at 0x0810 0000 (and aliased at 0x0010 0000)
Bits 7:3 Reserved, must be kept at reset value.
Bits 2:0 MEM_MODE: Memory mapping selection
Set and cleared by software. This bit controls the memory internal mapping at
address 0x0000 0000. After reset these bits take the value selected by the Boot
pins (except for FMC).
000: Main Flash memory mapped at 0x0000 0000
001: System Flash memory mapped at 0x0000 0000
010: FMC Bank1 (NOR/PSRAM 1 and 2) mapped at 0x0000 0000
011: Embedded SRAM (SRAM1) mapped at 0x0000 0000
100: FMC/SDRAM Bank 1 mapped at 0x0000 0000
Other configurations are reserved
Note: Refer to Section 2.3: Memory map for details about the memory mapping at
address 0x0000 0000.

9.3.2

SYSCFG peripheral mode configuration register (SYSCFG_PMC)
Address offset: 0x04
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

MII_RMII
_SEL

Reserved

21

20

19

14

13

12

11

10

9

8

7

6

5

4

17

16

ADCxDC2

Reserved

rw
15

18

3

rw

rw

rw

2

1

0

Reserved

Bits 31:24 Reserved, must be kept at reset value.
Bit 23 MII_RMII_SEL: Ethernet PHY interface selection
Set and Cleared by software.These bits control the PHY interface for the
Ethernet MAC.
0: MII interface is selected
1: RMII PHY interface is selected
Note: This configuration must be done while the MAC is under reset and before
enabling the MAC clocks.

296/1747

RM0090 Rev 17

RM0090

System configuration controller (SYSCFG)

Bits 22:19 Reserved, must be kept at reset value.
Bits 18:16 ADCxDC2:
0: No effect.
1: Refer to AN4073 on how to use this bit.
Note: These bits can be set only if the following conditions are met:
- ADC clock higher or equal to 30 MHz.
- Only one ADCxDC2 bit must be selected if ADC conversions do not start
at the same time and the sampling times differ.
- These bits must not be set when the ADCDC1 bit is set in PWR_CR
register.
Bits 15:0 Reserved, must be kept at reset value.

9.3.3

SYSCFG external interrupt configuration register 1
(SYSCFG_EXTICR1)
Address offset: 0x08
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

6

5

4

3

18

17

16

2

1

0

Reserved
15

14

13

12

11

EXTI3[3:0]
rw

rw

rw

10

9

8

7

EXTI2[3:0]
rw

rw

rw

rw

EXTI1[3:0]
rw

rw

rw

rw

EXTI0[3:0]
rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 EXTIx[3:0]: EXTI x configuration (x = 0 to 3)
These bits are written by software to select the source input for the EXTIx
external interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin
1000: PI[x] pin
1001: PJ[x] pin
1010: PK[x] pin

RM0090 Rev 17

297/1747
301

System configuration controller (SYSCFG)

9.3.4

RM0090

SYSCFG external interrupt configuration register 2
(SYSCFG_EXTICR2)
Address offset: 0x0C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

6

5

4

3

18

17

16

2

1

0

Reserved
15

14

13

12

11

EXTI7[3:0]
rw

rw

rw

10

9

8

7

EXTI6[3:0]
rw

rw

rw

rw

EXTI5[3:0]
rw

rw

rw

rw

EXTI4[3:0]
rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 EXTIx[3:0]: EXTI x configuration (x = 4 to 7)
These bits are written by software to select the source input for the EXTIx
external interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin
1000: PI[x] pin
1001: PJ[x] pin
1010: PK[x] pin

9.3.5

SYSCFG external interrupt configuration register 3
(SYSCFG_EXTICR3)
Address offset: 0x10
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

5

4

3

2

1

0

rw

rw

rw

Reserved
15

14

13

12

11

rw

rw

EXTI11[3:0]
rw

298/1747

rw

rw

10

9

8

7

6

rw

rw

rw

EXTI10[3:0]
rw

rw

EXTI9[3:0]

RM0090 Rev 17

rw

EXTI8[3:0]
rw

rw

RM0090

System configuration controller (SYSCFG)

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 EXTIx[3:0]: EXTI x configuration (x = 8 to 11)
These bits are written by software to select the source input for the EXTIx
external interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin
1000: PI[x] pin
1001: PJ[x] pin
Note: PK[11:8] are not used.

9.3.6

SYSCFG external interrupt configuration register 4
(SYSCFG_EXTICR4)
Address offset: 0x14
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

rw

rw

Reserved
15

14

13

12

11

rw

rw

EXTI15[3:0]
rw

rw

rw

10

9

8

7

rw

rw

EXTI14[3:0]
rw

rw

EXTI13[3:0]
rw

rw

EXTI12[3:0]
rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 EXTIx[3:0]: EXTI x configuration (x = 12 to 15)
These bits are written by software to select the source input for the EXTIx external
interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin
1000: PI[x] pin
1001: PJ[x] pin
Note: PK[15:12] are not used.

RM0090 Rev 17

299/1747
301

System configuration controller (SYSCFG)

9.3.7

RM0090

Compensation cell control register (SYSCFG_CMPCR)
Address offset: 0x20
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

6

5

4

3

2

1

16

Reserved
15

14

13

12

11

Reserved

10

9

8

7

READY
r

Reserved

Bits 31:9 Reserved, must be kept at reset value.
Bit 8 READY: Compensation cell ready flag
0: I/O compensation cell not ready
1: O compensation cell ready
Bits 7:2 Reserved, must be kept at reset value.
Bit 0 CMP_PD: Compensation cell power-down
0: I/O compensation cell power-down mode
1: I/O compensation cell enabled

300/1747

RM0090 Rev 17

0
CMP_PD
rw

RM0090

9.3.8

System configuration controller (SYSCFG)

SYSCFG register maps for STM32F42xxx and STM32F43xxx
The following table gives the SYSCFG register map and the reset values.

Reserved

0x0C
0x10
0x14

0x20

Reset value
SYSCFG_EXTICR2
Reset value
SYSCFG_EXTICR3
Reset value
SYSCFG_EXTICR4
Reset value
SYSCFG_CMPCR

0

0

0

0

0

0

0

0

EXTI7[3:0]

Reserved

0

Reserved
Reserved

Reserved

0

x

x

x

Reserved

EXTI3[3:0]

Reserved

0

MEM_
MODE

Reserved

0

0

0

EXTI2[3:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

EXTI5[3:0]

0

EXTI10[3:0]

0

0

EXTI6[3:0]

EXTI11[3:0]
0

EXTI1[3:0]

0

0

0

0

0

EXTI9[3:0]

0

0

0

0

0

EXTI0[3:0]
0

0

0

0

EXTI4[3:0]
0

0

0

0

EXTI8[3:0]
0

0

0

0

EXTI15[3:0]

EXTI14[3:0]

EXTI13[3:0]

EXTI12[3:0]

0

0

0

0

0

0

0

0

0

0

Reset value

0

0

0

0

0

Reserved

0

0
CMP_PD

0x08

Reserved

ADC1DC2

Reset value
SYSCFG_EXTICR1

Reserved

ADC2DC2

SYSCFG_PMC

ADC3DC2

0x04

0
MII_RMII_SEL

Reset value

FB_MODE

SYSCFG_
MEMRMP

READY

0x00

Reserved

Register

SWP_FMC

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 41. SYSCFG register map and reset values (STM32F42xxx and STM32F43xxx)

0

Refer to Section 2.3: Memory map for the register boundary addresses.

RM0090 Rev 17

301/1747
301

DMA controller (DMA)

10

RM0090

DMA controller (DMA)
This section applies to the whole STM32F4xx family, unless otherwise specified.

10.1

DMA introduction
Direct memory access (DMA) is used in order to provide high-speed data transfer between
peripherals and memory and between memory and memory. Data can be quickly moved by
DMA without any CPU action. This keeps CPU resources free for other operations.
The DMA controller combines a powerful dual AHB master bus architecture with
independent FIFO to optimize the bandwidth of the system, based on a complex bus matrix
architecture.
The two DMA controllers have 16 streams in total (8 for each controller), each dedicated to
managing memory access requests from one or more peripherals. Each stream can have
up to 8 channels (requests) in total. And each has an arbiter for handling the priority
between DMA requests.

10.2

DMA main features
The main DMA features are:
•

Dual AHB master bus architecture, one dedicated to memory accesses and one
dedicated to peripheral accesses

•

AHB slave programming interface supporting only 32-bit accesses

•

8 streams for each DMA controller, up to 8 channels (requests) per stream

•

Four-word depth 32 first-in, first-out memory buffers (FIFOs) per stream, that can be
used in FIFO mode or direct mode:
–

FIFO mode: with threshold level software selectable between 1/4, 1/2 or 3/4 of the
FIFO size

–

Direct mode
Each DMA request immediately initiates a transfer from/to the memory. When it is
configured in direct mode (FIFO disabled), to transfer data in memory-toperipheral mode, the DMA preloads only one data from the memory to the internal

302/1747

RM0090 Rev 17

RM0090

DMA controller (DMA)
FIFO to ensure an immediate data transfer as soon as a DMA request is triggered
by a peripheral.
•

Each stream can be configured by hardware to be:
–

a regular channel that supports peripheral-to-memory, memory-to-peripheral and
memory-to-memory transfers

–

a double buffer channel that also supports double buffering on the memory side

•

Each of the 8 streams are connected to dedicated hardware DMA channels (requests)

•

Priorities between DMA stream requests are software-programmable (4 levels
consisting of very high, high, medium, low) or hardware in case of equality (request 0
has priority over request 1, etc.)

•

Each stream also supports software trigger for memory-to-memory transfers (only
available for the DMA2 controller)

•

Each stream request can be selected among up to 8 possible channel requests. This
selection is software-configurable and allows several peripherals to initiate DMA
requests

•

The number of data items to be transferred can be managed either by the DMA
controller or by the peripheral:
–

DMA flow controller: the number of data items to be transferred is softwareprogrammable from 1 to 65535

–

Peripheral flow controller: the number of data items to be transferred is unknown
and controlled by the source or the destination peripheral that signals the end of
the transfer by hardware

•

Independent source and destination transfer width (byte, half-word, word): when the
data widths of the source and destination are not equal, the DMA automatically
packs/unpacks the necessary transfers to optimize the bandwidth. This feature is only
available in FIFO mode

•

Incrementing or non-incrementing addressing for source and destination

•

Supports incremental burst transfers of 4, 8 or 16 beats. The size of the burst is
software-configurable, usually equal to half the FIFO size of the peripheral

•

Each stream supports circular buffer management

•

5 event flags (DMA Half Transfer, DMA Transfer complete, DMA Transfer Error, DMA
FIFO Error, Direct Mode Error) logically ORed together in a single interrupt request for
each stream

RM0090 Rev 17

303/1747
338

DMA controller (DMA)

RM0090

10.3

DMA functional description

10.3.1

General description
Figure 32 shows the block diagram of a DMA.
Figure 32. DMA block diagram

3HULSKHUDOSRUW

),)2

675($0

0HPRU\SRUW

675($0

675($0
),)2
675($0

675($0
),)2
675($0

675($0
),)2
675($0

675($0
),)2
675($0

675($0

675($0
),)2
675($0

5(4B675B&+
5(4B675B&+

),)2

$UELWHU

675($0

5(4B675B&+

),)2

5(4B675($0
5(4B675($0
5(4B675($0
5(4B675($0
5(4B675($0
5(4B675($0
5(4B675($0
5(4B675($0

675($0

5(4B675B&+
5(4B675B&+

675($0

5(4B675B&+

$+%PDVWHU

5(4B675B&+
5(4B675B&+

$+%PDVWHU

'0$FRQWUROOHU

5(4B675B&+

&KDQQHO
VHOHFWLRQ

$+%VODYH
SURJUDPPLQJ
LQWHUIDFH

3URJUDPPLQJSRUW

DL

The DMA controller performs direct memory transfer: as an AHB master, it can take the
control of the AHB bus matrix to initiate AHB transactions.
It can carry out the following transactions:
•

peripheral-to-memory

•

memory-to-peripheral

•

memory-to-memory

The DMA controller provides two AHB master ports: the AHB memory port, intended to be
connected to memories and the AHB peripheral port, intended to be connected to
peripherals. However, to allow memory-to-memory transfers, the AHB peripheral port must
also have access to the memories.
The AHB slave port is used to program the DMA controller (it supports only 32-bit
accesses).
304/1747

RM0090 Rev 17

RM0090

DMA controller (DMA)
See Figure 33 and Figure 34 for the implementation of the system of two DMA controllers.
Figure 33. System implementation of the two DMA controllers
(STM32F405xx/07xx and STM32F415xx/17xx)
$#/$%

"US MATRIX
!("
MULTILAYER

&LASH
MEMORY

)#/$%

 +" 32! +" 32!-

!(" PERIPHERALS
!(" MEMORY
PORT
!(" PERIPH
PORT

&)&/

!RBITER

!(" SLAVE

4O !("
PERIPHERALS

$-! CONTROLLER 

!(" !0"
BRIDGE
DUAL !("

!(" !0"
BRIDGE
DUAL !("

!0"

!0"

!0"
PERIPHERALS

!0"
PERIPHERALS

!(" MEMORY
PORT

&)&/

!(" PERIPH
PORT

$-! CONTROLLER 

!RBITER

!(" SLAVE

4O !("
PERIPHERALS

$-! REQUEST
-!00).'

!(" PERIPHERALS

%XTERNAL MEMORY
CONTROLLER &3-#

-36

1. The DMA1 controller AHB peripheral port is not connected to the bus matrix like DMA2 controller. As a result, only DMA2
streams are able to perform memory-to-memory transfers.

RM0090 Rev 17

305/1747
338

DMA controller (DMA)

RM0090

Figure 34. System implementation of the two DMA controllers (STM32F42xxx and
STM32F43xxx)

"US -ATRIX
!(" MULTILAYER

$#/$%

&LASH
MEMORY

)#/$%

 +" 32! +" 32! +" 32!!(" PERIPHERALS
!(" MEMORY
PORT
!(" PERIPH
PORT

&)&/

!RBITER

!(" SLAVE

4O !("
PERIPHERALS

$-! CONTROLLER 

!(" !0"
BRIDGE
DUAL !("

!(" !0"
BRIDGE
DUAL !("

!0"

!0"

!0"
PERIPHERALS

!0"
PERIPHERALS

!(" MEMORY
PORT
!(" PERIPH
PORT

$-! CONTROLLER 

&)&/

!RBITER

!(" SLAVE

4O !("
PERIPHERALS

$-! REQUEST
-!00).'

!(" PERIPHERALS

%XTERNAL MEMORY
CONTROLLER &-#

-36

1. The DMA1 controller AHB peripheral port is not connected to the bus matrix like in the case of the DMA2
controller, thus only DMA2 streams are able to perform memory-to-memory transfers.

10.3.2

DMA transactions
A DMA transaction consists of a sequence of a given number of data transfers. The number
of data items to be transferred and their width (8-bit, 16-bit or 32-bit) are softwareprogrammable.
Each DMA transfer consists of three operations:

306/1747

•

A loading from the peripheral data register or a location in memory, addressed through
the DMA_SxPAR or DMA_SxM0AR register

•

A storage of the data loaded to the peripheral data register or a location in memory
addressed through the DMA_SxPAR or DMA_SxM0AR register

•

A post-decrement of the DMA_SxNDTR register, which contains the number of
transactions that still have to be performed

RM0090 Rev 17

RM0090

DMA controller (DMA)
After an event, the peripheral sends a request signal to the DMA controller. The DMA
controller serves the request depending on the channel priorities. As soon as the DMA
controller accesses the peripheral, an Acknowledge signal is sent to the peripheral by the
DMA controller. The peripheral releases its request as soon as it gets the Acknowledge
signal from the DMA controller. Once the request has been deasserted by the peripheral,
the DMA controller releases the Acknowledge signal. If there are more requests, the
peripheral can initiate the next transaction.

10.3.3

Channel selection
Each stream is associated with a DMA request that can be selected out of 8 possible
channel requests. The selection is controlled by the CHSEL[2:0] bits in the DMA_SxCR
register.
Figure 35. Channel selection
5(4B675[B&+
5(4B675[B&+
5(4B675[B&+
5(4B675($0[

5(4B675[B&+
5(4B675[B&+
5(4B675[B&+
5(4B675[B&+
5(4B675[B&+





'0$B6[&5





&+6(/>@
DLE

The 8 requests from the peripherals (TIM, ADC, SPI, I2C, etc.) are independently connected
to each channel and their connection depends on the product implementation.
See the following table(s) for examples of DMA request mappings.
Table 42. DMA1 request mapping
Peripheral
requests

Stream 0

Stream 1

Stream 2

Channel 0

SPI3_RX

SPI3_RX

Channel 1

I2C1_RX

TIM7_UP

Channel 2

TIM4_CH1

Channel 3
Channel 4
Channel 5

I2S3_EXT_
RX

I2S3_EXT_

TIM2_UP

RX

TIM2_CH3

UART5_RX

USART3_RX

UART8_TX

(1)

UART7_TX(1)

I2C3_RX
UART4_RX
TIM3_CH4
TIM3_UP

Stream 3
SPI2_RX

TIM4_CH2
I2S2_EXT_
RX
USART3_TX
UART7_RX(1)

RM0090 Rev 17

Stream 4

Stream 5

SPI2_TX

SPI3_TX

TIM7_UP

I2C1_RX

I2S2_EXT_

I2S3_EXT_

TX

TX

I2C3_TX

TIM2_CH1

UART4_TX
TIM3_CH1
TIM3_TRIG

Stream 6

Stream 7
SPI3_TX

I2C1_TX

I2C1_TX

TIM4_UP

TIM4_CH3

TIM2_CH2

TIM2_UP

TIM2_CH4

TIM2_CH4

USART2_RX

USART2_TX

UART5_TX

TIM3_CH2

UART8_RX(1)

TIM3_CH3

307/1747
338

DMA controller (DMA)

RM0090
Table 42. DMA1 request mapping (continued)

Peripheral
requests

Stream 0

Channel 6

Stream 1

Stream 2

TIM5_CH3

TIM5_CH4

TIM5_UP

TIM5_TRIG

Channel 7

TIM5_CH1

TIM6_UP

Stream 3

Stream 4

TIM5_CH4

TIM5_CH2

TIM5_TRIG

I2C2_RX

Stream 5

I2C2_RX

USART3_TX

Stream 6

Stream 7

TIM5_UP
DAC1

DAC2

I2C2_TX

1. These requests are available on STM32F42xxx and STM32F43xxx only.

Table 43. DMA2 request mapping
Peripheral
requests

Stream 0

Channel 0

ADC1

Channel 1
Channel 2

ADC3

Channel 3

SPI1_RX

Channel 4

SPI4_RX(1)

Channel 5
Channel 6

TIM1_TRIG

Channel 7

Stream 1

Stream 2

Stream 3

Stream 4

Stream 5

Stream 6

Stream 7

SAI1_A(1)

TIM8_CH1
TIM8_CH2
TIM8_CH3

SAI1_A(1)

ADC1

SAI1_B(1)

TIM1_CH1
TIM1_CH2
TIM1_CH3

DCMI

ADC2

ADC2

SAI1_B(1)

SPI6_TX(1)

SPI6_RX(1)

DCMI

CRYP_OUT

CRYP_IN

HASH_IN

SDIO

USART1_TX

USART6_TX

USART6_TX

(1)

ADC3

SPI4_TX(1)

SPI5_RX

SPI5_TX

(1)

SPI1_RX

SPI1_TX

SPI1_TX

USART1_RX

SDIO

USART1_RX
(1)

(1)

SPI4_RX

SPI4_TX

TIM1_CH2

TIM1_CH1

TIM1_CH4
TIM1_TRIG
TIM1_COM

TIM1_UP

TIM1_CH3

TIM8_CH1

TIM8_CH2

TIM8_CH3

SPI5_RX(1)

SPI5_TX(1)

USART6_RX

USART6_RX

TIM1_CH1

TIM8_UP

TIM8_CH4
TIM8_TRIG
TIM8_COM

1. These requests are available on STM32F42xxx and STM32F43xxx.

10.3.4

Arbiter
An arbiter manages the 8 DMA stream requests based on their priority for each of the two
AHB master ports (memory and peripheral ports) and launches the peripheral/memory
access sequences.
Priorities are managed in two stages:
•

•

308/1747

Software: each stream priority can be configured in the DMA_SxCR register. There are
four levels:
–

Very high priority

–

High priority

–

Medium priority

–

Low priority

Hardware: If two requests have the same software priority level, the stream with the
lower number takes priority over the stream with the higher number. For example,
Stream 2 takes priority over Stream 4.

RM0090 Rev 17

RM0090

10.3.5

DMA controller (DMA)

DMA streams
Each of the 8 DMA controller streams provides a unidirectional transfer link between a
source and a destination.
Each stream can be configured to perform:
•

Regular type transactions: memory-to-peripherals, peripherals-to-memory or memoryto-memory transfers

•

Double-buffer type transactions: double buffer transfers using two memory pointers for
the memory (while the DMA is reading/writing from/to a buffer, the application can
write/read to/from the other buffer).

The amount of data to be transferred (up to 65535) is programmable and related to the
source width of the peripheral that requests the DMA transfer connected to the peripheral
AHB port. The register that contains the amount of data items to be transferred is
decremented after each transaction.

10.3.6

Source, destination and transfer modes
Both source and destination transfers can address peripherals and memories in the entire
4 GB area, at addresses comprised between 0x0000 0000 and 0xFFFF FFFF.
The direction is configured using the DIR[1:0] bits in the DMA_SxCR register and offers
three possibilities: memory-to-peripheral, peripheral-to-memory or memory-to-memory
transfers. Table 44 describes the corresponding source and destination addresses.
Table 44. Source and destination address
Bits DIR[1:0] of the
DMA_SxCR register

Direction

Source address

Destination address

00

Peripheral-to-memory

DMA_SxPAR

DMA_SxM0AR

01

Memory-to-peripheral

DMA_SxM0AR

DMA_SxPAR

10

Memory-to-memory

DMA_SxPAR

DMA_SxM0AR

11

reserved

-

-

When the data width (programmed in the PSIZE or MSIZE bits in the DMA_SxCR register)
is a half-word or a word, respectively, the peripheral or memory address written into the
DMA_SxPAR or DMA_SxM0AR/M1AR registers has to be aligned on a word or half-word
address boundary, respectively.

Peripheral-to-memory mode
Figure 36 describes this mode.
When this mode is enabled (by setting the bit EN in the DMA_SxCR register), each time a
peripheral request occurs, the stream initiates a transfer from the source to fill the FIFO.
When the threshold level of the FIFO is reached, the contents of the FIFO are drained and
stored into the destination.
The transfer stops once the DMA_SxNDTR register reaches zero, when the peripheral
requests the end of transfers (in case of a peripheral flow controller) or when the EN bit in
the DMA_SxCR register is cleared by software.

RM0090 Rev 17

309/1747
338

DMA controller (DMA)

RM0090

In direct mode (when the DMDIS value in the DMA_SxFCR register is ‘0’), the threshold
level of the FIFO is not used: after each single data transfer from the peripheral to the FIFO,
the corresponding data are immediately drained and stored into the destination.
The stream has access to the AHB source or destination port only if the arbitration of the
corresponding stream is won. This arbitration is performed using the priority defined for
each stream using the PL[1:0] bits in the DMA_SxCR register.
Figure 36. Peripheral-to-memory mode
'0$FRQWUROOHU

'0$B6[0$5
'0$B6[0$5 

$+%PHPRU\
SRUW

0HPRU\EXV
0HPRU\
GHVWLQDWLRQ

5(4B675($0[

$UELWHU

),)2
OHYHO

),)2

$+%SHULSKHUDO
SRUW

3HULSKHUDOEXV
3HULSKHUDO
VRXUFH

'0$B6[3$5

3HULSKHUDO'0$UHTXHVW
069

1. For double-buffer mode.

Memory-to-peripheral mode
Figure 37 describes this mode.
When this mode is enabled (by setting the EN bit in the DMA_SxCR register), the stream
immediately initiates transfers from the source to entirely fill the FIFO.
Each time a peripheral request occurs, the contents of the FIFO are drained and stored into
the destination. When the level of the FIFO is lower than or equal to the predefined
threshold level, the FIFO is fully reloaded with data from the memory.
The transfer stops once the DMA_SxNDTR register reaches zero, when the peripheral
requests the end of transfers (in case of a peripheral flow controller) or when the EN bit in
the DMA_SxCR register is cleared by software.
In direct mode (when the DMDIS value in the DMA_SxFCR register is '0'), the threshold
level of the FIFO is not used. Once the stream is enabled, the DMA preloads the first data to
transfer into an internal FIFO. As soon as the peripheral requests a data transfer, the DMA
transfers the preloaded value into the configured destination. It then reloads again the

310/1747

RM0090 Rev 17

RM0090

DMA controller (DMA)
empty internal FIFO with the next data to be transfer. The preloaded data size corresponds
to the value of the PSIZE bitfield in the DMA_SxCR register.
The stream has access to the AHB source or destination port only if the arbitration of the
corresponding stream is won. This arbitration is performed using the priority defined for
each stream using the PL[1:0] bits in the DMA_SxCR register.
Figure 37. Memory-to-peripheral mode
$-!?3X-!2

$-! CONTROLLER

$-!?3X-!2

!(" MEMORY
PORT

-EMORY BUS

-EMORY
SOURCE
2%1?342%!-X

!RBITER

&)&/
LEVEL

&)&/

!(" PERIPHERAL
PORT

$-!?3X0!2

0ERIPHERAL BUS

0ERIPHERAL
DESTINATION

0ERIPHERAL $-! REQUEST

AI

1. For double-buffer mode.

Memory-to-memory mode
The DMA channels can also work without being triggered by a request from a peripheral.
This is the memory-to-memory mode, described in Figure 38.
When the stream is enabled by setting the Enable bit (EN) in the DMA_SxCR register, the
stream immediately starts to fill the FIFO up to the threshold level. When the threshold level
is reached, the FIFO contents are drained and stored into the destination.
The transfer stops once the DMA_SxNDTR register reaches zero or when the EN bit in the
DMA_SxCR register is cleared by software.
The stream has access to the AHB source or destination port only if the arbitration of the
corresponding stream is won. This arbitration is performed using the priority defined for
each stream using the PL[1:0] bits in the DMA_SxCR register.
Note:

When memory-to-memory mode is used, the Circular and direct modes are not allowed.
Only the DMA2 controller is able to perform memory-to-memory transfers.

RM0090 Rev 17

311/1747
338

DMA controller (DMA)

RM0090
Figure 38. Memory-to-memory mode
$-!?3X-!2

$-! CONTROLLER

$-!?3X-!2

!(" MEMORY
PORT

-EMORY BUS

-EMORY 
DESTINATION
!RBITER
3TREAM ENABLE

&)&/
LEVEL

&)&/

!(" PERIPHERAL
PORT

$-!?3X0!2

0ERIPHERAL BUS

-EMORY 
SOURCE

AI

1. For double-buffer mode.

10.3.7

Pointer incrementation
Peripheral and memory pointers can optionally be automatically post-incremented or kept
constant after each transfer depending on the PINC and MINC bits in the DMA_SxCR
register.
Disabling the Increment mode is useful when the peripheral source or destination data are
accessed through a single register.
If the Increment mode is enabled, the address of the next transfer will be the address of the
previous one incremented by 1 (for bytes), 2 (for half-words) or 4 (for words) depending on
the data width programmed in the PSIZE or MSIZE bits in the DMA_SxCR register.
In order to optimize the packing operation, it is possible to fix the increment offset size for
the peripheral address whatever the size of the data transferred on the AHB peripheral port.
The PINCOS bit in the DMA_SxCR register is used to align the increment offset size with
the data size on the peripheral AHB port, or on a 32-bit address (the address is then
incremented by 4). The PINCOS bit has an impact on the AHB peripheral port only.
If PINCOS bit is set, the address of the next transfer is the address of the previous one
incremented by 4 (automatically aligned on a 32-bit address) whatever the PSIZE value.
The AHB memory port, however, is not impacted by this operation.

312/1747

RM0090 Rev 17

RM0090

10.3.8

DMA controller (DMA)

Circular mode
The Circular mode is available to handle circular buffers and continuous data flows (e.g.
ADC scan mode). This feature can be enabled using the CIRC bit in the DMA_SxCR
register.
When the circular mode is activated, the number of data items to be transferred is
automatically reloaded with the initial value programmed during the stream configuration
phase, and the DMA requests continue to be served.

Note:

In the circular mode, it is mandatory to respect the following rule in case of a burst mode
configured for memory:
DMA_SxNDTR = Multiple of ((Mburst beat) × (Msize)/(Psize)), where:
–

(Mburst beat) = 4, 8 or 16 (depending on the MBURST bits in the DMA_SxCR
register)

–

((Msize)/(Psize)) = 1, 2, 4, 1/2 or 1/4 (Msize and Psize represent the MSIZE and
PSIZE bits in the DMA_SxCR register. They are byte dependent)

–

DMA_SxNDTR = Number of data items to transfer on the AHB peripheral port

For example: Mburst beat = 8 (INCR8), MSIZE = ‘00’ (byte) and PSIZE = ‘01’ (half-word), in
this case: DMA_SxNDTR must be a multiple of (8 × 1/2 = 4).
If this formula is not respected, the DMA behavior and data integrity are not guaranteed.
NDTR must also be a multiple of the Peripheral burst size multiplied by the peripheral data
size, otherwise this could result in a bad DMA behavior.

10.3.9

Double buffer mode
This mode is available for all the DMA1 and DMA2 streams.
The Double buffer mode is enabled by setting the DBM bit in the DMA_SxCR register.
A double-buffer stream works as a regular (single buffer) stream with the difference that it
has two memory pointers. When the Double buffer mode is enabled, the Circular mode is
automatically enabled (CIRC bit in DMA_SxCR is don’t care) and at each end of transaction,
the memory pointers are swapped.
In this mode, the DMA controller swaps from one memory target to another at each end of
transaction. This allows the software to process one memory area while the second memory
area is being filled/used by the DMA transfer. The double-buffer stream can work in both
directions (the memory can be either the source or the destination) as described in
Table 45: Source and destination address registers in Double buffer mode (DBM=1).

Note:

In Double buffer mode, it is possible to update the base address for the AHB memory port
on-the-fly (DMA_SxM0AR or DMA_SxM1AR) when the stream is enabled, by respecting the
following conditions:
•

When the CT bit is ‘0’ in the DMA_SxCR register, the DMA_SxM1AR register can be
written. Attempting to write to this register while CT = '1' sets an error flag (TEIF) and
the stream is automatically disabled.

•

When the CT bit is ‘1’ in the DMA_SxCR register, the DMA_SxM0AR register can be
written. Attempting to write to this register while CT = '0', sets an error flag (TEIF) and
the stream is automatically disabled.

To avoid any error condition, it is advised to change the base address as soon as the TCIF
flag is asserted because, at this point, the targeted memory must have changed from

RM0090 Rev 17

313/1747
338

DMA controller (DMA)

RM0090

memory 0 to 1 (or from 1 to 0) depending on the value of CT in the DMA_SxCR register in
accordance with one of the two above conditions.
For all the other modes (except the Double buffer mode), the memory address registers are
write-protected as soon as the stream is enabled.
Table 45. Source and destination address registers in Double buffer mode (DBM=1)
Bits DIR[1:0] of the
DMA_SxCR register

Direction

Source address

Destination address

00

Peripheral-to-memory

DMA_SxPAR

DMA_SxM0AR /
DMA_SxM1AR

01

Memory-to-peripheral

DMA_SxM0AR /
DMA_SxM1AR

DMA_SxPAR

10

Not allowed(1)

11

Reserved

-

-

1. When the Double buffer mode is enabled, the Circular mode is automatically enabled. Since the memoryto-memory mode is not compatible with the Circular mode, when the Double buffer mode is enabled, it is
not allowed to configure the memory-to-memory mode.

10.3.10

Programmable data width, packing/unpacking, endianess
The number of data items to be transferred has to be programmed into DMA_SxNDTR
(number of data items to transfer bit, NDT) before enabling the stream (except when the
flow controller is the peripheral, PFCTRL bit in DMA_SxCR is set).
When using the internal FIFO, the data widths of the source and destination data are
programmable through the PSIZE and MSIZE bits in the DMA_SxCR register (can be 8-,
16- or 32-bit).
When PSIZE and MSIZE are not equal:
•

The data width of the number of data items to transfer, configured in the DMA_SxNDTR
register is equal to the width of the peripheral bus (configured by the PSIZE bits in the
DMA_SxCR register). For instance, in case of peripheral-to-memory, memory-toperipheral or memory-to-memory transfers and if the PSIZE[1:0] bits are configured for
half-word, the number of bytes to be transferred is equal to 2 × NDT.

•

The DMA controller only copes with little-endian addressing for both source and
destination. This is described in Table 46: Packing/unpacking & endian behavior (bit
PINC = MINC = 1).

This packing/unpacking procedure may present a risk of data corruption when the operation
is interrupted before the data are completely packed/unpacked. So, to ensure data
coherence, the stream may be configured to generate burst transfers: in this case, each
group of transfers belonging to a burst are indivisible (refer to Section 10.3.11: Single and
burst transfers).
In direct mode (DMDIS = 0 in the DMA_SxFCR register), the packing/unpacking of data is
not possible. In this case, it is not allowed to have different source and destination transfer
data widths: both are equal and defined by the PSIZE bits in the DMA_SxCR MSIZE bits are
don’t care).

314/1747

RM0090 Rev 17

RM0090

DMA controller (DMA)
Table 46. Packing/unpacking & endian behavior (bit PINC = MINC = 1)

Number
AHB
AHB
of data
Memory Memory port
memory
peripheral items to transfer address / byte
port
port width transfer number lane
width
(NDT)

8

8

8

16

16

16

32

8

16

32

8

16

32

8

4

2

1

4

1

0x0 / B0[7:0]

1

0x0 / B0[7:0]

0x0 / B0[7:0]

2

0x1 / B1[7:0]

2

0x4 / B1[7:0]

0x1 / B1[7:0]

3

0x2 / B2[7:0]

3

0x8 / B2[7:0]

0x2 / B2[7:0]

4

0x3 / B3[7:0]

4

0xC / B3[7:0]

0x3 / B3[7:0]

1

0x0 / B0[7:0]

1

0x0 / B1|B0[15:0]

0x0 / B1|B0[15:0]

2

0x1 / B1[7:0]

3

0x2 / B2[7:0]

2

0x4 / B3|B2[15:0]

0x2 / B3|B2[15:0]

4

0x3 / B3[7:0]

1

0x0 / B0[7:0]

1

0x0 / B3|B2|B1|B0[31:0] 0x0 / B3|B2|B1|B0[31:0]

2

0x1 / B1[7:0]

3

0x2 / B2[7:0]

4

0x3 / B3[7:0]

1

0x0 / B1|B0[15:0]

1

0x0 / B0[7:0]

0x0 / B0[7:0]

2

0x4 / B1[7:0]

0x1 / B1[7:0]

3

0x8 / B2[7:0]

0x2 / B2[7:0]

4

0xC / B3[7:0]

0x3 / B3[7:0]

2

0x2 / B3|B2[15:0]

1

0x0 / B1|B0[15:0]

1

0x0 / B1|B0[15:0]

0x0 / B1|B0[15:0]

2

0x2 / B1|B0[15:0]

2

0x4 / B3|B2[15:0]

0x2 / B3|B2[15:0]

1

0x0 / B1|B0[15:0]

1

0x0 / B3|B2|B1|B0[31:0] 0x0 / B3|B2|B1|B0[31:0]

2

0x2 / B3|B2[15:0]

1

0x0 / B3|B2|B1|B0[31:0] 1
2

0x0 / B0[7:0]

0x0 / B0[7:0]

0x4 / B1[7:0]

0x1 / B1[7:0]

3

0x8 / B2[7:0]

0x2 / B2[7:0]

4

0xC / B3[7:0]

0x3 / B3[7:0]

1

0x0 / B1|B0[15:0]

0x0 / B1|B0[15:0]

2

0x4 / B3|B2[15:0]

0x2 / B3|B2[15:0]

2

1

4

32

16

2

32

32

1

Note:

Peripheral port address / byte lane
Peripher
al
transfer
PINCOS = 1
PINCOS = 0
number

1

1

0x0 /B3|B2|B1|B0[31:0]

0x0 /B3|B2|B1|B0 [31:0] 1

0x0 /B3|B2|B1|B0 [31:0] 0x0 / B3|B2|B1|B0[31:0]

Peripheral port may be the source or the destination (it could also be the memory source in
the case of memory-to-memory transfer).
PSIZE, MSIZE and NDT[15:0] have to be configured so as to ensure that the last transfer
will not be incomplete. This can occur when the data width of the peripheral port (PSIZE
bits) is lower than the data width of the memory port (MSIZE bits). This constraint is
summarized in Table 47.

RM0090 Rev 17

315/1747
338

DMA controller (DMA)

RM0090
Table 47. Restriction on NDT versus PSIZE and MSIZE

10.3.11

PSIZE[1:0] of DMA_SxCR

MSIZE[1:0] of DMA_SxCR

NDT[15:0] of DMA_SxNDTR

00 (8-bit)

01 (16-bit)

must be a multiple of 2

00 (8-bit)

10 (32-bit)

must be a multiple of 4

01 (16-bit)

10 (32-bit)

must be a multiple of 2

Single and burst transfers
The DMA controller can generate single transfers or incremental burst transfers of 4, 8 or 16
beats.
The size of the burst is configured by software independently for the two AHB ports by using
the MBURST[1:0] and PBURST[1:0] bits in the DMA_SxCR register.
The burst size indicates the number of beats in the burst, not the number of bytes
transferred.
To ensure data coherence, each group of transfers that form a burst are indivisible: AHB
transfers are locked and the arbiter of the AHB bus matrix does not degrant the DMA master
during the sequence of the burst transfer.
Depending on the single or burst configuration, each DMA request initiates a different
number of transfers on the AHB peripheral port:
•

When the AHB peripheral port is configured for single transfers, each DMA request
generates a data transfer of a byte, half-word or word depending on the PSIZE[1:0] bits
in the DMA_SxCR register

•

When the AHB peripheral port is configured for burst transfers, each DMA request
generates 4,8 or 16 beats of byte, half word or word transfers depending on the
PBURST[1:0] and PSIZE[1:0] bits in the DMA_SxCR register.

The same as above has to be considered for the AHB memory port considering the
MBURST and MSIZE bits.
In direct mode, the stream can only generate single transfers and the MBURST[1:0] and
PBURST[1:0] bits are forced by hardware.
The address pointers (DMA_SxPAR or DMA_SxM0AR registers) must be chosen so as to
ensure that all transfers within a burst block are aligned on the address boundary equal to
the size of the transfer.
The burst configuration has to be selected in order to respect the AHB protocol, where
bursts must not cross the 1 KB address boundary because the minimum address space that
can be allocated to a single slave is 1 KB. This means that the 1 KB address boundary
should not be crossed by a burst block transfer, otherwise an AHB error would be
generated, that is not reported by the DMA registers.

316/1747

RM0090 Rev 17

RM0090

10.3.12

DMA controller (DMA)

FIFO
FIFO structure
The FIFO is used to temporarily store data coming from the source before transmitting them
to the destination.
Each stream has an independent 4-word FIFO and the threshold level is softwareconfigurable between 1/4, 1/2, 3/4 or full.
To enable the use of the FIFO threshold level, the direct mode must be disabled by setting
the DMDIS bit in the DMA_SxFCR register.
The structure of the FIFO differs depending on the source and destination data widths, and
is described in Figure 39: FIFO structure.
Figure 39. FIFO structure
 WORDS
%MPTY

3OURCE BYTE
" " " " " " " " " " " " " " " "







&ULL

BYTE LANE 

"

" 

"

"

BYTE LANE 

"

"

"

"

BYTE LANE 

"

"

"

"

BYTE LANE  7 "

7

"

"

7

7

$ESTINATION WORD
7 7 7 7

"

 WORDS
%MPTY

3OURCE BYTE
" " " " " " " " " " " " " " " "

BYTE LANE 







"

" 

"

BYTE LANE  ( "

( "

( "

"

"

"

BYTE LANE  ( "

( "

( "

BYTE LANE 

&ULL
"

(

$ESTINATION HALF WORD

"
( ( ( ( ( ( ( (

"
(

"

 WORDS
%MPTY

3OURCE HALF WORD

BYTE LANE 
( ( ( ( ( ( ( (







&ULL

BYTE LANE 
(

(

(

(

(

(

(

(

$ESTINATION WORD
7 7 7 7

BYTE LANE 
BYTE LANE  7

7

7

7

 WORDS
%MPTY

3OURCE HALF WORD
( ( ( ( ( ( ( (

BYTE LANE 







"

" 

"

BYTE LANE  ( "

( "

( "

"

"

"

BYTE LANE  ( "

( "

( "

BYTE LANE 

&ULL
"

(

"
(

$ESTINATION BYTE

"
" " " " " " " "
" " " " " " " "

"

AI

RM0090 Rev 17

317/1747
338

DMA controller (DMA)

RM0090

FIFO threshold and burst configuration
Caution is required when choosing the FIFO threshold (bits FTH[1:0] of the DMA_SxFCR
register) and the size of the memory burst (MBURST[1:0] of the DMA_SxCR register): The
content pointed by the FIFO threshold must exactly match to an integer number of memory
burst transfers. If this is not in the case, a FIFO error (flag FEIFx of the DMA_HISR or
DMA_LISR register) will be generated when the stream is enabled, then the stream will be
automatically disabled. The allowed and forbidden configurations are described in the
Table 48: FIFO threshold configurations.
Table 48. FIFO threshold configurations
MSIZE

Byte

Half-word

FIFO level MBURST = INCR4

MBURST = INCR8

1/4

1 burst of 4 beats

forbidden

1/2

2 bursts of 4 beats

1 burst of 8 beats

3/4

3 bursts of 4 beats

forbidden

Full

4 bursts of 4 beats

2 bursts of 8 beats

1/4

forbidden

1/2

1 burst of 4 beats

3/4

forbidden

Full

2 bursts of 4 beats

1/2

forbidden

3/4
Full

forbidden

1 burst of 16 beats

forbidden

1 burst of 8 beats

1/4
Word

MBURST = INCR16

forbidden

forbidden

1 burst of 4 beats

In all cases, the burst size multiplied by the data size must not exceed the FIFO size (data
size can be: 1 (byte), 2 (half-word) or 4 (word)).
Incomplete Burst transfer at the end of a DMA transfer may happen if one of the following
conditions occurs:
•

For the AHB peripheral port configuration: the total number of data items (set in the
DMA_SxNDTR register) is not a multiple of the burst size multiplied by the data size

•

For the AHB memory port configuration: the number of remaining data items in the
FIFO to be transferred to the memory is not a multiple of the burst size multiplied by the
data size

In such cases, the remaining data to be transferred will be managed in single mode by the
DMA, even if a burst transaction was requested during the DMA stream configuration.
Note:

When burst transfers are requested on the peripheral AHB port and the FIFO is used
(DMDIS = 1 in the DMA_SxCR register), it is mandatory to respect the following rule to
avoid permanent underrun or overrun conditions, depending on the DMA stream direction:
If (PBURST × PSIZE) = FIFO_SIZE (4 words), FIFO_Threshold = 3/4 is forbidden with
PSIZE = 1, 2 or 4 and PBURST = 4, 8 or 16.
This rule ensures that enough FIFO space at a time will be free to serve the request from
the peripheral.

318/1747

RM0090 Rev 17

RM0090

DMA controller (DMA)

FIFO flush
The FIFO can be flushed when the stream is disabled by resetting the EN bit in the
DMA_SxCR register and when the stream is configured to manage peripheral-to-memory or
memory-to-memory transfers: If some data are still present in the FIFO when the stream is
disabled, the DMA controller continues transferring the remaining data to the destination
(even though stream is effectively disabled). When this flush is completed, the transfer
complete status bit (TCIFx) in the DMA_LISR or DMA_HISR register is set.
The remaining data counter DMA_SxNDTR keeps the value in this case to indicate how
many data items are currently available in the destination memory.
Note that during the FIFO flush operation, if the number of remaining data items in the FIFO
to be transferred to memory (in bytes) is less than the memory data width (for example 2
bytes in FIFO while MSIZE is configured to word), data will be sent with the data width set in
the MSIZE bit in the DMA_SxCR register. This means that memory will be written with an
undesired value. The software may read the DMA_SxNDTR register to determine the
memory area that contains the good data (start address and last address).
If the number of remaining data items in the FIFO is lower than a burst size (if the MBURST
bits in DMA_SxCR register are set to configure the stream to manage burst on the AHB
memory port), single transactions will be generated to complete the FIFO flush.

Direct mode
By default, the FIFO operates in direct mode (DMDIS bit in the DMA_SxFCR is reset) and
the FIFO threshold level is not used. This mode is useful when the system requires an
immediate and single transfer to or from the memory after each DMA request.
When the DMA is configured in direct mode (FIFO disabled), to transfer data in memory-toperipheral mode, the DMA preloads one data from the memory to the internal FIFO to
ensure an immediate data transfer as soon as a DMA request is triggered by a peripheral.
To avoid saturating the FIFO, it is recommended to configure the corresponding stream with
a high priority.
This mode is restricted to transfers where:
•

The source and destination transfer widths are equal and both defined by the
PSIZE[1:0] bits in DMA_SxCR (MSIZE[1:0] bits are don’t care)

•

Burst transfers are not possible (PBURST[1:0] and MBURST[1:0] bits in DMA_SxCR
are don’t care)

Direct mode must not be used when implementing memory-to-memory transfers.

10.3.13

DMA transfer completion
Different events can generate an end of transfer by setting the TCIFx bit in the DMA_LISR
or DMA_HISR status register:
•

In DMA flow controller mode:
–

The DMA_SxNDTR counter has reached zero in the memory-to-peripheral mode

–

The stream is disabled before the end of transfer (by clearing the EN bit in the
DMA_SxCR register) and (when transfers are peripheral-to-memory or memory-

RM0090 Rev 17

319/1747
338

DMA controller (DMA)

RM0090

to-memory) all the remaining data have been flushed from the FIFO into the
memory
•

Note:

In Peripheral flow controller mode:
–

The last external burst or single request has been generated from the peripheral
and (when the DMA is operating in peripheral-to-memory mode) the remaining
data have been transferred from the FIFO into the memory

–

The stream is disabled by software, and (when the DMA is operating in peripheralto-memory mode) the remaining data have been transferred from the FIFO into
the memory

The transfer completion is dependent on the remaining data in FIFO to be transferred into
memory only in the case of peripheral-to-memory mode. This condition is not applicable in
memory-to-peripheral mode.
If the stream is configured in noncircular mode, after the end of the transfer (that is when the
number of data to be transferred reaches zero), the DMA is stopped (EN bit in DMA_SxCR
register is cleared by Hardware) and no DMA request is served unless the software
reprograms the stream and re-enables it (by setting the EN bit in the DMA_SxCR register).

10.3.14

DMA transfer suspension
At any time, a DMA transfer can be suspended to be restarted later on or to be definitively
disabled before the end of the DMA transfer.
There are two cases:

Note:

320/1747

•

The stream disables the transfer with no later-on restart from the point where it was
stopped. There is no particular action to do, except to clear the EN bit in the
DMA_SxCR register to disable the stream. The stream may take time to be disabled
(ongoing transfer is completed first). The transfer complete interrupt flag (TCIF in the
DMA_LISR or DMA_HISR register) is set in order to indicate the end of transfer. The
value of the EN bit in DMA_SxCR is now ‘0’ to confirm the stream interruption. The
DMA_SxNDTR register contains the number of remaining data items at the moment
when the stream was stopped so that the software can determine how many data items
have been transferred before the stream was interrupted.

•

The stream suspends the transfer before the number of remaining data items to be
transferred in the DMA_SxNDTR register reaches 0. The aim is to restart the transfer
later by re-enabling the stream. In order to restart from the point where the transfer was
stopped, the software has to read the DMA_SxNDTR register after disabling the stream
by writing the EN bit in DMA_SxCR register (and then checking that it is at ‘0’) to know
the number of data items already collected. Then:
–

The peripheral and/or memory addresses have to be updated in order to adjust
the address pointers

–

The SxNDTR register has to be updated with the remaining number of data items
to be transferred (the value read when the stream was disabled)

–

The stream may then be re-enabled to restart the transfer from the point it was
stopped

Note that a Transfer complete interrupt flag (TCIF in DMA_LISR or DMA_HISR) is set to
indicate the end of transfer due to the stream interruption.

RM0090 Rev 17

RM0090

10.3.15

DMA controller (DMA)

Flow controller
The entity that controls the number of data to be transferred is known as the flow controller.
This flow controller is configured independently for each stream using the PFCTRL bit in the
DMA_SxCR register.
The flow controller can be:
•

The DMA controller: in this case, the number of data items to be transferred is
programmed by software into the DMA_SxNDTR register before the DMA stream is
enabled.

•

The peripheral source or destination: this is the case when the number of data items to
be transferred is unknown. The peripheral indicates by hardware to the DMA controller
when the last data are being transferred. This feature is only supported for peripherals
which are able to signal the end of the transfer, that is:
–

SDIO

When the peripheral flow controller is used for a given stream, the value written into the
DMA_SxNDTR has no effect on the DMA transfer. Actually, whatever the value written, it will
be forced by hardware to 0xFFFF as soon as the stream is enabled, to respect the following
schemes:
•

Anticipated stream interruption: EN bit in DMA_SxCR register is reset to 0 by the
software to stop the stream before the last data hardware signal (single or burst) is sent
by the peripheral. In such a case, the stream is switched off and the FIFO flush is
triggered in the case of a peripheral-to-memory DMA transfer. The TCIFx flag of the
corresponding stream is set in the status register to indicate the DMA completion. To
know the number of data items transferred during the DMA transfer, read the
DMA_SxNDTR register and apply the following formula:
–

Note:

Number_of_data_transferred = 0xFFFF – DMA_SxNDTR

•

Normal stream interruption due to the reception of a last data hardware signal: the
stream is automatically interrupted when the peripheral requests the last transfer
(single or burst) and when this transfer is complete. the TCIFx flag of the corresponding
stream is set in the status register to indicate the DMA transfer completion. To know the
number of data items transferred, read the DMA_SxNDTR register and apply the same
formula as above.

•

The DMA_SxNDTR register reaches 0: the TCIFx flag of the corresponding stream is
set in the status register to indicate the forced DMA transfer completion. The stream is
automatically switched off even though the last data hardware signal (single or burst)
has not been yet asserted. The already transferred data will not be lost. This means
that a maximum of 65535 data items can be managed by the DMA in a single
transaction, even in peripheral flow control mode.

When configured in memory-to-memory mode, the DMA is always the flow controller and
the PFCTRL bit is forced to 0 by hardware.
The Circular mode is forbidden in the peripheral flow controller mode.

RM0090 Rev 17

321/1747
338

DMA controller (DMA)

10.3.16

RM0090

Summary of the possible DMA configurations
Table 49 summarizes the different possible DMA configurations.
Table 49. Possible DMA configurations

DMA transfer
mode

Peripheral-tomemory

Destination

Circular
mode

DMA

possible

Peripheral

forbidden

DMA

possible

Peripheral

forbidden

DMA only

forbidden

AHB
peripheral port

AHB
AHB
peripheral port memory port

Memory-tomemory

Flow
controller

AHB
AHB
peripheral port memory port

AHB
memory port

Memory-toperipheral

10.3.17

Source

Transfer
type

Direct
mode

single

possible

burst

forbidden

single

possible

burst

forbidden

single

possible

burst

forbidden

single

possible

burst

forbidden

single
burst

forbidden

Double
buffer mode
possible

forbidden

possible

forbidden

forbidden

Stream configuration procedure
The following sequence should be followed to configure a DMA stream x (where x is the
stream number):

322/1747

1.

If the stream is enabled, disable it by resetting the EN bit in the DMA_SxCR register,
then read this bit in order to confirm that there is no ongoing stream operation. Writing
this bit to 0 is not immediately effective since it is actually written to 0 once all the
current transfers have finished. When the EN bit is read as 0, this means that the
stream is ready to be configured. It is therefore necessary to wait for the EN bit to be
cleared before starting any stream configuration. All the stream dedicated bits set in the
status register (DMA_LISR and DMA_HISR) from the previous data block DMA
transfer should be cleared before the stream can be re-enabled.

2.

Set the peripheral port register address in the DMA_SxPAR register. The data will be
moved from/ to this address to/ from the peripheral port after the peripheral event.

3.

Set the memory address in the DMA_SxMA0R register (and in the DMA_SxMA1R
register in the case of a double buffer mode). The data will be written to or read from
this memory after the peripheral event.

4.

Configure the total number of data items to be transferred in the DMA_SxNDTR
register. After each peripheral event or each beat of the burst, this value is
decremented.

5.

Select the DMA channel (request) using CHSEL[2:0] in the DMA_SxCR register.

6.

If the peripheral is intended to be the flow controller and if it supports this feature, set
the PFCTRL bit in the DMA_SxCR register.

7.

Configure the stream priority using the PL[1:0] bits in the DMA_SxCR register.

8.

Configure the FIFO usage (enable or disable, threshold in transmission and reception)

9.

Configure the data transfer direction, peripheral and memory incremented/fixed mode,
single or burst transactions, peripheral and memory data widths, Circular mode,

RM0090 Rev 17

RM0090

DMA controller (DMA)
Double buffer mode and interrupts after half and/or full transfer, and/or errors in the
DMA_SxCR register.
10. Activate the stream by setting the EN bit in the DMA_SxCR register.
As soon as the stream is enabled, it can serve any DMA request from the peripheral
connected to the stream.
Once half the data have been transferred on the AHB destination port, the half-transfer flag
(HTIF) is set and an interrupt is generated if the half-transfer interrupt enable bit (HTIE) is
set. At the end of the transfer, the transfer complete flag (TCIF) is set and an interrupt is
generated if the transfer complete interrupt enable bit (TCIE) is set.

Warning:

10.3.18

To switch off a peripheral connected to a DMA stream
request, it is mandatory to, first, switch off the DMA stream to
which the peripheral is connected, then to wait for EN bit = 0.
Only then can the peripheral be safely disabled.

Error management
The DMA controller can detect the following errors:
•

•

•

Transfer error: the transfer error interrupt flag (TEIFx) is set when:
–

A bus error occurs during a DMA read or a write access

–

A write access is requested by software on a memory address register in Double
buffer mode whereas the stream is enabled and the current target memory is the
one impacted by the write into the memory address register (refer to
Section 10.3.9: Double buffer mode)

FIFO error: the FIFO error interrupt flag (FEIFx) is set if:
–

A FIFO underrun condition is detected

–

A FIFO overrun condition is detected (no detection in memory-to-memory mode
because requests and transfers are internally managed by the DMA)

–

The stream is enabled while the FIFO threshold level is not compatible with the
size of the memory burst (refer to Table 48: FIFO threshold configurations)

Direct mode error: the direct mode error interrupt flag (DMEIFx) can only be set in the
peripheral-to-memory mode while operating in direct mode and when the MINC bit in
the DMA_SxCR register is cleared. This flag is set when a DMA request occurs while
the previous data have not yet been fully transferred into the memory (because the
memory bus was not granted). In this case, the flag indicates that 2 data items were be
transferred successively to the same destination address, which could be an issue if
the destination is not able to manage this situation

In direct mode, the FIFO error flag can also be set under the following conditions:
•

In the peripheral-to-memory mode, the FIFO can be saturated (overrun) if the memory
bus is not granted for several peripheral requests

•

In the memory-to-peripheral mode, an underrun condition may occur if the memory bus
has not been granted before a peripheral request occurs

If the TEIFx or the FEIFx flag is set due to incompatibility between burst size and FIFO
threshold level, the faulty stream is automatically disabled through a hardware clear of its
EN bit in the corresponding stream configuration register (DMA_SxCR).
RM0090 Rev 17

323/1747
338

DMA controller (DMA)

RM0090

If the DMEIFx or the FEIFx flag is set due to an overrun or underrun condition, the faulty
stream is not automatically disabled and it is up to the software to disable or not the stream
by resetting the EN bit in the DMA_SxCR register. This is because there is no data loss
when this kind of errors occur.
When the stream's error interrupt flag (TEIF, FEIF, DMEIF) in the DMA_LISR or DMA_HISR
register is set, an interrupt is generated if the corresponding interrupt enable bit (TEIE,
FEIE, DMIE) in the DMA_SxCR or DMA_SxFCR register is set.
Note:

When a FIFO overrun or underrun condition occurs, the data are not lost because the
peripheral request is not acknowledged by the stream until the overrun or underrun
condition is cleared. If this acknowledge takes too much time, the peripheral itself may
detect an overrun or underrun condition of its internal buffer and data might be lost.

10.4

DMA interrupts
For each DMA stream, an interrupt can be produced on the following events:
•

Half-transfer reached

•

Transfer complete

•

Transfer error

•

Fifo error (overrun, underrun or FIFO level error)

•

Direct mode error

Separate interrupt enable control bits are available for flexibility as shown in Table 50.
Table 50. DMA interrupt requests
Interrupt event

Event flag

Enable control bit

Half-transfer

HTIF

HTIE

Transfer complete

TCIF

TCIE

Transfer error

TEIF

TEIE

FIFO overrun/underrun

FEIF

FEIE

DMEIF

DMEIE

Direct mode error

Note:

324/1747

Before setting an Enable control bit to ‘1’, the corresponding event flag should be cleared,
otherwise an interrupt is immediately generated.

RM0090 Rev 17

RM0090

DMA controller (DMA)

10.5

DMA registers
The DMA registers have to be accessed by words (32 bits).

10.5.1

DMA low interrupt status register (DMA_LISR)
Address offset: 0x00
Reset value: 0x0000 0000

31

30

29

28

Reserved
r

r

r

r

15

14

13

12

Reserved
r

r

r

r

27

26

25

TCIF3

HTIF3

TEIF3

r

r

r

11

10

9

TCIF1

HTIF1

TEIF1

r

r

r

24

23

22

DMEIF3 Reserv FEIF3
ed
r
r
8

7

6

DMEIF1 Reserv FEIF1
ed
r
r

21

20

19

TCIF2

HTIF2

TEIF2

r

r

r

5

4

3

TCIF0

HTIF0

TEIF0

r

r

r

18

17

16

DMEIF2 Reserv
ed
r
2

1

FEIF2
r
0

DMEIF0 Reserv
ed
r

FEIF0
r

Bits 31:28, 15:12 Reserved, must be kept at reset value.
Bits 27, 21, 11, 5 TCIFx: Stream x transfer complete interrupt flag (x = 3..0)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_LIFCR register.
0: No transfer complete event on stream x
1: A transfer complete event occurred on stream x
Bits 26, 20, 10, 4 HTIFx: Stream x half transfer interrupt flag (x=3..0)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_LIFCR register.
0: No half transfer event on stream x
1: A half transfer event occurred on stream x
Bits 25, 19, 9, 3 TEIFx: Stream x transfer error interrupt flag (x=3..0)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_LIFCR register.
0: No transfer error on stream x
1: A transfer error occurred on stream x
Bits 24, 18, 8, 2 DMEIFx: Stream x direct mode error interrupt flag (x=3..0)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_LIFCR register.
0: No Direct Mode Error on stream x
1: A Direct Mode Error occurred on stream x
Bits 23, 17, 7, 1 Reserved, must be kept at reset value.
Bits 22, 16, 6, 0 FEIFx: Stream x FIFO error interrupt flag (x=3..0)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_LIFCR register.
0: No FIFO Error event on stream x
1: A FIFO Error event occurred on stream x

RM0090 Rev 17

325/1747
338

DMA controller (DMA)

10.5.2

RM0090

DMA high interrupt status register (DMA_HISR)
Address offset: 0x04
Reset value: 0x0000 0000

31

30

29

28

Reserved
15

14

13

Reserved

12

27

26

25

TCIF7

HTIF7

TEIF7

r

r

r

11

10

9

TCIF5

HTIF5

TEIF5

r

r

r

24

23

22

DMEIF7 Reserv FEIF7
ed
r
r
8

7

6

DMEIF5 Reserv FEIF5
ed
r
r

21

20

19

TCIF6

HTIF6

TEIF6

r

r

r

5

4

3

TCIF4

HTIF4

TEIF4

r

r

r

18

17

DMEIF6 Reserv
ed
r
2

1

DMEIF4 Reserv
ed
r

16
FEIF6
r
0
FEIF4
r

Bits 31:28, 15:12 Reserved, must be kept at reset value.
Bits 27, 21, 11, 5 TCIFx: Stream x transfer complete interrupt flag (x=7..4)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_HIFCR register.
0: No transfer complete event on stream x
1: A transfer complete event occurred on stream x
Bits 26, 20, 10, 4 HTIFx: Stream x half transfer interrupt flag (x=7..4)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_HIFCR register.
0: No half transfer event on stream x
1: A half transfer event occurred on stream x
Bits 25, 19, 9, 3 TEIFx: Stream x transfer error interrupt flag (x=7..4)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_HIFCR register.
0: No transfer error on stream x
1: A transfer error occurred on stream x
Bits 24, 18, 8, 2 DMEIFx: Stream x direct mode error interrupt flag (x=7..4)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_HIFCR register.
0: No Direct mode error on stream x
1: A Direct mode error occurred on stream x
Bits 23, 17, 7, 1 Reserved, must be kept at reset value.
Bits 22, 16, 6, 0 FEIFx: Stream x FIFO error interrupt flag (x=7..4)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_HIFCR register.
0: No FIFO error event on stream x
1: A FIFO error event occurred on stream x

326/1747

RM0090 Rev 17

RM0090

DMA controller (DMA)

10.5.3

DMA low interrupt flag clear register (DMA_LIFCR)
Address offset: 0x08
Reset value: 0x0000 0000

31

30

29

28

Reserved
15

14

13

27

26

25

24

CTCIF3 CHTIF3 CTEIF3 CDMEIF3

12

w

w

w

w

11

10

9

8

CTCIF1 CHTIF1 CTEIF1 CDMEIF1

Reserved

w

w

w

w

23
Reserved

22

21

20

CFEIF3

CTCIF2

CHTIF2

w

w

w

w

w

6

5

4

3

2

CFEIF1

CTCIF0

CHTIF0

w

w

w

7
Reserved

19

18

CTEIF2 CDMEIF2

CTEIF0 CDMEIF0
w

w

17

16

Reserved
1

CFEIF2
w
0

Reserved

CFEIF0
w

Bits 31:28, 15:12 Reserved, must be kept at reset value.
Bits 27, 21, 11, 5 CTCIFx: Stream x clear transfer complete interrupt flag (x = 3..0)
Writing 1 to this bit clears the corresponding TCIFx flag in the DMA_LISR register
Bits 26, 20, 10, 4 CHTIFx: Stream x clear half transfer interrupt flag (x = 3..0)
Writing 1 to this bit clears the corresponding HTIFx flag in the DMA_LISR register
Bits 25, 19, 9, 3 CTEIFx: Stream x clear transfer error interrupt flag (x = 3..0)
Writing 1 to this bit clears the corresponding TEIFx flag in the DMA_LISR register
Bits 24, 18, 8, 2 CDMEIFx: Stream x clear direct mode error interrupt flag (x = 3..0)
Writing 1 to this bit clears the corresponding DMEIFx flag in the DMA_LISR register
Bits 23, 17, 7, 1 Reserved, must be kept at reset value.
Bits 22, 16, 6, 0 CFEIFx: Stream x clear FIFO error interrupt flag (x = 3..0)
Writing 1 to this bit clears the corresponding CFEIFx flag in the DMA_LISR register

10.5.4

DMA high interrupt flag clear register (DMA_HIFCR)
Address offset: 0x0C
Reset value: 0x0000 0000

31

30

29

28

Reserved
15

14

13

Reserved

27

26

25

24

CTCIF7 CHTIF7 CTEIF7 CDMEIF7

12

w

w

w

w

11

10

9

8

CTCIF5 CHTIF5 CTEIF5 CDMEIF5
w

w

w

w

23
Reserved

22

21

20

CFEIF7

CTCIF6

CHTIF6

w

w

w

7
Reserved

6

5

4

CFEIF5

CTCIF4

CHTIF4

w

w

w

19

18

CTEIF6 CDMEIF6
w

w

3

2

CTEIF4 CDMEIF4
w

w

17

16

Reserved
1

CFEIF6
w
0

Reserved

CFEIF4
w

Bits 31:28, 15:12 Reserved, must be kept at reset value.
Bits 27, 21, 11, 5 CTCIFx: Stream x clear transfer complete interrupt flag (x = 7..4)
Writing 1 to this bit clears the corresponding TCIFx flag in the DMA_HISR register
Bits 26, 20, 10, 4 CHTIFx: Stream x clear half transfer interrupt flag (x = 7..4)
Writing 1 to this bit clears the corresponding HTIFx flag in the DMA_HISR register
Bits 25, 19, 9, 3 CTEIFx: Stream x clear transfer error interrupt flag (x = 7..4)
Writing 1 to this bit clears the corresponding TEIFx flag in the DMA_HISR register

RM0090 Rev 17

327/1747
338

DMA controller (DMA)

RM0090

Bits 24, 18, 8, 2 CDMEIFx: Stream x clear direct mode error interrupt flag (x = 7..4)
Writing 1 to this bit clears the corresponding DMEIFx flag in the DMA_HISR register
Bits 23, 17, 7, 1 Reserved, must be kept at reset value.
Bits 22, 16, 6, 0 CFEIFx: Stream x clear FIFO error interrupt flag (x = 7..4)
Writing 1 to this bit clears the corresponding CFEIFx flag in the DMA_HISR register

10.5.5

DMA stream x configuration register (DMA_SxCR) (x = 0..7)
This register is used to configure the concerned stream.
Address offset: 0x10 + 0x18 × stream number
Reset value: 0x0000 0000

31

30

29

28

PINCOS
rw

14

13

25

12

24

23

21

19

18

17

DBM

rw

rw

rw

Reserved

CT

rw

rw

rw

rw

11

10

9

8

7

6

5

4

3

2

1

0

MINC

PINC

CIRC

PFCTRL

TCIE

HTIE

TEIE

DMEIE

EN

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

DIR[1:0]
rw

rw

PL[1:0]

Bits 31:28 Reserved, must be kept at reset value.
Bits 27:25 CHSEL[2:0]: Channel selection
These bits are set and cleared by software.
000: channel 0 selected
001: channel 1 selected
010: channel 2 selected
011: channel 3 selected
100: channel 4 selected
101: channel 5 selected
110: channel 6 selected
111: channel 7 selected
These bits are protected and can be written only if EN is ‘0’
Bits 24:23 MBURST: Memory burst transfer configuration
These bits are set and cleared by software.
00: single transfer
01: INCR4 (incremental burst of 4 beats)
10: INCR8 (incremental burst of 8 beats)
11: INCR16 (incremental burst of 16 beats)
These bits are protected and can be written only if EN is ‘0’
In direct mode, these bits are forced to 0x0 by hardware as soon as bit EN= '1'.
Bits 22:21 PBURST[1:0]: Peripheral burst transfer configuration
These bits are set and cleared by software.
00: single transfer
01: INCR4 (incremental burst of 4 beats)
10: INCR8 (incremental burst of 8 beats)
11: INCR16 (incremental burst of 16 beats)
These bits are protected and can be written only if EN is ‘0’
In direct mode, these bits are forced to 0x0 by hardware.
Bit 20 Reserved, must be kept at reset value.

328/1747

16

rw

rw

PBURST[1:0]

20

rw

PSIZE[1:0]

MBURST [1:0]

22

rw

MSIZE[1:0]
rw

26
CHSEL[2:0]

Reserved
15

27

RM0090 Rev 17

rw

RM0090

DMA controller (DMA)

Bit 19 CT: Current target (only in double buffer mode)
This bits is set and cleared by hardware. It can also be written by software.
0: The current target memory is Memory 0 (addressed by the DMA_SxM0AR pointer)
1: The current target memory is Memory 1 (addressed by the DMA_SxM1AR pointer)
This bit can be written only if EN is ‘0’ to indicate the target memory area of the first transfer.
Once the stream is enabled, this bit operates as a status flag indicating which memory area
is the current target.
Bit 18 DBM: Double buffer mode
This bits is set and cleared by software.
0: No buffer switching at the end of transfer
1: Memory target switched at the end of the DMA transfer
This bit is protected and can be written only if EN is ‘0’.
Bits 17:16 PL[1:0]: Priority level
These bits are set and cleared by software.
00: Low
01: Medium
10: High
11: Very high
These bits are protected and can be written only if EN is ‘0’.
Bit 15 PINCOS: Peripheral increment offset size
This bit is set and cleared by software
0: The offset size for the peripheral address calculation is linked to the PSIZE
1: The offset size for the peripheral address calculation is fixed to 4 (32-bit alignment).
This bit has no meaning if bit PINC = '0'.
This bit is protected and can be written only if EN = '0'.
This bit is forced low by hardware when the stream is enabled (bit EN = '1') if the direct
mode is selected or if PBURST are different from “00”.
Bits 14:13 MSIZE[1:0]: Memory data size
These bits are set and cleared by software.
00: byte (8-bit)
01: half-word (16-bit)
10: word (32-bit)
11: reserved
These bits are protected and can be written only if EN is ‘0’.
In direct mode, MSIZE is forced by hardware to the same value as PSIZE as soon as bit EN
= '1'.
Bits 12:11 PSIZE[1:0]: Peripheral data size
These bits are set and cleared by software.
00: Byte (8-bit)
01: Half-word (16-bit)
10: Word (32-bit)
11: reserved
These bits are protected and can be written only if EN is ‘0’
Bit 10 MINC: Memory increment mode
This bit is set and cleared by software.
0: Memory address pointer is fixed
1: Memory address pointer is incremented after each data transfer (increment is done
according to MSIZE)
This bit is protected and can be written only if EN is ‘0’.

RM0090 Rev 17

329/1747
338

DMA controller (DMA)

RM0090

Bit 9 PINC: Peripheral increment mode
This bit is set and cleared by software.
0: Peripheral address pointer is fixed
1: Peripheral address pointer is incremented after each data transfer (increment is done
according to PSIZE)
This bit is protected and can be written only if EN is ‘0’.
Bit 8 CIRC: Circular mode
This bit is set and cleared by software and can be cleared by hardware.
0: Circular mode disabled
1: Circular mode enabled
When the peripheral is the flow controller (bit PFCTRL=1) and the stream is enabled (bit
EN=1), then this bit is automatically forced by hardware to 0.
It is automatically forced by hardware to 1 if the DBM bit is set, as soon as the stream is
enabled (bit EN ='1').
Bits 7:6 DIR[1:0]: Data transfer direction
These bits are set and cleared by software.
00: Peripheral-to-memory
01: Memory-to-peripheral
10: Memory-to-memory
11: reserved
These bits are protected and can be written only if EN is ‘0’.
Bit 5 PFCTRL: Peripheral flow controller
This bit is set and cleared by software.
0: The DMA is the flow controller
1: The peripheral is the flow controller
This bit is protected and can be written only if EN is ‘0’.
When the memory-to-memory mode is selected (bits DIR[1:0]=10), then this bit is
automatically forced to 0 by hardware.
Bit 4 TCIE: Transfer complete interrupt enable
This bit is set and cleared by software.
0: TC interrupt disabled
1: TC interrupt enabled
Bit 3 HTIE: Half transfer interrupt enable
This bit is set and cleared by software.
0: HT interrupt disabled
1: HT interrupt enabled
Bit 2 TEIE: Transfer error interrupt enable
This bit is set and cleared by software.
0: TE interrupt disabled
1: TE interrupt enabled
Bit 1 DMEIE: Direct mode error interrupt enable
This bit is set and cleared by software.
0: DME interrupt disabled
1: DME interrupt enabled

330/1747

RM0090 Rev 17

RM0090

DMA controller (DMA)

Bit 0 EN: Stream enable / flag stream ready when read low
This bit is set and cleared by software.
0: Stream disabled
1: Stream enabled
This bit may be cleared by hardware:
–
on a DMA end of transfer (stream ready to be configured)
–
if a transfer error occurs on the AHB master buses
–
when the FIFO threshold on memory AHB port is not compatible with the size of the
burst
When this bit is read as 0, the software is allowed to program the Configuration and FIFO
bits registers. It is forbidden to write these registers when the EN bit is read as 1.
Note: Before setting EN bit to '1' to start a new transfer, the event flags corresponding to the
stream in DMA_LISR or DMA_HISR register must be cleared.

10.5.6

DMA stream x number of data register (DMA_SxNDTR) (x = 0..7)
Address offset: 0x14 + 0x18 × stream number
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

Reserved
15

14

13

12

11

10

9

8

7

NDT[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 NDT[15:0]: Number of data items to transfer
Number of data items to be transferred (0 up to 65535). This register can be written only
when the stream is disabled. When the stream is enabled, this register is read-only,
indicating the remaining data items to be transmitted. This register decrements after each
DMA transfer.
Once the transfer has completed, this register can either stay at zero (when the stream is in
normal mode) or be reloaded automatically with the previously programmed value in the
following cases:
–
when the stream is configured in Circular mode.
–
when the stream is enabled again by setting EN bit to '1'
If the value of this register is zero, no transaction can be served even if the stream is
enabled.

RM0090 Rev 17

331/1747
338

DMA controller (DMA)

10.5.7

RM0090

DMA stream x peripheral address register (DMA_SxPAR) (x = 0..7)
Address offset: 0x18 + 0x18 × stream number
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

PAR[31:16]
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

PAR[15:0]
rw

Bits 31:0 PAR[31:0]: Peripheral address
Base address of the peripheral data register from/to which the data will be read/written.
These bits are write-protected and can be written only when bit EN = '0' in the DMA_SxCR register.

10.5.8

DMA stream x memory 0 address register (DMA_SxM0AR) (x = 0..7)
Address offset: 0x1C + 0x18 × stream number
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

23

22

21

20

19

18

17

16

rw

rw

rw

rw

rw

rw

rw

rw

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

M0A[31:16]

M0A[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 M0A[31:0]: Memory 0 address
Base address of Memory area 0 from/to which the data will be read/written.
These bits are write-protected. They can be written only if:
–
the stream is disabled (bit EN= '0' in the DMA_SxCR register) or
–
the stream is enabled (EN=’1’ in DMA_SxCR register) and bit CT = '1' in the
DMA_SxCR register (in Double buffer mode).

10.5.9

DMA stream x memory 1 address register (DMA_SxM1AR) (x = 0..7)
Address offset: 0x20 + 0x18 × stream number
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

23

22

21

20

19

18

17

16

rw

rw

rw

rw

rw

rw

rw

rw

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

M1A[31:16]

M1A[15:0]
rw

332/1747

rw

rw

rw

rw

rw

rw

rw

rw

RM0090 Rev 17

RM0090

DMA controller (DMA)

Bits 31:0 M1A[31:0]: Memory 1 address (used in case of Double buffer mode)
Base address of Memory area 1 from/to which the data will be read/written.
This register is used only for the Double buffer mode.
These bits are write-protected. They can be written only if:
–
the stream is disabled (bit EN= '0' in the DMA_SxCR register) or
–
the stream is enabled (EN=’1’ in DMA_SxCR register) and bit CT = '0' in the
DMA_SxCR register.

10.5.10

DMA stream x FIFO control register (DMA_SxFCR) (x = 0..7)
Address offset: 0x24 + 0x24 × stream number
Reset value: 0x0000 0021

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

7

6

5

4

3

2

1

0

FEIE

Reser
ved

r

Reserved
15

14

13

12

11

Reserved

10

9

8

rw

FS[2:0]
r

DMDIS
r

rw

FTH[1:0]
rw

rw

Bits 31:8 Reserved, must be kept at reset value.
Bit 7 FEIE: FIFO error interrupt enable
This bit is set and cleared by software.
0: FE interrupt disabled
1: FE interrupt enabled
Bit 6 Reserved, must be kept at reset value.

RM0090 Rev 17

333/1747
338

DMA controller (DMA)

RM0090

Bits 5:3 FS[2:0]: FIFO status
These bits are read-only.
000: 0 < fifo_level < 1/4
001: 1/4 ≤ fifo_level < 1/2
010: 1/2 ≤ fifo_level < 3/4
011: 3/4 ≤ fifo_level < full
100: FIFO is empty
101: FIFO is full
others: no meaning
These bits are not relevant in the direct mode (DMDIS bit is zero).
Bit 2 DMDIS: Direct mode disable
This bit is set and cleared by software. It can be set by hardware.
0: Direct mode enabled
1: Direct mode disabled
This bit is protected and can be written only if EN is ‘0’.
This bit is set by hardware if the memory-to-memory mode is selected (DIR bit in
DMA_SxCR are “10”) and the EN bit in the DMA_SxCR register is ‘1’ because the direct
mode is not allowed in the memory-to-memory configuration.
Bits 1:0 FTH[1:0]: FIFO threshold selection
These bits are set and cleared by software.
00: 1/4 full FIFO
01: 1/2 full FIFO
10: 3/4 full FIFO
11: full FIFO
These bits are not used in the direct mode when the DMIS value is zero.
These bits are protected and can be written only if EN is ‘0’.

334/1747

RM0090 Rev 17

RM0090

DMA controller (DMA)

10.5.11

DMA register map
Table 51 summarizes the DMA registers.

0

0

0

0

0

0

0

0

0

0

0

1

0
FEIF0
FEIF4
CFEIF0

2

Reserved
Reserved
Reserved Reserved

3

DMEIF0
DMEIF4

CFEIF4

0

0

0

0

EN

0

DMEIE

CDMEIF4

CDMEIF0

4

TEIF0
TEIF4
CTEIF4

CTEIF0

5

HTIF0
HTIF4
CHTIF0

6

TCIF0
TCIF4
CTCIF0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

M0A[31:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

CIRC

PSIZE[1:0]

MSIZE[1:0]

PINCOS

0

PL[1:0]

0

PINC

0

MINC

0

DBM

0

CT

0

Reserved

MBURST[1:]

CHSEL
[2:0]

Reserved

PBURST[1:0]

0

0

0

FS[2:0]

FTH
[1:0]

1

0

0

0

0

1

EN

0

DMEIE

0

TEIE

0

HTIE

0

TCIE

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DIR[1:0]

0

DMDIS

M1A[31:0]

DMA_S0FCR

Reset value

0

PFCTRL

0x002C

0

PA[31:0]

DMA_S0M1AR

Reset value

0

Reserved

0

DMA_S0M0AR

DMA_S1NDTR

0

FEIE

Reserved

DMA_S0PAR

DMA_S1CR

0

NDT[15:.]

Reset value

0x0028

0

0

TEIE

0

0

HTIE

0

0

0

CHTIF4

0

0

0

TCIE

0

0

CTCIF4

0

0

0

PFCTRL

0

CTEIF5

0

CDMEIF5

0

0

PINC

PL[1:0]
0

0

CIRC

0

CHTIF5

0

0
PSIZE[1:0]

0
MSIZE[1:0]

0

PINCOS

0

Reserved

7

0

FEIF1

CTEIF1

CDMEIF1

0

FEIF5

CHTIF1
0

0

CFEIF1

CTCIF1
0

0

CFEIF5

0

0

DIR[1:0]

8

TEIF5

DMEIF5

0

Reserved

HTIF5
0

0

Reserved

9
TEIF1

DMEIF1

TCIF5
0

0

Reserved Reserved

11

10
HTIF1

12

TCIF1

13

14

16

15

17

FEIF2
FEIF6

0

0

Reserved

MINC

0

0

CTCIF5

0

0

0

CFEIF6

0

CFEIF2

18

0

Reserved

0

Reserved

0

CDMEIF6

0

0

Reserved

DBM

0

Reserved Reserved

19

DMEIF2
DMEIF6
CDMEIF2

20

TEIF2
TEIF6
CTEIF2

21

HTIF2
HTIF6
CHTIF2

22

TCIF2
TCIF6
CTCIF2

23

FEIF3
FEIF7
CFEIF3

Reserved

0

0

0

CTEIF6

0

0

CT

0

0

0

Reserved

0

CHTIF6

0

0

PBURST[1:0]

Reserved

0

Reserved

0

0

CTCIF6

0

0

CFEIF7

0

0

Reserved

24

0

0

Reserved

0

0

Reserved

25

DMEIF3
DMEIF7

26

TEIF3

CDMEIF3

27

HTIF3

0

Reset value

Reset value

0x0024

0

DMA_S0NDTR

Reset value
0x0020

0

CHSEL[2:0]

DMA_S0CR

Reset value
0x001C

0

0

MBURST[1:0]

Reserved

Reset value

0x0018

0

CDMEIF7

DMA_HIFCR

Reset value

0x0014

0

CTEIF7

Reset value

0x0010

0

Reserved

0x0008

0x000C

TEIF7

DMA_LIFCR

0

TEIF3

Reset value

0

HTIF7

28

Reserved

0

CHTIF3

DMA_HISR

0

CHTIF7

0x0004

TCIF3

Reset value

TCIF7

29

Reserved

CTCIF3

DMA_LISR

CTCIF7

0x0000

Register

30

Offset

31

Table 51. DMA register map and reset values

NDT[15:.]

Reserved

0

RM0090 Rev 17

0

0

0

0

0

0

0

0

335/1747
338

DMA controller (DMA)

RM0090

11

10

9

8

7

6

5

4

3

2

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

FEIE

Reserved

15

16

17

18

19

20

21

22

23

24

25

26

27

12

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

M0A[31:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0x0064

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

0

FTH
[1:0]

1

0

0

0

0

1

0

0

0

0

0

0

0

DIR[1:0]

0

CIRC

PSIZE[1:0]

MSIZE[1:0]

PINCOS

0

PL[1:0]

0

PINC

0

MINC

0

DBM

0

CT

0

Reserved

Reserved

PBURST[1:0]

CHSEL[2:0]

MBURST[1:0]

0

FS[2:0]

EN

0

DMEIE

0

DMDIS

0

TEIE

0

HTIE

0

NDT[15:.]

Reserved

0

DMA_S3PAR

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

PA[31:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMA_S3M0AR

336/1747

0

TCIE

0

Reset value

Reset value

0

PFCTRL

0

DMA_S3NDTR

Reset value

0

M1A[31:0]

DMA_S3CR

0x0060

1

M0A[31:0]

Reset value
0x005C

0

0

Reset value

0x0058

0

0

0

DIR
[1:0]

0

CIRC

PSIZE[1:0]

MSIZE[1:0]
0

0

PA[31:0]

DMA_S2FCR

0x0054

0

0

Reserved

Reserved

DMA_S2M1AR
Reset value

0

1

NDT[15:.]

DMA_S2M0AR
Reset value

PINCOS

0

PL[1:0]

0

PINC

0

MINC

0

DMA_S2PAR
Reset value

0x0050

0

Reset value

0x0048

0x004C

0

FTH
[1:0]

FEIE

0x0044

0

DBM

Reset value
DMA_S2NDTR

Reserved

Reserved

PBURST[1:0]

CHSEL
[2:0]

DMA_S2CR

0x0040

CT

0
MBURST[1:0]

Reset value

FS[2:0]

EN

0

DMEIE

0

TEIE

0

HTIE

0

TCIE

0

PFCTRL

0

DMDIS

M1A[31:0]

DMA_S1FCR

0x003C

13

0

DMA_S1M1AR
Reset value

0

PA[31:0]

DMA_S1M0AR
Reset value

14

Reset value

0x0038

28

DMA_S1PAR

0x0030

0x0034

29

Register

30

Offset

31

Table 51. DMA register map and reset values (continued)

0

0

0

M0A[31:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

RM0090 Rev 17

0

0

0

RM0090

DMA controller (DMA)

9

8

7

6

5

4

3

2

1

0

0

11

0

10

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

12

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMA_S3FCR

0

0

0

0

0

0

0

Reserved

0

0

0

0

0

CIRC

PSIZE[1:0]

MSIZE[1:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Reserved

0

0

0

0

0

0

0

0

0

0

0

CIRC

PSIZE[1:0]
0

0

0

DMA_S5PAR

0x00A8

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMA_S5M0AR

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

M0A[31:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMA_S5FCR

0

0

0

0

0

0

0

Reserved

0

0

0

0

0

0

0

0

0

0

CIRC

PSIZE[1:0]
0

0

0

FTH
[1:0]

1

0

0

0

0

1

0

0

0

0

0

0

0

NDT[15:.]

Reserved

Reset value

0

MSIZE[1:0]

PINCOS

0

PL[1:0]

0

PINC

0

MINC

0

DBM

0

CT

0

DMA_S6NDTR

Reserved

Reserved

PBURST[1:0]

CHSEL[2:0]

DMA_S6CR

MBURST[1:0]

0

FS[2:0]

EN

0

DMEIE

0

DMDIS

M1A[31:0]

TEIE

DMA_S5M1AR

0

DMA_S6PAR
Reset value

1

PA[31:0]

Reset value
0x00A4

0

FTH
[1:0]

NDT[15:.]

Reserved

Reset value

MSIZE[1:0]

PINCOS

0

PL[1:0]

0

PINC

0

MINC

0

DBM

0

CT

0

DMA_S5NDTR

Reserved

Reserved

PBURST[1:0]

CHSEL[2:0]

MBURST[1:0]

0

FS[2:0]

EN

0

DMEIE

0

DMDIS

0

TEIE

0

HTIE

0

Reset value

0x00A0

0

HTIE

0x009C

0

TCIE

0

DMA_S5CR

Reset value

0

TCIE

0x0098

0

M1A[31:0]

DMA_S4FCR

Reset value

0

Reserved

0x0094

0

PFCTRL

DMA_S4M1AR

Reset value

0

FEIE

0x0090

1

M0A[31:0]

Reset value
0x008C

0

0

Reset value

0x0088

0

0

PA[31:0]

DMA_S4M0AR

Reset value

0

0

NDT[15:.]

DMA_S4PAR

Reset value

0

PINC

0

Reserved

Reset value
Reset value

PINCOS

0

PL[1:0]

0

MINC

0

DBM

0

0

PFCTRL

0x0084

0

1

Reserved

0x0080

0

DIR[1:0]

0x007C

0

DIR[1:0]

0x0078

0

FTH
[1:0]

FEIE

0x0074

0

CT

Reset value
DMA_S4NDTR

Reserved

Reserved

PBURST[1:0]

DMA_S4CR

CHSEL[2:0]

0x0070

0
MBURST[1:0]

Reset value

FS[2:0]

EN

0

DMEIE

0

DMDIS

0

TEIE

0

HTIE

0

TCIE

0

PFCTRL

0

Reserved

0

FEIE

M1A[31:0]

DIR
[1:0]

Reset value

0x006C

29

DMA_S3M1AR

13

0x0068

Register

30

Offset

31

Table 51. DMA register map and reset values (continued)

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

PA[31:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

RM0090 Rev 17

0

0

337/1747
338

DMA controller (DMA)

RM0090

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Reserved

0

0

0

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMA_S7FCR

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

M1A[31:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Reserved

0

0

0

Reset value

0

Refer to Section 2.3: Memory map for the register boundary addresses.

338/1747

1

0

DIR[1:0]

0

CIRC

MSIZE[1:0]

PSIZE[1:0]
0

0

RM0090 Rev 17

FS[2:0]
1

0

0

DMDIS

0x00CC

0

0

M0A[31:0]

DMA_S7M1AR
Reset value

0

0

PA[31:0]

DMA_S7M0AR
Reset value

0

0

NDT[15:.]

DMA_S7PAR
Reset value

PINCOS

0

PL[1:0]

0

PINC

0

MINC

0

1

Reserved

0x00C8

0

Reset value

0x00C0
0x00C4

0

FTH
[1:0]

FEIE

0x00BC

0

DBM

Reset value
DMA_S7NDTR

Reserved

Reserved

PBURST[1:0]

CHSEL[2:0]

DMA_S7CR

0x00B8

CT

0
MBURST[1:0]

Reset value

FS[2:0]

EN

0

DMEIE

0

TEIE

0

HTIE

0

TCIE

0

PFCTRL

0

DMDIS

M1A[31:0]

DMA_S6FCR

0x00B4

0

M0A[31:0]

DMA_S6M1AR
Reset value

2

0

3

0

4

0

5

0

6

17

0

7

18

0

Reserved

19

0

8

20

0

FEIE

21

0

9

22

0

11

23

0

10

24

0

12

25

0

14

26

0

15

27

Reset value

16

28

0

DMA_S6M0AR

13

0x00B0

29

0x00AC

Register

30

Offset

31

Table 51. DMA register map and reset values (continued)

0

FTH
[1:0]
0

1

RM0090

Chrom-Art Accelerator™ controller (DMA2D)

11

Chrom-Art Accelerator™ controller (DMA2D)

11.1

DMA2D introduction
The Chrom-Art Accelerator™ (DMA2D) is a specialized DMA dedicated to image
manipulation. It can perform the following operations:
•

Filling a part or the whole of a destination image with a specific color

•

Copying a part or the whole of a source image into a part or the whole of a destination
image

•

Copying a part or the whole of a source image into a part or the whole of a destination
image with a pixel format conversion

•

Blending a part and/or two complete source images with different pixel format and copy
the result into a part or the whole of a destination image with a different color format.

All the classical color coding schemes are supported from 4-bit up to 32-bit per pixel with
indexed or direct color mode. The DMA2D has its own dedicated memories for CLUTs (color
look-up tables).

RM0090 Rev 17

339/1747
370

Chrom-Art Accelerator™ controller (DMA2D)

11.2

RM0090

DMA2D main features
The main DMA2D features are:
•

Single AHB master bus architecture.

•

AHB slave programming interface supporting 8/16/32-bit accesses (except for CLUT
accesses which are 32-bit).

•

User programmable working area size

•

User programmable offset for sources and destination areas

•

User programmable sources and destination addresses on the whole memory space

•

Up to 2 sources with blending operation

•

Alpha value can be modified (source value, fixed value or modulated value)

•

User programmable source and destination color format

•

Up to 11 color formats supported from 4-bit up to 32-bit per pixel with indirect or direct
color coding

•

2 internal memories for CLUT storage in indirect color mode

•

Automatic CLUT loading or CLUT programming via the CPU

•

User programmable CLUT size

•

Internal timer to control AHB bandwidth

•

4 operating modes: register-to-memory, memory-to-memory, memory-to-memory with
pixel format conversion, and memory-to-memory with pixel format conversion and
blending

•

Area filling with a fixed color

•

Copy from an area to another

•

Copy with pixel format conversion between source and destination images

•

Copy from two sources with independent color format and blending

•

Abort and suspend of DMA2D operations

•

Watermark interrupt on a user programmable destination line

•

Interrupt generation on bus error or access conflict

•

Interrupt generation on process completion

11.3

DMA2D functional description

11.3.1

General description
The DMA2D controller performs direct memory transfer. As an AHB master, it can take the
control of the AHB bus matrix to initiate AHB transactions.
The DMA2D can operate in the following modes:
•

Register-to-memory

•

Memory-to-memory

•

Memory-to-memory with Pixel Format Conversion

•

Memory-to-memory with Pixel Format Conversion and Blending

The AHB slave port is used to program the DMA2D controller.
The block diagram of the DMA2D is shown in Figure 40: DMA2D block diagram.

340/1747

RM0090 Rev 17

RM0090

Chrom-Art Accelerator™ controller (DMA2D)
Figure 40. DMA2D block diagram
!(" -!34%2

&' 0&#

A MODE

#OLOR MODE
&'
&)&/

 BIT A


A 



2'"

%XPANDER

8

A 



#,54 ITF

X BIT
2!-



#OLOR MODE

A
2ED
'REEN


#ONVERTER


#OLOR
/54
&)&/

"LUE

"' 0&#

A MODE

#OLOR MODE
"'
&)&/

/54 0&#

",%.$%2

 BIT A



A 



2'"

%XPANDER

8

A 

#,54 ITF

X BIT
2!-

!(" 3,!6%

-36

11.3.2

DMA2D control
The DMA2D controller is configured through the DMA2D Control Register (DMA2D_CR)
which allows selecting:
The user application can perform the following operations:

11.3.3

•

Select the operating mode

•

Enable/disable the DMA2D interrupt

•

Start/suspend/abort ongoing data transfers

DMA2D foreground and background FIFOs
The DMA2D foreground (FG) FG FIFO and background (BG) FIFO fetch the input data to
be copied and/or processed.
The FIFOs fetch the pixels according to the color format defined in their respective pixel
format converter (PFC).

RM0090 Rev 17

341/1747
370

Chrom-Art Accelerator™ controller (DMA2D)

RM0090

They are programmed through a set of control registers:
•

DMA2D foreground memory address register (DMA2D_FGMAR)

•

DMA2D foreground offset register (DMA2D_FGOR)

•

DMA2D background memory address register (DMA2D_BGMAR)

•

DMA2D background offset register (DMA2D_BGBOR)

•

DMA2D number of lines register (number of lines and pixel per lines) (DMA2D_NLR)

When the DMA2D operates in register-to-memory mode, none of the FIFOs is activated.
When the DMA2D operates in memory-to-memory mode (no pixel format conversion nor
blending operation), only the FG FIFO is activated and acts as a buffer.
When the DMA2D operates in memory-to-memory operation with pixel format conversion
(no blending operation), the BG FIFO is not activated.

11.3.4

DMA2D foreground and background pixel format converter (PFC)
DMA2D foreground pixel format converter (PFC) and background pixel format converter
perform the pixel format conversion to generate a 32-bit per pixel value. The PFC can also
modify the alpha channel.
The first stage of the converter converts the color format. The original color format of the
foreground pixel and background pixels are configured through the CM[3:0] bits of the
DMA2D_FGPFCCR and DMA2D_BGPFCCR, respectively.
The supported input formats are given in Table 52: Supported color mode in input.
Table 52. Supported color mode in input
CM[3:0]

342/1747

Color mode

0000

ARGB8888

0001

RGB888

0010

RGB565

0011

ARGB1555

0100

ARGB4444

0101

L8

0110

AL44

0111

AL88

1000

L4

1001

A8

1010

A4

RM0090 Rev 17

RM0090

Chrom-Art Accelerator™ controller (DMA2D)
The color format are coded as follows:
•

Alpha value field: transparency
0xFF value corresponds to an opaque pixel and 0x00 to a transparent one.

•

R field for Red

•

G field for Green

•

B field for Blue

•

L field: luminance
This field is the index to a CLUT to retrieve the three/four RGB/ARGB components.

If the original format was direct color mode, then the extension to 8-bit per channel is
performed by copying the MSBs into the LSBs. This ensures a perfect linearity of the
conversion.
If the original format does not include an alpha channel, the alpha value is automatically set
to 0xFF (opaque).
If the original format is indirect color mode, a CLUT is required and each pixel format
converter is associated with a 256 entry 32-bit CLUT.
For the specific alpha mode A4 and A8, no color information is stored nor indexed. The color
to be used for the image generation is fixed and is defined in the DMA2D_FGCOLR for
foreground pixels and in the DMA2D_BGCOLR register for background pixels.
The order of the fields in the system memory is defined in Table 53: Data order in memory.
Table 53. Data order in memory
Color Mode

@+3

@+2

@+1

@+0

ARGB8888

A0[7:0]

R0[7:0]

G0[7:0]

B0[7:0]

B1[7:0]

R0[7:0]

G0[7:0]

B0[7:0]

G2[7:0]

B2[7:0]

R1[7:0]

G1[7:0]

R3[7:0]

G3[7:0]

B3[7:0]

R2[7:0]

RGB565

R1[4:0]G1[5:3]

G1[2:0]B1[4:0]

R0[4:0]G0[5:3]

G0[2:0]B0[4:0]

ARGB1555

A1[0]R1[4:0]G1[4:3]

G1[2:0]B1[4:0]

A0[0]R0[4:0]G0[4:3]

G0[2:0]B0[4:0]

ARGB4444

A1[3:0]R1[3:0]

G1[3:0]B1[3:0]

A0[3:0]R0[3:0]

G0[3:0]B0[3:0]

L8

L3[7:0]

L2[7:0]

L1[7:0]

L0[7:0]

AL44

A3[3:0]L3[3:0]

A2[3:0]L2[3:0]

A1[3:0]L1[3:0]

A0[3:0]L0[3:0]

AL88

A1[7:0]

L1[7:0]

A0[7:0]

L0[7:0]

L4

L7[3:0]L6[3:0]

L5[3:0]L4[3:0]

L3[3:0]L2[3:0]

L1[3:0]L0[3:0]

A8

A3[7:0]

A2[7:0]

A1[7:0]

A0[7:0]

A4

A7[3:0]A6[3:0]

A5[3:0]A4[3:0]

A3[3:0]A2[3:0]

A1[3:0]A0[3:0]

RGB888

The 24-bit RGB888 aligned on 32-bit is supported through the ARGB8888 mode.
Once the 32-bit value is generated, the alpha channel can be modified according to the
AM[1:0] field of the DMA2D_FGPFCCR/DMA2D_BGPFCCR registers as shown in
Table 54: Alpha mode configuration.

RM0090 Rev 17

343/1747
370

Chrom-Art Accelerator™ controller (DMA2D)

RM0090

The alpha channel can be:
•

kept as it is (no modification),

•

replaced by the ALPHA[7:0] value of DMA2D_FGPFCCR/DMA2D_BGPFCCR,

•

or replaced by the original alpha value multiplied by the ALPHA[7:0] value of
DMA2D_FGPFCCR/DMA2D_BGPFCCR divided by 255.
Table 54. Alpha mode configuration
AM[1:0]

11.3.5

Alpha mode

00

No modification

01

Replaced by value in DMA2D_xxPFCCR

10

Replaced by original value multiplied by the value in DMA2D_xxPFCCR / 255

11

Reserved

DMA2D foreground and background CLUT interface
The CLUT interface manages the CLUT memory access and the automatic loading of the
CLUT.
Three kinds of accesses are possible:
•

CLUT read by the PFC during pixel format conversion operation

•

CLUT accessed through the AHB slave port when the CPU is reading or writing data
into the CLUT

•

CLUT written through the AHB master port when an automatic loading of the CLUT is
performed

The CLUT memory loading can be done in two different ways:
•

Automatic loading
The following sequence should be followed to load the CLUT:
a)

Program the CLUT address into the DMA2D_FGCMAR register (foreground
CLUT) or DMA2D_BGCMAR register (background CLUT)

b)

Program the CLUT size in the CS[7:0] field of the DMA2D_FGPFCCR register
(foreground CLUT) or DMA2D_BGPFCCR register (background CLUT).

c)

Set the START bit of the DMA2D_FGPFCCR register (foreground CLUT) or
DMA2D_BGPFCCR register (background CLUT) to start the transfer. During this
automatic loading process, the CLUT is not accessible by the CPU. If a conflict
occurs, a CLUT access error interrupt is raised assuming CAEIE is set to ‘1’ in
DMA2D_CR.

•

Manual loading
The application has to program the CLUT manually through the DMA2D AHB slave
port to which the local CLUT memory is mapped.The foreground CLUT is located at
address offset 0x0400 and the background CLUT at address offset 0x0800.

The CLUT format can be 24 or 32 bits. It is configured through the CCM bit of the
DMA2D_FGPFCCR register (foreground CLUT) or DMA2D_BGPFCCR register
(background CLUT) as shown in Table 55: Supported CLUT color mode.

344/1747

RM0090 Rev 17

RM0090

Chrom-Art Accelerator™ controller (DMA2D)
Table 55. Supported CLUT color mode
CCM

CLUT color mode

0

32-bit ARGB8888

1

24-bit RGB888

The way the CLUT data are organized in the system memory is specified in Table 56: CLUT
data order in memory.
Table 56. CLUT data order in memory
CLUT Color Mode
ARGB8888

RGB888

11.3.6

@+3

@+2

@+1

@+0

A0[7:0]

R0[7:0]

G0[7:0]

B0[7:0]

B1[7:0]

R0[7:0]

G0[7:0]

B0[7:0]

G2[7:0]

B2[7:0]

R1[7:0]

G1[7:0]

R3[7:0]

G3[7:0]

B3[7:0]

R2[7:0]

DMA2D blender
The DMA2D blender blends the source pixels by pair to compute the resulting pixel.
The blending is performed according to the following equation:

with αMult =

αFG . αBG
255

αOUT = αFG + αBG - αMult

COUT =

CFG.αFG + CBG.αBG - CBG.αMult
αOUT

with C = R or G or B

Division is rounded to the nearest lower integer

No configuration register is required by the blender. The blender usage depends on the
DMA2D operating mode defined in MODE[1:0] field of the DMA2D_CR register.

11.3.7

DMA2D output PFC
The output PFC performs the pixel format conversion from 32 bits to the output format
defined in the CM[2:0] field of the DMA2D output pixel format converter configuration
register (DMA2D_OPFCCR).
The supported output formats are given in Table 57: Supported color mode in output

RM0090 Rev 17

345/1747
370

Chrom-Art Accelerator™ controller (DMA2D)

RM0090

Table 57. Supported color mode in output
CM[2:0]

11.3.8

Color mode

000

ARGB8888

001

RGB888

010

RGB565

011

ARGB1555

100

ARGB4444

DMA2D output FIFO
The output FIFO programs the pixels according to the color format defined in the output
PFC.
The destination area is defined through a set of control registers:
•

DMA2D output memory address register (DMA2D_OMAR)

•

DMA2D output offset register (DMA2D_OOR)

•

DMA2D number of lines register (number of lines and pixel per lines) (DMA2D_NLR)

If the DMA2D operates in register-to-memory mode, the configured output rectangle is filled
by the color specified in the DMA2D output color register (DMA2D_OCOLR) which contains
a fixed 32-bit, 24-bit or 16-bit value. The format is selected by the CM[2:0] field of the
DMA2D_OPFCCR register.
The data are stored into the memory in the order defined in Table 58: Data order in memory
Table 58. Data order in memory
Color Mode

@+3

@+2

@+1

@+0

ARGB8888

A0[7:0]

R0[7:0]

G0[7:0]

B0[7:0]

B1[7:0]

R0[7:0]

G0[7:0]

B0[7:0]

G2[7:0]

B2[7:0]

R1[7:0]

G1[7:0]

R3[7:0]

G3[7:0]

B3[7:0]

R2[7:0]

RGB565

R1[4:0]G1[5:3]

G1[2:0]B1[4:0]

R0[4:0]G0[5:3]

G0[2:0]B0[4:0]

ARGB1555

A1[0]R1[4:0]G1[4:3]

G1[2:0]B1[4:0]

A0[0]R0[4:0]G0[4:3]

G0[2:0]B0[4:0]

ARGB4444

A1[3:0]R1[3:0]

G1[3:0]B1[3:0]

A0[3:0]R0[3:0]

G0[3:0]B0[3:0]

RGB888

The RGB888 aligned on 32-bit is supported through the ARGB8888 mode.

11.3.9

DMA2D AHB master port timer
An 8-bit timer is embedded into the AHB master port to provide an optional limitation of the
bandwidth on the crossbar.
This timer is clocked by the AHB clock and counts a dead time between two consecutive
accesses. This limits the bandwidth usage.

346/1747

RM0090 Rev 17

RM0090

Chrom-Art Accelerator™ controller (DMA2D)
The timer enabling and the dead time value are configured through the AHB master port
timer configuration register (DMA2D_AMPTCR).

11.3.10

DMA2D transactions
DMA2D transactions consist of a sequence of a given number of data transfers. The
number of data and the width can be programmed by software.
Each DMA2D data transfer is composed of up to 4 steps:

11.3.11

1.

Data loading from the memory location pointed by the DMA2D_FGMAR register and
pixel format conversion as defined in DMA2D_FGCR.

2.

Data loading from a memory location pointed by the DMA2D_BGMAR register and
pixel format conversion as defined in DMA2D_BGCR.

3.

Blending of all retrieved pixels according to the alpha channels resulting of the PFC
operation on alpha values.

4.

Pixel format conversion of the resulting pixels according to the DMA2D_OCR register
and programming of the data to the memory location addressed through the
DMA2D_OMAR register.

DMA2D configuration
Both source and destination data transfers can target peripherals and memories in the
whole 4 Gbyte memory area, at addresses ranging between 0x0000 0000 and
0xFFFF FFFF.
The DMA2D can operate in any of the four following modes selected through MODE[1:0]
bits of the DMA2D_CR register:
•

Register-to-memory

•

Memory-to-memory

•

Memory-to-memory with PFC

•

Memory-to-memory with PFC and blending

Register-to-memory
The register-to-memory mode is used to fill a user defined area with a predefined color.
The color format is set in the DMA2D_OPFCCR.
The DMA2D does not perform any data fetching from any source. It just writes the color
defined in the DMA2D_OCOLR register to the area located at the address pointed by the
DMA2D_OMAR and defined in the DMA2D_NLR and DMA2D_OOR.

Memory-to-memory
In memory-to-memory mode, the DMA2D does not perform any graphical data
transformation. The foreground input FIFO acts as a buffer and the data are transferred
from the source memory location defined in DMA2D_FGMAR to the destination memory
location pointed by DMA2D_OMAR.
The color mode programmed in the CM[3:0] bits of the DMA2D_FGPFCCR register defines
the number of bits per pixel for both input and output.

RM0090 Rev 17

347/1747
370

Chrom-Art Accelerator™ controller (DMA2D)

RM0090

The size of the area to be transferred is defined by the DMA2D_NLR and DMA2D_FGOR
registers for the source, and by DMA2D_NLR and DMA2D_OOR registers for the
destination.

Memory-to-memory with PFC
In this mode, the DMA2D performs a pixel format conversion of the source data and stores
them in the destination memory location.
The size of the areas to be transferred are defined by the DMA2D_NLR and DMA2D_FGOR
registers for the source, and by DMA2D_NLR and DMA2D_OOR registers for the
destination.
Data are fetched from the location defined in the DMA2D_FGMAR register and processed
by the foreground PFC. The original pixel format is configured through the
DMA2D_FGPFCCR register.
If the original pixel format is direct color mode, then the color channels are all expanded to 8
bits.
If the pixel format is indirect color mode, the associated CLUT has to be loaded into the
CLUT memory.
The CLUT loading can be done automatically by following the sequence below:
1.

Set the CLUT address into the DMA2D_FGCMAR.

2.

Set the CLUT size in the CS[7:0] bits of the DMA2D_FGPFCCR register.

3.

Set the CLUT format (24 or 32 bits) in the CCM bit of the DMA2D_FGPFCCR register.

4.

Start the CLUT loading by setting the START bit of the DMA2D_FGPFCCR register.

Once the CLUT loading is complete, the CTCIF flag of the DMA2D_IFR register is raised,
and an interrupt is generated if the CTCIE bit is set in DMA2D_CR. The automatic CLUT
loading process can not work in parallel with classical DMA2D transfers.
The CLUT can also be filled by the CPU or by any other master through the APB port. The
access to the CLUT is not possible when a DMA2D transfer is ongoing and uses the CLUT
(indirect color format).
In parallel to the color conversion process, the alpha value can be added or changed
depending on the value programmed in the DMA2D_FGPFCCR register. If the original
image does not have an alpha channel, a default alpha value of 0xFF is automatically added
to obtain a fully opaque pixel. The alpha value can be modified according to the AM[1:0] bits
of the DMA2D_FGPFCCR register:
•

It can be unchanged.

•

It can be replaced by the value defined in the ALPHA[7:0] value of the
DMA2D_FGPFCCR register.

•

It can be replaced by the original value multiplied by the ALPHA[7:0] value of the
DMA2D_FGPFCCR register divided by 255.

The resulting 32-bit data are encoded by the OUT PFC into the format specified by the
CM[2:0] field of the DMA2D_OPFCCR register. The output pixel format cannot be the
indirect mode since no CLUT generation process is supported.
The processed data are written into the destination memory location pointed by
DMA2D_OMAR.

348/1747

RM0090 Rev 17

RM0090

Chrom-Art Accelerator™ controller (DMA2D)

Memory-to-memory with PFC and blending
In this mode, 2 sources are fetched in the foreground FIFO and background FIFO from the
memory locations defined by DMA2D_FGMAR and DMA2D_BGMAR.
The two pixel format converters have to be configured as described in the memory-tomemory mode. Their configurations can be different as each pixel format converter are
independent and have their own CLUT memory.
Once each pixel has been converted into 32 bits by their respective PFCs, they are blended
according to the equation below:
with αMult =

αFG . αBG
255

αOUT = αFG + αBG - αMult

COUT =

CFG.αFG + CBG.αBG - CBG.αMult
αOUT

with C = R or G or B

Division are rounded to the nearest lower integer

The resulting 32-bit pixel value is encoded by the output PFC according to the specified
output format, and the data are written into the destination memory location pointed by
DMA2D_OMAR.

Configuration error detection
The DMA2D checks that the configuration is correct before any transfer. The configuration
error interrupt flag is set by hardware when a wrong configuration is detected when a new
transfer/automatic loading starts. An interrupt is then generated if the CEIE bit of the
DMA2D_CR is set.

RM0090 Rev 17

349/1747
370

Chrom-Art Accelerator™ controller (DMA2D)

RM0090

The wrong configurations that can be detected are listed below:

11.3.12

•

Foreground CLUT automatic loading: MA bits of DMA2D_FGCMAR not aligned with
CCM of DMA2D_FGPFCCR.

•

Background CLUT automatic loading: MA of DMA2D_BGCMAR not aligned with CCM
of DMA2D_BGPFCCR

•

Memory transfer (except in register-to-memory mode): MA of DMA2D_FGMAR not
aligned with CM of DMA2D_FGPFCCR

•

Memory transfer (except in register-to-memory mode): CM in DMA2D_FGPFCCR are
invalid

•

Memory transfer (except in register-to-memory mode): PL bits of DMA2D_NLR odd
while CM of DMA2D_FGPFCCR is A4 or L4

•

Memory transfer (except in register-to-memory mode): LO bits in DMA2D_FGOR odd
while CM of DMA2D_FGPFCCR is A4 or L4

•

Memory transfer (only in blending mode): MA bits in DMA2D_BGMAR are not aligned
with the CM of DMA2D_BGPFCCR

•

Memory transfer: CM of DMA2D_BGPFCCR invalid (only in blending mode)

•

Memory transfer (only in blending mode): PL bits of DMA2D_NLR odd while CM of
DMA2D_BGPFCCR is A4 or L4

•

Memory transfer (only in blending mode): LO bits of DMA2D_BGOR odd while CM of
DMA2D_BGPFCCR is A4 or L4

•

Memory transfer (except in memory to memory mode): MA bits in DMA2D_OMAR are
not aligned with CM bits in DMA2D_OPFCCR.

•

Memory transfer (except in memory to memory mode): CM bits in DMA2D_OPFCCR
invalid

•

Memory transfer: NL bits in DMA2D_NLR = 0

•

Memory transfer: PL bits in DMA2D_NLR = 0

DMA2D transfer control (start, suspend, abort and completion)
Once the DMA2D is configured, the transfer can be launched by setting the START bit of the
DMA2D_CR register. Once the transfer is completed, the START bit is automatically reset
and the TCIF flag of the DMA2D_ISR register is raised. An interrupt can be generated if the
TCIE bit of the DMA2D_CR is set.
The user application can suspend the DMA2D at any time by setting the SUSP bit of the
DMA2D_CR register. The transaction can then be aborted by setting the ABORT bit of the
DMA2D_CR register or can be restarted by resetting the SUSP bit of the DMA2D_CR
register.
The user application can abort at any time an ongoing transaction by setting the ABORT bit
of the DMA2D_CR register. In this case, the TCIF flag is not raised.
Automatic CLUT transfers can also be aborted or suspended by using the ABORT or the
SUSP bit of the DMA2D_CR register.

350/1747

RM0090 Rev 17

RM0090

11.3.13

Chrom-Art Accelerator™ controller (DMA2D)

Watermark
A watermark can be programmed to generate an interrupt when the last pixel of a given line
has been written to the destination memory area.
The line number is defined in the LW[15:0] field of the DMA2D_LWR register.
When the last pixel of this line has been transferred, the TWIF flag of the DMA2D_ISR
register is raised and an interrupt is generated if the TWIE bit of the DMA2D_CR is set.

11.3.14

Error management
Two kind of errors can be triggered:
•

AHB master port errors signalled by the TEIF flag of the DMA2D_ISR register.

•

Conflicts caused by CLUT access (CPU trying to access the CLUT while a CLUT
loading or a DMA2D transfer is ongoing) signalled by the CAEIF flag of the
DMA2D_ISR register.

Both flags are associated to their own interrupt enable flag in the DMA2D_CR register to
generate an interrupt if need be (TEIE and CAEIE).

11.3.15

AHB dead time
To limit the AHB bandwidth usage, a dead time between two consecutive AHB accesses
can be programmed.
This feature can be enabled by setting the EN bit in the DMA2D_AMTCR register.
The dead time value is stored in the DT[7:0] field of the DMA2D_AMTCR register. This
value represents the guaranteed minimum number of cycles between two consecutive
transactions on the AHB bus.
The update of the dead time value while the DMA2D is running will be taken into account for
the next AHB transfer.

11.4

DMA2D interrupts
An interrupt can be generated on the following events:
•

Configuration error

•

CLUT transfer complete

•

CLUT access error

•

Transfer watermark reached

•

Transfer complete

•

Transfer error

Separate interrupt enable bits are available for flexibility.
Table 59. DMA2D interrupt requests
Interrupt event
Configuration error
CLUT transfer complete

Event flag

Enable control bit

CEIF

CEIE

CTCIF

CTCIE

RM0090 Rev 17

351/1747
370

Chrom-Art Accelerator™ controller (DMA2D)

RM0090

Table 59. DMA2D interrupt requests (continued)
Interrupt event

Event flag

Enable control bit

CLUT access error

CAEIF

CAEIE

Transfer watermark

TWF

TWIE

Transfer complete

TCIF

TCIE

Transfer error

TEIF

TEIE

11.5

DMA2D registers

11.5.1

DMA2D control register (DMA2D_CR)
Address offset: 0x0000
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17
MODE

Reserved
15

14

Reserved

13
CEIE
rw

12

11

CTCIE CAEIE
rw

rw

10

9

8

TWIE

TCIE

TEIE

rw

rw

rw

16

7

6

5
Reserved

4

3

2

1

0

ABORT

SUSP

START

rs

rw

rs

Bits 31:18 Reserved, must be kept at reset value
Bits 17:16 MODE: DMA2D mode
This bit is set and cleared by software. It cannot be modified while a transfer is ongoing.
00: Memory-to-memory (FG fetch only)
01: Memory-to-memory with PFC (FG fetch only with FG PFC active)
10: Memory-to-memory with blending (FG and BG fetch with PFC and blending)
11: Register-to-memory (no FG nor BG, only output stage active)
Bits 15:14 Reserved, must be kept at reset value
Bit 13 CEIE: Configuration Error Interrupt Enable
This bit is set and cleared by software.
0: CE interrupt disable
1: CE interrupt enable
Bit 12 CTCIE: CLUT transfer complete interrupt enable
This bit is set and cleared by software.
0: CTC interrupt disable
1: CTC interrupt enable
Bit 11 CAEIE: CLUT access error interrupt enable
This bit is set and cleared by software.
0: CAE interrupt disable
1: CAE interrupt enable

352/1747

RM0090 Rev 17

RM0090

Chrom-Art Accelerator™ controller (DMA2D)

Bit 10 TWIE: Transfer watermark interrupt enable
This bit is set and cleared by software.
0: TW interrupt disable
1: TW interrupt enable
Bit 9 TCIE: Transfer complete interrupt enable
This bit is set and cleared by software.
0: TC interrupt disable
1: TC interrupt enable
Bit 8 TEIE: Transfer error interrupt enable
This bit is set and cleared by software.
0: TE interrupt disable
1: TE interrupt enable
Bits 7:3 Reserved, must be kept at reset value
Bit 2 ABORT: Abort
This bit can be used to abort the current transfer. This bit is set by software and is
automatically reset by hardware when the START bit is reset.
0: No transfer abort requested
1: Transfer abort requested
Bit 1 SUSP: Suspend
This bit can be used to suspend the current transfer. This bit is set and reset by
software. It is automatically reset by hardware when the START bit is reset.
0: Transfer not suspended
1: Transfer suspended
Bit 0 START: Start
This bit can be used to launch the DMA2D according to the parameters loaded in the
various configuration registers. This bit is automatically reset by the following events:
–
At the end of the transfer
–
When the data transfer is aborted by the user application by setting the ABORT
bit in DMA2D_CR
–
When a data transfer error occurs
–
When the data transfer has not started due to a configuration error or another
transfer operation already ongoing (automatic CLUT loading).

RM0090 Rev 17

353/1747
370

Chrom-Art Accelerator™ controller (DMA2D)

11.5.2

RM0090

DMA2D Interrupt Status Register (DMA2D_ISR)
Address offset: 0x0004
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

CEIF

CTCIF

CAEIF

TWIF

TCIF

TEIF

r

r

r

r

r

r

Reserved
15

14

13

12

11

10

9

8

7

Reserved

Bits 31:6 Reserved, must be kept at reset value
Bit 5 CEIF: Configuration error interrupt flag
This bit is set when the START bit of DMA2D_CR, DMA2DFGPFCCR or
DMA2D_BGPFCCR is set and a wrong configuration has been programmed.
Bit 4 CTCIF: CLUT transfer complete interrupt flag
This bit is set when the CLUT copy from a system memory area to the internal DMA2D
memory is complete.
Bit 3 CAEIF: CLUT access error interrupt flag
This bit is set when the CPU accesses the CLUT while the CLUT is being automatically
copied from a system memory to the internal DMA2D.
Bit 2 TWIF: Transfer watermark interrupt flag
This bit is set when the last pixel of the watermarked line has been transferred.
Bit 1 TCIF: Transfer complete interrupt flag
This bit is set when a DMA2D transfer operation is complete (data transfer only).
Bit 0 TEIF: Transfer error interrupt flag
This bit is set when an error occurs during a DMA transfer (data transfer or automatic
CLUT loading).

354/1747

RM0090 Rev 17

RM0090

11.5.3

Chrom-Art Accelerator™ controller (DMA2D)

DMA2D interrupt flag clear register (DMA2D_IFCR)
Address offset: 0x0008
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

CTWIF

CTCIF

CTEIF

rc_w1

rc_w1

rc_w1

Reserved
15

14

13

12

11

10

9

8

7

CCEIF

Reserved

rc_w1

CCTCIF CAECIF
rc_w1

rc_w1

Bits 31:6 Reserved, must be kept at reset value
Bit 5 CCEIF: Clear configuration error interrupt flag
Programming this bit to 1 clears the CEIF flag in the DMA2D_ISR register
Bit 4 CCTCIF: Clear CLUT transfer complete interrupt flag
Programming this bit to 1 clears the CTCIF flag in the DMA2D_ISR register
Bit 3 CAECIF: Clear CLUT access error interrupt flag
Programming this bit to 1 clears the CAEIF flag in the DMA2D_ISR register
Bit 2 CTWIF: Clear transfer watermark interrupt flag
Programming this bit to 1 clears the TWIF flag in the DMA2D_ISR register
Bit 1 CTCIF: Clear transfer complete interrupt flag
Programming this bit to 1 clears the TCIF flag in the DMA2D_ISR register
Bit 0 CTEIF: Clear Transfer error interrupt flag
Programming this bit to 1 clears the TEIF flag in the DMA2D_ISR register

RM0090 Rev 17

355/1747
370

Chrom-Art Accelerator™ controller (DMA2D)

11.5.4

RM0090

DMA2D foreground memory address register (DMA2D_FGMAR)
Address offset: 0x000C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

MA[31:16]
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

MA[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 MA[31: 0]: Memory address
Address of the data used for the foreground image. This register can only be written
when data transfers are disabled. Once the data transfer has started, this register is
read-only.
The address alignment must match the image format selected e.g. a 32-bit per pixel
format must be 32-bit aligned, a 16-bit per pixel format must be 16-bit aligned and a 4bit per pixel format must be 8-bit aligned.

11.5.5

DMA2D foreground offset register (DMA2D_FGOR)
Address offset: 0x0010
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

Reserved
15

14

13

12

11

10

9

8

7

LO[13:0]

Reserved
rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:14 Reserved, must be kept at reset value
Bits 13:0 LO[13: 0]: Line offset
Line offset used for the foreground expressed in pixel. This value is used to generate
the address. It is added at the end of each line to determine the starting address of the
next line.
These bits can only be written when data transfers are disabled. Once a data transfer
has started, they become read-only.
If the image format is 4-bit per pixel, the line offset must be even.

356/1747

RM0090 Rev 17

RM0090

Chrom-Art Accelerator™ controller (DMA2D)

11.5.6

DMA2D background memory address register (DMA2D_BGMAR)
Address offset: 0x0014
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

MA[31:16]
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

MA[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31: 0 MA[31: 0]: Memory address
Address of the data used for the background image. This register can only be written
when data transfers are disabled. Once a data transfer has started, this register is readonly.
The address alignment must match the image format selected e.g. a 32-bit per pixel
format must be 32-bit aligned, a 16-bit per pixel format must be 16-bit aligned and a 4bit per pixel format must be 8-bit aligned.

11.5.7

DMA2D background offset register (DMA2D_BGOR)
Address offset: 0x0018
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

Reserved
15

14

13

12

11

10

9

8

7

LO[13:0]

Reserved
rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:14 Reserved, must be kept at reset value
Bits 13:0 LO[13: 0]: Line offset
Line offset used for the background image (expressed in pixel). This value is used for
the address generation. It is added at the end of each line to determine the starting
address of the next line.
These bits can only be written when data transfers are disabled. Once data transfer has
started, they become read-only.
If the image format is 4-bit per pixel, the line offset must be even.

RM0090 Rev 17

357/1747
370

Chrom-Art Accelerator™ controller (DMA2D)

11.5.8

RM0090

DMA2D foreground PFC control register (DMA2D_FGPFCCR)
Address offset: 0x001C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

ALPHA[7:0]
rw

rw

rw

rw

rw

rw

rw

rw

14

13

12

11

10

9

8

CS[7:0]
rw

rw

rw

rw

19

18

17

7

6

Reserved
rw

rw

rw

5

4

START

CCM

rs

rw

16

AM[1:0]

Reserved

15

rw

20

3

rw

rw

1

0

rw

rw

2
CM[3:0]

rw

rw

Bits 31:24 ALPHA[7: 0]: Alpha value
These bits define a fixed alpha channel value which can replace the original alpha value
or be multiplied by the original alpha value according to the alpha mode selected
through the AM[1:0] bits.
These bits can only be written when data transfers are disabled. Once a transfer has
started, they become read-only.
Bits 23:18 Reserved, must be kept at reset value
Bits 17:16 AM[1: 0]: Alpha mode
These bits select the alpha channel value to be used for the foreground image. They
can only be written data the transfer are disabled. Once the transfer has started, they
become read-only.
00: No modification of the foreground image alpha channel value
01: Replace original foreground image alpha channel value by ALPHA[7: 0]
10: Replace original foreground image alpha channel value by ALPHA[7:0] multiplied
with original alpha channel value
other configurations are meaningless
Bits 15:8 CS[7: 0]: CLUT size
These bits define the size of the CLUT used for the foreground image. Once the CLUT
transfer has started, this field is read-only.
The number of CLUT entries is equal to CS[7:0] + 1.
Bits 7:6 Reserved, must be kept at reset value

358/1747

RM0090 Rev 17

RM0090

Chrom-Art Accelerator™ controller (DMA2D)

Bit 5 START: Start
This bit can be set to start the automatic loading of the CLUT. It is automatically reset:
–
at the end of the transfer
–
when the transfer is aborted by the user application by setting the ABORT bit in
DMA2D_CR
–
when a transfer error occurs
–
when the transfer has not started due to a configuration error or another
transfer operation already ongoing (data transfer or automatic background
CLUT transfer).
Bit 4 CCM: CLUT color mode
This bit defines the color format of the CLUT. It can only be written when the transfer is
disabled. Once the CLUT transfer has started, this bit is read-only.
0: ARGB8888
1: RGB888
others: meaningless
Bits 3:0 CM[3: 0]: Color mode
These bits defines the color format of the foreground image. They can only be written
when data transfers are disabled. Once the transfer has started, they are read-only.
0000: ARGB8888
0001: RGB888
0010: RGB565
0011: ARGB1555
0100: ARGB4444
0101: L8
0110: AL44
0111: AL88
1000: L4
1001: A8
1010: A4
others: meaningless

RM0090 Rev 17

359/1747
370

Chrom-Art Accelerator™ controller (DMA2D)

11.5.9

RM0090

DMA2D foreground color register (DMA2D_FGCOLR)
Address offset: 0x0020
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

14

13

12

11

10

9

8

rw

rw

rw

rw

17

16

rw

rw

rw

rw

rw

rw

rw

rw

7

6

5

4

3

2

1

0

rw

rw

rw

GREEN[7:0]
rw

18

RED[7:0]

Reserved
15

19

BLUE[7:0]
rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:24 Reserved, must be kept at reset value
Bits 23:16 RED[7: 0]: Red Value
These bits defines the red value for the A4 or A8 mode of the foreground image. They
can only be written when data transfers are disabled. Once the transfer has started,
they are read-only.
Bits 15:8 GREEN[7: 0]: Green Value
These bits defines the green value for the A4 or A8 mode of the foreground image. They
can only be written when data transfers are disabled. Once the transfer has started,
They are read-only.
Bits 7:0 BLUE[7: 0]: Blue Value
These bits defines the blue value for the A4 or A8 mode of the foreground image. They
can only be written when data transfers are disabled. Once the transfer has started,
They are read-only.

360/1747

RM0090 Rev 17

RM0090

Chrom-Art Accelerator™ controller (DMA2D)

11.5.10

DMA2D background PFC control register (DMA2D_BGPFCCR)
Address offset: 0x0024
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

ALPHA[7:0]
rw

rw

rw

rw

rw

rw

rw

rw

14

13

12

11

10

9

8

CS[7:0]
rw

rw

rw

rw

19

18

17

7

6

Reserved
rw

rw

rw

5

4

START

CCM

rs

rw

16

AM[1:0]

Reserved

15

rw

20

3

rw

rw

1

0

rw

rw

2
CM[3:0]

rw

rw

Bits 31:24 ALPHA[7: 0]: Alpha value
These bits define a fixed alpha channel value which can replace the original alpha value
or be multiplied with the original alpha value according to the alpha mode selected with
bits AM[1: 0]. These bits can only be written when data transfers are disabled. Once the
transfer has started, they are read-only.
Bits 23:18 Reserved, must be kept at reset value
Bits 17:16 AM[1: 0]: Alpha mode
These bits define which alpha channel value to be used for the background image.
These bits can only be written when data transfers are disabled. Once the transfer has
started, they are read-only.
00: No modification of the foreground image alpha channel value
01: Replace original background image alpha channel value by ALPHA[7: 0]
10: Replace original background image alpha channel value by ALPHA[7:0] multiplied
with original alpha channel value
others: meaningless
Bits 15:8 CS[7: 0]: CLUT size
These bits define the size of the CLUT used for the BG. Once the CLUT transfer has
started, this field is read-only.
The number of CLUT entries is equal to CS[7:0] + 1.
Bits 7:6 Reserved, must be kept at reset value

RM0090 Rev 17

361/1747
370

Chrom-Art Accelerator™ controller (DMA2D)

RM0090

Bit 5 START: Start
This bit is set to start the automatic loading of the CLUT. This bit is automatically reset:
–
at the end of the transfer
–
when the transfer is aborted by the user application by setting the ABORT bit in
the DMA2D_CR
–
when a transfer error occurs
–
when the transfer has not started due to a configuration error or another
transfer operation already on going (data transfer or automatic foreground
CLUT transfer).
Bit 4 CCM: CLUT Color mode
These bits define the color format of the CLUT. This register can only be written when
the transfer is disabled. Once the CLUT transfer has started, this bit is read-only.
0: ARGB8888
1: RGB888
others: meaningless
Bits 3:0 CM[3: 0]: Color mode
These bits define the color format of the foreground image. These bits can only be
written when data transfers are disabled. Once the transfer has started, they are readonly.
0000: ARGB8888
0001: RGB888
0010: RGB565
0011: ARGB1555
0100: ARGB4444
0101: L8
0110: AL44
0111: AL88
1000: L4
1001: A8
1010: A4
others: meaningless

362/1747

RM0090 Rev 17

RM0090

Chrom-Art Accelerator™ controller (DMA2D)

11.5.11

DMA2D background color register (DMA2D_BGCOLR)
Address offset: 0x0028
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

14

13

12

11

10

9

8

rw

rw

rw

rw

17

16

rw

rw

rw

rw

rw

rw

rw

rw

7

6

5

4

3

2

1

0

rw

rw

rw

GREEN[7:0]
rw

18

RED[7:0]

Reserved
15

19

BLUE[7:0]
rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:24 Reserved, must be kept at reset value
Bits 23:16 RED[7: 0]: Red Value
These bits define the red value for the A4 or A8 mode of the background. These bits
can only be written when data transfers are disabled. Once the transfer has started,
they are read-only.
Bits 15:8 GREEN[7: 0]: Green Value
These bits define the green value for the A4 or A8 mode of the background. These bits
can only be written when data transfers are disabled. Once the transfer has started,
they are read-only.
Bits 7:0 BLUE[7: 0]: Blue Value
These bits define the blue value for the A4 or A8 mode of the background. These bits
can only be written when data transfers are disabled. Once the transfer has started,
they are read-only.

11.5.12

DMA2D foreground CLUT memory address register
(DMA2D_FGCMAR)
Address offset: 0x002C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

MA[31:16]
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

MA[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31: 0 MA[31: 0]: Memory Address
Address of the data used for the CLUT address dedicated to the foreground image. This
register can only be written when no transfer is ongoing. Once the CLUT transfer has
started, this register is read-only.
If the foreground CLUT format is 32-bit, the address must be 32-bit aligned.

RM0090 Rev 17

363/1747
370

Chrom-Art Accelerator™ controller (DMA2D)

11.5.13

RM0090

DMA2D background CLUT memory address register
(DMA2D_BGCMAR)
Address offset: 0x0030
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

MA[31:16]
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

MA[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31: 0 MA[31: 0]: Memory address
Address of the data used for the CLUT address dedicated to the background image.
This register can only be written when no transfer is on going. Once the CLUT transfer
has started, this register is read-only.
If the background CLUT format is 32-bit, the address must be 32-bit aligned.

11.5.14

DMA2D output PFC control register (DMA2D_OPFCCR)
Address offset: 0x0034
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

Reserved
15

14

13

12

11

10

9

8

7

CM[2:0]

Reserved
rw

rw

rw

Bits 31: 3 Reserved, must be kept at reset value
Bits 2: 0 CM[2: 0]: Color mode
These bits define the color format of the output image. These bits can only be written
when data transfers are disabled. Once the transfer has started, they are read-only.
000: ARGB8888
001: RGB888
010: RGB565
011: ARGB1555
100: ARGB4444
others: meaningless

364/1747

RM0090 Rev 17

RM0090

Chrom-Art Accelerator™ controller (DMA2D)

11.5.15

DMA2D output color register (DMA2D_OCOLR)
Address offset: 0x0038
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

ALPHA[7:0]

19

18

17

16

RED[7:0]

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

GREEN[7:0]

BLUE[7:0]

RED[4:0]
A

GREEN[5:0]

RED[4:0]

GREEN[4:0]

ALPHA[3:0]
rw

rw

rw

RED[3:0]
rw

BLUE[4:0]

rw

rw

rw

BLUE[4:0]
GREEN[3:0]

rw

rw

rw

rw

BLUE[3:0]
rw

rw

rw

rw

rw

Bits 31:24 ALPHA[7: 0]: Alpha Channel Value
These bits define the alpha channel of the output color. These bits can only be written
when data transfers are disabled. Once the transfer has started, they are read-only.
Bits 23:16 RED[7: 0]: Red Value
These bits define the red value of the output image. These bits can only be written when
data transfers are disabled. Once the transfer has started, they are read-only.
Bits 15:8 GREEN[7: 0]: Green Value
These bits define the green value of the output image. These bits can only be written
when data transfers are disabled. Once the transfer has started, they are read-only.
Bits 7:0 BLUE[7: 0]: Blue Value
These bits define the blue value of the output image. These bits can only be written
when data transfers are disabled. Once the transfer has started, they are read-only.

RM0090 Rev 17

365/1747
370

Chrom-Art Accelerator™ controller (DMA2D)

11.5.16

RM0090

DMA2D output memory address register (DMA2D_OMAR)
Address offset: 0x003C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

MA[31:16]
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

MA[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31: 0 MA[31: 0]: Memory Address
Address of the data used for the output FIFO. These bits can only be written when data
transfers are disabled. Once the transfer has started, they are read-only.
The address alignment must match the image format selected e.g. a 32-bit per pixel
format must be 32-bit aligned and a 16-bit per pixel format must be 16-bit aligned.

366/1747

RM0090 Rev 17

RM0090

Chrom-Art Accelerator™ controller (DMA2D)

11.5.17

DMA2D output offset register (DMA2D_OOR)
Address offset: 0x0040
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

Reserved
15

14

13

12

11

10

9

8

7

LO[13:0]

Reserved
rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:14 Reserved, must be kept at reset value
Bits 13:0 LO[13: 0]: Line Offset
Line offset used for the output (expressed in pixels). This value is used for the address
generation. It is added at the end of each line to determine the starting address of the
next line. These bits can only be written when data transfers are disabled. Once the
transfer has started, they are read-only.

11.5.18

DMA2D number of line register (DMA2D_NLR)
Address offset: 0x0044
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

14

21

20

19

18

17

16

PL[13:0]

Reserved
15

22

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

NL[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:30 Reserved, must be kept at reset value
Bits 29:16 PL[13: 0]: Pixel per lines
Number of pixels per lines of the area to be transferred. These bits can only be written
when data transfers are disabled. Once the transfer has started, they are read-only.
If any of the input image format is 4-bit per pixel, pixel per lines must be even.
Bits 15:0 NL[15: 0]: Number of lines
Number of lines of the area to be transferred. These bits can only be written when data
transfers are disabled. Once the transfer has started, they are read-only.

RM0090 Rev 17

367/1747
370

Chrom-Art Accelerator™ controller (DMA2D)

11.5.19

RM0090

DMA2D line watermark register (DMA2D_LWR)
Address offset: 0x0048
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

Reserved
15

14

13

12

11

10

9

8

7
LW[15:0]

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value
Bits 15:0 LW[15:0]: Line watermark
These bits allow to configure the line watermark for interrupt generation.
An interrupt is raised when the last pixel of the watermarked line has been transferred.
These bits can only be written when data transfers are disabled. Once the transfer has
started, they are read-only.

11.5.20

DMA2D AHB master timer configuration register (DMA2D_AMTCR)
Address offset: 0x004C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

Reserved
15

14

13

12

11

10

9

8

7

DT[7:0]
rw

rw

rw

rw

rw

Reserved
rw

rw

rw

EN
rw

Bits 31:16 Reserved
Bits 15:8 DT[7: 0]: Dead Time
Dead time value in the AHB clock cycle inserted between two consecutive accesses on
the AHB master port. These bits represent the minimum guaranteed number of cycles
between two consecutive AHB accesses.
Bits 7:1 Reserved
Bit 0 EN: Enable
Enables the dead time functionality.

368/1747

RM0090 Rev 17

RM0090

Chrom-Art Accelerator™ controller (DMA2D)

11.5.21

DMA2D register map
The following table summarizes the DMA2D registers. Refer to Section 2.3: Memory map
for the DMA2D register base address.

Reserved

0

0

0

0

CAECIF

CTWIF

CTCIF

CTEIF

0

0

0

0

0

0

0x0024

0x002C
0x0030
0x0034

0

0

0

0

0

0

0

Reset value

0

0

0

0

0

0

0

0

0

0

ALPHA[7:0]
0

0

DMA2D_FGCOLR

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0
0

0
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Res

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Res

0

0

0

GREEN[7:0]
0

0

0

0

CM[3:0]
0

0

0

0

0

0

BLUE[7:0]

CS[7:0]

RED[7:0]
0

0

GREEN[7:0]

Reserved

APLHA[7:0]
0

0

0

CS[7:0]

RED[7:0]

DMA2D_FGCMAR

0

0

CM[3:0]
0

0

0

0

BLUE[7:0]

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

MA[31:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMA2D_BGCMAR

0

0

MA[31:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

DMA2D_OPFCCR

0

0

0

0

CM[2:0]

Reserved

Reset value

0

RED[7:0]

GREEN[7:0]

Reserved

DMA2D_OCOLR

0

0

0

0

0

0

0

RED[4:0]

Reserved

A

Reserved

ALPHA[3:0]

0

0

0

0

0

0

0

0

DMA2D_OMAR
Reset value

0

0

0

Reserved

ALPHA[7:0]

DMA2D_BGCOLR

Reset value

0

LO[13:0]

APLHA[7:0]

DMA2D_BGPFCCR

Reset value

0

Reserved

DMA2D_FGPFCCR

Reset value

0

MA[31:0]

Reset value

Reset value

0

Reserved

DMA2D_BGOR

Reset value
0x003C

0

LO[13:0]

APLHA[7:0]
0x0038

0

CCM

0

DMA2D_BGMAR

Reset value
0x0028

0

Reset value

Reset value
0x0020

0

START

0x001C

0

CCM

0x0018

0

START

0x0014

0

DMA2D_FGOR

AM[1:0]

0x0010

Reset value

MA[31:0]

AM[1:0]

0x000C

SUSP

0

Reset value
DMA2D_FGMAR

0

CCEIF

Reserved

0

CCTCIF

DMA2D_IFCR

0

0

Reset value
0x0008

START

0

TEIF

0

ABORT

0

TCIF

0

TWIF

TEIE

0

CAEIF

TCIE

0

CEIF

DMA2D_ISR

0

Reserved

CTCIF

0x0004

0

TWIE

Reset value

Reserved

CAEIE

Reserved

CEIE

DMA2D_CR

CTCIE

0x0000

Register

MODE[1:0]

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 60. DMA2D register map and reset values

0

0

0

0

BLUE[7:0]
GREEN[6:0]

RED[4:0]

BLUE[4:0]

GREEN[4:0]
RED[3:0]

BLUE[4:0]

GREEN[3:0]

BLUE[3:0]

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

MA[31:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

RM0090 Rev 17

0

0

369/1747
370

Chrom-Art Accelerator™ controller (DMA2D)

RM0090

Register

DMA2D_OOR

0x0040

DMA2D_NLR

0x0044

Reset value

0x004C

Res

0

0

0

0

0

DMA2D_AMTCR

0

0

0

0

0

0

0

0

0x08000x0BFF

DMA2D_BGCLUT

Reset value

370/1747

Reset value

0

0
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DT[7:0]
0

DMA2D_FGCLUT

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

LW[15:0]

Reserved

Reset value

0x04000x07FF

0

NL[15:0]

Reserved

Reset value

0x0050Ox03FF

0

PL[13:0]

DMA2D_LWR

0x0048

LO[13:0]

Reserved

Reset value

0

0

0

0

0

Reserved

EN

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 60. DMA2D register map and reset values (continued)

0

Reserved
APLHA[7:0][255:0]

RED[7:0][255:0]

GREEN[7:0][255:0]

BLUE[7:0][255:0]

X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X
APLHA[7:0][255:0]

RED[7:0][255:0]

GREEN[7:0][255:0]

BLUE[7:0][255:0]

X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X

RM0090 Rev 17

RM0090

12

Interrupts and events

Interrupts and events
This Section applies to the whole STM32F4xx family, unless otherwise specified.

12.1

Nested vectored interrupt controller (NVIC)

12.1.1

NVIC features
The nested vector interrupt controller NVIC includes the following features:
•

82 maskable interrupt channels for STM32F405xx/07xx and STM32F415xx/17xx, and
up to 91 maskable interrupt channels for STM32F42xxx and STM32F43xxx (not
including the 16 interrupt lines of Cortex®-M4 with FPU)

•

16 programmable priority levels (4 bits of interrupt priority are used)

•

low-latency exception and interrupt handling

•

power management control

•

implementation of system control registers

The NVIC and the processor core interface are closely coupled, which enables low latency
interrupt processing and efficient processing of late arriving interrupts.
All interrupts including the core exceptions are managed by the NVIC. For more information
on exceptions and NVIC programming, refer to programming manual PM0214.

12.1.2

SysTick calibration value register
The SysTick calibration value is fixed to 18750, which gives a reference time base of 1 ms
with the SysTick clock set to 18.75 MHz (HCLK/8, with HCLK set to 150 MHz).

12.1.3

Interrupt and exception vectors
See Table 62 and Table 62, for the vector table for the STM32F405xx/07xx and
STM32F415xx/17xx and STM32F42xxx and STM32F43xxx devices.

12.2

External interrupt/event controller (EXTI)
The external interrupt/event controller consists of up to 23 edge detectors for generating
event/interrupt requests. Each input line can be independently configured to select the type
(interrupt or event) and the corresponding trigger event (rising or falling or both). Each line
can also masked independently. A pending register maintains the status line of the interrupt
requests.

RM0090 Rev 17

371/1747
387

Interrupts and events

RM0090

Priority

Position

Table 61. Vector table for STM32F405xx/07xx and STM32F415xx/17xx

Type of
priority

-

-

-3

Acronym

Description

Address

-

Reserved

0x0000 0000

fixed

Reset

Reset

0x0000 0004

-2

fixed

NMI

Non maskable interrupt. The RCC
Clock Security System (CSS) is linked
to the NMI vector.

0x0000 0008

-1

fixed

HardFault

All class of fault

0x0000 000C

0

settable

MemManage

Memory management

0x0000 0010

1

settable

BusFault

Pre-fetch fault, memory access fault

0x0000 0014

2

settable

UsageFault

Undefined instruction or illegal state

0x0000 0018

-

-

-

Reserved

3

settable

SVCall

System service call via SWI
instruction

0x0000 002C

4

settable

Debug Monitor

Debug Monitor

0x0000 0030

-

-

-

Reserved

0x0000 0034

5

settable

PendSV

Pendable request for system service

0x0000 0038

6

settable

SysTick

System tick timer

0x0000 003C

0

7

settable

WWDG

Window Watchdog interrupt

0x0000 0040

1

8

settable

PVD

PVD through EXTI line detection
interrupt

0x0000 0044

2

9

settable

TAMP_STAMP

Tamper and TimeStamp interrupts
through the EXTI line

0x0000 0048

3

10

settable

RTC_WKUP

RTC Wakeup interrupt through the
EXTI line

0x0000 004C

4

11

settable

FLASH

Flash global interrupt

0x0000 0050

5

12

settable

RCC

RCC global interrupt

0x0000 0054

6

13

settable

EXTI0

EXTI Line0 interrupt

0x0000 0058

7

14

settable

EXTI1

EXTI Line1 interrupt

0x0000 005C

8

15

settable

EXTI2

EXTI Line2 interrupt

0x0000 0060

9

16

settable

EXTI3

EXTI Line3 interrupt

0x0000 0064

10

17

settable

EXTI4

EXTI Line4 interrupt

0x0000 0068

11

18

settable

DMA1_Stream0

DMA1 Stream0 global interrupt

0x0000 006C

372/1747

RM0090 Rev 17

0x0000 001C - 0x0000
002B

RM0090

Interrupts and events

Position

Priority

Table 61. Vector table for STM32F405xx/07xx and STM32F415xx/17xx (continued)

Type of
priority

12

19

settable

DMA1_Stream1

DMA1 Stream1 global interrupt

0x0000 0070

13

20

settable

DMA1_Stream2

DMA1 Stream2 global interrupt

0x0000 0074

14

21

settable

DMA1_Stream3

DMA1 Stream3 global interrupt

0x0000 0078

15

22

settable

DMA1_Stream4

DMA1 Stream4 global interrupt

0x0000 007C

16

23

settable

DMA1_Stream5

DMA1 Stream5 global interrupt

0x0000 0080

17

24

settable

DMA1_Stream6

DMA1 Stream6 global interrupt

0x0000 0084

18

25

settable

ADC

ADC1, ADC2 and ADC3 global
interrupts

0x0000 0088

19

26

settable

CAN1_TX

CAN1 TX interrupts

0x0000 008C

20

27

settable

CAN1_RX0

CAN1 RX0 interrupts

0x0000 0090

21

28

settable

CAN1_RX1

CAN1 RX1 interrupt

0x0000 0094

22

29

settable

CAN1_SCE

CAN1 SCE interrupt

0x0000 0098

23

30

settable

EXTI9_5

EXTI Line[9:5] interrupts

0x0000 009C

24

31

settable

TIM1_BRK_TIM9

TIM1 Break interrupt and TIM9 global
interrupt

0x0000 00A0

25

32

settable

TIM1_UP_TIM10

TIM1 Update interrupt and TIM10
global interrupt

0x0000 00A4

26

33

settable

TIM1_TRG_COM_TIM11

TIM1 Trigger and Commutation
interrupts and TIM11 global interrupt

0x0000 00A8

27

34

settable

TIM1_CC

TIM1 Capture Compare interrupt

0x0000 00AC

28

35

settable

TIM2

TIM2 global interrupt

0x0000 00B0

29

36

settable

TIM3

TIM3 global interrupt

0x0000 00B4

30

37

settable

TIM4

TIM4 global interrupt

0x0000 00B8

31

38

settable

I2C1_EV

I2C1 event interrupt

0x0000 00BC

I2C1_ER

I2C1

0x0000 00C0

32
33

39
40

settable
settable

Acronym

I2C2_EV

Description

error interrupt

2

I C2 event interrupt
2C2

error interrupt

Address

0x0000 00C4
0x0000 00C8

34

41

settable

I2C2_ER

I

35

42

settable

SPI1

SPI1 global interrupt

0x0000 00CC

36

43

settable

SPI2

SPI2 global interrupt

0x0000 00D0

37

44

settable

USART1

USART1 global interrupt

0x0000 00D4

38

45

settable

USART2

USART2 global interrupt

0x0000 00D8

39

46

settable

USART3

USART3 global interrupt

0x0000 00DC

RM0090 Rev 17

373/1747
387

Interrupts and events

RM0090

Position

Priority

Table 61. Vector table for STM32F405xx/07xx and STM32F415xx/17xx (continued)

Type of
priority

40

47

settable

EXTI15_10

EXTI Line[15:10] interrupts

0x0000 00E0

41

48

settable

RTC_Alarm

RTC Alarms (A and B) through EXTI
line interrupt

0x0000 00E4

42

49

settable

OTG_FS_WKUP

USB On-The-Go FS Wakeup through
EXTI line interrupt

0x0000 00E8

43

50

settable

TIM8_BRK_TIM12

TIM8 Break interrupt and TIM12
global interrupt

0x0000 00EC

44

51

settable

TIM8_UP_TIM13

TIM8 Update interrupt and TIM13
global interrupt

0x0000 00F0

45

52

settable

TIM8_TRG_COM_TIM14

TIM8 Trigger and Commutation
interrupts and TIM14 global interrupt

0x0000 00F4

46

53

settable

TIM8_CC

TIM8 Capture Compare interrupt

0x0000 00F8

47

54

settable

DMA1_Stream7

DMA1 Stream7 global interrupt

0x0000 00FC

48

55

settable

FSMC

FSMC global interrupt

0x0000 0100

49

56

settable

SDIO

SDIO global interrupt

0x0000 0104

50

57

settable

TIM5

TIM5 global interrupt

0x0000 0108

51

58

settable

SPI3

SPI3 global interrupt

0x0000 010C

52

59

settable

UART4

UART4 global interrupt

0x0000 0110

53

60

settable

UART5

UART5 global interrupt

0x0000 0114

54

61

settable

TIM6_DAC

TIM6 global interrupt,
DAC1 and DAC2 underrun error
interrupts

0x0000 0118

55

62

settable

TIM7

TIM7 global interrupt

0x0000 011C

56

63

settable

DMA2_Stream0

DMA2 Stream0 global interrupt

0x0000 0120

57

64

settable

DMA2_Stream1

DMA2 Stream1 global interrupt

0x0000 0124

58

65

settable

DMA2_Stream2

DMA2 Stream2 global interrupt

0x0000 0128

59

66

settable

DMA2_Stream3

DMA2 Stream3 global interrupt

0x0000 012C

60

67

settable

DMA2_Stream4

DMA2 Stream4 global interrupt

0x0000 0130

61

68

settable

ETH

Ethernet global interrupt

0x0000 0134

62

69

settable

ETH_WKUP

Ethernet Wakeup through EXTI line
interrupt

0x0000 0138

63

70

settable

CAN2_TX

CAN2 TX interrupts

0x0000 013C

64

71

settable

CAN2_RX0

CAN2 RX0 interrupts

0x0000 0140

374/1747

Acronym

Description

RM0090 Rev 17

Address

RM0090

Interrupts and events

Position

Priority

Table 61. Vector table for STM32F405xx/07xx and STM32F415xx/17xx (continued)

Type of
priority

65

72

settable

CAN2_RX1

CAN2 RX1 interrupt

0x0000 0144

66

73

settable

CAN2_SCE

CAN2 SCE interrupt

0x0000 0148

67

74

settable

OTG_FS

USB On The Go FS global interrupt

0x0000 014C

68

75

settable

DMA2_Stream5

DMA2 Stream5 global interrupt

0x0000 0150

69

76

settable

DMA2_Stream6

DMA2 Stream6 global interrupt

0x0000 0154

70

77

settable

DMA2_Stream7

DMA2 Stream7 global interrupt

0x0000 0158

71

78

settable

USART6

USART6 global interrupt

0x0000 015C

72

79

settable

I2C3_EV

I2C3 event interrupt

0x0000 0160

Acronym

Description

2C3

error interrupt

Address

0x0000 0164

73

80

settable

I2C3_ER

I

74

81

settable

OTG_HS_EP1_OUT

USB On The Go HS End Point 1 Out
global interrupt

0x0000 0168

75

82

settable

OTG_HS_EP1_IN

USB On The Go HS End Point 1 In
global interrupt

0x0000 016C

76

83

settable

OTG_HS_WKUP

USB On The Go HS Wakeup through
EXTI interrupt

0x0000 0170

77

84

settable

OTG_HS

USB On The Go HS global interrupt

0x0000 0174

78

85

settable

DCMI

DCMI global interrupt

0x0000 0178

79

86

settable

CRYP

CRYP crypto global interrupt

0x0000 017C

80

87

settable

HASH_RNG

Hash and Rng global interrupt

0x0000 0180

81

88

settable

FPU

FPU global interrupt

0x0000 0184

Priority

Position

Table 62. Vector table for STM32F42xxx and STM32F43xxx

Type of
priority

Acronym

Description

Address

-

-

-

Reserved

0x0000 0000

-3

fixed

Reset

Reset

0x0000 0004

-2

fixed

NMI

Non maskable interrupt, Clock Security
System

0x0000 0008

RM0090 Rev 17

375/1747
387

Interrupts and events

RM0090

Priority

Position

Table 62. Vector table for STM32F42xxx and STM32F43xxx (continued)

Type of
priority

Acronym

Description

Address

-1

fixed

HardFault

All class of fault

0x0000 000C

0

settable

MemManage

Memory management

0x0000 0010

1

settable

BusFault

Pre-fetch fault, memory access fault

0x0000 0014

2

settable

UsageFault

Undefined instruction or illegal state

0x0000 0018

-

-

-

Reserved

0x0000 001C 0x0000 002B

3

settable

SVCall

System Service call via SWI instruction

0x0000 002C

4

settable

Debug Monitor

Debug Monitor

0x0000 0030

-

-

Reserved

0x0000 0034

5

settable

PendSV

Pendable request for system service

0x0000 0038

6

settable

Systick

System tick timer

0x0000 003C

0

7

settable

WWDG

Window Watchdog interrupt

0x0000 0040

1

8

settable

PVD

PVD through EXTI line detection interrupt

0x0000 0044

2

9

settable

TAMP_STAMP

Tamper and TimeStamp interrupts through
the EXTI line

0x0000 0048

3

10

settable

RTC_WKUP

RTC Wakeup interrupt through the EXTI
line

0x0000 004C

4

11

settable

FLASH

Flash global interrupt

0x0000 0050

5

12

settable

RCC

RCC global interrupt

0x0000 0054

6

13

settable

EXTI0

EXTI Line0 interrupt

0x0000 0058

7

14

settable

EXTI1

EXTI Line1 interrupt

0x0000 005C

8

15

settable

EXTI2

EXTI Line2 interrupt

0x0000 0060

9

16

settable

EXTI3

EXTI Line3 interrupt

0x0000 0064

10

17

settable

EXTI4

EXTI Line4 interrupt

0x0000 0068

11

18

settable

DMA1_Stream0

DMA1 Stream0 global interrupt

0x0000 006C

12

19

settable

DMA1_Stream1

DMA1 Stream1 global interrupt

0x0000 0070

13

20

settable

DMA1_Stream2

DMA1 Stream2 global interrupt

0x0000 0074

14

21

settable

DMA1_Stream3

DMA1 Stream3 global interrupt

0x0000 0078

15

22

settable

DMA1_Stream4

DMA1 Stream4 global interrupt

0x0000 007C

16

23

settable

DMA1_Stream5

DMA1 Stream5 global interrupt

0x0000 0080

17

24

settable

DMA1_Stream6

DMA1 Stream6 global interrupt

0x0000 0084

376/1747

RM0090 Rev 17

RM0090

Interrupts and events

Position

Priority

Table 62. Vector table for STM32F42xxx and STM32F43xxx (continued)

Type of
priority

18

25

settable

19

26

settable

CAN1_TX

CAN1 TX interrupts

0x0000 008C

20

27

settable

CAN1_RX0

CAN1 RX0 interrupts

0x0000 0090

21

28

settable

CAN1_RX1

CAN1 RX1 interrupt

0x0000 0094

22

29

settable

CAN1_SCE

CAN1 SCE interrupt

0x0000 0098

23

30

settable

EXTI9_5

EXTI Line[9:5] interrupts

0x0000 009C

24

31

settable

TIM1_BRK_TIM9

TIM1 Break interrupt and TIM9 global
interrupt

0x0000 00A0

25

32

settable

TIM1_UP_TIM10

TIM1 Update interrupt and TIM10 global
interrupt

0x0000 00A4

26

33

settable

TIM1_TRG_COM_TIM11

TIM1 Trigger and Commutation interrupts
and TIM11 global interrupt

0x0000 00A8

27

34

settable

TIM1_CC

TIM1 Capture Compare interrupt

0x0000 00AC

28

35

settable

TIM2

TIM2 global interrupt

0x0000 00B0

29

36

settable

TIM3

TIM3 global interrupt

0x0000 00B4

30

37

settable

TIM4

TIM4 global interrupt

0x0000 00B8

Acronym

Description

Address

ADC

ADC1, ADC2 and ADC3 global interrupts

0x0000 0088

2C1

31

38

settable

I2C1_EV

I

event interrupt

0x0000 00BC

32

39

settable

I2C1_ER

I2C1 error interrupt

0x0000 00C0

I2C2_EV

I2C2

event interrupt

0x0000 00C4

33

40

settable

2

34

41

settable

I2C2_ER

I C2 error interrupt

0x0000 00C8

35

42

settable

SPI1

SPI1 global interrupt

0x0000 00CC

36

43

settable

SPI2

SPI2 global interrupt

0x0000 00D0

37

44

settable

USART1

USART1 global interrupt

0x0000 00D4

38

45

settable

USART2

USART2 global interrupt

0x0000 00D8

39

46

settable

40

47

settable

EXTI15_10

EXTI Line[15:10] interrupts

0x0000 00E0

41

48

settable

RTC_Alarm

RTC Alarms (A and B) through EXTI line
interrupt

0x0000 00E4

42

49

settable

OTG_FS_WKUP

USB On-The-Go FS Wakeup through EXTI
line interrupt

0x0000 00E8

43

50

settable

TIM8 Break interrupt and TIM12 global
interrupt

0x0000 00EC

USART3

TIM8_BRK_TIM12

USART3 global interrupt

RM0090 Rev 17

0x0000 00DC

377/1747
387

Interrupts and events

RM0090

Position

Priority

Table 62. Vector table for STM32F42xxx and STM32F43xxx (continued)

Type of
priority

44

51

settable

TIM8_UP_TIM13

TIM8 Update interrupt and TIM13 global
interrupt

0x0000 00F0

45

52

settable

TIM8_TRG_COM_TIM1
4

TIM8 Trigger and Commutation interrupts
and TIM14 global interrupt

0x0000 00F4

46

53

settable

TIM8_CC

TIM8 Capture Compare interrupt

0x0000 00F8

47

54

settable

48

55

settable

49

56

settable

SDIO

SDIO global interrupt

0x0000 0104

50

57

settable

TIM5

TIM5 global interrupt

0x0000 0108

51

58

settable

SPI3

SPI3 global interrupt

0x0000 010C

52

59

settable

UART4

UART4 global interrupt

0x0000 0110

53

60

settable

UART5

UART5 global interrupt

0x0000 0114

54

61

settable

TIM6_DAC

TIM6 global interrupt,
DAC1 and DAC2 underrun error interrupts

0x0000 0118

55

62

settable

TIM7

TIM7 global interrupt

0x0000 011C

56

63

settable

DMA2_Stream0

DMA2 Stream0 global interrupt

0x0000 0120

57

64

settable

DMA2_Stream1

DMA2 Stream1 global interrupt

0x0000 0124

58

65

settable

DMA2_Stream2

DMA2 Stream2 global interrupt

0x0000 0128

59

66

settable

DMA2_Stream3

DMA2 Stream3 global interrupt

0x0000 012C

60

67

settable

DMA2_Stream4

DMA2 Stream4 global interrupt

0x0000 0130

61

68

settable

ETH

Ethernet global interrupt

0x0000 0134

62

69

settable

ETH_WKUP

Ethernet Wakeup through EXTI line
interrupt

0x0000 0138

63

70

settable

CAN2_TX

CAN2 TX interrupts

0x0000 013C

64

71

settable

CAN2_RX0

CAN2 RX0 interrupts

0x0000 0140

65

72

settable

CAN2_RX1

CAN2 RX1 interrupt

0x0000 0144

66

73

settable

CAN2_SCE

CAN2 SCE interrupt

0x0000 0148

67

74

settable

OTG_FS

USB On The Go FS global interrupt

0x0000 014C

68

75

settable

DMA2_Stream5

DMA2 Stream5 global interrupt

0x0000 0150

69

76

settable

DMA2_Stream6

DMA2 Stream6 global interrupt

0x0000 0154

70

77

settable

DMA2_Stream7

DMA2 Stream7 global interrupt

0x0000 0158

71

78

settable

USART6

USART6 global interrupt

0x0000 015C

378/1747

Acronym

DMA1_Stream7
FSMC

Description

DMA1 Stream7 global interrupt
FSMC global interrupt

RM0090 Rev 17

Address

0x0000 00FC
0x0000 0100

RM0090

Interrupts and events

Position

Priority

Table 62. Vector table for STM32F42xxx and STM32F43xxx (continued)

Type of
priority

72

79

settable

Acronym

Description

Address

I2C3_EV

I2C3 event interrupt

0x0000 0160

I2C3_ER

I2

0x0000 0164

73

80

settable

74

81

settable

OTG_HS_EP1_OUT

USB On The Go HS End Point 1 Out global
interrupt

0x0000 0168

75

82

settable

OTG_HS_EP1_IN

USB On The Go HS End Point 1 In global
interrupt

0x0000 016C

76

83

settable

OTG_HS_WKUP

USB On The Go HS Wakeup through EXTI
interrupt

0x0000 0170

77

84

settable

OTG_HS

USB On The Go HS global interrupt

0x0000 0174

78

85

settable

DCMI

DCMI global interrupt

0x0000 0178

79

86

settable

CRYP

CRYP crypto global interrupt

0x0000 017C

80

87

settable

HASH_RNG

Hash and Rng global interrupt

0x0000 0180

81

88

Settable

82

89

settable

UART7

UART 7 global interrupt

0x0000 0188

83

90

settable

UART8

UART 8 global interrupt

0x0000 018C

84

91

settable

85

92

settable

SPI5

SPI 5 global interrupt

0x0000 0194

86

93

settable

SPI6

SPI 6 global interrupt

0x0000 0198

87

94

settable

SAI1

SAI1 global interrupt

0x0000 019C

88

95

settable

LCD-TFT

LTDC global interrupt

0x0000 01A0

89

96

settable

LCD-TFT

LTDC global Error interrupt

0x0000 01A4

90

97

settable

DMA2D

DMA2D global interrupt

0x0000 01A8

12.2.1

FPU

SPI4

C3 error interrupt

FPU global interrupt

SPI 4 global interrupt

0x0000 0184

0x0000 0190

EXTI main features
The main features of the EXTI controller are the following:
•

independent trigger and mask on each interrupt/event line

•

dedicated status bit for each interrupt line

•

generation of up to 23 software event/interrupt requests

•

detection of external signals with a pulse width lower than the APB2 clock period. Refer
to the electrical characteristics section of the STM32F4xx datasheets for details on this
parameter.

RM0090 Rev 17

379/1747
387

Interrupts and events

12.2.2

RM0090

EXTI block diagram
Figure 41 shows the block diagram.
Figure 41. External interrupt/event controller block diagram
$0%$$3%EXV
3&/.

3HULSKHUDOLQWHUIDFH



3HQGLQJ
UHTXHVW
UHJLVWHU



,QWHUUXSW
PDVN
UHJLVWHU



6RIWZDUH
LQWHUUXSW
HYHQW
UHJLVWHU



7R19,&LQWHUUXSW
FRQWUROOHU







5LVLQJ
WULJJHU
VHOHFWLRQ
UHJLVWHU


)DOOLQJ
WULJJHU
VHOHFWLRQ
UHJLVWHU






3XOVH
JHQHUDWRU





(GJHGHWHFW
FLUFXLW

,QSXW
OLQH

(YHQW
PDVN
UHJLVWHU
069

12.2.3

Wakeup event management
The STM32F4xx are able to handle external or internal events in order to wake up the core
(WFE). The wakeup event can be generated either by:
•

enabling an interrupt in the peripheral control register but not in the NVIC, and enabling
the SEVONPEND bit in the Cortex®-M4 with FPU System Control register. When the
MCU resumes from WFE, the peripheral interrupt pending bit and the peripheral NVIC
IRQ channel pending bit (in the NVIC interrupt clear pending register) have to be
cleared.

•

or configuring an external or internal EXTI line in event mode. When the CPU resumes
from WFE, it is not necessary to clear the peripheral interrupt pending bit or the NVIC
IRQ channel pending bit as the pending bit corresponding to the event line is not set.

To use an external line as a wakeup event, refer to Section 12.2.4: Functional description.

12.2.4

Functional description
To generate the interrupt, the interrupt line should be configured and enabled. This is done
by programming the two trigger registers with the desired edge detection and by enabling
the interrupt request by writing a ‘1’ to the corresponding bit in the interrupt mask register.
When the selected edge occurs on the external interrupt line, an interrupt request is

380/1747

RM0090 Rev 17

RM0090

Interrupts and events
generated. The pending bit corresponding to the interrupt line is also set. This request is
reset by writing a ‘1’ in the pending register.
To generate the event, the event line should be configured and enabled. This is done by
programming the two trigger registers with the desired edge detection and by enabling the
event request by writing a ‘1’ to the corresponding bit in the event mask register. When the
selected edge occurs on the event line, an event pulse is generated. The pending bit
corresponding to the event line is not set.
An interrupt/event request can also be generated by software by writing a ‘1’ in the software
interrupt/event register.

Hardware interrupt selection
To configure the 23 lines as interrupt sources, use the following procedure:
•

Configure the mask bits of the 23 interrupt lines (EXTI_IMR)

•

Configure the Trigger selection bits of the interrupt lines (EXTI_RTSR and EXTI_FTSR)

•

Configure the enable and mask bits that control the NVIC IRQ channel mapped to the
external interrupt controller (EXTI) so that an interrupt coming from one of the 23 lines
can be correctly acknowledged.

Hardware event selection
To configure the 23 lines as event sources, use the following procedure:
•

Configure the mask bits of the 23 event lines (EXTI_EMR)

•

Configure the Trigger selection bits of the event lines (EXTI_RTSR and EXTI_FTSR)

Software interrupt/event selection
The 23 lines can be configured as software interrupt/event lines. The following is the
procedure to generate a software interrupt.
•

Configure the mask bits of the 23 interrupt/event lines (EXTI_IMR, EXTI_EMR)

•

Set the required bit in the software interrupt register (EXTI_SWIER)

RM0090 Rev 17

381/1747
387

Interrupts and events

12.2.5

RM0090

External interrupt/event line mapping
Up to 140 GPIOs (STM32F405xx/07xx and STM32F415xx/17xx), 168 GPIOs
(STM32F42xxx and STM32F43xxx) are connected to the 16 external interrupt/event lines in
the following manner:
Figure 42. External interrupt/event GPIO mapping (STM32F405xx/07xx and
STM32F415xx/17xx)
%84);= BITS IN THE 393#&'?%84)#2 REGISTER
0!
0"
0#
0$
0%
0&
0'
0(
0)

%84)

%84);= BITS IN THE 393#&'?%84)#2 REGISTER
0!
0"
0#
0$
0%
0&
0'
0(
0)

%84)


%84);= BITS IN THE 393#&'?%84)#2 REGISTER
0!
0"
0#
0$
0%
0&
0'
0(

%84)

AI

382/1747

RM0090 Rev 17

RM0090

Interrupts and events
Figure 43. External interrupt/event GPIO mapping (STM32F42xxx and STM32F43xxx)
(;7,>@ELWVLQWKH6<6&)*B(;7,&5UHJLVWHU
3$
3%
3&
3'
3(
3)
3*
3+
3,
3-
3.

(;7,

(;7,>@ELWVLQWKH6<6&)*B(;7,&5UHJLVWHU
3$
3%
3&
3'
3(
3)
3*
3+
3,
3-
3.

(;7,

(;7,>@ELWVLQWKH6<6&)*B(;7,&5UHJLVWHU

3$
3%
3&
3'
3(
3)
3*
3+
3-

(;7,

069

The seven other EXTI lines are connected as follows:
•

EXTI line 16 is connected to the PVD output

•

EXTI line 17 is connected to the RTC Alarm event

•

EXTI line 18 is connected to the USB OTG FS Wakeup event

•

EXTI line 19 is connected to the Ethernet Wakeup event

•

EXTI line 20 is connected to the USB OTG HS (configured in FS) Wakeup event

•

EXTI line 21 is connected to the RTC Tamper and TimeStamp events

•

EXTI line 22 is connected to the RTC Wakeup event

RM0090 Rev 17

383/1747
387

Interrupts and events

12.3

RM0090

EXTI registers
Refer to Section 1.1: List of abbreviations for registers for a list of abbreviations used in
register descriptions.

12.3.1

Interrupt mask register (EXTI_IMR)
Address offset: 0x00
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

Reserved

22

21

20

19

18

17

16

MR22

MR21

MR20

MR19

MR18

MR17

MR16

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

MR15

MR14

MR13

MR12

MR11

MR10

MR9

MR8

MR7

MR6

MR5

MR4

MR3

MR2

MR1

MR0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:23 Reserved, must be kept at reset value.
Bits 22:0 MRx: Interrupt mask on line x
0: Interrupt request from line x is masked
1: Interrupt request from line x is not masked

12.3.2

Event mask register (EXTI_EMR)
Address offset: 0x04
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

Reserved

22

21

20

19

18

17

16

MR22

MR21

MR20

MR19

MR18

MR17

MR16

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

MR15

MR14

MR13

MR12

MR11

MR10

MR9

MR8

MR7

MR6

MR5

MR4

MR3

MR2

MR1

MR0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:23 Reserved, must be kept at reset value.
Bits 22:0 MRx: Event mask on line x
0: Event request from line x is masked
1: Event request from line x is not masked

384/1747

RM0090 Rev 17

RM0090

Interrupts and events

12.3.3

Rising trigger selection register (EXTI_RTSR)
Address offset: 0x08
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

Reserved

22

21

20

19

18

17

16

TR22

TR21

TR20

TR19

TR18

TR17

TR16

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

TR15

TR14

TR13

TR12

TR11

TR10

TR9

TR8

TR7

TR6

TR5

TR4

TR3

TR2

TR1

TR0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:23 Reserved, must be kept at reset value.
Bits 22:0 TRx: Rising trigger event configuration bit of line x
0: Rising trigger disabled (for Event and Interrupt) for input line
1: Rising trigger enabled (for Event and Interrupt) for input line

Note:

The external wakeup lines are edge triggered, no glitch must be generated on these lines.
If a rising edge occurs on the external interrupt line while writing to the EXTI_RTSR register,
the pending bit is be set.
Rising and falling edge triggers can be set for the same interrupt line. In this configuration,
both generate a trigger condition.

12.3.4

Falling trigger selection register (EXTI_FTSR)
Address offset: 0x0C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

Reserved

22

21

20

19

18

17

16

TR22

TR21

TR20

TR19

TR18

TR17

TR16

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

TR15

TR14

TR13

TR12

TR11

TR10

TR9

TR8

TR7

TR6

TR5

TR4

TR3

TR2

TR1

TR0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:23 Reserved, must be kept at reset value.
Bits 22:0 TRx: Falling trigger event configuration bit of line x
0: Falling trigger disabled (for Event and Interrupt) for input line
1: Falling trigger enabled (for Event and Interrupt) for input line.

Note:

The external wakeup lines are edge triggered, no glitch must be generated on these lines.
If a falling edge occurs on the external interrupt line while writing to the EXTI_FTSR register,
the pending bit is not set.
Rising and falling edge triggers can be set for the same interrupt line. In this configuration,
both generate a trigger condition.

RM0090 Rev 17

385/1747
387

Interrupts and events

12.3.5

RM0090

Software interrupt event register (EXTI_SWIER)
Address offset: 0x10
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

Reserved
15

14

13

12

11

10

9

SWIER SWIER SWIER SWIER SWIER SWIER SWIER
15
14
13
12
11
10
9
rw

rw

rw

rw

rw

22

21

20

19

18

17

16

SWIER SWIER SWIER SWIER SWIER SWIER SWIER
22
21
20
19
18
17
16

rw

rw

8

7

rw

rw

rw

rw

rw

rw

rw

6

5

4

3

2

1

0

SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER
8
7
6
5
4
3
2
1
0
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:23 Reserved, must be kept at reset value.
Bits 22:0 SWIERx: Software Interrupt on line x
If interrupt are enabled on line x in the EXTI_IMR register, writing '1' to SWIERx bit when it is
set at '0' sets the corresponding pending bit in the EXTI_PR register, thus resulting in an
interrupt request generation.
This bit is cleared by clearing the corresponding bit in EXTI_PR (by writing a 1 to the bit).

12.3.6

Pending register (EXTI_PR)
Address offset: 0x14
Reset value: undefined

31

30

29

28

27

26

25

24

23

Reserved

22

21

20

19

18

17

16

PR22

PR21

PR20

PR19

PR18

PR17

PR16

rc_w1

rc_w1

rc_w1

rc_w1

rc_w1

rc_w1

rc_w1

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

PR15

PR14

PR13

PR12

PR11

PR10

PR9

PR8

PR7

PR6

PR5

PR4

PR3

PR2

PR1

PR0

rc_w1

rc_w1

rc_w1

rc_w1

rc_w1

rc_w1

rc_w1

rc_w1

rc_w1

rc_w1

rc_w1

rc_w1

rc_w1

rc_w1

rc_w1

rc_w1

Bits 31:23 Reserved, must be kept at reset value.
Bits 22:0 PRx: Pending bit
0: No trigger request occurred
1: selected trigger request occurred
This bit is set when the selected edge event arrives on the external interrupt line.
This bit is cleared by programming it to ‘1’.

386/1747

RM0090 Rev 17

RM0090

12.3.7

Interrupts and events

EXTI register map
Table 64 gives the EXTI register map and the reset values.

Offset

0x00

Register

EXTI_IMR

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 63. External interrupt/event controller register map and reset values

Reset value

0x04

EXTI_EMR

0

EXTI_RTSR

EXTI_FTSR

EXTI_SWIER

EXTI_PR
Reset value

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

TR[22:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

TR[22:0]

Reserved
0

0

0

0

0

0

0

0

0

0

0

0

0

SWIER[22:0]

Reserved

Reset value

0x14

0

Reserved

Reset value

0x10

0

MR[22:0]
0

Reset value

0x0C

0

Reserved

Reset value

0x08

MR[22:0]

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

0

PR[22:0]

Reserved
0

0

0

0

0

0

0

0

0

0

0

0

0

Refer to Section 2.3: Memory map for the register boundary addresses.

RM0090 Rev 17

387/1747
387

Analog-to-digital converter (ADC)

13

RM0090

Analog-to-digital converter (ADC)
This section applies to the whole STM32F4xx family, unless otherwise specified.

13.1

ADC introduction
The 12-bit ADC is a successive approximation analog-to-digital converter. It has up to 19
multiplexed channels allowing it to measure signals from 16 external sources, two internal
sources, and the VBAT channel. The A/D conversion of the channels can be performed in
single, continuous, scan or discontinuous mode. The result of the ADC is stored into a leftor right-aligned 16-bit data register.
The analog watchdog feature allows the application to detect if the input voltage goes
beyond the user-defined, higher or lower thresholds.

13.2

ADC main features
•

12-bit, 10-bit, 8-bit or 6-bit configurable resolution

•

Interrupt generation at the end of conversion, end of injected conversion, and in case of
analog watchdog or overrun events

•

Single and continuous conversion modes

•

Scan mode for automatic conversion of channel 0 to channel ‘n’

•

Data alignment with in-built data coherency

•

Channel-wise programmable sampling time

•

External trigger option with configurable polarity for both regular and injected
conversions

•

Discontinuous mode

•

Dual/Triple mode (on devices with 2 ADCs or more)

•

Configurable DMA data storage in Dual/Triple ADC mode

•

Configurable delay between conversions in Dual/Triple interleaved mode

•

ADC conversion type (refer to the datasheets)

•

ADC supply requirements: 2.4 V to 3.6 V at full speed and down to 1.8 V at slower
speed

•

ADC input range: VREF– ≤VIN ≤VREF+

•

DMA request generation during regular channel conversion

Figure 44 shows the block diagram of the ADC.
Note:

388/1747

VREF–, if available (depending on package), must be tied to VSSA.

RM0090 Rev 17

RM0090

13.3

Analog-to-digital converter (ADC)

ADC functional description
Figure 44 shows a single ADC block diagram and Table 65 gives the ADC pin description.
Figure 44. Single ADC block diagram
)NTERRUPT
ENABLE BITS

&LAGS
$-! OVERRUN
%ND OF CONVERSION
%ND OF INJECTED CONVERSION
!NALOG WATCHDOG EVENT

/62

/62)%

%/#

%/#)%

*%/#

*%/#)%

!7$

!7$)%

!$# )NTERRUPT TO .6)#

!NALOG WATCHDOG
#OMPARE RESULT
(IGHER THRESHOLD  BITS

)NJECTED DATA REGISTERS
 X  BITS

6 2%&
6 2%&

2EGULAR DATA REGISTER
 BITS

6 $$!
6 33!

!NALOG
MUX

!DDRESSDATA BUS

,OWER THRESHOLD  BITS

$-! REQUEST

!$#X?).
!$#X?).

'0)/
PORTS

UP TO 
UP TO 

!$#X?).

!$##,+

)NJECTED
CHANNELS

!NALOG TO DIGITAL

2EGULAR
CHANNELS

CONVERTER

4EMP SENSOR
6 2%&).4
6 "!4
&ROM !$# PRESCALER
%843%,;= BITS

*%843%,;= BITS
4)-?#(
4)-?42'/
4)-?#(
4)-?42'/
4)-?#(
4)-?#(
4)-?#(
4)-?#(
4)-?#(
4)-?42'/
4)-?#(
4)-?42'/
4)-?#(
4)-?#(
4)-?#(

%84%.
;= BITS

*%84%.
;= BITS

3TART TRIGGER
INJECTED GROUP

3TART TRIGGER
REGULAR GROUP

4)-?#(
4)-?#(
4)-?#(
4)-?#(
4)-?#(
4)-?#(
4)-?42'/
4)-?#(
4)-?42'/
4)-?#(
4)-?#(
4)-?#(
4)-?#(
4)-?#(
4)-?42'/

%84)?
%84)?
AI

RM0090 Rev 17

389/1747
432

Analog-to-digital converter (ADC)

RM0090
Table 65. ADC pins

Name

13.3.1

Signal type

Remarks

VREF+

Input, analog reference
positive

The higher/positive reference voltage for the ADC,
1.8 V ≤VREF+ ≤VDDA

VDDA

Input, analog supply

Analog power supply equal to VDD and
2.4 V ≤VDDA ≤VDD (3.6 V) for full speed
1.8 V ≤VDDA ≤VDD (3.6 V) for reduced speed

VREF–

Input, analog reference
negative

The lower/negative reference voltage for the ADC,
VREF– = VSSA

VSSA

Input, analog supply
ground

Ground for analog power supply equal to VSS

ADCx_IN[15:0]

Analog input signals

16 analog input channels

ADC on-off control
The ADC is powered on by setting the ADON bit in the ADC_CR2 register. When the ADON
bit is set for the first time, it wakes up the ADC from the Power-down mode.
Conversion starts when either the SWSTART or the JSWSTART bit is set.
You can stop conversion and put the ADC in power down mode by clearing the ADON bit. In
this mode the ADC consumes almost no power (only a few µA).

13.3.2

ADC clock
The ADC features two clock schemes:
•

Clock for the analog circuitry: ADCCLK, common to all ADCs
This clock is generated from the APB2 clock divided by a programmable prescaler that
allows the ADC to work at fPCLK2/2, /4, /6 or /8. Refer to the datasheets for the
maximum value of ADCCLK.

•

Clock for the digital interface (used for registers read/write access)
This clock is equal to the APB2 clock. The digital interface clock can be
enabled/disabled individually for each ADC through the RCC APB2 peripheral clock
enable register (RCC_APB2ENR).

13.3.3

Channel selection
There are 16 multiplexed channels. It is possible to organize the conversions in two groups:
regular and injected. A group consists of a sequence of conversions that can be done on
any channel and in any order. For instance, it is possible to implement the conversion
sequence in the following order: ADC_IN3, ADC_IN8, ADC_IN2, ADC_IN2, ADC_IN0,
ADC_IN2, ADC_IN2, ADC_IN15.

390/1747

•

A regular group is composed of up to 16 conversions. The regular channels and their
order in the conversion sequence must be selected in the ADC_SQRx registers. The
total number of conversions in the regular group must be written in the L[3:0] bits in the
ADC_SQR1 register.

•

An injected group is composed of up to 4 conversions. The injected channels and
their order in the conversion sequence must be selected in the ADC_JSQR register.

RM0090 Rev 17

RM0090

Analog-to-digital converter (ADC)
The total number of conversions in the injected group must be written in the L[1:0] bits
in the ADC_JSQR register.
If the ADC_SQRx or ADC_JSQR registers are modified during a conversion, the current
conversion is reset and a new start pulse is sent to the ADC to convert the newly chosen
group.

Temperature sensor, VREFINT and VBAT internal channels
•

For the STM32F40x and STM32F41x devices, the temperature sensor is internally
connected to channel ADC1_IN16.
The internal reference voltage VREFINT is connected to ADC1_IN17.

•

For the STM32F42x and STM32F43x devices, the temperature sensor is internally
connected to ADC1_IN18 channel which is shared with VBAT. Only one conversion,
temperature sensor or VBAT, must be selected at a time. When the temperature sensor
and VBAT conversion are set simultaneously, only the VBAT conversion is performed.
The internal reference voltage VREFINT is connected to ADC1_IN17.

The VBAT channel (connected to channel ADC1_IN18) can also be converted as an injected
or regular channel.
Note:

The temperature sensor, VREFINT and the VBAT channel are available only on the master
ADC1 peripheral.

13.3.4

Single conversion mode
In Single conversion mode the ADC does one conversion. This mode is started with the
CONT bit at 0 by either:
•

setting the SWSTART bit in the ADC_CR2 register (for a regular channel only)

•

setting the JSWSTART bit (for an injected channel)

•

external trigger (for a regular or injected channel)

Once the conversion of the selected channel is complete:
•

•

If a regular channel was converted:
–

The converted data are stored into the 16-bit ADC_DR register

–

The EOC (end of conversion) flag is set

–

An interrupt is generated if the EOCIE bit is set

If an injected channel was converted:
–

The converted data are stored into the 16-bit ADC_JDR1 register

–

The JEOC (end of conversion injected) flag is set

–

An interrupt is generated if the JEOCIE bit is set

Then the ADC stops.

RM0090 Rev 17

391/1747
432

Analog-to-digital converter (ADC)

13.3.5

RM0090

Continuous conversion mode
In continuous conversion mode, the ADC starts a new conversion as soon as it finishes one.
This mode is started with the CONT bit at 1 either by external trigger or by setting the
SWSTRT bit in the ADC_CR2 register (for regular channels only).
After each conversion:
•

If a regular group of channels was converted:
–

The last converted data are stored into the 16-bit ADC_DR register

–

The EOC (end of conversion) flag is set

–

An interrupt is generated if the EOCIE bit is set

Note:

Injected channels cannot be converted continuously. The only exception is when an injected
channel is configured to be converted automatically after regular channels in continuous
mode (using JAUTO bit), refer to Auto-injection section).

13.3.6

Timing diagram
As shown in Figure 45, the ADC needs a stabilization time of tSTAB before it starts
converting accurately. After the start of the ADC conversion and after 15 clock cycles, the
EOC flag is set and the 16-bit ADC data register contains the result of the conversion.
Figure 45. Timing diagram

!$#?#,+

!$/.

3734!24
*3734!24
3TART ST CONVERSION

!$# CONVERSION

!$#
T34!"

3TART NEXT CONVERSION

.EXT !$# CONVERSION

#ONVERSION TIME
TOTAL CONV TIME

%/#
3OFTWARE CLEARS THE %/# BIT
AIB

13.3.7

Analog watchdog
The AWD analog watchdog status bit is set if the analog voltage converted by the ADC is
below a lower threshold or above a higher threshold. These thresholds are programmed in
the 12 least significant bits of the ADC_HTR and ADC_LTR 16-bit registers. An interrupt can
be enabled by using the AWDIE bit in the ADC_CR1 register.
The threshold value is independent of the alignment selected by the ALIGN bit in the
ADC_CR2 register. The analog voltage is compared to the lower and higher thresholds
before alignment.
Table 66 shows how the ADC_CR1 register should be configured to enable the analog
watchdog on one or more channels.

392/1747

RM0090 Rev 17

RM0090

Analog-to-digital converter (ADC)
Figure 46. Analog watchdog’s guarded area
!NALOG VOLTAGE
(42

(IGHER THRESHOLD
'UARDED AREA
,OWER THRESHOLD

,42
AI

Table 66. Analog watchdog channel selection
Channels guarded by the analog
watchdog

ADC_CR1 register control bits (x = don’t care)
AWDSGL bit

AWDEN bit

JAWDEN bit

None

x

0

0

All injected channels

0

0

1

All regular channels

0

1

0

All regular and injected channels

0

1

1

Single(1)

1

0

1

1

1

0

1

1

1

injected channel

Single(1) regular channel
Single

(1)

regular or injected channel

1. Selected by the AWDCH[4:0] bits

13.3.8

Scan mode
This mode is used to scan a group of analog channels.
The Scan mode is selected by setting the SCAN bit in the ADC_CR1 register. Once this bit
has been set, the ADC scans all the channels selected in the ADC_SQRx registers (for
regular channels) or in the ADC_JSQR register (for injected channels). A single conversion
is performed for each channel of the group. After each end of conversion, the next channel
in the group is converted automatically. If the CONT bit is set, regular channel conversion
does not stop at the last selected channel in the group but continues again from the first
selected channel.
If the DMA bit is set, the direct memory access (DMA) controller is used to transfer the data
converted from the regular group of channels (stored in the ADC_DR register) to SRAM
after each regular channel conversion.
The EOC bit is set in the ADC_SR register:
•

At the end of each regular group sequence if the EOCS bit is cleared to 0

•

At the end of each regular channel conversion if the EOCS bit is set to 1

The data converted from an injected channel are always stored into the ADC_JDRx
registers.

RM0090 Rev 17

393/1747
432

Analog-to-digital converter (ADC)

13.3.9

RM0090

Injected channel management
Triggered injection
To use triggered injection, the JAUTO bit must be cleared in the ADC_CR1 register.

Note:

1.

Start the conversion of a group of regular channels either by external trigger or by
setting the SWSTART bit in the ADC_CR2 register.

2.

If an external injected trigger occurs or if the JSWSTART bit is set during the
conversion of a regular group of channels, the current conversion is reset and the
injected channel sequence switches to Scan-once mode.

3.

Then, the regular conversion of the regular group of channels is resumed from the last
interrupted regular conversion.
If a regular event occurs during an injected conversion, the injected conversion is not
interrupted but the regular sequence is executed at the end of the injected sequence.
Figure 47 shows the corresponding timing diagram.

When using triggered injection, one must ensure that the interval between trigger events is
longer than the injection sequence. For instance, if the sequence length is 30 ADC clock
cycles (that is two conversions with a sampling time of 3 clock periods), the minimum
interval between triggers must be 31 ADC clock cycles.

Auto-injection
If the JAUTO bit is set, then the channels in the injected group are automatically converted
after the regular group of channels. This can be used to convert a sequence of up to 20
conversions programmed in the ADC_SQRx and ADC_JSQR registers.
In this mode, external trigger on injected channels must be disabled.
If the CONT bit is also set in addition to the JAUTO bit, regular channels followed by injected
channels are continuously converted.
Note:

It is not possible to use both the auto-injected and discontinuous modes simultaneously.
Figure 47. Injected conversion latency
!$##,+

)NJECTION EVENT

2ESET !$#

3/#

MAX LATENCY 

AI

1. The maximum latency value can be found in the electrical characteristics of the STM32F40x and
STM32F41x datasheets.

394/1747

RM0090 Rev 17

RM0090

13.3.10

Analog-to-digital converter (ADC)

Discontinuous mode
Regular group
This mode is enabled by setting the DISCEN bit in the ADC_CR1 register. It can be used to
convert a short sequence of n conversions (n ≤8) that is part of the sequence of conversions
selected in the ADC_SQRx registers. The value of n is specified by writing to the
DISCNUM[2:0] bits in the ADC_CR1 register.
When an external trigger occurs, it starts the next n conversions selected in the ADC_SQRx
registers until all the conversions in the sequence are done. The total sequence length is
defined by the L[3:0] bits in the ADC_SQR1 register.
Example:

Note:

•

n = 3, channels to be converted = 0, 1, 2, 3, 6, 7, 9, 10

•

1st trigger: sequence converted 0, 1, 2. An EOC event is generated at each
conversion.

•

2nd trigger: sequence converted 3, 6, 7. An EOC event is generated at each
conversion

•

3rd trigger: sequence converted 9, 10.An EOC event is generated at each conversion

•

4th trigger: sequence converted 0, 1, 2. An EOC event is generated at each conversion

When a regular group is converted in discontinuous mode, no rollover occurs.
When all subgroups are converted, the next trigger starts the conversion of the first
subgroup. In the example above, the 4th trigger reconverts the channels 0, 1 and 2 in the
1st subgroup.

Injected group
This mode is enabled by setting the JDISCEN bit in the ADC_CR1 register. It can be used to
convert the sequence selected in the ADC_JSQR register, channel by channel, after an
external trigger event.
When an external trigger occurs, it starts the next channel conversions selected in the
ADC_JSQR registers until all the conversions in the sequence are done. The total sequence
length is defined by the JL[1:0] bits in the ADC_JSQR register.
Example:
n = 1, channels to be converted = 1, 2, 3
1st trigger: channel 1 converted
2nd trigger: channel 2 converted
3rd trigger: channel 3 converted and JEOC event generated
4th trigger: channel 1
Note:

When all injected channels are converted, the next trigger starts the conversion of the first
injected channel. In the example above, the 4th trigger reconverts the 1st injected channel
1.
It is not possible to use both the auto-injected and discontinuous modes simultaneously.
Discontinuous mode must not be set for regular and injected groups at the same time.
Discontinuous mode must be enabled only for the conversion of one group.

RM0090 Rev 17

395/1747
432

Analog-to-digital converter (ADC)

13.4

RM0090

Data alignment
The ALIGN bit in the ADC_CR2 register selects the alignment of the data stored after
conversion. Data can be right- or left-aligned as shown in Figure 48 and Figure 49.
The converted data value from the injected group of channels is decreased by the userdefined offset written in the ADC_JOFRx registers so the result can be a negative value.
The SEXT bit represents the extended sign value.
For channels in a regular group, no offset is subtracted so only twelve bits are significant.
Figure 48. Right alignment of 12-bit data
)NJECTED GROUP
3%84 3%84 3%84 3%84

$

$

$

$

$

$

$

$

$

$

$

$

$

$

$

$

$

$

$

$

$

$

$

$

2EGULAR GROUP








AI

Figure 49. Left alignment of 12-bit data
)NJECTED GROUP
3%84

$

$

$

$

$

$

$

$

$

$

$

$

$

$

$

$

$

$

$















2EGULAR GROUP
$

$

$

$

$

AI

Special case: when left-aligned, the data are aligned on a half-word basis except when the
resolution is set to 6-bit. in that case, the data are aligned on a byte basis as shown in
Figure 50.
Figure 50. Left alignment of 6-bit data
)NJECTED GROUP
3%84 3%84 3%84

3%84

3%84 3%84 3%84

3%84 3%84

$

$

$

$

$

$

$

$



$





2EGULAR GROUP
















$

$

$

AI

396/1747

RM0090 Rev 17

RM0090

13.5

Analog-to-digital converter (ADC)

Channel-wise programmable sampling time
The ADC samples the input voltage for a number of ADCCLK cycles that can be modified
using the SMP[2:0] bits in the ADC_SMPR1 and ADC_SMPR2 registers. Each channel can
be sampled with a different sampling time.
The total conversion time is calculated as follows:
Tconv = Sampling time + 12 cycles
Example:
With ADCCLK = 30 MHz and sampling time = 3 cycles:
Tconv = 3 + 12 = 15 cycles = 0.5 µs with APB2 at 60 MHz

13.6

Conversion on external trigger and trigger polarity
Conversion can be triggered by an external event (e.g. timer capture, EXTI line). If the
EXTEN[1:0] control bits (for a regular conversion) or JEXTEN[1:0] bits (for an injected
conversion) are different from “0b00”, then external events are able to trigger a conversion
with the selected polarity. Table 67 provides the correspondence between the EXTEN[1:0]
and JEXTEN[1:0] values and the trigger polarity.
Table 67. Configuring the trigger polarity
Source

Note:

EXTEN[1:0] / JEXTEN[1:0]

Trigger detection disabled

00

Detection on the rising edge

01

Detection on the falling edge

10

Detection on both the rising and falling edges

11

The polarity of the external trigger can be changed on the fly.
The EXTSEL[3:0] and JEXTSEL[3:0] control bits are used to select which out of 16 possible
events can trigger conversion for the regular and injected groups.
Table 68 gives the possible external trigger for regular conversion.

RM0090 Rev 17

397/1747
432

Analog-to-digital converter (ADC)

RM0090

Table 68. External trigger for regular channels
Source

Type

EXTSEL[3:0]

TIM1_CH1 event

0000

TIM1_CH2 event

0001

TIM1_CH3 event

0010

TIM2_CH2 event

0011

TIM2_CH3 event

0100

TIM2_CH4 event

0101

TIM2_TRGO event
TIM3_CH1 event

0110
Internal signal from on-chip
timers

0111

TIM3_TRGO event

1000

TIM4_CH4 event

1001

TIM5_CH1 event

1010

TIM5_CH2 event

1011

TIM5_CH3 event

1100

TIM8_CH1 event

1101

TIM8_TRGO event

1110

EXTI line11

External pin

1111

Table 69 gives the possible external trigger for injected conversion.

398/1747

RM0090 Rev 17

RM0090

Analog-to-digital converter (ADC)
Table 69. External trigger for injected channels
Source

Connection type

JEXTSEL[3:0]

TIM1_CH4 event

0000

TIM1_TRGO event

0001

TIM2_CH1 event

0010

TIM2_TRGO event

0011

TIM3_CH2 event

0100

TIM3_CH4 event

0101

TIM4_CH1 event
TIM4_CH2 event

0110
Internal signal from on-chip
timers

0111

TIM4_CH3 event

1000

TIM4_TRGO event

1001

TIM5_CH4 event

1010

TIM5_TRGO event

1011

TIM8_CH2 event

1100

TIM8_CH3 event

1101

TIM8_CH4 event

1110

EXTI line15

External pin

1111

Software source trigger events can be generated by setting SWSTART (for regular
conversion) or JSWSTART (for injected conversion) in ADC_CR2.
A regular group conversion can be interrupted by an injected trigger.
Note:

The trigger selection can be changed on the fly. However, when the selection changes,
there is a time frame of 1 APB clock cycle during which the trigger detection is disabled.
This is to avoid spurious detection during transitions.

13.7

Fast conversion mode
It is possible to perform faster conversion by reducing the ADC resolution. The RES bits are
used to select the number of bits available in the data register. The minimum conversion
time for each resolution is then as follows:
•

12 bits: 3 + 12 = 15 ADCCLK cycles

•

10 bits: 3 + 10 = 13 ADCCLK cycles

•

8 bits: 3 + 8 = 11 ADCCLK cycles

•

6 bits: 3 + 6 = 9 ADCCLK cycles

RM0090 Rev 17

399/1747
432

Analog-to-digital converter (ADC)

13.8

Data management

13.8.1

Using the DMA

RM0090

Since converted regular channel values are stored into a unique data register, it is useful to
use DMA for conversion of more than one regular channel. This avoids the loss of the data
already stored in the ADC_DR register.
When the DMA mode is enabled (DMA bit set to 1 in the ADC_CR2 register), after each
conversion of a regular channel, a DMA request is generated. This allows the transfer of the
converted data from the ADC_DR register to the destination location selected by the
software.
Despite this, if data are lost (overrun), the OVR bit in the ADC_SR register is set and an
interrupt is generated (if the OVRIE enable bit is set). DMA transfers are then disabled and
DMA requests are no longer accepted. In this case, if a DMA request is made, the regular
conversion in progress is aborted and further regular triggers are ignored. It is then
necessary to clear the OVR flag and the DMAEN bit in the used DMA stream, and to reinitialize both the DMA and the ADC to have the wanted converted channel data transferred
to the right memory location. Only then can the conversion be resumed and the data
transfer, enabled again. Injected channel conversions are not impacted by overrun errors.
When OVR = 1 in DMA mode, the DMA requests are blocked after the last valid data have
been transferred, which means that all the data transferred to the RAM can be considered
as valid.
At the end of the last DMA transfer (number of transfers configured in the DMA controller’s
DMA_SxNTR register):
•

No new DMA request is issued to the DMA controller if the DDS bit is cleared to 0 in the
ADC_CR2 register (this avoids generating an overrun error). However the DMA bit is
not cleared by hardware. It must be written to 0, then to 1 to start a new transfer.

•

Requests can continue to be generated if the DDS bit is set to 1. This allows
configuring the DMA in double-buffer circular mode.

To recover the ADC from OVR state when the DMA is used, follow the steps below:

13.8.2

1.

Reinitialize the DMA (adjust destination address and NDTR counter)

2.

Clear the ADC OVR bit in ADC_SR register

3.

Trigger the ADC to start the conversion.

Managing a sequence of conversions without using the DMA
If the conversions are slow enough, the conversion sequence can be handled by the
software. In this case the EOCS bit must be set in the ADC_CR2 register for the EOC status
bit to be set at the end of each conversion, and not only at the end of the sequence. When
EOCS = 1, overrun detection is automatically enabled. Thus, each time a conversion is
complete, EOC is set and the ADC_DR register can be read. The overrun management is
the same as when the DMA is used.
To recover the ADC from OVR state when the EOCS is set, follow the steps below:

400/1747

1.

Clear the ADC OVR bit in ADC_SR register

2.

Trigger the ADC to start the conversion.

RM0090 Rev 17

RM0090

13.8.3

Analog-to-digital converter (ADC)

Conversions without DMA and without overrun detection
It may be useful to let the ADC convert one or more channels without reading the data each
time (if there is an analog watchdog for instance). For that, the DMA must be disabled
(DMA = 0) and the EOC bit must be set at the end of a sequence only (EOCS = 0). In this
configuration, overrun detection is disabled.

13.9

Multi ADC mode
In devices with two ADCs or more, the Dual (with two ADCs) and Triple (with three ADCs)
ADC modes can be used (see Figure 51).
In multi ADC mode, the start of conversion is triggered alternately or simultaneously by the
ADC1 master to the ADC2 and ADC3 slaves, depending on the mode selected by the
MULTI[4:0] bits in the ADC_CCR register.

Note:

In multi ADC mode, when configuring conversion trigger by an external event, the
application must set trigger by the master only and disable trigger by slaves to prevent
spurious triggers that would start unwanted slave conversions.
The four possible modes below are implemented:
•

Injected simultaneous mode

•

Regular simultaneous mode

•

Interleaved mode

•

Alternate trigger mode

It is also possible to use the previous modes combined in the following ways:

Note:

•

Injected simultaneous mode + Regular simultaneous mode

•

Regular simultaneous mode + Alternate trigger mode

In multi ADC mode, the converted data can be read on the multi-mode data register
(ADC_CDR). The status bits can be read in the multi-mode status register (ADC_CSR).

RM0090 Rev 17

401/1747
432

Analog-to-digital converter (ADC)

RM0090
Figure 51. Multi ADC block diagram(1)
2EGULAR DATA REGISTER
BITS
BITS

)NJECTED DATA REGISTERS
 X  BITS

2EGULAR
CHANNELS

!$# 3LAVE

)NJECTED
CHANNELS

2EGULAR DATA REGISTER

BITS
BITS
)NJECTED DATA REGISTERS
 X  BITS

2EGULAR
CHANNELS

INTERNAL TRIGGERS
#OMMON REGULAR DATA REGISTER
 BITS 

$UAL4RIPLE
MODE CONTROL

!DDRESSDATA BUS

!$# 3LAVE

)NJECTED
CHANNELS

#OMMON PART

2EGULAR DATA REGISTER
 BITS
)NJECTED DATA REGISTERS
 X  BITS

!$#X?).
!$#X?).

'0)/
0ORTS

2EGULAR
CHANNELS

!$#X?).

)NJECTED
CHANNELS

4EMP SENSOR
62%&).4
6"!4
%84)?

!$# -ASTER
3TART TRIGGER MUX
REGULAR GROUP

%84)?
3TART TRIGGER MUX
INJECTED GROUP

AI

1. Although external triggers are present on ADC2 and ADC3 they are not shown in this diagram.
2. In the Dual ADC mode, the ADC3 slave part is not present.
3. In Triple ADC mode, the ADC common data register (ADC_CDR) contains the ADC1, ADC2 and ADC3’s
regular converted data. All 32 register bits are used according to a selected storage order.
In Dual ADC mode, the ADC common data register (ADC_CDR) contains both the ADC1 and ADC2’s
regular converted data. All 32 register bits are used.

402/1747

RM0090 Rev 17

RM0090

Analog-to-digital converter (ADC)
•

DMA requests in Multi ADC mode:
In Multi ADC mode the DMA may be configured to transfer converted data in three
different modes. In all cases, the DMA streams to use are those connected to the ADC:
–

DMA mode 1: On each DMA request (one data item is available), a half-word
representing an ADC-converted data item is transferred.
In Triple ADC mode, ADC1 data are transferred on the first request, ADC2 data
are transferred on the second request and ADC3 data are transferred on the third
request; the sequence is repeated. So the DMA first transfers ADC1 data followed
by ADC2 data followed by ADC3 data and so on.
DMA mode 1 is used in regular simultaneous triple mode only.
Example:
Regular simultaneous triple mode: 3 consecutive DMA requests are generated
(one for each converted data item)
1st request: ADC_CDR[31:0] = ADC1_DR[15:0]
2nd request: ADC_CDR[31:0] = ADC2_DR[15:0]
3rd request: ADC_CDR[31:0] = ADC3_DR[15:0]
4th request: ADC_CDR[31:0] = ADC1_DR[15:0]

–

DMA mode 2: On each DMA request (two data items are available) two halfwords representing two ADC-converted data items are transferred as a word.
In Dual ADC mode, both ADC2 and ADC1 data are transferred on the first request
(ADC2 data take the upper half-word and ADC1 data take the lower half-word) and
so on.
In Triple ADC mode, three DMA requests are generated. On the first request, both
ADC2 and ADC1 data are transferred (ADC2 data take the upper half-word and
ADC1 data take the lower half-word). On the second request, both ADC1 and
ADC3 data are transferred (ADC1 data take the upper half-word and ADC3 data
take the lower half-word).On the third request, both ADC3 and ADC2 data are
transferred (ADC3 data take the upper half-word and ADC2 data take the lower
half-word) and so on.
DMA mode 2 is used in interleaved mode and in regular simultaneous mode (for
Dual ADC mode only).
Example:

a)

Interleaved dual mode: a DMA request is generated each time 2 data items are
available:
1st request: ADC_CDR[31:0] = ADC2_DR[15:0] | ADC1_DR[15:0]
2nd request: ADC_CDR[31:0] = ADC2_DR[15:0] | ADC1_DR[15:0]

b)

Interleaved triple mode: a DMA request is generated each time 2 data items are
available
1st request: ADC_CDR[31:0] = ADC2_DR[15:0] | ADC1_DR[15:0]
2nd request: ADC_CDR[31:0] = ADC1_DR[15:0] | ADC3_DR[15:0]
3rd request: ADC_CDR[31:0] = ADC3_DR[15:0] | ADC2_DR[15:0]
4th request: ADC_CDR[31:0] = ADC2_DR[15:0] | ADC1_DR[15:0]

–

DMA mode 3: This mode is similar to the DMA mode 2. The only differences are
that the on each DMA request (two data items are available) two bytes

RM0090 Rev 17

403/1747
432

Analog-to-digital converter (ADC)

RM0090

representing two ADC converted data items are transferred as a half-word. The
data transfer order is similar to that of the DMA mode 2.
DMA mode 3 is used in interleaved mode in 6-bit and 8-bit resolutions (dual and
triple mode).
Example:
a)

Interleaved dual mode: a DMA request is generated each time 2 data items are
available
1st request: ADC_CDR[15:0] = ADC2_DR[7:0] | ADC1_DR[7:0]
2nd request: ADC_CDR[15:0] = ADC2_DR[7:0] | ADC1_DR[7:0]

b)

Interleaved triple mode: a DMA request is generated each time 2 data items are
available
1st request: ADC_CDR[15:0] = ADC2_DR[7:0] | ADC1_DR[7:0]
2nd request: ADC_CDR[15:0] = ADC1_DR[7:0] | ADC3_DR[7:0]
3rd request: ADC_CDR[15:0] = ADC3_DR[7:0] | ADC2_DR[7:0]
4th request: ADC_CDR[15:0] = ADC2_DR[7:0] | ADC1_DR[7:0]

Overrun detection: If an overrun is detected on one of the concerned ADCs (ADC1 and
ADC2 in dual and triple modes, ADC3 in triple mode only), the DMA requests are no longer
issued to ensure that all the data transferred to the RAM are valid. It may happen that the
EOC bit corresponding to one ADC remains set because the data register of this ADC
contains valid data.

13.9.1

Injected simultaneous mode
This mode converts an injected group of channels. The external trigger source comes from
the injected group multiplexer of ADC1 (selected by the JEXTSEL[3:0] bits in the ADC1_CR2
register). A simultaneous trigger is provided to ADC2 and ADC3.

Note:

Do not convert the same channel on the two/three ADCs (no overlapping sampling times for
the two/three ADCs when converting the same channel).
In simultaneous mode, one must convert sequences with the same length or ensure that the
interval between triggers is longer than the longer of the 2 sequences (Dual ADC mode) /3
sequences (Triple ADC mode). Otherwise, the ADC with the shortest sequence may restart
while the ADC with the longest sequence is completing the previous conversions.
Regular conversions can be performed on one or all ADCs. In that case, they are
independent of each other and are interrupted when an injected event occurs. They are
resumed at the end of the injected conversion group.

404/1747

RM0090 Rev 17

RM0090

Analog-to-digital converter (ADC)

Dual ADC mode
At the end of conversion event on ADC1 or ADC2:
•

The converted data are stored into the ADC_JDRx registers of each ADC interface.

•

A JEOC interrupt is generated (if enabled on one of the two ADC interfaces) when the
ADC1/ADC2’s injected channels have all been converted.
Figure 52. Injected simultaneous mode on 4 channels: dual ADC mode


!$#

#(

#(

#(

#(

!$#

#(

#(

#(

#( 

#(
#(
%ND OF CONVERSION ON !$# AND !$#

4RIGGER
3AMPLING
#ONVERSION

AI

Triple ADC mode
At the end of conversion event on ADC1, ADC2 or ADC3:
•

The converted data are stored into the ADC_JDRx registers of each ADC interface.

•

A JEOC interrupt is generated (if enabled on one of the three ADC interfaces) when the
ADC1/ADC2/ADC3’s injected channels have all been converted.
Figure 53. Injected simultaneous mode on 4 channels: triple ADC mode
!$#

#(

#(



#(

#(

#( 
#( 

!$#

#(

#(

#(

!$#

#(

#(

#(

#(
#(
#(

%ND OF CONVERSION ON !$# !$# AND !$#

4RIGGER
3AMPLING
#ONVERSION

13.9.2

AI

Regular simultaneous mode
This mode is performed on a regular group of channels. The external trigger source comes
from the regular group multiplexer of ADC1 (selected by the EXTSEL[3:0] bits in the
ADC1_CR2 register). A simultaneous trigger is provided to ADC2 and ADC3.

Note:

Do not convert the same channel on the two/three ADCs (no overlapping sampling times for
the two/three ADCs when converting the same channel).
In regular simultaneous mode, one must convert sequences with the same length or ensure
that the interval between triggers is longer than the long conversion time of the 2 sequences
(Dual ADC mode) /3 sequences (Triple ADC mode). Otherwise, the ADC with the shortest
sequence may restart while the ADC with the longest sequence is completing the previous
conversions.
Injected conversions must be disabled.

RM0090 Rev 17

405/1747
432

Analog-to-digital converter (ADC)

RM0090

Dual ADC mode
At the end of conversion event on ADC1 or ADC2:
•

A 32-bit DMA transfer request is generated (if DMA[1:0] bits in the ADC_CCR register
are equal to 0b10). This request transfers the ADC2 converted data stored in the upper
half-word of the ADC_CDR 32-bit register to the SRAM and then the ADC1 converted
data stored in the lower half-word of ADC_CCR to the SRAM.

•

An EOC interrupt is generated (if enabled on one of the two ADC interfaces) when the
ADC1/ADC2’s regular channels have all been converted.
Figure 54. Regular simultaneous mode on 16 channels: dual ADC mode


!$#

#(

#(

#(

#(

!$#

#(

#(

#(

#( 

#(
#(
%ND OF CONVERSION ON !$# AND !$#

4RIGGER
3AMPLING
#ONVERSION

AI

Triple ADC mode
At the end of conversion event on ADC1, ADC2 or ADC3:
•

Three 32-bit DMA transfer requests are generated (if DMA[1:0] bits in the ADC_CCR
register are equal to 0b01). Three transfers then take place from the ADC_CDR 32-bit
register to SRAM: first the ADC1 converted data, then the ADC2 converted data and
finally the ADC3 converted data. The process is repeated for each new three
conversions.

•

An EOC interrupt is generated (if enabled on one of the three ADC interfaces) when the
ADC1/ADC2/ADC3’s regular channels are have all been converted.
Figure 55. Regular simultaneous mode on 16 channels: triple ADC mode


!$#

#(

#(

#(

#(

!$#

#(

#(

#(

!$#

#(

#(

#(

#( 
#( 

#(
#(
#(

%ND OF CONVERSION ON !$# !$# AND !$#

4RIGGER
3AMPLING
#ONVERSION

406/1747

AI

RM0090 Rev 17

RM0090

13.9.3

Analog-to-digital converter (ADC)

Interleaved mode
This mode can be started only on a regular group (usually one channel). The external
trigger source comes from the regular channel multiplexer of ADC1.

Dual ADC mode
After an external trigger occurs:
•

ADC1 starts immediately

•

ADC2 starts after a delay of several ADC clock cycles

The minimum delay which separates 2 conversions in interleaved mode is configured in the
DELAY bits in the ADC_CCR register. However, an ADC cannot start a conversion if the
complementary ADC is still sampling its input (only one ADC can sample the input signal at
a given time). In this case, the delay becomes the sampling time + 2 ADC clock cycles. For
instance, if DELAY = 5 clock cycles and the sampling takes 15 clock cycles on both ADCs,
then 17 clock cycles will separate conversions on ADC1 and ADC2).
If the CONT bit is set on both ADC1 and ADC2, the selected regular channels of both ADCs
are continuously converted.
Note:

If the conversion sequence is interrupted (for instance when DMA end of transfer occurs),
the multi-ADC sequencer must be reset by configuring it in independent mode first (bits
DUAL[4:0] = 00000) before reprogramming the interleaved mode.
After an EOC interrupt is generated by ADC2 (if enabled through the EOCIE bit) a 32-bit
DMA transfer request is generated (if the DMA[1:0] bits in ADC_CCR are equal to 0b10).
This request first transfers the ADC2 converted data stored in the upper half-word of the
ADC_CDR 32-bit register into SRAM, then the ADC1 converted data stored in the register’s
lower half-word into SRAM.
Figure 56. Interleaved mode on 1 channel in continuous conversion mode: dual ADC
mode
%ND OF CONVERSION ON !$#


#(

!$#
!$#
4RIGGER

#ONVERSION

#(


#(

3AMPLING

#(

%ND OF CONVERSION ON !$#
 !$##,+
CYCLES

AI

Triple ADC mode
After an external trigger occurs:
•

ADC1 starts immediately and

•

ADC2 starts after a delay of several ADC clock cycles

•

ADC3 starts after a delay of several ADC clock cycles referred to the ADC2 conversion

The minimum delay which separates 2 conversions in interleaved mode is configured in the
DELAY bits in the ADC_CCR register. However, an ADC cannot start a conversion if the
complementary ADC is still sampling its input (only one ADC can sample the input signal at

RM0090 Rev 17

407/1747
432

Analog-to-digital converter (ADC)

RM0090

a given time). In this case, the delay becomes the sampling time + 2 ADC clock cycles. For
instance, if DELAY = 5 clock cycles and the sampling takes 15 clock cycles on the three
ADCs, then 17 clock cycles will separate the conversions on ADC1, ADC2 and ADC3).
If the CONT bit is set on ADC1, ADC2 and ADC3, the selected regular channels of all ADCs
are continuously converted.
Note:

If the conversion sequence is interrupted (for instance when DMA end of transfer occurs),
the multi-ADC sequencer must be reset by configuring it in independent mode first (bits
DUAL[4:0] = 00000) before reprogramming the interleaved mode.
In this mode a DMA request is generated each time 2 data items are available, (if the
DMA[1:0] bits in the ADC_CCR register are equal to 0b10). The request first transfers the
first converted data stored in the lower half-word of the ADC_CDR 32-bit register to SRAM,
then it transfers the second converted data stored in ADC_CDR’s upper half-word to SRAM.
The sequence is the following:
•

1st request: ADC_CDR[31:0] = ADC2_DR[15:0] | ADC1_DR[15:0]

•

2nd request: ADC_CDR[31:0] = ADC1_DR[15:0] | ADC3_DR[15:0]

•

3rd request: ADC_CDR[31:0] = ADC3_DR[15:0] | ADC2_DR[15:0]

•

4th request: ADC_CDR[31:0] = ADC2_DR[15:0] | ADC1_DR[15:0], ...

Figure 57. Interleaved mode on 1 channel in continuous conversion mode: triple ADC
mode
%ND OF CONVERSION ON !$#
$-! REQUEST EVERY  CONVERSIONS
!$#
!$#

#(
#(

!$#



#(
#(

#(

#(

#(




#(
#(

3AMPLING
#ONVERSION

%ND OF CONVERSION ON !$#

4RIGGER

%ND OF CONVERSION ON !$#
 !$##,+
CYCLES

13.9.4

AI

Alternate trigger mode
This mode can be started only on an injected group. The source of external trigger comes
from the injected group multiplexer of ADC1.

Note:

408/1747

Regular conversions can be enabled on one or all ADCs. In this case the regular
conversions are independent of each other. A regular conversion is interrupted when the

RM0090 Rev 17

RM0090

Analog-to-digital converter (ADC)
ADC has to perform an injected conversion. It is resumed when the injected conversion is
finished.
If the conversion sequence is interrupted (for instance when DMA end of transfer occurs),
the multi-ADC sequencer must be reset by configuring it in independent mode first (bits
DUAL[4:0] = 00000) before reprogramming the interleaved mode.
The time interval between 2 trigger events must be greater than or equal to 1 ADC clock
period. The minimum time interval between 2 trigger events that start conversions on the
same ADC is the same as in the single ADC mode.

Dual ADC mode
•

When the 1st trigger occurs, all injected ADC1 channels in the group are converted

•

When the 2nd trigger occurs, all injected ADC2 channels in the group are converted

•

and so on

A JEOC interrupt, if enabled, is generated after all injected ADC1 channels in the group
have been converted.
A JEOC interrupt, if enabled, is generated after all injected ADC2 channels in the group
have been converted.
If another external trigger occurs after all injected channels in the group have been
converted then the alternate trigger process restarts by converting the injected ADC1
channels in the group.
Figure 58. Alternate trigger: injected group of each ADC
ST TRIGGER

%/# *%/#
ON !$#

RD TRIGGER

N TH TRIGGER

%/# *%/#
ON !$#

!$#



!$#

ND TRIGGER

%/# *%/#
ON !$#
TH TRIGGER

%/# *%/#
ON !$#

3AMPLING
N  TH TRIGGER

#ONVERSION
AI

If the injected discontinuous mode is enabled for both ADC1 and ADC2:
•

When the 1st trigger occurs, the first injected ADC1 channel is converted.

•

When the 2nd trigger occurs, the first injected ADC2 channel are converted

•

and so on

A JEOC interrupt, if enabled, is generated after all injected ADC1 channels in the group
have been converted.
A JEOC interrupt, if enabled, is generated after all injected ADC2 channels in the group
have been converted.
If another external trigger occurs after all injected channels in the group have been
converted then the alternate trigger process restarts.

RM0090 Rev 17

409/1747
432

Analog-to-digital converter (ADC)

RM0090

Figure 59. Alternate trigger: 4 injected channels (each ADC) in discontinuous mode
ST TRIGGER

RD TRIGGER

TH TRIGGER

TH TRIGGER
*%/# ON !$#

3AMPLING
#ONVERSION

!$#
!$#
*%/# ON !$#
ND TRIGGER

TH TRIGGER

TH TRIGGER

TH TRIGGER

AI

Triple ADC mode
•

When the 1st trigger occurs, all injected ADC1 channels in the group are converted.

•

When the 2nd trigger occurs, all injected ADC2 channels in the group are converted.

•

When the 3rd trigger occurs, all injected ADC3 channels in the group are converted.

•

and so on

A JEOC interrupt, if enabled, is generated after all injected ADC1 channels in the group
have been converted.
A JEOC interrupt, if enabled, is generated after all injected ADC2 channels in the group
have been converted.
A JEOC interrupt, if enabled, is generated after all injected ADC3 channels in the group
have been converted.
If another external trigger occurs after all injected channels in the group have been
converted then the alternate trigger process restarts by converting the injected ADC1
channels in the group.
Figure 60. Alternate trigger: injected group of each ADC
ST TRIGGER

%/# *%/#
ON !$#

TH TRIGGER

%/# *%/#
ON !$#

!$#

N TH TRIGGER

3AMPLING
#ONVERSION



!$#

ND TRIGGER
N  TH TRIGGER

RD TRIGGER
%/# *%/#
ON !$#

N  TH TRIGGER

TH TRIGGER
%/# *%/#
ON !$#

AI

13.9.5

Combined regular/injected simultaneous mode
It is possible to interrupt the simultaneous conversion of a regular group to start the
simultaneous conversion of an injected group.

Note:

410/1747

In combined regular/injected simultaneous mode, one must convert sequences with the
same length or ensure that the interval between triggers is longer than the long conversion
time of the 2 sequences (Dual ADC mode) /3 sequences (Triple ADC mode). Otherwise, the

RM0090 Rev 17

RM0090

Analog-to-digital converter (ADC)
ADC with the shortest sequence may restart while the ADC with the longest sequence is
completing the previous conversions.

13.9.6

Combined regular simultaneous + alternate trigger mode
It is possible to interrupt the simultaneous conversion of a regular group to start the alternate
trigger conversion of an injected group. Figure 61 shows the behavior of an alternate trigger
interrupting a simultaneous regular conversion.
The injected alternate conversion is immediately started after the injected event. If regular
conversion is already running, in order to ensure synchronization after the injected
conversion, the regular conversion of all (master/slave) ADCs is stopped and resumed
synchronously at the end of the injected conversion.

Note:

In combined regular simultaneous + alternate trigger mode, one must convert sequences
with the same length or ensure that the interval between triggers is longer than the long
conversion time of the 2 sequences (Dual ADC mode) /3 sequences (Triple ADC mode).
Otherwise, the ADC with the shortest sequence may restart while the ADC with the longest
sequence is completing the previous conversions.
If the conversion sequence is interrupted (for instance when DMA end of transfer occurs),
the multi-ADC sequencer must be reset by configuring it in independent mode first (bits
DUAL[4:0] = 00000) before reprogramming the interleaved mode.
Figure 61. Alternate + regular simultaneous
ST TRIGGER

!$# REG

#(

#(

#(

#(

#(

#(

#(

#(

#(

#(

#(

!$# INJ
!$# REG

#(

#(

#(

#(

#(

!$# INJ

SYNCHRO NOT LOST
ND TRIGGER

AI

If a trigger occurs during an injected conversion that has interrupted a regular conversion, it
is ignored. Figure 62 shows the behavior in this case (2nd trigger is ignored).

RM0090 Rev 17

411/1747
432

Analog-to-digital converter (ADC)

RM0090

Figure 62. Case of trigger occurring during injected conversion
ST TRIGGER

#(

!$# REG

#(

RD TRIGGER

#(

#(

#(

#(

#(

!$# INJ
#(

!$# REG

#(

#(

#(
#(

#(

#(

#(

#(

#(

!$# INJ

ND TRIGGER

ND TRIGGER
AI

13.10

Temperature sensor
The temperature sensor can be used to measure the ambient temperature (TA) of the
device.
•

On STM32F40x and STM32F41x devices, the temperature sensor is internally
connected to ADC1_IN16 channel which is used to convert the sensor output voltage
to a digital value.

•

On STM32F42x and STM32F43x devices, the temperature sensor is internally
connected to the same input channel, ADC1_IN18, as VBAT: ADC1_IN18 is used to
convert the sensor output voltage or VBAT into a digital value. Only one conversion,
temperature sensor or VBAT, must be selected at a time. When the temperature sensor
and the VBAT conversion are set simultaneously, only the VBAT conversion is
performed.

Figure 63 shows the block diagram of the temperature sensor.
When not in use, the sensor can be put in power down mode.
Note:

The TSVREFE bit must be set to enable the conversion of both internal channels: the
ADC1_IN16 or ADC1_IN18 (temperature sensor) and the ADC1_IN17 (VREFINT).

Main features

412/1747

•

Supported temperature range: –40 to 125 °C

•

Precision: ±1.5 °C

RM0090 Rev 17

RM0090

Analog-to-digital converter (ADC)
Figure 63. Temperature sensor and VREFINT channel block diagram
4362%&% CONTROL BIT

6 3%.3%

!$#?).
!$#?).
CONVERTED DATA
!$#

62%&).4

)NTERNAL
POWER BLOCK

!DDRESSDATA BUS

4EMPERATURE
SENSOR

!$#?).

-36

1. VSENSE is input to ADC1_IN16 for the STM23F40x and STM32F41x devices and to ADC1_IN18 for the
STM32F42x and STM32F43x devices.

Reading the temperature
To use the sensor:
3.

Select ADC1_IN16 or ADC1_IN18 input channel.

4.

Select a sampling time greater than the minimum sampling time specified in the
datasheet.

5.

Set the TSVREFE bit in the ADC_CCR register to wake up the temperature sensor
from power down mode

6.

Start the ADC conversion by setting the SWSTART bit (or by external trigger)

7.

Read the resulting VSENSE data in the ADC data register

8.

Calculate the temperature using the following formula:
Temperature (in °C) = {(VSENSE – V25) / Avg_Slope} + 25
Where:
–

V25 = VSENSE value for 25° C

–

Avg_Slope = average slope of the temperature vs. VSENSE curve (given in mV/°C
or µV/°C)

Refer to the datasheet’s electrical characteristics section for the actual values of V25
and Avg_Slope.
Note:

The sensor has a startup time after waking from power down mode before it can output
VSENSE at the correct level. The ADC also has a startup time after power-on, so to minimize
the delay, the ADON and TSVREFE bits should be set at the same time.
The temperature sensor output voltage changes linearly with temperature. The offset of this
linear function depends on each chip due to process variation (up to 45 °C from one chip to
another).

RM0090 Rev 17

413/1747
432

Analog-to-digital converter (ADC)

RM0090

The internal temperature sensor is more suited for applications that detect temperature
variations instead of absolute temperatures. If accurate temperature reading is required, an
external temperature sensor should be used.

13.11

Battery charge monitoring
The VBATE bit in the ADC_CCR register is used to switch to the battery voltage. As the
VBAT voltage could be higher than VDDA, to ensure the correct operation of the ADC, the
VBAT pin is internally connected to a bridge divider.
When the VBATE is set, the bridge is automatically enabled to connect:
•

VBAT/2 to the ADC1_IN18 input channel, on STM32F40xx and STM32F41xx devices

•

VBAT/4 to the ADC1_IN18 input channel, on STM32F42xx and STM32F43xx devices

Note:

On STM32F42xx and STM32F43xx devices, VBAT and temperature sensor are connected
to the same ADC internal channel (ADC1_IN18). Only one conversion, either temperature
sensor or VBAT, must be selected at a time. When both conversion are enabled
simultaneously, only the VBAT conversion is performed.

13.12

ADC interrupts
An interrupt can be produced on the end of conversion for regular and injected groups,
when the analog watchdog status bit is set and when the overrun status bit is set. Separate
interrupt enable bits are available for flexibility.
Two other flags are present in the ADC_SR register, but there is no interrupt associated with
them:
•

JSTRT (Start of conversion for channels of an injected group)

•

STRT (Start of conversion for channels of a regular group)
Table 70. ADC interrupts
Interrupt event

414/1747

Event flag

Enable control bit

End of conversion of a regular group

EOC

EOCIE

End of conversion of an injected group

JEOC

JEOCIE

Analog watchdog status bit is set

AWD

AWDIE

Overrun

OVR

OVRIE

RM0090 Rev 17

RM0090

13.13

Analog-to-digital converter (ADC)

ADC registers
Refer to Section 1.1: List of abbreviations for registers for registers for a list of abbreviations
used in register descriptions.
The peripheral registers must be written at word level (32 bits). Read accesses can be done
by bytes (8 bits), half-words (16 bits) or words (32 bits).

13.13.1

ADC status register (ADC_SR)
Address offset: 0x00
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

Reserved
15

14

13

12

11

10

9

8

7

Reserved

6

5

4

3

2

1

0

OVR

STRT

JSTRT

JEOC

EOC

AWD

rc_w0

rc_w0

rc_w0

rc_w0

rc_w0

rc_w0

Bits 31:6 Reserved, must be kept at reset value.
Bit 5 OVR: Overrun
This bit is set by hardware when data are lost (either in single mode or in dual/triple mode). It
is cleared by software. Overrun detection is enabled only when DMA = 1 or EOCS = 1.
0: No overrun occurred
1: Overrun has occurred
Bit 4 STRT: Regular channel start flag
This bit is set by hardware when regular channel conversion starts. It is cleared by software.
0: No regular channel conversion started
1: Regular channel conversion has started
Bit 3 JSTRT: Injected channel start flag
This bit is set by hardware when injected group conversion starts. It is cleared by software.
0: No injected group conversion started
1: Injected group conversion has started
Bit 2 JEOC: Injected channel end of conversion
This bit is set by hardware at the end of the conversion of all injected channels in the group.
It is cleared by software.
0: Conversion is not complete
1: Conversion complete
Bit 1 EOC: Regular channel end of conversion
This bit is set by hardware at the end of the conversion of a regular group of channels. It is
cleared by software or by reading the ADC_DR register.
0: Conversion not complete (EOCS=0), or sequence of conversions not complete (EOCS=1)
1: Conversion complete (EOCS=0), or sequence of conversions complete (EOCS=1)
Bit 0 AWD: Analog watchdog flag
This bit is set by hardware when the converted voltage crosses the values programmed in
the ADC_LTR and ADC_HTR registers. It is cleared by software.
0: No analog watchdog event occurred
1: Analog watchdog event occurred

RM0090 Rev 17

415/1747
432

Analog-to-digital converter (ADC)

13.13.2

RM0090

ADC control register 1 (ADC_CR1)
Address offset: 0x04
Reset value: 0x0000 0000

31

30

29

28

27

Reserved
15

14

13

DISCNUM[2:0]
rw

rw

rw

26

25

OVRIE

24
RES

23

22

21

20

AWDEN JAWDEN

rw

rw

rw

rw

rw

19

18

17

16

1

0

rw

rw

Reserved

12

11

10

9

8

7

6

5

JDISCE
N

DISC
EN

JAUTO

AWDSG
L

SCAN

JEOCIE

AWDIE

EOCIE

rw

rw

rw

rw

rw

rw

rw

rw

4

3

2
AWDCH[4:0]

rw

rw

rw

Bits 31:27 Reserved, must be kept at reset value.
Bit 26 OVRIE: Overrun interrupt enable
This bit is set and cleared by software to enable/disable the Overrun interrupt.
0: Overrun interrupt disabled
1: Overrun interrupt enabled. An interrupt is generated when the OVR bit is set.
Bits 25:24 RES[1:0]: Resolution
These bits are written by software to select the resolution of the conversion.
00: 12-bit (15 ADCCLK cycles)
01: 10-bit (13 ADCCLK cycles)
10: 8-bit (11 ADCCLK cycles)
11: 6-bit (9 ADCCLK cycles)
Bit 23 AWDEN: Analog watchdog enable on regular channels
This bit is set and cleared by software.
0: Analog watchdog disabled on regular channels
1: Analog watchdog enabled on regular channels
Bit 22 JAWDEN: Analog watchdog enable on injected channels
This bit is set and cleared by software.
0: Analog watchdog disabled on injected channels
1: Analog watchdog enabled on injected channels
Bits 21:16 Reserved, must be kept at reset value.
Bits 15:13 DISCNUM[2:0]: Discontinuous mode channel count
These bits are written by software to define the number of regular channels to be converted
in discontinuous mode, after receiving an external trigger.
000: 1 channel
001: 2 channels
...
111: 8 channels
Bit 12 JDISCEN: Discontinuous mode on injected channels
This bit is set and cleared by software to enable/disable discontinuous mode on the injected
channels of a group.
0: Discontinuous mode on injected channels disabled
1: Discontinuous mode on injected channels enabled

416/1747

RM0090 Rev 17

RM0090

Analog-to-digital converter (ADC)

Bit 11 DISCEN: Discontinuous mode on regular channels
This bit is set and cleared by software to enable/disable Discontinuous mode on regular
channels.
0: Discontinuous mode on regular channels disabled
1: Discontinuous mode on regular channels enabled
Bit 10 JAUTO: Automatic injected group conversion
This bit is set and cleared by software to enable/disable automatic injected group conversion
after regular group conversion.
0: Automatic injected group conversion disabled
1: Automatic injected group conversion enabled
Bit 9 AWDSGL: Enable the watchdog on a single channel in scan mode
This bit is set and cleared by software to enable/disable the analog watchdog on the channel
identified by the AWDCH[4:0] bits.
0: Analog watchdog enabled on all channels
1: Analog watchdog enabled on a single channel
Bit 8 SCAN: Scan mode
This bit is set and cleared by software to enable/disable the Scan mode. In Scan mode, the
inputs selected through the ADC_SQRx or ADC_JSQRx registers are converted.
0: Scan mode disabled
1: Scan mode enabled
Note: An EOC interrupt is generated if the EOCIE bit is set:
–
At the end of each regular group sequence if the EOCS bit is cleared to 0
–
At the end of each regular channel conversion if the EOCS bit is set to 1
Note: A JEOC interrupt is generated only on the end of conversion of the last channel if the
JEOCIE bit is set.
Bit 7 JEOCIE: Interrupt enable for injected channels
This bit is set and cleared by software to enable/disable the end of conversion interrupt for
injected channels.
0: JEOC interrupt disabled
1: JEOC interrupt enabled. An interrupt is generated when the JEOC bit is set.
Bit 6 AWDIE: Analog watchdog interrupt enable
This bit is set and cleared by software to enable/disable the analog watchdog interrupt.
0: Analog watchdog interrupt disabled
1: Analog watchdog interrupt enabled
Bit 5 EOCIE: Interrupt enable for EOC
This bit is set and cleared by software to enable/disable the end of conversion interrupt.
0: EOC interrupt disabled
1: EOC interrupt enabled. An interrupt is generated when the EOC bit is set.
Bits 4:0 AWDCH[4:0]: Analog watchdog channel select bits
These bits are set and cleared by software. They select the input channel to be guarded by
the analog watchdog.
Note: 00000: ADC analog input Channel0
00001: ADC analog input Channel1
...
01111: ADC analog input Channel15
10000: ADC analog input Channel16
10001: ADC analog input Channel17
10010: ADC analog input Channel18
Other values reserved

RM0090 Rev 17

417/1747
432

Analog-to-digital converter (ADC)

13.13.3

RM0090

ADC control register 2 (ADC_CR2)
Address offset: 0x08
Reset value: 0x0000 0000

31

30

29

SWST
reserved ART
15

28

27

EXTEN

26

25

24

22

JSWST
ART
reserved

EXTSEL[3:0]

rw

rw

rw

rw

rw

rw

rw

14

13

12

11

10

9

8

ALIGN

EOCS

DDS

DMA

rw

rw

rw

rw

reserved

23

7

21

20

19

JEXTEN

18

17

16

JEXTSEL[3:0]

rw

rw

rw

rw

rw

rw

rw

6

5

4

3

2

1

0

CONT

ADON

rw

rw

Reserved

Bit 31 Reserved, must be kept at reset value.
Bit 30 SWSTART: Start conversion of regular channels
This bit is set by software to start conversion and cleared by hardware as soon as the
conversion starts.
0: Reset state
1: Starts conversion of regular channels
Note: This bit can be set only when ADON = 1 otherwise no conversion is launched.
Bits 29:28 EXTEN: External trigger enable for regular channels
These bits are set and cleared by software to select the external trigger polarity and enable
the trigger of a regular group.
00: Trigger detection disabled
01: Trigger detection on the rising edge
10: Trigger detection on the falling edge
11: Trigger detection on both the rising and falling edges
Bits 27:24 EXTSEL[3:0]: External event select for regular group
These bits select the external event used to trigger the start of conversion of a regular group:
0000: Timer 1 CC1 event
0001: Timer 1 CC2 event
0010: Timer 1 CC3 event
0011: Timer 2 CC2 event
0100: Timer 2 CC3 event
0101: Timer 2 CC4 event
0110: Timer 2 TRGO event
0111: Timer 3 CC1 event
1000: Timer 3 TRGO event
1001: Timer 4 CC4 event
1010: Timer 5 CC1 event
1011: Timer 5 CC2 event
1100: Timer 5 CC3 event
1101: Timer 8 CC1 event
1110: Timer 8 TRGO event
1111: EXTI line11
Bit 23 Reserved, must be kept at reset value.

418/1747

RM0090 Rev 17

RM0090

Analog-to-digital converter (ADC)

Bit 22 JSWSTART: Start conversion of injected channels
This bit is set by software and cleared by hardware as soon as the conversion starts.
0: Reset state
1: Starts conversion of injected channels
Note: This bit can be set only when ADON = 1 otherwise no conversion is launched.
Bits 21:20 JEXTEN: External trigger enable for injected channels
These bits are set and cleared by software to select the external trigger polarity and enable
the trigger of an injected group.
00: Trigger detection disabled
01: Trigger detection on the rising edge
10: Trigger detection on the falling edge
11: Trigger detection on both the rising and falling edges
Bits 19:16 JEXTSEL[3:0]: External event select for injected group
These bits select the external event used to trigger the start of conversion of an injected
group.
0000: Timer 1 CC4 event
0001: Timer 1 TRGO event
0010: Timer 2 CC1 event
0011: Timer 2 TRGO event
0100: Timer 3 CC2 event
0101: Timer 3 CC4 event
0110: Timer 4 CC1 event
0111: Timer 4 CC2 event
1000: Timer 4 CC3 event
1001: Timer 4 TRGO event
1010: Timer 5 CC4 event
1011: Timer 5 TRGO event
1100: Timer 8 CC2 event
1101: Timer 8 CC3 event
1110: Timer 8 CC4 event
1111: EXTI line15
Bits 15:12 Reserved, must be kept at reset value.
Bit 11 ALIGN: Data alignment
This bit is set and cleared by software. Refer to Figure 48 and Figure 49.
0: Right alignment
1: Left alignment
Bit 10 EOCS: End of conversion selection
This bit is set and cleared by software.
0: The EOC bit is set at the end of each sequence of regular conversions. Overrun detection
is enabled only if DMA=1.
1: The EOC bit is set at the end of each regular conversion. Overrun detection is enabled.
Bit 9 DDS: DMA disable selection (for single ADC mode)
This bit is set and cleared by software.
0: No new DMA request is issued after the last transfer (as configured in the DMA controller)
1: DMA requests are issued as long as data are converted and DMA=1
Bit 8 DMA: Direct memory access mode (for single ADC mode)
This bit is set and cleared by software. Refer to the DMA controller chapter for more details.
0: DMA mode disabled
1: DMA mode enabled

RM0090 Rev 17

419/1747
432

Analog-to-digital converter (ADC)

RM0090

Bits 7:2 Reserved, must be kept at reset value.
Bit 1 CONT: Continuous conversion
This bit is set and cleared by software. If it is set, conversion takes place continuously until it
is cleared.
0: Single conversion mode
1: Continuous conversion mode
Bit 0 ADON: A/D Converter ON / OFF
This bit is set and cleared by software.
Note: 0: Disable ADC conversion and go to power down mode
1: Enable ADC

13.13.4

ADC sample time register 1 (ADC_SMPR1)
Address offset: 0x0C
Reset value: 0x0000 0000

31

30

29

28

27

26

14

SMP15_0

13

12

11

SMP14[2:0]

rw

rw

rw

24

23

SMP18[2:0]

Reserved
15

25

rw

21

20

SMP17[2:0]

19

18

SMP16[2:0]

17

16

SMP15[2:1]

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

10

9

8

7

6

5

4

3

2

1

0

SMP13[2:0]
rw

22

rw

SMP12[2:0]
rw

rw

rw

SMP11[2:0]
rw

rw

rw

SMP10[2:0]
rw

rw

rw

rw

Bits 31: 27 Reserved, must be kept at reset value.
Bits 26:0 SMPx[2:0]: Channel x sampling time selection
These bits are written by software to select the sampling time individually for each channel.
During sampling cycles, the channel selection bits must remain unchanged.
Note: 000: 3 cycles
001: 15 cycles
010: 28 cycles
011: 56 cycles
100: 84 cycles
101: 112 cycles
110: 144 cycles
111: 480 cycles

13.13.5

ADC sample time register 2 (ADC_SMPR2)
Address offset: 0x10
Reset value: 0x0000 0000

31

30

Reserved
15

14

SMP
5_0
rw

29

28

27

420/1747

25

24

23

SMP8[2:0]

22

21

20

SMP7[2:0]

19

18

SMP6[2:0]

17

16

SMP5[2:1]

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

13

12

11

10

9

8

7

6

5

4

3

2

1

0

SMP4[2:0]
rw

26

SMP9[2:0]

rw

SMP3[2:0]
rw

rw

rw

SMP2[2:0]
rw

rw

rw

RM0090 Rev 17

SMP1[2:0]
rw

rw

rw

SMP0[2:0]
rw

rw

rw

rw

RM0090

Analog-to-digital converter (ADC)

Bits 31:30 Reserved, must be kept at reset value.
Bits 29:0 SMPx[2:0]: Channel x sampling time selection
These bits are written by software to select the sampling time individually for each channel.
During sample cycles, the channel selection bits must remain unchanged.
Note: 000: 3 cycles
001: 15 cycles
010: 28 cycles
011: 56 cycles
100: 84 cycles
101: 112 cycles
110: 144 cycles
111: 480 cycles

13.13.6

ADC injected channel data offset register x (ADC_JOFRx) (x=1..4)
Address offset: 0x14-0x20
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

Reserved
15

14

13

12

11

10

9

8

7

JOFFSETx[11:0]

Reserved

rw

rw

rw

rw

rw

rw

rw

Bits 31:12 Reserved, must be kept at reset value.
Bits 11:0 JOFFSETx[11:0]: Data offset for injected channel x
These bits are written by software to define the offset to be subtracted from the raw
converted data when converting injected channels. The conversion result can be read from
in the ADC_JDRx registers.

13.13.7

ADC watchdog higher threshold register (ADC_HTR)
Address offset: 0x24
Reset value: 0x0000 0FFF

31

30

29

28

27

26

25

24

15

14

13

12

11

10

9

8

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

Reserved

Reserved

7

HT[11:0]
rw

rw

rw

rw

rw

rw

rw

Bits 31:12 Reserved, must be kept at reset value.
Bits 11:0 HT[11:0]: Analog watchdog higher threshold
These bits are written by software to define the higher threshold for the analog watchdog.

RM0090 Rev 17

421/1747
432

Analog-to-digital converter (ADC)

RM0090

Note:

The software can write to these registers when an ADC conversion is ongoing. The
programmed value will be effective when the next conversion is complete. Writing to this
register is performed with a write delay that can create uncertainty on the effective time at
which the new value is programmed.

13.13.8

ADC watchdog lower threshold register (ADC_LTR)
Address offset: 0x28
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

Reserved
15

14

13

12

11

10

9

8

7

LT[11:0]

Reserved

rw

rw

rw

rw

rw

rw

rw

Bits 31:12 Reserved, must be kept at reset value.
Bits 11:0 LT[11:0]: Analog watchdog lower threshold
These bits are written by software to define the lower threshold for the analog watchdog.

Note:

The software can write to these registers when an ADC conversion is ongoing. The
programmed value will be effective when the next conversion is complete. Writing to this
register is performed with a write delay that can create uncertainty on the effective time at
which the new value is programmed.

13.13.9

ADC regular sequence register 1 (ADC_SQR1)
Address offset: 0x2C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

14

13

SQ16_0
rw

12

11

10

9

rw

rw

rw
7

8

SQ15[4:0]
rw

22

21

20

19

18

rw

rw

rw

rw

rw

rw

rw

6

5

4

3

2

1

0

rw

rw

L[3:0]

Reserved
15

23

rw

rw

rw

16

SQ16[4:1]

SQ14[4:0]
rw

17

SQ13[4:0]
rw

rw

rw

Bits 31:24 Reserved, must be kept at reset value.
Bits 23:20 L[3:0]: Regular channel sequence length
These bits are written by software to define the total number of conversions in the regular
channel conversion sequence.
0000: 1 conversion
0001: 2 conversions
...
1111: 16 conversions
Bits 19:15 SQ16[4:0]: 16th conversion in regular sequence
These bits are written by software with the channel number (0..18) assigned as the 16th in
the conversion sequence.

422/1747

RM0090 Rev 17

RM0090

Analog-to-digital converter (ADC)

Bits 14:10 SQ15[4:0]: 15th conversion in regular sequence
Bits 9:5 SQ14[4:0]: 14th conversion in regular sequence
Bits 4:0 SQ13[4:0]: 13th conversion in regular sequence

13.13.10 ADC regular sequence register 2 (ADC_SQR2)
Address offset: 0x30
Reset value: 0x0000 0000
31

30

Reserved
15

14

29

28

rw

rw

13

12

26

25

24

23

rw

rw

rw

rw

rw

11

10

9

8

7

SQ12[4:0]

SQ10_0
rw

27

rw

rw

21

20

19

18

rw

rw

rw

rw

rw

rw

rw

6

5

4

3

2

1

0

rw

rw

SQ11[4:0]

SQ9[4:0]
rw

22

rw

rw

rw

16

SQ10[4:1]

SQ8[4:0]
rw

17

SQ7[4:0]

rw

rw

rw

rw

rw

rw

Bits 31:30 Reserved, must be kept at reset value.
Bits 29:26 SQ12[4:0]: 12th conversion in regular sequence
These bits are written by software with the channel number (0..18) assigned as the 12th in
the sequence to be converted.
Bits 24:20 SQ11[4:0]: 11th conversion in regular sequence
Bits 19:15 SQ10[4:0]: 10th conversion in regular sequence
Bits 14:10 SQ9[4:0]: 9th conversion in regular sequence
Bits 9:5 SQ8[4:0]: 8th conversion in regular sequence
Bits 4:0 SQ7[4:0]: 7th conversion in regular sequence

13.13.11 ADC regular sequence register 3 (ADC_SQR3)
Address offset: 0x34
Reset value: 0x0000 0000
31

30

Reserved
15

14

29

28

rw

rw

13

12

26

25

24

23

rw

rw

rw

rw

rw

11

10

9

8

7

SQ6[4:0]

SQ4_0
rw

27

rw

rw

21

20

19

18

rw

rw

rw

rw

rw

rw

rw

6

5

4

3

2

1

0

rw

rw

SQ5[4:0]

SQ3[4:0]
rw

22

rw

rw

rw

rw

16

SQ4[4:1]

SQ2[4:0]
rw

17

SQ1[4:0]
rw

rw

rw

rw

rw

Bits 31:30 Reserved, must be kept at reset value.
Bits 29:25 SQ6[4:0]: 6th conversion in regular sequence
These bits are written by software with the channel number (0..18) assigned as the 6th in the
sequence to be converted.
Bits 24:20 SQ5[4:0]: 5th conversion in regular sequence

RM0090 Rev 17

423/1747
432

Analog-to-digital converter (ADC)

RM0090

Bits 19:15 SQ4[4:0]: 4th conversion in regular sequence
Bits 14:10 SQ3[4:0]: 3rd conversion in regular sequence
Bits 9:5 SQ2[4:0]: 2nd conversion in regular sequence
Bits 4:0 SQ1[4:0]: 1st conversion in regular sequence

13.13.12 ADC injected sequence register (ADC_JSQR)
Address offset: 0x38
Reset value: 0x0000 0000
31

30

29

28

27

26

25

24

23

22

14

13

rw

rw

JSQ4[0]
rw

12

11

10

9

8

rw

rw

rw

rw

JSQ3[4:0]
rw

20

19

JL[1:0]

Reserved
15

21

7

17

16

JSQ4[4:1]

rw

rw

rw

rw

rw

rw

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

JSQ2[4:0]
rw

18

JSQ1[4:0]
rw

Bits 31:22 Reserved, must be kept at reset value.
Bits 21:20 JL[1:0]: Injected sequence length
These bits are written by software to define the total number of conversions in the injected
channel conversion sequence.
00: 1 conversion
01: 2 conversions
10: 3 conversions
11: 4 conversions
Bits 19:15 JSQ4[4:0]: 4th conversion in injected sequence (when JL[1:0]=3, see note below)
These bits are written by software with the channel number (0..18) assigned as the 4th in the
sequence to be converted.
Bits 14:10 JSQ3[4:0]: 3rd conversion in injected sequence (when JL[1:0]=3, see note below)
Bits 9:5 JSQ2[4:0]: 2nd conversion in injected sequence (when JL[1:0]=3, see note below)
Bits 4:0 JSQ1[4:0]: 1st conversion in injected sequence (when JL[1:0]=3, see note below)

Note:

When JL[1:0]=3 (4 injected conversions in the sequencer), the ADC converts the channels
in the following order: JSQ1[4:0], JSQ2[4:0], JSQ3[4:0], and JSQ4[4:0].
When JL=2 (3 injected conversions in the sequencer), the ADC converts the channels in the
following order: JSQ2[4:0], JSQ3[4:0], and JSQ4[4:0].
When JL=1 (2 injected conversions in the sequencer), the ADC converts the channels in
starting from JSQ3[4:0], and then JSQ4[4:0].
When JL=0 (1 injected conversion in the sequencer), the ADC converts only JSQ4[4:0]
channel.

424/1747

RM0090 Rev 17

RM0090

Analog-to-digital converter (ADC)

13.13.13 ADC injected data register x (ADC_JDRx) (x= 1..4)
Address offset: 0x3C - 0x48
Reset value: 0x0000 0000
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

r

r

r

r

r

r

r

Reserved
15

14

13

12

11

10

9

8

7

JDATA[15:0]
r

r

r

r

r

r

r

r

r

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 JDATA[15:0]: Injected data
These bits are read-only. They contain the conversion result from injected channel x. The
data are left -or right-aligned as shown in Figure 48 and Figure 49.

13.13.14 ADC regular data register (ADC_DR)
Address offset: 0x4C
Reset value: 0x0000 0000
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

r

r

r

r

r

r

r

Reserved
15

14

13

12

11

10

9

8

7

DATA[15:0]
r

r

r

r

r

r

r

r

r

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 DATA[15:0]: Regular data
These bits are read-only. They contain the conversion result from the regular
channels. The data are left- or right-aligned as shown in Figure 48 and
Figure 49.

RM0090 Rev 17

425/1747
432

Analog-to-digital converter (ADC)

RM0090

13.13.15 ADC Common status register (ADC_CSR)
Address offset: 0x00 (this offset address is relative to ADC1 base address + 0x300)
Reset value: 0x0000 0000
This register provides an image of the status bits of the different ADCs. Nevertheless it is
read-only and does not allow to clear the different status bits. Instead each status bit must
be cleared by writing it to 0 in the corresponding ADC_SR register.
31

30

29

28

27

26

25

24

23

22

21
OVR3

20

19

Reserved

15

14

13

12

OVR2 STRT2

11

10

JSTRT
JEOC2
2

Reserved
r

r

9

8

EOC2

AWD2

7

6

r

r

r

r

r

5

4

3

2

1

0

OVR1

STRT1 JSTRT1 JEOC 1 EOC1

r

r

r

Bit 21 OVR3: Overrun flag of ADC3
This bit is a copy of the OVR bit in the ADC3_SR register.
Bit 20 STRT3: Regular channel Start flag of ADC3
This bit is a copy of the STRT bit in the ADC3_SR register.
Bit 19 JSTRT3: Injected channel Start flag of ADC3
This bit is a copy of the JSTRT bit in the ADC3_SR register.
Bit 18 JEOC3: Injected channel end of conversion of ADC3
This bit is a copy of the JEOC bit in the ADC3_SR register.
Bit 17 EOC3: End of conversion of ADC3
This bit is a copy of the EOC bit in the ADC3_SR register.
Bit 16 AWD3: Analog watchdog flag of ADC3
This bit is a copy of the AWD bit in the ADC3_SR register.
Bits 15:14 Reserved, must be kept at reset value.
Bit 13 OVR2: Overrun flag of ADC2
This bit is a copy of the OVR bit in the ADC2_SR register.
Bit 12 STRT2: Regular channel Start flag of ADC2
This bit is a copy of the STRT bit in the ADC2_SR register.
Bit 11 JSTRT2: Injected channel Start flag of ADC2
This bit is a copy of the JSTRT bit in the ADC2_SR register.
Bit 10 JEOC2: Injected channel end of conversion of ADC2
This bit is a copy of the JEOC bit in the ADC2_SR register.
Bit 9 EOC2: End of conversion of ADC2
This bit is a copy of the EOC bit in the ADC2_SR register.
Bit 8 AWD2: Analog watchdog flag of ADC2
This bit is a copy of the AWD bit in the ADC2_SR register.

RM0090 Rev 17

AWD1

ADC1
r

Bits 31:22 Reserved, must be kept at reset value.

426/1747

16
AWD3

r

Reserved
r

17

ADC3

ADC2
r

18

STRT3 JSTRT3 JEOC 3 EOC3

r

r

r

r

RM0090

Analog-to-digital converter (ADC)

Bits 7:6 Reserved, must be kept at reset value.
Bit 5 OVR1: Overrun flag of ADC1
This bit is a copy of the OVR bit in the ADC1_SR register.
Bit 4 STRT1: Regular channel Start flag of ADC1
This bit is a copy of the STRT bit in the ADC1_SR register.
Bit 3 JSTRT1: Injected channel Start flag of ADC1
This bit is a copy of the JSTRT bit in the ADC1_SR register.
Bit 2 JEOC1: Injected channel end of conversion of ADC1
This bit is a copy of the JEOC bit in the ADC1_SR register.
Bit 1 EOC1: End of conversion of ADC1
This bit is a copy of the EOC bit in the ADC1_SR register.
Bit 0 AWD1: Analog watchdog flag of ADC1
This bit is a copy of the AWD bit in the ADC1_SR register.

13.13.16 ADC common control register (ADC_CCR)
Address offset: 0x04 (this offset address is relative to ADC1 base address + 0x300)
Reset value: 0x0000 0000
31

30

29

28

27

26

25

24

Reserved
15

14

DMA[1:0]
rw

rw

13
DDS
rw

12
Res.

11

23

10

9

8

rw

rw

7

6

DELAY[3:0]
rw

22

21

TSVREFE VBATE

rw

rw

rw

Reserved

20

19

18

4

3

16

ADCPRE

Reserved
5

17

2

rw

rw

1

0

rw

rw

MULTI[4:0]
rw

rw

rw

Bits 31:24 Reserved, must be kept at reset value.
Bit 23 TSVREFE: Temperature sensor and VREFINT enable
This bit is set and cleared by software to enable/disable the temperature sensor and the
VREFINT channel.
0: Temperature sensor and VREFINT channel disabled
1: Temperature sensor and VREFINT channel enabled
Note: On STM32F42x and STM32F43x devices, VBATE must be disabled when TSVREFE is
set. If both bits are set, only the VBAT conversion is performed.
Bit 22 VBATE: VBAT enable
This bit is set and cleared by software to enable/disable the VBAT channel.
0: VBAT channel disabled
1: VBAT channel enabled
Bits 21:18 Reserved, must be kept at reset value.

RM0090 Rev 17

427/1747
432

Analog-to-digital converter (ADC)

RM0090

Bits 17:16 ADCPRE: ADC prescaler
Set and cleared by software to select the frequency of the clock to the ADC. The clock is
common for all the ADCs.
Note: 00: PCLK2 divided by 2
01: PCLK2 divided by 4
10: PCLK2 divided by 6
11: PCLK2 divided by 8
Bits 15:14 DMA: Direct memory access mode for multi ADC mode
This bit-field is set and cleared by software. Refer to the DMA controller section for more
details.
00: DMA mode disabled
01: DMA mode 1 enabled (2 / 3 half-words one by one - 1 then 2 then 3)
10: DMA mode 2 enabled (2 / 3 half-words by pairs - 2&1 then 1&3 then 3&2)
11: DMA mode 3 enabled (2 / 3 bytes by pairs - 2&1 then 1&3 then 3&2)
Bit 13 DDS: DMA disable selection (for multi-ADC mode)
This bit is set and cleared by software.
0: No new DMA request is issued after the last transfer (as configured in the DMA
controller). DMA bits are not cleared by hardware, however they must have been cleared
and set to the wanted mode by software before new DMA requests can be generated.
1: DMA requests are issued as long as data are converted and DMA = 01, 10 or 11.
Bit 12 Reserved, must be kept at reset value.

428/1747

RM0090 Rev 17

RM0090

Analog-to-digital converter (ADC)

Bit 11:8 DELAY: Delay between 2 sampling phases
Set and cleared by software. These bits are used in dual or triple interleaved modes.
0000: 5 * TADCCLK
0001: 6 * TADCCLK
0010: 7 * TADCCLK
...
1111: 20 * TADCCLK
Bits 7:5 Reserved, must be kept at reset value.
Bits 4:0 MULTI[4:0]: Multi ADC mode selection
These bits are written by software to select the operating mode.
– All the ADCs independent:
00000: Independent mode
– 00001 to 01001: Dual mode, ADC1 and ADC2 working together, ADC3 is independent
00001: Combined regular simultaneous + injected simultaneous mode
00010: Combined regular simultaneous + alternate trigger mode
00011: Reserved
00101: Injected simultaneous mode only
00110: Regular simultaneous mode only
00111: interleaved mode only
01001: Alternate trigger mode only
– 10001 to 11001: Triple mode: ADC1, 2 and 3 working together
10001: Combined regular simultaneous + injected simultaneous mode
10010: Combined regular simultaneous + alternate trigger mode
10011: Reserved
10101: Injected simultaneous mode only
10110: Regular simultaneous mode only
10111: interleaved mode only
11001: Alternate trigger mode only
All other combinations are reserved and must not be programmed
Note: In multi mode, a change of channel configuration generates an abort that can cause a
loss of synchronization. It is recommended to disable the multi ADC mode before any
configuration change.

RM0090 Rev 17

429/1747
432

Analog-to-digital converter (ADC)

RM0090

13.13.17 ADC common regular data register for dual and triple modes
(ADC_CDR)
Address offset: 0x08 (this offset address is relative to ADC1 base address + 0x300)
Reset value: 0x0000 0000
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

DATA2[15:0]
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

DATA1[15:0]
r

Bits 31:16 DATA2[15:0]: 2nd data item of a pair of regular conversions
– In dual mode, these bits contain the regular data of ADC2. Refer to Dual ADC mode.
– In triple mode, these bits contain alternatively the regular data of ADC2, ADC1 and ADC3.
Refer to Triple ADC mode.
Bits 15:0 DATA1[15:0]: 1st data item of a pair of regular conversions
– In dual mode, these bits contain the regular data of ADC1. Refer to Dual ADC mode
– In triple mode, these bits contain alternatively the regular data of ADC1, ADC3 and ADC2.
Refer to Triple ADC mode.

13.13.18 ADC register map
The following table summarizes the ADC registers.
Table 71. ADC global register map
Offset

Register

0x000 - 0x04C

ADC1

0x050 - 0x0FC

Reserved

0x100 - 0x14C

ADC2

0x118 - 0x1FC

Reserved

0x200 - 0x24C

ADC3

0x250 - 0x2FC

Reserved

0x300 - 0x308

Common registers

430/1747

RM0090 Rev 17

RM0090

Analog-to-digital converter (ADC)

0x14
0x18
0x1C
0x20
0x24
0x28
0x2C

Reset value

0x38
0x3C
0x40
0x44
0x48
0x4C

0

0

0

0

0

Reset value

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

ADC_JOFR4
ADC_HTR

0

0

0

ADC_LTR
ADC_SQR1

Reset value
ADC_JDR4
Reset value
ADC_DR
Reset value

0

0

0

0

0

0

0

0

0

0

EOC

AWD

EOCIE

JEOC

AWDIE

DMA

SCAN

JAUTO

AWD SGL
DDS

0

JEOCIE

DISCEN

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

AWDCH[4:0]
0

0

0

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

1

1

1

0

0

0

0

0

JOFFSET1[11:0]
0

0

0

0

0

0

0

0

0

0

0

JOFFSET2[11:0]
0

0

0

0

0

0

0

0

JOFFSET3[11:0]
0

0

0

0

0

0

0

0

JOFFSET4[11:0]
0

0

0

0

0

0

0

HT[11:0]
1

1

1

1

1

1

1

LT[11:0]
0

L[3:0]

Reserved

Reset value

ADC_JDR3

0

Reserved

Reset value

Reset value

0

Reserved

Reset value

ADC_JDR2

0

Reserved

Reset value

Reset value

Reserved

Reserved

Reset value

ADC_JDR1

0

Reserved

ADC_JOFR3

Reset value

0

Reserved

Reset value

ADC_JSQR

0

0

Sample time bits SMPx_x

ADC_JOFR2

Reset value

0

0

0

0

Sample time bits SMPx_x

Reset value

Reset value

0

JDISCEN

JEXTEN[1:0]

AWDEN
0

JEXTSEL
[3:0]

0

0

ADON

0

ADC_JOFR1

ADC_SQR3
0x34

0

Re
se
rv
ed

0

ADC_SMPR2

ADC_SQR2
0x30

0

0

EOCS

0x10

0

EXTSEL [3:0]

0

ADC_SMPR1

Reserved Reserved

0x0C

0

0

DISC
NUM [2:0]

Reserved

ALIGN

Reset value

0

JAWDEN

Re
se
rv
ed

EXTEN[1:0]

0x08

ADC_CR2

0
SWSTART

Reset value

JSWSTART

Reserved

RES[1:0]

ADC_CR1

OVRIE

0x04

0

CONT

Reserved

Reset value

JSTRT

ADC_SR

OVR

0x00

Register

STRT

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 72. ADC register map and reset values for each ADC

0

0

0

0

0

0

Regular channel sequence SQx_x bits
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Regular channel sequence SQx_x bits
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Regular channel sequence SQx_x bits
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

JL[1:0]

Reserved

0

0

0

0

0

0

0

Injected channel sequence JSQx_x bits
0

0

0

0

0

0

0

0

0

JDATA[15:0]

Reserved

0

0

0

0

0

0

0

0

0

0

JDATA[15:0]

Reserved

0

0

0

0

0

0

0

0

0

0

JDATA[15:0]

Reserved

0

0

0

0

0

0

0

0

0

0

JDATA[15:0]

Reserved

0

0

0

0

0

0

0

0

0

0

Regular DATA[15:0]

Reserved

0

RM0090 Rev 17

0

0

0

0

0

0

0

0

0

0

431/1747
432

Analog-to-digital converter (ADC)

RM0090

Reset value
0x08

ADC_CDR
Reset value

0

0

0

DDS

0

Reserved

0

0

DELAY [3:0]

0

Regular DATA2[15:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

EOC

AWD

0

0

0

0

0

ADC1

MULTI [4:0]

Reserved

0

0

0

0

0

0

0

0

0

0

0

Regular DATA1[15:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

Refer to Section 2.3: Memory map for the register boundary addresses.

432/1747

0

JEOC

0

JSTRT

0

OVR

0

STRT

EOC

AWD

0

Reserved

JEOC

0

ADC2

DMA[1:0]

0

ADCPRE[1:0]

Reserved

VBATE

ADC_CCR

TSVREFE

0x04

0

JSTRT

0

OVR

0

STRT

0

Reserved

0

ADC3

Reserved

EOC

0

AWD

0

JEOC

Reserved

Reset value

JSTRT

ADC_CSR
0x00

OVR

Register

STRT

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 73. ADC register map and reset values (common ADC registers)

RM0090 Rev 17

0

0

RM0090

14

Digital-to-analog converter (DAC)

Digital-to-analog converter (DAC)
This section applies to the whole STM32F4xx family, unless otherwise specified.

14.1

DAC introduction
The DAC module is a 12-bit, voltage output digital-to-analog converter. The DAC can be
configured in 8- or 12-bit mode and may be used in conjunction with the DMA controller. In
12-bit mode, the data could be left- or right-aligned. The DAC has two output channels, each
with its own converter. In dual DAC channel mode, conversions could be done
independently or simultaneously when both channels are grouped together for synchronous
update operations. An input reference pin, VREF+ (shared with ADC) is available for better
resolution.

14.2

DAC main features
•

Two DAC converters: one output channel each

•

Left or right data alignment in 12-bit mode

•

Synchronized update capability

•

Noise-wave generation

•

Triangular-wave generation

•

Dual DAC channel for independent or simultaneous conversions

•

DMA capability for each channel

•

DMA underrun error detection

•

External triggers for conversion

•

Input voltage reference, VREF+

Figure 64 shows the block diagram of a DAC channel and Table 74 gives the pin
description.

RM0090 Rev 17

433/1747
454

Digital-to-analog converter (DAC)

RM0090
Figure 64. DAC channel block diagram

'$&FRQWUROUHJLVWHU

(;7,B

6:75 ,*[
7,0B7 5*2
7,0B7 5*2
7,0B7 5*2
7,0B7 5*2
7,0B7 5*2
7,0B7 5*2

7ULJJHUVHOHFWRU[

76(/[>@ELWV

'0$(1[

'0 $UHT XH VW[
'+5[

ELW

&RQWUROORJLF[
/)65[

WULDQJOH[

7(1[
0$03[>@ELWV
:$9(1[>@ELWV

ELW
'25[
ELW
9''$
'$&B 28 7[

'LJLWDOWRDQDORJ
FRQYHUWHU[

966$
95 ()

DLE

Table 74. DAC pins
Name

Note:

434/1747

Signal type

Remarks

VREF+

Input, analog reference
positive

The higher/positive reference voltage for the DAC,
1.8 V ≤ VREF+ ≤ VDDA

VDDA

Input, analog supply

Analog power supply

VSSA

Input, analog supply ground

Ground for analog power supply

DAC_OUTx

Analog output signal

DAC channelx analog output

Once the DAC channelx is enabled, the corresponding GPIO pin (PA4 or PA5) is
automatically connected to the analog converter output (DAC_OUTx). In order to avoid
parasitic consumption, the PA4 or PA5 pin should first be configured to analog (AIN).

RM0090 Rev 17

RM0090

Digital-to-analog converter (DAC)

14.3

DAC functional description

14.3.1

DAC channel enable
Each DAC channel can be powered on by setting its corresponding ENx bit in the DAC_CR
register. The DAC channel is then enabled after a startup time tWAKEUP.

Note:

The ENx bit enables the analog DAC Channelx macrocell only. The DAC Channelx digital
interface is enabled even if the ENx bit is reset.

14.3.2

DAC output buffer enable
The DAC integrates two output buffers that can be used to reduce the output impedance,
and to drive external loads directly without having to add an external operational amplifier.
Each DAC channel output buffer can be enabled and disabled using the corresponding
BOFFx bit in the DAC_CR register.

14.3.3

DAC data format
Depending on the selected configuration mode, the data have to be written into the specified
register as described below:
•

Single DAC channelx, there are three possibilities:
–

8-bit right alignment: the software has to load data into the DAC_DHR8Rx [7:0]
bits (stored into the DHRx[11:4] bits)

–

12-bit left alignment: the software has to load data into the DAC_DHR12Lx [15:4]
bits (stored into the DHRx[11:0] bits)

–

12-bit right alignment: the software has to load data into the DAC_DHR12Rx [11:0]
bits (stored into the DHRx[11:0] bits)

Depending on the loaded DAC_DHRyyyx register, the data written by the user is shifted and
stored into the corresponding DHRx (data holding registerx, which are internal non-memorymapped registers). The DHRx register is then loaded into the DORx register either
automatically, by software trigger or by an external event trigger.

RM0090 Rev 17

435/1747
454

Digital-to-analog converter (DAC)

RM0090

Figure 65. Data registers in single DAC channel mode










ELWULJKWDOLJQHG
ELWOHIWDOLJQHG
ELWULJKWDOLJQHG
DLE

•

Dual DAC channels, there are three possibilities:
–

8-bit right alignment: data for DAC channel1 to be loaded into the DAC_DHR8RD
[7:0] bits (stored into the DHR1[11:4] bits) and data for DAC channel2 to be loaded
into the DAC_DHR8RD [15:8] bits (stored into the DHR2[11:4] bits)

–

12-bit left alignment: data for DAC channel1 to be loaded into the DAC_DHR12LD
[15:4] bits (stored into the DHR1[11:0] bits) and data for DAC channel2 to be
loaded into the DAC_DHR12LD [31:20] bits (stored into the DHR2[11:0] bits)

–

12-bit right alignment: data for DAC channel1 to be loaded into the
DAC_DHR12RD [11:0] bits (stored into the DHR1[11:0] bits) and data for DAC
channel2 to be loaded into the DAC_DHR12LD [27:16] bits (stored into the
DHR2[11:0] bits)

Depending on the loaded DAC_DHRyyyD register, the data written by the user is shifted
and stored into DHR1 and DHR2 (data holding registers, which are internal non-memorymapped registers). The DHR1 and DHR2 registers are then loaded into the DOR1 and
DOR2 registers, respectively, either automatically, by software trigger or by an external
event trigger.
Figure 66. Data registers in dual DAC channel mode



31

24



15

7

0


8-bit right aligned

ELWULJKWDOLJQHG

12-bit left aligned

ELWOHIWDOLJQHG

12-bit right aligned

ELWULJKWDOLJQHG
ai14709

DLE

14.3.4

DAC conversion
The DAC_DORx cannot be written directly and any data transfer to the DAC channelx must
be performed by loading the DAC_DHRx register (write to DAC_DHR8Rx, DAC_DHR12Lx,
DAC_DHR12Rx, DAC_DHR8RD, DAC_DHR12LD or DAC_DHR12LD).
Data stored in the DAC_DHRx register are automatically transferred to the DAC_DORx
register after one APB1 clock cycle, if no hardware trigger is selected (TENx bit in DAC_CR
register is reset). However, when a hardware trigger is selected (TENx bit in DAC_CR
register is set) and a trigger occurs, the transfer is performed three APB1 clock cycles later.

436/1747

RM0090 Rev 17

RM0090

Digital-to-analog converter (DAC)
When DAC_DORx is loaded with the DAC_DHRx contents, the analog output voltage
becomes available after a time tSETTLING that depends on the power supply voltage and the
analog output load.
Figure 67. Timing diagram for conversion with trigger disabled TEN = 0
!0"?#,+

$(2

$/2

X!#

/UTPUT VOLTAGE
AVAILABLE ON $!#?/54 PIN

X!#
T3%44,).'

14.3.5

AIC

DAC output voltage
Digital inputs are converted to output voltages on a linear conversion between 0 and VREF+.
The analog output voltages on each DAC channel pin are determined by the following
equation:
DOR
DACoutput = V REF × -------------4096

14.3.6

DAC trigger selection
If the TENx control bit is set, conversion can then be triggered by an external event (timer
counter, external interrupt line). The TSELx[2:0] control bits determine which out of 8
possible events will trigger conversion as shown in Table 75.
Table 75. External triggers
Source

Type

TSEL[2:0]

Timer 6 TRGO event

000

Timer 8 TRGO event

001

Timer 7 TRGO event
Timer 5 TRGO event

Internal signal from on-chip
timers

010
011

Timer 2 TRGO event

100

Timer 4 TRGO event

101

EXTI line9

External pin

110

SWTRIG

Software control bit

111

Each time a DAC interface detects a rising edge on the selected timer TRGO output, or on
the selected external interrupt line 9, the last data stored into the DAC_DHRx register are
transferred into the DAC_DORx register. The DAC_DORx register is updated three APB1
cycles after the trigger occurs.

RM0090 Rev 17

437/1747
454

Digital-to-analog converter (DAC)

RM0090

If the software trigger is selected, the conversion starts once the SWTRIG bit is set.
SWTRIG is reset by hardware once the DAC_DORx register has been loaded with the
DAC_DHRx register contents.
Note:

TSELx[2:0] bit cannot be changed when the ENx bit is set.
When software trigger is selected, the transfer from the DAC_DHRx register to the
DAC_DORx register takes only one APB1 clock cycle.

14.3.7

DMA request
Each DAC channel has a DMA capability. Two DMA channels are used to service DAC
channel DMA requests.
A DAC DMA request is generated when an external trigger (but not a software trigger)
occurs while the DMAENx bit is set. The value of the DAC_DHRx register is then transferred
into the DAC_DORx register.
In dual mode, if both DMAENx bits are set, two DMA requests are generated. If only one
DMA request is needed, the user should set only the corresponding DMAENx bit. In this
way, the application can manage both DAC channels in dual mode by using one DMA
request and a unique DMA channel.

DMA underrun
The DAC DMA request is not queued so that if a second external trigger arrives before the
acknowledgement for the first external trigger is received (first request), then no new
request is issued and the DMA channelx underrun flag DMAUDRx in the DAC_SR register
is set, reporting the error condition. DMA data transfers are then disabled and no further
DMA request is treated. The DAC channelx continues to convert old data.
The software should clear the DMAUDRx flag by writing “1”, clear the DMAEN bit of the
used DMA stream and re-initialize both DMA and DAC channelx to restart the transfer
correctly. The software should modify the DAC trigger conversion frequency or lighten the
DMA workload to avoid a new DMA underrun. Finally, the DAC conversion could be
resumed by enabling both DMA data transfer and conversion trigger.
For each DAC channelx, an interrupt is also generated if its corresponding DMAUDRIEx bit
in the DAC_CR register is enabled.

14.3.8

Noise generation
In order to generate a variable-amplitude pseudonoise, an LFSR (linear feedback shift
register) is available. DAC noise generation is selected by setting WAVEx[1:0] to “01”. The
preloaded value in LFSR is 0xAAA. This register is updated three APB1 clock cycles after
each trigger event, following a specific calculation algorithm.

438/1747

RM0090 Rev 17

RM0090

Digital-to-analog converter (DAC)
Figure 68. DAC LFSR register calculation algorithm

;25
;
; 













;




;

;









125

DLF

The LFSR value, that may be masked partially or totally by means of the MAMPx[3:0] bits in
the DAC_CR register, is added up to the DAC_DHRx contents without overflow and this
value is then stored into the DAC_DORx register.
If LFSR is 0x0000, a ‘1 is injected into it (antilock-up mechanism).
It is possible to reset LFSR wave generation by resetting the WAVEx[1:0] bits.
Figure 69. DAC conversion (SW trigger enabled) with LFSR wave generation

$3%B&/.

'+5

[

[$$$

'25

['

6:75,*
DLE

Note:

The DAC trigger must be enabled for noise generation by setting the TENx bit in the
DAC_CR register.

14.3.9

Triangle-wave generation
It is possible to add a small-amplitude triangular waveform on a DC or slowly varying signal.
DAC triangle-wave generation is selected by setting WAVEx[1:0] to “10”. The amplitude is
configured through the MAMPx[3:0] bits in the DAC_CR register. An internal triangle counter
is incremented three APB1 clock cycles after each trigger event. The value of this counter is
then added to the DAC_DHRx register without overflow and the sum is stored into the
DAC_DORx register. The triangle counter is incremented as long as it is less than the
maximum amplitude defined by the MAMPx[3:0] bits. Once the configured amplitude is
reached, the counter is decremented down to 0, then incremented again and so on.
RM0090 Rev 17

439/1747
454

Digital-to-analog converter (DAC)

RM0090

It is possible to reset triangle wave generation by resetting the WAVEx[1:0] bits.
Figure 70. DAC triangle wave generation

N
TIO
TA
M
EN

N

)N

TIO

TA

EN

CR
E

M

RE

C
$E

-!-0X;= MAX AMPLITUDE
$!#?$(2X BASE VALUE

$!#?$(2X BASE VALUE

AIC

Figure 71. DAC conversion (SW trigger enabled) with triangle wave generation
$3%B&/.

'+5

'25

[

[$$$

['

6:75,*
DLE

Note:

The DAC trigger must be enabled for noise generation by setting the TENx bit in the
DAC_CR register.
The MAMPx[3:0] bits must be configured before enabling the DAC, otherwise they cannot
be changed.

14.4

Dual DAC channel conversion
To efficiently use the bus bandwidth in applications that require the two DAC channels at the
same time, three dual registers are implemented: DHR8RD, DHR12RD and DHR12LD. A
unique register access is then required to drive both DAC channels at the same time.
Eleven possible conversion modes are possible using the two DAC channels and these dual
registers. All the conversion modes can nevertheless be obtained using separate DHRx
registers if needed.
All modes are described in the paragraphs below.

440/1747

RM0090 Rev 17

RM0090

14.4.1

Digital-to-analog converter (DAC)

Independent trigger without wave generation
To configure the DAC in this conversion mode, the following sequence is required:
•

Set the two DAC channel trigger enable bits TEN1 and TEN2

•

Configure different trigger sources by setting different values in the TSEL1[2:0] and
TSEL2[2:0] bits

•

Load the dual DAC channel data into the desired DHR register (DAC_DHR12RD,
DAC_DHR12LD or DAC_DHR8RD)

When a DAC channel1 trigger arrives, the DHR1 register is transferred into DAC_DOR1
(three APB1 clock cycles later).
When a DAC channel2 trigger arrives, the DHR2 register is transferred into DAC_DOR2
(three APB1 clock cycles later).

14.4.2

Independent trigger with single LFSR generation
To configure the DAC in this conversion mode, the following sequence is required:
•

Set the two DAC channel trigger enable bits TEN1 and TEN2

•

Configure different trigger sources by setting different values in the TSEL1[2:0] and
TSEL2[2:0] bits

•

Configure the two DAC channel WAVEx[1:0] bits as “01” and the same LFSR mask
value in the MAMPx[3:0] bits

•

Load the dual DAC channel data into the desired DHR register (DHR12RD, DHR12LD
or DHR8RD)

When a DAC channel1 trigger arrives, the LFSR1 counter, with the same mask, is added to
the DHR1 register and the sum is transferred into DAC_DOR1 (three APB1 clock cycles
later). Then the LFSR1 counter is updated.
When a DAC channel2 trigger arrives, the LFSR2 counter, with the same mask, is added to
the DHR2 register and the sum is transferred into DAC_DOR2 (three APB1 clock cycles
later). Then the LFSR2 counter is updated.

14.4.3

Independent trigger with different LFSR generation
To configure the DAC in this conversion mode, the following sequence is required:
•

Set the two DAC channel trigger enable bits TEN1 and TEN2

•

Configure different trigger sources by setting different values in the TSEL1[2:0] and
TSEL2[2:0] bits

•

Configure the two DAC channel WAVEx[1:0] bits as “01” and set different LFSR masks
values in the MAMP1[3:0] and MAMP2[3:0] bits

•

Load the dual DAC channel data into the desired DHR register (DAC_DHR12RD,
DAC_DHR12LD or DAC_DHR8RD)

When a DAC channel1 trigger arrives, the LFSR1 counter, with the mask configured by
MAMP1[3:0], is added to the DHR1 register and the sum is transferred into DAC_DOR1
(three APB1 clock cycles later). Then the LFSR1 counter is updated.
When a DAC channel2 trigger arrives, the LFSR2 counter, with the mask configured by
MAMP2[3:0], is added to the DHR2 register and the sum is transferred into DAC_DOR2
(three APB1 clock cycles later). Then the LFSR2 counter is updated.

RM0090 Rev 17

441/1747
454

Digital-to-analog converter (DAC)

14.4.4

RM0090

Independent trigger with single triangle generation
To configure the DAC in this conversion mode, the following sequence is required:
•

Set the two DAC channel trigger enable bits TEN1 and TEN2

•

Configure different trigger sources by setting different values in the TSEL1[2:0] and
TSEL2[2:0] bits

•

Configure the two DAC channel WAVEx[1:0] bits as “1x” and the same maximum
amplitude value in the MAMPx[3:0] bits

•

Load the dual DAC channel data into the desired DHR register (DAC_DHR12RD,
DAC_DHR12LD or DAC_DHR8RD)

When a DAC channel1 trigger arrives, the DAC channel1 triangle counter, with the same
triangle amplitude, is added to the DHR1 register and the sum is transferred into
DAC_DOR1 (three APB1 clock cycles later). The DAC channel1 triangle counter is then
updated.
When a DAC channel2 trigger arrives, the DAC channel2 triangle counter, with the same
triangle amplitude, is added to the DHR2 register and the sum is transferred into
DAC_DOR2 (three APB1 clock cycles later). The DAC channel2 triangle counter is then
updated.

14.4.5

Independent trigger with different triangle generation
To configure the DAC in this conversion mode, the following sequence is required:
•

Set the two DAC channel trigger enable bits TEN1 and TEN2

•

Configure different trigger sources by setting different values in the TSEL1[2:0] and
TSEL2[2:0] bits

•

Configure the two DAC channel WAVEx[1:0] bits as “1x” and set different maximum
amplitude values in the MAMP1[3:0] and MAMP2[3:0] bits

•

Load the dual DAC channel data into the desired DHR register (DAC_DHR12RD,
DAC_DHR12LD or DAC_DHR8RD)

When a DAC channel1 trigger arrives, the DAC channel1 triangle counter, with a triangle
amplitude configured by MAMP1[3:0], is added to the DHR1 register and the sum is
transferred into DAC_DOR1 (three APB1 clock cycles later). The DAC channel1 triangle
counter is then updated.
When a DAC channel2 trigger arrives, the DAC channel2 triangle counter, with a triangle
amplitude configured by MAMP2[3:0], is added to the DHR2 register and the sum is
transferred into DAC_DOR2 (three APB1 clock cycles later). The DAC channel2 triangle
counter is then updated.

14.4.6

Simultaneous software start
To configure the DAC in this conversion mode, the following sequence is required:
•

Load the dual DAC channel data to the desired DHR register (DAC_DHR12RD,
DAC_DHR12LD or DAC_DHR8RD)

In this configuration, one APB1 clock cycle later, the DHR1 and DHR2 registers are
transferred into DAC_DOR1 and DAC_DOR2, respectively.

442/1747

RM0090 Rev 17

RM0090

14.4.7

Digital-to-analog converter (DAC)

Simultaneous trigger without wave generation
To configure the DAC in this conversion mode, the following sequence is required:
•

Set the two DAC channel trigger enable bits TEN1 and TEN2

•

Configure the same trigger source for both DAC channels by setting the same value in
the TSEL1[2:0] and TSEL2[2:0] bits

•

Load the dual DAC channel data to the desired DHR register (DAC_DHR12RD,
DAC_DHR12LD or DAC_DHR8RD)

When a trigger arrives, the DHR1 and DHR2 registers are transferred into DAC_DOR1 and
DAC_DOR2, respectively (after three APB1 clock cycles).

14.4.8

Simultaneous trigger with single LFSR generation
To configure the DAC in this conversion mode, the following sequence is required:
•

Set the two DAC channel trigger enable bits TEN1 and TEN2

•

Configure the same trigger source for both DAC channels by setting the same value in
the TSEL1[2:0] and TSEL2[2:0] bits

•

Configure the two DAC channel WAVEx[1:0] bits as “01” and the same LFSR mask
value in the MAMPx[3:0] bits

•

Load the dual DAC channel data to the desired DHR register (DHR12RD, DHR12LD or
DHR8RD)

When a trigger arrives, the LFSR1 counter, with the same mask, is added to the DHR1
register and the sum is transferred into DAC_DOR1 (three APB1 clock cycles later). The
LFSR1 counter is then updated. At the same time, the LFSR2 counter, with the same mask,
is added to the DHR2 register and the sum is transferred into DAC_DOR2 (three APB1
clock cycles later). The LFSR2 counter is then updated.

14.4.9

Simultaneous trigger with different LFSR generation
To configure the DAC in this conversion mode, the following sequence is required:
•

Set the two DAC channel trigger enable bits TEN1 and TEN2

•

Configure the same trigger source for both DAC channels by setting the same value in
the TSEL1[2:0] and TSEL2[2:0] bits

•

Configure the two DAC channel WAVEx[1:0] bits as “01” and set different LFSR mask
values using the MAMP1[3:0] and MAMP2[3:0] bits

•

Load the dual DAC channel data into the desired DHR register (DAC_DHR12RD,
DAC_DHR12LD or DAC_DHR8RD)

When a trigger arrives, the LFSR1 counter, with the mask configured by MAMP1[3:0], is
added to the DHR1 register and the sum is transferred into DAC_DOR1 (three APB1 clock
cycles later). The LFSR1 counter is then updated.
At the same time, the LFSR2 counter, with the mask configured by MAMP2[3:0], is added to
the DHR2 register and the sum is transferred into DAC_DOR2 (three APB1 clock cycles
later). The LFSR2 counter is then updated.

RM0090 Rev 17

443/1747
454

Digital-to-analog converter (DAC)

14.4.10

RM0090

Simultaneous trigger with single triangle generation
To configure the DAC in this conversion mode, the following sequence is required:
•

Set the two DAC channel trigger enable bits TEN1 and TEN2

•

Configure the same trigger source for both DAC channels by setting the same value in
the TSEL1[2:0] and TSEL2[2:0] bits

•

Configure the two DAC channel WAVEx[1:0] bits as “1x” and the same maximum
amplitude value using the MAMPx[3:0] bits

•

Load the dual DAC channel data into the desired DHR register (DAC_DHR12RD,
DAC_DHR12LD or DAC_DHR8RD)

When a trigger arrives, the DAC channel1 triangle counter, with the same triangle
amplitude, is added to the DHR1 register and the sum is transferred into DAC_DOR1 (three
APB1 clock cycles later). The DAC channel1 triangle counter is then updated.
At the same time, the DAC channel2 triangle counter, with the same triangle amplitude, is
added to the DHR2 register and the sum is transferred into DAC_DOR2 (three APB1 clock
cycles later). The DAC channel2 triangle counter is then updated.

14.4.11

Simultaneous trigger with different triangle generation
To configure the DAC in this conversion mode, the following sequence is required:
•

Set the two DAC channel trigger enable bits TEN1 and TEN2

•

Configure the same trigger source for both DAC channels by setting the same value in
the TSEL1[2:0] and TSEL2[2:0] bits

•

Configure the two DAC channel WAVEx[1:0] bits as “1x” and set different maximum
amplitude values in the MAMP1[3:0] and MAMP2[3:0] bits

•

Load the dual DAC channel data into the desired DHR register (DAC_DHR12RD,
DAC_DHR12LD or DAC_DHR8RD)

When a trigger arrives, the DAC channel1 triangle counter, with a triangle amplitude
configured by MAMP1[3:0], is added to the DHR1 register and the sum is transferred into
DAC_DOR1 (three APB1 clock cycles later). Then the DAC channel1 triangle counter is
updated.
At the same time, the DAC channel2 triangle counter, with a triangle amplitude configured
by MAMP2[3:0], is added to the DHR2 register and the sum is transferred into DAC_DOR2
(three APB1 clock cycles later). Then the DAC channel2 triangle counter is updated.

444/1747

RM0090 Rev 17

RM0090

Digital-to-analog converter (DAC)

14.5

DAC registers
Refer to Section 1.1: List of abbreviations for registers for registers for a list of abbreviations
used in register descriptions.
The peripheral registers have to be accessed by words (32 bits).

14.5.1

DAC control register (DAC_CR)
Address offset: 0x00
Reset value: 0x0000 0000

31

30

Reserved
15

14

Reserved

29

28

DMAU
DRIE2

DMA
EN2

27

26

25

24

MAMP2[3:0]

23

22

21

WAVE2[1:0]

20

19

TSEL2[2:0]

18

17

16

TEN2

BOFF2

EN2

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

13

12

11

10

9

8

7

6

5

4

3

2

1

0

DMAU
DRIE1

DMA
EN1

TEN1

BOFF1

EN1

rw

rw

rw

rw

rw

MAMP1[3:0]
rw

rw

rw

WAVE1[1:0]
rw

rw

rw

TSEL1[2:0]
rw

rw

rw

Bits 31:30 Reserved, must be kept at reset value.
Bits 29 DMAUDRIE2: DAC channel2 DMA underrun interrupt enable
This bit is set and cleared by software.
0: DAC channel2 DMA underrun interrupt disabled
1: DAC channel2 DMA underrun interrupt enabled
Bit 28 DMAEN2: DAC channel2 DMA enable
This bit is set and cleared by software.
0: DAC channel2 DMA mode disabled
1: DAC channel2 DMA mode enabled
Bits 27:24 MAMP2[3:0]: DAC channel2 mask/amplitude selector
These bits are written by software to select mask in wave generation mode or amplitude in
triangle generation mode.
0000: Unmask bit0 of LFSR/ triangle amplitude equal to 1
0001: Unmask bits[1:0] of LFSR/ triangle amplitude equal to 3
0010: Unmask bits[2:0] of LFSR/ triangle amplitude equal to 7
0011: Unmask bits[3:0] of LFSR/ triangle amplitude equal to 15
0100: Unmask bits[4:0] of LFSR/ triangle amplitude equal to 31
0101: Unmask bits[5:0] of LFSR/ triangle amplitude equal to 63
0110: Unmask bits[6:0] of LFSR/ triangle amplitude equal to 127
0111: Unmask bits[7:0] of LFSR/ triangle amplitude equal to 255
1000: Unmask bits[8:0] of LFSR/ triangle amplitude equal to 511
1001: Unmask bits[9:0] of LFSR/ triangle amplitude equal to 1023
1010: Unmask bits[10:0] of LFSR/ triangle amplitude equal to 2047
≥ 1011: Unmask bits[11:0] of LFSR/ triangle amplitude equal to 4095
Bits 23:22 WAVE2[1:0]: DAC channel2 noise/triangle wave generation enable
These bits are set/reset by software.
00: wave generation disabled
01: Noise wave generation enabled
1x: Triangle wave generation enabled
Note: Only used if bit TEN2 = 1 (DAC channel2 trigger enabled)

RM0090 Rev 17

445/1747
454

Digital-to-analog converter (DAC)

RM0090

Bits 21:19 TSEL2[2:0]: DAC channel2 trigger selection
These bits select the external event used to trigger DAC channel2
000: Timer 6 TRGO event
001: Timer 8 TRGO event
010: Timer 7 TRGO event
011: Timer 5 TRGO event
100: Timer 2 TRGO event
101: Timer 4 TRGO event
110: External line9
111: Software trigger
Note: Only used if bit TEN2 = 1 (DAC channel2 trigger enabled).
Bit 18 TEN2: DAC channel2 trigger enable
This bit is set and cleared by software to enable/disable DAC channel2 trigger
0: DAC channel2 trigger disabled and data written into the DAC_DHRx register are
transferred one APB1 clock cycle later to the DAC_DOR2 register
1: DAC channel2 trigger enabled and data from the DAC_DHRx register are transferred
three APB1 clock cycles later to the DAC_DOR2 register
Note: When software trigger is selected, the transfer from the DAC_DHRx register to the
DAC_DOR2 register takes only one APB1 clock cycle.
Bit 17 BOFF2: DAC channel2 output buffer disable
This bit is set and cleared by software to enable/disable DAC channel2 output buffer.
0: DAC channel2 output buffer enabled
1: DAC channel2 output buffer disabled
Bit 16 EN2: DAC channel2 enable
This bit is set and cleared by software to enable/disable DAC channel2.
0: DAC channel2 disabled
1: DAC channel2 enabled
Bits 15:14 Reserved, must be kept at reset value.
Bit 13 DMAUDRIE1: DAC channel1 DMA Underrun Interrupt enable
This bit is set and cleared by software.
0: DAC channel1 DMA Underrun Interrupt disabled
1: DAC channel1 DMA Underrun Interrupt enabled
Bit 12 DMAEN1: DAC channel1 DMA enable
This bit is set and cleared by software.
0: DAC channel1 DMA mode disabled
1: DAC channel1 DMA mode enabled

446/1747

RM0090 Rev 17

RM0090

Digital-to-analog converter (DAC)

Bits 11:8 MAMP1[3:0]: DAC channel1 mask/amplitude selector
These bits are written by software to select mask in wave generation mode or amplitude in
triangle generation mode.
0000: Unmask bit0 of LFSR/ triangle amplitude equal to 1
0001: Unmask bits[1:0] of LFSR/ triangle amplitude equal to 3
0010: Unmask bits[2:0] of LFSR/ triangle amplitude equal to 7
0011: Unmask bits[3:0] of LFSR/ triangle amplitude equal to 15
0100: Unmask bits[4:0] of LFSR/ triangle amplitude equal to 31
0101: Unmask bits[5:0] of LFSR/ triangle amplitude equal to 63
0110: Unmask bits[6:0] of LFSR/ triangle amplitude equal to 127
0111: Unmask bits[7:0] of LFSR/ triangle amplitude equal to 255
1000: Unmask bits[8:0] of LFSR/ triangle amplitude equal to 511
1001: Unmask bits[9:0] of LFSR/ triangle amplitude equal to 1023
1010: Unmask bits[10:0] of LFSR/ triangle amplitude equal to 2047
≥ 1011: Unmask bits[11:0] of LFSR/ triangle amplitude equal to 4095
Bits 7:6 WAVE1[1:0]: DAC channel1 noise/triangle wave generation enable
These bits are set and cleared by software.
00: wave generation disabled
01: Noise wave generation enabled
1x: Triangle wave generation enabled
Note: Only used if bit TEN1 = 1 (DAC channel1 trigger enabled).
Bits 5:3 TSEL1[2:0]: DAC channel1 trigger selection
These bits select the external event used to trigger DAC channel1.
000: Timer 6 TRGO event
001: Timer 8 TRGO event
010: Timer 7 TRGO event
011: Timer 5 TRGO event
100: Timer 2 TRGO event
101: Timer 4 TRGO event
110: External line9
111: Software trigger
Note: Only used if bit TEN1 = 1 (DAC channel1 trigger enabled).
Bit 2 TEN1: DAC channel1 trigger enable
This bit is set and cleared by software to enable/disable DAC channel1 trigger.
0: DAC channel1 trigger disabled and data written into the DAC_DHRx register are
transferred one APB1 clock cycle later to the DAC_DOR1 register
1: DAC channel1 trigger enabled and data from the DAC_DHRx register are transferred
three APB1 clock cycles later to the DAC_DOR1 register
Note: When software trigger is selected, the transfer from the DAC_DHRx register to the
DAC_DOR1 register takes only one APB1 clock cycle.
Bit 1 BOFF1: DAC channel1 output buffer disable
This bit is set and cleared by software to enable/disable DAC channel1 output buffer.
0: DAC channel1 output buffer enabled
1: DAC channel1 output buffer disabled
Bit 0 EN1: DAC channel1 enable
This bit is set and cleared by software to enable/disable DAC channel1.
0: DAC channel1 disabled
1: DAC channel1 enabled

RM0090 Rev 17

447/1747
454

Digital-to-analog converter (DAC)

14.5.2

RM0090

DAC software trigger register (DAC_SWTRIGR)
Address offset: 0x04
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

6

5

4

3

2

17

16

1

0

Reserved
15

14

13

12

11

10

9

8

7

SWTRIG2 SWTRIG1

Reserved

w

w

Bits 31:2 Reserved, must be kept at reset value.
Bit 1 SWTRIG2: DAC channel2 software trigger
This bit is set and cleared by software to enable/disable the software trigger.
0: Software trigger disabled
1: Software trigger enabled
Note: This bit is cleared by hardware (one APB1 clock cycle later) once the DAC_DHR2
register value has been loaded into the DAC_DOR2 register.
Bit 0 SWTRIG1: DAC channel1 software trigger
This bit is set and cleared by software to enable/disable the software trigger.
0: Software trigger disabled
1: Software trigger enabled
Note: This bit is cleared by hardware (one APB1 clock cycle later) once the DAC_DHR1
register value has been loaded into the DAC_DOR1 register.

14.5.3

DAC channel1 12-bit right-aligned data holding register
(DAC_DHR12R1)
Address offset: 0x08
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

Reserved
15

14

13

Reserved

12

11

10

9

8

7

rw

rw

rw

rw

rw

DACC1DHR[11:0]
rw

rw

Bits 31:12 Reserved, must be kept at reset value.
Bits 11:0 DACC1DHR[11:0]: DAC channel1 12-bit right-aligned data
These bits are written by software which specifies 12-bit data for DAC channel1.

448/1747

RM0090 Rev 17

RM0090

Digital-to-analog converter (DAC)

14.5.4

DAC channel1 12-bit left aligned data holding register
(DAC_DHR12L1)
Address offset: 0x0C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

3

2

1

0

Reserved
15

14

13

12

11

10

9

8

7

6

5

4

rw

rw

rw

rw

rw

DACC1DHR[11:0]
rw

rw

rw

rw

rw

rw

rw

Reserved

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:4 DACC1DHR[11:0]: DAC channel1 12-bit left-aligned data
These bits are written by software which specifies 12-bit data for DAC channel1.
Bits 3:0 Reserved, must be kept at reset value.

14.5.5

DAC channel1 8-bit right aligned data holding register
(DAC_DHR8R1)
Address offset: 0x10
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

6

5

20

19

18

17

16

4

3

2

1

0

rw

rw

rw

Reserved
15

14

13

12

11

Reserved

10

9

8

7

DACC1DHR[7:0]
rw

rw

rw

rw

rw

Bits 31:8 Reserved, must be kept at reset value.
Bits 7:0 DACC1DHR[7:0]: DAC channel1 8-bit right-aligned data
These bits are written by software which specifies 8-bit data for DAC channel1.

RM0090 Rev 17

449/1747
454

Digital-to-analog converter (DAC)

14.5.6

RM0090

DAC channel2 12-bit right aligned data holding register
(DAC_DHR12R2)
Address offset: 0x14
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

Reserved
15

14

13

12

11

10

9

8

7

DACC2DHR[11:0]

Reserved

rw

rw

rw

rw

rw

rw

rw

Bits 31:12 Reserved, must be kept at reset value.
Bits 11:0 DACC2DHR[11:0]: DAC channel2 12-bit right-aligned data
These bits are written by software which specifies 12-bit data for DAC channel2.

14.5.7

DAC channel2 12-bit left aligned data holding register
(DAC_DHR12L2)
Address offset: 0x18
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

15

14

13

12

11

10

9

8

rw

23

22

21

20

19

18

17

16

7

6

5

4

3

2

1

0

rw

rw

rw

rw

Reserved

DACC2DHR[11:0]
rw

rw

rw

rw

rw

rw

rw

Reserved

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:4 DACC2DHR[11:0]: DAC channel2 12-bit left-aligned data
These bits are written by software which specify 12-bit data for DAC channel2.
Bits 3:0 Reserved, must be kept at reset value.

14.5.8

DAC channel2 8-bit right-aligned data holding register
(DAC_DHR8R2)
Address offset: 0x1C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

15

14

13

12

11

10

9

8

23

22

21

6

5

20

19

18

17

16

4

3

2

1

0

rw

rw

rw

Reserved

Reserved

7

DACC2DHR[7:0]
rw

rw

rw

rw

rw

Bits 31:8 Reserved, must be kept at reset value.
Bits 7:0 DACC2DHR[7:0]: DAC channel2 8-bit right-aligned data
These bits are written by software which specifies 8-bit data for DAC channel2.

450/1747

RM0090 Rev 17

RM0090

Digital-to-analog converter (DAC)

14.5.9

Dual DAC 12-bit right-aligned data holding register
(DAC_DHR12RD)
Address offset: 0x20
Reset value: 0x0000 0000

31

30

29

28

14

13

26

25

24

23

22

21

20

19

18

17

16

DACC2DHR[11:0]

Reserved
15

27

12

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

17

16

DACC1DHR[11:0]

Reserved

rw

rw

rw

rw

rw

rw

rw

Bits 31:28 Reserved, must be kept at reset value.
Bits 27:16 DACC2DHR[11:0]: DAC channel2 12-bit right-aligned data
These bits are written by software which specifies 12-bit data for DAC channel2.
Bits 15:12 Reserved, must be kept at reset value.
Bits 11:0 DACC1DHR[11:0]: DAC channel1 12-bit right-aligned data
These bits are written by software which specifies 12-bit data for DAC channel1.

14.5.10

DUAL DAC 12-bit left aligned data holding register
(DAC_DHR12LD)
Address offset: 0x24
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

DACC2DHR[11:0]
rw

rw

rw

rw

rw

15

14

13

12

11

rw

rw

rw

rw

rw

rw

rw

10

9

8

7

6

5

4

rw

rw

rw

rw

rw

DACC1DHR[11:0]
rw

rw

rw

rw

rw

rw

rw

18

Reserved
3

2

1

0

Reserved

Bits 31:20 DACC2DHR[11:0]: DAC channel2 12-bit left-aligned data
These bits are written by software which specifies 12-bit data for DAC channel2.
Bits 19:16 Reserved, must be kept at reset value.
Bits 15:4 DACC1DHR[11:0]: DAC channel1 12-bit left-aligned data
These bits are written by software which specifies 12-bit data for DAC channel1.
Bits 3:0 Reserved, must be kept at reset value.

RM0090 Rev 17

451/1747
454

Digital-to-analog converter (DAC)

14.5.11

RM0090

DUAL DAC 8-bit right aligned data holding register
(DAC_DHR8RD)
Address offset: 0x28
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

6

5

20

19

18

17

16

4

3

2

1

0

rw

rw

rw

Reserved
15

14

13

12

11

10

9

8

7

DACC2DHR[7:0]
rw

rw

rw

rw

rw

DACC1DHR[7:0]
rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:8 DACC2DHR[7:0]: DAC channel2 8-bit right-aligned data
These bits are written by software which specifies 8-bit data for DAC channel2.
Bits 7:0 DACC1DHR[7:0]: DAC channel1 8-bit right-aligned data
These bits are written by software which specifies 8-bit data for DAC channel1.

14.5.12

DAC channel1 data output register (DAC_DOR1)
Address offset: 0x2C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

15

14

13

12

11

10

9

8

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

r

r

r

r

r

Reserved
7

DACC1DOR[11:0]

Reserved

r

r

r

r

r

r

r

Bits 31:12 Reserved, must be kept at reset value.
Bit 11:0 DACC1DOR[11:0]: DAC channel1 data output
These bits are read-only, they contain data output for DAC channel1.

14.5.13

DAC channel2 data output register (DAC_DOR2)
Address offset: 0x30
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

r

r

r

r

r

Reserved
15

14

13

Reserved

12

11

10

9

8

7

DACC2DOR[11:0]
r

r

r

r

r

r

r

Bits 31:12 Reserved, must be kept at reset value.
Bits 11:0 DACC2DOR[11:0]: DAC channel2 data output
These bits are read-only, they contain data output for DAC channel2.

452/1747

RM0090 Rev 17

RM0090

Digital-to-analog converter (DAC)

14.5.14

DAC status register (DAC_SR)
Address offset: 0x34
Reset value: 0x0000 0000

31

30

Reserved
15

29

27

26

25

24

23

22

21

20

19

18

17

16

5

4

3

2

1

0

Reserved

rc_w1

14

Reserved

28

DMAUDR2

13

12

11

10

9

8

7

6

DMAUDR1

Reserved

rc_w1

Bits 31:30 Reserved, must be kept at reset value.
Bit 29 DMAUDR2: DAC channel2 DMA underrun flag
This bit is set by hardware and cleared by software (by writing it to 1).
0: No DMA underrun error condition occurred for DAC channel2
1: DMA underrun error condition occurred for DAC channel2 (the currently selected trigger is
driving DAC channel2 conversion at a frequency higher than the DMA service capability rate)
Bits 28:14 Reserved, must be kept at reset value.
Bit 13 DMAUDR1: DAC channel1 DMA underrun flag
This bit is set by hardware and cleared by software (by writing it to 1).
0: No DMA underrun error condition occurred for DAC channel1
1: DMA underrun error condition occurred for DAC channel1 (the currently selected trigger is
driving DAC channel1 conversion at a frequency higher than the DMA service capability rate)
Bits 12:0 Reserved, must be kept at reset value.

14.5.15

DAC register map
Table 76 summarizes the DAC registers.

DAC_
DHR12L2

19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

20

21

22

Reserved

DACC1DHR[11:0]
DACC1DHR[11:0]

Reserved
Reserved
Reserved

Reserved

DACC1DHR[7:0]
DACC2DHR[11:0]
DACC2DHR[11:0]

RM0090 Rev 17

EN1

0x18

Reserved

SWTRIG1

DAC_
DHR12R2

Reserved

TEN1

0x14

WAVE TSEL1[2
1[2:0]
:0]

BOFF1

DAC_
DHR8R1

MAMP1[3:0]

SWTRIG2

0x10

DMAEN1

DAC_
DHR12L1

DMAUDRIE1

0x0C

Reserved

DAC_
DHR12R1

TSEL2[2:0]

EN2

0x08

WAVE
2[2:0]

BOFF2

DAC_
SWTRIGR

MAMP2[3:0]

TEN2

0x04

DMAEN2

DAC_CR

DMAUDRIE2

0x00

Reserved

Offset Register

31
30
29
28
27
26
25
24
23

Table 76. DAC register map

Reserved

453/1747
454

Digital-to-analog converter (DAC)

RM0090

19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

20

21

22

Offset Register

31
30
29
28
27
26
25
24
23

Table 76. DAC register map (continued)

0x1C

DAC_
DHR8R2

0x20

DAC_
DHR12RD

0x24

DAC_
DHR12LD

0x28

DAC_
DHR8RD

0x2C

DAC_
DOR1

Reserved

DACC1DOR[11:0]

0x30

DAC_
DOR2

Reserved

DACC2DOR[11:0]

0x34

DAC_SR

Reserved
DACC2DHR[11:0]

Reserved
Reserved

DACC2DHR[11:0]

DMAUDR2

Reserved

Reserved

DACC1DHR[11:0]
DACC1DHR[11:0]

DACC2DHR[7:0]

Reserved

DMAUDR1

Reserved

DACC2DHR[7:0]

DACC1DHR[7:0]

Reserved

Refer to Section 2.3: Memory map for the register boundary addresses.

454/1747

RM0090 Rev 17

Reserved

RM0090

15

Digital camera interface (DCMI)

Digital camera interface (DCMI)
This section applies to all STM32F4xx devices, unless otherwise specified.

15.1

DCMI introduction
The digital camera is a synchronous parallel interface able to receive a high-speed data flow
from an external 8-, 10-, 12- or 14-bit CMOS camera module. It supports different data
formats: YCbCr4:2:2/RGB565 progressive video and compressed data (JPEG).
This interface is for use with black & white cameras, X24 and X5 cameras, and it is
assumed that all pre-processing like resizing is performed in the camera module.

15.2

15.3

DCMI main features
•

8-, 10-, 12- or 14-bit parallel interface

•

Embedded/external line and frame synchronization

•

Continuous or snapshot mode

•

Crop feature

•

Supports the following data formats:
–

8/10/12/14- bit progressive video: either monochrome or raw bayer

–

YCbCr 4:2:2 progressive video

–

RGB 565 progressive video

–

Compressed data: JPEG

DCMI pins
Table 77 shows the DCMI pins.
Table 77. DCMI pins
Name

15.4

Signal type

D[0:13]

Data inputs

HSYNC

Horizontal synchronization input

VSYNC

Vertical synchronization input

PIXCLK

Pixel clock input

DCMI clocks
The digital camera interface uses two clock domains PIXCLK and HCLK. The signals
generated with PIXCLK are sampled on the rising edge of HCLK once they are stable. An
enable signal is generated in the HCLK domain, to indicate that data coming from the
camera are stable and can be sampled. The minimum PIXCLK period must be higher than
2.5 HCLK periods.

RM0090 Rev 17

455/1747
479

Digital camera interface (DCMI)

15.5

RM0090

DCMI functional overview
The digital camera interface is a synchronous parallel interface that can receive high-speed
(up to 54 Mbytes/s) data flows. It consists of up to 14 data lines (D13-D0) and a pixel clock
line (PIXCLK). The pixel clock has a programmable polarity, so that data can be captured on
either the rising or the falling edge of the pixel clock.
The data are packed into a 32-bit data register (DCMI_DR) and then transferred through a
general-purpose DMA channel. The image buffer is managed by the DMA, not by the
camera interface.
The data received from the camera can be organized in lines/frames (raw YUB/RGB/Bayer
modes) or can be a sequence of JPEG images. To enable JPEG image reception, the JPEG
bit (bit 3 of DCMI_CR register) must be set.
The data flow is synchronized either by hardware using the optional HSYNC (horizontal
synchronization) and VSYNC (vertical synchronization) signals or by synchronization codes
embedded in the data flow.
Figure 72 shows the DCMI block diagram.
Figure 72. DCMI block diagram

'0$
LQWHUIDFH

&RQWURO6WDWXV
UHJLVWHU

$+%
LQWHUIDFH
),)2
'DWD
IRUPDWWHU

'DWD
H[WUDFWLRQ

6\QFKURQL]HU

'&0,B3,;&/.

'&0,B'>@'&0,B+6<1&'&0,B96<1&
DLE

456/1747

RM0090 Rev 17

RM0090

Digital camera interface (DCMI)
Figure 73. Top-level block diagram
$#-)?$;=
$#-)?0)8#,+

(#,+

$#-)?(39.#

%XTERNAL
INTERFACE

$#-)?639.#
$#-)
)NTERRUPT
CONTROLLER

$#-)?)4

$-!?2%1

AIB

15.5.1

DMA interface
The DMA interface is active when the CAPTURE bit in the DCMI_CR register is set. A DMA
request is generated each time the camera interface receives a complete 32-bit data block
in its register.

15.5.2

DCMI physical interface
The interface is composed of 11/13/15/17 inputs. Only the Slave mode is supported.
The camera interface can capture 8-bit, 10-bit, 12-bit or 14-bit data depending on the
EDM[1:0] bits in the DCMI_CR register. If less than 14 bits are used, the unused data pins
must not be assigned to DCMI interface through GPIO alternate functions.
Table 78. DCMI signals
Signal name
8 bits
10 bits
12 bits
14 bits

D[0..7]
D[0..9]
D[0..11]
D[0..13]

Signal description

Data

PIXCLK

Pixel clock

HSYNC

Horizontal synchronization / Data valid

VSYNC

Vertical synchronization

The data are synchronous with PIXCLK and change on the rising/falling edge of the pixel
clock depending on the polarity.
The HSYNC signal indicates the start/end of a line.
The VSYNC signal indicates the start/end of a frame

RM0090 Rev 17

457/1747
479

Digital camera interface (DCMI)

RM0090
Figure 74. DCMI signal waveforms

$#-)?0)8#,+

$#-)?$2;=

$#-)?(39.#

$#-)?639.#
AIB

1. The capture edge of DCMI_PIXCLK is the falling edge, the active state of DCMI_HSYNC and
DCMI_VSYNC is 1.
1. DCMI_HSYNC and DCMI_VSYNC can change states at the same time.

8-bit data
When EDM[1:0] in DCMI_CR are programmed to “00” the interface captures 8 LSB’s at its
input (D[0:7]) and stores them as 8-bit data. The D[13:8] inputs are ignored. In this case, to
capture a 32-bit word, the camera interface takes four pixel clock cycles.
The first captured data byte is placed in the LSB position in the 32-bit word and the 4th
captured data byte is placed in the MSB position in the 32-bit word. Table 79 gives an
example of the positioning of captured data bytes in two 32-bit words.
Table 79. Positioning of captured data bytes in 32-bit words (8-bit width)
Byte address

31:24

23:16

15:8

7:0

0

Dn+3[7:0]

Dn+2[7:0]

Dn+1[7:0]

Dn[7:0]

4

Dn+7[7:0]

Dn+6[7:0]

Dn+5[7:0]

Dn+4[7:0]

10-bit data
When EDM[1:0] in DCMI_CR are programmed to “01”, the camera interface captures 10-bit
data at its input D[0..9] and stores them as the 10 least significant bits of a 16-bit word. The
remaining most significant bits in the DCMI_DR register (bits 11 to 15) are cleared to zero.
So, in this case, a 32-bit data word is made up every two pixel clock cycles.
The first captured data are placed in the LSB position in the 32-bit word and the 2nd
captured data are placed in the MSB position in the 32-bit word as shown in Table 80.
Table 80. Positioning of captured data bytes in 32-bit words (10-bit width)

458/1747

Byte address

31:26

25:16

15:10

9:0

0

0

Dn+1[9:0]

0

Dn[9:0]

4

0

Dn+3[9:0]

0

Dn+2[9:0]

RM0090 Rev 17

RM0090

Digital camera interface (DCMI)

12-bit data
When EDM[1:0] in DCMI_CR are programmed to “10”, the camera interface captures the
12-bit data at its input D[0..11] and stores them as the 12 least significant bits of a 16-bit
word. The remaining most significant bits are cleared to zero. So, in this case a 32-bit data
word is made up every two pixel clock cycles.
The first captured data are placed in the LSB position in the 32-bit word and the 2nd
captured data are placed in the MSB position in the 32-bit word as shown in Table 81.
Table 81. Positioning of captured data bytes in 32-bit words (12-bit width)
Byte address

31:28

27:16

15:12

11:0

0

0

Dn+1[11:0]

0

Dn[11:0]

4

0

Dn+3[11:0]

0

Dn+2[11:0]

14-bit data
When EDM[1:0] in DCMI_CR are programmed to “11”, the camera interface captures the
14-bit data at its input D[0..13] and stores them as the 14 least significant bits of a 16-bit
word. The remaining most significant bits are cleared to zero. So, in this case a 32-bit data
word is made up every two pixel clock cycles.
The first captured data are placed in the LSB position in the 32-bit word and the 2nd
captured data are placed in the MSB position in the 32-bit word as shown in Table 82.
Table 82. Positioning of captured data bytes in 32-bit words (14-bit width)

15.5.3

Byte address

31:30

29:16

15:14

13:0

0

0

Dn+1[13:0]

0

Dn[13:0]

4

0

Dn+3[13:0]

0

Dn+2[13:0]

Synchronization
The digital camera interface supports embedded or hardware (HSYNC & VSYNC)
synchronization. When embedded synchronization is used, it is up to the digital camera
module to make sure that the 0x00 and 0xFF values are used ONLY for synchronization
(not in data). Embedded synchronization codes are supported only for the 8-bit parallel data
interface width (that is, in the DCMI_CR register, the EDM[1:0] bits should be cleared to
“00”).
For compressed data, the DCMI supports only the hardware synchronization mode. In this
case, VSYNC is used as a start/end of the image, and HSYNC is used as a Data Valid
signal. Figure 75 shows the corresponding timing diagram.

RM0090 Rev 17

459/1747
479

Digital camera interface (DCMI)

RM0090
Figure 75. Timing diagram
3DGGLQJGDWD
DWWKHHQGRIWKH-3(*VWUHDP

%HJLQQLQJRI-3(*VWUHDP

3URJUDPPDEOH
-3(*SDFNHWVL]H

-3(*GDWD
(QGRI-3(*VWUHDP
'&0,B+6<1&

'&0,B96<1&

3DFNHWGLVSDWFKLQJGHSHQGVRQWKHLPDJHFRQWHQW
7KLVUHVXOWVLQDYDULDEOHEODQNLQJGXUDWLRQ

-3(*SDFNHWGDWD
DLE

460/1747

RM0090 Rev 17

RM0090

Digital camera interface (DCMI)

Hardware synchronization mode
In hardware synchronisation mode, the two synchronization signals (HSYNC/VSYNC) are
used.
Depending on the camera module/mode, data may be transmitted during horizontal/vertical
synchronisation periods. The HSYNC/VSYNC signals act like blanking signals since all the
data received during HSYNC/VSYNC active periods are ignored.
In order to correctly transfer images into the DMA/RAM buffer, data transfer is synchronized
with the VSYNC signal. When the hardware synchronisation mode is selected, and capture
is enabled (CAPTURE bit set in DCMI_CR), data transfer is synchronized with the
deactivation of the VSYNC signal (next start of frame).
Transfer can then be continuous, with successive frames transferred by DMA to successive
buffers or the same/circular buffer. To allow the DMA management of successive frames, a
VSIF (Vertical synchronization interrupt flag) is activated at the end of each frame.

Embedded data synchronization mode
In this synchronisation mode, the data flow is synchronised using 32-bit codes embedded in
the data flow. These codes use the 0x00/0xFF values that are not used in data anymore.
There are 4 types of codes, all with a 0xFF0000XY format. The embedded synchronization
codes are supported only in 8-bit parallel data width capture (in the DCMI_CR register, the
EDM[1:0] bits should be programmed to “00”). For other data widths, this mode generates
unpredictable results and must not be used.
Note:

Camera modules can have 8 such codes (in interleaved mode). For this reason, the
interleaved mode is not supported by the camera interface (otherwise, every other halfframe would be discarded).
•

Mode 2
Four embedded codes signal the following events
–

Frame start (FS)

–

Frame end (FE)

–

Line start (LS)

–

Line end (LE)

The XY values in the 0xFF0000XY format of the four codes are programmable (see
Section 15.8.7: DCMI embedded synchronization code register (DCMI_ESCR)).
A 0xFF value programmed as a “frame end” means that all the unused codes are
interpreted as valid frame end codes.
In this mode, once the camera interface has been enabled, the frame capture starts
after the first occurrence of the frame end (FE) code followed by a frame start (FS)
code.
•

Mode 1
An alternative coding is the camera mode 1. This mode is ITU656 compatible.
The codes signal another set of events:
–

SAV (active line) - line start

–

EAV (active line) - line end

–

SAV (blanking) - end of line during interframe blanking period

–

EAV (blanking) - end of line during interframe blanking period

RM0090 Rev 17

461/1747
479

Digital camera interface (DCMI)

RM0090

This mode can be supported by programming the following codes:
•

FS ≤ 0xFF

•

FE ≤ 0xFF

•

LS ≤ SAV (active)

•

LE ≤ EAV (active)

An embedded unmask code is also implemented for frame/line start and frame/line end
codes. Using it, it is possible to compare only the selected unmasked bits with the
programmed code. You can therefore select a bit to compare in the embedded code and
detect a frame/line start or frame/line end. This means that there can be different codes for
the frame/line start and frame/line end with the unmasked bit position remaining the same.

Example
FS = 0xA5
Unmask code for FS = 0x10
In this case the frame start code is embedded in the bit 4 of the frame start code.

15.5.4

Capture modes
This interface supports two types of capture: snapshot (single frame) and continuous grab.

Snapshot mode (single frame)
In this mode, a single frame is captured (CM = ‘1’ in the DCMI_CR register). After the
CAPTURE bit is set in DCMI_CR, the interface waits for the detection of a start of frame
before sampling the data. The camera interface is automatically disabled (CAPTURE bit
cleared in DCMI_CR) after receiving the first complete frame. An interrupt is generated
(IT_FRAME) if it is enabled.
In case of an overrun, the frame is lost and the CAPTURE bit is cleared.
Figure 76. Frame capture waveforms in Snapshot mode

'&0,B+6<1&

'&0,B96<1&

)UDPHFDSWXUHG

)UDPH
QRWFDSWXUHG

DLE

1. Here, the active state of DCMI_HSYNC and DCMI_VSYNC is 1.
2. DCMI_HSYNC and DCMI_VSYNC can change states at the same time.

462/1747

RM0090 Rev 17

RM0090

Digital camera interface (DCMI)

Continuous grab mode
In this mode (CM bit = ‘0’ in DCMI_CR), once the CAPTURE bit has been set in DCMI_CR,
the grabbing process starts on the next VSYNC or embedded frame start depending on the
mode. The process continues until the CAPTURE bit is cleared in DCMI_CR. Once the
CAPTURE bit has been cleared, the grabbing process continues until the end of the current
frame.
Figure 77. Frame capture waveforms in continuous grab mode

'&0,B+6<1&

'&0,B96<1&

)UDPHFDSWXUHG

)UDPHFDSWXUHG

DLE

1. Here, the active state of DCMI_HSYNC and DCMI_VSYNC is 1.
2. DCMI_HSYNC and DCMI_VSYNC can change states at the same time.

In continuous grab mode, you can configure the FCRC bits in DCMI_CR to grab all pictures,
every second picture or one out of four pictures to decrease the frame capture rate.
Note:

In the hardware synchronization mode (ESS = ‘0’ in DCMI_CR), the IT_VSYNC interrupt is
generated (if enabled) even when CAPTURE = ‘0’ in DCMI_CR so, to reduce the frame
capture rate even further, the IT_VSYNC interrupt can be used to count the number of
frames between 2 captures in conjunction with the Snapshot mode. This is not allowed by
embedded data synchronization mode.

15.5.5

Crop feature
With the crop feature, the camera interface can select a rectangular window from the
received image. The start (upper left corner) coordinates and size (horizontal dimension in
number of pixel clocks and vertical dimension in number of lines) are specified using two 32bit registers (DCMI_CWSTRT and DCMI_CWSIZE). The size of the window is specified in
number of pixel clocks (horizontal dimension) and in number of lines (vertical dimension).

RM0090 Rev 17

463/1747
479

Digital camera interface (DCMI)

RM0090

Figure 78. Coordinates and size of the window after cropping

634 BIT IN $#-)?#3424

6,).% BIT IN $#-)?#3):%
(/&&#.4 BIT IN $#-)?#3424
#!0#.4 BIT IN $#-)?#3):%
AI

These registers specify the coordinates of the starting point of the capture window as a line
number (in the frame, starting from 0) and a number of pixel clocks (on the line, starting from
0), and the size of the window as a line number and a number of pixel clocks. The CAPCNT
value can only be a multiple of 4 (two least significant bits are forced to 0) to allow the
correct transfer of data through the DMA.
If the VSYNC signal goes active before the number of lines is specified in the
DCMI_CWSIZE register, then the capture stops and an IT_FRAME interrupt is generated
when enabled.
Figure 79. Data capture waveforms

$#-)?(39.#

$#-)?639.#
(/&&#.4
#!0#.4

$ATA NOT CAPTURED IN THIS PHASE
$ATA CAPTURED IN THIS PHASE
1. Here, the active state of DCMI_HSYNC and DCMI_VSYNC is 1.
2. DCMI_HSYNC and DCMI_VSYNC can change states at the same time.

464/1747

RM0090 Rev 17

AI

RM0090

15.5.6

Digital camera interface (DCMI)

JPEG format
To allow JPEG image reception, it is necessary to set the JPEG bit in the DCMI_CR register.
JPEG images are not stored as lines and frames, so the VSYNC signal is used to start the
capture while HSYNC serves as a data enable signal. The number of bytes in a line may not
be a multiple of 4, you should therefore be careful when handling this case since a DMA
request is generated each time a complete 32-bit word has been constructed from the
captured data. When an end of frame is detected and the 32-bit word to be transferred has
not been completely received, the remaining data are padded with ‘0s’ and a DMA request
is generated.
The crop feature and embedded synchronization codes cannot be used in the JPEG format.

15.5.7

FIFO
A four-word FIFO is implemented to manage data rate transfers on the AHB. The DCMI
features a simple FIFO controller with a read pointer incremented each time the camera
interface reads from the AHB, and a write pointer incremented each time the camera
interface writes to the FIFO. There is no overrun protection to prevent the data from being
overwritten if the AHB interface does not sustain the data transfer rate.
In case of overrun or errors in the synchronization signals, the FIFO is reset and the DCMI
interface waits for a new start of frame.

15.6

Data format description

15.6.1

Data formats
Three types of data are supported:
•

8-bit progressive video: either monochrome or raw Bayer format

•

YCbCr 4:2:2 progressive video

•

RGB565 progressive video. A pixel coded in 16 bits (5 bits for blue, 5 bits for red, 6 bits
for green) takes two clock cycles to be transferred.

Compressed data: JPEG
For B&W, YCbCr or RGB data, the maximum input size is 2048 × 2048 pixels. No limit in
JPEG compressed mode.
For monochrome, RGB & YCbCr, the frame buffer is stored in raster mode. 32-bit words are
used. Only the little endian format is supported.
Figure 80. Pixel raster scan order
7ORD  7ORD 7ORD

0IXEL ROW 

0IXEL RASTER
SCAN ORDER
INCREASING
ADDRESSES

0IXEL ROW N n 
AI

RM0090 Rev 17

465/1747
479

Digital camera interface (DCMI)

15.6.2

RM0090

Monochrome format
Characteristics:
•

Raster format

•

8 bits per pixel

Table 83 shows how the data are stored.
Table 83. Data storage in monochrome progressive video format

15.6.3

Byte address

31:24

23:16

15:8

7:0

0

n+3

n+2

n+1

n

4

n+7

n+6

n+5

n+4

RGB format
Characteristics:
•

Raster format

•

RGB

•

Interleaved: one buffer: R, G & B interleaved: BRGBRGBRG, etc.

•

Optimized for display output

The RGB planar format is compatible with standard OS frame buffer display formats.
Only 16 BPP (bits per pixel): RGB565 (2 pixels per 32-bit word) is supported.
The 24 BPP (palletized format) and grayscale formats are not supported. Pixels are stored
in a raster scan order, that is from top to bottom for pixel rows, and from left to right within a
pixel row. Pixel components are R (red), G (green) and B (blue). All components have the
same spatial resolution (4:4:4 format). A frame is stored in a single part, with the
components interleaved on a pixel basis.
Table 84 shows how the data are stored.
Table 84. Data storage in RGB progressive video format

15.6.4

Byte address

31:27

26:21

20:16

15:11

10:5

4:0

0

Red n + 1

Green n + 1

Blue n + 1

Red n

Green n

Blue n

4

Red n + 4

Green n + 3

Blue n + 3

Red n + 2

Green n + 2

Blue n + 2

YCbCr format
Characteristics:
•

Raster format

•

YCbCr 4:2:2

•

Interleaved: one Buffer: Y, Cb & Cr interleaved: CbYCrYCbYCr, etc.

Pixel components are Y (luminance or “luma”), Cb and Cr (chrominance or “chroma” blue
and red). Each component is encoded in 8 bits. Luma and chroma are stored together
(interleaved) as shown in Table 85.

466/1747

RM0090 Rev 17

RM0090

Digital camera interface (DCMI)
Table 85. Data storage in YCbCr progressive video format

15.7

Byte address

31:24

23:16

15:8

7:0

0

Yn+1

Cr n

Yn

Cb n

4

Yn+3

Cr n + 2

Yn+2

Cb n + 2

DCMI interrupts
Five interrupts are generated. All interrupts are maskable by software. The global interrupt
(IT_DCMI) is the OR of all the individual interrupts. Table 86 gives the list of all interrupts.
Table 86. DCMI interrupts
Interrupt name

15.8

Interrupt event

IT_LINE

Indicates the end of line

IT_FRAME

Indicates the end of frame capture

IT_OVR

indicates the overrun of data reception

IT_VSYNC

Indicates the synchronization frame

IT_ERR

Indicates the detection of an error in the embedded synchronization frame
detection

IT_DCMI

Logic OR of the previous interrupts

DCMI register description
All DCMI registers have to be accessed as 32-bit words, otherwise a bus error occurs.

15.8.1

DCMI control register 1 (DCMI_CR)
Address offset: 0x00

3

2

CM

rw

rw

4

CAPTURE

rw

5

rw

rw

rw

ESS

FCRC

6

JPEG

rw

7

HSPOL

rw

8

PCKPOL

rw

EDM

9

VSPOL

Reserved

Reserved

ENABLE

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

CROP

Reset value: 0x0000 0x0000

rw

rw

rw

rw

1

0

Bit 31:15 Reserved, must be kept at reset value.
Bit 14 ENABLE: DCMI enable
0: DCMI disabled
1: DCMI enabled
Note: The DCMI configuration registers should be programmed correctly before
enabling this Bit
Bit 13: 12 Reserved, must be kept at reset value.

RM0090 Rev 17

467/1747
479

Digital camera interface (DCMI)

RM0090

Bits 11:10 EDM[1:0]: Extended data mode
00: Interface captures 8-bit data on every pixel clock
01: Interface captures 10-bit data on every pixel clock
10: Interface captures 12-bit data on every pixel clock
11: Interface captures 14-bit data on every pixel clock
Bits 9:8 FCRC[1:0]: Frame capture rate control
These bits define the frequency of frame capture. They are meaningful only in
Continuous grab mode. They are ignored in snapshot mode.
00: All frames are captured
01: Every alternate frame captured (50% bandwidth reduction)
10: One frame in 4 frames captured (75% bandwidth reduction)
11: reserved
Bit 7 VSPOL: Vertical synchronization polarity
This bit indicates the level on the VSYNC pin when the data are not valid on the
parallel interface.
0: VSYNC active low
1: VSYNC active high
Bit 6 HSPOL: Horizontal synchronization polarity
This bit indicates the level on the HSYNC pin when the data are not valid on the
parallel interface.
0: HSYNC active low
1: HSYNC active high
Bit 5 PCKPOL: Pixel clock polarity
This bit configures the capture edge of the pixel clock
0: Falling edge active.
1: Rising edge active.
Bit 4 ESS: Embedded synchronization select
0: Hardware synchronization data capture (frame/line start/stop) is synchronized
with the HSYNC/VSYNC signals.
1: Embedded synchronization data capture is synchronized with synchronization
codes embedded in the data flow.
Note: Valid only for 8-bit parallel data. HSPOL/VSPOL are ignored when the ESS
bit is set.
This bit is disabled in JPEG mode.
Bit 3 JPEG: JPEG format
0: Uncompressed video format
1: This bit is used for JPEG data transfers. The HSYNC signal is used as data
enable. The crop and embedded synchronization features (ESS bit) cannot be
used in this mode.
Bits 2 CROP: Crop feature
0: The full image is captured. In this case the total number of bytes in an image
frame should be a multiple of 4
1: Only the data inside the window specified by the crop register will be captured.
If the size of the crop window exceeds the picture size, then only the picture size
is captured.

468/1747

RM0090 Rev 17

RM0090

Digital camera interface (DCMI)

Bit 1 CM: Capture mode
0: Continuous grab mode - The received data are transferred into the destination
memory through the DMA. The buffer location and mode (linear or circular
buffer) is controlled through the system DMA.
1: Snapshot mode (single frame) - Once activated, the interface waits for the
start of frame and then transfers a single frame through the DMA. At the end of
the frame, the CAPTURE bit is automatically reset.
Bit 0 CAPTURE: Capture enable
0: Capture disabled.
1: Capture enabled.
The camera interface waits for the first start of frame, then a DMA request is
generated to transfer the received data into the destination memory.
In snapshot mode, the CAPTURE bit is automatically cleared at the end of the
1st frame received.
In continuous grab mode, if the software clears this bit while a capture is
ongoing, the bit will be effectively cleared after the frame end.
Note: The DMA controller and all DCMI configuration registers should be
programmed correctly before enabling this bit.

RM0090 Rev 17

469/1747
479

Digital camera interface (DCMI)

15.8.2

RM0090

DCMI status register (DCMI_SR)
Address offset: 0x04

8

7

6

5

4

3

2

1

0

VSYNC

Reserved

9

HSYNC

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

FNE

Reset value: 0x0000 0x0000

r

r

r

Bits 31:3 Reserved, must be kept at reset value.
Bit 2 FNE: FIFO not empty
This bit gives the status of the FIFO
1: FIFO contains valid data
0: FIFO empty
Bit 1 VSYNC
This bit gives the state of the VSYNC pin with the correct programmed polarity.
When embedded synchronization codes are used, the meaning of this bit is the
following:
0: active frame
1: synchronization between frames
In case of embedded synchronization, this bit is meaningful only if the
CAPTURE bit in DCMI_CR is set.
Bit 0 HSYNC
This bit gives the state of the HSYNC pin with the correct programmed polarity.
When embedded synchronization codes are used, the meaning of this bit is the
following:
0: active line
1: synchronization between lines
In case of embedded synchronization, this bit is meaningful only if the
CAPTURE bit in DCMI_CR is set.

470/1747

RM0090 Rev 17

RM0090

15.8.3

Digital camera interface (DCMI)

DCMI raw interrupt status register (DCMI_RIS)
Address offset: 0x08

6

5

4

3

2

1

0

OVR_RIS

7

FRAME_RIS

8

ERR_RIS

9

LINE_RIS

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

VSYNC_RIS

Reset value: 0x0000 0x0000

r

r

r

r

r

DCMI_RIS gives the raw interrupt status and is accessible in read only. When read, this
register returns the status of the corresponding interrupt before masking with the DCMI_IER
register value.
Bits 31:5 Reserved, must be kept at reset value.
Bit 4 LINE_RIS: Line raw interrupt status
This bit gets set when the HSYNC signal changes from the inactive state to the
active state. It goes high even if the line is not valid.
In the case of embedded synchronization, this bit is set only if the CAPTURE bit
in DCMI_CR is set.
It is cleared by writing a ‘1’ to the LINE_ISC bit in DCMI_ICR.
Bit 3 VSYNC_RIS: VSYNC raw interrupt status
This bit is set when the VSYNC signal changes from the inactive state to the
active state.
In the case of embedded synchronization, this bit is set only if the CAPTURE bit
is set in DCMI_CR.
It is cleared by writing a ‘1’ to the VSYNC_ISC bit in DCMI_ICR.
Bit 2 ERR_RIS: Synchronization error raw interrupt status
0: No synchronization error detected
1: Embedded synchronization characters are not received in the correct order.
This bit is valid only in the embedded synchronization mode. It is cleared by
writing a ‘1’ to the ERR_ISC bit in DCMI_ICR.
Note: This bit is available only in embedded synchronization mode.
Bit 1 OVR_RIS: Overrun raw interrupt status
0: No data buffer overrun occurred
1: A data buffer overrun occurred and the data FIFO is corrupted.
This bit is cleared by writing a ‘1’ to the OVR_ISC bit in DCMI_ICR.
Bit 0 FRAME_RIS: Capture complete raw interrupt status
0: No new capture
1: A frame has been captured.
This bit is set when a frame or window has been captured.
In case of a cropped window, this bit is set at the end of line of the last line in the
crop. It is set even if the captured frame is empty (e.g. window cropped outside
the frame).
This bit is cleared by writing a ‘1’ to the FRAME_ISC bit in DCMI_ICR.

RM0090 Rev 17

471/1747
479

Digital camera interface (DCMI)

15.8.4

RM0090

DCMI interrupt enable register (DCMI_IER)
Address offset: 0x0C

6

5

3

2

OVR_IE

7

FRAME_IE

8

ERR_IE

Reserved

9

LINE_IE

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

VSYNC_IE

Reset value: 0x0000 0x0000
4

1

0

rw

rw

rw

rw

rw

The DCMI_IER register is used to enable interrupts. When one of the DCMI_IER bits is set,
the corresponding interrupt is enabled. This register is accessible in both read and write.
Bits 31:5 Reserved, must be kept at reset value.
Bit 4 LINE_IE: Line interrupt enable
0: No interrupt generation when the line is received
1: An Interrupt is generated when a line has been completely received
Bit 3 VSYNC_IE: VSYNC interrupt enable
0: No interrupt generation
1: An interrupt is generated on each VSYNC transition from the inactive to the
active state
The active state of the VSYNC signal is defined by the VSPOL bit.
Bit 2 ERR_IE: Synchronization error interrupt enable
0: No interrupt generation
1: An interrupt is generated if the embedded synchronization codes are not
received in the correct order.
Note: This bit is available only in embedded synchronization mode.
Bit 1 OVR_IE: Overrun interrupt enable
0: No interrupt generation
1: An interrupt is generated if the DMA was not able to transfer the last data
before new data (32-bit) are received.
Bit 0 FRAME_IE: Capture complete interrupt enable
0: No interrupt generation
1: An interrupt is generated at the end of each received frame/crop window (in
crop mode).

472/1747

RM0090 Rev 17

RM0090

15.8.5

Digital camera interface (DCMI)

DCMI masked interrupt status register (DCMI_MIS)
This DCMI_MIS register is a read-only register. When read, it returns the current masked
status value (depending on the value in DCMI_IER) of the corresponding interrupt. A bit in
this register is set if the corresponding enable bit in DCMI_IER is set and the corresponding
bit in DCMI_RIS is set.
Address offset: 0x10

6

5

4

3

2

1

0

OVR_MIS

7

FRAME_MIS

8

ERR_MIS

Reserved

9

LINE_MIS

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

VSYNC_MIS

Reset value: 0x0000

r

r

r

r

r

Bits 31:5 Reserved, must be kept at reset value.
Bit 4 LINE_MIS: Line masked interrupt status
This bit gives the status of the masked line interrupt
0: No interrupt generation when the line is received
1: An Interrupt is generated when a line has been completely received and the
LINE_IE bit is set in DCMI_IER.
Bit 3 VSYNC_MIS: VSYNC masked interrupt status
This bit gives the status of the masked VSYNC interrupt
0: No interrupt is generated on VSYNC transitions
1: An interrupt is generated on each VSYNC transition from the inactive to the
active state and the VSYNC_IE bit is set in DCMI_IER.
The active state of the VSYNC signal is defined by the VSPOL bit.
Bit 2 ERR_MIS: Synchronization error masked interrupt status
This bit gives the status of the masked synchronization error interrupt
0: No interrupt is generated on a synchronization error
1: An interrupt is generated if the embedded synchronization codes are not
received in the correct order and the ERR_IE bit in DCMI_IER is set.
Note: This bit is available only in embedded synchronization mode.
Bit 1 OVR_MIS: Overrun masked interrupt status
This bit gives the status of the masked overflow interrupt
0: No interrupt is generated on overrun
1: An interrupt is generated if the DMA was not able to transfer the last data
before new data (32-bit) are received and the OVR_IE bit is set in DCMI_IER.
Bit 0 FRAME_MIS: Capture complete masked interrupt status
This bit gives the status of the masked capture complete interrupt
0: No interrupt is generated after a complete capture
1: An interrupt is generated at the end of each received frame/crop window (in
crop mode) and the FRAME_IE bit is set in DCMI_IER.

RM0090 Rev 17

473/1747
479

Digital camera interface (DCMI)

15.8.6

RM0090

DCMI interrupt clear register (DCMI_ICR)
Address offset: 0x14

6

Reserved

5

4

3

2

1

0

OVR_ISC

7

FRAME_ISC

8

ERR_ISC

9

LINE_ISC

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

VSYNC_ISC

Reset value: 0x0000 0x0000

w w w w w

The DCMI_ICR register is write-only. Writing a ‘1’ into a bit of this register clears the
corresponding bit in the DCMI_RIS and DCMI_MIS registers. Writing a ‘0’ has no effect.
Bits 15:5 Reserved, must be kept at reset value.
Bit 4 LINE_ISC: line interrupt status clear
Writing a ‘1’ into this bit clears LINE_RIS in the DCMI_RIS register
Bit 3 VSYNC_ISC: Vertical synch interrupt status clear
Writing a ‘1’ into this bit clears the VSYNC_RIS bit in DCMI_RIS
Bit 2 ERR_ISC: Synchronization error interrupt status clear
Writing a ‘1’ into this bit clears the ERR_RIS bit in DCMI_RIS
Note: This bit is available only in embedded synchronization mode.
Bit 1 OVR_ISC: Overrun interrupt status clear
Writing a ‘1’ into this bit clears the OVR_RIS bit in DCMI_RIS
Bits 0 FRAME_ISC: Capture complete interrupt status clear
Writing a ‘1’ into this bit clears the FRAME_RIS bit in DCMI_RIS

474/1747

RM0090 Rev 17

RM0090

Digital camera interface (DCMI)

15.8.7

DCMI embedded synchronization code register (DCMI_ESCR)
Address offset: 0x18
Reset value: 0x0000 0x0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
FEC
rw

rw

rw

rw

rw

LEC
rw

rw

rw

rw

rw

rw

rw

rw

9

8

7

6

5

LSC
rw

rw

rw

rw

rw

rw

rw

rw

4

3

2

1

0

rw

rw

rw

FSC
rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:24 FEC: Frame end delimiter code
This byte specifies the code of the frame end delimiter. The code consists of 4
bytes in the form of 0xFF, 0x00, 0x00, FEC.
If FEC is programmed to 0xFF, all the unused codes (0xFF0000XY) are
interpreted as frame end delimiters.
Bits 23:16 LEC: Line end delimiter code
This byte specifies the code of the line end delimiter. The code consists of 4
bytes in the form of 0xFF, 0x00, 0x00, LEC.
Bits 15:8 LSC: Line start delimiter code
This byte specifies the code of the line start delimiter. The code consists of 4
bytes in the form of 0xFF, 0x00, 0x00, LSC.
Bits 7:0 FSC: Frame start delimiter code
This byte specifies the code of the frame start delimiter. The code consists of 4
bytes in the form of 0xFF, 0x00, 0x00, FSC.
If FSC is programmed to 0xFF, no frame start delimiter is detected. But, the 1st
occurrence of LSC after an FEC code will be interpreted as a start of frame
delimiter.

RM0090 Rev 17

475/1747
479

Digital camera interface (DCMI)

15.8.8

RM0090

DCMI embedded synchronization unmask register (DCMI_ESUR)
Address offset: 0x1C
Reset value: 0x0000 0x0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
FEU
rw

rw

rw

rw

rw

LEU
rw

rw

rw

rw

rw

rw

rw

rw

9

8

7

6

5

LSU
rw

rw

rw

rw

rw

rw

rw

rw

4

3

2

1

0

rw

rw

rw

FSU
rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:24 FEU: Frame end delimiter unmask
This byte specifies the mask to be applied to the code of the frame end delimiter.
0: The corresponding bit in the FEC byte in DCMI_ESCR is masked while
comparing the frame end delimiter with the received data.
1: The corresponding bit in the FEC byte in DCMI_ESCR is compared while
comparing the frame end delimiter with the received data
Bits 23:16 LEU: Line end delimiter unmask
This byte specifies the mask to be applied to the code of the line end delimiter.
0: The corresponding bit in the LEC byte in DCMI_ESCR is masked while
comparing the line end delimiter with the received data
1: The corresponding bit in the LEC byte in DCMI_ESCR is compared while
comparing the line end delimiter with the received data
Bits 15:8 LSU: Line start delimiter unmask
This byte specifies the mask to be applied to the code of the line start delimiter.
0: The corresponding bit in the LSC byte in DCMI_ESCR is masked while
comparing the line start delimiter with the received data
1: The corresponding bit in the LSC byte in DCMI_ESCR is compared while
comparing the line start delimiter with the received data
Bits 7:0 FSU: Frame start delimiter unmask
This byte specifies the mask to be applied to the code of the frame start
delimiter.
0: The corresponding bit in the FSC byte in DCMI_ESCR is masked while
comparing the frame start delimiter with the received data
1: The corresponding bit in the FSC byte in DCMI_ESCR is compared while
comparing the frame start delimiter with the received data

476/1747

RM0090 Rev 17

RM0090

Digital camera interface (DCMI)

15.8.9

DCMI crop window start (DCMI_CWSTRT)
Address offset: 0x20
Reset value: 0x0000 0x0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

VST[12:0
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Reserv
ed
rw
rw

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

HOFFCNT[13:0]
rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:29 Reserved, must be kept at reset value.
Bits 28:16 VST[12:0]: Vertical start line count
The image capture starts with this line number. Previous line data are ignored.
0x0000 => line 1
0x0001 => line 2
0x0002 => line 3
....
Bits 15:14 Reserved, must be kept at reset value.
Bit 13:0 HOFFCNT[13:0]: Horizontal offset count
This value gives the number of pixel clocks to count before starting a capture.

15.8.10

DCMI crop window size (DCMI_CWSIZE)
Address offset: 0x24
Reset value: 0x0000 0x0000

VLINE13:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Reserved

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

CAPCNT[13:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:30 Reserved, must be kept at reset value.
Bitst 29:16 VLINE[13:0]: Vertical line count
This value gives the number of lines to be captured from the starting point.
0x0000 => 1 line
0x0001 => 2 lines
0x0002 => 3 lines
....
Bits 15:14 Reserved, must be kept at reset value.
Bits 13:0 CAPCNT[13:0]: Capture count
This value gives the number of pixel clocks to be captured from the starting
point on the same line. It value should corresponds to word-aligned data for
different widths of parallel interfaces.
0x0000 => 1 pixel
0x0001 => 2 pixels
0x0002 => 3 pixels
....

RM0090 Rev 17

477/1747
479

Digital camera interface (DCMI)

15.8.11

RM0090

DCMI data register (DCMI_DR)
Address offset: 0x28
Reset value: 0x0000 0x0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Byte3
r

r

r

r

r

Byte2
r

r

r

r

r

r

r

9

8

7

6

5

Byte1

r

r

r

r

r

r

r

r

4

3

2

1

0

r

r

r

Byte0

r

r

r

r

r

r

r

r

r

Bits 31:24 Data byte 3
Bits 23:16 Data byte 2
Bits 15:8 Data byte 1
Bits 7:0 Data byte 0

The digital camera Interface packages all the received data in 32-bit format before
requesting a DMA transfer. A 4-word deep FIFO is available to leave enough time for DMA
transfers and avoid DMA overrun conditions.

15.8.12

DCMI register map
Table 87 summarizes the DCMI registers.

2

1

0

CM

CAPTURE

0

0

0

0

0

0

DCMI_SR
Reserved

0x04

478/1747

RM0090 Rev 17

ERR_RIS

OVR_RIS

FRAME_RIS
FRAME_IE

VSYNC_RIS

0

0

0

0

0
FRAME_MIS

Reset value

OVR_IE

Reserved

0

OVR_MIS

DCMI_MIS
0x10

0

ERR_IE

Reset value

0

ERR_MIS

Reserved

0
VSYNC_IE

DCMI_IER
0x0C

0

VSYNC_MIS

Reset value

0

LINE_IE

Reserved

0

LINE_MIS

DCMI_RIS
0x08

0
LINE_RIS

Reset value

HSYNC

3

CROP

0

VSYNC

4

JPEG

0

FNE

5

ESS

0

6

0

0

PCKPOL

8

9

11

12

13

10

0

7

0

FCR
C

HSPOL

Reset value

EDM

VSPOL

Reserved

Reserved

14

15

16

17

18

20

21

22

23

24

25

26

27

28

19

ENABLE

DCMI_CR
0x00

29

Register

30

Offset

31

Table 87. DCMI register map and reset values

0

0

0

0

0

RM0090

Digital camera interface (DCMI)

4

3

2

1

0

ERR_ISC

OVR_ISC

FRAME_ISC

Reserved

VSYNC_ISC

5

DCMI_ICR
0x14

LINE_ISC

6

7

8

9

11

10

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

Register

30

Offset

31

Table 87. DCMI register map and reset values (continued)

0

0

0

0

0

Reset value
FEC

DCMI_ESCR

LEC

LSC

FSC

0x18
Reset value

0

0

0

0

0

0

0

0

0

0

0

FEU

DCMI_ESUR

0

0

0

0

0

0

0

0

LEU

0

0

0

0

0

0

0

0

LSU

0

0

0

0

0

FSU

0x1C
0

0

0

0

0

0

0

0x20

T

E
Reset value

0
Reserved

0x24

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

VLINE13:0]
0

0

0

0

0

0

0

0

0

Byte3

DCMI_DR

0

0

0

0

0

0

VST[12:0

Reserve
d

Reset value
DCMI_CWSIZ

0

Reserved

0

Reserved

Reset value
DCMI_CWSTR

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

HOFFCNT[13:0]
0

0

0

0

0

0

0

0

0

CAPCNT[13:0]
0

Byte2

0

0

0

0

0

0

0

0

Byte1

Byte0

0x28
Reset value

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Refer to Section 2.3: Memory map for the register boundary addresses.

RM0090 Rev 17

479/1747
479

LCD-TFT controller (LTDC)

16

RM0090

LCD-TFT controller (LTDC)
This section applies only to STM32F429xx/439xx devices.

16.1

Introduction
The LCD-TFT (Liquid Crystal Display - Thin Film Transistor) display controller provides a
parallel digital RGB (Red, Green, Blue) and signals for horizontal, vertical synchronisation,
Pixel Clock and Data Enable as output to interface directly to a variety of LCD and TFT
panels.

16.2

LTDC main features
•

24-bit RGB Parallel Pixel Output; 8 bits-per-pixel (RGB888)

•

2 display layers with dedicated FIFO (64x32-bit)

•

Color Look-Up Table (CLUT) up to 256 color (256x24-bit) per layer

•

Supports up to XGA (1024x768) resolution

•

Programmable timings for different display panels

•

Programmable Background color

•

Programmable polarity for HSync, VSync and Data Enable

•

Up to 8 Input color formats selectable per layer

•

–

ARGB8888

–

RGB888

–

RGB565

–

ARGB1555

–

ARGB4444

–

L8 (8-bit Luminance or CLUT)

–

AL44 (4-bit alpha + 4-bit luminance)

–

AL88 (8-bit alpha + 8-bit luminance)

Pseudo-random dithering output for low bits per channel
–

480/1747

Dither width 2-bits for Red, Green, Blue

•

Flexible blending between two layers using alpha value (per pixel or constant)

•

Color Keying (transparency color)

•

Programmable Window position and size

•

Supports thin film transistor (TFT) color displays

•

AHB master interface with burst of 16 words

•

Up to 4 programmable interrupt events

RM0090 Rev 17

RM0090

LCD-TFT controller (LTDC)

16.3

LTDC functional description

16.3.1

LTDC block diagram
The block diagram of the LTDC is shown in Figure 81: LTDC block diagram.
Figure 81. LTDC block diagram

$3%FORFNGRPDLQ

$+%FORFNGRPDLQ

3L[HOFORFNGRPDLQ

/D\HU
),)2

3)&

/&'B+6<1&
%OHQGLQJ 'LWKHULQJ
XQLW
XQLW

$+%
LQWHUIDFH
/D\HU
),)2

3)&

/&'B96<1&
/&'B'(
/&'B&/.
/&'B5>@

/&'7)7
SDQHO

/&'B*>@
&RQILJXUDWLRQ
DQGVWDWXV
UHJLVWHUV

7LPLQJ
JHQHUDWRU

/&'B%>@

,QWHUUXSWV
06Y9

Layer FIFO: One FIFO 64x32 bit per layer.
PFC: Pixel Format Convertor performing the pixel format conversion from the selected input
pixel format of a layer to words.
AHB interface: For data transfer from memories to the FIFO.
Blending, Dithering unit and Timings Generator: Refer to Section 16.4.1 and Section 16.4.2.

16.3.2

LTDC reset and clocks
The LCD-TFT controller peripheral uses 3 clock domains:
•

The AHB clock domain (HCLK) is used for data transfer from the memories to the
Layer FIFO and frame buffer configuration register

•

The APB2 clock domain (PCLK2) is used for global configuration register and interrupt
registers

•

The Pixel Clock domain (LCD_CLK) is used to generate LCD-TFT interface signals,
pixel data generation and layer configuration. The LCD_CLK output should be
configured following the panel requirements. The LCD_CLK is configured through the
PLLSAI (refer to RCC section).

Table 88 summarizes the clock domain for each register.

RM0090 Rev 17

481/1747
514

LCD-TFT controller (LTDC)

RM0090
Table 88. LTDC registers versus clock domain

LTDC registers

Clock domain

LTDC_LxCR
LTDC_LxCFBAR

HCLK

LTDC_LxCFBLR
LTDC_LxCFBLNR
LTDC_SRCR
LTDC_IER

PCLK2

LTDC_ISR
LTDC_ICR
LTDC_SSCR
LTDC_BPCR
LTDC_AWCR
LTDC_TWCR
LTDC_GCR
LTDC_BCCR
LTDC_LIPCR
LTDC_CPSR

Pixel Clock (LCD_CLK)

LTDC_CDSR
LTDC_LxWHPCR
LTDC_LxWVPCR
LTDC_LxCKCR
LTDC_LxPFCR
LTDC_LxCACR
LTDC_LxDCCR
LTDC_LxBFCR
LTDC_LxCLUTWR

Care must be taken when accessing the LTDC registers since the APB2 bus is stalling when
the following operations are ongoing:
•

Register write access and update for 6 xPCKL2 period + 5x LCD_CLK period (5x
HCLK period for register on AHB clock domain)

•

Register read access for 7xPCKL2 period + 5x LCD_CLK period (5x HCLK period for
register on AHB clock domain).

For registers on PCLK2 clock domain, APB2 bus is stalling during the register write access
for 6 xPCKL2 period and 7xPCKL2 period for read access.
The LCD controller can be reset by setting the corresponding bit in the RCC_APB2RSTR
register. It resets the three clock domains.

482/1747

RM0090 Rev 17

RM0090

16.3.3

LCD-TFT controller (LTDC)

LCD-TFT pins and signal interface
The Table below summarizes the LTDC signal interface:
Table 89. LCD-TFT pins and signal interface
LCD-TFT
signals

I/O

LCD_CLK

O

Clock Output

LCD_HSYNC

O

Horizontal Synchronization

LCD_VSYNC

O

Vertical Synchronization

LCD_DE

O

Not Data Enable

LCD_R[7:0]

O

Data: 8-bit Red data

LCD_G[7:0]

O

Data: 8-bit Green data

LCD_B[7:0]

O

Data: 8-bit Blue data

Description

The LCD-TFT controller pins must be configured by the user application. The unused pins
can be used for other purposes.
For LTDC outputs up to 24-bit (RGB888), if less than 8bpp are used to output for example
RGB565 or RGB666 to interface on 16b-bit or 18-bit displays, the RGB display data lines
must be connected to the MSB of the LCD-TFT controller RGB data lines. As an example, in
the case of an LCD-TFT controller interfacing with a RGB565 16-bit display, the LCD display
R[4:0], G[5:0] and B[4:0] data lines pins must be connected to LCD-TFT controller
LCD_R[7:3], LCD_G[7:2] and LCD_B[7:3].

16.4

LTDC programmable parameters
The LCD-TFT controller provides flexible configurable parameters. It can be enabled or
disabled through the LTDC_GCR register.

16.4.1

LTDC Global configuration parameters
Synchronous Timings:
Figure 82 presents the configurable timing parameters generated by the Synchronous
Timings Generator block presented in the block diagram Figure 81. It generates the
Horizontal and Vertical Synchronization timings panel signals, the Pixel Clock and the Data
Enable signals.

RM0090 Rev 17

483/1747
514

LCD-TFT controller (LTDC)

RM0090
Figure 82. LCD-TFT Synchronous timings

+6<1&
ZLGWK

7RWDOZLGWK
+%3

+)3
$FWLYHZLGWK

96<1&ZLGWK
9%3

7RWDOKHLJKW

'DWD/LQH

$FWLYHGLVSOD\DUHD

$FWLYHKHLJKW

'DWD Q /LQH Q
9)3

06Y9

Note:

The HBP and HFP are respectively the Horizontal back porch and front porch period.
The VBP and the VFP are respectively the Vertical back porch and front porch period.
The LCD-TFT programmable synchronous timings are:

Note:

484/1747

–

HSYNC and VSYNC Width: Horizontal and Vertical Synchronization width
configured by programming a value of HSYNC Width - 1 and VSYNC Width - 1 in
the LTDC_SSCR register.

–

HBP and VBP: Horizontal and Vertical Synchronization back porch width
configured by programming the accumulated value HSYNC Width + HBP - 1 and
the accumulated value VSYNC Width + VBP - 1 in the LTDC_BPCR register.

–

Active Width and Active Height: The Active Width and Active Height are
configured by programming the accumulated value HSYNC Width + HBP +
Active Width - 1 and the accumulated value VSYNC Width + VBP + Active
Height - 1 in the LTDC_AWCR register (only up to 1024x768 is supported).

–

Total Width: The Total width is configured by programming the accumulated value
HSYNC Width + HBP + Active Width + HFP - 1 in the LTDC_TWCR register. The
HFP is the Horizontal front porch period.

–

Total Height: The Total Height is configured by programming the accumulated
value VSYNC Height + VBP + Active Height + VFP - 1 in the LTDC_TWCR
register. The VFP is the Vertical front porch period.

When the LTDC is enabled, the timings generated start with X/Y=0/0 position as the first
horizontal synchronization pixel in the vertical synchronization area and following the back
porch, active data display area and the front porch.

RM0090 Rev 17

RM0090

LCD-TFT controller (LTDC)
When the LTDC is disabled, the timing generator block is reset to X=Total Width - 1,
Y=Total Height - 1 and held the last pixel before the vertical synchronization phase and the
FIFO are flushed. Therefore only blanking data is output continuously.

Example of Synchronous timings configuration
TFT-LCD timings (should be extracted from Panel datasheet):
•

Horizontal and Vertical Synchronization width: 0x8 pixels and 0x4 lines

•

Horizontal and Vertical back porch: 0x7 pixels and 0x2 lines

•

Active Width and Active Height: 0x280 pixels, 0x1E0 lines (640x480)

•

Horizontal front porch: 0x6 pixels

•

Vertical front porch: 0x2 lines

The programmed values in the LTDC Timings registers will be:
•

LTDC_SSCR register: to be programmed to 0x00070003. (HSW[11:0] is 0x7 and
VSH[10:0] is 0x3)

•

LTDC_BPCR register: to be programmed to 0x000E0005. (AHBP[11:0] is 0xE(0x8 +
0x6) and AVBP[10:0] is 0x5(0x4 + 0x1))

•

LTDC_AWCR register: to be programmed to 0x028E01E5. (AAW[11:0] is 0x28E(0x8
+0x7 +0x27F) and AAH[10:0] is 0x1E5(0x4 +0x2 + 0x1DF)

•

LTDC_TWCR register: to be programmed to 0x00000294. (TOTALW[11:0] is
0x294(0x8 +0x7 +0x280 + 0x5)

•

LTDC_THCR register: to be programmed to 0x000001E7. (TOTALH[10:0] is
0x1E7(0x4 +0x2 + 0x1E0 + 1)

Programmable polarity
The Horizontal and Vertical Synchronization, Data Enable and Pixel Clock output signals
polarity can be programmed to active high or active low through the LTDC_GCR register.

Background Color
A constant background color (RGB888) can programmed through the LTDC_BCCR
register. It is used for blending with the bottom layer.

Dithering
The Dithering pseudo-random technique using an LFSR is used to add a small random
value (threshold) to each pixel color channel (R, G or B) value, thus rounding up the MSB in
some cases when displaying a 24-bit data on 18-bit display. Thus the Dithering technique is
used to round data which is different from one frame to the other.
The Dither pseudo-random technique is the same as comparing LSBs against a threshold
value and adding a 1 to the MSB part only, if the LSB part is >= the threshold. The LSBs are
typically dropped once dithering was applied.
The width of the added pseudo-random value is 2 bits for each color channel; 2 bits for Red,
2 bits for Green and 2 bits for Blue.
Once the LCD-TFT controller is enabled, the LFSR starts running with the first active pixel
and it is kept running even during blanking periods and when dithering is switched off. If the
LTDC is disabled, the LFSR is reset.
The Dithering can be switched On and Off on the fly through the LTDC_GCR register.

RM0090 Rev 17

485/1747
514

LCD-TFT controller (LTDC)

RM0090

Reload Shadow registers
Some configuration registers are shadowed. The shadow registers values can be reloaded
immediately to the active registers when writing to these registers or at the beginning of the
vertical blanking period following the configuration in the LTDC_SRCR register. If the
immediate reload configuration is selected, the reload should be only activated when all new
registers have been written.
The shadow registers should not be modified again before the reload has been done.
Reading from the shadow registers returns the actual active value. The new written value
can only be read after the reload has taken place.
A register reload interrupt can be generated if enabled in the LTDC_IER register.
The shadowed registers are all the Layer 1 and Layer 2 registers except the
LTDC_LxCLUTWR register.

Interrupt generation event
Refer to Section 16.5: LTDC interrupts for interrupt configuration.

16.4.2

Layer programmable parameters
Up to two layers can be enabled, disabled and configured separately. The layer display
order is fixed and it is bottom up. If two layers are enabled, the Layer2 is the top displayed
window.

Windowing
Every layer can be positioned and resized and it must be inside the Active Display area.
The window position and size are configured through the top-left and bottom-right X/Y
positions and the Internal timing generator which includes the synchronous, back porch size
and the active data area. Refer to LTDC_LxWHPCR and LTDC_WVPCR registers.
The programmable layer position and size defines the first/last visible pixel of a line and the
first/last visible line in the window. It allows to display either the full image frame or only a
part of the image frame. Refer to Figure 83
• The first and the last visible pixel in the layer are set by configuring the WHSTPOS[11:0]
and WHSPPOS[11:0] in the LTDC_LxWHPCR register.
• The first and the last visible lines in the layer are set by configuring the WVSTPOS[10:0]
and WVSPPOS[10:0] in the LTDC_LxWVPCR register.

486/1747

RM0090 Rev 17

RM0090

LCD-TFT controller (LTDC)
Figure 83. Layer window programmable parameters:
$FWLYHGDWDDUHD

:967326ELWVLQ
/7'&B/[:93&5

:+67326ELWVLQ
/7'&B/[:+3&5

:963326ELWVLQ
/7'&B/[:93&5

:LQGRZ

:+63326ELWVLQ
/7'&B/[:+3&5
06Y9

Pixel input Format
The programmable pixel format is used for the data stored in the frame buffer of a layer.
Up to 8 input pixel formats can be configured for every layer through the LTDC_LxPFCR
register
The pixel data is read from the frame buffer and then transformed to the internal 8888
(ARGB) format as follows:
•

Components which have a width of less than 8 bits get expanded to 8 bits by bit
replication. The selected bit range is concatenated multiple times until it is longer than
8 bits. Of the resulting vector, the 8 MSB bits are chosen. Example: 5 bits of an
RGB565 red channel become (bit positions): 43210432 (the 3 LSBs are filled with the 3
MSBs of the 5 bits)

The figure below describes the pixel data mapping depending on the selected format.
Table 90. Pixel Data mapping versus Color Format
ARGB8888
@+3
Ax[7:0]

@+2
Rx[7:0]

@+1
Gx[7:0]

@
Bx[7:0]

@+7
Ax+1[7:0]

@+6
Rx+1[7:0]

@+5
Gx+1[7:0]

@+4
Bx+1[7:0]

RGB888
@+3
Bx+1[7:0]

@+2
Rx[7:0]

@+1
Gx[7:0]

@
Bx[7:0]

@+7
Gx+2[7:0]

@+6
Bx+2[7:0]

@+5
Rx+1[7:0]

@+4
Gx+1[7:0]

@+1
Rx[4:0] Gx[5:3]

@
Gx[2:0] Bx[4:0]

RGB565
@+3
Rx+1[4:0] Gx+1[5:3]

@+2
Gx+1[2:0] Bx+1[4:0]

RM0090 Rev 17

487/1747
514

LCD-TFT controller (LTDC)

RM0090

Table 90. Pixel Data mapping versus Color Format (continued)
ARGB8888
@+7
Rx+3[4:0] Gx+3[5:3]

@+6
Gx+3[2:0] Bx+3[4:0]

@+5
Rx+2[4:0] Gx+2[5:3]

@+4
Gx+2[2:0] Bx+2[4:0]

ARGB1555
@+3
Ax+1[0]Rx+1[4:0]
Gx+1[4:3]

@+2
Gx+1[2:0] Bx+1[4:0]

@+1
Ax[0] Rx[4:0] Gx[4:3]

@
Gx[2:0] Bx[4:0]

@+7
Ax+3[0]Rx+3[4:0]
Gx+3[4:3]

@+6
Gx+3[2:0] Bx+3[4:0]

@+5
Ax+2[0]Rx+2[4:0]Gx+2[4:
3]

@+4
Gx+2[2:0] Bx+2[4:0]

ARGB4444
@+3
Ax+1[3:0]Rx+1[3:0]

@+2
Gx+1[3:0] Bx+1[3:0]

@+1
Ax[3:0] Rx[3:0]

@
Gx[3:0] Bx[3:0]

@+7
Ax+3[3:0]Rx+3[3:0]

@+6
Gx+3[3:0] Bx+3[3:0]

@+5
Ax+2[3:0]Rx+2[3:0]

@+4
Gx+2[3:0] Bx+2[3:0]

L8
@+3
Lx+3[7:0]

@+2
Lx+2[7:0]

@+1
Lx+1[7:0]

@
Lx[7:0]

@+7
Lx+7[7:0]

@+6
Lx+6[7:0]

@+5
Lx+5[7:0]

@+4
Lx+4[7:0]

AL44
@+3
Ax+3[3:0] Lx+3[3:0]

@+2
Ax+2[3:0] Lx+2[3:0]

@+1
Ax+1[3:0] Lx+1[3:0]

@
Ax[3:0] Lx[3:0]

@+7
Ax+7[3:0] Lx+7[3:0]

@+6
Ax+6[3:0] Lx+6[3:0]

@+5
Ax+5[3:0] Lx+5[3:0]

@+4
Ax+4[3:0] Lx+4[3:0]

AL88
@+3
Ax+1[7:0]

@+2
Lx+1[7:0]

@+1
Ax[7:0]

@
Lx[7:0]

@+7
Ax+3[7:0]

@+6
Lx+3[7:0]

@+5
Ax+2[7:0]

@+4
Lx+2[7:0]

Color Look-Up Table (CLUT)
The CLUT can be enabled at run-time for every layer through the LTDC_LxCR register and
it is only useful in case of indexed color when using the L8, AL44 and AL88 input pixel
format.
First, the CLUT has to be loaded with the R, G and B values that will replace the original R,
G, B values of that pixel (indexed color). Each color (RGB value) has its own address which
is the position within the CLUT.

488/1747

RM0090 Rev 17

RM0090

LCD-TFT controller (LTDC)
The R, G and B values and their own respective address are programmed through the
LTDC_LxCLUTWR register.
• In case of L8 and AL88 input pixel format, the CLUT has to be loaded by 256 colors. The
address of each color is configured in the CLUTADD bits in the LTDC_LxCLUTWR
register.
• In case of AL44 input pixel format, the CLUT has to be only loaded by 16 colors. The
address of each color must be filled by replicating the 4-bit L channel to 8-bit as follows:
– L0 (indexed color 0), at address 0x00
– L1, at address 0x11
– L2, at address 0x22
– .....
– L15, at address 0xFF

Color Frame Buffer Address
Every Layer has a start address for the color frame buffer configured through the
LTDC_LxCFBAR register.
When a layer is enabled, the data is fetched from the Color Frame Buffer.

Color Frame Buffer Length
Every layer has a total line length setting for the color frame buffer in bytes and a number of
lines in the frame buffer configurable in the LTDC_LxCFBLR and LTDC_LxCFBLNR
register respectively.
The line length and the number of lines settings are used to stop the prefetching of data to
the layer FIFO at the end of the frame buffer.
•

If it is set to less bytes than required, a FIFO underrun interrupt is generated if it has
been previously enabled.

•

If it is set to more bytes than actually required, the useless data read from the FIFO is
discarded. The useless data is not displayed.

Color Frame Buffer Pitch
Every layer has a configurable pitch for the color frame buffer, which is the distance between
the start of one line and the beginning of the next line in bytes. It is configured through the
LTDC_LxCFBLR register.

Layer Blending
The blending is always active and the two layers can be blended following the blending
factors configured through the LTDC_LxBFCR register.
The blending order is fixed and it is bottom up. If two layers are enabled, first the Layer1 is
blended with the Background color, then the Layer2 is blended with the result of blended
color of Layer1 and the background. Refer to Figure 84.

RM0090 Rev 17

489/1747
514

LCD-TFT controller (LTDC)

RM0090
Figure 84. Blending two layers with background

/D\HU
/D\HU

/D\HU

%*

/D\HU%*

/D\HU
/D\HU%*

069

Default color
Every layer can have a default color in the format ARGB which is used outside the defined
layer window or when a layer is disabled.
The default color is configured through the LTDC_LxDCCR register.
The blending is always performed between the two layers even when a layer is disabled. To
avoid displaying the default color when a layer is disabled, keep the blending factors of this
layer in the LTDC_LxBFCR register to their reset value.

Color Keying
A color key (RGB) can be configured to be representative for a transparent pixel.
If the Color Keying is enabled, the current pixels (after format conversion and before
blending) are compared to the color key. If they match for the programmed RGB value, all
channels (ARGB) of that pixel are set to 0.
The Color Key value can be configured and used at run-time to replace the pixel RGB value.
The Color Keying is enabled through the LTDC_LxCKCR register.

16.5

LTDC interrupts
The LTDC provides four maskable interrupts logically ORed to two interrupt vectors.
The interrupt sources can be enabled or disabled separately through the LTDC_IER
register. Setting the appropriate mask bit to 1 enables the corresponding interrupt.
The two interrupts are generated on the following events:
•

Line interrupt: generated when a programmed line is reached. The line interrupt
position is programmed in the LTDC_LIPCR register

•

Register Reload interrupt: generated when the shadow registers reload was performed
during the vertical blanking period

•

FIFO Underrun interrupt: generated when a pixel is requested from an empty layer
FIFO

•

Transfer Error interrupt: generated when an AHB bus error occurs during data transfer

Those interrupts events are connected to the NVIC controller as described in the figure
below.

490/1747

RM0090 Rev 17

RM0090

LCD-TFT controller (LTDC)
Figure 85. Interrupt events

/LQH

/7'&JOREDOLQWHUUXSW

5HJLVWHUUHORDG
),)2XQGHUUXQ

/7'&JOREDOHUURULQWHUUXSW

7UDQVIHUHUURU

069

Table 91. LTDC interrupt requests
Interrupt event

Event flag

Enable Control bit

LIF

LIE

Register Reload

RRIF

RRIEN

FIFO Underrun

FUDERRIF

FUDERRIE

Transfer Error

TERRIF

TERRIE

Line

RM0090 Rev 17

491/1747
514

LCD-TFT controller (LTDC)

16.6

Note:

492/1747

RM0090

LTDC programming procedure
•

Enable the LTDC clock in the RCC register

•

Configure the required Pixel clock following the panel datasheet

•

Configure the Synchronous timings: VSYNC, HSYNC, Vertical and Horizontal back
porch, active data area and the front porch timings following the panel datasheet as
described in the Section 16.4.1: LTDC Global configuration parameters

•

Configure the synchronous signals and clock polarity in the LTDC_GCR register

•

If needed, configure the background color in the LTDC_BCCR register

•

Configure the needed interrupts in the LTDC_IER and LTDC_LIPCR register

•

Configure the Layer1/2 parameters by programming:
–

The Layer window horizontal and vertical position in the LTDC_LxWHPCR and
LTDC_WVPCR registers. The layer window must be in the active data area.

–

The pixel input format in the LTDC_LxPFCR register

–

The color frame buffer start address in the LTDC_LxCFBAR register

–

The line length and pitch of the color frame buffer in the LTDC_LxCFBLR register

–

The number of lines of the color frame buffer in the LTDC_LxCFBLNR register

–

if needed, load the CLUT with the RGB values and its address in the
LTDC_LxCLUTWR register

–

If needed, configure the default color and the blending factors respectively in the
LTDC_LxDCCR and LTDC_LxBFCR registers

•

Enable Layer1/2 and if needed the CLUT in the LTDC_LxCR register

•

If needed, dithering and color keying can be enabled respectively in the LTDC_GCR
and LTDC_LxCKCR registers. It can be also enabled on the fly.

•

Reload the shadow registers to active register through the LTDC_SRCR register.

•

Enable the LCD-TFT controller in the LTDC_GCR register.

•

All layer parameters can be modified on the fly except the CLUT. The new configuration
has to be either reloaded immediately or during vertical blanking period by configuring
the LTDC_SRCR register.

All layer’s registers are shadowed. Once a register is written, it should not be modified again
before the reload has been done. Thus, a new write to the same register will override the
previous configuration if not yet reloaded.

RM0090 Rev 17

RM0090

LCD-TFT controller (LTDC)

16.7

LTDC registers

16.7.1

LTDC Synchronization Size Configuration Register (LTDC_SSCR)
This register defines the number of Horizontal Synchronization pixels minus 1 and the
number of Vertical Synchronization lines minus 1. Refer to Figure 82 and Section 16.4:
LTDC programmable parameters for an example of configuration.
Address offset: 0x08
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

14

13

20

19

18

17

16

HSW[11:0]

Reserved
15

21

12

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

VSH[10:0]

Reserved
rw

rw

rw

rw

rw

rw

Bits 31:28 Reserved, must be kept at reset value
Bits 27:16 HSW[11:0]: Horizontal Synchronization Width (in units of pixel clock period)
These bits define the number of Horizontal Synchronization pixel minus 1.
Bits 15:11 Reserved, must be kept at reset value
Bits 10:0 VSH[10:0]: Vertical Synchronization Height (in units of horizontal scan line)

These bits define the vertical Synchronization height minus 1. It represents the
number
of horizontal synchronization lines.

16.7.2

LTDC Back Porch Configuration Register (LTDC_BPCR)
This register defines the accumulated number of Horizontal Synchronization and back porch
pixels minus 1 (HSYNC Width + HBP- 1) and the accumulated number of Vertical
Synchronization and back porch lines minus 1 (VSYNC Height + VBP - 1). Refer to
Figure 82 and Section 16.4: LTDC programmable parameters for an example of
configuration.
Address offset: 0x0C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

14

13

20

19

18

17

16

AHBP[11:0]

Reserved
15

21

12

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

AVBP[10:0]

Reserved
rw

rw

rw

rw

rw

RM0090 Rev 17

rw

493/1747
514

LCD-TFT controller (LTDC)

RM0090

Bits 31:28 Reserved, must be kept at reset value
Bits 27:16 AHBP[11:0]: Accumulated Horizontal back porch (in units of pixel clock period)
These bits define the Accumulated Horizontal back porch width which includes the
Horizontal Synchronization and Horizontal back porch pixels minus 1.
The Horizontal back porch is the period between Horizontal Synchronization going
inactive and the start of the active display part of the next scan line.
Bits 15:11 Reserved, must be kept at reset value
Bits 10:0 AVBP[10:0]: Accumulated Vertical back porch (in units of horizontal scan line)
These bits define the accumulated Vertical back porch width which includes the Vertical
Synchronization and Vertical back porch lines minus 1.
The Vertical back porch is the number of horizontal scan lines at a start of frame to the
start of the first active scan line of the next frame.

16.7.3

LTDC Active Width Configuration Register (LTDC_AWCR)
This register defines the accumulated number of Horizontal Synchronization, back porch
and Active pixels minus 1 (HSYNC width + HBP + Active Width - 1) and the accumulated
number of Vertical Synchronization, back porch lines and Active lines minus 1 (VSYNC
Height+ BVBP + Active Height - 1). Refer to Figure 82 and Section 16.4: LTDC
programmable parameters for an example of configuration.
Address offset: 0x10
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

14

13

20

19

18

17

16

AAW[11:0]

Reserved
15

21

12

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

AAH[10:0]

Reserved
rw

rw

rw

rw

rw

rw

Bits 31:28 Reserved, must be kept at reset value
Bits 27:16 AAW[11:0]: Accumulated Active Width (in units of pixel clock period)
These bits define the Accumulated Active Width which includes the Horizontal
Synchronization, Horizontal back porch and Active pixels minus 1.
The Active Width is the number of pixels in active display area of the panel scan line. The
maximum Active Width supported is 0x400.
Bits 15:11 Reserved, must be kept at reset value
Bits 10:0 AAH[10:0]: Accumulated Active Height (in units of horizontal scan line)
These bits define the Accumulated Height which includes the Vertical Synchronization,
Vertical back porch and the Active Height lines minus 1. The Active Height is the number
of active lines in the panel. The maximum Active Height supported is 0x300.

494/1747

RM0090 Rev 17

RM0090

LCD-TFT controller (LTDC)

16.7.4

LTDC Total Width Configuration Register (LTDC_TWCR)
This register defines the accumulated number of Horizontal Synchronization, back porch,
Active and front porch pixels minus 1 (HSYNC Width + HBP + Active Width + HFP - 1) and
the accumulated number of Vertical Synchronization, back porch lines, Active and Front
lines minus 1 (VSYNC Height+ BVBP + Active Height + VFP - 1). Refer to Figure 82 and
Section 16.4: LTDC programmable parameters for an example of configuration.
Address offset: 0x14
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

14

21

20

19

18

17

16

TOTALW[11:0]

Reserved
16

22

13

12

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

TOTALH[10:0]

Reserved
rw

rw

rw

rw

rw

rw

Bits 31:28 Reserved, must be kept at reset value
Bits 27:16 TOTALW[11:0]: Total Width (in units of pixel clock period)
These bits defines the accumulated Total Width which includes the Horizontal
Synchronization, Horizontal back porch, Active Width and Horizontal front porch pixels
minus 1.
Bits 15:11 Reserved, must be kept at reset value
Bits 10:0 TOTALH[10:0]: Total Height (in units of horizontal scan line)
These bits defines the accumulated Height which includes the Vertical Synchronization,
Vertical back porch, the Active Height and Vertical front porch Height lines minus 1.

16.7.5

LTDC Global Control Register (LTDC_GCR)
This register defines the global configuration of the LCD-TFT controller.
Address offset: 0x18
Reset value: 0x0000 2220

31
HSPOL

30

29

28

26

25

24

23

VSPOL DEPOL PCPOL

rw

rw

rw

rw

15

14

13

12

Reserve
d

27

DRW[2:0]
r

r

r

22

21

20

19

18

17

16
DEN

Reserved

rw
11
Reser
ved

10

9

8

DGW[2:0]
r

r

r

7
Reser
ved

RM0090 Rev 17

6

5

4

DBW[2:0]
r

r

3

2
Reserved

r

1

0
LTDCEN
rw

495/1747
514

LCD-TFT controller (LTDC)

RM0090

Bit 31 HSPOL: Horizontal Synchronization Polarity
This bit is set and cleared by software.
0: Horizontal Synchronization polarity is active low
1: Horizontal Synchronization polarity is active high
Bit 30 VSPOL: Vertical Synchronization Polarity
This bit is set and cleared by software.
0: Vertical Synchronization is active low
1: Vertical Synchronization is active high
Bit 29 DEPOL: Data Enable Polarity
This bit is set and cleared by software.
0: Data Enable polarity is active low
1: Data Enable polarity is active high
Bit 28 PCPOL: Pixel Clock Polarity
This bit is set and cleared by software.
0: input pixel clock
1: inverted input pixel clock
Bits 27:17 Reserved, must be kept at reset value
Bit 16 DEN: Dither Enable
This bit is set and cleared by software.
0: Dither disable
1: Dither enable
Bit 15 Reserved, must be kept at reset value
Bits 14:12 DRW[2:0]: Dither Red Width
These bits return the Dither Red Bits
Bit 11 Reserved, must be kept at reset value
Bits 10:8 DGW[2:0]: Dither Green Width
These bits return the Dither Green Bits
Bit 7 Reserved, must be kept at reset value
Bits 6:4 DBW[2:0]: Dither Blue Width
These bits return the Dither Blue Bits
Bits 3:1 Reserved, must be kept at reset value
Bit 0 LTDCEN: LCD-TFT controller enable bit
This bit is set and cleared by software.
0: LTDC disable
1: LTDC enable

496/1747

RM0090 Rev 17

RM0090

LCD-TFT controller (LTDC)

16.7.6

LTDC Shadow Reload Configuration Register (LTDC_SRCR)
This register allows to reload either immediately or during the vertical blanking period, the
shadow registers values to the active registers. The shadow registers are all Layer1 and
Layer2 registers except the LTDC_L1CLUTWR and the LTDC_L2CLUTWR.
Address offset: 0x24
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

VBR

IMR

rw

rw

Reserved
15

14

13

12

11

10

9

8

7

Reserved

Bits 31:2 Reserved, must be kept at reset value
Bit 1 VBR: Vertical Blanking Reload
This bit is set by software and cleared only by hardware after reload. (it cannot be cleared
through register write once it is set)
0: No effect
1: The shadow registers are reloaded during the vertical blanking period (at the
beginning of the first line after the Active Display Area)
Bit 0 IMR: Immediate Reload
This bit is set by software and cleared only by hardware after reload.
0: No effect
1: The shadow registers are reloaded immediately

Note:

The shadow registers read back the active values. Until the reload has been done, the 'old'
value will be read.

16.7.7

LTDC Background Color Configuration Register (LTDC_BCCR)
This register defines the background color (RGB888).
Address offset: 0x2C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

14

13

12

11

18

17

16

10

9

8

rw

rw

rw

rw

rw

rw

rw

rw

7

6

5

4

3

2

1

0

rw

rw

rw

BCGREEN[7:0]
rw

19

BCRED[7:0]

Reserved
15

20

BCBLUE[7:0]
rw

rw

RM0090 Rev 17

rw

rw

rw

497/1747
514

LCD-TFT controller (LTDC)

RM0090

Bits 31:24 Reserved, must be kept at reset value
Bits 23:16 BCRED[7:0]: Background Color Red value
These bits configure the background red value
Bits 15:8 BCGREEN[7:0]: Background Color Green value
These bits configure the background green value
Bits 7:0 BCBLUE[7:0]: Background Color Blue value
These bits configure the background blue value

16.7.8

LTDC Interrupt Enable Register (LTDC_IER)
This register determines which status flags generate an interrupt request by setting the
corresponding bit to 1.
Address offset: 0x34
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

RRIE

TERRIE

FUIE

LIE

rw

rw

rw

rw

Reserved
15

14

13

12

11

10

9

8

7

Reserved

Bits 31:4 Reserved, must be kept at reset value
Bit 3 RRIE: Register Reload interrupt enable
This bit is set and cleared by software
0: Register Reload interrupt disable
1: Register Reload interrupt enable
Bit 2 TERRIE: Transfer Error Interrupt Enable
This bit is set and cleared by software
0: Transfer Error interrupt disable
1: Transfer Error interrupt enable
Bit 1 FUIE: FIFO Underrun Interrupt Enable
This bit is set and cleared by software
0: FIFO Underrun interrupt disable
1: FIFO Underrun Interrupt enable
Bit 0 LIE: Line Interrupt Enable
This bit is set and cleared by software
0: Line interrupt disable
1: Line Interrupt enable

498/1747

RM0090 Rev 17

RM0090

LCD-TFT controller (LTDC)

16.7.9

LTDC Interrupt Status Register (LTDC_ISR)
This register returns the interrupt status flag
Address offset: 0x38
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

RRIF

TERRIF

FUIF

LIF

r

r

r

r

Reserved
15

14

13

12

11

10

9

8

7

Reserved

Bits 31:24 Reserved, must be kept at reset value
Bit 3 RRIF: Register Reload Interrupt Flag
0: No Register Reload interrupt generated
1: Register Reload interrupt generated when a vertical blanking reload occurs (and the
first line after the active area is reached)
Bit 2 TERRIF: Transfer Error interrupt flag
0: No Transfer Error interrupt generated
1: Transfer Error interrupt generated when a Bus error occurs
Bit 1 FUIF: FIFO Underrun Interrupt flag
0: NO FIFO Underrun interrupt generated.
1: A FIFO underrun interrupt is generated, if one of the layer FIFOs is empty and pixel
data is read from the FIFO
Bit 0 LIF: Line Interrupt flag
0: No Line interrupt generated
1: A Line interrupt is generated, when a programmed line is reached

16.7.10

LTDC Interrupt Clear Register (LTDC_ICR)
Address offset: 0x3C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

CFUIF

CLIF

w

w

Reserved
15

14

13

12

11

10

9

8

7

CRRIF CTERRIF

Reserved

w

RM0090 Rev 17

w

499/1747
514

LCD-TFT controller (LTDC)

RM0090

Bits 31:24 Reserved, must be kept at reset value
Bit 3 CRRIF: Clears Register Reload Interrupt Flag
0: No effect
1: Clears the RRIF flag in the LTDC_ISR register
Bit 2 CTERRIF: Clears the Transfer Error Interrupt Flag
0: No effect
1: Clears the TERRIF flag in the LTDC_ISR register.
Bit 1 CFUIF: Clears the FIFO Underrun Interrupt flag
0: No effect
1: Clears the FUDERRIF flag in the LTDC_ISR register.
Bit 0 CLIF: Clears the Line Interrupt Flag
0: No effect
1: Clears the LIF flag in the LTDC_ISR register.

16.7.11

LTDC Line Interrupt Position Configuration Register (LTDC_LIPCR)
This register defines the position of the line interrupt. The line value to be programmed
depends on the timings parameters. Refer to Figure 82.
Address offset: 0x40
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

18

17

16

Reserved
15

14

13

12

11

10

9

8

7

LIPOS[10:0]

Reserved
rw

rw

rw

rw

rw

rw

Bits 31:11 Reserved, must be kept at reset value
Bits 10:0 LIPOS[10:0]: Line Interrupt Position
These bits configure the line interrupt position

16.7.12

LTDC Current Position Status Register (LTDC_CPSR)
Address offset: 0x44
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

CXPOS[15:0]
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

CYPOS{15;0]
r

500/1747

r

r

r

r

r

r

r

r

RM0090 Rev 17

RM0090

LCD-TFT controller (LTDC)

Bits 31:16: CXPOS[15:0]: Current X Position
These bits return the current X position
Bits 15:0 CYPOS[15:0]: Current Y Position
These bits return the current Y position

16.7.13

LTDC Current Display Status Register (LTDC_CDSR)
This register returns the status of the current display phase which is controlled by the
HSYNC, VSYNC, and Horizontal/Vertical DE signals.
Example: if the current display phase is the vertical synchronization, the VSYNCS bit is set
(active high). If the current display phase is the horizontal synchronization, the HSYNCS bit
is active high.
Address offset: 0x48
Reset value: 0x0000 000F

31

30

29

28

27

26

25

24

23

22

21

20

6

5

4

19

18

17

16

3

2

1

0

HDES

VDES

r

r

Reserved
15

14

13

12

11

10

9

8

7

HSYNC VSYNC
S
S

Reserved

r

r

Bits 31:24 Reserved, must be kept at reset value
Bit 3 HSYNCS: Horizontal Synchronization display Status
0: Active low
1: Active high
Bit 2 VSYNCS: Vertical Synchronization display Status
0: Active low
1: Active high
Bit 1 HDES: Horizontal Data Enable display Status
0: Active low
1: Active high
Bit 0 VDES: Vertical Data Enable display Status
0: Active low
1: Active high

Note:

The returned status does not depend on the configured polarity in the LTDC_GCR register,
instead it returns the current active display phase.

RM0090 Rev 17

501/1747
514

LCD-TFT controller (LTDC)

16.7.14

RM0090

LTDC Layerx Control Register (LTDC_LxCR) (where x=1..2)
Address offset: 0x84 + 0x80 x (Layerx -1), Layerx = 1 or 2
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

COLKEN

LEN

rw

rw

Reserved
15

14

13

12

11

10

9

8

7

CLUTEN

Reserved

rw

Reserved

Bits 31:5 Reserved, must be kept at reset value
Bit 4 CLUTEN: Color Look-Up Table Enable
This bit is set and cleared by software.
0: Color Look-Up Table disable
1: Color Look-Up Table enable
The CLUT is only meaningful for L8, AL44 and AL88 pixel format. Refer to Color Look-Up
Table (CLUT) on page 488
Bit 3 Reserved, must be kept at reset value
Bit 2 Reserved, must be kept at reset value
Bit 1 COLKEN: Color Keying Enable
This bit is set and cleared by software.
0: Color Keying disable
1: Color Keying enable
Bit 0 LEN: Layer Enable
This bit is set and cleared by software.
0: Layer disable
1: Layer enable

502/1747

RM0090 Rev 17

RM0090

LCD-TFT controller (LTDC)

16.7.15

LTDC Layerx Window Horizontal Position Configuration Register
(LTDC_LxWHPCR) (where x=1..2)
This register defines the Horizontal Position (first and last pixel) of the layer 1 or 2 window.
The first visible pixel of a line is the programmed value of AHBP[10:0] bits + 1 in the
LTDC_BPCR register.
The last visible pixel of a line is the programmed value of AAW[10:0] bits in the
LTDC_AWCR register.
Address offset: 0x88 + 0x80 x (Layerx -1), Layerx = 1 or 2
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

14

13

21

20

19

18

17

16

WHSPPOS[11:0]

Reserved
15

22

12

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

WHSTPOS[11:0]

Reserved
rw

rw

rw

rw

rw

rw

rw

Bits 31:28 Reserved, must be kept at reset value
Bits 27:16 WHSPPOS[11:0]: Window Horizontal Stop Position
These bits configure the last visible pixel of a line of the layer window.
The following condition must be respected:
WHSPPOS[11:0] ≥ AHBP[10:0] bits + 1 (programmed in LTDC_BPCR register)
Bits 15:12 Reserved, must be kept at reset value
Bits 11:0 WHSTPOS[11:0]: Window Horizontal Start Position
These bits configure the first visible pixel of a line of the layer window.
The following condition must be respected:
WHSTPOS[11:0] must be ≤ AAW[10:0] bits (programmed in the LTDC_AWCR
register).

Example:
The LTDC_BPCR register is configured to 0x000E0005(AHBP[11:0] is 0xE) and the
LTDC_AWCR register is configured to 0x028E01E5(AAW[11:0] is 0x28E). To configure the
horizontal position of a window size of 630x460, with horizontal start offset of 5 pixels in the
Active data area.
1.

Layer window first pixel: WHSTPOS[11:0] should be programmed to 0x14 (0xE+1+0x5)

2.

Layer window last pixel: WHSPPOS[11:0] should be programmed to 0x28A

RM0090 Rev 17

503/1747
514

LCD-TFT controller (LTDC)

16.7.16

RM0090

LTDC Layerx Window Vertical Position Configuration Register
(LTDC_LxWVPCR) (where x=1..2)
This register defines the vertical position (first and last line) of the layer1 or 2 window.
The first visible line of a frame is the programmed value of AVBP[10:0] bits + 1 in the
register LTDC_BPCR register.
The last visible line of a frame is the programmed value of AAH[10:0] bits in the
LTDC_AWCR register.
Address offset: 0x8C + 0x80 x (Layerx -1), Layerx = 1 or 2
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

14

13

20

19

18

17

16

WVSPPOS[10:0]

Reserved
15

21

12

11

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

WVSTPOS[10:0]

Reserved
rw

rw

rw

rw

rw

rw

rw

Bits 31:27 Reserved, must be kept at reset value
Bits 26:16 WVSPPOS[10:0]: Window Vertical Stop Position
These bits configures the last visible line of the layer window.
The following condition must be respected:
WHSPPOS[11:0] must be ≥ AVBP[10:0] bits + 1 (programmed in LTDC_BPCR
register)
Bits 15:11 Reserved, must be kept at reset value
Bits 10:0 WVSTPOS[10:0]: Window Vertical Start Position
These bits configure the first visible line of the layer window.
The following condition must be respected:
WHSTPOS[11:0] must be ≤ AAH[10:0] bits (programmed in the LTDC_AWCR register)

Example:
The LTDC_BPCR register is configured to 0x000E0005 (AVBP[10:0] is 0x5) and the
LTDC_AWCR register is configured to 0x028E01E5 (AAH[10:0] is 0x1E5). To configure the
vertical position of a window size of 630x460, with vertical start offset of 8 lines in the Active
data area:

504/1747

1.

Layer window first line: WVSTPOS[10:0] should be programmed to 0xE (0x5 + 1 + 0x8)

2.

Layer window last line: WVSPPOS[10:0] should be programmed to 0x1DA

RM0090 Rev 17

RM0090

LCD-TFT controller (LTDC)

16.7.17

LTDC Layerx Color Keying Configuration Register (LTDC_LxCKCR)
(where x=1..2)
This register defines the color key value (RGB), which is used by the Color Keying.
Address offset: 0x90 + 0x80 x (Layerx -1), Layerx = 1 or 2
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

14

13

12

11

19

18

17

16

CKRED[7:0]

Reserved
15

20

10

9

8

rw

rw

rw

rw

rw

rw

rw

rw

7

6

5

4

3

2

1

0

rw

rw

rw

CKGREEN[7:0]

CKBLUE[7:0]

rw

rw

rw

rw

rw

rw

Bits 31:24 Reserved, must be kept at reset value
Bits 23:16 CKRED[7:0]: Color Key Red value
Bits 15:8 CKGREEN[7:0]: Color Key Green value
Bits 7:0 CKBLUE[7:0]: Color Key Blue value

16.7.18

LTDC Layerx Pixel Format Configuration Register (LTDC_LxPFCR)
(where x=1..2)
This register defines the pixel format which is used for the stored data in the frame buffer of
a layer. The pixel data is read from the frame buffer and then transformed to the internal
format 8888 (ARGB).
Address offset: 0x94 + 0x80 x (Layerx -1), Layerx = 1 or 2
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

Reserved
15

14

13

12

11

10

9

8

7

PF[2:0]

Reserved
rw

RM0090 Rev 17

rw

rw

505/1747
514

LCD-TFT controller (LTDC)

RM0090

Bits 31:3 Reserved, must be kept at reset value
Bits 2:0 PF[2:0]: Pixel Format
These bits configures the Pixel format
000: ARGB8888
001: RGB888
010: RGB565
011: ARGB1555
100: ARGB4444
101: L8 (8-Bit Luminance)
110: AL44 (4-Bit Alpha, 4-Bit Luminance)
111: AL88 (8-Bit Alpha, 8-Bit Luminance)

16.7.19

LTDC Layerx Constant Alpha Configuration Register (LTDC_LxCACR)
(where x=1..2)
This register defines the constant alpha value (divided by 255 by Hardware), which is used
in the alpha blending. Refer to LTDC_LxBFCR register.
Address offset: 0x98 + 0x80 x (Layerx -1), Layerx = 1 or 2
Reset value: (Layerx -1) 0x0000 00FF

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

rw

rw

rw

Reserved
15

14

13

12

11

10

9

8

7

CONSTA[7:0]

Reserved
rw

rw

rw

rw

rw

Bits 31:8 Reserved, must be kept at reset value
Bits 7:0 CONSTA[7:0]: Constant Alpha
These bits configure the Constant Alpha used for blending. The Constant Alpha is divided
by 255 by hardware.
Example: if the programmed Constant Alpha is 0xFF, the Constant Alpha value is
255/255=1

16.7.20

LTDC Layerx Default Color Configuration Register (LTDC_LxDCCR)
(where x=1..2)
This register defines the default color of a layer in the format ARGB. The default color is
used outside the defined layer window or when a layer is disabled. The reset value of
0x00000000 defines a transparent black color.
Address offset: 0x9C + 0x80 x (Layerx -1), Layerx = 1 or 2
Reset value: 0x0000 0000

506/1747

RM0090 Rev 17

RM0090

31

LCD-TFT controller (LTDC)

30

29

28

27

26

25

24

23

22

21

DCALPHA[7:0]

14

13

12

11

10

9

8

18

17

16

rw

rw

rw

rw

rw

rw

rw

rw

7

6

5

4

3

2

1

0

rw

rw

rw

DCGREEN[7:0]
rw

19

DCRED[7:0]

rw
15

20

DCBLUE[7:0]
rw

rw

rw

rw

rw

Bits 31:24 DCALPHA[7:0]: Default Color Alpha
These bits configure the default alpha value
Bits 23:16 DCRED[7:0]: Default Color Red
These bits configure the default red value
Bits 15:8 DCGREEN[7:0]: Default Color Green
These bits configure the default green value
Bits 7:0 DCBLUE[7:0]: Default Color Blue
These bits configure the default blue value

RM0090 Rev 17

507/1747
514

LCD-TFT controller (LTDC)

16.7.21

RM0090

LTDC Layerx Blending Factors Configuration Register (LTDC_LxBFCR)
(where x=1..2)
This register defines the blending factors F1 and F2.
The general blending formula is: BC = BF1 x C + BF2 x Cs
•

BC = Blended color

•

BF1 = Blend Factor 1

•

C = Current layer color

•

BF2 = Blend Factor 2

•

Cs = subjacent layers blended color

Address offset: 0xA0 + 0x80 x (Layerx -1), Layerx = 1 or 2
Reset value: 0x0000 0607
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

Reserved
15

14

13

12

11

10

9

8

7

BF1[2:0]

Reserved
rw

rw

rw

Bits 31:11 Reserved, must be kept at reset value
Bits 10:8 BF1[2:0]: Blending Factor 1
These bits select the blending factor F1
000: Reserved
001: Reserved
010: Reserved
011: Reserved
100: Constant Alpha
101: Reserved
110: Pixel Alpha x Constant Alpha
111:Reserved
Bits 7:3 Reserved, must be kept at reset value
Bits 2:0 BF2[2:0]: Blending Factor 2
These bits select the blending factor F2
000: Reserved
001: Reserved
010: Reserved
011: Reserved
100: Reserved
101: 1 - Constant Alpha
110: Reserved
111: 1 - (Pixel Alpha x Constant Alpha)

508/1747

BF2[2:0]

Reserved

RM0090 Rev 17

rw

rw

rw

RM0090

LCD-TFT controller (LTDC)

Note:

The Constant Alpha value, is the programmed value in the LxCACR register divided by 255
by hardware.
Example: Only layer1 is enabled, BF1 configured to Constant Alpha
BF2 configured to 1 - Constant Alpha
Constant Alpha: The Constant Alpha programmed in the LxCACR register is 240 (0xF0).
Thus, the Constant Alpha value is 240/255 = 0.94
C: Current Layer Color is 128
Cs: Background color is 48
Layer1 is blended with the background color.
BC = Constant Alpha x C + (1 - Constant Alpha) x Cs = 0.94 x 128 + (1- 0.94) x 48 = 123.

16.7.22

LTDC Layerx Color Frame Buffer Address Register (LTDC_LxCFBAR)
(where x=1..2)
This register defines the color frame buffer start address which has to point to the address
where the pixel data of the top left pixel of a layer is stored in the frame buffer.
Address offset: 0xAC + 0x80 x (Layerx -1), Layerx = 1 or 2
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

CFBADD[31:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

CFBADD[31:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 CFBADD[31:0]: Color Frame Buffer Start Address
These bits defines the color frame buffer start address.

16.7.23

LTDC Layerx Color Frame Buffer Length Register (LTDC_LxCFBLR)
(where x=1..2)
This register defines the color frame buffer line length and pitch.
Address offset: 0xB0 + 0x80 x (Layerx -1), Layerx = 1 or 2
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

14

21

20

19

18

17

16

CFBP[12:0]

Reserved
15

22

13

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

CFBLL[12:0]

Reserved
rw

rw

rw

rw

rw

rw

rw

RM0090 Rev 17

509/1747
514

LCD-TFT controller (LTDC)

RM0090

Bits 31:29 Reserved, must be kept at reset valuer
Bits 28:16 CFBP[12:0]: Color Frame Buffer Pitch in bytes
These bits define the pitch which is the increment from the start of one line of pixels to the
start of the next line in bytes.
Bits 15:13 Reserved, must be kept at reset value
Bits 12:0 CFBLL[12:0]: Color Frame Buffer Line Length
These bits define the length of one line of pixels in bytes + 3.
The line length is computed as follows: Active high width x number of bytes per pixel + 3.

Example:

16.7.24

•

A frame buffer having the format RGB565 (2 bytes per pixel) and a width of 256 pixels
(total number of bytes per line is 256x2=512 bytes), where pitch = line length requires a
value of 0x02000203 to be written into this register.

•

A frame buffer having the format RGB888 (3 bytes per pixel) and a width of 320 pixels
(total number of bytes per line is 320x3=960), where pitch = line length requires a value
of 0x03C003C3 to be written into this register.

LTDC Layerx ColorFrame Buffer Line Number Register
(LTDC_LxCFBLNR) (where x=1..2)
This register defines the number of lines in the color frame buffer.
Address offset: 0xB4 + 0x80 x (Layerx -1), Layerx = 1 or 2
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

Reserved
15

14

13

12

11

10

9

8

7

CFBLNBR[10:0]

Reserved
rw

rw

rw

rw

rw

rw

Bits 31:11 Reserved, must be kept at reset value
Bits 10:0 CFBLNBR[10:0]: Frame Buffer Line Number
These bits define the number of lines in the frame buffer which corresponds to the Active
high width.

Note:

The number of lines and line length settings define how much data is fetched per frame for
every layer. If it is configured to less bytes than required, a FIFO underrun interrupt will be
generated if enabled.
The start address and pitch settings on the other hand define the correct start of every line in
memory.

510/1747

RM0090 Rev 17

RM0090

LCD-TFT controller (LTDC)

16.7.25

LTDC Layerx CLUT Write Register (LTDC_LxCLUTWR)
(where x=1..2)
This register defines the CLUT address and the RGB value.
Address offset: 0xC4 + 0x80 x (Layerx -1), Layerx = 1 or 2
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

CLUTADD[7:0]

19

18

17

16

RED[7:0]

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

w

w

w

GREEN[7:0]
w

w

w

w

w

BLUE[7:0]
w

w

w

w

w

w

w

w

Bits 31:24 CLUTADD[7:0]: CLUT Address
These bits configure the CLUT address (color position within the CLUT) of each RGB
value
Bits 23:16 RED[7:0]: Red value
These bits configure the red value
Bits 15:8 GREEN[7:0]: Green value
These bits configure the green value
Bits 7:0 BLUE[7:0]: Blue value
These bits configure the blue value

Note:

The CLUT write register should only be configured during blanking period or if the layer is
disabled. The CLUT can be enabled or disabled in the LTDC_LxCR register.
The CLUT is only meaningful for L8, AL44 and AL88 pixel format.

RM0090 Rev 17

511/1747
514

LCD-TFT controller (LTDC)

16.7.26

RM0090

LTDC register map
The following table summarizes the LTDC registers. Refer to the register boundary
addresses table for the LTDC register base address.

0

0

0

0

0

9

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DEPOL

PCPOL

0

0

0

0

0

0

0

0

0

0

0

Reserve

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

RRIF

FUIF

LIF

0

0

0

0

CRRIF

CFUIF

CLIF

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1
CLUTEN

LTDC_L1CR

Reserved

Reset value

0
WHSPPOS[11:0]
0

0

0

0

0

0

0

Reserved

0

0

CYPOS[15:0]

LTDC_CDSR

Reserved

0

0

0

0

0

0

RM0090 Rev 17

0

Reserved

VDES

0

0

HDES

0

0

HSYNCS

0

CXPOS[15:0]
0

0

LIPOS[10:0]

Reserved

0

0

CTERRIF

Reserved

LTDC_CPSR

512/1747

0

TERRIF

Reserved

Reset value

Reset value

0

0

Reset value

0x0088

0

LIE

0

Reset value

LTDC_L1WHPCR

0

FUIE

0

Reserved

LTDC_LIPCR

0x0084

0

RRIE

0

LTDC_ICR

0x0048

0

0

Reset value

Reset value

0

TERRIE

0

LTDC_ISR

0x0044

0

0

Reset value

0x0040

0

VSYNCS

0

LTDC_IER

0x003C

0

BC[23:0]

Reserved

Reset value

0x0038

0

1

1

1

Reserved

LTDC_BCCR

0x0034

0

0

Reserved

Reset value

0x002C

0

0

0

TOTALH[10:0]

Reserved

1

0

AAH[10:0]

Reserved

0

0

AVBP[10:0]

Reserved

0

0

LTDC_SRCR

0x0024

0

0

LEN

VSPOL

Reset value

0

0

COLKEN

LTDC_GCR

0

VSH[10:0]

Reserved

LTDCEN

0

IMR

0

Reserved

0

VBR

0

DBW[2:0]

0

0

HSPOL

0x0018

0

TOTALW[11:0]

Reserved

Reset value

0

Reserved

LTDC_TWCR

0x0014

0

AAV[11:0]

Reserved

Reset value

0

DGW[2:0]

LTDC_AWCR

0x0010

0

AHBP[11:0]

Reserved

Reset value

0

Reserved

LTDC_BPCR

0x000C

HSW[9:0]

Reserved

Reset value

DRW[2:0]

LTDC_SSCR

0x0008

DEN

Register

Reserved

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 92. LTDC register map and reset values

0

0

0

0

WHSTPOS[11:0]
0

0

0

0

0

0

0

0

0

0

RM0090

LCD-TFT controller (LTDC)

0x0098

0x009C

0x00A0

0x00AC

0x00B0

0x00B4

0x00C4

0x0104

LTDC_L1CKCR

0

0

0

0

0

0

0

0

0

0

0

0

LTDC_L1PFCR

LTDC_L1CACR

LTDC_L1DCCR

DCALPHA[7:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CKGREEN[7:0]

0

0

0

0

0

0

0

0

0

0x010C

0x0110

0x0114

0x0118

0x011C

0

0

0

0

0x0120

0x012C

0

0

0

0

Reserved

0

0

0

0

0

0

0x0130

0

0

0

1

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

1

1

1

1

0

0

DCBLUE[7:0]
0

0

BF1[2:0]

0

0

0

0

0

0

0

0

0

0

0

CLUTADD[7:0]

0

0

0

0

0

0

0

0

0

0

Reserved

1

1

0

0

0

0

0

0

0

0

BF2[2:0]

Reserved

1

1

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CFBLL[12:0]
0

0

0

0

0

0

0

0

CFBLNBR[10:0]
0

RED[7:0]
0

LTDC_L2CR

0

0

0

0

0

GREEN[7:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0
WHSPPOS[11:0]

Reserved

Reset value
LTDC_L2WVPCR

0

LTDC_L2CKCR

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

LTDC_L2PFCR

0

0

0

0

0

0

0

0

0

0

0

0

CKGREEN[7:0]
0

0

0

0

0

0

0

0

0

LTDC_L2CACR

0

0

0

0

LTDC_L2DCCR

DCALPHA[7:0]
0

0

0

0

0

0

0

0

LTDC_L2BFCR

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

0

0

CONSTA[7:0]

0

1

1

0

0

DCGREEN[7:0]
0

0

0

0

0

0

0

0

LTDC_L2CFBAR

1

1

1

1

DCBLUE[7:0]
0

0

BF1[2:0]

Reserved

Reset value

0

0

DCRED[7:0]
0

0

PF[2:0]

Reserved

Reset value

0

CKBLUE[7:0]

Reserved

Reset value

0

WVSTPOS[10:0]

Reserved

0

0

WHSTPOS[11:0]

Reserved

0

CKRED[7:0]

Reserved

Reset value

0

WVSPPOS[10:0]

Reserved

Reset value

0

0

BLUE[7:0]

Reserved

LTDC_L2WHPCR

Reset value

0

PF[2:0]

Reserved

LTDC_L1CLUTWR

LTDC_L2CFBLR

0

0

DCGREEN[7:0]

0

CFBP[12:0]

Reset value

Reset value

0

CFBADD[31:0]

LTDC_L1CFBLNR

Reset value

0

CKBLUE[7:0]

Reset value
0x0108

0

CONSTA[7:0]

LTDC_L1CFBAR

Reset value

0

0

Reserved

Reset value

Reset value

0

WVSTPOS[10:0]

Reserved

DCRED[7:0]
0

LTDC_L1BFCR

Reset value

0

Reserved

Reset value

LTDC_L1CFBLR

0

Reserved

Reset value

Reset value

0

CKRED[7:0]

Reserved

Reset value

0

LEN

0x0094

WVSPPOS[10:0]

Reserved

Reset value

COLKEN

0x0090

LTDC_L1WVPCR

Reserved

0x008C

Register

CLUTEN

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 92. LTDC register map and reset values (continued)

1

1

0

0

0

0

0

BF2[2:0]

Reserved

1

1

1

CFBADD[31:0]
0

0

0

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CFBP[12:0]
0

0

0

RM0090 Rev 17

0

0

0

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CFBLL[12:0]
0

0

0

513/1747
514

LCD-TFT controller (LTDC)

RM0090

Offset
0x0134

0x0144

Register

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 92. LTDC register map and reset values (continued)

LTDC_L2CFBLNR

LTDC_L2CLUTWR

514/1747

Reset value

CFBLNBR[10:0]

Reserved

Reset value
CLUTADD[7:0]
0

0

0

0

0

0

0

RED[7:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

GREEN[7:0]
0

0

RM0090 Rev 17

0

0

0

0

0

0

0

0

0

0

0

0

0

0

BLUE[7:0]
0

0

0

RM0090

17

Advanced-control timers (TIM1 and TIM8)

Advanced-control timers (TIM1 and TIM8)
This section applies to the whole STM32F4xx family, unless otherwise specified.

17.1

TIM1 and TIM8 introduction
The advanced-control timers (TIM1 and TIM8) consist of a 16-bit auto-reload counter driven
by a programmable prescaler.
It may be used for a variety of purposes, including measuring the pulse lengths of input
signals (input capture) or generating output waveforms (output compare, PWM,
complementary PWM with dead-time insertion).
Pulse lengths and waveform periods can be modulated from a few microseconds to several
milliseconds using the timer prescaler and the RCC clock controller prescalers.
The advanced-control (TIM1 and TIM8) and general-purpose (TIMx) timers are completely
independent, and do not share any resources. They can be synchronized together as
described in Section 17.3.20.

RM0090 Rev 17

515/1747
588

Advanced-control timers (TIM1 and TIM8)

17.2

RM0090

TIM1 and TIM8 main features
TIM1 and TIM8 timer features include:

516/1747

•

16-bit up, down, up/down auto-reload counter.

•

16-bit programmable prescaler allowing dividing (also “on the fly”) the counter clock
frequency either by any factor between 1 and 65536.

•

Up to 4 independent channels for:
–

Input Capture

–

Output Compare

–

PWM generation (Edge and Center-aligned Mode)

–

One-pulse mode output

•

Complementary outputs with programmable dead-time

•

Synchronization circuit to control the timer with external signals and to interconnect
several timers together.

•

Repetition counter to update the timer registers only after a given number of cycles of
the counter.

•

Break input to put the timer’s output signals in reset state or in a known state.

•

Interrupt/DMA generation on the following events:
–

Update: counter overflow/underflow, counter initialization (by software or
internal/external trigger)

–

Trigger event (counter start, stop, initialization or count by internal/external trigger)

–

Input capture

–

Output compare

–

Break input

•

Supports incremental (quadrature) encoder and hall-sensor circuitry for positioning
purposes

•

Trigger input for external clock or cycle-by-cycle current management

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)

7,0[B(75

Figure 86. Advanced-control timer block diagram
&.B7,0IURP5&&
(75

,QWHUQDOFORFN &.B,17
7ULJJHU
FRQWUROOHU

(75)

3RODULW\VHOHFWLRQ
(GJHGHWHFWRUDQG3UHVFDOHU

(753
75*2

7RRWKHUWLPHUV
7R'$&DQG$'&

,QSXWILOWHU

,75

7*,

,75
75*,

,75
,75

6ODYHPRGH
FRQWUROOHU

7,)B('

5(35HJLVWHU

7,0[B&+
7,0[B&+

'7*>@UHJLVWHUV

,&

3UHVFDOHU

7,)3
75&

7,

,QSXWILOWHU 
(GJHGHWHFWRU

7,)3

&&,
,&

3UHVFDOHU

7,

,QSXWILOWHU 
(GJHGHWHFWRU

7,)3
7,

,QSXWILOWHU 
(GJHGHWHFWRU

&&,
,&

3UHVFDOHU

75&

7,)3

3UHVFDOHU

7,)3

3RODULW\VHOHFWLRQ

&&,
&DSWXUH&RPSDUH 2&5()
5HJLVWHU
2&1

8
&&,

2&

&DSWXUH&RPSDUH 2&5()
5HJLVWHU
8
&&,

,&36

&DSWXUH&RPSDUH 2&5()
5HJLVWHU
8

75&
%5.

8

,&36

&&,
,&

&DSWXUH&RPSDUH 2&5()
5HJLVWHU

,&36

7,)3
7,)3

&&,

,&36

2&

7,)3
,QSXWILOWHU 
(GJHGHWHFWRU

2&
%,

&ORFNIDLOXUHHYHQWIURPFORFNFRQWUROOHU
&66 &ORFN6HFXULW\6\VWHP

,QWHUUXSW '0$RXWSXW

/HJHQG
5HJ

(YHQW

'7*

7,0[B&+

&17
FRXQWHU

&&,
7,

75&

7,0[B%.,1

&.B&17

2&1

7,0[B&+
7,0[B&+

36&
SUHVFDOHU

7,0[B&+1

5HSHWLWLRQFRXQWHU
&.B36&

7,0[B&+1

2&

$XWR5HORDG
5HJLVWHU

7,0[B&+1 7,0[B&+

8

7,0[B&+

8

7,0[B&+

(QFRGHU
LQWHUIDFH

7,)3

8,

2&1

7,)3

5HVHW
(QDEOH
8S'RZQ
&RXQW

2XWSXWFRQWURO

3UHORDGUHJLVWHUVWUDQVIHUUHGWRDFWLYHUHJLVWHUVRQ8HYHQWDFFRUGLQJWRFRQWUROELW
069

RM0090 Rev 17

517/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

17.3

TIM1 and TIM8 functional description

17.3.1

Time-base unit
The main block of the programmable advanced-control timer is a 16-bit counter with its
related auto-reload register. The counter can count up, down or both up and down. The
counter clock can be divided by a prescaler.
The counter, the auto-reload register and the prescaler register can be written or read by
software. This is true even when the counter is running.
The time-base unit includes:
•

Counter register (TIMx_CNT)

•

Prescaler register (TIMx_PSC)

•

Auto-reload register (TIMx_ARR)

•

Repetition counter register (TIMx_RCR)

The auto-reload register is preloaded. Writing to or reading from the auto-reload register
accesses the preload register. The content of the preload register are transferred into the
shadow register permanently or at each update event (UEV), depending on the auto-reload
preload enable bit (ARPE) in TIMx_CR1 register. The update event is sent when the counter
reaches the overflow (or underflow when downcounting) and if the UDIS bit equals 0 in the
TIMx_CR1 register. It can also be generated by software. The generation of the update
event is described in detailed for each configuration.
The counter is clocked by the prescaler output CK_CNT, which is enabled only when the
counter enable bit (CEN) in TIMx_CR1 register is set (refer also to the slave mode controller
description to get more details on counter enabling).
Note that the counter starts counting 1 clock cycle after setting the CEN bit in the TIMx_CR1
register.

Prescaler description
The prescaler can divide the counter clock frequency by any factor between 1 and 65536. It
is based on a 16-bit counter controlled through a 16-bit register (in the TIMx_PSC register).
It can be changed on the fly as this control register is buffered. The new prescaler ratio is
taken into account at the next update event.
Figure 87 and Figure 88 give some examples of the counter behavior when the prescaler
ratio is changed on the fly:

518/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)
Figure 87. Counter timing diagram with prescaler division change from 1 to 2

&.B36&

&(1
7LPHUFORFN

&.B&17


&RXQWHUUHJLVWHU

)

)

)

)$ )% )&









8SGDWHHYHQW 8(9

3UHVFDOHUFRQWUROUHJLVWHU





:ULWHDQHZYDOXHLQ7,0[B36&
3UHVFDOHUEXIIHU



3UHVFDOHUFRXQWHU





















069

Figure 88. Counter timing diagram with prescaler division change from 1 to 4

&.B36&
&(1

7LPHUFORFN

&.B&17


&RXQWHUUHJLVWHU

)

)

)

)$ )% )&





8SGDWHHYHQW 8(9

3UHVFDOHUFRQWUROUHJLVWHU





:ULWHDQHZYDOXHLQ7,0[B36&
3UHVFDOHUEXIIHU

3UHVFDOHUFRXQWHU






















069

RM0090 Rev 17

519/1747
588

Advanced-control timers (TIM1 and TIM8)

17.3.2

RM0090

Counter modes
Upcounting mode
In upcounting mode, the counter counts from 0 to the auto-reload value (content of the
TIMx_ARR register), then restarts from 0 and generates a counter overflow event.
If the repetition counter is used, the update event (UEV) is generated after upcounting is
repeated for the number of times programmed in the repetition counter register plus one
(TIMx_RCR+1). Else the update event is generated at each counter overflow.
Setting the UG bit in the TIMx_EGR register (by software or by using the slave mode
controller) also generates an update event.
The UEV event can be disabled by software by setting the UDIS bit in the TIMx_CR1
register. This is to avoid updating the shadow registers while writing new values in the
preload registers. Then no update event occurs until the UDIS bit has been written to 0.
However, the counter restarts from 0, as well as the counter of the prescaler (but the
prescale rate does not change). In addition, if the URS bit (update request selection) in
TIMx_CR1 register is set, setting the UG bit generates an update event UEV but without
setting the UIF flag (thus no interrupt or DMA request is sent). This is to avoid generating
both update and capture interrupts when clearing the counter on the capture event.
When an update event occurs, all the registers are updated and the update flag (UIF bit in
TIMx_SR register) is set (depending on the URS bit):
•

The repetition counter is reloaded with the content of TIMx_RCR register,

•

The auto-reload shadow register is updated with the preload value (TIMx_ARR),

•

The buffer of the prescaler is reloaded with the preload value (content of the TIMx_PSC
register).

The following figures show some examples of the counter behavior for different clock
frequencies when TIMx_ARR=0x36.
Figure 89. Counter timing diagram, internal clock divided by 1
<ͺW^

EdͺE

dŝŵĞƌĐůŽĐŬс<ͺEd
ŽƵŶƚĞƌƌĞŐŝƐƚĞƌ

ϯϭ

ϯϮ

ϯϯ

ϯϰ

ϯϱ

ϯϲ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

ŽƵŶƚĞƌŽǀĞƌĨůŽǁ

hƉĚĂƚĞĞǀĞŶƚ;hsͿ

hƉĚĂƚĞŝŶƚĞƌƌƵƉƚĨůĂŐ;h/&Ϳ
069

520/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)
Figure 90. Counter timing diagram, internal clock divided by 2
&.B36&
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU















&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

Figure 91. Counter timing diagram, internal clock divided by 4
&.B36&
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU









&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

Figure 92. Counter timing diagram, internal clock divided by N
&.B36&

7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU

)





&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

RM0090 Rev 17

521/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Figure 93. Counter timing diagram, update event when ARPE=0 (TIMx_ARR not
preloaded)
<ͺW^

E

dŝŵĞƌĐůŽĐŬс<ͺEd
ŽƵŶƚĞƌƌĞŐŝƐƚĞƌ

ϯϭ

ϯϯ

ϯϮ

ϯϰ

ϯϱ

ϯϲ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

ŽƵŶƚĞƌŽǀĞƌĨůŽǁ

hƉĚĂƚĞĞǀĞŶƚ;hsͿ

hƉĚĂƚĞŝŶƚĞƌƌƵƉƚĨůĂŐ;h/&Ϳ

ƵƚŽͲƌĞůŽĂĚƉƌĞůŽĂĚƌĞŐŝƐƚĞƌ

&&

ϯϲ

tƌŝƚĞĂŶĞǁǀĂůƵĞŝŶd/DdžͺZZ

069

Figure 94. Counter timing diagram, update event when ARPE=1 (TIMx_ARR
preloaded)
&.B36&

&(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU

)

)

)

)

) )

















&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ
 8,)
$XWRUHORDGSUHORDG
UHJLVWHU

)



$XWRUHORDGVKDGRZ
UHJLVWHU

)

:ULWHDQHZYDOXHLQ7,0[B$55

522/1747

RM0090 Rev 17



069

RM0090

Advanced-control timers (TIM1 and TIM8)

Downcounting mode
In downcounting mode, the counter counts from the auto-reload value (content of the
TIMx_ARR register) down to 0, then restarts from the auto-reload value and generates a
counter underflow event.
If the repetition counter is used, the update event (UEV) is generated after downcounting is
repeated for the number of times programmed in the repetition counter register plus one
(TIMx_RCR+1). Else the update event is generated at each counter underflow.
Setting the UG bit in the TIMx_EGR register (by software or by using the slave mode
controller) also generates an update event.
The UEV update event can be disabled by software by setting the UDIS bit in TIMx_CR1
register. This is to avoid updating the shadow registers while writing new values in the
preload registers. Then no update event occurs until UDIS bit has been written to 0.
However, the counter restarts from the current auto-reload value, whereas the counter of the
prescaler restarts from 0 (but the prescale rate doesn’t change).
In addition, if the URS bit (update request selection) in TIMx_CR1 register is set, setting the
UG bit generates an update event UEV but without setting the UIF flag (thus no interrupt or
DMA request is sent). This is to avoid generating both update and capture interrupts when
clearing the counter on the capture event.
When an update event occurs, all the registers are updated and the update flag (UIF bit in
TIMx_SR register) is set (depending on the URS bit):
•

The repetition counter is reloaded with the content of TIMx_RCR register

•

The buffer of the prescaler is reloaded with the preload value (content of the TIMx_PSC
register)

•

The auto-reload active register is updated with the preload value (content of the
TIMx_ARR register). Note that the auto-reload is updated before the counter is
reloaded, so that the next period is the expected one

The following figures show some examples of the counter behavior for different clock
frequencies when TIMx_ARR=0x36.

RM0090 Rev 17

523/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Figure 95. Counter timing diagram, internal clock divided by 1

&.B36&

&17B(1

7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU



























)

&RXQWHUXQGHUIORZ
FQWBXGI

8SGDWHHYHQW 8(9

8SGDWHLQWHUUXSWIODJ
 8,)
069

Figure 96. Counter timing diagram, internal clock divided by 2

&.B36&

&17B(1

7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU















&RXQWHUXQGHUIORZ

8SGDWHHYHQW 8(9

8SGDWHLQWHUUXSWIODJ
 8,)

069

524/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)
Figure 97. Counter timing diagram, internal clock divided by 4
&.B36&
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU









&RXQWHUXQGHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)

069

Figure 98. Counter timing diagram, internal clock divided by N

&.B36&

7LPHUFORFN &.B&17

&RXQWHUUHJLVWHU



)





&RXQWHUXQGHUIORZ

8SGDWHHYHQW 8(9

8SGDWHLQWHUUXSWIODJ
 8,)
069

RM0090 Rev 17

525/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Figure 99. Counter timing diagram, update event when repetition counter is not used
&.B36&

&(1

7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU









 

  





  )

&RXQWHUXQGHUIORZ

8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ
 8,)
$XWRUHORDGSUHORDG
UHJLVWHU

))



:ULWHDQHZYDOXHLQ7,0[B$55
069

Center-aligned mode (up/down counting)
In center-aligned mode, the counter counts from 0 to the auto-reload value (content of the
TIMx_ARR register) – 1, generates a counter overflow event, then counts from the autoreload value down to 1 and generates a counter underflow event. Then it restarts counting
from 0.
Center-aligned mode is active when the CMS bits in TIMx_CR1 register are not equal to
'00'. The Output compare interrupt flag of channels configured in output is set when: the
counter counts down (Center aligned mode 1, CMS = "01"), the counter counts up (Center
aligned mode 2, CMS = "10") the counter counts up and down (Center aligned mode 3,
CMS = "11").
In this mode, the DIR direction bit in the TIMx_CR1 register cannot be written. It is updated
by hardware and gives the current direction of the counter.
The update event can be generated at each counter overflow and at each counter underflow
or by setting the UG bit in the TIMx_EGR register (by software or by using the slave mode
controller) also generates an update event. In this case, the counter restarts counting from
0, as well as the counter of the prescaler.
The UEV update event can be disabled by software by setting the UDIS bit in the TIMx_CR1
register. This is to avoid updating the shadow registers while writing new values in the
preload registers. Then no update event occurs until UDIS bit has been written to 0.
However, the counter continues counting up and down, based on the current auto-reload
value.
In addition, if the URS bit (update request selection) in TIMx_CR1 register is set, setting the
UG bit generates an UEV update event but without setting the UIF flag (thus no interrupt or
DMA request is sent). This is to avoid generating both update and capture interrupts when
clearing the counter on the capture event.
526/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)
When an update event occurs, all the registers are updated and the update flag (UIF bit in
TIMx_SR register) is set (depending on the URS bit):
•

The repetition counter is reloaded with the content of TIMx_RCR register

•

The buffer of the prescaler is reloaded with the preload value (content of the TIMx_PSC
register)

•

The auto-reload active register is updated with the preload value (content of the
TIMx_ARR register). Note that if the update source is a counter overflow, the autoreload is updated before the counter is reloaded, so that the next period is the expected
one (the counter is loaded with the new value).

The following figures show some examples of the counter behavior for different clock
frequencies.
Figure 100. Counter timing diagram, internal clock divided by 1, TIMx_ARR = 0x6
&.B36&
&(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU







 





 





 



&RXQWHUXQGHUIORZ
&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

1. Here, center-aligned mode 1 is used (for more details refer to Section 17.4: TIM1 and TIM8 registers).

Figure 101. Counter timing diagram, internal clock divided by 2
&.B36&

&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU















&RXQWHUXQGHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

RM0090 Rev 17

527/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Figure 102. Counter timing diagram, internal clock divided by 4, TIMx_ARR=0x36

&.B36&

&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU









&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

1.

Center-aligned mode 2 or 3 is used with an UIF on overflow.

Figure 103. Counter timing diagram, internal clock divided by N
&.B36&

7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU



)





&RXQWHUXQGHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

528/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)
Figure 104. Counter timing diagram, update event with ARPE=1 (counter underflow)
&.B36&
&(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU





 

 











  

&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
$XWRUHORDGSUHORDGUHJLVWHU

)'



:ULWHDQHZYDOXHLQ7,0[B$55

$XWRUHORDGDFWLYHUHJLVWHU

)'


069

Figure 105. Counter timing diagram, Update event with ARPE=1 (counter overflow)
&.B36&
&(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU

)

)

)

)$

)% )& 

 







 )

&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
$XWRUHORDGSUHORDGUHJLVWHU

)'



:ULWHDQHZYDOXHLQ7,0[B$55

$XWRUHORDGDFWLYHUHJLVWHU

)'


069

17.3.3

Repetition counter
Section 17.3.1: Time-base unit describes how the update event (UEV) is generated with
respect to the counter overflows/underflows. It is actually generated only when the repetition
counter has reached zero. This can be useful when generating PWM signals.
This means that data are transferred from the preload registers to the shadow registers
(TIMx_ARR auto-reload register, TIMx_PSC prescaler register, but also TIMx_CCRx
capture/compare registers in compare mode) every N+1 counter overflows or underflows,
where N is the value in the TIMx_RCR repetition counter register.

RM0090 Rev 17

529/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

The repetition counter is decremented:
•

At each counter overflow in upcounting mode,

•

At each counter underflow in downcounting mode,

•

At each counter overflow and at each counter underflow in center-aligned mode.
Although this limits the maximum number of repetition to 128 PWM cycles, it makes it
possible to update the duty cycle twice per PWM period. When refreshing compare
registers only once per PWM period in center-aligned mode, maximum resolution is
2xTck, due to the symmetry of the pattern.

The repetition counter is an auto-reload type; the repetition rate is maintained as defined by
the TIMx_RCR register value (refer to Figure 106). When the update event is generated by
software (by setting the UG bit in TIMx_EGR register) or by hardware through the slave
mode controller, it occurs immediately whatever the value of the repetition counter is and the
repetition counter is reloaded with the content of the TIMx_RCR register.
In center-aligned mode, for odd values of RCR, the update event occurs either on the
overflow or on the underflow depending on when the RCR register was written and when
the counter was started. If the RCR was written before starting the counter, the UEV occurs
on the overflow. If the RCR was written after starting the counter, the UEV occurs on the
underflow. For example for RCR = 3, the UEV is generated on each 4th overflow or
underflow event depending on when RCR was written.

530/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)
Figure 106. Update rate examples depending on mode and TIMx_RCR register
settings
(GJHDOLJQHGPRGH

&RXQWHUDOLJQHGPRGH
8SFRXQWLQJ

&RXQWHU
7,0[B&17

7,0[B5&5 

7,0[B5&5 

7,0[B5&5 

7,0[B5&5 

7,0[B5&5 
DQG
UHV\QFKURQL]DWLRQ

8(9

8(9

8(9

8(9

8(9
E\6:

8(9

'RZQFRXQWLQJ

E\6:

E\6:

8SGDWHHYHQW3UHORDGUHJLVWHUVWUDQVIHUUHGWRDFWLYHUHJLVWHUVDQGXSGDWHLQWHUUXSWJHQHUDWHG
8SGDWH(YHQWLIWKHUHSHWLWLRQFRXQWHUXQGHUIORZRFFXUVZKHQWKHFRXQWHULVHTXDOWRWKHDXWRUHORDGYDOXH

06Y9

RM0090 Rev 17

531/1747
588

Advanced-control timers (TIM1 and TIM8)

17.3.4

RM0090

Clock selection
The counter clock can be provided by the following clock sources:
•

Internal clock (CK_INT)

•

External clock mode1: external input pin

•

External clock mode2: external trigger input ETR

•

Internal trigger inputs (ITRx): using one timer as prescaler for another timer, for
example, the user can configure Timer 1 to act as a prescaler for Timer 2. Refer to
Using one timer as prescaler for another timer for more details.

Internal clock source (CK_INT)
If the slave mode controller is disabled (SMS=000), then the CEN, DIR (in the TIMx_CR1
register) and UG bits (in the TIMx_EGR register) are actual control bits and can be changed
only by software (except UG which remains cleared automatically). As soon as the CEN bit
is written to 1, the prescaler is clocked by the internal clock CK_INT.
Figure 107 shows the behavior of the control circuit and the upcounter in normal mode,
without prescaler.
Figure 107. Control circuit in normal mode, internal clock divided by 1

,QWHUQDOFORFN
&(1 &17B(1
8*

&17B,1,7

&RXQWHUFORFN &.B&17 &.B36&
&RXQWHUUHJLVWHU









 

 



  




069

External clock source mode 1
This mode is selected when SMS=111 in the TIMx_SMCR register. The counter can count at
each rising or falling edge on a selected input.

532/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)
Figure 108. TI2 external clock connection example
7,0[B60&5
76>@
RU 7,)
7,)
,75[
7,B('

7,

)LOWHU

(GJH
GHWHFWRU

7,)B5LVLQJ
7,)B)DOOLQJ

7,)3



7,)3
(75)

RU
RU

[[


75*,





(75)
&.B,17

,&)>@

&&3

7,0[B&&05

7,0[B&&(5

LQWHUQDOFORFN

(QFRGHU
PRGH
([WHUQDOFORFN
PRGH

([WHUQDOFORFN
PRGH


&.B36&

,QWHUQDOFORFN
PRGH

(&(

606>@

7,0[B60&5

069

For example, to configure the upcounter to count in response to a rising edge on the TI2
input, use the following procedure:

Note:

1.

Configure channel 2 to detect rising edges on the TI2 input by writing CC2S = ‘01’ in
the TIMx_CCMR1 register.

2.

Configure the input filter duration by writing the IC2F[3:0] bits in the TIMx_CCMR1
register (if no filter is needed, keep IC2F=0000).

3.

Select rising edge polarity by writing CC2P=0 and CC2NP=0 in the TIMx_CCER
register.

4.

Configure the timer in external clock mode 1 by writing SMS=111 in the TIMx_SMCR
register.

5.

Select TI2 as the trigger input source by writing TS=110 in the TIMx_SMCR register.

6.

Enable the counter by writing CEN=1 in the TIMx_CR1 register.

The capture prescaler is not used for triggering, so the user does not need to configure it.
When a rising edge occurs on TI2, the counter counts once and the TIF flag is set.
The delay between the rising edge on TI2 and the actual clock of the counter is due to the
resynchronization circuit on TI2 input.

RM0090 Rev 17

533/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Figure 109. Control circuit in external clock mode 1

7,
&17B(1
&RXQWHUFORFN &.B&17 &.B36&
&RXQWHUUHJLVWHU







7,)

:ULWH7,) 

069

External clock source mode 2
This mode is selected by writing ECE=1 in the TIMx_SMCR register.
The counter can count at each rising or falling edge on the external trigger input ETR.
Figure 110 gives an overview of the external trigger input block.
Figure 110. External trigger input block
RU 7,)
7,)

(75

RU
RU

75*,


(75SLQ


'LYLGHU


(753
I'76

)LOWHU
GRZQFRXQWHU

(75)
&.B,17

(73

(736>@

(7)>@

7,0[B60&5

7,0[B60&5

7,0[B60&5

LQWHUQDOFORFN

(QFRGHU
PRGH
([WHUQDOFORFN
PRGH

([WHUQDOFORFN
PRGH


&.B36&

,QWHUQDOFORFN
PRGH

(&(

606>@

7,0[B60&5

069

534/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)
For example, to configure the upcounter to count each 2 rising edges on ETR, use the
following procedure:
1.

As no filter is needed in this example, write ETF[3:0]=0000 in the TIMx_SMCR register.

2.

Set the prescaler by writing ETPS[1:0]=01 in the TIMx_SMCR register

3.

Select rising edge detection on the ETR pin by writing ETP=0 in the TIMx_SMCR
register

4.

Enable external clock mode 2 by writing ECE=1 in the TIMx_SMCR register.

5.

Enable the counter by writing CEN=1 in the TIMx_CR1 register.

The counter counts once each 2 ETR rising edges.
The delay between the rising edge on ETR and the actual clock of the counter is due to the
resynchronization circuit on the ETRP signal.
Figure 111. Control circuit in external clock mode 2

I &.B,17
&17B(1
(75

(753

(75)
&RXQWHUFORFN 
&.B,17 &.B36&
&RXQWHUUHJLVWHU







069

17.3.5

Capture/compare channels
Each Capture/Compare channel is built around a capture/compare register (including a
shadow register), a input stage for capture (with digital filter, multiplexing and prescaler) and
an output stage (with comparator and output control).
Figure 112 to Figure 115 give an overview of one Capture/Compare channel.
The input stage samples the corresponding TIx input to generate a filtered signal TIxF.
Then, an edge detector with polarity selection generates a signal (TIxFPx) which can be
used as trigger input by the slave mode controller or as the capture command. It is
prescaled before the capture register (ICxPS).

RM0090 Rev 17

535/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Figure 112. Capture/compare channel (example: channel 1 input stage)
7,)B('
7RWKHVODYHPRGHFRQWUROOHU
7,)B5LVLQJ

7,

)LOWHU
7,)
GRZQFRXQWHU

I'76

(GJH
GHWHFWRU

7,)B)DOOLQJ

 7,)3




7,)3

&&3&&13

,&)>@

75&

7,0[B&&(5

7,0[B&&05

7,)B5LVLQJ

IURPFKDQQHO
7,)B)DOOLQJ

IURPFKDQQHO

,&



IURPVODYHPRGH
FRQWUROOHU

,&36
'LYLGHU




&&6>@ ,&36>@

&&(

7,0[B&&05

7,0[B&&(5

069

The output stage generates an intermediate waveform that is then used for reference:
OCxRef (active high). The polarity acts at the end of the chain.
Figure 113. Capture/compare channel 1 main circuit
$3%%XV

5HDG&&5/

UHDGBLQBSURJUHVV

&&6>@
,&36

ZULWHBLQBSURJUHVV

,QSXW
PRGH

6 ZULWH&&5+

&DSWXUHFRPSDUHSUHORDGUHJLVWHU

5

5
FRPSDUHBWUDQVIHU

FDSWXUHBWUDQVIHU
&&6>@


ORZ

5HDG&&5+ 6

KLJK



LIELW

0&8SHULSKHUDOLQWHUIDFH

2XWSXW
PRGH

&17!&&5
&RXQWHU

&&*

&&6>@

8(9
&RPSDUDWRU

&&(

&&6>@

2&3(

&DSWXUHFRPSDUHVKDGRZUHJLVWHU
&DSWXUH

ZULWH&&5/

IURPWLPH
EDVHXQLW

2&3(

7,0B&&05

&17 &&5

7,0B(*5

069

536/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)
Figure 114. Output stage of capture/compare channel (channel 1 to 3)
7RWKHPDVWHUPRGH
FRQWUROOHU


(75)
µ¶
2&5()

2&B'7

&17!&&5 2XWSXW
PRGH
&17 &&5
FRQWUROOHU
2&[5()
2&5()

2&5()&

2XWSXW
VHOHFWRU

'HDGWLPH
JHQHUDWRU





[


2XWSXW
HQDEOH
FLUFXLW

2&

2XWSXW
HQDEOH
FLUFXLW

2&1

&&3



7,0B&&(5
2&1B'7
µ¶





[



&&1( &&( 7,0B&&(5
2&&( 2&0>@

'7*>@

&&1( &&(

&&13

02( 266, 2665

7,0B&&05

7,0B%'75

7,0B&&(5

7,0B&&(5

7,0B%'75

069

Figure 115. Output stage of capture/compare channel (channel 4)

7RWKHPDVWHU
PRGHFRQWUROOHU

(75




&17!&&5
&17!&&5

2XWSXW
PRGH
FRQWUROOHU

2XWSXW
HQDEOH
FLUFXLW

2&

&&3

2&5()

7,0B&&(5

&&(
02(
2&0>@
7,0B&&05

266,

2,6

7,0B&&(5
7,0B%'75
7,0B&5
069

The capture/compare block is made of one preload register and one shadow register. Write
and read always access the preload register.
In capture mode, captures are actually done in the shadow register, which is copied into the
preload register.
In compare mode, the content of the preload register is copied into the shadow register
which is compared to the counter.

RM0090 Rev 17

537/1747
588

Advanced-control timers (TIM1 and TIM8)

17.3.6

RM0090

Input capture mode
In Input capture mode, the Capture/Compare Registers (TIMx_CCRx) are used to latch the
value of the counter after a transition detected by the corresponding ICx signal. When a
capture occurs, the corresponding CCXIF flag (TIMx_SR register) is set and an interrupt or
a DMA request can be sent if they are enabled. If a capture occurs while the CCxIF flag was
already high, then the over-capture flag CCxOF (TIMx_SR register) is set. CCxIF can be
cleared by software by writing it to ‘0’ or by reading the captured data stored in the
TIMx_CCRx register. CCxOF is cleared when written to ‘0’.
The following example shows how to capture the counter value in TIMx_CCR1 when TI1
input rises. To do this, use the following procedure:
•

Select the active input: TIMx_CCR1 must be linked to the TI1 input, so write the CC1S
bits to 01 in the TIMx_CCMR1 register. As soon as CC1S becomes different from 00,
the channel is configured in input and the TIMx_CCR1 register becomes read-only.

•

Program the needed input filter duration with respect to the signal connected to the
timer (by programming ICxF bits in the TIMx_CCMRx register if the input is a TIx input).
Let’s imagine that, when toggling, the input signal is not stable during at must five
internal clock cycles. We must program a filter duration longer than these five clock
cycles. We can validate a transition on TI1 when 8 consecutive samples with the new
level have been detected (sampled at fDTS frequency). Then write IC1F bits to 0011 in
the TIMx_CCMR1 register.

•

Select the edge of the active transition on the TI1 channel by writing CC1P and CC1NP
bits to 0 in the TIMx_CCER register (rising edge in this case).

•

Program the input prescaler. In our example, we wish the capture to be performed at
each valid transition, so the prescaler is disabled (write IC1PS bits to ‘00’ in the
TIMx_CCMR1 register).

•

Enable capture from the counter into the capture register by setting the CC1E bit in the
TIMx_CCER register.

•

If needed, enable the related interrupt request by setting the CC1IE bit in the
TIMx_DIER register, and/or the DMA request by setting the CC1DE bit in the
TIMx_DIER register.

When an input capture occurs:
•

The TIMx_CCR1 register gets the value of the counter on the active transition.

•

CC1IF flag is set (interrupt flag). CC1OF is also set if at least two consecutive captures
occurred whereas the flag was not cleared.

•

An interrupt is generated depending on the CC1IE bit.

•

A DMA request is generated depending on the CC1DE bit.

In order to handle the overcapture, it is recommended to read the data before the
overcapture flag. This is to avoid missing an overcapture which could happen after reading
the flag and before reading the data.
Note:

538/1747

IC interrupt and/or DMA requests can be generated by software by setting the
corresponding CCxG bit in the TIMx_EGR register.

RM0090 Rev 17

RM0090

17.3.7

Advanced-control timers (TIM1 and TIM8)

PWM input mode
This mode is a particular case of input capture mode. The procedure is the same except:
•

Two ICx signals are mapped on the same TIx input.

•

These 2 ICx signals are active on edges with opposite polarity.

•

One of the two TIxFP signals is selected as trigger input and the slave mode controller
is configured in reset mode.

For example, user can measure the period (in TIMx_CCR1 register) and the duty cycle (in
TIMx_CCR2 register) of the PWM applied on TI1 using the following procedure (depending
on CK_INT frequency and prescaler value):
•

Select the active input for TIMx_CCR1: write the CC1S bits to 01 in the TIMx_CCMR1
register (TI1 selected).

•

Select the active polarity for TI1FP1 (used both for capture in TIMx_CCR1 and counter
clear): write the CC1P and CC1NP bits to ‘0’ (active on rising edge).

•

Select the active input for TIMx_CCR2: write the CC2S bits to 10 in the TIMx_CCMR1
register (TI1 selected).

•

Select the active polarity for TI1FP2 (used for capture in TIMx_CCR2): write the CC2P
and CC2NP bits to ‘1’ (active on falling edge).

•

Select the valid trigger input: write the TS bits to 101 in the TIMx_SMCR register
(TI1FP1 selected).

•

Configure the slave mode controller in reset mode: write the SMS bits to 100 in the
TIMx_SMCR register.

•

Enable the captures: write the CC1E and CC2E bits to ‘1’ in the TIMx_CCER register.
Figure 116. PWM input mode timing

17.3.8

Forced output mode
In output mode (CCxS bits = 00 in the TIMx_CCMRx register), each output compare signal
(OCxREF and then OCx/OCxN) can be forced to active or inactive level directly by software,
independently of any comparison between the output compare register and the counter.
To force an output compare signal (OCXREF/OCx) to its active level, the user just needs to
write 101 in the OCxM bits in the corresponding TIMx_CCMRx register. Thus OCXREF is

RM0090 Rev 17

539/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

forced high (OCxREF is always active high) and OCx get opposite value to CCxP polarity
bit.
For example: CCxP=0 (OCx active high) => OCx is forced to high level.
The OCxREF signal can be forced low by writing the OCxM bits to 100 in the TIMx_CCMRx
register.
Anyway, the comparison between the TIMx_CCRx shadow register and the counter is still
performed and allows the flag to be set. Interrupt and DMA requests can be sent
accordingly. This is described in the output compare mode section below.

17.3.9

Output compare mode
This function is used to control an output waveform or indicating when a period of time has
elapsed.
When a match is found between the capture/compare register and the counter, the output
compare function:
•

Assigns the corresponding output pin to a programmable value defined by the output
compare mode (OCxM bits in the TIMx_CCMRx register) and the output polarity (CCxP
bit in the TIMx_CCER register). The output pin can keep its level (OCXM=000), be set
active (OCxM=001), be set inactive (OCxM=010) or can toggle (OCxM=011) on match.

•

Sets a flag in the interrupt status register (CCxIF bit in the TIMx_SR register).

•

Generates an interrupt if the corresponding interrupt mask is set (CCXIE bit in the
TIMx_DIER register).

•

Sends a DMA request if the corresponding enable bit is set (CCxDE bit in the
TIMx_DIER register, CCDS bit in the TIMx_CR2 register for the DMA request
selection).

The TIMx_CCRx registers can be programmed with or without preload registers using the
OCxPE bit in the TIMx_CCMRx register.
In output compare mode, the update event UEV has no effect on OCxREF and OCx output.
The timing resolution is one count of the counter. Output compare mode can also be used to
output a single pulse (in One Pulse mode).
Procedure:
1.

Select the counter clock (internal, external, prescaler).

2.

Write the desired data in the TIMx_ARR and TIMx_CCRx registers.

3.

Set the CCxIE bit if an interrupt request is to be generated.

4.

5.

Select the output mode. For example:
–

Write OCxM = 011 to toggle OCx output pin when CNT matches CCRx

–

Write OCxPE = 0 to disable preload register

–

Write CCxP = 0 to select active high polarity

–

Write CCxE = 1 to enable the output

Enable the counter by setting the CEN bit in the TIMx_CR1 register.

The TIMx_CCRx register can be updated at any time by software to control the output
waveform, provided that the preload register is not enabled (OCxPE=’0’, else TIMx_CCRx
shadow register is updated only at the next update event UEV). An example is given in
Figure 117.

540/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)
Figure 117. Output compare mode, toggle on OC1.
:ULWH%KLQWKH&&5UHJLVWHU

7,0B&17



$

7,0B&&5

%

%

%

%

$

2&5() 2&

0DWFKGHWHFWHGRQ&&5
,QWHUUXSWJHQHUDWHGLIHQDEOHG
069

17.3.10

PWM mode
Pulse Width Modulation mode allows generating a signal with a frequency determined by
the value of the TIMx_ARR register and a duty cycle determined by the value of the
TIMx_CCRx register.
The PWM mode can be selected independently on each channel (one PWM per OCx
output) by writing ‘110’ (PWM mode 1) or ‘111’ (PWM mode 2) in the OCxM bits in the
TIMx_CCMRx register. The corresponding preload register must be enabled by setting the
OCxPE bit in the TIMx_CCMRx register, and eventually the auto-reload preload register (in
upcounting or center-aligned modes) by setting the ARPE bit in the TIMx_CR1 register.
As the preload registers are transferred to the shadow registers only when an update event
occurs, before starting the counter, the user must initialize all the registers by setting the UG
bit in the TIMx_EGR register.
OCx polarity is software programmable using the CCxP bit in the TIMx_CCER register. It
can be programmed as active high or active low. OCx output is enabled by a combination of
the CCxE, CCxNE, MOE, OSSI and OSSR bits (TIMx_CCER and TIMx_BDTR registers).
Refer to the TIMx_CCER register description for more details.
In PWM mode (1 or 2), TIMx_CNT and TIMx_CCRx are always compared to determine
whether TIMx_CCRx ≤TIMx_CNT or TIMx_CNT ≤TIMx_CCRx (depending on the direction
of the counter).
The timer is able to generate PWM in edge-aligned mode or center-aligned mode
depending on the CMS bits in the TIMx_CR1 register.

PWM edge-aligned mode
•

Upcounting configuration
Upcounting is active when the DIR bit in the TIMx_CR1 register is low. Refer to
Upcounting mode.
In the following example, we consider PWM mode 1. The reference PWM signal
OCxREF is high as long as TIMx_CNT < TIMx_CCRx else it becomes low. If the

RM0090 Rev 17

541/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

compare value in TIMx_CCRx is greater than the auto-reload value (in TIMx_ARR)
then OCxREF is held at ‘1’. If the compare value is 0 then OCxRef is held at ‘0’.
Figure 118 shows some edge-aligned PWM waveforms in an example where
TIMx_ARR=8.
Figure 118. Edge-aligned PWM waveforms (ARR=8)



&RXQWHUUHJLVWHU

&&5[ 





















2&;5()
&&[,)

2&;5()
&&5[ 
&&[,)

2&;5()

µ¶

&&5[!
&&[,)

2&;5()

µ¶

&&5[ 
&&[,)

069

•

Downcounting configuration
Downcounting is active when DIR bit in TIMx_CR1 register is high. Refer to
Downcounting mode
In PWM mode 1, the reference signal OCxRef is low as long as
TIMx_CNT > TIMx_CCRx else it becomes high. If the compare value in TIMx_CCRx is
greater than the auto-reload value in TIMx_ARR, then OCxREF is held at ‘1’. 0% PWM
is not possible in this mode.

PWM center-aligned mode
Center-aligned mode is active when the CMS bits in TIMx_CR1 register are different from
‘00’ (all the remaining configurations having the same effect on the OCxRef/OCx signals).
The compare flag is set when the counter counts up, when it counts down or both when it
counts up and down depending on the CMS bits configuration. The direction bit (DIR) in the
TIMx_CR1 register is updated by hardware and must not be changed by software. Refer to

542/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)
Center-aligned mode (up/down counting).
Figure 119 shows some center-aligned PWM waveforms in an example where:
•

TIMx_ARR=8,

•

PWM mode is the PWM mode 1,

•

The flag is set when the counter counts down corresponding to the center-aligned
mode 1 selected for CMS=01 in TIMx_CR1 register.
Figure 119. Center-aligned PWM waveforms (ARR=8)
#OUNTER REGISTER





































/#X2%&
##2X  
#-3
#-3
#-3

##X)&

/#X2%&
##2X  

#-3 OR 

##X)&
/#X2%&
##2X  

gg

#-3
#-3
#-3

##X)&

/#X2%&
##2X  

gg

#-3
#-3
#-3

##X)&

/#X2%&
##2X  
##X)&

gg

#-3
#-3
#-3
AIB

Hints on using center-aligned mode:
•

When starting in center-aligned mode, the current up-down configuration is used. It
means that the counter counts up or down depending on the value written in the DIR bit

RM0090 Rev 17

543/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

in the TIMx_CR1 register. Moreover, the DIR and CMS bits must not be changed at the
same time by the software.
•

•

17.3.11

Writing to the counter while running in center-aligned mode is not recommended as it
can lead to unexpected results. In particular:
–

The direction is not updated if the user writes a value in the counter greater than
the auto-reload value (TIMx_CNT>TIMx_ARR). For example, if the counter was
counting up, it will continue to count up.

–

The direction is updated if the user writes 0 or write the TIMx_ARR value in the
counter but no Update Event UEV is generated.

The safest way to use center-aligned mode is to generate an update by software
(setting the UG bit in the TIMx_EGR register) just before starting the counter and not to
write the counter while it is running.

Complementary outputs and dead-time insertion
The advanced-control timers (TIM1 and TIM8) can output two complementary signals and
manage the switching-off and the switching-on instants of the outputs.
This time is generally known as dead-time and it has to be adjust it depending on the
devices connected to the outputs and their characteristics (intrinsic delays of level-shifters,
delays due to power switches...)
User can select the polarity of the outputs (main output OCx or complementary OCxN)
independently for each output. This is done by writing to the CCxP and CCxNP bits in the
TIMx_CCER register.
The complementary signals OCx and OCxN are activated by a combination of several
control bits: the CCxE and CCxNE bits in the TIMx_CCER register and the MOE, OISx,
OISxN, OSSI and OSSR bits in the TIMx_BDTR and TIMx_CR2 registers. Refer to Table 95
for more details. In particular, the dead-time is activated when switching to the IDLE state
(MOE falling down to 0).
Dead-time insertion is enabled by setting both CCxE and CCxNE bits, and the MOE bit if the
break circuit is present. DTG[7:0] bits of the TIMx_BDTR register are used to control the
dead-time generation for all channels. From a reference waveform OCxREF, it generates 2
outputs OCx and OCxN. If OCx and OCxN are active high:
•

The OCx output signal is the same as the reference signal except for the rising edge,
which is delayed relative to the reference rising edge.

•

The OCxN output signal is the opposite of the reference signal except for the rising
edge, which is delayed relative to the reference falling edge.

If the delay is greater than the width of the active output (OCx or OCxN) then the
corresponding pulse is not generated.
The following figures show the relationships between the output signals of the dead-time
generator and the reference signal OCxREF. (we suppose CCxP=0, CCxNP=0, MOE=1,
CCxE=1 and CCxNE=1 in these examples)

544/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)
Figure 120. Complementary output with dead-time insertion.
2&[5()
2&[
GHOD\
2&[1
GHOD\

069

Figure 121. Dead-time waveforms with delay greater than the negative pulse.

2&[5()
2&[
GHOD\
2&[1

069

Figure 122. Dead-time waveforms with delay greater than the positive pulse.
2&[5()
2&[
2&[1
GHOD\

069

The dead-time delay is the same for each of the channels and is programmable with the
DTG bits in the TIMx_BDTR register. Refer to Section 17.4.18: TIM1 and TIM8 break and
dead-time register (TIMx_BDTR) for delay calculation.

Re-directing OCxREF to OCx or OCxN
In output mode (forced, output compare or PWM), OCxREF can be re-directed to the OCx
output or to OCxN output by configuring the CCxE and CCxNE bits in the TIMx_CCER
register.
This allows the user to send a specific waveform (such as PWM or static active level) on
one output while the complementary remains at its inactive level. Other possibilities are to

RM0090 Rev 17

545/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

have both outputs at inactive level or both outputs active and complementary with
dead-time.
Note:

When only OCxN is enabled (CCxE=0, CCxNE=1), it is not complemented and becomes
active as soon as OCxREF is high. For example, if CCxNP=0 then OCxN=OCxRef. On the
other hand, when both OCx and OCxN are enabled (CCxE=CCxNE=1) OCx becomes
active when OCxREF is high whereas OCxN is complemented and becomes active when
OCxREF is low.

17.3.12

Using the break function
When using the break function, the output enable signals and inactive levels are modified
according to additional control bits (MOE, OSSI and OSSR bits in the TIMx_BDTR register,
OISx and OISxN bits in the TIMx_CR2 register). In any case, the OCx and OCxN outputs
cannot be set both to active level at a given time. Refer to Table 95 for more details.
The break source can be either the break input pin or a clock failure event, generated by the
Clock Security System (CSS), from the Reset Clock Controller. For further information on
the Clock Security System, refer to Section 7.2.7: Clock security system (CSS).
When exiting from reset, the break circuit is disabled and the MOE bit is low. User can
enable the break function by setting the BKE bit in the TIMx_BDTR register. The break input
polarity can be selected by configuring the BKP bit in the same register. BKE and BKP can
be modified at the same time. When the BKE and BKP bits are written, a delay of 1 APB
clock cycle is applied before the writing is effective. Consequently, it is necessary to wait 1
APB clock period to correctly read back the bit after the write operation.
Because MOE falling edge can be asynchronous, a resynchronization circuit has been
inserted between the actual signal (acting on the outputs) and the synchronous control bit
(accessed in the TIMx_BDTR register). It results in some delays between the asynchronous
and the synchronous signals. In particular, if MOE is written to 1 whereas it was low, a delay
(dummy instruction) must be inserted before reading it correctly. This is because the user
writes an asynchronous signal, but reads a synchronous signal.
When a break occurs (selected level on the break input):

546/1747

•

The MOE bit is cleared asynchronously, putting the outputs in inactive state, idle state
or in reset state (selected by the OSSI bit). This feature functions even if the MCU
oscillator is off.

•

Each output channel is driven with the level programmed in the OISx bit in the
TIMx_CR2 register as soon as MOE=0. If OSSI=0 then the timer releases the enable
output else the enable output remains high.

•

When complementary outputs are used:
–

The outputs are first put in reset state inactive state (depending on the polarity).
This is done asynchronously so that it works even if no clock is provided to the
timer.

–

If the timer clock is still present, then the dead-time generator is reactivated in
order to drive the outputs with the level programmed in the OISx and OISxN bits
after a dead-time. Even in this case, OCx and OCxN cannot be driven to their

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)
active level together. Note that because of the resynchronization on MOE, the
dead-time duration is a bit longer than usual (around 2 ck_tim clock cycles).
–

Note:

If OSSI=0 then the timer releases the enable outputs else the enable outputs
remain or become high as soon as one of the CCxE or CCxNE bits is high.

•

The break status flag (BIF bit in the TIMx_SR register) is set. An interrupt can be
generated if the BIE bit in the TIMx_DIER register is set. A DMA request can be sent if
the BDE bit in the TIMx_DIER register is set.

•

If the AOE bit in the TIMx_BDTR register is set, the MOE bit is automatically set again
at the next update event UEV. This can be used to perform a regulation, for instance.
Else, MOE remains low until it is written to ‘1’ again. In this case, it can be used for
security and the break input can be connected to an alarm from power drivers, thermal
sensors or any security components.

The break inputs is acting on level. Thus, the MOE cannot be set while the break input is
active (neither automatically nor by software). In the meantime, the status flag BIF cannot
be cleared.
The break can be generated by the BRK input which has a programmable polarity and an
enable bit BKE in the TIMx_BDTR Register.
There are two solutions to generate a break:
•

By using the BRK input which has a programmable polarity and an enable bit BKE in
the TIMx_BDTR register

•

By software through the BG bit of the TIMx_EGR register.

In addition to the break input and the output management, a write protection has been
implemented inside the break circuit to safeguard the application. It allows freezing the
configuration of several parameters (dead-time duration, OCx/OCxN polarities and state
when disabled, OCxM configurations, break enable and polarity). The user can choose from
three levels of protection selected by the LOCK bits in the TIMx_BDTR register. Refer to
Section 17.4.18: TIM1 and TIM8 break and dead-time register (TIMx_BDTR). The LOCK
bits can be written only once after an MCU reset.

RM0090 Rev 17

547/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Figure 123 shows an example of behavior of the outputs in response to a break.
Figure 123. Output behavior in response to a break.
%5($. 02( 

2&[5()

2&[
2&[1QRWLPSOHPHQWHG&&[3 2,6[ 

2&[
2&[1QRWLPSOHPHQWHG&&[3 2,6[ 

2&[
2&[1QRWLPSOHPHQWHG&&[3 2,6[ 

2&[
2&[1QRWLPSOHPHQWHG&&[3 2,6[ 

2&[
GHOD\
GHOD\
2&[1
&&[( &&[3 2,6[ &&[1( &&[13 2,6[1 

GHOD\

2&[
GHOD\
GHOD\
2&[1
&&[( &&[3 2,6[ &&[1( &&[13 2,6[1 

GHOD\

2&[
2&[1
&&[( &&[3 2,6[ &&[1( &&[13 2,6[1 

GHOD\

2&[
2&[1
&&[( &&[3 2,6[ &&[1( &&[13 2,6[1 

GHOD\

2&[
2&[1
&&[( &&[3 &&[1( &&[13 2,6[ 2,6[1 RU2,6[ 2,6[1 

069

548/1747

RM0090 Rev 17

RM0090

17.3.13

Advanced-control timers (TIM1 and TIM8)

Clearing the OCxREF signal on an external event
The OCxREF signal for a given channel can be driven Low by applying a High level to the
ETRF input (OCxCE enable bit of the corresponding TIMx_CCMRx register set to ‘1’). The
OCxREF signal remains Low until the next update event, UEV, occurs.
This function can only be used in output compare and PWM modes, and does not work in
forced mode.
For example, the ETR signal can be connected to the output of a comparator to be used for
current handling. In this case, the ETR must be configured as follow:
1.

The External Trigger Prescaler should be kept off: bits ETPS[1:0] of the TIMx_SMCR
register set to ‘00’.

2.

The external clock mode 2 must be disabled: bit ECE of the TIMx_SMCR register set to
‘0’.

3.

The External Trigger Polarity (ETP) and the External Trigger Filter (ETF) can be
configured according to the user needs.

Figure 124 shows the behavior of the OCxREF signal when the ETRF Input becomes High,
for both values of the enable bit OCxCE. In this example, the timer TIMx is programmed in
PWM mode.
Figure 124. Clearing TIMx OCxREF

&&5[
&RXQWHU &17

(75)

2&[5() 2&[&( µ¶
2&[5() 2&[&( µ¶

(75)EHFRPHVKLJK

(75)VWLOOKLJK
06Y9

Note:

In case of a PWM with a 100% duty cycle (if CCRx>ARR), then OCxREF is enabled again at
the next counter overflow.

RM0090 Rev 17

549/1747
588

Advanced-control timers (TIM1 and TIM8)

17.3.14

RM0090

6-step PWM generation
When complementary outputs are used on a channel, preload bits are available on the
OCxM, CCxE and CCxNE bits. The preload bits are transferred to the shadow bits at the
COM commutation event. The user can thus program in advance the configuration for the
next step and change the configuration of all the channels at the same time. COM can be
generated by software by setting the COM bit in the TIMx_EGR register or by hardware (on
TRGI rising edge).
A flag is set when the COM event occurs (COMIF bit in the TIMx_SR register), which can
generate an interrupt (if the COMIE bit is set in the TIMx_DIER register) or a DMA request
(if the COMDE bit is set in the TIMx_DIER register).
Figure 125 describes the behavior of the OCx and OCxN outputs when a COM event
occurs, in 3 different examples of programmed configurations.
Figure 125. 6-step generation, COM example (OSSR=1)

550/1747

RM0090 Rev 17

RM0090

One-pulse mode
One-pulse mode (OPM) is a particular case of the previous modes. It allows the counter to
be started in response to a stimulus and to generate a pulse with a programmable length
after a programmable delay.
Starting the counter can be controlled through the slave mode controller. Generating the
waveform can be done in output compare mode or PWM mode. Select One-pulse mode by
setting the OPM bit in the TIMx_CR1 register. This makes the counter stop automatically at
the next update event UEV.
A pulse can be correctly generated only if the compare value is different from the counter
initial value. Before starting (when the timer is waiting for the trigger), the configuration must
be:
•

In upcounting: CNT < CCRx ≤ ARR (in particular, 0 < CCRx)

•

In downcounting: CNT > CCRx
Figure 126. Example of one pulse mode.
7,
2&5()
2&

7,0B$55

&RXQWHU

17.3.15

Advanced-control timers (TIM1 and TIM8)

7,0B&&5



ƚ>z

ƚWh>^

ƚ
069

For example the user may want to generate a positive pulse on OC1 with a length of tPULSE
and after a delay of tDELAY as soon as a positive edge is detected on the TI2 input pin.
Let’s use TI2FP2 as trigger 1:
•

Map TI2FP2 to TI2 by writing CC2S=’01’ in the TIMx_CCMR1 register.

•

TI2FP2 must detect a rising edge, write CC2P=’0’ and CC2NP=’0’ in the TIMx_CCER
register.

•

Configure TI2FP2 as trigger for the slave mode controller (TRGI) by writing TS=’110’ in
the TIMx_SMCR register.

•

TI2FP2 is used to start the counter by writing SMS to ‘110’ in the TIMx_SMCR register
(trigger mode).

RM0090 Rev 17

551/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

The OPM waveform is defined by writing the compare registers (taking into account the
clock frequency and the counter prescaler).
•

The tDELAY is defined by the value written in the TIMx_CCR1 register.

•

The tPULSE is defined by the difference between the auto-reload value and the compare
value (TIMx_ARR - TIMx_CCR1).

•

Let us say the user wants to build a waveform with a transition from ‘0’ to ‘1’ when a
compare match occurs and a transition from ‘1’ to ‘0’ when the counter reaches the
auto-reload value. To do this, enable PWM mode 2 by writing OC1M=111 in the
TIMx_CCMR1 register. The user can optionally enable the preload registers by writing
OC1PE=’1’ in the TIMx_CCMR1 register and ARPE in the TIMx_CR1 register. In this
case the compare value must be written in the TIMx_CCR1 register, the auto-reload
value in the TIMx_ARR register, generate an update by setting the UG bit and wait for
external trigger event on TI2. CC1P is written to ‘0’ in this example.

In our example, the DIR and CMS bits in the TIMx_CR1 register should be low.
The user only wants one pulse (Single mode), so '1’ must be written in the OPM bit in the
TIMx_CR1 register to stop the counter at the next update event (when the counter rolls over
from the auto-reload value back to 0). When OPM bit in the TIMx_CR1 register is set to '0',
so the Repetitive Mode is selected.
Particular case: OCx fast enable:
In One-pulse mode, the edge detection on TIx input set the CEN bit which enables the
counter. Then the comparison between the counter and the compare value makes the
output toggle. But several clock cycles are needed for these operations and it limits the
minimum delay tDELAY min we can get.
If the user wants to output a waveform with the minimum delay, the OCxFE bit in the
TIMx_CCMRx register must be set. Then OCxRef (and OCx) are forced in response to the
stimulus, without taking in account the comparison. Its new level is the same as if a compare
match had occurred. OCxFE acts only if the channel is configured in PWM1 or PWM2
mode.

17.3.16

Encoder interface mode
To select Encoder Interface mode write SMS=‘001’ in the TIMx_SMCR register if the
counter is counting on TI2 edges only, SMS=’010’ if it is counting on TI1 edges only and
SMS=’011’ if it is counting on both TI1 and TI2 edges.
Select the TI1 and TI2 polarity by programming the CC1P and CC2P bits in the TIMx_CCER
register. When needed, the user can program the input filter as well. CC1NP and CC2NP
must be kept low.
The two inputs TI1 and TI2 are used to interface to an incremental encoder. Refer to
Table 93. The counter is clocked by each valid transition on TI1FP1 or TI2FP2 (TI1 and TI2
after input filter and polarity selection, TI1FP1=TI1 if not filtered and not inverted,
TI2FP2=TI2 if not filtered and not inverted) assuming that it is enabled (CEN bit in
TIMx_CR1 register written to ‘1’). The sequence of transitions of the two inputs is evaluated
and generates count pulses as well as the direction signal. Depending on the sequence the
counter counts up or down, the DIR bit in the TIMx_CR1 register is modified by hardware
accordingly. The DIR bit is calculated at each transition on any input (TI1 or TI2), whatever
the counter is counting on TI1 only, TI2 only or both TI1 and TI2.
Encoder interface mode acts simply as an external clock with direction selection. This
means that the counter just counts continuously between 0 and the auto-reload value in the

552/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)
TIMx_ARR register (0 to ARR or ARR down to 0 depending on the direction). So user must
configure TIMx_ARR before starting. in the same way, the capture, compare, prescaler,
repetition counter, trigger output features continue to work as normal. Encoder mode and
External clock mode 2 are not compatible and must not be selected together.
In this mode, the counter is modified automatically following the speed and the direction of
the incremental encoder and its content, therefore, always represents the encoder’s
position. The count direction correspond to the rotation direction of the connected sensor.
Table 93 summarizes the possible combinations, assuming TI1 and TI2 do not switch at the
same time.
Table 93. Counting direction versus encoder signals
Active
edge

Level on opposite signal
(TI1FP1 for TI2, TI2FP2 for TI1)

Counting on
TI1 only

TI1FP1 signal

TI2FP2 signal

Rising

Falling

Rising

Falling

High

Down

Up

No Count

No Count

Low

Up

Down

No Count

No Count

Counting on
TI2 only

High

No Count

No Count

Up

Down

Low

No Count

No Count

Down

Up

Counting on
TI1 and TI2

High

Down

Up

Up

Down

Low

Up

Down

Down

Up

An external incremental encoder can be connected directly to the MCU without external
interface logic. However, comparators are normally be used to convert the encoder’s
differential outputs to digital signals. This greatly increases noise immunity. The third
encoder output which indicate the mechanical zero position, may be connected to an
external interrupt input and trigger a counter reset.
Figure 127 gives an example of counter operation, showing count signal generation and
direction control. It also shows how input jitter is compensated where both edges are
selected. This might occur if the sensor is positioned near to one of the switching points. For
this example we assume that the configuration is the following:
•

CC1S=’01’ (TIMx_CCMR1 register, TI1FP1 mapped on TI1).

•

CC2S=’01’ (TIMx_CCMR2 register, TI1FP2 mapped on TI2).

•

CC1P=’0’, CC1NP=’0’, and IC1F = ‘0000’ (TIMx_CCER register, TI1FP1 non-inverted,
TI1FP1=TI1).

•

CC2P=’0’, CC2NP=’0’, and IC2F = ‘0000’ (TIMx_CCER register, TI1FP2 non-inverted,
TI1FP2= TI2).

•

SMS=’011’ (TIMx_SMCR register, both inputs are active on both rising and falling
edges).

•

CEN=’1’ (TIMx_CR1 register, Counter enabled).

RM0090 Rev 17

553/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Figure 127. Example of counter operation in encoder interface mode.
IRUZDUG

MLWWHU

EDFNZDUG

MLWWHU

IRUZDUG

7,
7,

&RXQWHU

XS

GRZQ

XS

069

Figure 128 gives an example of counter behavior when TI1FP1 polarity is inverted (same
configuration as above except CC1P=’1’).
Figure 128. Example of encoder interface mode with TI1FP1 polarity inverted.
IRUZDUG

MLWWHU

EDFNZDUG

MLWWHU

IRUZDUG

7,
7,

&RXQWHU

GRZQ

XS

GRZQ

069

The timer, when configured in Encoder Interface mode provides information on the sensor’s
current position.The user can obtain dynamic information (speed, acceleration,
deceleration) by measuring the period between two encoder events using a second timer
configured in capture mode. The output of the encoder which indicates the mechanical zero
can be used for this purpose. Depending on the time between two events, the counter can
also be read at regular times. This can be done by latching the counter value into a third
input capture register if available (then the capture signal must be periodic and can be
generated by another timer). when available, it is also possible to read its value through a
DMA request generated by a real-time clock.

554/1747

RM0090 Rev 17

RM0090

17.3.17

Advanced-control timers (TIM1 and TIM8)

Timer input XOR function
The TI1S bit in the TIMx_CR2 register, allows the input filter of channel 1 to be connected to
the output of a XOR gate, combining the three input pins TIMx_CH1, TIMx_CH2 and
TIMx_CH3.
The XOR output can be used with all the timer input functions such as trigger or input
capture. An example of this feature used to interface Hall sensors is given in
Section 17.3.18.

17.3.18

Interfacing with Hall sensors
This is done using the advanced-control timers (TIM1 or TIM8) to generate PWM signals to
drive the motor and another timer TIMx (TIM2, TIM3, TIM4 or TIM5) referred to as
“interfacing timer” in Figure 129. The “interfacing timer” captures the 3 timer input pins
(TIMx_CH1, TIMx_CH2, and TIMx_CH3) connected through a XOR to the TI1 input channel
(selected by setting the TI1S bit in the TIMx_CR2 register).
The slave mode controller is configured in reset mode; the slave input is TI1F_ED. Thus,
each time one of the 3 inputs toggles, the counter restarts counting from 0. This creates a
time base triggered by any change on the Hall inputs.
On the “interfacing timer”, capture/compare channel 1 is configured in capture mode,
capture signal is TRC (see Figure 112). The captured value, which corresponds to the time
elapsed between 2 changes on the inputs, gives information about motor speed.
The “interfacing timer” can be used in output mode to generate a pulse which changes the
configuration of the channels of the advanced-control timer (TIM1 or TIM8) (by triggering a
COM event). The TIM1 timer is used to generate PWM signals to drive the motor. To do this,
the interfacing timer channel must be programmed so that a positive pulse is generated
after a programmed delay (in output compare or PWM mode). This pulse is sent to the
advanced-control timer (TIM1 or TIM8) through the TRGO output.
Example: the user wants to change the PWM configuration of the advanced-control timer
TIM1 after a programmed delay each time a change occurs on the Hall inputs connected to
one of the TIMx timers.
•

Configure 3 timer inputs ORed to the TI1 input channel by writing the TI1S bit in the
TIMx_CR2 register to ‘1’,

•

Program the time base: write the TIMx_ARR to the max value (the counter must be
cleared by the TI1 change. Set the prescaler to get a maximum counter period longer
than the time between 2 changes on the sensors,

•

Program channel 1 in capture mode (TRC selected): write the CC1S bits in the
TIMx_CCMR1 register to ‘11’. The user can also program the digital filter if needed,

•

Program channel 2 in PWM 2 mode with the desired delay: write the OC2M bits to ‘111’
and the CC2S bits to ‘00’ in the TIMx_CCMR1 register,

•

Select OC2REF as trigger output on TRGO: write the MMS bits in the TIMx_CR2
register to ‘101’,

In the advanced-control timer TIM1, the right ITR input must be selected as trigger input, the
timer is programmed to generate PWM signals, the capture/compare control signals are
preloaded (CCPC=1 in the TIMx_CR2 register) and the COM event is controlled by the
trigger input (CCUS=1 in the TIMx_CR2 register). The PWM control bits (CCxE, OCxM) are
written after a COM event for the next step (this can be done in an interrupt subroutine
generated by the rising edge of OC2REF).

RM0090 Rev 17

555/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Figure 129 describes this example.
Figure 129. Example of Hall sensor interface

7,+

,QWHUIDFLQJWLPHU

7,+
7,+

&RXQWHU &17
&&5
&&5

&$

&$

&

&$

&$%

&

75*2 2&5()

&20
2&

$GYDQFHGFRQWUROWLPHUV

2&1
2&

2&1

2&
2&1

:ULWH&&[(&&[1(
DQG2&[0IRUQH[WVWHS
DLE

556/1747

RM0090 Rev 17

RM0090

17.3.19

Advanced-control timers (TIM1 and TIM8)

TIMx and external trigger synchronization
The TIMx timer can be synchronized with an external trigger in several modes: Reset mode,
Gated mode and Trigger mode.

Slave mode: Reset mode
The counter and its prescaler can be reinitialized in response to an event on a trigger input.
Moreover, if the URS bit from the TIMx_CR1 register is low, an update event UEV is
generated. Then all the preloaded registers (TIMx_ARR, TIMx_CCRx) are updated.
In the following example, the upcounter is cleared in response to a rising edge on TI1 input:
•

Configure the channel 1 to detect rising edges on TI1. Configure the input filter duration
(in this example, we don’t need any filter, so we keep IC1F=0000). The capture
prescaler is not used for triggering, so there’s no need to configure it. The CC1S bits
select the input capture source only, CC1S = 01 in the TIMx_CCMR1 register. Write
CC1P=0 and CC1NP=’0’ in TIMx_CCER register to validate the polarity (and detect
rising edges only).

•

Configure the timer in reset mode by writing SMS=100 in TIMx_SMCR register. Select
TI1 as the input source by writing TS=101 in TIMx_SMCR register.

•

Start the counter by writing CEN=1 in the TIMx_CR1 register.

The counter starts counting on the internal clock, then behaves normally until TI1 rising
edge. When TI1 rises, the counter is cleared and restarts from 0. In the meantime, the
trigger flag is set (TIF bit in the TIMx_SR register) and an interrupt request, or a DMA
request can be sent if enabled (depending on the TIE and TDE bits in TIMx_DIER register).
The following figure shows this behavior when the auto-reload register TIMx_ARR=0x36.
The delay between the rising edge on TI1 and the actual reset of the counter is due to the
resynchronization circuit on TI1 input.
Figure 130. Control circuit in reset mode
7,

8*
&RXQWHUFORFN &.B&17 &.B36&
&RXQWHUUHJLVWHU































7,)
069

RM0090 Rev 17

557/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Slave mode: Gated mode
The counter can be enabled depending on the level of a selected input.
In the following example, the upcounter counts only when TI1 input is low:
•

Configure the channel 1 to detect low levels on TI1. Configure the input filter duration
(in this example, we don’t need any filter, so we keep IC1F=0000). The capture
prescaler is not used for triggering, so the user does not need to configure it. The
CC1S bits select the input capture source only, CC1S=01 in TIMx_CCMR1 register.
Write CC1P=1 and CC1NP=’0’ in TIMx_CCER register to validate the polarity (and
detect low level only).

•

Configure the timer in gated mode by writing SMS=101 in TIMx_SMCR register. Select
TI1 as the input source by writing TS=101 in TIMx_SMCR register.

•

Enable the counter by writing CEN=1 in the TIMx_CR1 register (in gated mode, the
counter doesn’t start if CEN=0, whatever is the trigger input level).

The counter starts counting on the internal clock as long as TI1 is low and stops as soon as
TI1 becomes high. The TIF flag in the TIMx_SR register is set both when the counter starts
or stops.
The delay between the rising edge on TI1 and the actual stop of the counter is due to the
resynchronization circuit on TI1 input.
Figure 131. Control circuit in gated mode
7,
&17B(1
&RXQWHUFORFN FNBFQW FNBSVF
&RXQWHUUHJLVWHU

   



   

7,)
:ULWH7,) 
069

558/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)

Slave mode: Trigger mode
The counter can start in response to an event on a selected input.
In the following example, the upcounter starts in response to a rising edge on TI2 input:
•

Configure the channel 2 to detect rising edges on TI2. Configure the input filter duration
(in this example, we don’t need any filter, so we keep IC2F=0000). The capture
prescaler is not used for triggering, so there’s no need to configure it. The CC2S bits
are configured to select the input capture source only, CC2S=01 in TIMx_CCMR1
register. Write CC2P=1 and CC2NP=0 in TIMx_CCER register to validate the polarity
(and detect low level only).

•

Configure the timer in trigger mode by writing SMS=110 in TIMx_SMCR register. Select
TI2 as the input source by writing TS=110 in TIMx_SMCR register.

When a rising edge occurs on TI2, the counter starts counting on the internal clock and the
TIF flag is set.
The delay between the rising edge on TI2 and the actual start of the counter is due to the
resynchronization circuit on TI2 input.
Figure 132. Control circuit in trigger mode
7,
&17B(1
&RXQWHUFORFN FNBFQW FNBSVF
&RXQWHUUHJLVWHU





  

7,)
069

Slave mode: external clock mode 2 + trigger mode
The external clock mode 2 can be used in addition to another slave mode (except external
clock mode 1 and encoder mode). In this case, the ETR signal is used as external clock
input, and another input can be selected as trigger input (in reset mode, gated mode or
trigger mode). It is recommended not to select ETR as TRGI through the TS bits of
TIMx_SMCR register.
In the following example, the upcounter is incremented at each rising edge of the ETR
signal as soon as a rising edge of TI1 occurs:
1.

2.

Configure the external trigger input circuit by programming the TIMx_SMCR register as
follows:
–

ETF = 0000: no filter

–

ETPS = 00: prescaler disabled

–

ETP = 0: detection of rising edges on ETR and ECE=1 to enable the external clock
mode 2.

Configure the channel 1 as follows, to detect rising edges on TI:
–

IC1F=0000: no filter.

–

The capture prescaler is not used for triggering and does not need to be
configured.

–

CC1S=01 in TIMx_CCMR1 register to select only the input capture source
RM0090 Rev 17

559/1747
588

Advanced-control timers (TIM1 and TIM8)
–
3.

RM0090

CC1P=0 and CC1NP=’0’ in TIMx_CCER register to validate the polarity (and
detect rising edge only).

Configure the timer in trigger mode by writing SMS=110 in TIMx_SMCR register. Select
TI1 as the input source by writing TS=101 in TIMx_SMCR register.

A rising edge on TI1 enables the counter and sets the TIF flag. The counter then counts on
ETR rising edges.
The delay between the rising edge of the ETR signal and the actual reset of the counter is
due to the resynchronization circuit on ETRP input.
Figure 133. Control circuit in external clock mode 2 + trigger mode
7,

&(1&17B(1

(75

&RXQWHUFORFN &.B&17 &.B36&

&RXQWHUUHJLVWHU







7,)
069

17.3.20

Timer synchronization
The TIM timers are linked together internally for timer synchronization or chaining. Refer to
Section 18.3.15: Timer synchronization for details.

Note:

The clock of the slave timer must be enabled prior to receive events from the master timer,
and must not be changed on-the-fly while triggers are received from the master timer.

17.3.21

Debug mode
When the microcontroller enters debug mode (Cortex®-M4 with FPU core halted), the TIMx
counter either continues to work normally or stops, depending on DBG_TIMx_STOP
configuration bit in DBG module. For more details, refer to Section 38.16.2: Debug support
for timers, watchdog, bxCAN and I2C.
For safety purposes, when the counter is stopped (DBG_TIMx_STOP = 1 in
DBGMCU_APBx_FZ register), the outputs are disabled (as if the MOE bit was reset). The
outputs can either be forced to an inactive state (OSSI bit = 1), or have their control taken
over by the GPIO controller (OSSI bit = 0) to force them to Hi-Z.

560/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)

17.4

TIM1 and TIM8 registers
Refer to Section 2.2 for a list of abbreviations used in register descriptions.
The peripheral registers must be written by half-words (16 bits) or words (32 bits). Read
accesses can be done by bytes (8 bits), half-word (16 bits) or words (32 bits).

17.4.1

TIM1 and TIM8 control register 1 (TIMx_CR1)
Address offset: 0x00
Reset value: 0x0000

15

14

13

12

Reserved

11

10

9

8

CKD[1:0]
rw

7
ARPE

rw

rw

6

5

CMS[1:0]
rw

rw

4

3

2

1

0

DIR

OPM

URS

UDIS

CEN

rw

rw

rw

rw

rw

Bits 15:10 Reserved, must be kept at reset value.
Bits 9:8 CKD[1:0]: Clock division
This bit-field indicates the division ratio between the timer clock (CK_INT) frequency and the
dead-time and sampling clock (tDTS)used by the dead-time generators and the digital filters
(ETR, TIx),
00: tDTS=tCK_INT
01: tDTS=2*tCK_INT
10: tDTS=4*tCK_INT
11: Reserved, do not program this value
Bit 7 ARPE: Auto-reload preload enable
0: TIMx_ARR register is not buffered
1: TIMx_ARR register is buffered
Bits 6:5 CMS[1:0]: Center-aligned mode selection
00: Edge-aligned mode. The counter counts up or down depending on the direction bit
(DIR).
01: Center-aligned mode 1. The counter counts up and down alternatively. Output compare
interrupt flags of channels configured in output (CCxS=00 in TIMx_CCMRx register) are set
only when the counter is counting down.
10: Center-aligned mode 2. The counter counts up and down alternatively. Output compare
interrupt flags of channels configured in output (CCxS=00 in TIMx_CCMRx register) are set
only when the counter is counting up.
11: Center-aligned mode 3. The counter counts up and down alternatively. Output compare
interrupt flags of channels configured in output (CCxS=00 in TIMx_CCMRx register) are set
both when the counter is counting up or down.
Note: It is not allowed to switch from edge-aligned mode to center-aligned mode as long as
the counter is enabled (CEN=1)
Bit 4 DIR: Direction
0: Counter used as upcounter
1: Counter used as downcounter
Note: This bit is read only when the timer is configured in Center-aligned mode or Encoder
mode.
Bit 3 OPM: One pulse mode
0: Counter is not stopped at update event
1: Counter stops counting at the next update event (clearing the bit CEN)

RM0090 Rev 17

561/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Bit 2 URS: Update request source
This bit is set and cleared by software to select the UEV event sources.
0: Any of the following events generate an update interrupt or DMA request if enabled.
These events can be:
–
Counter overflow/underflow
–
Setting the UG bit
–
Update generation through the slave mode controller
1: Only counter overflow/underflow generates an update interrupt or DMA request if
enabled.
Bit 1 UDIS: Update disable
This bit is set and cleared by software to enable/disable UEV event generation.
0: UEV enabled. The Update (UEV) event is generated by one of the following events:
–
Counter overflow/underflow
–
Setting the UG bit
–
Update generation through the slave mode controller
Buffered registers are then loaded with their preload values.
1: UEV disabled. The Update event is not generated, shadow registers keep their value
(ARR, PSC, CCRx). However the counter and the prescaler are reinitialized if the UG bit is
set or if a hardware reset is received from the slave mode controller.
Bit 0 CEN: Counter enable
0: Counter disabled
1: Counter enabled
Note: External clock, gated mode and encoder mode can work only if the CEN bit has been
previously set by software. However trigger mode can set the CEN bit automatically by
hardware.

17.4.2

TIM1 and TIM8 control register 2 (TIMx_CR2)
Address offset: 0x04
Reset value: 0x0000

15
Res.

14

13

12

11

10

9

8

7

OIS4

OIS3N

OIS3

OIS2N

OIS2

OIS1N

OIS1

TI1S

rw

rw

rw

rw

rw

rw

rw

rw

Bit 15 Reserved, must be kept at reset value.
Bit 14 OIS4: Output Idle state 4 (OC4 output)
refer to OIS1 bit
Bit 13 OIS3N: Output Idle state 3 (OC3N output)
refer to OIS1N bit
Bit 12 OIS3: Output Idle state 3 (OC3 output)
refer to OIS1 bit
Bit 11 OIS2N: Output Idle state 2 (OC2N output)
refer to OIS1N bit
Bit 10 OIS2: Output Idle state 2 (OC2 output)
refer to OIS1 bit

562/1747

RM0090 Rev 17

6

5

4

MMS[2:0]
rw

rw

rw

3

2

CCDS

CCUS

rw

rw

1
Res.

0
CCPC
rw

RM0090

Advanced-control timers (TIM1 and TIM8)

Bit 9 OIS1N: Output Idle state 1 (OC1N output)
0: OC1N=0 after a dead-time when MOE=0
1: OC1N=1 after a dead-time when MOE=0
Note: This bit can not be modified as long as LOCK level 1, 2 or 3 has been programmed
(LOCK bits in TIMx_BDTR register).
Bit 8 OIS1: Output Idle state 1 (OC1 output)
0: OC1=0 (after a dead-time if OC1N is implemented) when MOE=0
1: OC1=1 (after a dead-time if OC1N is implemented) when MOE=0
Note: This bit can not be modified as long as LOCK level 1, 2 or 3 has been programmed
(LOCK bits in TIMx_BDTR register).
Bit 7 TI1S: TI1 selection
0: The TIMx_CH1 pin is connected to TI1 input
1: The TIMx_CH1, CH2 and CH3 pins are connected to the TI1 input (XOR combination)
Bits 6:4 MMS[2:0]: Master mode selection
These bits allow to select the information to be sent in master mode to slave timers for
synchronization (TRGO). The combination is as follows:
000: Reset - the UG bit from the TIMx_EGR register is used as trigger output (TRGO). If the
reset is generated by the trigger input (slave mode controller configured in reset mode) then
the signal on TRGO is delayed compared to the actual reset.
001: Enable - the Counter Enable signal CNT_EN is used as trigger output (TRGO). It is
useful to start several timers at the same time or to control a window in which a slave timer is
enable. The Counter Enable signal is generated by a logic OR between CEN control bit and
the trigger input when configured in gated mode. When the Counter Enable signal is
controlled by the trigger input, there is a delay on TRGO, except if the master/slave mode is
selected (see the MSM bit description in TIMx_SMCR register).
010: Update - The update event is selected as trigger output (TRGO). For instance a master
timer can then be used as a prescaler for a slave timer.
011: Compare Pulse - The trigger output send a positive pulse when the CC1IF flag is to be
set (even if it was already high), as soon as a capture or a compare match occurred.
(TRGO).
100: Compare - OC1REF signal is used as trigger output (TRGO)
101: Compare - OC2REF signal is used as trigger output (TRGO)
110: Compare - OC3REF signal is used as trigger output (TRGO)
111: Compare - OC4REF signal is used as trigger output (TRGO)
Note: The clock of the slave timer and ADC must be enabled prior to receiving events from
the master timer, and must not be changed on-the-fly while triggers are received from
the master timer.
Bit 3 CCDS: Capture/compare DMA selection
0: CCx DMA request sent when CCx event occurs
1: CCx DMA requests sent when update event occurs

RM0090 Rev 17

563/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Bit 2 CCUS: Capture/compare control update selection
0: When capture/compare control bits are preloaded (CCPC=1), they are updated by setting
the COMG bit only
1: When capture/compare control bits are preloaded (CCPC=1), they are updated by setting
the COMG bit or when an rising edge occurs on TRGI
Note: This bit acts only on channels that have a complementary output.
Bit 1 Reserved, must be kept at reset value.
Bit 0 CCPC: Capture/compare preloaded control
0: CCxE, CCxNE and OCxM bits are not preloaded
1: CCxE, CCxNE and OCxM bits are preloaded, after having been written, they are updated
only when a commutation event (COM) occurs (COMG bit set or rising edge detected on
TRGI, depending on the CCUS bit).
Note: This bit acts only on channels that have a complementary output.

564/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)

17.4.3

TIM1 and TIM8 slave mode control register (TIMx_SMCR)
Address offset: 0x08
Reset value: 0x0000

15

14

ETP

ECE

rw

rw

13

12

11

ETPS[1:0]
rw

rw

10

9

8

ETF[3:0]
rw

rw

rw

7

6

MSM
rw

rw

5

4

TS[2:0]
rw

rw

3

2

Res.
rw

Res.

1

0

SMS[2:0]
rw

rw

rw

Bit 15 ETP: External trigger polarity
This bit selects whether ETR or ETR is used for trigger operations
0: ETR is non-inverted, active at high level or rising edge.
1: ETR is inverted, active at low level or falling edge.
Bit 14 ECE: External clock enable
This bit enables External clock mode 2.
0: External clock mode 2 disabled
1: External clock mode 2 enabled. The counter is clocked by any active edge on the ETRF
signal.
Note: 1: Setting the ECE bit has the same effect as selecting external clock mode 1 with
TRGI connected to ETRF (SMS=111 and TS=111).
2: It is possible to simultaneously use external clock mode 2 with the following slave
modes: reset mode, gated mode and trigger mode. Nevertheless, TRGI must not be
connected to ETRF in this case (TS bits must not be 111).
3: If external clock mode 1 and external clock mode 2 are enabled at the same time,
the external clock input is ETRF.
Bits 13:12 ETPS[1:0]: External trigger prescaler
External trigger signal ETRP frequency must be at most 1/4 of TIMxCLK frequency. A
prescaler can be enabled to reduce ETRP frequency. It is useful when inputting fast external
clocks.
00: Prescaler OFF
01: ETRP frequency divided by 2
10: ETRP frequency divided by 4
11: ETRP frequency divided by 8

RM0090 Rev 17

565/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Bits 11:8 ETF[3:0]: External trigger filter
This bit-field then defines the frequency used to sample ETRP signal and the length of the
digital filter applied to ETRP. The digital filter is made of an event counter in which N
consecutive events are needed to validate a transition on the output:
0000: No filter, sampling is done at fDTS
0001: fSAMPLING=fCK_INT, N=2
0010: fSAMPLING=fCK_INT, N=4
0011: fSAMPLING=fCK_INT, N=8
0100: fSAMPLING=fDTS/2, N=6
0101: fSAMPLING=fDTS/2, N=8
0110: fSAMPLING=fDTS/4, N=6
0111: fSAMPLING=fDTS/4, N=8
1000: fSAMPLING=fDTS/8, N=6
1001: fSAMPLING=fDTS/8, N=8
1010: fSAMPLING=fDTS/16, N=5
1011: fSAMPLING=fDTS/16, N=6
1100: fSAMPLING=fDTS/16, N=8
1101: fSAMPLING=fDTS/32, N=5
1110: fSAMPLING=fDTS/32, N=6
1111: fSAMPLING=fDTS/32, N=8
Bit 7 MSM: Master/slave mode
0: No action
1: The effect of an event on the trigger input (TRGI) is delayed to allow a perfect
synchronization between the current timer and its slaves (through TRGO). It is useful if we
want to synchronize several timers on a single external event.
Bits 6:4 TS[2:0]: Trigger selection
This bit-field selects the trigger input to be used to synchronize the counter.
000: Internal Trigger 0 (ITR0)
001: Internal Trigger 1 (ITR1)
010: Internal Trigger 2 (ITR2)
011: Internal Trigger 3 (ITR3)
100: TI1 Edge Detector (TI1F_ED)
101: Filtered Timer Input 1 (TI1FP1)
110: Filtered Timer Input 2 (TI2FP2)
111: External Trigger input (ETRF)
See Table 94 for more details on ITRx meaning for each Timer.
Note: These bits must be changed only when they are not used (e.g. when SMS=000) to
avoid wrong edge detections at the transition.
Bit 3 Reserved, must be kept at reset value.

566/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)

Bits 2:0 SMS: Slave mode selection
When external signals are selected the active edge of the trigger signal (TRGI) is linked to
the polarity selected on the external input (see Input Control register and Control Register
description.
000: Slave mode disabled - if CEN = ‘1’ then the prescaler is clocked directly by the internal
clock.
001: Encoder mode 1 - Counter counts up/down on TI2FP1 edge depending on TI1FP2
level.
010: Encoder mode 2 - Counter counts up/down on TI1FP2 edge depending on TI2FP1
level.
011: Encoder mode 3 - Counter counts up/down on both TI1FP1 and TI2FP2 edges
depending on the level of the other input.
100: Reset Mode - Rising edge of the selected trigger input (TRGI) reinitializes the counter
and generates an update of the registers.
101: Gated Mode - The counter clock is enabled when the trigger input (TRGI) is high. The
counter stops (but is not reset) as soon as the trigger becomes low. Both start and stop of
the counter are controlled.
110: Trigger Mode - The counter starts at a rising edge of the trigger TRGI (but it is not
reset). Only the start of the counter is controlled.
111: External Clock Mode 1 - Rising edges of the selected trigger (TRGI) clock the counter.
Note: The gated mode must not be used if TI1F_ED is selected as the trigger input
(TS=’100’). Indeed, TI1F_ED outputs 1 pulse for each transition on TI1F, whereas the
gated mode checks the level of the trigger signal.
The clock of the slave timer must be enabled prior to receiving events from the master
timer, and must not be changed on-the-fly while triggers are received from the master
timer.

Table 94. TIMx Internal trigger connection

17.4.4

Slave TIM

ITR0 (TS = 000)

ITR1 (TS = 001)

ITR2 (TS = 010)

ITR3 (TS = 011)

TIM1

TIM5_TRGO

TIM2_TRGO

TIM3_TRGO

TIM4_TRGO

TIM8

TIM1_TRGO

TIM2_TRGO

TIM4_TRGO

TIM5_TRGO

TIM1 and TIM8 DMA/interrupt enable register (TIMx_DIER)
Address offset: 0x0C
Reset value: 0x0000

15
Res.

14
TDE
rw

13

12

11

10

9

COMDE CC4DE CC3DE CC2DE CC1DE
rw

rw

rw

rw

rw

8

7

6

5

4

3

2

1

0

UDE

BIE

TIE

COMIE

CC4IE

CC3IE

CC2IE

CC1IE

UIE

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bit 15 Reserved, must be kept at reset value.
Bit 14 TDE: Trigger DMA request enable
0: Trigger DMA request disabled
1: Trigger DMA request enabled
Bit 13 COMDE: COM DMA request enable
0: COM DMA request disabled
1: COM DMA request enabled

RM0090 Rev 17

567/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Bit 12 CC4DE: Capture/Compare 4 DMA request enable
0: CC4 DMA request disabled
1: CC4 DMA request enabled
Bit 11 CC3DE: Capture/Compare 3 DMA request enable
0: CC3 DMA request disabled
1: CC3 DMA request enabled
Bit 10 CC2DE: Capture/Compare 2 DMA request enable
0: CC2 DMA request disabled
1: CC2 DMA request enabled
Bit 9 CC1DE: Capture/Compare 1 DMA request enable
0: CC1 DMA request disabled
1: CC1 DMA request enabled
Bit 8 UDE: Update DMA request enable
0: Update DMA request disabled
1: Update DMA request enabled
Bit 7 BIE: Break interrupt enable
0: Break interrupt disabled
1: Break interrupt enabled
Bit 6 TIE: Trigger interrupt enable
0: Trigger interrupt disabled
1: Trigger interrupt enabled
Bit 5 COMIE: COM interrupt enable
0: COM interrupt disabled
1: COM interrupt enabled
Bit 4 CC4IE: Capture/Compare 4 interrupt enable
0: CC4 interrupt disabled
1: CC4 interrupt enabled
Bit 3 CC3IE: Capture/Compare 3 interrupt enable
0: CC3 interrupt disabled
1: CC3 interrupt enabled
Bit 2 CC2IE: Capture/Compare 2 interrupt enable
0: CC2 interrupt disabled
1: CC2 interrupt enabled
Bit 1 CC1IE: Capture/Compare 1 interrupt enable
0: CC1 interrupt disabled
1: CC1 interrupt enabled
Bit 0 UIE: Update interrupt enable
0: Update interrupt disabled
1: Update interrupt enabled

568/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)

17.4.5

TIM1 and TIM8 status register (TIMx_SR)
Address offset: 0x10
Reset value: 0x0000

15

14
Reserved

13

12

11

10

9

CC4OF CC3OF CC2OF CC1OF
rc_w0

rc_w0

rc_w0

rc_w0

8

7

6

5

4

3

2

1

0

Res.

BIF

TIF

COMIF

CC4IF

CC3IF

CC2IF

CC1IF

UIF

Res.

rc_w0

rc_w0

rc_w0

rc_w0

rc_w0

rc_w0

rc_w0

rc_w0

Bits 15:13 Reserved, must be kept at reset value.
Bit 12 CC4OF: Capture/Compare 4 overcapture flag
refer to CC1OF description
Bit 11 CC3OF: Capture/Compare 3 overcapture flag
refer to CC1OF description
Bit 10 CC2OF: Capture/Compare 2 overcapture flag
refer to CC1OF description
Bit 9 CC1OF: Capture/Compare 1 overcapture flag
This flag is set by hardware only when the corresponding channel is configured in input
capture mode. It is cleared by software by writing it to ‘0’.
0: No overcapture has been detected.
1: The counter value has been captured in TIMx_CCR1 register while CC1IF flag was
already set
Bit 8 Reserved, must be kept at reset value.
Bit 7 BIF: Break interrupt flag
This flag is set by hardware as soon as the break input goes active. It can be cleared by
software if the break input is not active.
0: No break event occurred.
1: An active level has been detected on the break input.
Bit 6 TIF: Trigger interrupt flag
This flag is set by hardware on trigger event (active edge detected on TRGI input when the
slave mode controller is enabled in all modes but gated mode. It is set when the counter
starts or stops when gated mode is selected. It is cleared by software.
0: No trigger event occurred.
1: Trigger interrupt pending.
Bit 5 COMIF: COM interrupt flag
This flag is set by hardware on COM event (when Capture/compare Control bits - CCxE,
CCxNE, OCxM - have been updated). It is cleared by software.
0: No COM event occurred.
1: COM interrupt pending.
Bit 4 CC4IF: Capture/Compare 4 interrupt flag
refer to CC1IF description
Bit 3 CC3IF: Capture/Compare 3 interrupt flag
refer to CC1IF description

RM0090 Rev 17

569/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Bit 2 CC2IF: Capture/Compare 2 interrupt flag
refer to CC1IF description
Bit 1 CC1IF: Capture/Compare 1 interrupt flag
If channel CC1 is configured as output:
This flag is set by hardware when the counter matches the compare value, with some
exception in center-aligned mode (refer to the CMS bits in the TIMx_CR1 register
description). It is cleared by software.
0: No match.
1: The content of the counter TIMx_CNT matches the content of the TIMx_CCR1 register.
When the contents of TIMx_CCR1 are greater than the contents of TIMx_ARR, the CC1IF
bit goes high on the counter overflow (in upcounting and up/down-counting modes) or
underflow (in downcounting mode)
If channel CC1 is configured as input:
This bit is set by hardware on a capture. It is cleared by software or by reading the
TIMx_CCR1 register.
0: No input capture occurred
1: The counter value has been captured in TIMx_CCR1 register (An edge has been
detected on IC1 which matches the selected polarity)
Bit 0 UIF: Update interrupt flag
This bit is set by hardware on an update event. It is cleared by software.
0: No update occurred.
1: Update interrupt pending. This bit is set by hardware when the registers are updated:
– At overflow or underflow regarding the repetition counter value (update if repetition counter
= 0) and if the UDIS=0 in the TIMx_CR1 register.
– When CNT is reinitialized by software using the UG bit in TIMx_EGR register, if URS=0 and
UDIS=0 in the TIMx_CR1 register.
– When CNT is reinitialized by a trigger event (refer to Section 17.4.3: TIM1 and TIM8 slave
mode control register (TIMx_SMCR)), if URS=0 and UDIS=0 in the TIMx_CR1 register.

17.4.6

TIM1 and TIM8 event generation register (TIMx_EGR)
Address offset: 0x14
Reset value: 0x0000

15

14

13

12

11

Reserved

10

9

8

7

6

5

4

3

2

1

0

BG

TG

COMG

CC4G

CC3G

CC2G

CC1G

UG

w

w

w

w

w

w

w

w

Bits 15:8 Reserved, must be kept at reset value.
Bit 7 BG: Break generation
This bit is set by software in order to generate an event, it is automatically cleared by
hardware.
0: No action
1: A break event is generated. MOE bit is cleared and BIF flag is set. Related interrupt or
DMA transfer can occur if enabled.
Bit 6 TG: Trigger generation
This bit is set by software in order to generate an event, it is automatically cleared by
hardware.
0: No action
1: The TIF flag is set in TIMx_SR register. Related interrupt or DMA transfer can occur if
enabled.

570/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)

Bit 5 COMG: Capture/Compare control update generation
This bit can be set by software, it is automatically cleared by hardware
0: No action
1: When CCPC bit is set, it allows to update CCxE, CCxNE and OCxM bits
Note: This bit acts only on channels having a complementary output.
Bit 4 CC4G: Capture/Compare 4 generation
refer to CC1G description
Bit 3 CC3G: Capture/Compare 3 generation
refer to CC1G description
Bit 2 CC2G: Capture/Compare 2 generation
refer to CC1G description
Bit 1 CC1G: Capture/Compare 1 generation
This bit is set by software in order to generate an event, it is automatically cleared by
hardware.
0: No action
1: A capture/compare event is generated on channel 1:
If channel CC1 is configured as output:
CC1IF flag is set, Corresponding interrupt or DMA request is sent if enabled.
If channel CC1 is configured as input:
The current value of the counter is captured in TIMx_CCR1 register. The CC1IF flag is set,
the corresponding interrupt or DMA request is sent if enabled. The CC1OF flag is set if the
CC1IF flag was already high.
Bit 0 UG: Update generation
This bit can be set by software, it is automatically cleared by hardware.
0: No action
1: Reinitialize the counter and generates an update of the registers. Note that the prescaler
counter is cleared too (anyway the prescaler ratio is not affected). The counter is cleared if
the center-aligned mode is selected or if DIR=0 (upcounting), else it takes the auto-reload
value (TIMx_ARR) if DIR=1 (downcounting).

RM0090 Rev 17

571/1747
588

Advanced-control timers (TIM1 and TIM8)

17.4.7

RM0090

TIM1 and TIM8 capture/compare mode register 1 (TIMx_CCMR1)
Address offset: 0x18
Reset value: 0x0000
The channels can be used in input (capture mode) or in output (compare mode). The
direction of a channel is defined by configuring the corresponding CCxS bits. All the other
bits of this register have a different function in input and in output mode. For a given bit,
OCxx describes its function when the channel is configured in output, ICxx describes its
function when the channel is configured in input. So the user must take care that the same
bit can have a different meaning for the input stage and for the output stage.

15

14

OC2
CE

13

12

OC2M[2:0]
IC2F[3:0]

rw

rw

rw

11

10

OC2
PE

OC2
FE

9

8

CC2S[1:0]

7

IC2PSC[1:0]
rw

rw

rw

6

OC1
CE

5

4

OC1M[2:0]
IC1F[3:0]

rw

rw

rw

rw

rw

3

2

OC1
PE

OC1
FE

1

0

CC1S[1:0]

IC1PSC[1:0]
rw

rw

rw

rw

rw

Output compare mode:
Bit 15 OC2CE: Output Compare 2 clear enable
Bits 14:12 OC2M[2:0]: Output Compare 2 mode
Bit 11 OC2PE: Output Compare 2 preload enable
Bit 10 OC2FE: Output Compare 2 fast enable
Bits 9:8 CC2S[1:0]: Capture/Compare 2 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC2 channel is configured as output
01: CC2 channel is configured as input, IC2 is mapped on TI2
10: CC2 channel is configured as input, IC2 is mapped on TI1
11: CC2 channel is configured as input, IC2 is mapped on TRC. This mode is working only if
an internal trigger input is selected through the TS bit (TIMx_SMCR register)
Note: CC2S bits are writable only when the channel is OFF (CC2E = ‘0’ in TIMx_CCER).
Bit 7 OC1CE: Output Compare 1 clear enable
OC1CE: Output Compare 1 Clear Enable
0: OC1Ref is not affected by the ETRF Input
1: OC1Ref is cleared as soon as a High level is detected on ETRF input

572/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)

Bits 6:4 OC1M: Output Compare 1 mode
These bits define the behavior of the output reference signal OC1REF from which OC1 and
OC1N are derived. OC1REF is active high whereas OC1 and OC1N active level depends
on CC1P and CC1NP bits.
000: Frozen - The comparison between the output compare register TIMx_CCR1 and the
counter TIMx_CNT has no effect on the outputs.(this mode is used to generate a timing
base).
001: Set channel 1 to active level on match. OC1REF signal is forced high when the counter
TIMx_CNT matches the capture/compare register 1 (TIMx_CCR1).
010: Set channel 1 to inactive level on match. OC1REF signal is forced low when the
counter TIMx_CNT matches the capture/compare register 1 (TIMx_CCR1).
011: Toggle - OC1REF toggles when TIMx_CNT=TIMx_CCR1.
100: Force inactive level - OC1REF is forced low.
101: Force active level - OC1REF is forced high.
110: PWM mode 1 - In upcounting, channel 1 is active as long as TIMx_CNTTIMx_CCR1 else active (OC1REF=’1’).
111: PWM mode 2 - In upcounting, channel 1 is inactive as long as TIMx_CNTTIMx_CCR1 else
inactive.
Note: 1: These bits can not be modified as long as LOCK level 3 has been programmed
(LOCK bits in TIMx_BDTR register) and CC1S=’00’ (the channel is configured in
output).
2: In PWM mode 1 or 2, the OCREF level changes only when the result of the
comparison changes or when the output compare mode switches from “frozen” mode
to “PWM” mode.
3: On channels having a complementary output, this bit field is preloaded. If the CCPC
bit is set in the TIMx_CR2 register then the OC1M active bits take the new value from
the preloaded bits only when a COM event is generated.
Bit 3 OC1PE: Output Compare 1 preload enable
0: Preload register on TIMx_CCR1 disabled. TIMx_CCR1 can be written at anytime, the
new value is taken in account immediately.
1: Preload register on TIMx_CCR1 enabled. Read/Write operations access the preload
register. TIMx_CCR1 preload value is loaded in the active register at each update event.
Note: 1: These bits can not be modified as long as LOCK level 3 has been programmed
(LOCK bits in TIMx_BDTR register) and CC1S=’00’ (the channel is configured in
output).
2: The PWM mode can be used without validating the preload register only in one
pulse mode (OPM bit set in TIMx_CR1 register). Else the behavior is not guaranteed.
Bit 2 OC1FE: Output Compare 1 fast enable
This bit is used to accelerate the effect of an event on the trigger in input on the CC output.
0: CC1 behaves normally depending on counter and CCR1 values even when the trigger is
ON. The minimum delay to activate CC1 output when an edge occurs on the trigger input is
5 clock cycles.
1: An active edge on the trigger input acts like a compare match on CC1 output. Then, OC is
set to the compare level independently from the result of the comparison. Delay to sample
the trigger input and to activate CC1 output is reduced to 3 clock cycles. OCFE acts only if
the channel is configured in PWM1 or PWM2 mode.

RM0090 Rev 17

573/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Bits 1:0 CC1S: Capture/Compare 1 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC1 channel is configured as output
01: CC1 channel is configured as input, IC1 is mapped on TI1
10: CC1 channel is configured as input, IC1 is mapped on TI2
11: CC1 channel is configured as input, IC1 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC1S bits are writable only when the channel is OFF (CC1E = ‘0’ in TIMx_CCER).

Input capture mode
Bits 15:12 IC2F: Input capture 2 filter
Bits 11:10 IC2PSC[1:0]: Input capture 2 prescaler
Bits 9:8 CC2S: Capture/Compare 2 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC2 channel is configured as output
01: CC2 channel is configured as input, IC2 is mapped on TI2
10: CC2 channel is configured as input, IC2 is mapped on TI1
11: CC2 channel is configured as input, IC2 is mapped on TRC. This mode is working only if an
internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC2S bits are writable only when the channel is OFF (CC2E = ‘0’ in TIMx_CCER).
Bits 7:4 IC1F[3:0]: Input capture 1 filter
This bit-field defines the frequency used to sample TI1 input and the length of the digital filter applied
to TI1. The digital filter is made of an event counter in which N consecutive events are needed to
validate a transition on the output:
0000: No filter, sampling is done at fDTS
0001: fSAMPLING=fCK_INT, N=2
0010: fSAMPLING=fCK_INT, N=4
0011: fSAMPLING=fCK_INT, N=8
0100: fSAMPLING=fDTS/2, N=6
0101: fSAMPLING=fDTS/2, N=8
0110: fSAMPLING=fDTS/4, N=6
0111: fSAMPLING=fDTS/4, N=8
1000: fSAMPLING=fDTS/8, N=6
1001: fSAMPLING=fDTS/8, N=8
1010: fSAMPLING=fDTS/16, N=5
1011: fSAMPLING=fDTS/16, N=6
1100: fSAMPLING=fDTS/16, N=8
1101: fSAMPLING=fDTS/32, N=5
1110: fSAMPLING=fDTS/32, N=6
1111: fSAMPLING=fDTS/32, N=8
Bits 3:2 IC1PSC: Input capture 1 prescaler
This bit-field defines the ratio of the prescaler acting on CC1 input (IC1).
The prescaler is reset as soon as CC1E=’0’ (TIMx_CCER register).
00: no prescaler, capture is done each time an edge is detected on the capture input
01: capture is done once every 2 events
10: capture is done once every 4 events
11: capture is done once every 8 events

574/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)

Bits 1:0 CC1S: Capture/Compare 1 Selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC1 channel is configured as output
01: CC1 channel is configured as input, IC1 is mapped on TI1
10: CC1 channel is configured as input, IC1 is mapped on TI2
11: CC1 channel is configured as input, IC1 is mapped on TRC. This mode is working only if an
internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC1S bits are writable only when the channel is OFF (CC1E = ‘0’ in TIMx_CCER).

17.4.8

TIM1 and TIM8 capture/compare mode register 2 (TIMx_CCMR2)
Address offset: 0x1C
Reset value: 0x0000
Refer to the above CCMR1 register description.

15

14

OC4
CE

13

12

OC4M[2:0]
IC4F[3:0]

rw

rw

rw

11

10

OC4
PE

OC4
FE

9

8

CC4S[1:0]

7

IC4PSC[1:0]
rw

rw

rw

6

OC3
CE.

5

4

OC3M[2:0]
IC3F[3:0]

rw

rw

rw

rw

rw

3

2

OC3
PE

OC3
FE

1

0

CC3S[1:0]

IC3PSC[1:0]
rw

rw

rw

rw

rw

Output compare mode
Bit 15 OC4CE: Output compare 4 clear enable
Bits 14:12 OC4M: Output compare 4 mode
Bit 11 OC4PE: Output compare 4 preload enable
Bit 10 OC4FE: Output compare 4 fast enable
Bits 9:8 CC4S: Capture/Compare 4 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC4 channel is configured as output
01: CC4 channel is configured as input, IC4 is mapped on TI4
10: CC4 channel is configured as input, IC4 is mapped on TI3
11: CC4 channel is configured as input, IC4 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC4S bits are writable only when the channel is OFF (CC4E = ‘0’ in TIMx_CCER).
Bit 7 OC3CE: Output compare 3 clear enable
Bits 6:4 OC3M: Output compare 3 mode
Bit 3 OC3PE: Output compare 3 preload enable
Bit 2 OC3FE: Output compare 3 fast enable
Bits 1:0 CC3S: Capture/Compare 3 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC3 channel is configured as output
01: CC3 channel is configured as input, IC3 is mapped on TI3
10: CC3 channel is configured as input, IC3 is mapped on TI4
11: CC3 channel is configured as input, IC3 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC3S bits are writable only when the channel is OFF (CC3E = ‘0’ in TIMx_CCER).

RM0090 Rev 17

575/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Input capture mode
Bits 15:12 IC4F: Input capture 4 filter
Bits 11:10 IC4PSC: Input capture 4 prescaler
Bits 9:8 CC4S: Capture/Compare 4 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC4 channel is configured as output
01: CC4 channel is configured as input, IC4 is mapped on TI4
10: CC4 channel is configured as input, IC4 is mapped on TI3
11: CC4 channel is configured as input, IC4 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC4S bits are writable only when the channel is OFF (CC4E = ‘0’ in TIMx_CCER).
Bits 7:4 IC3F: Input capture 3 filter
Bits 3:2 IC3PSC: Input capture 3 prescaler
Bits 1:0 CC3S: Capture/compare 3 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC3 channel is configured as output
01: CC3 channel is configured as input, IC3 is mapped on TI3
10: CC3 channel is configured as input, IC3 is mapped on TI4
11: CC3 channel is configured as input, IC3 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC3S bits are writable only when the channel is OFF (CC3E = ‘0’ in TIMx_CCER).

17.4.9

TIM1 and TIM8 capture/compare enable register (TIMx_CCER)
Address offset: 0x20
Reset value: 0x0000

15

14

Reserved

13

12

CC4P

CC4E

rw

rw

11

10

CC3NP CC3NE
rw

rw

9

8

CC3P

CC3E

rw

rw

7

6

CC2NP CC2NE
rw

rw

5

4

CC2P

CC2E

rw

rw

Bits 15:14 Reserved, must be kept at reset value.
Bit 13 CC4P: Capture/Compare 4 output polarity
refer to CC1P description
Bit 12 CC4E: Capture/Compare 4 output enable
refer to CC1E description
Bit 11 CC3NP: Capture/Compare 3 complementary output polarity
refer to CC1NP description
Bit 10 CC3NE: Capture/Compare 3 complementary output enable
refer to CC1NE description
Bit 9 CC3P: Capture/Compare 3 output polarity
refer to CC1P description
Bit 8 CC3E: Capture/Compare 3 output enable
refer to CC1E description

576/1747

RM0090 Rev 17

3

2

CC1NP CC1NE
rw

rw

1

0

CC1P

CC1E

rw

rw

RM0090

Advanced-control timers (TIM1 and TIM8)

Bit 7 CC2NP: Capture/Compare 2 complementary output polarity
refer to CC1NP description
Bit 6 CC2NE: Capture/Compare 2 complementary output enable
refer to CC1NE description
Bit 5 CC2P: Capture/Compare 2 output polarity
refer to CC1P description
Bit 4 CC2E: Capture/Compare 2 output enable
refer to CC1E description
Bit 3 CC1NP: Capture/Compare 1 complementary output polarity
CC1 channel configured as output:
0: OC1N active high.
1: OC1N active low.
CC1 channel configured as input:
This bit is used in conjunction with CC1P to define the polarity of TI1FP1 and TI2FP1. Refer
to CC1P description.
Note: On channels having a complementary output, this bit is preloaded. If the CCPC bit is
set in the TIMx_CR2 register then the CC1NP active bit takes the new value from the
preloaded bit only when a Commutation event is generated.
Note: This bit is not writable as soon as LOCK level 2 or 3 has been programmed (LOCK bits
in TIMx_BDTR register) and CC1S=”00” (the channel is configured in output).
Bit 2 CC1NE: Capture/Compare 1 complementary output enable
0: Off - OC1N is not active. OC1N level is then function of MOE, OSSI, OSSR, OIS1, OIS1N
and CC1E bits.
1: On - OC1N signal is output on the corresponding output pin depending on MOE, OSSI,
OSSR, OIS1, OIS1N and CC1E bits.
Note: On channels having a complementary output, this bit is preloaded. If the CCPC bit is
set in the TIMx_CR2 register then the CC1NE active bit takes the new value from the
preloaded bit only when a Commutation event is generated.

RM0090 Rev 17

577/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Bit 1 CC1P: Capture/Compare 1 output polarity
CC1 channel configured as output:
0: OC1 active high
1: OC1 active low
CC1 channel configured as input:
CC1NP/CC1P bits select the active polarity of TI1FP1 and TI2FP1 for trigger or capture
operations.
00: non-inverted/rising edge
The circuit is sensitive to TIxFP1 rising edge (capture or trigger operations in reset, external
clock or trigger mode), TIxFP1 is not inverted (trigger operation in gated mode or encoder
mode).
01: inverted/falling edge
The circuit is sensitive to TIxFP1 falling edge (capture or trigger operations in reset, external
clock or trigger mode), TIxFP1 is inverted (trigger operation in gated mode or encoder
mode).
10: reserved, do not use this configuration.
11: non-inverted/both edges
The circuit is sensitive to both TIxFP1 rising and falling edges (capture or trigger operations
in reset, external clock or trigger mode), TIxFP1 is not inverted (trigger operation in gated
mode). This configuration must not be used in encoder mode.
Note: On channels having a complementary output, this bit is preloaded. If the CCPC bit is
set in the TIMx_CR2 register then the CC1P active bit takes the new value from the
preloaded bit only when a Commutation event is generated.
Note: This bit is not writable as soon as LOCK level 2 or 3 has been programmed (LOCK bits
in TIMx_BDTR register).
Bit 0 CC1E: Capture/Compare 1 output enable
CC1 channel configured as output:
0: Off - OC1 is not active. OC1 level is then function of MOE, OSSI, OSSR, OIS1, OIS1N
and CC1NE bits.
1: On - OC1 signal is output on the corresponding output pin depending on MOE, OSSI,
OSSR, OIS1, OIS1N and CC1NE bits.
CC1 channel configured as input:
This bit determines if a capture of the counter value can actually be done into the input
capture/compare register 1 (TIMx_CCR1) or not.
0: Capture disabled.
1: Capture enabled.
Note: On channels having a complementary output, this bit is preloaded. If the CCPC bit is
set in the TIMx_CR2 register then the CC1E active bit takes the new value from the
preloaded bit only when a Commutation event is generated.

578/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)
Table 95. Output control bits for complementary OCx and OCxN channels with
break feature
Output states(1)

Control bits
MOE
bit

1

OSSI
bit

X

OSSR
bit

CCxE
bit

CCxNE
bit

0

0

0

Output Disabled (not driven by Output Disabled (not driven by the
the timer), OCx=0, OCx_EN=0 timer), OCxN=0, OCxN_EN=0

0

0

1

Output Disabled (not driven by OCxREF + Polarity OCxN=OCxREF
the timer), OCx=0, OCx_EN=0 xor CCxNP, OCxN_EN=1

0

1

0

OCxREF + Polarity
OCx=OCxREF xor CCxP,
OCx_EN=1

0

1

1

Complementary to OCREF (not
OCREF + Polarity + dead-time
OCREF) + Polarity + dead-time
OCx_EN=1
OCxN_EN=1

1

0

0

Output Disabled (not driven by
the timer)
OCx=CCxP, OCx_EN=0

Output Disabled (not driven by the
timer)
OCxN=CCxNP, OCxN_EN=0

1

0

1

Off-State (output enabled with
inactive state)
OCx=CCxP, OCx_EN=1

OCxREF + Polarity
OCxN=OCxREF xor CCxNP,
OCxN_EN=1

1

1

0

OCxREF + Polarity
OCx=OCxREF xor CCxP,
OCx_EN=1

Off-State (output enabled with
inactive state)
OCxN=CCxNP, OCxN_EN=1

1

1

1

Complementary to OCREF (not
OCREF + Polarity + dead-time
OCREF) + Polarity + dead-time
OCx_EN=1
OCxN_EN=1

X

X

0

0

0

1

1

0

1

1

0

0

1

X

OCx output state

OCxN output state

Output Disabled (not driven by the
timer)
OCxN=0, OCxN_EN=0

Output disabled (not driven by the timer anymore). The output state is
defined by the GPIO controller and can be High, Low or Hi-Z.
Off-State (output enabled with inactive state)
Asynchronously: OCx=CCxP, OCx_EN=1, OCxN=CCxNP,
OCxN_EN=1
Then if the clock is present: OCx=OISx and OCxN=OISxN after a
dead-time, assuming that OISx and OISxN do not correspond to OCX
and OCxN both in active state

1. When both outputs of a channel are not used (CCxE = CCxNE = 0), the OISx, OISxN, CCxP and CCxNP bits must be kept
cleared.

Note:

The state of the external I/O pins connected to the complementary OCx and OCxN channels
depends on the OCx and OCxN channel state and the GPIO registers.

RM0090 Rev 17

579/1747
588

Advanced-control timers (TIM1 and TIM8)

17.4.10

RM0090

TIM1 and TIM8 counter (TIMx_CNT)
Address offset: 0x24
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

CNT[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 15:0 CNT[15:0]: Counter value

17.4.11

TIM1 and TIM8 prescaler (TIMx_PSC)
Address offset: 0x28
Reset value: 0x0000

15

14

13

12

11

10

9

8

rw

rw

rw

rw

rw

rw

rw

rw

7

PSC[15:0]
rw

Bits 15:0 PSC[15:0]: Prescaler value
The counter clock frequency (CK_CNT) is equal to fCK_PSC / (PSC[15:0] + 1).
PSC contains the value to be loaded in the active prescaler register at each update event
(including when the counter is cleared through UG bit of TIMx_EGR register or through
trigger controller when configured in “reset mode”).

17.4.12

TIM1 and TIM8 auto-reload register (TIMx_ARR)
Address offset: 0x2C
Reset value: 0xFFFF

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

ARR[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 15:0 ARR[15:0]: Auto-reload value
ARR is the value to be loaded in the actual auto-reload register.
Refer to Section 17.3.1: Time-base unit for more details about ARR update and behavior.
The counter is blocked while the auto-reload value is null.

580/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)

17.4.13

TIM1 and TIM8 repetition counter register (TIMx_RCR)
Address offset: 0x30
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

REP[7:0]

Reserved

rw

rw

rw

rw

rw

Bits 15:8 Reserved, must be kept at reset value.
Bits 7:0 REP[7:0]: Repetition counter value
These bits allow the user to set-up the update rate of the compare registers (i.e. periodic
transfers from preload to active registers) when preload registers are enable, as well as the
update interrupt generation rate, if this interrupt is enable.
Each time the REP_CNT related downcounter reaches zero, an update event is generated
and it restarts counting from REP value. As REP_CNT is reloaded with REP value only at
the repetition update event U_RC, any write to the TIMx_RCR register is not taken in
account until the next repetition update event.
It means in PWM mode (REP+1) corresponds to:
–
the number of PWM periods in edge-aligned mode
–
the number of half PWM period in center-aligned mode.

17.4.14

TIM1 and TIM8 capture/compare register 1 (TIMx_CCR1)
Address offset: 0x34
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

CCR1[15:0]
rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

Bits 15:0 CCR1[15:0]: Capture/Compare 1 value
If channel CC1 is configured as output:
CCR1 is the value to be loaded in the actual capture/compare 1 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR1 register
(bit OC1PE). Else the preload value is copied in the active capture/compare 1 register when
an update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signaled on OC1 output.
If channel CC1 is configured as input:
CCR1 is the counter value transferred by the last input capture 1 event (IC1). The
TIMx_CCR1 register is read-only and cannot be programmed.

RM0090 Rev 17

581/1747
588

Advanced-control timers (TIM1 and TIM8)

17.4.15

RM0090

TIM1 and TIM8 capture/compare register 2 (TIMx_CCR2)
Address offset: 0x38
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

CCR2[15:0]
rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

Bits 15:0 CCR2[15:0]: Capture/Compare 2 value
If channel CC2 is configured as output:
CCR2 is the value to be loaded in the actual capture/compare 2 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR2 register
(bit OC2PE). Else the preload value is copied in the active capture/compare 2 register when
an update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signalled on OC2 output.
If channel CC2 is configured as input:
CCR2 is the counter value transferred by the last input capture 2 event (IC2). The
TIMx_CCR2 register is read-only and cannot be programmed.

17.4.16

TIM1 and TIM8 capture/compare register 3 (TIMx_CCR3)
Address offset: 0x3C
Reset value: 0x0000

15

14

13

12

11

10

9

8

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

7

6

5

4

3

2

1

0

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

CCR3[15:0]
rw/ro

Bits 15:0 CCR3[15:0]: Capture/Compare value
If channel CC3 is configured as output:
CCR3 is the value to be loaded in the actual capture/compare 3 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR3 register
(bit OC3PE). Else the preload value is copied in the active capture/compare 3 register when
an update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signalled on OC3 output.
If channel CC3 is configured as input:
CCR3 is the counter value transferred by the last input capture 3 event (IC3). The
TIMx_CCR3 register is read-only and cannot be programmed.

582/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)

17.4.17

TIM1 and TIM8 capture/compare register 4 (TIMx_CCR4)
Address offset: 0x40
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

CCR4[15:0]
rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

Bits 15:0 CCR4[15:0]: Capture/Compare value
If channel CC4 is configured as output:
CCR4 is the value to be loaded in the actual capture/compare 4 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR4 register
(bit OC4PE). Else the preload value is copied in the active capture/compare 4 register when
an update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signalled on OC4 output.
If channel CC4 is configured as input:
CCR4 is the counter value transferred by the last input capture 4 event (IC4). The
TIMx_CCR3 register is read-only and cannot be programmed.

17.4.18

TIM1 and TIM8 break and dead-time register (TIMx_BDTR)
Address offset: 0x44
Reset value: 0x0000

15

14

13

12

11

10

MOE

AOE

BKP

BKE

OSSR

OSSI

rw

rw

rw

rw

rw

rw

Note:

9

8

7

6

5

4

rw

rw

rw

rw

LOCK[1:0]
rw

rw

3

2

1

0

rw

rw

rw

DTG[7:0]
rw

As the bits AOE, BKP, BKE, OSSI, OSSR and DTG[7:0] can be write-locked depending on
the LOCK configuration, it can be necessary to configure all of them during the first write
access to the TIMx_BDTR register.
Bit 15 MOE: Main output enable
This bit is cleared asynchronously by hardware as soon as the break input is active. It is set
by software or automatically depending on the AOE bit. It is acting only on the channels
which are configured in output.
0: OC and OCN outputs are disabled or forced to idle state.
1: OC and OCN outputs are enabled if their respective enable bits are set (CCxE, CCxNE in
TIMx_CCER register).
See OC/OCN enable description for more details (Section 17.4.9: TIM1 and TIM8
capture/compare enable register (TIMx_CCER)).
Bit 14 AOE: Automatic output enable
0: MOE can be set only by software
1: MOE can be set by software or automatically at the next update event (if the break input is
not be active)
Note: This bit can not be modified as long as LOCK level 1 has been programmed (LOCK bits
in TIMx_BDTR register).

RM0090 Rev 17

583/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

Bit 13 BKP: Break polarity
0: Break input BRK is active low
1: Break input BRK is active high
Note: This bit can not be modified as long as LOCK level 1 has been programmed (LOCK bits
in TIMx_BDTR register).
Note: Any write operation to this bit takes a delay of 1 APB clock cycle to become effective.
Bit 12 BKE: Break enable
0: Break inputs (BRK and CSS clock failure event) disabled
1; Break inputs (BRK and CSS clock failure event) enabled
Note: This bit cannot be modified when LOCK level 1 has been programmed (LOCK bits in
TIMx_BDTR register).
Note: Any write operation to this bit takes a delay of 1 APB clock cycle to become effective.
Bit 11 OSSR: Off-state selection for Run mode
This bit is used when MOE=1 on channels having a complementary output which are
configured as outputs. OSSR is not implemented if no complementary output is implemented
in the timer.
See OC/OCN enable description for more details (Section 17.4.9: TIM1 and TIM8
capture/compare enable register (TIMx_CCER)).
0: When inactive, OC/OCN outputs are disabled (OC/OCN enable output signal=0).
1: When inactive, OC/OCN outputs are enabled with their inactive level as soon as CCxE=1
or CCxNE=1. Then, OC/OCN enable output signal=1
Note: This bit can not be modified as soon as the LOCK level 2 has been programmed (LOCK
bits in TIMx_BDTR register).
Bit 10 OSSI: Off-state selection for Idle mode
This bit is used when MOE=0 on channels configured as outputs.
See OC/OCN enable description for more details (Section 17.4.9: TIM1 and TIM8
capture/compare enable register (TIMx_CCER)).
0: When inactive, OC/OCN outputs are disabled (OC/OCN enable output signal=0).
1: When inactive, OC/OCN outputs are forced first with their idle level as soon as CCxE=1 or
CCxNE=1. OC/OCN enable output signal=1)
Note: This bit can not be modified as soon as the LOCK level 2 has been programmed (LOCK
bits in TIMx_BDTR register).
Bits 9:8 LOCK[1:0]: Lock configuration
These bits offer a write protection against software errors.
00: LOCK OFF - No bit is write protected.
01: LOCK Level 1 = DTG bits in TIMx_BDTR register, OISx and OISxN bits in TIMx_CR2
register and BKE/BKP/AOE bits in TIMx_BDTR register can no longer be written.
10: LOCK Level 2 = LOCK Level 1 + CC Polarity bits (CCxP/CCxNP bits in TIMx_CCER
register, as long as the related channel is configured in output through the CCxS bits) as well
as OSSR and OSSI bits can no longer be written.
11: LOCK Level 3 = LOCK Level 2 + CC Control bits (OCxM and OCxPE bits in
TIMx_CCMRx registers, as long as the related channel is configured in output through the
CCxS bits) can no longer be written.
Note: The LOCK bits can be written only once after the reset. Once the TIMx_BDTR register
has been written, their content is frozen until the next reset.

584/1747

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)

Bits 7:0 DTG[7:0]: Dead-time generator setup
This bit-field defines the duration of the dead-time inserted between the complementary
outputs. DT correspond to this duration.
DTG[7:5]=0xx => DT=DTG[7:0]x tdtg with tdtg=tDTS.
DTG[7:5]=10x => DT=(64+DTG[5:0])xtdtg with Tdtg=2xtDTS.
DTG[7:5]=110 => DT=(32+DTG[4:0])xtdtg with Tdtg=8xtDTS.
DTG[7:5]=111 => DT=(32+DTG[4:0])xtdtg with Tdtg=16xtDTS.
Example if TDTS=125ns (8MHz), dead-time possible values are:
0 to 15875 ns by 125 ns steps,
16 us to 31750 ns by 250 ns steps,
32 us to 63us by 1 us steps,
64 us to 126 us by 2 us steps
Note: This bit-field can not be modified as long as LOCK level 1, 2 or 3 has been programmed
(LOCK bits in TIMx_BDTR register).

17.4.19

TIM1 and TIM8 DMA control register (TIMx_DCR)
Address offset: 0x48
Reset value: 0x0000

15

14

13

Reserved

12

11

10

9

8

DBL[4:0]
rw

rw

rw

rw

rw

7

6
Reserved

5

4

3

2

1

0

rw

rw

DBA[4:0]
rw

rw

rw

Bits 15:13 Reserved, must be kept at reset value.
Bits 12:8 DBL[4:0]: DMA burst length
This 5-bit vector defines the number of DMA transfers (the timer detects a burst transfer
when a read or a write access to the TIMx_DMAR register address is performed).
the TIMx_DMAR address)
00000: 1 transfer
00001: 2 transfers
00010: 3 transfers
...
10001: 18 transfers
Bits 7:5 Reserved, must be kept at reset value.
Bits 4:0 DBA[4:0]: DMA base address
This 5-bits vector defines the base-address for DMA transfers (when read/write access are
done through the TIMx_DMAR address). DBA is defined as an offset starting from the
address of the TIMx_CR1 register.
Example:
00000: TIMx_CR1,
00001: TIMx_CR2,
00010: TIMx_SMCR,
...
Example: Let us consider the following transfer: DBL = 7 transfers and DBA = TIMx_CR1. In
this case the transfer is done to/from 7 registers starting from the TIMx_CR1 address.

RM0090 Rev 17

585/1747
588

Advanced-control timers (TIM1 and TIM8)

17.4.20

RM0090

TIM1 and TIM8 DMA address for full transfer (TIMx_DMAR)
Address offset: 0x4C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

DMAB[31:16]
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

DMAB[15:0]
rw

rw

Bits 31:0 DMAB[31:0]: DMA register for burst accesses
A read or write operation to the DMAR register accesses the register located at the address
(TIMx_CR1 address) + (DBA + DMA index) x 4
where TIMx_CR1 address is the address of the control register 1, DBA is the DMA base
address configured in TIMx_DCR register, DMA index is automatically controlled by the
DMA transfer, and ranges from 0 to DBL (DBL configured in TIMx_DCR).

Example of how to use the DMA burst feature
In this example the timer DMA burst feature is used to update the contents of the CCRx
registers (x = 2, 3, 4) with the DMA transferring half words into the CCRx registers.
This is done in the following steps:
1.

Note:

586/1747

Configure the corresponding DMA channel as follows:
–

DMA channel peripheral address is the DMAR register address

–

DMA channel memory address is the address of the buffer in the RAM containing
the data to be transferred by DMA into CCRx registers.

–

Number of data to transfer = 3 (See note below).

–

Circular mode disabled.

2.

Configure the DCR register by configuring the DBA and DBL bit fields as follows:
DBL = 3 transfers, DBA = 0xE.

3.

Enable the TIMx update DMA request (set the UDE bit in the DIER register).

4.

Enable TIMx

5.

Enable the DMA channel

This example is for the case where every CCRx register to be updated once. If every CCRx
register is to be updated twice for example, the number of data to transfer should be 6. Let's
take the example of a buffer in the RAM containing data1, data2, data3, data4, data5 and
data6. The data is transferred to the CCRx registers as follows: on the first update DMA
request, data1 is transferred to CCR2, data2 is transferred to CCR3, data3 is transferred to
CCR4 and on the second update DMA request, data4 is transferred to CCR2, data5 is
transferred to CCR3 and data6 is transferred to CCR4.

RM0090 Rev 17

RM0090

Advanced-control timers (TIM1 and TIM8)

17.4.21

TIM1 and TIM8 register map
TIM1 and TIM8 registers are mapped as 16-bit addressable registers as described in the
table below:

Reserved

0

3

2

1

0

URS

UDIS

CEN

CCUS

Reserved

CCPC

CC2IF

CC1IF

UIF

CC2G

CC1G

UG

0

0

0

0

0

0

CC4S
[1:0]
0

0

CC4S
[1:0]

0

0

0

IC1F[3:0]

0

IC4
PSC
[1:0]

OC1M
[2:0]

0

0

0

OC3M
[2:0]

0

0

0

0

IC3F[3:0]

OC1FE

0

OC1PE

0

OC1CE

OC2FE
0

0

0

0

CC1
S
[1:0]
0

0

IC1
PSC
[1:0]

CC1
S
[1:0]

0

0

0

0

0

0

CC3
S
[1:0]
0

0

IC3
PSC
[1:0]

CC3
S
[1:0]

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CC1E

0

CC1P

0

CC1NE

0

CC1NP

Reserved

0

0

CC2E

0

0

0

CC3E

0

0

0

CC3P

IC4F[3:0]

Reserved

0

0

OC3FE

0

CC2S
[1:0]

0

CC2P

0

0

0

OC3PE

OC4M
[2:0]

0

0

OC3CE

0

0

IC2
PSC
[1:0]

0

CC2NE

TIMx_PSC

0

0

OC4FE

0

OC2PE

OC2CE
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CNT[15:0]

Reserved

Reset value
0x28

0

0

CC2NP

TIMx_CNT

0

OC4PE

0

CC2S
[1:0]

CC3IF

0

CC3G

0

CC4IF

0

CC4G

0

COMIF

0

COMG

0

TIF

0

TG

0
BIF

0

BG

UIE

0

CC1IE

0

CC2IE

CC1OF

0

OC2M
[2:0]

Reset value
0x24

4

Reserved

CC2OF

0

CC3IE

0

CC4IE

0

COMIE

0

0

TIE

CC1DE

0

0

BIE

CC2DE

0

0

CC3NE

TIMx_CCER

5

0

SMS[2:0]

CC3NP

0x20

OPM

TI1S

0

MSM

0

0

Reset value

CCDS

OIS1

0

IC2F[3:0]

Reset value
TIMx_CCMR2
Input Capture
mode

DIR

OIS1N

0

TS[2:0]

0

CC4E

0x1C

6

OIS2

0

Reserved

Reserved

0

0

O24CE

TIMx_CCMR2
Output
Compare mode

0

0

0

0

Reset value

0

0

0

Reserved

Reset value
TIMx_CCMR1
Input Capture
mode

MMS[2:0]

CC4P

0x18

0

0

Reserved

Reserved

0

0

ETF[3:0]

Reset value
TIMx_CCMR1
Output
Compare mode

0

0

Reserved

TIMx_EGR

7

OIS2N

0

Reset value

0x14

8

OIS3

0

CC3DE

TIMx_SR

0

0

0

0

Reset value

0x10

0

CC3OF

TIMx_DIER

0

CC4DE

0x0C

0

ETP
S
[1:0]

CC4OF

Reset value

0

COMDE

Reserved

0

0

0

ECE

TIMx_SMCR

0

TDE

0x08

0

ETP

Reset value

0
OIS3N

Reserved

CMS
[1:0]

OIS4

TIMx_CR2

ARPE

CKD
[1:0]

Reserved

Reset value

0x04

9

11

10

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

TIMx_CR1

UDE

0x00

Register

30

Offset

31

Table 96. TIM1 and TIM8 register map and reset values

0

0

0

0

0

0

0

0

0

PSC[15:0]

Reserved

Reset value

0

RM0090 Rev 17

0

0

0

0

0

0

0

0

587/1747
588

Advanced-control timers (TIM1 and TIM8)

RM0090

TIMx_RCR

TIMx_CCR1

0

0

1

1

1

0

0

0

0

0

0

0

6

1

1

1

1

1

0

0

0

0

0

0

OSSI

0

OSSR

0

BKE

0

BKP

0

AOE

Reserved

0

MOE

0

TIMx_DCR

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CCR4[15:0]

Reserved

TIMx_BDTR

0

CCR3[15:0]

Reserved

TIMx_CCR4

0

0

0

0

0

0

0

0

TIMx_DMAR

0

0

LOCK
[1:0]
0

0

DBL[4:0]

Reserved

Reset value

DT[7:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DBA[4:0]

Reserved
0

0

0

0

0

0

0

0

0

0

DMAB[31:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Refer to Section 2.3: Memory map for the register boundary addresses.

588/1747

1

CCR2[15:0]
0

TIMx_CCR3

Reset value

0

Reserved

Reset value

0x4C

0

1

CCR1[15:0]
0

TIMx_CCR2

0x48

1

1

Reserved

Reset value
0x44

2

1

0

Reset value
0x40

3

1

REP[7:0]

Reset value
0x3C

4

1

Reserved

Reset value
0x38

7

8

9

11

10

12

13

14

15

16

17

18

19

20

21

22

23

1

Reset value
0x34

1

ARR[15:0]

Reserved

Reset value
0x30

5

TIMx_ARR

0x2C

24

25

26

27

28

29

Register

30

Offset

31

Table 96. TIM1 and TIM8 register map and reset values (continued)

RM0090 Rev 17

0

0

RM0090

18

General-purpose timers (TIM2 to TIM5)

General-purpose timers (TIM2 to TIM5)
This section applies to the whole STM32F4xx family, unless otherwise specified.

18.1

TIM2 to TIM5 introduction
The general-purpose timers consist of a 16-bit or 32-bit auto-reload counter driven by a
programmable prescaler.
They may be used for a variety of purposes, including measuring the pulse lengths of input
signals (input capture) or generating output waveforms (output compare and PWM).
Pulse lengths and waveform periods can be modulated from a few microseconds to several
milliseconds using the timer prescaler and the RCC clock controller prescalers.
The timers are completely independent, and do not share any resources. They can be
synchronized together as described in Section 18.3.15.

18.2

TIM2 to TIM5 main features
General-purpose TIMx timer features include:
•

16-bit (TIM3 and TIM4) or 32-bit (TIM2 and TIM5) up, down, up/down auto-reload
counter.

•

16-bit programmable prescaler used to divide (also “on the fly”) the counter clock
frequency by any factor between 1 and 65536.

•

Up to 4 independent channels for:
–

Input capture

–

Output compare

–

PWM generation (Edge- and Center-aligned modes)

–

One-pulse mode output

•

Synchronization circuit to control the timer with external signals and to interconnect
several timers.

•

Interrupt/DMA generation on the following events:
–

Update: counter overflow/underflow, counter initialization (by software or
internal/external trigger)

–

Trigger event (counter start, stop, initialization or count by internal/external trigger)

–

Input capture

–

Output compare

•

Supports incremental (quadrature) encoder and hall-sensor circuitry for positioning
purposes

•

Trigger input for external clock or cycle-by-cycle current management

RM0090 Rev 17

589/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Figure 134. General-purpose timer block diagram

,QWHUQDOFORFN &.B,17
7,0[&/.IURP5&&
(75)
(75

7,0[B(75

7ULJJHU
FRQWUROOHU
75*2

3RODULW\VHOHFWLRQ HGJH (753
,QSXWILOWHU
GHWHFWRU SUHVFDOHU
,75
,75
,75
,75

WRRWKHUWLPHUV
WR'$&$'&
7*,

,75
75&

75*,

6ODYH
FRQWUROOHU 5HVHWHQDEOHXSFRXQW
PRGH

7,)B('

(QFRGHU
LQWHUIDFH

7,)3
7,)3

8

$XWRUHORDGUHJLVWHU

6WRSFOHDURUXSGRZQ
&.B36&
;25
7,
7,0[B&+
7,

7,0[B&+

,QSXWILOWHU 
HGJHGHWHFWRU

,QSXWILOWHU 
HGJHGHWHFWRU

7,)3
7,)3

,&

36& &.B&17
SUHVFDOHU
&&, 8
3UHVFDOHU

75&
7,)3
7,)3
75&

,&36
&&,

,&

3UHVFDOHU

7,

7,0[B&+

7,

7,0[B&+

,QSXWILOWHU 
HGJHGHWHFWRU

,QSXWILOWHU 
HGJHGHWHFWRU

,&

3UHVFDOHU

75&
7,)3
7,)3

&17FRXQWHU
&&,

&DSWXUH&RPSDUHUHJLVWHU
 

&DSWXUH&RPSDUHUHJLVWHU 2&5()

&DSWXUH&RPSDUHUHJLVWHU
 

3UHVFDOHU

7,0[B&+

2XWSXW 2&
FRQWURO

7,0[B&+

2&5()

2XWSXW 2&
FRQWURO

7,0[B&+

2XWSXW 2&
FRQWURO

7,0[B&+

&&,

8
,&36

2XWSXW 2&
FRQWURO

&&,

8

,&36
&&,

,&

2&5()

&&,

8

,&36
&&,

7,)3
7,)3



8,
8

&DSWXUH&RPSDUHUHJLVWHU 2&5()

75&
(75)

1RWHV
5HJ

3UHORDGUHJLVWHUVWUDQVIHUUHG
WRDFWLYHUHJLVWHUVRQ8HYHQW
DFFRUGLQJWRFRQWUROELW
(YHQW
,QWHUUXSW '0$RXWSXW

069

590/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)

18.3

TIM2 to TIM5 functional description

18.3.1

Time-base unit
The main block of the programmable timer is a 16-bit/32-bit counter with its related autoreload register. The counter can count up. The counter clock can be divided by a prescaler.
The counter, the auto-reload register and the prescaler register can be written or read by
software. This is true even when the counter is running.
The time-base unit includes:
•

Counter Register (TIMx_CNT)

•

Prescaler Register (TIMx_PSC):

•

Auto-Reload Register (TIMx_ARR)

The auto-reload register is preloaded. Writing to or reading from the auto-reload register
accesses the preload register. The content of the preload register are transferred into the
shadow register permanently or at each update event (UEV), depending on the auto-reload
preload enable bit (ARPE) in TIMx_CR1 register. The update event is sent when the counter
reaches the overflow (or underflow when downcounting) and if the UDIS bit equals 0 in the
TIMx_CR1 register. It can also be generated by software. The generation of the update
event is described in detail for each configuration.
The counter is clocked by the prescaler output CK_CNT, which is enabled only when the
counter enable bit (CEN) in TIMx_CR1 register is set (refer also to the slave mode controller
description to get more details on counter enabling).
Note that the actual counter enable signal CNT_EN is set 1 clock cycle after CEN.

Prescaler description
The prescaler can divide the counter clock frequency by any factor between 1 and 65536. It
is based on a 16-bit counter controlled through a 16-bit/32-bit register (in the TIMx_PSC
register). It can be changed on the fly as this control register is buffered. The new prescaler
ratio is taken into account at the next update event.
Figure 135 and Figure 136 give some examples of the counter behavior when the prescaler
ratio is changed on the fly:

RM0090 Rev 17

591/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Figure 135. Counter timing diagram with prescaler division change from 1 to 2
&.B36&
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU

)

) ) )$ )% )&









8SGDWHHYHQW 8(9
3UHVFDOHUFRQWUROUHJLVWHU





:ULWHDQHZYDOXHLQ7,0[B36&

3UHVFDOHUEXIIHU



3UHVFDOHUFRXQWHU




















069

Figure 136. Counter timing diagram with prescaler division change from 1 to 4
&.B36&
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU

)

) ) )$ )% )&





8SGDWHHYHQW 8(9
3UHVFDOHUFRQWUROUHJLVWHU





:ULWHDQHZYDOXHLQ7,0[B36&

3UHVFDOHUEXIIHU



3UHVFDOHUFRXQWHU




















069

18.3.2

Counter modes
Upcounting mode
In upcounting mode, the counter counts from 0 to the auto-reload value (content of the
TIMx_ARR register), then restarts from 0 and generates a counter overflow event.
An Update event can be generated at each counter overflow or by setting the UG bit in the
TIMx_EGR register (by software or by using the slave mode controller).
The UEV event can be disabled by software by setting the UDIS bit in TIMx_CR1 register.
This is to avoid updating the shadow registers while writing new values in the preload
registers. Then no update event occurs until the UDIS bit has been written to 0. However,
the counter restarts from 0, as well as the counter of the prescaler (but the prescale rate

592/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)
does not change). In addition, if the URS bit (update request selection) in TIMx_CR1
register is set, setting the UG bit generates an update event UEV but without setting the UIF
flag (thus no interrupt or DMA request is sent). This is to avoid generating both update and
capture interrupts when clearing the counter on the capture event.
When an update event occurs, all the registers are updated and the update flag (UIF bit in
TIMx_SR register) is set (depending on the URS bit):
•

The buffer of the prescaler is reloaded with the preload value (content of the TIMx_PSC
register)

•

The auto-reload shadow register is updated with the preload value (TIMx_ARR)

The following figures show some examples of the counter behavior for different clock
frequencies when TIMx_ARR=0x36.
Figure 137. Counter timing diagram, internal clock divided by 1
&.B,17

&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU





























&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

Figure 138. Counter timing diagram, internal clock divided by 2
&.B,17
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU















&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

RM0090 Rev 17

593/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Figure 139. Counter timing diagram, internal clock divided by 4
&.B,17
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU









&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
06Y9

Figure 140. Counter timing diagram, internal clock divided by N
&.B,17

7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU

)





&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
06Y9

594/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)
Figure 141. Counter timing diagram, Update event when ARPE=0 (TIMx_ARR not
preloaded)
&.B,17
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU



            

&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
$XWRUHORDGUHJLVWHU

))



:ULWHDQHZYDOXHLQ7,0[B$55
06Y9

Figure 142. Counter timing diagram, Update event when ARPE=1 (TIMx_ARR
preloaded)
&.B36&
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU

)

) ) ) ) )        

&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
$XWRUHORDGSUHORDGUHJLVWHU

$XWRUHORDGVKDGRZUHJLVWHU

)



)



:ULWHDQHZYDOXHLQ7,0[B$55
06Y9

Downcounting mode
In downcounting mode, the counter counts from the auto-reload value (content of the
TIMx_ARR register) down to 0, then restarts from the auto-reload value and generates a
counter underflow event.
An Update event can be generate at each counter underflow or by setting the UG bit in the
TIMx_EGR register (by software or by using the slave mode controller)
The UEV update event can be disabled by software by setting the UDIS bit in TIMx_CR1
register. This is to avoid updating the shadow registers while writing new values in the

RM0090 Rev 17

595/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

preload registers. Then no update event occurs until UDIS bit has been written to 0.
However, the counter restarts from the current auto-reload value, whereas the counter of the
prescaler restarts from 0 (but the prescale rate doesn’t change).
In addition, if the URS bit (update request selection) in TIMx_CR1 register is set, setting the
UG bit generates an update event UEV but without setting the UIF flag (thus no interrupt or
DMA request is sent). This is to avoid generating both update and capture interrupts when
clearing the counter on the capture event.
When an update event occurs, all the registers are updated and the update flag (UIF bit in
TIMx_SR register) is set (depending on the URS bit):
•

The buffer of the prescaler is reloaded with the preload value (content of the TIMx_PSC
register).

•

The auto-reload active register is updated with the preload value (content of the
TIMx_ARR register). Note that the auto-reload is updated before the counter is
reloaded, so that the next period is the expected one.

The following figures show some examples of the counter behavior for different clock
frequencies when TIMx_ARR=0x36.
Figure 143. Counter timing diagram, internal clock divided by 1
&.B,17

&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU



    

       )

&RXQWHUXQGHUIORZ FQWBXGI 
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
06Y9

596/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)
Figure 144. Counter timing diagram, internal clock divided by 2
&.B,17
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU















&RXQWHUXQGHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
06Y9

Figure 145. Counter timing diagram, internal clock divided by 4
&.B,17
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU









&RXQWHUXQGHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)

069

Figure 146. Counter timing diagram, internal clock divided by N
&.B,17

7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU



)





&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

RM0090 Rev 17

597/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Figure 147. Counter timing diagram, Update event
&.B,17
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU



 

  

   

   )

&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
$XWRUHORDGSUHORDGUHJLVWHU

))

:ULWHDQHZYDOXHLQ7,0[B$55



069

Center-aligned mode (up/down counting)
In center-aligned mode, the counter counts from 0 to the auto-reload value (content of the
TIMx_ARR register) – 1, generates a counter overflow event, then counts from the autoreload value down to 1 and generates a counter underflow event. Then it restarts counting
from 0.
Center-aligned mode is active when the CMS bits in TIMx_CR1 register are not equal to
'00'. The Output compare interrupt flag of channels configured in output is set when: the
counter counts down (Center aligned mode 1, CMS = "01"), the counter counts up (Center
aligned mode 2, CMS = "10") the counter counts up and down (Center aligned mode 3,
CMS = "11").
In this mode, the direction bit (DIR from TIMx_CR1 register) cannot be written. It is updated
by hardware and gives the current direction of the counter.
The update event can be generated at each counter overflow and at each counter underflow
or by setting the UG bit in the TIMx_EGR register (by software or by using the slave mode
controller) also generates an update event. In this case, the counter restarts counting from
0, as well as the counter of the prescaler.
The UEV update event can be disabled by software by setting the UDIS bit in TIMx_CR1
register. This is to avoid updating the shadow registers while writing new values in the
preload registers. Then no update event occurs until the UDIS bit has been written to 0.
However, the counter continues counting up and down, based on the current auto-reload
value.
In addition, if the URS bit (update request selection) in TIMx_CR1 register is set, setting the
UG bit generates an update event UEV but without setting the UIF flag (thus no interrupt or
DMA request is sent). This is to avoid generating both update and capture interrupt when
clearing the counter on the capture event.

598/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)
When an update event occurs, all the registers are updated and the update flag (UIF bit in
TIMx_SR register) is set (depending on the URS bit):
•

The buffer of the prescaler is reloaded with the preload value (content of the TIMx_PSC
register).

•

The auto-reload active register is updated with the preload value (content of the
TIMx_ARR register). Note that if the update source is a counter overflow, the autoreload is updated before the counter is reloaded, so that the next period is the expected
one (the counter is loaded with the new value).

The following figures show some examples of the counter behavior for different clock
frequencies.
Figure 148. Counter timing diagram, internal clock divided by 1, TIMx_ARR=0x6
&.B,17
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU



            

&RXQWHUXQGHUIORZ
&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

1. Here, center-aligned mode 1 is used, for more details refer to Section 18.4.1: TIMx control register 1 (TIMx_CR1).

Figure 149. Counter timing diagram, internal clock divided by 2
&.B,17

&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU















&RXQWHUXQGHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

RM0090 Rev 17

599/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Figure 150. Counter timing diagram, internal clock divided by 4, TIMx_ARR=0x36
&.B,17
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU









&RXQWHURYHUIORZ FQWBRYI
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)

069

1. Center-aligned mode 2 or 3 is used with an UIF on overflow.

Figure 151. Counter timing diagram, internal clock divided by N
&.B,17

7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU



)





&RXQWHUXQGHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

600/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)
Figure 152. Counter timing diagram, Update event with ARPE=1 (counter underflow)
&.B,17
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU



            

&RXQWHUXQGHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
$XWRUHORDGSUHORDGUHJLVWHU

)'



:ULWHDQHZYDOXHLQ7,0[B$55
$XWRUHORDGDFWLYHUHJLVWHU

)'


069

Figure 153. Counter timing diagram, Update event with ARPE=1 (counter overflow)
&.B,17
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU

)

) ) )$ )% )&        )

&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
$XWRUHORDGSUHORDGUHJLVWHU

)'



:ULWHDQHZYDOXHLQ7,0[B$55
$XWRUHORDGDFWLYHUHJLVWHU

)'


069

RM0090 Rev 17

601/1747
649

General-purpose timers (TIM2 to TIM5)

18.3.3

RM0090

Clock selection
The counter clock can be provided by the following clock sources:
•

Internal clock (CK_INT)

•

External clock mode1: external input pin (TIx)

•

External clock mode2: external trigger input (ETR) available on TIM2, TIM3 and TIM4
only.

•

Internal trigger inputs (ITRx): using one timer as prescaler for another timer, for
example, Timer3 can be configured to act as a prescaler for Timer 2. Refer to Using
one timer as prescaler for another timer for more details.

Internal clock source (CK_INT)
If the slave mode controller is disabled (SMS=000 in the TIMx_SMCR register), then the
CEN, DIR (in the TIMx_CR1 register) and UG bits (in the TIMx_EGR register) are actual
control bits and can be changed only by software (except UG which remains cleared
automatically). As soon as the CEN bit is written to 1, the prescaler is clocked by the internal
clock CK_INT.
Figure 154 shows the behavior of the control circuit and the upcounter in normal mode,
without prescaler.
Figure 154. Control circuit in normal mode, internal clock divided by 1
,QWHUQDOFORFN
&(1 &17B(1
8*

&17B,1,7

&RXQWHUFORFN &.B&17 &.B36&
&RXQWHUUHJLVWHU









 

 



  




069

External clock source mode 1
This mode is selected when SMS=111 in the TIMx_SMCR register. The counter can count at
each rising or falling edge on a selected input.

602/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)
Figure 155. TI2 external clock connection example
7,0[B60&5
76>@
RU 7,)
7,)
,75[
7,B('

7,

)LOWHU

(GJH
GHWHFWRU

7,)B5LVLQJ
7,)B)DOOLQJ

7,)3



7,)3
(75)

RU
RU

[[


75*,





(75)
&.B,17

,&)>@

&&3

7,0[B&&05

7,0[B&&(5

LQWHUQDOFORFN

(QFRGHU
PRGH
([WHUQDOFORFN
PRGH

([WHUQDOFORFN
PRGH


&.B36&

,QWHUQDOFORFN
PRGH

(&(

606>@

7,0[B60&5

069

For example, to configure the upcounter to count in response to a rising edge on the TI2
input, use the following procedure:

Note:

1.

Configure channel 2 to detect rising edges on the TI2 input by writing CC2S= ‘01 in the
TIMx_CCMR1 register.

2.

Configure the input filter duration by writing the IC2F[3:0] bits in the TIMx_CCMR1
register (if no filter is needed, keep IC2F=0000).

The capture prescaler is not used for triggering, so there’s no need to configure it.
3.

Select rising edge polarity by writing CC2P=0 and CC2NP=0 in the TIMx_CCER
register.

4.

Configure the timer in external clock mode 1 by writing SMS=111 in the TIMx_SMCR
register.

5.

Select TI2 as the input source by writing TS=110 in the TIMx_SMCR register.

6.

Enable the counter by writing CEN=1 in the TIMx_CR1 register.

When a rising edge occurs on TI2, the counter counts once and the TIF flag is set.
The delay between the rising edge on TI2 and the actual clock of the counter is due to the
resynchronization circuit on TI2 input.

RM0090 Rev 17

603/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Figure 156. Control circuit in external clock mode 1

7,
&17B(1
&RXQWHUFORFN &.B&17 &.B36&
&RXQWHUUHJLVWHU







7,)

:ULWH7,) 

069

External clock source mode 2
This mode is selected by writing ECE=1 in the TIMx_SMCR register.
The counter can count at each rising or falling edge on the external trigger input ETR.
Figure 157 gives an overview of the external trigger input block.
Figure 157. External trigger input block
RU 7,)
7,)

(75
(75SLQ




'LYLGHU


(73

(736>@

(753
&.B,17

7,0[B60&5 7,0[B60&5

)LOWHU
GRZQFRXQWHU
(7)>@

RU
RU

(QFRGHU
PRGH

75*,

([WHUQDOFORFN
&.B36&
PRGH

(75)

([WHUQDOFORFN
PRGH

&.B,17
,QWHUQDOFORFN
LQWHUQDOFORFN PRGH

7,0[B60&5
(&( 606>@
7,0[B60&5
069

For example, to configure the upcounter to count each 2 rising edges on ETR, use the
following procedure:
1.

As no filter is needed in this example, write ETF[3:0]=0000 in the TIMx_SMCR register.

2.

Set the prescaler by writing ETPS[1:0]=01 in the TIMx_SMCR register

3.

Select rising edge detection on the ETR pin by writing ETP=0 in the TIMx_SMCR
register

4.

Enable external clock mode 2 by writing ECE=1 in the TIMx_SMCR register.

5.

Enable the counter by writing CEN=1 in the TIMx_CR1 register.

The counter counts once each 2 ETR rising edges.

604/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)
The delay between the rising edge on ETR and the actual clock of the counter is due to the
resynchronization circuit on the ETRP signal.
Figure 158. Control circuit in external clock mode 2

&.B,17
&17B(1
(75
(753
(75)
&RXQWHUFORFN &.B,17 &.B36&


&RXQWHUUHJLVWHU




069

18.3.4

Capture/compare channels
Each Capture/Compare channel (see Figure 159) is built around a capture/compare register
(including a shadow register), an input stage for capture (with digital filter, multiplexing and
prescaler) and an output stage (with comparator and output control).
The input stage samples the corresponding TIx input to generate a filtered signal TIxF.
Then, an edge detector with polarity selection generates a signal (TIxFPx) which can be
used as trigger input by the slave mode controller or as the capture command. It is
prescaled before the capture register (ICxPS).
Figure 159. Capture/compare channel (example: channel 1 input stage)
TI1F_ED
to the slave mode controller

TI1
fDTS

filter
downcounter

TI1F

TI1F_Rising
Edge
Detector

TI1FP1
TI1F_Falling
TI2FP1

ICF[3:0]

CC1P/CC1NP

TIMx_CCMR1

TIMx_CCER
TI2F_rising
(from channel 2)
TI2F_falling
(from channel 2)

01
10

IC1

divider
/1, /2, /4, /8

IC1PS

TRC
11
(from slave mode
controller)
CC1S[1:0] ICPS[1:0]
TIMx_CCMR1

CC1E
TIMx_CCER

The output stage generates an intermediate waveform which is then used for reference:
OCxRef (active high). The polarity acts at the end of the chain.

RM0090 Rev 17

605/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Figure 160. Capture/compare channel 1 main circuit
$3%%XV

5HDG&&5/

UHDGBLQBSURJUHVV

&&6>@

ZULWHBLQBSURJUHVV

,QSXW
PRGH

6 ZULWH&&5+

&DSWXUHFRPSDUHSUHORDGUHJLVWHU

ZULWH&&5/

5

5
FRPSDUHBWUDQVIHU

FDSWXUHBWUDQVIHU
&&6>@


ORZ

5HDG&&5+ 6

KLJK



LIELW

0&8SHULSKHUDOLQWHUIDFH

2XWSXW
PRGH

&&6>@
&&6>@
2&3(

&DSWXUHFRPSDUHVKDGRZUHJLVWHU
&RPSDUDWRU

&DSWXUH

,&36

8(9

&&(

&17!&&5
&RXQWHU

IURPWLPH
EDVHXQLW

2&3(

7,0[B&&05

&17 &&5

&&*
7,0[B(*5

069

Figure 161. Output stage of capture/compare channel (channel 1)
%42&
4O THE MASTER MODE
CONTROLLER



/UTPUT
%NABLE
#IRCUIT

/#

##0

4)-X?#.4  4)-X?##2

/UTPUT MODE OCREF
4)-X?#.4  4)-X?##2 CONTROLLER

4)-X?##%2

##%
/#-;=
4)-X?##-2

4)-X?##%2
AIB

The capture/compare block is made of one preload register and one shadow register. Write
and read always access the preload register.
In capture mode, captures are actually done in the shadow register, which is copied into the
preload register.
In compare mode, the content of the preload register is copied into the shadow register
which is compared to the counter.

606/1747

RM0090 Rev 17

RM0090

18.3.5

General-purpose timers (TIM2 to TIM5)

Input capture mode
In Input capture mode, the Capture/Compare Registers (TIMx_CCRx) are used to latch the
value of the counter after a transition detected by the corresponding ICx signal. When a
capture occurs, the corresponding CCXIF flag (TIMx_SR register) is set and an interrupt or
a DMA request can be sent if they are enabled. If a capture occurs while the CCxIF flag was
already high, then the over-capture flag CCxOF (TIMx_SR register) is set. CCxIF can be
cleared by software by writing it to 0 or by reading the captured data stored in the
TIMx_CCRx register. CCxOF is cleared when written to 0.
The following example shows how to capture the counter value in TIMx_CCR1 when TI1
input rises. To do this, use the following procedure:
•

Select the active input: TIMx_CCR1 must be linked to the TI1 input, so write the CC1S
bits to 01 in the TIMx_CCMR1 register. As soon as CC1S becomes different from 00,
the channel is configured in input and the TIMx_CCR1 register becomes read-only.

•

Program the needed input filter duration with respect to the signal connected to the
timer (by programming the ICxF bits in the TIMx_CCMRx register if the input is one of
the TIx inputs). Let’s imagine that, when toggling, the input signal is not stable during at
must five internal clock cycles. We must program a filter duration longer than these five
clock cycles. We can validate a transition on TI1 when eight consecutive samples with
the new level have been detected (sampled at fDTS frequency). Then write IC1F bits to
0011 in the TIMx_CCMR1 register.

•

Select the edge of the active transition on the TI1 channel by writing the CC1P and
CC1NP bits to 00 in the TIMx_CCER register (rising edge in this case).

•

Program the input prescaler. In our example, we wish the capture to be performed at
each valid transition, so the prescaler is disabled (write IC1PS bits to 00 in the
TIMx_CCMR1 register).

•

Enable capture from the counter into the capture register by setting the CC1E bit in the
TIMx_CCER register.

•

If needed, enable the related interrupt request by setting the CC1IE bit in the
TIMx_DIER register, and/or the DMA request by setting the CC1DE bit in the
TIMx_DIER register.

When an input capture occurs:
•

The TIMx_CCR1 register gets the value of the counter on the active transition.

•

CC1IF flag is set (interrupt flag). CC1OF is also set if at least two consecutive captures
occurred whereas the flag was not cleared.

•

An interrupt is generated depending on the CC1IE bit.

•

A DMA request is generated depending on the CC1DE bit.

In order to handle the overcapture, it is recommended to read the data before the
overcapture flag. This is to avoid missing an overcapture which could happen after reading
the flag and before reading the data.
Note:

IC interrupt and/or DMA requests can be generated by software by setting the
corresponding CCxG bit in the TIMx_EGR register.

RM0090 Rev 17

607/1747
649

General-purpose timers (TIM2 to TIM5)

18.3.6

RM0090

PWM input mode
This mode is a particular case of input capture mode. The procedure is the same except:
•

Two ICx signals are mapped on the same TIx input.

•

These 2 ICx signals are active on edges with opposite polarity.

•

One of the two TIxFP signals is selected as trigger input and the slave mode controller
is configured in reset mode.

For example, the user can measure the period (in TIMx_CCR1 register) and the duty cycle
(in TIMx_CCR2 register) of the PWM applied on TI1 using the following procedure
(depending on CK_INT frequency and prescaler value):
•

Select the active input for TIMx_CCR1: write the CC1S bits to 01 in the TIMx_CCMR1
register (TI1 selected).

•

Select the active polarity for TI1FP1 (used both for capture in TIMx_CCR1 and counter
clear): write the CC1P to ‘0’ and the CC1NP bit to ‘0’ (active on rising edge).

•

Select the active input for TIMx_CCR2: write the CC2S bits to 10 in the TIMx_CCMR1
register (TI1 selected).

•

Select the active polarity for TI1FP2 (used for capture in TIMx_CCR2): write the CC2P
bit to ‘1’ and the CC2NP bit to ’0’(active on falling edge).

•

Select the valid trigger input: write the TS bits to 101 in the TIMx_SMCR register
(TI1FP1 selected).

•

Configure the slave mode controller in reset mode: write the SMS bits to 100 in the
TIMx_SMCR register.

•

Enable the captures: write the CC1E and CC2E bits to ‘1 in the TIMx_CCER register.
Figure 162. PWM input mode timing
7,

7,0[B&17









7,0[B&&5



7,0[B&&5



,&FDSWXUH
,&FDSWXUH
UHVHWFRXQWHU





,&FDSWXUH
SX
O HVZLG WK
PHDVXUHPHQW



,&FDSWXUH
SHULRG
PHDVXUHPHQW
DL

608/1747

RM0090 Rev 17

RM0090

18.3.7

General-purpose timers (TIM2 to TIM5)

Forced output mode
In output mode (CCxS bits = 00 in the TIMx_CCMRx register), each output compare signal
(OCxREF and then OCx) can be forced to active or inactive level directly by software,
independently of any comparison between the output compare register and the counter.
To force an output compare signal (ocxref/OCx) to its active level, the user just needs to
write 101 in the OCxM bits in the corresponding TIMx_CCMRx register. Thus ocxref is
forced high (OCxREF is always active high) and OCx get opposite value to CCxP polarity
bit.
e.g.: CCxP=0 (OCx active high) => OCx is forced to high level.
ocxref signal can be forced low by writing the OCxM bits to 100 in the TIMx_CCMRx
register.
Anyway, the comparison between the TIMx_CCRx shadow register and the counter is still
performed and allows the flag to be set. Interrupt and DMA requests can be sent
accordingly. This is described in the Output Compare Mode section.

18.3.8

Output compare mode
This function is used to control an output waveform or indicating when a period of time has
elapsed.
When a match is found between the capture/compare register and the counter, the output
compare function:
•

Assigns the corresponding output pin to a programmable value defined by the output
compare mode (OCxM bits in the TIMx_CCMRx register) and the output polarity (CCxP
bit in the TIMx_CCER register). The output pin can keep its level (OCXM=000), be set
active (OCxM=001), be set inactive (OCxM=010) or can toggle (OCxM=011) on match.

•

Sets a flag in the interrupt status register (CCxIF bit in the TIMx_SR register).

•

Generates an interrupt if the corresponding interrupt mask is set (CCXIE bit in the
TIMx_DIER register).

•

Sends a DMA request if the corresponding enable bit is set (CCxDE bit in the
TIMx_DIER register, CCDS bit in the TIMx_CR2 register for the DMA request
selection).

The TIMx_CCRx registers can be programmed with or without preload registers using the
OCxPE bit in the TIMx_CCMRx register.
In output compare mode, the update event UEV has no effect on ocxref and OCx output.
The timing resolution is one count of the counter. Output compare mode can also be used to
output a single pulse (in One-pulse mode).
Procedure:
1.

Select the counter clock (internal, external, prescaler).

2.

Write the desired data in the TIMx_ARR and TIMx_CCRx registers.

3.

Set the CCxIE and/or CCxDE bits if an interrupt and/or a DMA request is to be
generated.

4.

Select the output mode. For example, the user must write OCxM=011, OCxPE=0,
CCxP=0 and CCxE=1 to toggle OCx output pin when CNT matches CCRx, CCRx
preload is not used, OCx is enabled and active high.

5.

Enable the counter by setting the CEN bit in the TIMx_CR1 register.

RM0090 Rev 17

609/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

The TIMx_CCRx register can be updated at any time by software to control the output
waveform, provided that the preload register is not enabled (OCxPE=0, else TIMx_CCRx
shadow register is updated only at the next update event UEV). An example is given in
Figure 163.
Figure 163. Output compare mode, toggle on OC1
:ULWH%KLQWKH&&5UHJLVWHU

7,0[B&17



7,0[B&&5

$

%
$

%

%

%

2&5() 2&

0DWFKGHWHFWHGRQ&&5
,QWHUUXSWJHQHUDWHGLIHQDEOHG
069

18.3.9

PWM mode
Pulse width modulation mode allows generating a signal with a frequency determined by the
value of the TIMx_ARR register and a duty cycle determined by the value of the
TIMx_CCRx register.
The PWM mode can be selected independently on each channel (one PWM per OCx
output) by writing 110 (PWM mode 1) or ‘111 (PWM mode 2) in the OCxM bits in the
TIMx_CCMRx register. The user must enable the corresponding preload register by setting
the OCxPE bit in the TIMx_CCMRx register, and eventually the auto-reload preload register
by setting the ARPE bit in the TIMx_CR1 register.
As the preload registers are transferred to the shadow registers only when an update event
occurs, before starting the counter, the user has to initialize all the registers by setting the
UG bit in the TIMx_EGR register.
OCx polarity is software programmable using the CCxP bit in the TIMx_CCER register. It
can be programmed as active high or active low. OCx output is enabled by the CCxE bit in
the TIMx_CCER register. Refer to the TIMx_CCERx register description for more details.
In PWM mode (1 or 2), TIMx_CNT and TIMx_CCRx are always compared to determine
whether TIMx_CCRx≤TIMx_CNT or TIMx_CNT≤TIMx_CCRx (depending on the direction of
the counter). However, to comply with the ETRF (OCREF can be cleared by an external
event through the ETR signal until the next PWM period), the OCREF signal is asserted
only:
•

When the result of the comparison changes, or

•

When the output compare mode (OCxM bits in TIMx_CCMRx register) switches from
the “frozen” configuration (no comparison, OCxM=‘000) to one of the PWM modes
(OCxM=‘110 or ‘111).

This forces the PWM by software while the timer is running.

610/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)
The timer is able to generate PWM in edge-aligned mode or center-aligned mode
depending on the CMS bits in the TIMx_CR1 register.

PWM edge-aligned mode
Upcounting configuration
Upcounting is active when the DIR bit in the TIMx_CR1 register is low. Refer to Upcounting
mode.
In the following example, we consider PWM mode 1. The reference PWM signal OCxREF is
high as long as TIMx_CNT TIMx_CCRx else
it becomes high. If the compare value in TIMx_CCRx is greater than the auto-reload value in
TIMx_ARR, then ocxref is held at ‘1. 0% PWM is not possible in this mode.

PWM center-aligned mode
Center-aligned mode is active when the CMS bits in TIMx_CR1 register are different from
‘00 (all the remaining configurations having the same effect on the ocxref/OCx signals). The
compare flag is set when the counter counts up, when it counts down or both when it counts
RM0090 Rev 17

611/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

up and down depending on the CMS bits configuration. The direction bit (DIR) in the
TIMx_CR1 register is updated by hardware and must not be changed by software. Refer to
Center-aligned mode (up/down counting).
Figure 165 shows some center-aligned PWM waveforms in an example where:
•

TIMx_ARR=8,

•

PWM mode is the PWM mode 1,

•

The flag is set when the counter counts down corresponding to the center-aligned
mode 1 selected for CMS=01 in TIMx_CR1 register.
Figure 165. Center-aligned PWM waveforms (ARR=8)
#OUNTER REGISTER





































/#X2%&
##2X  
#-3
#-3
#-3

##X)&

/#X2%&
##2X  

#-3 OR 

##X)&
/#X2%&
##2X  

gg

#-3
#-3
#-3

##X)&

/#X2%&
##2X  

gg

#-3
#-3
#-3

##X)&

/#X2%&
##2X  
##X)&

gg

#-3
#-3
#-3
AIB

Hints on using center-aligned mode:
•

612/1747

When starting in center-aligned mode, the current up-down configuration is used. It
means that the counter counts up or down depending on the value written in the DIR bit

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)
in the TIMx_CR1 register. Moreover, the DIR and CMS bits must not be changed at the
same time by the software.
•

•

–

The direction is not updated if the user writes a value in the counter that is greater
than the auto-reload value (TIMx_CNT>TIMx_ARR). For example, if the counter
was counting up, it continues to count up.

–

The direction is updated if the user writes 0 or write the TIMx_ARR value in the
counter but no Update Event UEV is generated.

The safest way to use center-aligned mode is to generate an update by software
(setting the UG bit in the TIMx_EGR register) just before starting the counter and not to
write the counter while it is running.

One-pulse mode
One-pulse mode (OPM) is a particular case of the previous modes. It allows the counter to
be started in response to a stimulus and to generate a pulse with a programmable length
after a programmable delay.
Starting the counter can be controlled through the slave mode controller. Generating the
waveform can be done in output compare mode or PWM mode. Select One-pulse mode by
setting the OPM bit in the TIMx_CR1 register. This makes the counter stop automatically at
the next update event UEV.
A pulse can be correctly generated only if the compare value is different from the counter
initial value. Before starting (when the timer is waiting for the trigger), the configuration must
be:
•

In upcounting: CNTCCRx.
Figure 166. Example of one-pulse mode
7,
2&5()
2&

7,0B$55
&RXQWHU

18.3.10

Writing to the counter while running in center-aligned mode is not recommended as it
can lead to unexpected results. In particular:

7,0B&&5


W'(/$<

W38/6(

W

069

For example the user may want to generate a positive pulse on OC1 with a length of tPULSE
and after a delay of tDELAY as soon as a positive edge is detected on the TI2 input pin.

RM0090 Rev 17

613/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Let’s use TI2FP2 as trigger 1:
•

Map TI2FP2 on TI2 by writing CC2S=01 in the TIMx_CCMR1 register.

•

TI2FP2 must detect a rising edge, write CC2P=0 and CC2NP=’0’ in the TIMx_CCER
register.

•

Configure TI2FP2 as trigger for the slave mode controller (TRGI) by writing TS=110 in
the TIMx_SMCR register.

•

TI2FP2 is used to start the counter by writing SMS to ‘110 in the TIMx_SMCR register
(trigger mode).

The OPM waveform is defined by writing the compare registers (taking into account the
clock frequency and the counter prescaler).
•

The tDELAY is defined by the value written in the TIMx_CCR1 register.

•

The tPULSE is defined by the difference between the auto-reload value and the compare
value (TIMx_ARR - TIMx_CCR + 1).

•

Let us say user wants to build a waveform with a transition from ‘0 to ‘1 when a
compare match occurs and a transition from ‘1 to ‘0 when the counter reaches the
auto-reload value. To do this enable PWM mode 2 by writing OC1M=111 in the
TIMx_CCMR1 register. The user can optionally enable the preload registers by writing
OC1PE=1 in the TIMx_CCMR1 register and ARPE in the TIMx_CR1 register. In this
case write the compare value in the TIMx_CCR1 register, the auto-reload value in the
TIMx_ARR register, generate an update by setting the UG bit and wait for external
trigger event on TI2. CC1P is written to ‘0 in this example.

In our example, the DIR and CMS bits in the TIMx_CR1 register should be low.
User only wants one pulse (Single mode), so write '1 in the OPM bit in the TIMx_CR1
register to stop the counter at the next update event (when the counter rolls over from the
auto-reload value back to 0). When OPM bit in the TIMx_CR1 register is set to '0', so the
Repetitive Mode is selected.

Particular case: OCx fast enable:
In One-pulse mode, the edge detection on TIx input set the CEN bit which enables the
counter. Then the comparison between the counter and the compare value makes the
output toggle. But several clock cycles are needed for these operations and it limits the
minimum delay tDELAY min we can get.
To output a waveform with the minimum delay, the user can set the OCxFE bit in the
TIMx_CCMRx register. Then OCxRef (and OCx) is forced in response to the stimulus,
without taking in account the comparison. Its new level is the same as if a compare match
had occurred. OCxFE acts only if the channel is configured in PWM1 or PWM2 mode.

18.3.11

Clearing the OCxREF signal on an external event
The OCxREF signal for a given channel can be driven Low by applying a High level to the
ETRF input (OCxCE enable bit of the corresponding TIMx_CCMRx register set to '1'). The
OCxREF signal remains Low until the next update event, UEV, occurs.
This function can only be used in output compare and PWM modes, and does not work in
forced mode.
For example, the ETR signal can be connected to the output of a comparator to be used for
current handling. In this case, ETR must be configured as follows:

614/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)
1.

The external trigger prescaler should be kept off: bits ETPS[1:0] in the TIMx_SMCR
register are cleared to 00.

2.

The external clock mode 2 must be disabled: bit ECE in the TIM1_SMCR register is
cleared to 0.

3.

The external trigger polarity (ETP) and the external trigger filter (ETF) can be
configured according to the application’s needs.

Figure 167 shows the behavior of the OCxREF signal when the ETRF input becomes high,
for both values of the OCxCE enable bit. In this example, the timer TIMx is programmed in
PWM mode.
Figure 167. Clearing TIMx OCxREF
&&5[

&RXQWHU &17
(75)
2&[5() 2&[&( µ¶
2&[5() 2&[&( µ¶
(75)EHFRPHVKLJK

(75)VWLOOKLJK
069

1. In case of a PWM with a 100% duty cycle (if CCRx>ARR), OCxREF is enabled again at the next counter
overflow.

18.3.12

Encoder interface mode
To select Encoder Interface mode write SMS=‘001 in the TIMx_SMCR register if the counter
is counting on TI2 edges only, SMS=010 if it is counting on TI1 edges only and SMS=011 if
it is counting on both TI1 and TI2 edges.
Select the TI1 and TI2 polarity by programming the CC1P and CC2P bits in the TIMx_CCER
register. When needed, program the input filter as well.
The two inputs TI1 and TI2 are used to interface to an incremental encoder. Refer to
Table 97. The counter is clocked by each valid transition on TI1FP1 or TI2FP2 (TI1 and TI2
after input filter and polarity selection, TI1FP1=TI1 if not filtered and not inverted,
TI2FP2=TI2 if not filtered and not inverted) assuming that it is enabled (CEN bit in
TIMx_CR1 register written to ‘1). The sequence of transitions of the two inputs is evaluated
and generates count pulses as well as the direction signal. Depending on the sequence the
counter counts up or down, the DIR bit in the TIMx_CR1 register is modified by hardware
accordingly. The DIR bit is calculated at each transition on any input (TI1 or TI2), whatever
the counter is counting on TI1 only, TI2 only or both TI1 and TI2.
Encoder interface mode acts simply as an external clock with direction selection. This
means that the counter just counts continuously between 0 and the auto-reload value in the
TIMx_ARR register (0 to ARR or ARR down to 0 depending on the direction). So the user
must configure TIMx_ARR before starting. In the same way, the capture, compare,
prescaler, trigger output features continue to work as normal.
RM0090 Rev 17

615/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

In this mode, the counter is modified automatically following the speed and the direction of
the incremental encoder and its content, therefore, always represents the encoder’s
position. The count direction correspond to the rotation direction of the connected sensor.
The table summarizes the possible combinations, assuming TI1 and TI2 don’t switch at the
same time.
Table 97. Counting direction versus encoder signals
Level on opposite
signal (TI1FP1 for
TI2, TI2FP2 for TI1)

Rising

Falling

Rising

Falling

Counting on
TI1 only

High

Down

Up

No Count

No Count

Low

Up

Down

No Count

No Count

Counting on
TI2 only

High

No Count

No Count

Up

Down

Low

No Count

No Count

Down

Up

Counting on
TI1 and TI2

High

Down

Up

Up

Down

Low

Up

Down

Down

Up

Active edge

TI1FP1 signal

TI2FP2 signal

An external incremental encoder can be connected directly to the MCU without external
interface logic. However, comparators are normally be used to convert the encoder’s
differential outputs to digital signals. This greatly increases noise immunity. The third
encoder output which indicate the mechanical zero position, may be connected to an
external interrupt input and trigger a counter reset.
Figure 168 gives an example of counter operation, showing count signal generation and
direction control. It also shows how input jitter is compensated where both edges are
selected. This might occur if the sensor is positioned near to one of the switching points. For
this example we assume that the configuration is the following:

616/1747

•

CC1S= ‘01’ (TIMx_CCMR1 register, TI1FP1 mapped on TI1)

•

CC2S= ‘01’ (TIMx_CCMR2 register, TI2FP2 mapped on TI2)

•

CC1P= ‘0’, CC1NP = ‘0’, IC1F =’0000’ (TIMx_CCER register, TI1FP1 noninverted,
TI1FP1=TI1)

•

CC2P= ‘0’, CC2NP = ‘0’, IC2F =’0000’ (TIMx_CCER register, TI2FP2 noninverted,
TI2FP2=TI2)

•

SMS= ‘011’ (TIMx_SMCR register, both inputs are active on both rising and falling
edges)

•

CEN = 1 (TIMx_CR1 register, Counter is enabled)

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)
Figure 168. Example of counter operation in encoder interface mode
IRUZDUG

MLWWHU

EDFNZDUG

MLWWHU

IRUZDUG

7,
7,

&RXQWHU

XS

GRZQ

XS

069

Figure 169 gives an example of counter behavior when TI1FP1 polarity is inverted (same
configuration as above except CC1P=1).
Figure 169. Example of encoder interface mode with TI1FP1 polarity inverted
IRUZDUG

MLWWHU

EDFNZDUG

MLWWHU

IRUZDUG

7,
7,

&RXQWHU

GRZQ

XS

GRZQ

069

The timer, when configured in Encoder Interface mode provides information on the sensor’s
current position. The user can obtain dynamic information (speed, acceleration,
deceleration) by measuring the period between two encoder events using a second timer
configured in capture mode. The output of the encoder which indicates the mechanical zero
can be used for this purpose. Depending on the time between two events, the counter can
also be read at regular times. The user can do this by latching the counter value into a third
input capture register if available (then the capture signal must be periodic and can be
generated by another timer). when available, it is also possible to read its value through a
DMA request generated by a Real-Time clock.

RM0090 Rev 17

617/1747
649

General-purpose timers (TIM2 to TIM5)

18.3.13

RM0090

Timer input XOR function
The TI1S bit in the TIM_CR2 register, allows the input filter of channel 1 to be connected to
the output of a XOR gate, combining the three input pins TIMx_CH1 to TIMx_CH3.
The XOR output can be used with all the timer input functions such as trigger or input
capture.

18.3.14

Timers and external trigger synchronization
The TIMx Timers can be synchronized with an external trigger in several modes: Reset
mode, Gated mode and Trigger mode.

Slave mode: Reset mode
The counter and its prescaler can be reinitialized in response to an event on a trigger input.
Moreover, if the URS bit from the TIMx_CR1 register is low, an update event UEV is
generated. Then all the preloaded registers (TIMx_ARR, TIMx_CCRx) are updated.
In the following example, the upcounter is cleared in response to a rising edge on TI1 input:
•

Configure the channel 1 to detect rising edges on TI1. Configure the input filter duration
(in this example, we don’t need any filter, so we keep IC1F=0000). The capture
prescaler is not used for triggering, so the user does not need to configure it. The
CC1S bits select the input capture source only, CC1S = 01 in the TIMx_CCMR1
register. Write CC1P=0 and CC1NP=0 in TIMx_CCER register to validate the polarity
(and detect rising edges only).

•

Configure the timer in reset mode by writing SMS=100 in TIMx_SMCR register. Select
TI1 as the input source by writing TS=101 in TIMx_SMCR register.

•

Start the counter by writing CEN=1 in the TIMx_CR1 register.

The counter starts counting on the internal clock, then behaves normally until TI1 rising
edge. When TI1 rises, the counter is cleared and restarts from 0. In the meantime, the
trigger flag is set (TIF bit in the TIMx_SR register) and an interrupt request, or a DMA
request can be sent if enabled (depending on the TIE and TDE bits in TIMx_DIER register).
Figure 170 shows this behavior when the auto-reload register TIMx_ARR=0x36. The delay
between the rising edge on TI1 and the actual reset of the counter is due to the
resynchronization circuit on TI1 input.
Figure 170. Control circuit in reset mode

7,

8*
&RXQWHUFORFN &.B&17 &.B36&
&RXQWHUUHJLVWHU

              

7,)
069

618/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)

Slave mode: Gated mode
The counter can be enabled depending on the level of a selected input.
In the following example, the upcounter counts only when TI1 input is low:
•

Configure the channel 1 to detect low levels on TI1. Configure the input filter duration
(in this example, we don’t need any filter, so we keep IC1F=0000). The capture
prescaler is not used for triggering, so the user does not need to configure it. The
CC1S bits select the input capture source only, CC1S=01 in TIMx_CCMR1 register.
Write CC1P=1 in TIMx_CCER register to validate the polarity (and detect low level
only).

•

Configure the timer in gated mode by writing SMS=101 in TIMx_SMCR register. Select
TI1 as the input source by writing TS=101 in TIMx_SMCR register.

•

Enable the counter by writing CEN=1 in the TIMx_CR1 register (in gated mode, the
counter doesn’t start if CEN=0, whatever is the trigger input level).

The counter starts counting on the internal clock as long as TI1 is low and stops as soon as
TI1 becomes high. The TIF flag in the TIMx_SR register is set both when the counter starts
or stops.
The delay between the rising edge on TI1 and the actual stop of the counter is due to the
resynchronization circuit on TI1 input.
Figure 171. Control circuit in gated mode

7,
&17B(1

&RXQWHUFORFN &.B&17 &.B36&
&RXQWHUUHJLVWHU



















7,)

:ULWH7,) 
069

1. The configuration “CCxP=CCxNP=1” (detection of both rising and falling edges) does not have any effect
in gated mode because gated mode acts on a level and not on an edge.

Slave mode: Trigger mode
The counter can start in response to an event on a selected input.
In the following example, the upcounter starts in response to a rising edge on TI2 input:
•

Configure the channel 2 to detect rising edges on TI2. Configure the input filter duration
(in this example, we don’t need any filter, so we keep IC2F=0000). The capture
prescaler is not used for triggering, so the user does not need to configure it. CC2S bits
are selecting the input capture source only, CC2S=01 in TIMx_CCMR1 register. Write
CC2P=1 in TIMx_CCER register to validate the polarity (and detect low level only).

•

Configure the timer in trigger mode by writing SMS=110 in TIMx_SMCR register. Select
TI2 as the input source by writing TS=110 in TIMx_SMCR register.

RM0090 Rev 17

619/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

When a rising edge occurs on TI2, the counter starts counting on the internal clock and the
TIF flag is set.
The delay between the rising edge on TI2 and the actual start of the counter is due to the
resynchronization circuit on TI2 input.
Figure 172. Control circuit in trigger mode

7,
&17B(1
&RXQWHUFORFN &.B&17 &.B36&
&RXQWHUUHJLVWHU



   

7,)
069

Slave mode: External Clock mode 2 + trigger mode
The external clock mode 2 can be used in addition to another slave mode (except external
clock mode 1 and encoder mode). In this case, the ETR signal is used as external clock
input, and another input can be selected as trigger input when operating in reset mode,
gated mode or trigger mode. It is recommended not to select ETR as TRGI through the TS
bits of TIMx_SMCR register.
In the following example, the upcounter is incremented at each rising edge of the ETR
signal as soon as a rising edge of TI1 occurs:
1.

2.

3.

Configure the external trigger input circuit by programming the TIMx_SMCR register as
follows:
–

ETF = 0000: no filter

–

ETPS = 00: prescaler disabled

–

ETP = 0: detection of rising edges on ETR and ECE=1 to enable the external clock
mode 2.

Configure the channel 1 as follows, to detect rising edges on TI:
–

IC1F = 0000: no filter.

–

The capture prescaler is not used for triggering and does not need to be
configured.

–

CC1S = 01 in TIMx_CCMR1 register to select only the input capture source

–

CC1P = 0 in TIMx_CCER register to validate the polarity (and detect rising edge
only).

Configure the timer in trigger mode by writing SMS=110 in TIMx_SMCR register. Select
TI1 as the input source by writing TS=101 in TIMx_SMCR register.

A rising edge on TI1 enables the counter and sets the TIF flag. The counter then counts on
ETR rising edges.
The delay between the rising edge of the ETR signal and the actual reset of the counter is
due to the resynchronization circuit on ETRP input.

620/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)
Figure 173. Control circuit in external clock mode 2 + trigger mode
7,

&(1&17B(1

(75

&RXQWHUFORFN &.B&17 &.B36&

&RXQWHUUHJLVWHU







7,)
069

18.3.15

Timer synchronization
The TIMx timers are linked together internally for timer synchronization or chaining. When
one Timer is configured in Master Mode, it can reset, start, stop or clock the counter of
another Timer configured in Slave Mode.
Figure 174 presents an overview of the trigger selection and the master mode selection
blocks.

Note:

The clock of the slave timer must be enabled prior to receiving events from the master timer,
and must not be changed on-the-fly while triggers are received from the master timer.

Using one timer as prescaler for another timer
Figure 174. Master/Slave timer example

7,0

7,0

&ORFN

76

006
8(9

3UHVFDOHU

&RXQWHU

0DVWHU
PRGH
FRQWURO

75*2

,75

606

6ODYH
&.B36&
PRGH
FRQWURO
3UHVFDOHU

&RXQWHU

,QSXWWULJJHU
VHOHFWLRQ
069

RM0090 Rev 17

621/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

For example, the user can configure Timer 1 to act as a prescaler for Timer 2 (see
Figure 174). To do this:

Note:

•

Configure Timer 1 in master mode so that it outputs a periodic trigger signal on each
update event UEV. If you write MMS=010 in the TIM1_CR2 register, a rising edge is
output on TRGO1 each time an update event is generated.

•

To connect the TRGO1 output of Timer 1 to Timer 2, Timer 2 must be configured in
slave mode using ITR0 as internal trigger. You select this through the TS bits in the
TIM2_SMCR register (writing TS=000).

•

Then you put the slave mode controller in external clock mode 1 (write SMS=111 in the
TIM2_SMCR register). This causes Timer 2 to be clocked by the rising edge of the
periodic Timer 1 trigger signal (which correspond to the timer 1 counter overflow).

•

Finally both timers must be enabled by setting their respective CEN bits (TIMx_CR1
register).

If OCx is selected on Timer 1 as trigger output (MMS=1xx), its rising edge is used to clock
the counter of timer 2.

Using one timer to enable another timer
In this example, we control the enable of Timer 2 with the output compare 1 of Timer 1.
Refer to Figure 174 for connections. Timer 2 counts on the divided internal clock only when
OC1REF of Timer 1 is high. Both counter clock frequencies are divided by 3 by the
prescaler compared to CK_INT (fCK_CNT = fCK_INT/3).

Note:

•

Configure Timer 1 master mode to send its Output Compare 1 Reference (OC1REF)
signal as trigger output (MMS=100 in the TIM1_CR2 register).

•

Configure the Timer 1 OC1REF waveform (TIM1_CCMR1 register).

•

Configure Timer 2 to get the input trigger from Timer 1 (TS=000 in the TIM2_SMCR
register).

•

Configure Timer 2 in gated mode (SMS=101 in TIM2_SMCR register).

•

Enable Timer 2 by writing ‘1 in the CEN bit (TIM2_CR1 register).

•

Start Timer 1 by writing ‘1 in the CEN bit (TIM1_CR1 register).

The counter 2 clock is not synchronized with counter 1, this mode only affects the Timer 2
counter enable signal.
Figure 175. Gating timer 2 with OC1REF of timer 1

&.B,17
7,0(52&5()
7,0(5&17

)&

7,0(5&17



)'

)(


))







7,0(57,)
:ULWH7,) 

069

In the example in Figure 175, the Timer 2 counter and prescaler are not initialized before
being started. So they start counting from their current value. It is possible to start from a
given value by resetting both timers before starting Timer 1. You can then write any value

622/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)
you want in the timer counters. The timers can easily be reset by software using the UG bit
in the TIMx_EGR registers.
In the next example, we synchronize Timer 1 and Timer 2. Timer 1 is the master and starts
from 0. Timer 2 is the slave and starts from 0xE7. The prescaler ratio is the same for both
timers. Timer 2 stops when Timer 1 is disabled by writing ‘0 to the CEN bit in the TIM1_CR1
register:
•

Configure Timer 1 master mode to send its Output Compare 1 Reference (OC1REF)
signal as trigger output (MMS=100 in the TIM1_CR2 register).

•

Configure the Timer 1 OC1REF waveform (TIM1_CCMR1 register).

•

Configure Timer 2 to get the input trigger from Timer 1 (TS=000 in the TIM2_SMCR
register).

•

Configure Timer 2 in gated mode (SMS=101 in TIM2_SMCR register).

•

Reset Timer 1 by writing ‘1 in UG bit (TIM1_EGR register).

•

Reset Timer 2 by writing ‘1 in UG bit (TIM2_EGR register).

•

Initialize Timer 2 to 0xE7 by writing ‘0xE7’ in the timer 2 counter (TIM2_CNTL).

•

Enable Timer 2 by writing ‘1 in the CEN bit (TIM2_CR1 register).

•

Start Timer 1 by writing ‘1 in the CEN bit (TIM1_CR1 register).

•

Stop Timer 1 by writing ‘0 in the CEN bit (TIM1_CR1 register).
Figure 176. Gating timer 2 with Enable of timer 1

&.B,17
7,0(5&(1 &17B(1
7,0(5&17B,1,7
7,0(5&17
7,0(5&17





$%



(





(

(

7,0(5&17B,1,7
7,0(5ZULWH&17
7,0(57,)
:ULWH7,) 
069

RM0090 Rev 17

623/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Using one timer to start another timer
In this example, we set the enable of Timer 2 with the update event of Timer 1. Refer to
Figure 174 for connections. Timer 2 starts counting from its current value (which can be
nonzero) on the divided internal clock as soon as the update event is generated by Timer 1.
When Timer 2 receives the trigger signal its CEN bit is automatically set and the counter
counts until we write ‘0 to the CEN bit in the TIM2_CR1 register. Both counter clock
frequencies are divided by 3 by the prescaler compared to CK_INT (fCK_CNT = fCK_INT/3).
•

Configure Timer 1 master mode to send its Update Event (UEV) as trigger output
(MMS=010 in the TIM1_CR2 register).

•

Configure the Timer 1 period (TIM1_ARR registers).

•

Configure Timer 2 to get the input trigger from Timer 1 (TS=000 in the TIM2_SMCR
register).

•

Configure Timer 2 in trigger mode (SMS=110 in TIM2_SMCR register).

•

Start Timer 1 by writing ‘1 in the CEN bit (TIM1_CR1 register).
Figure 177. Triggering timer 2 with update of timer 1

&.B,17
7,0(58(9
7,0(5&17
7,0(5&17

)'

)(



))













7,0(5&(1 &17B(1
7,0(57,)
:ULWH7,) 
069

As in the previous example, the user can initialize both counters before starting counting.
Figure 178 shows the behavior with the same configuration as in Figure 177 but in trigger
mode instead of gated mode (SMS=110 in the TIM2_SMCR register).

624/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)
Figure 178. Triggering timer 2 with Enable of timer 1

&.B,17
7,0(5&(1 &17B(1
7,0(5&17B,1,7
7,0(5&17
7,0(5&17





&'





(

(


(

($

7,0(5&17B,1,7
7,0(5ZULWH&17
7,0(57,)
:ULWH7,) 
069

Starting 2 timers synchronously in response to an external trigger
In this example, we set the enable of timer 1 when its TI1 input rises, and the enable of
Timer 2 with the enable of Timer 1. Refer to Figure 174 for connections. To ensure the
counters are aligned, Timer 1 must be configured in Master/Slave mode (slave with respect
to TI1, master with respect to Timer 2):
•

Configure Timer 1 master mode to send its Enable as trigger output (MMS=001 in the
TIM1_CR2 register).

•

Configure Timer 1 slave mode to get the input trigger from TI1 (TS=100 in the
TIM1_SMCR register).

•

Configure Timer 1 in trigger mode (SMS=110 in the TIM1_SMCR register).

•

Configure the Timer 1 in Master/Slave mode by writing MSM=1 (TIM1_SMCR register).

•

Configure Timer 2 to get the input trigger from Timer 1 (TS=000 in the TIM2_SMCR
register).

•

Configure Timer 2 in trigger mode (SMS=110 in the TIM2_SMCR register).

When a rising edge occurs on TI1 (Timer 1), both counters starts counting synchronously on
the internal clock and both TIF flags are set.
Note:

In this example both timers are initialized before starting (by setting their respective UG
bits). Both counters starts from 0, but you can easily insert an offset between them by
writing any of the counter registers (TIMx_CNT). You can see that the master/slave mode
insert a delay between CNT_EN and CK_PSC on timer 1.

RM0090 Rev 17

625/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Figure 179. Triggering timer 1 and 2 with timer 1 TI1 input

&.B,17
7,0(57,
7,0(5&(1 &17B(1
7,0(5&.B36&
7,0(5&17



        



        

7,0(57,)
7,0(5&(1 &17B(1
7,0(5&.B36&
7,0(5&17
7,0(57,)
069

18.3.16

Debug mode
When the microcontroller enters debug mode (Cortex®-M4 with FPU core - halted), the
TIMx counter either continues to work normally or stops, depending on DBG_TIMx_STOP
configuration bit in DBGMCU module. For more details, refer to Section 38.16.2: Debug
support for timers, watchdog, bxCAN and I2C.

626/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)

18.4

TIM2 to TIM5 registers
Refer to Section 2.2 for a list of abbreviations used in register descriptions.
The 32-bit peripheral registers have to be written by words (32 bits). All other peripheral
registers have to be written by half-words (16 bits) or words (32 bits). Read accesses can be
done by bytes (8 bits), half-words (16 bits) or words (32 bits).

18.4.1

TIMx control register 1 (TIMx_CR1)
Address offset: 0x00
Reset value: 0x0000

15

14

13

12

Reserved

11

10

9

8

CKD[1:0]
rw

7

6

ARPE

rw

rw

5
CMS

rw

rw

4

3

2

1

0

DIR

OPM

URS

UDIS

CEN

rw

rw

rw

rw

rw

Bits 15:10 Reserved, must be kept at reset value.
Bits 9:8 CKD: Clock division
This bit-field indicates the division ratio between the timer clock (CK_INT) frequency and
sampling clock used by the digital filters (ETR, TIx),
00: tDTS = tCK_INT
01: tDTS = 2 × tCK_INT
10: tDTS = 4 × tCK_INT
11: Reserved
Bit 7 ARPE: Auto-reload preload enable
0: TIMx_ARR register is not buffered
1: TIMx_ARR register is buffered
Bits 6:5 CMS: Center-aligned mode selection
00: Edge-aligned mode. The counter counts up or down depending on the direction bit
(DIR).
01: Center-aligned mode 1. The counter counts up and down alternatively. Output compare
interrupt flags of channels configured in output (CCxS=00 in TIMx_CCMRx register) are set
only when the counter is counting down.
10: Center-aligned mode 2. The counter counts up and down alternatively. Output compare
interrupt flags of channels configured in output (CCxS=00 in TIMx_CCMRx register) are set
only when the counter is counting up.
11: Center-aligned mode 3. The counter counts up and down alternatively. Output compare
interrupt flags of channels configured in output (CCxS=00 in TIMx_CCMRx register) are set
both when the counter is counting up or down.
Note: It is not allowed to switch from edge-aligned mode to center-aligned mode as long as
the counter is enabled (CEN=1)
Bit 4 DIR: Direction
0: Counter used as upcounter
1: Counter used as downcounter
Note: This bit is read only when the timer is configured in Center-aligned mode or Encoder
mode.
Bit 3 OPM: One-pulse mode
0: Counter is not stopped at update event
1: Counter stops counting at the next update event (clearing the bit CEN)

RM0090 Rev 17

627/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Bit 2 URS: Update request source
This bit is set and cleared by software to select the UEV event sources.
0: Any of the following events generate an update interrupt or DMA request if enabled.
These events can be:
–
Counter overflow/underflow
–
Setting the UG bit
–
Update generation through the slave mode controller
1: Only counter overflow/underflow generates an update interrupt or DMA request if
enabled.
Bit 1 UDIS: Update disable
This bit is set and cleared by software to enable/disable UEV event generation.
0: UEV enabled. The Update (UEV) event is generated by one of the following events:
–
Counter overflow/underflow
–
Setting the UG bit
–
Update generation through the slave mode controller
Buffered registers are then loaded with their preload values.
1: UEV disabled. The Update event is not generated, shadow registers keep their value
(ARR, PSC, CCRx). However the counter and the prescaler are reinitialized if the UG bit is
set or if a hardware reset is received from the slave mode controller.
Bit 0 CEN: Counter enable
0: Counter disabled
1: Counter enabled
Note: External clock, gated mode and encoder mode can work only if the CEN bit has been
previously set by software. However trigger mode can set the CEN bit automatically by
hardware.
CEN is cleared automatically in one-pulse mode, when an update event occurs.

628/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)

18.4.2

TIMx control register 2 (TIMx_CR2)
Address offset: 0x04
Reset value: 0x0000

15

14

13

12

11

Reserved

10

9

8

7

6

TI1S
rw

5

4

MMS[2:0]
rw

rw

3
CCDS

rw

rw

2

1

0

Reserved

Bits 15:8 Reserved, must be kept at reset value.
Bit 7 TI1S: TI1 selection
0: The TIMx_CH1 pin is connected to TI1 input
1: The TIMx_CH1, CH2 and CH3 pins are connected to the TI1 input (XOR combination)
Bits 6:4 MMS[2:0]: Master mode selection
These bits allow to select the information to be sent in master mode to slave timers for
synchronization (TRGO). The combination is as follows:
000: Reset - the UG bit from the TIMx_EGR register is used as trigger output (TRGO). If the
reset is generated by the trigger input (slave mode controller configured in reset mode) then
the signal on TRGO is delayed compared to the actual reset.
001: Enable - the Counter enable signal, CNT_EN, is used as trigger output (TRGO). It is
useful to start several timers at the same time or to control a window in which a slave timer is
enabled. The Counter Enable signal is generated by a logic OR between CEN control bit
and the trigger input when configured in gated mode.
When the Counter Enable signal is controlled by the trigger input, there is a delay on TRGO,
except if the master/slave mode is selected (see the MSM bit description in TIMx_SMCR
register).
010: Update - The update event is selected as trigger output (TRGO). For instance a master
timer can then be used as a prescaler for a slave timer.
011: Compare Pulse - The trigger output send a positive pulse when the CC1IF flag is to be
set (even if it was already high), as soon as a capture or a compare match occurred.
(TRGO)
100: Compare - OC1REF signal is used as trigger output (TRGO)
101: Compare - OC2REF signal is used as trigger output (TRGO)
110: Compare - OC3REF signal is used as trigger output (TRGO)
111: Compare - OC4REF signal is used as trigger output (TRGO)
Note: The clock of the slave timer and ADC must be enabled prior to receiving events from
the master timer, and must not be changed on-the-fly while triggers are received from
the master timer.
Bit 3 CCDS: Capture/compare DMA selection
0: CCx DMA request sent when CCx event occurs
1: CCx DMA requests sent when update event occurs
Bits 2:0 Reserved, must be kept at reset value.

RM0090 Rev 17

629/1747
649

General-purpose timers (TIM2 to TIM5)

18.4.3

RM0090

TIMx slave mode control register (TIMx_SMCR)
Address offset: 0x08
Reset value: 0x0000

15

14

ETP

ECE

13

rw

rw

12

11

ETPS[1:0]
rw

rw

10

9

8

ETF[3:0]
rw

rw

rw

7

6

MSM
rw

rw

5

4

TS[2:0]
rw

rw

rw

3
Res.

2

1

0

SMS[2:0]
rw

rw

rw

Bit 15 ETP: External trigger polarity
This bit selects whether ETR or ETR is used for trigger operations
0: ETR is noninverted, active at high level or rising edge
1: ETR is inverted, active at low level or falling edge
Bit 14 ECE: External clock enable
This bit enables External clock mode 2.
0: External clock mode 2 disabled
1: External clock mode 2 enabled. The counter is clocked by any active edge on the ETRF
signal.
1: Setting the ECE bit has the same effect as selecting external clock mode 1 with TRGI
connected to ETRF (SMS=111 and TS=111).
2: It is possible to simultaneously use external clock mode 2 with the following slave modes:
reset mode, gated mode and trigger mode. Nevertheless, TRGI must not be connected to
ETRF in this case (TS bits must not be 111).
3: If external clock mode 1 and external clock mode 2 are enabled at the same time, the
external clock input is ETRF.
Bits 13:12 ETPS: External trigger prescaler
External trigger signal ETRP frequency must be at most 1/4 of CK_INT frequency. A
prescaler can be enabled to reduce ETRP frequency. It is useful when inputting fast external
clocks.
00: Prescaler OFF
01: ETRP frequency divided by 2
10: ETRP frequency divided by 4
11: ETRP frequency divided by 8
Bits 11:8 ETF[3:0]: External trigger filter
This bit-field then defines the frequency used to sample ETRP signal and the length of the
digital filter applied to ETRP. The digital filter is made of an event counter in which N
consecutive events are needed to validate a transition on the output:
0000: No filter, sampling is done at fDTS
0001: fSAMPLING=fCK_INT, N=2
0010: fSAMPLING=fCK_INT, N=4
0011: fSAMPLING=fCK_INT, N=8
0100: fSAMPLING=fDTS/2, N=6
0101: fSAMPLING=fDTS/2, N=8
0110: fSAMPLING=fDTS/4, N=6
0111: fSAMPLING=fDTS/4, N=8
1000: fSAMPLING=fDTS/8, N=6
1001: fSAMPLING=fDTS/8, N=8
1010: fSAMPLING=fDTS/16, N=5
1011: fSAMPLING=fDTS/16, N=6
1100: fSAMPLING=fDTS/16, N=8
1101: fSAMPLING=fDTS/32, N=5
1110: fSAMPLING=fDTS/32, N=6
1111: fSAMPLING=fDTS/32, N=8

630/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)

Bit 7 MSM: Master/Slave mode
0: No action
1: The effect of an event on the trigger input (TRGI) is delayed to allow a perfect
synchronization between the current timer and its slaves (through TRGO). It is useful if we
want to synchronize several timers on a single external event.
Bits 6:4 TS: Trigger selection
This bit-field selects the trigger input to be used to synchronize the counter.
000: Internal Trigger 0 (ITR0)
001: Internal Trigger 1 (ITR1).
010: Internal Trigger 2 (ITR2).
011: Internal Trigger 3 (ITR3).
100: TI1 Edge Detector (TI1F_ED)
101: Filtered Timer Input 1 (TI1FP1)
110: Filtered Timer Input 2 (TI2FP2)
111: External Trigger input (ETRF)
See Table 98: TIMx internal trigger connection on page 632 for more details on ITRx
meaning for each Timer.
Note: These bits must be changed only when they are not used (e.g. when SMS=000) to
avoid wrong edge detections at the transition.
Bit 3 Reserved, must be kept at reset value.
Bits 2:0 SMS: Slave mode selection
When external signals are selected the active edge of the trigger signal (TRGI) is linked to
the polarity selected on the external input (see Input Control register and Control Register
description.
000: Slave mode disabled - if CEN = ‘1 then the prescaler is clocked directly by the internal
clock.
001: Encoder mode 1 - Counter counts up/down on TI2FP1 edge depending on TI1FP2
level.
010: Encoder mode 2 - Counter counts up/down on TI1FP2 edge depending on TI2FP1
level.
011: Encoder mode 3 - Counter counts up/down on both TI1FP1 and TI2FP2 edges
depending on the level of the other input.
100: Reset Mode - Rising edge of the selected trigger input (TRGI) reinitializes the counter
and generates an update of the registers.
101: Gated Mode - The counter clock is enabled when the trigger input (TRGI) is high. The
counter stops (but is not reset) as soon as the trigger becomes low. Both start and stop of
the counter are controlled.
110: Trigger Mode - The counter starts at a rising edge of the trigger TRGI (but it is not
reset). Only the start of the counter is controlled.
111: External Clock Mode 1 - Rising edges of the selected trigger (TRGI) clock the counter.
Note: The gated mode must not be used if TI1F_ED is selected as the trigger input (TS=100).
Indeed, TI1F_ED outputs 1 pulse for each transition on TI1F, whereas the gated mode
checks the level of the trigger signal.
The clock of the slave timer must be enabled prior to receiving events from the master
timer, and must not be changed on-the-fly while triggers are received from the master
timer.

RM0090 Rev 17

631/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Table 98. TIMx internal trigger connection

18.4.4

Slave TIM

ITR0 (TS = 000)

ITR1 (TS = 001)

ITR2 (TS = 010)

ITR3 (TS = 011)

TIM2

TIM1_TRGO

TIM8_TRGO

TIM3_TRGO

TIM4_TRGO

TIM3

TIM1_TRGO

TIM2_TRGO

TIM5_TRGO

TIM4_TRGO

TIM4

TIM1_TRGO

TIM2_TRGO

TIM3_TRGO

TIM8_TRGO

TIM5

TIM2_TRGO

TIM3_TRGO

TIM4_TRGO

TIM8_TRGO

TIMx DMA/Interrupt enable register (TIMx_DIER)
Address offset: 0x0C
Reset value: 0x0000

15
Res.

14
TDE
rw

13
Res

Bit 15

12

11

10

9

8

CC4DE CC3DE CC2DE CC1DE
rw

rw

rw

rw

UDE
rw

7
Res.

6
TIE
rw

Reserved, must be kept at reset value.

Bit 14 TDE: Trigger DMA request enable
0: Trigger DMA request disabled.
1: Trigger DMA request enabled.
Bit 13

Reserved, always read as 0

Bit 12 CC4DE: Capture/Compare 4 DMA request enable
0: CC4 DMA request disabled.
1: CC4 DMA request enabled.
Bit 11 CC3DE: Capture/Compare 3 DMA request enable
0: CC3 DMA request disabled.
1: CC3 DMA request enabled.
Bit 10 CC2DE: Capture/Compare 2 DMA request enable
0: CC2 DMA request disabled.
1: CC2 DMA request enabled.
Bit 9 CC1DE: Capture/Compare 1 DMA request enable
0: CC1 DMA request disabled.
1: CC1 DMA request enabled.
Bit 8 UDE: Update DMA request enable
0: Update DMA request disabled.
1: Update DMA request enabled.
Bit 7

Reserved, must be kept at reset value.

Bit 6 TIE: Trigger interrupt enable
0: Trigger interrupt disabled.
1: Trigger interrupt enabled.
Bit 5

Reserved, must be kept at reset value.

Bit 4 CC4IE: Capture/Compare 4 interrupt enable
0: CC4 interrupt disabled.
1: CC4 interrupt enabled.

632/1747

RM0090 Rev 17

5
Res

4

3

2

1

0

CC4IE

CC3IE

CC2IE

CC1IE

UIE

rw

rw

rw

rw

rw

RM0090

General-purpose timers (TIM2 to TIM5)

Bit 3 CC3IE: Capture/Compare 3 interrupt enable
0: CC3 interrupt disabled
1: CC3 interrupt enabled
Bit 2 CC2IE: Capture/Compare 2 interrupt enable
0: CC2 interrupt disabled
1: CC2 interrupt enabled
Bit 1 CC1IE: Capture/Compare 1 interrupt enable
0: CC1 interrupt disabled
1: CC1 interrupt enabled
Bit 0 UIE: Update interrupt enable
0: Update interrupt disabled
1: Update interrupt enabled

18.4.5

TIMx status register (TIMx_SR)
Address offset: 0x10
Reset value: 0x0000

15

14

13

12

11

10

9

8

CC4OF CC3OF CC2OF CC1OF

Reserved

rc_w0

Bits 15:13

rc_w0

rc_w0

rc_w0

7

Reserved

6
TIF
rc_w0

5
Res

4

3

2

1

0

CC4IF

CC3IF

CC2IF

CC1IF

UIF

rc_w0

rc_w0

rc_w0

rc_w0

rc_w0

Reserved, must be kept at reset value.

Bit 12 CC4OF: Capture/Compare 4 overcapture flag
refer to CC1OF description
Bit 11 CC3OF: Capture/Compare 3 overcapture flag
refer to CC1OF description
Bit 10 CC2OF: Capture/compare 2 overcapture flag
refer to CC1OF description
Bit 9 CC1OF: Capture/Compare 1 overcapture flag
This flag is set by hardware only when the corresponding channel is configured in input
capture mode. It is cleared by software by writing it to ‘0.
0: No overcapture has been detected
1: The counter value has been captured in TIMx_CCR1 register while CC1IF flag was
already set
Bits 8:7

Reserved, must be kept at reset value.

Bit 6 TIF: Trigger interrupt flag
This flag is set by hardware on trigger event (active edge detected on TRGI input when the
slave mode controller is enabled in all modes but gated mode. It is set when the counter
starts or stops when gated mode is selected. It is cleared by software.
0: No trigger event occurred
1: Trigger interrupt pending
Bit 5

Reserved, must be kept at reset value.

Bit 4 CC4IF: Capture/Compare 4 interrupt flag
refer to CC1IF description

RM0090 Rev 17

633/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Bit 3 CC3IF: Capture/Compare 3 interrupt flag
refer to CC1IF description
Bit 2 CC2IF: Capture/Compare 2 interrupt flag
refer to CC1IF description
Bit 1 CC1IF: Capture/compare 1 interrupt flag
If channel CC1 is configured as output:
This flag is set by hardware when the counter matches the compare value, with some
exception in center-aligned mode (refer to the CMS bits in the TIMx_CR1 register
description). It is cleared by software.
0: No match
1: The content of the counter TIMx_CNT matches the content of the TIMx_CCR1 register.
When the contents of TIMx_CCR1 are greater than the contents of TIMx_ARR, the CC1IF bit
goes high on the counter overflow (in upcounting and up/down-counting modes) or underflow
(in downcounting mode)
If channel CC1 is configured as input:
This bit is set by hardware on a capture. It is cleared by software or by reading the
TIMx_CCR1 register.
0: No input capture occurred
1: The counter value has been captured in TIMx_CCR1 register (An edge has been detected
on IC1 which matches the selected polarity)
Bit 0 UIF: Update interrupt flag
″
This bit is set by hardware on an update event. It is cleared by software.
0: No update occurred.
1: Update interrupt pending. This bit is set by hardware when the registers are updated:
″
At overflow or underflow (for TIM2 to TIM5) and if UDIS=0 in the TIMx_CR1 register.
″
When CNT is reinitialized by software using the UG bit in TIMx_EGR register, if URS=0
and UDIS=0 in the TIMx_CR1 register.
When CNT is reinitialized by a trigger event (refer to the synchro control register description),
if URS=0 and UDIS=0 in the TIMx_CR1 register.

634/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)

18.4.6

TIMx event generation register (TIMx_EGR)
Address offset: 0x14
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6
TG

Reserved

w

5
Res.

4

3

2

1

0

CC4G

CC3G

CC2G

CC1G

UG

w

w

w

w

w

Bits 15:7 Reserved, must be kept at reset value.
Bit 6 TG: Trigger generation
This bit is set by software in order to generate an event, it is automatically cleared by
hardware.
0: No action
1: The TIF flag is set in TIMx_SR register. Related interrupt or DMA transfer can occur if
enabled.
Bit 5 Reserved, must be kept at reset value.
Bit 4 CC4G: Capture/compare 4 generation
refer to CC1G description
Bit 3 CC3G: Capture/compare 3 generation
refer to CC1G description
Bit 2 CC2G: Capture/compare 2 generation
refer to CC1G description
Bit 1 CC1G: Capture/compare 1 generation
This bit is set by software in order to generate an event, it is automatically cleared by
hardware.
0: No action
1: A capture/compare event is generated on channel 1:
If channel CC1 is configured as output:
CC1IF flag is set, Corresponding interrupt or DMA request is sent if enabled.
If channel CC1 is configured as input:
The current value of the counter is captured in TIMx_CCR1 register. The CC1IF flag is set,
the corresponding interrupt or DMA request is sent if enabled. The CC1OF flag is set if the
CC1IF flag was already high.
Bit 0 UG: Update generation
This bit can be set by software, it is automatically cleared by hardware.
0: No action
1: Re-initialize the counter and generates an update of the registers. Note that the prescaler
counter is cleared too (anyway the prescaler ratio is not affected). The counter is cleared if
the center-aligned mode is selected or if DIR=0 (upcounting), else it takes the auto-reload
value (TIMx_ARR) if DIR=1 (downcounting).

RM0090 Rev 17

635/1747
649

General-purpose timers (TIM2 to TIM5)

18.4.7

RM0090

TIMx capture/compare mode register 1 (TIMx_CCMR1)
Address offset: 0x18
Reset value: 0x0000
The channels can be used in input (capture mode) or in output (compare mode). The
direction of a channel is defined by configuring the corresponding CCxS bits. All the other
bits of this register have a different function in input and in output mode. For a given bit,
OCxx describes its function when the channel is configured in output, ICxx describes its
function when the channel is configured in input. Take care that the same bit can have a
different meaning for the input stage and for the output stage.

15

14

OC2CE

13

12

OC2M[2:0]
IC2F[3:0]

rw

rw

rw

11

10

OC2PE OC2FE
IC2PSC[1:0]
rw

rw

rw

9

8

CC2S[1:0]
rw

7

6

OC1CE

rw

5

4

OC1M[2:0]
IC1F[3:0]

rw

rw

rw

3

2

OC1PE OC1FE
IC1PSC[1:0]
rw

rw

rw

1

0

CC1S[1:0]
rw

rw

Output compare mode
Bit 15 OC2CE: Output compare 2 clear enable
Bits 14:12 OC2M[2:0]: Output compare 2 mode
Bit 11 OC2PE: Output compare 2 preload enable
Bit 10 OC2FE: Output compare 2 fast enable
Bits 9:8 CC2S[1:0]: Capture/Compare 2 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC2 channel is configured as output
01: CC2 channel is configured as input, IC2 is mapped on TI2
10: CC2 channel is configured as input, IC2 is mapped on TI1
11: CC2 channel is configured as input, IC2 is mapped on TRC. This mode is working only if
an internal trigger input is selected through the TS bit (TIMx_SMCR register)
Note: CC2S bits are writable only when the channel is OFF (CC2E = 0 in TIMx_CCER).
Bit 7 OC1CE: Output compare 1 clear enable
OC1CE: Output Compare 1 Clear Enable
0: OC1Ref is not affected by the ETRF input
1: OC1Ref is cleared as soon as a High level is detected on ETRF input

636/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)

Bits 6:4 OC1M: Output compare 1 mode
These bits define the behavior of the output reference signal OC1REF from which OC1 and
OC1N are derived. OC1REF is active high whereas OC1 and OC1N active level depends
on CC1P and CC1NP bits.
000: Frozen - The comparison between the output compare register TIMx_CCR1 and the
counter TIMx_CNT has no effect on the outputs.(this mode is used to generate a timing
base).
001: Set channel 1 to active level on match. OC1REF signal is forced high when the counter
TIMx_CNT matches the capture/compare register 1 (TIMx_CCR1).
010: Set channel 1 to inactive level on match. OC1REF signal is forced low when the
counter TIMx_CNT matches the capture/compare register 1 (TIMx_CCR1).
011: Toggle - OC1REF toggles when TIMx_CNT=TIMx_CCR1.
100: Force inactive level - OC1REF is forced low.
101: Force active level - OC1REF is forced high.
110: PWM mode 1 - In upcounting, channel 1 is active as long as TIMx_CNTTIMx_CCR1 else active (OC1REF=1).
111: PWM mode 2 - In upcounting, channel 1 is inactive as long as TIMx_CNTTIMx_CCR1 else
inactive.
Note: In PWM mode 1 or 2, the OCREF level changes only when the result of the
comparison changes or when the output compare mode switches from “frozen” mode
to “PWM” mode.
Bit 3 OC1PE: Output compare 1 preload enable
0: Preload register on TIMx_CCR1 disabled. TIMx_CCR1 can be written at anytime, the
new value is taken in account immediately.
1: Preload register on TIMx_CCR1 enabled. Read/Write operations access the preload
register. TIMx_CCR1 preload value is loaded in the active register at each update event.
Note: 1: These bits can not be modified as long as LOCK level 3 has been programmed
(LOCK bits in TIMx_BDTR register) and CC1S=00 (the channel is configured in
output).
2: The PWM mode can be used without validating the preload register only in onepulse mode (OPM bit set in TIMx_CR1 register). Else the behavior is not guaranteed.
Bit 2 OC1FE: Output compare 1 fast enable
This bit is used to accelerate the effect of an event on the trigger in input on the CC output.
0: CC1 behaves normally depending on counter and CCR1 values even when the trigger is
ON. The minimum delay to activate CC1 output when an edge occurs on the trigger input is
5 clock cycles.
1: An active edge on the trigger input acts like a compare match on CC1 output. Then, OC
is set to the compare level independently from the result of the comparison. Delay to
sample the trigger input and to activate CC1 output is reduced to 3 clock cycles. OCFE acts
only if the channel is configured in PWM1 or PWM2 mode.
Bits 1:0 CC1S: Capture/Compare 1 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC1 channel is configured as output.
01: CC1 channel is configured as input, IC1 is mapped on TI1.
10: CC1 channel is configured as input, IC1 is mapped on TI2.
11: CC1 channel is configured as input, IC1 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC1S bits are writable only when the channel is OFF (CC1E = 0 in TIMx_CCER).

RM0090 Rev 17

637/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Input capture mode
Bits 15:12 IC2F: Input capture 2 filter
Bits 11:10 IC2PSC[1:0]: Input capture 2 prescaler
Bits 9:8 CC2S: Capture/compare 2 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC2 channel is configured as output.
01: CC2 channel is configured as input, IC2 is mapped on TI2.
10: CC2 channel is configured as input, IC2 is mapped on TI1.
11: CC2 channel is configured as input, IC2 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC2S bits are writable only when the channel is OFF (CC2E = 0 in TIMx_CCER).
Bits 7:4 IC1F: Input capture 1 filter
This bit-field defines the frequency used to sample TI1 input and the length of the digital filter
applied to TI1. The digital filter is made of an event counter in which N consecutive events
are needed to validate a transition on the output:
0000: No filter, sampling is done at fDTS
0001: fSAMPLING=fCK_INT, N=2
0010: fSAMPLING=fCK_INT, N=4
0011: fSAMPLING=fCK_INT, N=8
0100: fSAMPLING=fDTS/2, N=6
0101: fSAMPLING=fDTS/2, N=8
0110: fSAMPLING=fDTS/4, N=6
0111: fSAMPLING=fDTS/4, N=8
1000: fSAMPLING=fDTS/8, N=6
1001: fSAMPLING=fDTS/8, N=8
1010: fSAMPLING=fDTS/16, N=5
1011: fSAMPLING=fDTS/16, N=6
1100: fSAMPLING=fDTS/16, N=8
1101: fSAMPLING=fDTS/32, N=5
1110: fSAMPLING=fDTS/32, N=6
1111: fSAMPLING=fDTS/32, N=8
Bits 3:2 IC1PSC: Input capture 1 prescaler
This bit-field defines the ratio of the prescaler acting on CC1 input (IC1).
The prescaler is reset as soon as CC1E=0 (TIMx_CCER register).
00: no prescaler, capture is done each time an edge is detected on the capture input
01: capture is done once every 2 events
10: capture is done once every 4 events
11: capture is done once every 8 events
Bits 1:0 CC1S: Capture/Compare 1 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC1 channel is configured as output
01: CC1 channel is configured as input, IC1 is mapped on TI1
10: CC1 channel is configured as input, IC1 is mapped on TI2
11: CC1 channel is configured as input, IC1 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC1S bits are writable only when the channel is OFF (CC1E = 0 in TIMx_CCER).

638/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)

18.4.8

TIMx capture/compare mode register 2 (TIMx_CCMR2)
Address offset: 0x1C
Reset value: 0x0000
Refer to the above CCMR1 register description.

15

14

OC4CE

13

12

OC4M[2:0]
IC4F[3:0]

rw

rw

rw

11

10

OC4PE OC4FE
IC4PSC[1:0]
rw

rw

rw

9

8

CC4S[1:0]
rw

7

6

OC3CE

rw

5

4

OC3M[2:0]
IC3F[3:0]

rw

rw

rw

3

2

OC3PE OC3FE
IC3PSC[1:0]
rw

rw

rw

1

0

CC3S[1:0]
rw

rw

Output compare mode
Bit 15 OC4CE: Output compare 4 clear enable
Bits 14:12 OC4M: Output compare 4 mode
Bit 11 OC4PE: Output compare 4 preload enable
Bit 10 OC4FE: Output compare 4 fast enable
Bits 9:8 CC4S: Capture/Compare 4 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC4 channel is configured as output
01: CC4 channel is configured as input, IC4 is mapped on TI4
10: CC4 channel is configured as input, IC4 is mapped on TI3
11: CC4 channel is configured as input, IC4 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC4S bits are writable only when the channel is OFF (CC4E = 0 in TIMx_CCER).
Bit 7 OC3CE: Output compare 3 clear enable
Bits 6:4 OC3M: Output compare 3 mode
Bit 3 OC3PE: Output compare 3 preload enable
Bit 2 OC3FE: Output compare 3 fast enable
Bits 1:0 CC3S: Capture/Compare 3 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC3 channel is configured as output
01: CC3 channel is configured as input, IC3 is mapped on TI3
10: CC3 channel is configured as input, IC3 is mapped on TI4
11: CC3 channel is configured as input, IC3 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC3S bits are writable only when the channel is OFF (CC3E = 0 in TIMx_CCER).

RM0090 Rev 17

639/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Input capture mode
Bits 15:12 IC4F: Input capture 4 filter
Bits 11:10 IC4PSC: Input capture 4 prescaler
Bits 9:8 CC4S: Capture/Compare 4 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC4 channel is configured as output
01: CC4 channel is configured as input, IC4 is mapped on TI4
10: CC4 channel is configured as input, IC4 is mapped on TI3
11: CC4 channel is configured as input, IC4 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC4S bits are writable only when the channel is OFF (CC4E = 0 in TIMx_CCER).
Bits 7:4 IC3F: Input capture 3 filter
Bits 3:2 IC3PSC: Input capture 3 prescaler
Bits 1:0 CC3S: Capture/Compare 3 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC3 channel is configured as output
01: CC3 channel is configured as input, IC3 is mapped on TI3
10: CC3 channel is configured as input, IC3 is mapped on TI4
11: CC3 channel is configured as input, IC3 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC3S bits are writable only when the channel is OFF (CC3E = 0 in TIMx_CCER).

18.4.9

TIMx capture/compare enable register (TIMx_CCER)
Address offset: 0x20
Reset value: 0x0000

15
CC4NP
rw

14
Res.

13

12

11

CC4P

CC4E

CC3NP

rw

rw

rw

10
Res.

9

8

7

CC3P

CC3E

CC2NP

rw

rw

rw

Bit 15 CC4NP: Capture/Compare 4 output Polarity.
Refer to CC1NP description
Bit 14

Reserved, must be kept at reset value.

Bit 13 CC4P: Capture/Compare 4 output Polarity.
refer to CC1P description
Bit 12 CC4E: Capture/Compare 4 output enable.
refer to CC1E description
Bit 11 CC3NP: Capture/Compare 3 output Polarity.
refer to CC1NP description
Bit 10

Reserved, must be kept at reset value.

Bit 9 CC3P: Capture/Compare 3 output Polarity.
refer to CC1P description

640/1747

RM0090 Rev 17

6
Res.

5

4

3

CC2P

CC2E

CC1NP

rw

rw

rw

2
Res.

1

0

CC1P

CC1E

rw

rw

RM0090

General-purpose timers (TIM2 to TIM5)

Bit 8 CC3E: Capture/Compare 3 output enable.
refer to CC1E description
Bit 7 CC2NP: Capture/Compare 2 output Polarity.
refer to CC1NP description
Bit 6

Reserved, must be kept at reset value.

Bit 5 CC2P: Capture/Compare 2 output Polarity.
refer to CC1P description
Bit 4 CC2E: Capture/Compare 2 output enable.
refer to CC1E description
Bit 3 CC1NP: Capture/Compare 1 output Polarity.
CC1 channel configured as output:
CC1NP must be kept cleared in this case.
CC1 channel configured as input:
This bit is used in conjunction with CC1P to define TI1FP1/TI2FP1 polarity. refer to CC1P
description.
Bit 2

Reserved, must be kept at reset value.

Bit 1 CC1P: Capture/Compare 1 output Polarity.
CC1 channel configured as output:
0: OC1 active high
1: OC1 active low
CC1 channel configured as input:
CC1NP/CC1P bits select TI1FP1 and TI2FP1 polarity for trigger or capture operations.
00: noninverted/rising edge
Circuit is sensitive to TIxFP1 rising edge (capture, trigger in reset, external clock or trigger
mode), TIxFP1 is not inverted (trigger in gated mode, encoder mode).
01: inverted/falling edge
Circuit is sensitive to TIxFP1 falling edge (capture, trigger in reset, external clock or trigger
mode), TIxFP1 is inverted (trigger in gated mode, encoder mode).
10: reserved, do not use this configuration.
11: noninverted/both edges
Circuit is sensitive to both TIxFP1 rising and falling edges (capture, trigger in reset, external
clock or trigger mode), TIxFP1 is not inverted (trigger in gated mode). This configuration
must not be used for encoder mode.
Bit 0 CC1E: Capture/Compare 1 output enable.
CC1 channel configured as output:
0: Off - OC1 is not active
1: On - OC1 signal is output on the corresponding output pin
CC1 channel configured as input:
This bit determines if a capture of the counter value can actually be done into the input
capture/compare register 1 (TIMx_CCR1) or not.
0: Capture disabled
1: Capture enabled

Table 99. Output control bit for standard OCx channels
CCxE bit

OCx output state

0

Output Disabled (OCx=0, OCx_EN=0)

1

OCx=OCxREF + Polarity, OCx_EN=1

RM0090 Rev 17

641/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Note:

The state of the external IO pins connected to the standard OCx channels depends on the
OCx channel state and the GPIO registers.

18.4.10

TIMx counter (TIMx_CNT)
Address offset: 0x24
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

CNT[31:16] (depending on timers)
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

CNT[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:16 CNT[31:16]: High counter value (on TIM2 and TIM5).
Bits 15:0 CNT[15:0]: Counter value

18.4.11

TIMx prescaler (TIMx_PSC)
Address offset: 0x28
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

PSC[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 15:0 PSC[15:0]: Prescaler value
The counter clock frequency CK_CNT is equal to fCK_PSC / (PSC[15:0] + 1).
PSC contains the value to be loaded in the active prescaler register at each update event
(including when the counter is cleared through UG bit of TIMx_EGR register or through
trigger controller when configured in “reset mode”).

18.4.12

TIMx auto-reload register (TIMx_ARR)
Address offset: 0x2C
Reset value: 0xFFFF FFFF

31

30

29

28

27

26

25

24

23

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

22

21

20

19

18

17

16

rw

rw

rw

rw

rw

rw

rw

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

ARR[31:16] (depending on timers)

ARR[15:0]
rw

642/1747

rw

rw

rw

rw

rw

rw

rw

rw

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)

Bits 31:16 ARR[31:16]: High auto-reload value (on TIM2 and TIM5).
Bits 15:0

18.4.13

ARR[15:0]: Auto-reload value
ARR is the value to be loaded in the actual auto-reload register.
Refer to the Section 18.3.1: Time-base unit for more details about ARR update and
behavior.
The counter is blocked while the auto-reload value is null.

TIMx capture/compare register 1 (TIMx_CCR1)
Address offset: 0x34
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

CCR1[31:16] (depending on timers)
rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

CCR1[15:0]
rw/ro

Bits 31:16 CCR1[31:16]: High Capture/Compare 1 value (on TIM2 and TIM5).
Bits 15:0 CCR1[15:0]: Low Capture/Compare 1 value
If channel CC1 is configured as output:
CCR1 is the value to be loaded in the actual capture/compare 1 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR1 register
(bit OC1PE). Else the preload value is copied in the active capture/compare 1 register when
an update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signaled on OC1 output.
If channel CC1is configured as input:
CCR1 is the counter value transferred by the last input capture 1 event (IC1). The
TIMx_CCR1 register is read-only and cannot be programmed.

18.4.14

TIMx capture/compare register 2 (TIMx_CCR2)
Address offset: 0x38
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

CCR2[31:16] (depending on timers)
rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

15

14

13

12

11

10

9

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

8

7

6

5

4

3

2

1

0

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

CCR2[15:0]
rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

RM0090 Rev 17

643/1747
649

General-purpose timers (TIM2 to TIM5)

RM0090

Bits 31:16 CCR2[31:16]: High Capture/Compare 2 value (on TIM2 and TIM5).
Bits 15:0 CCR2[15:0]: Low Capture/Compare 2 value
If channel CC2 is configured as output:
CCR2 is the value to be loaded in the actual capture/compare 2 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR register (bit
OC2PE). Else the preload value is copied in the active capture/compare 2 register when an
update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signalled on OC2 output.
If channel CC2 is configured as input:
CCR2 is the counter value transferred by the last input capture 2 event (IC2). The
TIMx_CCR2 register is read-only and cannot be programmed.

18.4.15

TIMx capture/compare register 3 (TIMx_CCR3)
Address offset: 0x3C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

CCR3[31:16] (depending on timers)
rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

CCR3[15:0]
rw/ro

Bits 31:16 CCR3[31:16]: High Capture/Compare 3 value (on TIM2 and TIM5).
Bits 15:0 CCR3[15:0]: Low Capture/Compare value
If channel CC3 is configured as output:
CCR3 is the value to be loaded in the actual capture/compare 3 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR register (bit
OC3PE). Else the preload value is copied in the active capture/compare 3 register when an
update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signaled on OC3 output.
If channel CC3 is configured as input:
CCR3 is the counter value transferred by the last input capture 3 event (IC3). The
TIMx_CCR3 register is read-only and cannot be programmed.

18.4.16

TIMx capture/compare register 4 (TIMx_CCR4)
Address offset: 0x40
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

CCR4[31:16] (depending on timers)
rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

15

14

13

12

11

10

9

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

8

7

6

5

4

3

2

1

0

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

CCR4[15:0]
rw/ro

rw/ro

644/1747

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)

Bits 31:16 CCR4[31:16]: High Capture/Compare 4 value (on TIM2 and TIM5).
Bits 15:0 CCR4[15:0]: Low Capture/Compare value
1.
if CC4 channel is configured as output (CC4S bits):
CCR4 is the value to be loaded in the actual capture/compare 4 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR register
(bit OC4PE). Else the preload value is copied in the active capture/compare 4 register
when an update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signalled on OC4 output.
2.
if CC4 channel is configured as input (CC4S bits in TIMx_CCMR4 register):
CCR4 is the counter value transferred by the last input capture 4 event (IC4). The
TIMx_CCR4 register is read-only and cannot be programmed.

18.4.17

TIMx DMA control register (TIMx_DCR)
Address offset: 0x48
Reset value: 0x0000

15

14

13

Reserved

12

11

10

9

8

DBL[4:0]
rw

rw

rw

rw

rw

7

6
Reserved

5

4

3

2

1

0

rw

rw

DBA[4:0]
rw

rw

rw

Bits 15:13 Reserved, must be kept at reset value.
Bits 12:8 DBL[4:0]: DMA burst length
This 5-bit vector defines the number of DMA transfers (the timer recognizes a burst transfer
when a read or a write access is done to the TIMx_DMAR address).
00000: 1 transfer,
00001: 2 transfers,
00010: 3 transfers,
...
10001: 18 transfers.
Bits 7:5 Reserved, must be kept at reset value.
Bits 4:0 DBA[4:0]: DMA base address
This 5-bit vector defines the base-address for DMA transfers (when read/write access are
done through the TIMx_DMAR address). DBA is defined as an offset starting from the
address of the TIMx_CR1 register.
Example:
00000: TIMx_CR1,
00001: TIMx_CR2,
00010: TIMx_SMCR,
...
Example: Let us consider the following transfer: DBL = 7 transfers & DBA = TIMx_CR1. In this
case the transfer is done to/from 7 registers starting from the TIMx_CR1 address.

RM0090 Rev 17

645/1747
649

General-purpose timers (TIM2 to TIM5)

18.4.18

RM0090

TIMx DMA address for full transfer (TIMx_DMAR)
Address offset: 0x4C
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

DMAB[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 15:0 DMAB[15:0]: DMA register for burst accesses
A read or write operation to the DMAR register accesses the register located at the address
(TIMx_CR1 address) + (DBA + DMA index) x 4
where TIMx_CR1 address is the address of the control register 1, DBA is the DMA base
address configured in TIMx_DCR register, DMA index is automatically controlled by the
DMA transfer, and ranges from 0 to DBL (DBL configured in TIMx_DCR).

Example of how to use the DMA burst feature
In this example the timer DMA burst feature is used to update the contents of the CCRx
registers (x = 2, 3, 4) with the DMA transferring half words into the CCRx registers.
This is done in the following steps:
1.

Note:

646/1747

Configure the corresponding DMA channel as follows:
–

DMA channel peripheral address is the DMAR register address

–

DMA channel memory address is the address of the buffer in the RAM containing
the data to be transferred by DMA into CCRx registers.

–

Number of data to transfer = 3 (See note below).

–

Circular mode disabled.

2.

Configure the DCR register by configuring the DBA and DBL bit fields as follows:
DBL = 3 transfers, DBA = 0xE.

3.

Enable the TIMx update DMA request (set the UDE bit in the DIER register).

4.

Enable TIMx

5.

Enable the DMA channel

This example is for the case where every CCRx register to be updated once. If every CCRx
register is to be updated twice for example, the number of data to transfer should be 6. Let's
take the example of a buffer in the RAM containing data1, data2, data3, data4, data5 and
data6. The data is transferred to the CCRx registers as follows: on the first update DMA
request, data1 is transferred to CCR2, data2 is transferred to CCR3, data3 is transferred to
CCR4 and on the second update DMA request, data4 is transferred to CCR2, data5 is
transferred to CCR3 and data6 is transferred to CCR4.

RM0090 Rev 17

RM0090

General-purpose timers (TIM2 to TIM5)

18.4.19

TIM2 option register (TIM2_OR)
Address offset: 0x50
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

ITR1_RMP

Reserved

rw

5

4

3

2

1

0

3

2

1

0

Reserved

rw

Bits 15:12 Reserved, must be kept at reset value.
Bits 11:10 ITR1_RMP: Internal trigger 1 remap
Set and cleared by software.
00: TIM8_TRGOUT
01: PTP trigger output is connected to TIM2_ITR1
10: OTG FS SOF is connected to the TIM2_ITR1 input
11: OTG HS SOF is connected to the TIM2_ITR1 input
Bits 9:0 Reserved, must be kept at reset value.

18.4.20

TIM5 option register (TIM5_OR)
Address offset: 0x50
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

TI4_RMP

Reserved

rw

rw

5

4

Reserved

Bits 15:8 Reserved, must be kept at reset value.
Bits 7:6 TI4_RMP: Timer Input 4 remap
Set and cleared by software.
00: TIM5 Channel4 is connected to the GPIO: Refer to the Alternate function mapping table
in the datasheets.
01: the LSI internal clock is connected to the TIM5_CH4 input for calibration purposes
10: the LSE internal clock is connected to the TIM5_CH4 input for calibration purposes
11: the RTC wakeup interrupt is connected to TIM5_CH4 input for calibration purposes.
Wakeup interrupt should be enabled.
Bits 5:0 Reserved, must be kept at reset value.

RM0090 Rev 17

647/1747
649

General-purpose timers (TIM2 to TIM5)

18.4.21

RM0090

TIMx register map
TIMx registers are mapped as described in the table below:

4

3

2

1

0

OPM

URS

UDIS

CEN

CCDS

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

UIE

0

0

0

0

0

0

0

0

0

0

0

OC3M
[2:0]

0

0

0

0

0

0

IC1
PSC
[1:0]

CC1S
[1:0]

0

0

0

0

0

0

CC3S
[1:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CC3NP

Reserved

CC3P

CC3E

CC2NP

Reserved

Reserved

CC3S
[1:0]

CC4E

IC3
PSC
[1:0]

CC4P

IC3F[3:0]

CC1S
[1:0]

Reserved

CC4S
[1:0]

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CC1E

0

OC1CE

CC4S
[1:0]
0

0

IC1F[3:0]
0

IC4
PSC
[1:0]

0

OC1M
[2:0]

0

0

CNT[31:16]
(TIM2 and TIM5 only, reserved on the other timers)

0

0

0

0

0

0

0

0

CNT[15:0]
0

0

0

0

0

0

0

0

0

0

PSC[15:0]

Reserved

Reset value

648/1747

0

CC2S
[1:0]

0

CC1P

TIMx_PSC

0

0

0

CC1NP

0x28

0

0

TIMx_CNT
Reset value

OC4M
[2:0]

0

0

CC4NP

Reserved

Reset value
0x24

0

IC4F[3:0]

Reserved

TIMx_CCER

0

0

IC2
PSC
[1:0]

0

CC2E

Reset value

Reset value

0x20

0

0

0

CC2P

Reserved

TIMx_CCMR2
Input Capture
mode

0

OC3CE

0
O24CE

Reset value

0x1C

0

IC2F[3:0]

Reserved

TIMx_CCMR2
Output Compare
mode

0

OC2FE

0

TIMx_CCMR1
Input Capture
mode

CC2S
[1:0]

OC4FE

Reset value

0x18

OC2M
[2:0]

OC4PE

Reserved

OC2PE

0
OC2CE

Reset value
TIMx_CCMR1
Output Compare
mode

0

UIF

Reserved

0

0

UG

TIF
0
TG

TIMx_EGR

CC1IE

CC1OF
0

0

CC1IF

CC2OF
0

0

0

CC1G

CC3OF
0

CC2IE

CC4OF
0

Reserved

Reset value

0

CC2IF

UDE
0

0

CC2G

CC1DE
0

0

OC1FE

CC2DE
0

0

OC3FE

CC3DE
0

Reserved

Reserved

CC4DE
0

Reserved

COMDE

TIMx_SR

0

CC3IE

0

0

CC3IF

0

SMS[2:0]

CC3G

0

TS[2:0]

OC1PE

0

Reserved

0

OC3PE

0

0

CC4IE

0

0

CC4IF

0

Reserved

0

CC4G

0

ETF[3:0]

5

DIR

6

MMS[2:0]

0

Reserved

0

Reset value

0x14

0

TIE

ECE

0

TDE

TIMx_DIER

ETPS
[1:0]

ETP

Reserved

Reset value

0x10

0

0

TIMx_SMCR

0x0C

0

0

MSM

Reset value
0x08

0

0

Reserved

0

Reserved

0

CMS
[1:0]

Reserved

TIMx_CR2

7

8

0

TI1S

Reset value
0x04

9

11

10

12

13

14

15

16

17

19

20

18

CKD
[1:0]

Reserved

ARPE

TIMx_CR1

0x00

21

22

23

24

25

26

27

28

29

Register

30

Offset

31

Table 100. TIM2 to TIM5 register map and reset values

0

RM0090 Rev 17

0

0

0

0

0

0

0

0

RM0090

General-purpose timers (TIM2 to TIM5)

1

1

1

1

1

1

1

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

0

0

0

0

0

6

7

8

11

9
1

1

1

1

1

1

1

1

1

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CCR2[15:0]
0

0

0

0

0

0

0

0

0

0

0

CCR3[15:0]
0

0

0

0

0

0

0

0

0

0

0

CCR4[15:0]
0

0

0

0

0

0

0

0

0

0

0

Reserved

TIMx_DCR

DBL[4:0]

Reserved

TIMx_DMAR

0

TIM2_OR

0

0

DBA[4:0]

Reserved
0

0

0

0

0

0

0

0

0

0

0

DMAB[15:0]
0

Reserved

0

0

0

Reserved

Reset value
TIM5_OR

0

Reserved

Reset value

0x50

1

CCR1[15:0]

Reset value

0x50

1

CCR4[31:16]
(TIM2 and TIM5 only, reserved on the other timers)

TIMx_CCR4

0x44

0x4C

10

12

13

14

16

17

15

1

CCR3[31:16]
(TIM2 and TIM5 only, reserved on the other timers)

TIMx_CCR3

Reset value

0x48

1

CCR2[31:16]
(TIM2 and TIM5 only, reserved on the other timers)

TIMx_CCR2

Reset value
0x40

1

CCR1[31:16]
(TIM2 and TIM5 only, reserved on the other timers)

TIMx_CCR1

Reset value
0x3C

1

ARR[15:0]

Reserved

Reset value
0x38

1

0

1

0x30

0x34

18

19

20

21

22

23

24

25

26

27

28
1

1

1

2

1

3

1

4

Reset value

29

ARR[31:16]
(TIM2 and TIM5 only, reserved on the other timers)

TIMx_ARR

5

0x2C

Register

30

Offset

31

Table 100. TIM2 to TIM5 register map and reset values (continued)

0

0
Reserved

Reserved

Reset value

0

0

0

0

0

ITR1_
RMP

0

Reserved

0
IT4_
RMP
0

Reserved

0

Refer to Section 2.3: Memory map for the register boundary addresses.

RM0090 Rev 17

649/1747
649

General-purpose timers (TIM9 to TIM14)

19

RM0090

General-purpose timers (TIM9 to TIM14)
This section applies to the whole STM32F4xx family, unless otherwise specified.

19.1

TIM9 to TIM14 introduction
The TIM9 to TIM14 general-purpose timers consist of a 16-bit auto-reload counter driven by
a programmable prescaler.
They may be used for a variety of purposes, including measuring the pulse lengths of input
signals (input capture) or generating output waveforms (output compare, PWM).
Pulse lengths and waveform periods can be modulated from a few microseconds to several
milliseconds using the timer prescaler and the RCC clock controller prescalers.
The TIM9 to TIM14 timers are completely independent, and do not share any resources.
They can be synchronized together as described in Section 19.3.12.

19.2

TIM9 to TIM14 main features

19.2.1

TIM9/TIM12 main features
The features of the TIM9 to TIM14 general-purpose timers include:

650/1747

•

16-bit auto-reload upcounter

•

16-bit programmable prescaler used to divide the counter clock frequency by any factor
between 1 and 65536 (can be changed “on the fly”)

•

Up to 2 independent channels for:
–

Input capture

–

Output compare

–

PWM generation (edge-aligned mode)

–

One-pulse mode output

•

Synchronization circuit to control the timer with external signals and to interconnect
several timers together

•

Interrupt generation on the following events:
–

Update: counter overflow, counter initialization (by software or internal trigger)

–

Trigger event (counter start, stop, initialization or count by internal trigger)

–

Input capture

–

Output compare

RM0090 Rev 17

RM0090

General-purpose timers (TIM9 to TIM14)
Figure 180. General-purpose timer block diagram (TIM9 and TIM12)

,QWHUQDOFORFN &.B,17
7ULJJHU
FRQWUROOHU
,75
,75
,75
,75

7*,

,75
75&
7,)B('

75*,

6ODYH
FRQWUROOHU 5HVHWHQDEOHXSFRXQW
PRGH

7,)3
7,)3

8

$XWRUHORDGUHJLVWHU
6WRS&OHDU

&.B36&

7,

7,0[B&+

7,

7,0[B&+

,QSXWILOWHU 
HGJHGHWHFWRU

,QSXWILOWHU 
HGJHGHWHFWRU

7,)3
7,)3

,&

36& &.B&17
SUHVFDOHU
&&, 8
3UHVFDOHU

75&
7,)3
7,)3
75&

,&36
&&,

,&

3UHVFDOHU



&17FRXQWHU
&&,

&DSWXUH&RPSDUHUHJLVWHU

2&5()

2XWSXW
FRQWURO

2&

7,0[B&+

&&,

8

,&36

8,
8

&DSWXUH&RPSDUHUHJLVWHU 2&5()

2XWSXW 2&
FRQWURO

7,0[B&+

1RWHV
5HJ

3UHORDGUHJLVWHUVWUDQVIHUUHG
WRDFWLYHUHJLVWHUVRQ8HYHQW
DFFRUGLQJWRFRQWUROELW
(YHQW
DLE

,QWHUUXSW

19.2.2

TIM10/TIM11 and TIM13/TIM14 main features
The features of general-purpose timers TIM10/TIM11 and TIM13/TIM14 include:
•

16-bit auto-reload upcounter

•

16-bit programmable prescaler used to divide the counter clock frequency by any factor
between 1 and 65536 (can be changed “on the fly”)

•

independent channel for:

•

–

Input capture

–

Output compare

–

PWM generation (edge-aligned mode)

–

One-pulse mode output

Interrupt generation on the following events:
–

Update: counter overflow, counter initialization (by software)

–

Input capture

–

Output compare

RM0090 Rev 17

651/1747
695

General-purpose timers (TIM9 to TIM14)

RM0090

Figure 181. General-purpose timer block diagram (TIM10/11/13/14)

)NTERNAL CLOCK #+?).4
4RIGGER
#ONTROLLER

5

%NABLE
COUNTER

!UTORELOAD REGISTER
3TOP #LEAR

#+?03#

03#
PRESCALER

#+?#.4

5)
5

#.4
COUNTER



##)
4)
4)-X?#(

)NPUT FILTER 
EDGE DETECTOR

4)&0 )#

##)

5

0RESCALER

)#03

#APTURE#OMPARE  REGISTER

/#2%&

OUTPUT /#
CONTROL

4)-X?#(

.OTES
2EG

0RELOAD REGISTERS TRANSFERRED
TO ACTIVE REGISTERS ON 5 EVENT
ACCORDING TO CONTROL BIT
EVENT
INTERRUPT  $-! OUTPUT

AIC

652/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM9 to TIM14)

19.3

TIM9 to TIM14 functional description

19.3.1

Time-base unit
The main block of the timer is a 16-bit counter with its related auto-reload register. The
counters counts up.
The counter clock can be divided by a prescaler.
The counter, the auto-reload register and the prescaler register can be written or read by
software. This is true even when the counter is running.
The time-base unit includes:
•

Counter register (TIMx_CNT)

•

Prescaler register (TIMx_PSC)

•

Auto-reload register (TIMx_ARR)

The auto-reload register is preloaded. Writing to or reading from the auto-reload register
accesses the preload register. The content of the preload register are transferred into the
shadow register permanently or at each update event (UEV), depending on the auto-reload
preload enable bit (ARPE) in TIMx_CR1 register. The update event is sent when the counter
reaches the overflow and if the UDIS bit equals 0 in the TIMx_CR1 register. It can also be
generated by software. The generation of the update event is described in details for each
configuration.
The counter is clocked by the prescaler output CK_CNT, which is enabled only when the
counter enable bit (CEN) in TIMx_CR1 register is set (refer also to the slave mode controller
description to get more details on counter enabling).
Note that the counter starts counting 1 clock cycle after setting the CEN bit in the TIMx_CR1
register.

Prescaler description
The prescaler can divide the counter clock frequency by any factor between 1 and 65536. It
is based on a 16-bit counter controlled through a 16-bit register (in the TIMx_PSC register).
It can be changed on the fly as this control register is buffered. The new prescaler ratio is
taken into account at the next update event.
Figure 182 and Figure 183 give some examples of the counter behavior when the prescaler
ratio is changed on the fly.

RM0090 Rev 17

653/1747
695

General-purpose timers (TIM9 to TIM14)

RM0090

Figure 182. Counter timing diagram with prescaler division change from 1 to 2

&.B36&

&(1
7LPHUFORFN

&.B&17


&RXQWHUUHJLVWHU

)

)

)

)$ )% )&









8SGDWHHYHQW 8(9

3UHVFDOHUFRQWUROUHJLVWHU





:ULWHDQHZYDOXHLQ7,0[B36&
3UHVFDOHUEXIIHU



3UHVFDOHUFRXQWHU





















069

Figure 183. Counter timing diagram with prescaler division change from 1 to 4

&.B36&
&(1

7LPHUFORFN

&.B&17


&RXQWHUUHJLVWHU

)

)

)

)$ )% )&





8SGDWHHYHQW 8(9

3UHVFDOHUFRQWUROUHJLVWHU





:ULWHDQHZYDOXHLQ7,0[B36&
3UHVFDOHUEXIIHU

3UHVFDOHUFRXQWHU






















069

654/1747

RM0090 Rev 17

RM0090

19.3.2

General-purpose timers (TIM9 to TIM14)

Counter modes
Upcounting mode
In upcounting mode, the counter counts from 0 to the auto-reload value (content of the
TIMx_ARR register), then restarts from 0 and generates a counter overflow event.
Setting the UG bit in the TIMx_EGR register (by software or by using the slave mode
controller on TIM9 and TIM12) also generates an update event.
The UEV event can be disabled by software by setting the UDIS bit in the TIMx_CR1
register. This is to avoid updating the shadow registers while writing new values in the
preload registers. Then no update event occurs until the UDIS bit has been written to 0.
However, the counter restarts from 0, as well as the counter of the prescaler (but the
prescale rate does not change). In addition, if the URS bit (update request selection) in
TIMx_CR1 register is set, setting the UG bit generates an update event UEV but without
setting the UIF flag (thus no interrupt is sent). This is to avoid generating both update and
capture interrupts when clearing the counter on the capture event.
When an update event occurs, all the registers are updated and the update flag (UIF bit in
TIMx_SR register) is set (depending on the URS bit):
•

The auto-reload shadow register is updated with the preload value (TIMx_ARR),

•

The buffer of the prescaler is reloaded with the preload value (content of the TIMx_PSC
register).

The following figures show some examples of the counter behavior for different clock
frequencies when TIMx_ARR=0x36.
Figure 184. Counter timing diagram, internal clock divided by 1
<ͺW^

EdͺE

dŝŵĞƌĐůŽĐŬс<ͺEd
ŽƵŶƚĞƌƌĞŐŝƐƚĞƌ

ϯϭ

ϯϮ

ϯϯ

ϯϰ

ϯϱ

ϯϲ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

ŽƵŶƚĞƌŽǀĞƌĨůŽǁ

hƉĚĂƚĞĞǀĞŶƚ;hsͿ

hƉĚĂƚĞŝŶƚĞƌƌƵƉƚĨůĂŐ;h/&Ϳ
069

RM0090 Rev 17

655/1747
695

General-purpose timers (TIM9 to TIM14)

RM0090

Figure 185. Counter timing diagram, internal clock divided by 2
&.B36&
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU















&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

Figure 186. Counter timing diagram, internal clock divided by 4
&.B36&
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU









&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

Figure 187. Counter timing diagram, internal clock divided by N
&.B36&

7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU

)





&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

656/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM9 to TIM14)
Figure 188. Counter timing diagram, update event when ARPE=0 (TIMx_ARR not
preloaded)
<ͺW^

E

dŝŵĞƌĐůŽĐŬс<ͺEd
ŽƵŶƚĞƌƌĞŐŝƐƚĞƌ

ϯϭ

ϯϯ

ϯϮ

ϯϰ

ϯϱ

ϯϲ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

ŽƵŶƚĞƌŽǀĞƌĨůŽǁ

hƉĚĂƚĞĞǀĞŶƚ;hsͿ

hƉĚĂƚĞŝŶƚĞƌƌƵƉƚĨůĂŐ;h/&Ϳ

ƵƚŽͲƌĞůŽĂĚƉƌĞůŽĂĚƌĞŐŝƐƚĞƌ

&&

ϯϲ

tƌŝƚĞĂŶĞǁǀĂůƵĞŝŶd/DdžͺZZ

069

Figure 189. Counter timing diagram, update event when ARPE=1 (TIMx_ARR
preloaded)
&.B36&

&(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU

)

)

)

)

) )

















&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ
 8,)
$XWRUHORDGSUHORDG
UHJLVWHU

)



$XWRUHORDGVKDGRZ
UHJLVWHU

)

:ULWHDQHZYDOXHLQ7,0[B$55

RM0090 Rev 17



069

657/1747
695

General-purpose timers (TIM9 to TIM14)

19.3.3

RM0090

Clock selection
The counter clock can be provided by the following clock sources:
•

Internal clock (CK_INT)

•

External clock mode1 (for TIM9 and TIM12): external input pin (TIx)

•

Internal trigger inputs (ITRx) (for TIM9 and TIM12): connecting the trigger output from
another timer. Refer to Using one timer as prescaler for another timer for more details.

Internal clock source (CK_INT)
The internal clock source is the default clock source for TIM10/TIM11 and TIM13/TIM14.
For TIM9 and TIM12, the internal clock source is selected when the slave mode controller is
disabled (SMS=’000’). The CEN bit in the TIMx_CR1 register and the UG bit in the
TIMx_EGR register are then used as control bits and can be changed only by software
(except for UG which remains cleared). As soon as the CEN bit is programmed to 1, the
prescaler is clocked by the internal clock CK_INT.
Figure 190 shows the behavior of the control circuit and of the upcounter in normal mode,
without prescaler.
Figure 190. Control circuit in normal mode, internal clock divided by 1
,QWHUQDOFORFN
&(1 &17B(1
8*

&17B,1,7

&RXQWHUFORFN &.B&17 &.B36&
&RXQWHUUHJLVWHU









 

 



  




069

External clock source mode 1(TIM9 and TIM12)
This mode is selected when SMS=’111’ in the TIMx_SMCR register. The counter can count
at each rising or falling edge on a selected input.

658/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM9 to TIM14)
Figure 191. TI2 external clock connection example
7,0[B60&5
76>@
RU 7,)
7,)
,75[
7,B('

7,

(GJH
GHWHFWRU

)LOWHU

7,)B5LVLQJ
7,)B)DOOLQJ

7,)3


7,)3



RU
RU

[[


75*,



&.B,17

,&)>@

&&3

7,0[B&&05

7,0[B&&(5

&.B36&

([WHUQDOFORFN
PRGH

LQWHUQDOFORFN

,QWHUQDOFORFN
PRGH

606>@
7,0[B60&5
069

For example, to configure the upcounter to count in response to a rising edge on the TI2
input, use the following procedure:

Note:

1.

Configure channel 2 to detect rising edges on the TI2 input by writing CC2S = ‘01’ in
the TIMx_CCMR1 register.

2.

Configure the input filter duration by writing the IC2F[3:0] bits in the TIMx_CCMR1
register (if no filter is needed, keep IC2F=’0000’).

3.

Select the rising edge polarity by writing CC2P=’0’ and CC2NP=’0’ in the TIMx_CCER
register.

4.

Configure the timer in external clock mode 1 by writing SMS=’111’ in the TIMx_SMCR
register.

5.

Select TI2 as the trigger input source by writing TS=’110’ in the TIMx_SMCR register.

6.

Enable the counter by writing CEN=’1’ in the TIMx_CR1 register.

The capture prescaler is not used for triggering, so no need to configure it.
When a rising edge occurs on TI2, the counter counts once and the TIF flag is set.
The delay between the rising edge on TI2 and the actual clock of the counter is due to the
resynchronization circuit on TI2 input.
Figure 192. Control circuit in external clock mode 1
7,
&17B(1
&RXQWHUFORFN &.B&17 &.B36&
&RXQWHUUHJLVWHU







7,)

:ULWH7,) 

RM0090 Rev 17

069

659/1747
695

General-purpose timers (TIM9 to TIM14)

19.3.4

RM0090

Capture/compare channels
Each Capture/Compare channel is built around a capture/compare register (including a
shadow register), a input stage for capture (with digital filter, multiplexing and prescaler) and
an output stage (with comparator and output control).
Figure 193 to Figure 195 give an overview of a capture/compare channel.
The input stage samples the corresponding TIx input to generate a filtered signal TIxF.
Then, an edge detector with polarity selection generates a signal (TIxFPx) which can be
used as trigger input by the slave mode controller or as the capture command. It is
prescaled before the capture register (ICxPS).
Figure 193. Capture/compare channel (example: channel 1 input stage)
7,)B('
7RWKHVODYHPRGHFRQWUROOHU
7,
I'76

7,)B5LVLQJ
)LOWHU
7,)
GRZQFRXQWHU

(GJH
GHWHFWRU

7,)B)DOOLQJ

 7,)3




7,)3

,&)>@
7,0[B&&05

&&3&&13
7,0[B&&(5
7,)B5LVLQJ

IURPFKDQQHO
7,)B)DOOLQJ

IURPFKDQQHO



75&
IURPVODYHPRGH
FRQWUROOHU

,&

,&36
'LYLGHU




&&6>@ ,&36>@

&&(

7,0[B&&05

7,0[B&&(5

069

The output stage generates an intermediate waveform which is then used for reference:
OCxRef (active high). The polarity acts at the end of the chain.

660/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM9 to TIM14)
Figure 194. Capture/compare channel 1 main circuit
$3%%XV

5HDG&&5/

UHDGBLQBSURJUHVV

&&6>@

ZULWHBLQBSURJUHVV

,QSXW
PRGH

6 ZULWH&&5+

&DSWXUHFRPSDUHSUHORDGUHJLVWHU

5

5
FRPSDUHBWUDQVIHU

FDSWXUHBWUDQVIHU
&&6>@


ORZ

5HDG&&5+ 6

KLJK



LIELW

0&8SHULSKHUDOLQWHUIDFH

2XWSXW
PRGH

&&(

&17!&&5
&RXQWHU

&&*

&&6>@

8(9
&RPSDUDWRU

&DSWXUH

&&6>@

2&3(

&DSWXUHFRPSDUHVKDGRZUHJLVWHU

,&36

ZULWH&&5/

IURPWLPH
EDVHXQLW

2&3(

7,0B&&05

&17 &&5

7,0B(*5

069

Figure 195. Output stage of capture/compare channel (channel 1)
7RWKHPDVWHU
PRGHFRQWUROOHU

&17!&&5

2XWSXW
PRGH
&17 &&5
FRQWUROOHU

2&B5()




2XWSXW
HQDEOH
FLUFXLW

2&

&&3
7,0[B&&(5

2&0>@

&&( 7,0[B&&(5

7,0[B&&05
DL

The capture/compare block is made of one preload register and one shadow register. Write
and read always access the preload register.
In capture mode, captures are actually done in the shadow register, which is copied into the
preload register.
In compare mode, the content of the preload register is copied into the shadow register
which is compared to the counter.

19.3.5

Input capture mode
In Input capture mode, the Capture/Compare Registers (TIMx_CCRx) are used to latch the
value of the counter after a transition detected by the corresponding ICx signal. When a

RM0090 Rev 17

661/1747
695

General-purpose timers (TIM9 to TIM14)

RM0090

capture occurs, the corresponding CCXIF flag (TIMx_SR register) is set and an interrupt or
a DMA request can be sent if they are enabled. If a capture occurs while the CCxIF flag was
already high, then the over-capture flag CCxOF (TIMx_SR register) is set. CCxIF can be
cleared by software by writing it to ‘0’ or by reading the captured data stored in the
TIMx_CCRx register. CCxOF is cleared when the user writes it to ‘0’.
The following example shows how to capture the counter value in TIMx_CCR1 when TI1
input rises. To do this, use the following procedure:
1.

Select the active input: TIMx_CCR1 must be linked to the TI1 input, so write the CC1S
bits to ‘01’ in the TIMx_CCMR1 register. As soon as CC1S becomes different from ‘00’,
the channel is configured in input mode and the TIMx_CCR1 register becomes readonly.

2.

Program the needed input filter duration with respect to the signal connected to the
timer (by programming the ICxF bits in the TIMx_CCMRx register if the input is one of
the TIx inputs). Let’s imagine that, when toggling, the input signal is not stable during at
must 5 internal clock cycles. We must program a filter duration longer than these 5
clock cycles. We can validate a transition on TI1 when 8 consecutive samples with the
new level have been detected (sampled at fDTS frequency). Then write IC1F bits to
‘0011’ in the TIMx_CCMR1 register.

3.

Select the edge of the active transition on the TI1 channel by programming CC1P and
CC1NP bits to ‘00’ in the TIMx_CCER register (rising edge in this case).

4.

Program the input prescaler. In our example, we wish the capture to be performed at
each valid transition, so the prescaler is disabled (write IC1PS bits to ‘00’ in the
TIMx_CCMR1 register).

5.

Enable capture from the counter into the capture register by setting the CC1E bit in the
TIMx_CCER register.

6.

If needed, enable the related interrupt request by setting the CC1IE bit in the
TIMx_DIER register.

When an input capture occurs:
•

The TIMx_CCR1 register gets the value of the counter on the active transition.

•

CC1IF flag is set (interrupt flag). CC1OF is also set if at least two consecutive captures
occurred whereas the flag was not cleared.

•

An interrupt is generated depending on the CC1IE bit.

In order to handle the overcapture, it is recommended to read the data before the
overcapture flag. This is to avoid missing an overcapture which could happen after reading
the flag and before reading the data.
Note:

662/1747

IC interrupt requests can be generated by software by setting the corresponding CCxG bit in
the TIMx_EGR register.

RM0090 Rev 17

RM0090

19.3.6

General-purpose timers (TIM9 to TIM14)

PWM input mode (only for TIM9/12)
This mode is a particular case of input capture mode. The procedure is the same except:
•

Two ICx signals are mapped on the same TIx input.

•

These 2 ICx signals are active on edges with opposite polarity.

•

One of the two TIxFP signals is selected as trigger input and the slave mode controller
is configured in reset mode.

For example, the user can measure the period (in TIMx_CCR1 register) and the duty cycle
(in TIMx_CCR2 register) of the PWM applied on TI1 using the following procedure
(depending on CK_INT frequency and prescaler value):
1.

Select the active input for TIMx_CCR1: write the CC1S bits to ‘01’ in the TIMx_CCMR1
register (TI1 selected).

2.

Select the active polarity for TI1FP1 (used both for capture in TIMx_CCR1 and counter
clear): program the CC1P and CC1NP bits to ‘00’ (active on rising edge).

3.

Select the active input for TIMx_CCR2: write the CC2S bits to ‘10’ in the TIMx_CCMR1
register (TI1 selected).

4.

Select the active polarity for TI1FP2 (used for capture in TIMx_CCR2): program the
CC2P and CC2NP bits to ‘11’ (active on falling edge).

5.

Select the valid trigger input: write the TS bits to ‘101’ in the TIMx_SMCR register
(TI1FP1 selected).

6.

Configure the slave mode controller in reset mode: write the SMS bits to ‘100’ in the
TIMx_SMCR register.

7.

Enable the captures: write the CC1E and CC2E bits to ‘1’ in the TIMx_CCER register.
Figure 196. PWM input mode timing
7,

7,0[B&17









7,0[B&&5



7,0[B&&5



,&FDSWXUH
,&FDSWXUH
UHVHWFRXQWHU





,&FDSWXUH
SX
O HVZLG WK
PHDVXUHPHQW



,&FDSWXUH
SHULRG
PHDVXUHPHQW
DL

1. The PWM input mode can be used only with the TIMx_CH1/TIMx_CH2 signals due to the fact that only
TI1FP1 and TI2FP2 are connected to the slave mode controller.

RM0090 Rev 17

663/1747
695

General-purpose timers (TIM9 to TIM14)

19.3.7

RM0090

Forced output mode
In output mode (CCxS bits = ‘00’ in the TIMx_CCMRx register), each output compare signal
(OCxREF and then OCx) can be forced to active or inactive level directly by software,
independently of any comparison between the output compare register and the counter.
To force an output compare signal (OCXREF/OCx) to its active level, the user just needs to
write ‘101’ in the OCxM bits in the corresponding TIMx_CCMRx register. Thus OCXREF is
forced high (OCxREF is always active high) and OCx get opposite value to CCxP polarity
bit.
For example: CCxP=’0’ (OCx active high) => OCx is forced to high level.
The OCxREF signal can be forced low by writing the OCxM bits to ‘100’ in the
TIMx_CCMRx register.
Anyway, the comparison between the TIMx_CCRx shadow register and the counter is still
performed and allows the flag to be set. Interrupt requests can be sent accordingly. This is
described in the output compare mode section below.

19.3.8

Output compare mode
This function is used to control an output waveform or indicating when a period of time has
elapsed.
When a match is found between the capture/compare register and the counter, the output
compare function:
1.

Assigns the corresponding output pin to a programmable value defined by the output
compare mode (OCxM bits in the TIMx_CCMRx register) and the output polarity (CCxP
bit in the TIMx_CCER register). The output pin can keep its level (OCXM=’000’), be set
active (OCxM=’001’), be set inactive (OCxM=’010’) or can toggle (OCxM=’011’) on
match.

2.

Sets a flag in the interrupt status register (CCxIF bit in the TIMx_SR register).

3.

Generates an interrupt if the corresponding interrupt mask is set (CCXIE bit in the
TIMx_DIER register).

The TIMx_CCRx registers can be programmed with or without preload registers using the
OCxPE bit in the TIMx_CCMRx register.
In output compare mode, the update event UEV has no effect on OCxREF and OCx output.
The timing resolution is one count of the counter. Output compare mode can also be used to
output a single pulse (in One-pulse mode).
Procedure:
1.

Select the counter clock (internal, external, prescaler).

2.

Write the desired data in the TIMx_ARR and TIMx_CCRx registers.

3.

Set the CCxIE bit if an interrupt request is to be generated.

4.

5.

664/1747

Select the output mode. For example:
–

Write OCxM = ‘011’ to toggle OCx output pin when CNT matches CCRx

–

Write OCxPE = ‘0’ to disable preload register

–

Write CCxP = ‘0’ to select active high polarity

–

Write CCxE = ‘1’ to enable the output

Enable the counter by setting the CEN bit in the TIMx_CR1 register.

RM0090 Rev 17

RM0090

General-purpose timers (TIM9 to TIM14)
The TIMx_CCRx register can be updated at any time by software to control the output
waveform, provided that the preload register is not enabled (OCxPE=’0’, else TIMx_CCRx
shadow register is updated only at the next update event UEV). An example is given in
Figure 197.
Figure 197. Output compare mode, toggle on OC1.
:ULWH%KLQWKH&&5UHJLVWHU

7,0B&17



$

7,0B&&5

%

%

%

%

$

2&5() 2&

0DWFKGHWHFWHGRQ&&5
,QWHUUXSWJHQHUDWHGLIHQDEOHG
069

19.3.9

PWM mode
Pulse Width Modulation mode allows the user to generate a signal with a frequency
determined by the value of the TIMx_ARR register and a duty cycle determined by the value
of the TIMx_CCRx register.
The PWM mode can be selected independently on each channel (one PWM per OCx
output) by writing ‘110’ (PWM mode 1) or ‘111’ (PWM mode 2) in the OCxM bits in the
TIMx_CCMRx register. Enable the corresponding preload register by setting the OCxPE bit
in the TIMx_CCMRx register, and eventually the auto-reload preload register by setting the
ARPE bit in the TIMx_CR1 register.
As the preload registers are transferred to the shadow registers only when an update event
occurs, before starting the counter, the user has to initialize all the registers by setting the
UG bit in the TIMx_EGR register.
The OCx polarity is software programmable using the CCxP bit in the TIMx_CCER register.
It can be programmed as active high or active low. The OCx output is enabled by the CCxE
bit in the TIMx_CCER register. Refer to the TIMx_CCERx register description for more
details.
In PWM mode (1 or 2), TIMx_CNT and TIMx_CCRx are always compared to determine
whether TIMx_CNT ≤TIMx_CCRx.
The timer is able to generate PWM in edge-aligned mode only since the counter is
upcounting.

PWM edge-aligned mode
In the following example, we consider PWM mode 1. The reference PWM signal OCxREF is
high as long as TIMx_CNT < TIMx_CCRx else it becomes low. If the compare value in

RM0090 Rev 17

665/1747
695

General-purpose timers (TIM9 to TIM14)

RM0090

TIMx_CCRx is greater than the auto-reload value (in TIMx_ARR) then OCxREF is held at
‘1’. If the compare value is 0 then OCxRef is held at ‘0’. Figure 198 shows some edgealigned PWM waveforms in an example where TIMx_ARR=8.
Figure 198. Edge-aligned PWM waveforms (ARR=8)



&RXQWHUUHJLVWHU

&&5[ 





















2&;5()
&&[,)

2&;5()
&&5[ 
&&[,)

2&;5()

µ¶

&&5[!
&&[,)

2&;5()

µ¶

&&5[ 
&&[,)

069

19.3.10

One-pulse mode
One-pulse mode (OPM) is a particular case of the previous modes. It allows the counter to
be started in response to a stimulus and to generate a pulse with a programmable length
after a programmable delay.
Starting the counter can be controlled through the slave mode controller. Generating the
waveform can be done in output compare mode or PWM mode. Select One-pulse mode by
setting the OPM bit in the TIMx_CR1 register. This makes the counter stop automatically at
the next update event UEV.
A pulse can be correctly generated only if the compare value is different from the counter
initial value. Before starting (when the timer is waiting for the trigger), the configuration must
be as follows:
CNT < CCRx≤ ARR (in particular, 0 < CCRx)

666/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM9 to TIM14)
Figure 199. Example of one pulse mode.
7,
2&5()
2&

&RXQWHU

7,0B$55
7,0B&&5


W'(/$<

W38/6(

W

069

For example the user may want to generate a positive pulse on OC1 with a length of tPULSE
and after a delay of tDELAY as soon as a positive edge is detected on the TI2 input pin.
Use TI2FP2 as trigger 1:
1.

Map TI2FP2 to TI2 by writing CC2S=’01’ in the TIMx_CCMR1 register.

2.

TI2FP2 must detect a rising edge, write CC2P=’0’ and CC2NP = ‘0’ in the TIMx_CCER
register.

3.

Configure TI2FP2 as trigger for the slave mode controller (TRGI) by writing TS=’110’ in
the TIMx_SMCR register.

4.

TI2FP2 is used to start the counter by writing SMS to ‘110’ in the TIMx_SMCR register
(trigger mode).

The OPM waveform is defined by writing the compare registers (taking into account the
clock frequency and the counter prescaler).
•

The tDELAY is defined by the value written in the TIMx_CCR1 register.

•

The tPULSE is defined by the difference between the auto-reload value and the compare
value (TIMx_ARR - TIMx_CCR1).

•

Let us say the user wants to build a waveform with a transition from ‘0’ to ‘1’ when a
compare match occurs and a transition from ‘1’ to ‘0’ when the counter reaches the
auto-reload value. To do this enable PWM mode 2 by writing OC1M=’111’ in the
TIMx_CCMR1 register. The user can optionally enable the preload registers by writing
OC1PE=’1’ in the TIMx_CCMR1 register and ARPE in the TIMx_CR1 register. In this
case the user has to write the compare value in the TIMx_CCR1 register, the autoreload value in the TIMx_ARR register, generate an update by setting the UG bit and
wait for external trigger event on TI2. CC1P is written to ‘0’ in this example.

The user only wants one pulse (Single mode), so write '1 in the OPM bit in the TIMx_CR1
register to stop the counter at the next update event (when the counter rolls over from the
auto-reload value back to 0). When OPM bit in the TIMx_CR1 register is set to '0', so the
Repetitive Mode is selected.

RM0090 Rev 17

667/1747
695

General-purpose timers (TIM9 to TIM14)

RM0090

Particular case: OCx fast enable
In One-pulse mode, the edge detection on TIx input set the CEN bit which enables the
counter. Then the comparison between the counter and the compare value makes the
output toggle. But several clock cycles are needed for these operations and it limits the
minimum delay tDELAY min we can get.
If the user wants to output a waveform with the minimum delay, set the OCxFE bit in the
TIMx_CCMRx register. Then OCxRef (and OCx) are forced in response to the stimulus,
without taking in account the comparison. Its new level is the same as if a compare match
had occurred. OCxFE acts only if the channel is configured in PWM1 or PWM2 mode.

19.3.11

TIM9/12 external trigger synchronization
The TIM9/12 timers can be synchronized with an external trigger in several modes: Reset
mode, Gated mode and Trigger mode.

Slave mode: Reset mode
The counter and its prescaler can be reinitialized in response to an event on a trigger input.
Moreover, if the URS bit from the TIMx_CR1 register is low, an update event UEV is
generated. Then all the preloaded registers (TIMx_ARR, TIMx_CCRx) are updated.
In the following example, the upcounter is cleared in response to a rising edge on TI1 input:
1.

Configure the channel 1 to detect rising edges on TI1. Configure the input filter duration
(in this example, we don’t need any filter, so we keep IC1F=’0000’). The capture
prescaler is not used for triggering, so there’s no need to configure it. The CC1S bits
select the input capture source only, CC1S = ‘01’ in the TIMx_CCMR1 register.
Program CC1P and CC1NP to ‘00’ in TIMx_CCER register to validate the polarity (and
detect rising edges only).

2.

Configure the timer in reset mode by writing SMS=’100’ in TIMx_SMCR register. Select
TI1 as the input source by writing TS=’101’ in TIMx_SMCR register.

3.

Start the counter by writing CEN=’1’ in the TIMx_CR1 register.

The counter starts counting on the internal clock, then behaves normally until TI1 rising
edge. When TI1 rises, the counter is cleared and restarts from 0. In the meantime, the
trigger flag is set (TIF bit in the TIMx_SR register) and an interrupt request can be sent if
enabled (depending on the TIE bit in TIMx_DIER register).
The following figure shows this behavior when the auto-reload register TIMx_ARR=0x36.
The delay between the rising edge on TI1 and the actual reset of the counter is due to the
resynchronization circuit on TI1 input.

668/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM9 to TIM14)
Figure 200. Control circuit in reset mode
7,
8*
&RXQWHUFORFN FNBFQW FNBSVF
&RXQWHUUHJLVWHU

          

   

7,)

069

Slave mode: Gated mode
The counter can be enabled depending on the level of a selected input.
In the following example, the upcounter counts only when TI1 input is low:
1.

Configure the channel 1 to detect low levels on TI1. Configure the input filter duration
(in this example, we don’t need any filter, so we keep IC1F=’0000’). The capture
prescaler is not used for triggering, so there’s no need to configure it. The CC1S bits
select the input capture source only, CC1S=’01’ in TIMx_CCMR1 register. Program
CC1P=’1’ and CC1NP= ‘0’ in TIMx_CCER register to validate the polarity (and detect
low level only).

2.

Configure the timer in gated mode by writing SMS=’101’ in TIMx_SMCR register.
Select TI1 as the input source by writing TS=’101’ in TIMx_SMCR register.

3.

Enable the counter by writing CEN=’1’ in the TIMx_CR1 register (in gated mode, the
counter doesn’t start if CEN=’0’, whatever is the trigger input level).

The counter starts counting on the internal clock as long as TI1 is low and stops as soon as
TI1 becomes high. The TIF flag in the TIMx_SR register is set both when the counter starts
or stops.
The delay between the rising edge on TI1 and the actual stop of the counter is due to the
resynchronization circuit on TI1 input.

RM0090 Rev 17

669/1747
695

General-purpose timers (TIM9 to TIM14)

RM0090

Figure 201. Control circuit in gated mode

7,
FQWBHQ
&RXQWHUFORFN FNBFQW FNBSVF
&RXQWHUUHJLVWHU

   



   

7,)

:ULWH7,) 
069

Slave mode: Trigger mode
The counter can start in response to an event on a selected input.
In the following example, the upcounter starts in response to a rising edge on TI2 input:
1.

Configure the channel 2 to detect rising edges on TI2. Configure the input filter duration
(in this example, we don’t need any filter, so we keep IC2F=’0000’). The capture
prescaler is not used for triggering, so there’s no need to configure it. The CC2S bits
are configured to select the input capture source only, CC2S=’01’ in TIMx_CCMR1
register. Program CC2P=’1’ and CC2NP=’0’ in TIMx_CCER register to validate the
polarity (and detect low level only).

2.

Configure the timer in trigger mode by writing SMS=’110’ in TIMx_SMCR register.
Select TI2 as the input source by writing TS=’110’ in TIMx_SMCR register.

When a rising edge occurs on TI2, the counter starts counting on the internal clock and the
TIF flag is set.
The delay between the rising edge on TI2 and the actual start of the counter is due to the
resynchronization circuit on TI2 input.
Figure 202. Control circuit in trigger mode
7,
FQWBHQ
&RXQWHUFORFN FNBFQW FNBSVF
&RXQWHUUHJLVWHU





  

7,)
069

670/1747

RM0090 Rev 17

RM0090

19.3.12

General-purpose timers (TIM9 to TIM14)

Timer synchronization (TIM9/12)
The TIM timers are linked together internally for timer synchronization or chaining. Refer to
Section 18.3.15: Timer synchronization for details.

Note:

The clock of the slave timer must be enabled prior to receive events from the master timer,
and must not be changed on-the-fly while triggers are received from the master timer.

19.3.13

Debug mode
When the microcontroller enters debug mode (Cortex®-M4 with FPU core halted), the TIMx
counter either continues to work normally or stops, depending on DBG_TIMx_STOP
configuration bit in DBG module. For more details, refer to Section 38.16.2: Debug support
for timers, watchdog, bxCAN and I2C.

RM0090 Rev 17

671/1747
695

General-purpose timers (TIM9 to TIM14)

19.4

RM0090

TIM9 and TIM12 registers
Refer to Section 2.2 for a list of abbreviations used in register descriptions.
The peripheral registers have to be written by half-words (16 bits) or words (32 bits). Read
accesses can be done by bytes (8 bits), half-words (16 bits) or words (32 bits).

19.4.1

TIM9/12 control register 1 (TIMx_CR1)
Address offset: 0x00
Reset value: 0x0000

15

14

13

12

Reserved

11

10

9

8

CKD[1:0]
rw

7
ARPE

rw

rw

6

5
Reserved

4

3

2

1

0

OPM

URS

UDIS

CEN

rw

rw

rw

rw

Bits 15:10 Reserved, must be kept at reset value.
Bits 9:8 CKD: Clock division
This bit-field indicates the division ratio between the timer clock (CK_INT) frequency and
sampling clock used by the digital filters (TIx),
00: tDTS = tCK_INT
01: tDTS = 2 × tCK_INT
10: tDTS = 4 × tCK_INT
11: Reserved
Bit 7 ARPE: Auto-reload preload enable
0: TIMx_ARR register is not buffered.
1: TIMx_ARR register is buffered.
Bits 6:4 Reserved, must be kept at reset value.
Bit 3 OPM: One-pulse mode
0: Counter is not stopped on the update event
1: Counter stops counting on the next update event (clearing the CEN bit).
Bit 2 URS: Update request source
This bit is set and cleared by software to select the UEV event sources.
0: Any of the following events generates an update interrupt if enabled:
–
Counter overflow
–
Setting the UG bit
1: Only counter overflow generates an update interrupt if enabled.
Bit 1 UDIS: Update disable
This bit is set and cleared by software to enable/disable update event (UEV) generation.
0: UEV enabled. An UEV is generated by one of the following events:
–
Counter overflow
–
Setting the UG bit
Buffered registers are then loaded with their preload values.
1: UEV disabled. No UEV is generated, shadow registers keep their value (ARR, PSC,
CCRx). The counter and the prescaler are reinitialized if the UG bit is set.
Bit 0 CEN: Counter enable
0: Counter disabled
1: Counter enabled
CEN is cleared automatically in one-pulse mode, when an update event occurs.

672/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM9 to TIM14)

19.4.2

TIM9/12 slave mode control register (TIMx_SMCR)
Address offset: 0x08
Reset value: 0x0000

15

14

13

12

11

Reserved

10

9

8

7

6

MSM
rw

5

4

TS[2:0]
rw

rw

rw

3
Res.

2

1

0

SMS[2:0]
rw

rw

rw

Bits 15:8 Reserved, must be kept at reset value.
Bit 7 MSM: Master/Slave mode
0: No action
1: The effect of an event on the trigger input (TRGI) is delayed to allow a perfect
synchronization between the current timer and its slaves (through TRGO). It is useful in
order to synchronize several timers on a single external event.
Bits 6:4 TS: Trigger selection
This bit field selects the trigger input to be used to synchronize the counter.
000: Internal Trigger 0 (ITR0)
001: Internal Trigger 1 (ITR1)
010: Internal Trigger 2 (ITR2)
011: Internal Trigger 3 (ITR3)
100: TI1 Edge Detector (TI1F_ED)
101: Filtered Timer Input 1 (TI1FP1)
110: Filtered Timer Input 2 (TI2FP2)
111: Reserved.
See Table 101 for more details on the meaning of ITRx for each timer.
Note: These bits must be changed only when they are not used (e.g. when SMS=’000’) to
avoid wrong edge detections at the transition.

RM0090 Rev 17

673/1747
695

General-purpose timers (TIM9 to TIM14)

RM0090

Bit 3 Reserved, must be kept at reset value.
Bits 2:0 SMS: Slave mode selection
When external signals are selected, the active edge of the trigger signal (TRGI) is linked to
the polarity selected on the external input (see Input control register and Control register
descriptions.
000: Slave mode disabled - if CEN = 1 then the prescaler is clocked directly by the internal
clock
001: Reserved
010: Reserved
011: Reserved
100: Reset mode - Rising edge of the selected trigger input (TRGI) reinitializes the counter
and generates an update of the registers
101: Gated mode - The counter clock is enabled when the trigger input (TRGI) is high. The
counter stops (but is not reset) as soon as the trigger becomes low. Counter starts and stops
are both controlled
110: Trigger mode - The counter starts on a rising edge of the trigger TRGI (but it is not
reset). Only the start of the counter is controlled
111: External clock mode 1 - Rising edges of the selected trigger (TRGI) clock the counter
Note: The Gated mode must not be used if TI1F_ED is selected as the trigger input
(TS=’100’). Indeed, TI1F_ED outputs 1 pulse for each transition on TI1F, whereas the
Gated mode checks the level of the trigger signal.
Note: The clock of the slave timer must be enabled prior to receive events from the master
timer, and must not be changed on-the-fly while triggers are received from the master
timer.

Table 101. TIMx internal trigger connection

19.4.3

Slave TIM

ITR0 (TS = 000)

ITR1 (TS = 001)

ITR2 (TS = 010)

ITR3 (TS = 011)

TIM9

TIM2_TRGO

TIM3_TRGO

TIM10_OC

TIM11_OC

TIM12

TIM4_TRGO

TIM5_TRGO

TIM13_OC

TIM14_OC

TIM9/12 Interrupt enable register (TIMx_DIER)
Address offset: 0x0C
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

TIE

Reserved

Bits 15:7

rw

Reserved, must be kept at reset value.

Bit 6 TIE: Trigger interrupt enable
0: Trigger interrupt disabled.
1: Trigger interrupt enabled.
Bit 5:3

674/1747

6

Reserved, must be kept at reset value.

RM0090 Rev 17

5

4
Res

3

2

1

0

CC2IE

CC1IE

UIE

rw

rw

rw

RM0090

General-purpose timers (TIM9 to TIM14)

Bit 2 CC2IE: Capture/Compare 2 interrupt enable
0: CC2 interrupt disabled.
1: CC2 interrupt enabled.
Bit 1 CC1IE: Capture/Compare 1 interrupt enable
0: CC1 interrupt disabled.
1: CC1 interrupt enabled.
Bit 0 UIE: Update interrupt enable
0: Update interrupt disabled.
1: Update interrupt enabled.

RM0090 Rev 17

675/1747
695

General-purpose timers (TIM9 to TIM14)

19.4.4

RM0090

TIM9/12 status register (TIMx_SR)
Address offset: 0x10
Reset value: 0x0000

15

14

13
Reserved

Bits 15:11

12

11

10

9

8

CC2OF CC1OF
rc_w0

rc_w0

7

Reserved

6
TIF
rc_w0

5

4
Reserved

3

2

1

0

CC2IF

CC1IF

UIF

rc_w0

rc_w0

rc_w0

Reserved, must be kept at reset value.

Bit 10 CC2OF: Capture/compare 2 overcapture flag
refer to CC1OF description
Bit 9 CC1OF: Capture/Compare 1 overcapture flag
This flag is set by hardware only when the corresponding channel is configured in input
capture mode. It is cleared by software by writing it to ‘0’.
0: No overcapture has been detected.
1: The counter value has been captured in TIMx_CCR1 register while CC1IF flag was
already set
Bits 8:7

Reserved, must be kept at reset value.

Bit 6 TIF: Trigger interrupt flag
This flag is set by hardware on trigger event (active edge detected on TRGI input when the
slave mode controller is enabled in all modes but gated mode. It is set when the counter
starts or stops when gated mode is selected. It is cleared by software.
0: No trigger event occurred.
1: Trigger interrupt pending.
Bits 5:3

Reserved, must be kept at reset value.

Bit 2 CC2IF: Capture/Compare 2 interrupt flag
refer to CC1IF description
Bit 1 CC1IF: Capture/compare 1 interrupt flag
If channel CC1 is configured as output:
This flag is set by hardware when the counter matches the compare value. It is cleared by
software.
0: No match.
1: The content of the counter TIMx_CNT matches the content of the TIMx_CCR1 register.
When the contents of TIMx_CCR1 are greater than the contents of TIMx_ARR, the CC1IF
bit goes high on the counter overflow.
If channel CC1 is configured as input:
This bit is set by hardware on a capture. It is cleared by software or by reading the
TIMx_CCR1 register.
0: No input capture occurred.
1: The counter value has been captured in TIMx_CCR1 register (an edge has been detected
on IC1 which matches the selected polarity).

676/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM9 to TIM14)

Bit 0 UIF: Update interrupt flag
This bit is set by hardware on an update event. It is cleared by software.
0: No update occurred.
1: Update interrupt pending. This bit is set by hardware when the registers are updated:
– At overflow and if UDIS=’0’ in the TIMx_CR1 register.
– When CNT is reinitialized by software using the UG bit in TIMx_EGR register, if URS=’0’ and
UDIS=’0’ in the TIMx_CR1 register.
– When CNT is reinitialized by a trigger event (refer to the synchro control register
description), if URS=’0’ and UDIS=’0’ in the TIMx_CR1 register.

19.4.5

TIM9/12 event generation register (TIMx_EGR)
Address offset: 0x14
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6
TG

Reserved

w

5

4
Reserved

3

2

1

0

CC2G

CC1G

UG

w

w

w

Bits 15:7 Reserved, must be kept at reset value.
Bit 6 TG: Trigger generation
This bit is set by software in order to generate an event, it is automatically cleared by
hardware.
0: No action
1: The TIF flag is set in the TIMx_SR register. Related interrupt can occur if enabled
Bits 5:3 Reserved, must be kept at reset value.
Bit 2 CC2G: Capture/compare 2 generation
refer to CC1G description
Bit 1 CC1G: Capture/compare 1 generation
This bit is set by software to generate an event, it is automatically cleared by hardware.
0: No action
1: A capture/compare event is generated on channel 1:
If channel CC1 is configured as output:
the CC1IF flag is set, the corresponding interrupt is sent if enabled.
If channel CC1 is configured as input:
The current counter value is captured in the TIMx_CCR1 register. The CC1IF flag is set, the
corresponding interrupt is sent if enabled. The CC1OF flag is set if the CC1IF flag was
already high.
Bit 0 UG: Update generation
This bit can be set by software, it is automatically cleared by hardware.
0: No action
1: Re-initializes the counter and generates an update of the registers. The prescaler counter
is also cleared and the prescaler ratio is not affected. The counter is cleared.

RM0090 Rev 17

677/1747
695

General-purpose timers (TIM9 to TIM14)

19.4.6

RM0090

TIM9/12 capture/compare mode register 1 (TIMx_CCMR1)
Address offset: 0x18
Reset value: 0x0000
The channels can be used in input (capture mode) or in output (compare mode). The
direction of a channel is defined by configuring the corresponding CCxS bits. All the other
bits in this register have different functions in input and output modes. For a given bit, OCxx
describes its function when the channel is configured in output mode, ICxx describes its
function when the channel is configured in input mode. Take care that the same bit can have
different meanings for the input stage and the output stage.

15

14

Res.

13

12

OC2M[2:0]
IC2F[3:0]

rw

rw

rw

11

10

OC2PE OC2FE
IC2PSC[1:0]
rw

rw

rw

9

8

CC2S[1:0]
rw

rw

7

6

Res.

5

4

OC1M[2:0]
IC1F[3:0]

rw

rw

rw

3

2

OC1PE OC1FE
IC1PSC[1:0]
rw

rw

rw

1

0

CC1S[1:0]
rw

rw

Output compare mode
Bit 15

Reserved, must be kept at reset value.

Bits 14:12 OC2M[2:0]: Output compare 2 mode
Bit 11 OC2PE: Output compare 2 preload enable
Bit 10 OC2FE: Output compare 2 fast enable
Bits 9:8 CC2S[1:0]: Capture/Compare 2 selection
This bitfield defines the direction of the channel (input/output) as well as the used input.
00: CC2 channel is configured as output
01: CC2 channel is configured as input, IC2 is mapped on TI2
10: CC2 channel is configured as input, IC2 is mapped on TI1
11: CC2 channel is configured as input, IC2 is mapped on TRC. This mode works only if an
internal trigger input is selected through the TS bit (TIMx_SMCR register
Note: The CC2S bits are writable only when the channel is OFF (CC2E = 0 in TIMx_CCER).
Bit 7

678/1747

Reserved, must be kept at reset value.

RM0090 Rev 17

RM0090

General-purpose timers (TIM9 to TIM14)

Bits 6:4 OC1M: Output compare 1 mode
These bits define the behavior of the output reference signal OC1REF from which OC1 and
OC1N are derived. OC1REF is active high whereas the active levels of OC1 and OC1N
depend on the CC1P and CC1NP bits, respectively.
000: Frozen - The comparison between the output compare register TIMx_CCR1 and the
counter TIMx_CNT has no effect on the outputs.(this mode is used to generate a timing
base).
001: Set channel 1 to active level on match. The OC1REF signal is forced high when the
TIMx_CNT counter matches the capture/compare register 1 (TIMx_CCR1).
010: Set channel 1 to inactive level on match. The OC1REF signal is forced low when the
TIMx_CNT counter matches the capture/compare register 1 (TIMx_CCR1).
011: Toggle - OC1REF toggles when TIMx_CNT=TIMx_CCR1
100: Force inactive level - OC1REF is forced low
101: Force active level - OC1REF is forced high
110: PWM mode 1 - In upcounting, channel 1 is active as long as TIMx_CNTTIMx_CCR1, else it is active (OC1REF=’1’)
111: PWM mode 2 - In upcounting, channel 1 is inactive as long as TIMx_CNTTIMx_CCR1
else it is inactive.
Note: In PWM mode 1 or 2, the OCREF level changes only when the result of the
comparison changes or when the output compare mode switches from “frozen” mode
to “PWM” mode.
Bit 3 OC1PE: Output compare 1 preload enable
0: Preload register on TIMx_CCR1 disabled. TIMx_CCR1 can be written at anytime, the
new value is taken into account immediately
1: Preload register on TIMx_CCR1 enabled. Read/Write operations access the preload
register. TIMx_CCR1 preload value is loaded into the active register at each update event
Note: The PWM mode can be used without validating the preload register only in one-pulse
mode (OPM bit set in the TIMx_CR1 register). Else the behavior is not guaranteed.
Bit 2 OC1FE: Output compare 1 fast enable
This bit is used to accelerate the effect of an event on the trigger in input on the CC output.
0: CC1 behaves normally depending on the counter and CCR1 values even when the
trigger is ON. The minimum delay to activate the CC1 output when an edge occurs on the
trigger input is 5 clock cycles
1: An active edge on the trigger input acts like a compare match on the CC1 output. Then,
OC is set to the compare level independently of the result of the comparison. Delay to
sample the trigger input and to activate CC1 output is reduced to 3 clock cycles. OC1FE
acts only if the channel is configured in PWM1 or PWM2 mode.
Bits 1:0 CC1S: Capture/Compare 1 selection
This bitfield defines the direction of the channel (input/output) as well as the used input.
00: CC1 channel is configured as output
01: CC1 channel is configured as input, IC1 is mapped on TI1
10: CC1 channel is configured as input, IC1 is mapped on TI2
11: CC1 channel is configured as input, IC1 is mapped on TRC. This mode works only if an
internal trigger input is selected through the TS bit (TIMx_SMCR register)
Note: The CC1S bits are writable only when the channel is OFF (CC1E = 0 in TIMx_CCER).

RM0090 Rev 17

679/1747
695

General-purpose timers (TIM9 to TIM14)

RM0090

Input capture mode
Bits 15:12 IC2F: Input capture 2 filter
Bits 11:10 IC2PSC[1:0]: Input capture 2 prescaler
Bits 9:8 CC2S: Capture/compare 2 selection
This bitfield defines the direction of the channel (input/output) as well as the used input.
00: CC2 channel is configured as output
01: CC2 channel is configured as input, IC2 is mapped on TI2
10: CC2 channel is configured as input, IC2 is mapped on TI1
11: CC2 channel is configured as input, IC2 is mapped on TRC. This mode works only if an
internal trigger input is selected through the TS bit (TIMx_SMCR register)
Note: The CC2S bits are writable only when the channel is OFF (CC2E = 0 in TIMx_CCER).
Bits 7:4 IC1F: Input capture 1 filter
This bitfield defines the frequency used to sample the TI1 input and the length of the digital
filter applied to TI1. The digital filter is made of an event counter in which N consecutive
events are needed to validate a transition on the output:
0000: No filter, sampling is done at fDTS
0001: fSAMPLING=fCK_INT, N=2
0010: fSAMPLING=fCK_INT, N=4
0011: fSAMPLING=fCK_INT, N=8
0100: fSAMPLING=fDTS/2, N=6
0101: fSAMPLING=fDTS/2, N=8
0110: fSAMPLING=fDTS/4, N=6
0111: fSAMPLING=fDTS/4, N=8
1000: fSAMPLING=fDTS/8, N=6
1001: fSAMPLING=fDTS/8, N=8
1010: fSAMPLING=fDTS/16, N=5
1011: fSAMPLING=fDTS/16, N=6
1100: fSAMPLING=fDTS/16, N=8
1101: fSAMPLING=fDTS/32, N=5
1110: fSAMPLING=fDTS/32, N=6
1111: fSAMPLING=fDTS/32, N=8
Bits 3:2 IC1PSC: Input capture 1 prescaler
This bitfield defines the ratio of the prescaler acting on the CC1 input (IC1).
The prescaler is reset as soon as CC1E=’0’ (TIMx_CCER register).
00: no prescaler, capture is done each time an edge is detected on the capture input
01: capture is done once every 2 events
10: capture is done once every 4 events
11: capture is done once every 8 events
Bits 1:0 CC1S: Capture/Compare 1 selection
This bitfield defines the direction of the channel (input/output) as well as the used input.
00: CC1 channel is configured as output
01: CC1 channel is configured as input, IC1 is mapped on TI1
10: CC1 channel is configured as input, IC1 is mapped on TI2
11: CC1 channel is configured as input, IC1 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: The CC1S bits are writable only when the channel is OFF (CC1E = 0 in TIMx_CCER).

680/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM9 to TIM14)

19.4.7

TIM9/12 capture/compare enable register (TIMx_CCER)
Address offset: 0x20
Reset value: 0x0000

15

14

13

12

11

Reserved

10

9

8

7
CC2NP
rw

6
Res.

5

4

3

CC2P

CC2E

CC1NP

rw

rw

rw

2
Res.

1

0

CC1P

CC1E

rw

rw

Bits 15:8 Reserved, must be kept at reset value.
Bit 7 CC2NP: Capture/Compare 2 output Polarity
refer to CC1NP description
Bits 6 Reserved, must be kept at reset value.
Bit 5 CC2P: Capture/Compare 2 output Polarity
refer to CC1P description
Bit 4 CC2E: Capture/Compare 2 output enable
refer to CC1E description
Bit 3 CC1NP: Capture/Compare 1 complementary output Polarity
CC1 channel configured as output: CC1NP must be kept cleared
CC1 channel configured as input: CC1NP is used in conjunction with CC1P to define
TI1FP1/TI2FP1 polarity (refer to CC1P description).
Bit 2 Reserved, must be kept at reset value.
Bit 1 CC1P: Capture/Compare 1 output Polarity.
CC1 channel configured as output:
0: OC1 active high.
1: OC1 active low.
CC1 channel configured as input:
CC1NP/CC1P bits select TI1FP1 and TI2FP1 polarity for trigger or capture operations.
00: noninverted/rising edge
Circuit is sensitive to TIxFP1 rising edge (capture, trigger in reset, external clock or trigger
mode), TIxFP1 is not inverted (trigger in gated mode, encoder mode).
01: inverted/falling edge
Circuit is sensitive to TIxFP1 falling edge (capture, trigger in reset, external clock or trigger
mode), TIxFP1 is inverted (trigger in gated mode, encoder mode).
10: reserved, do not use this configuration.
Note: 11: noninverted/both edges
Circuit is sensitive to both TIxFP1 rising and falling edges (capture, trigger in reset,
external clock or trigger mode), TIxFP1 is not inverted (trigger in gated mode). This
configuration must not be used for encoder mode.
Bit 0 CC1E: Capture/Compare 1 output enable.
CC1 channel configured as output:
0: Off - OC1 is not active.
1: On - OC1 signal is output on the corresponding output pin.
CC1 channel configured as input:
This bit determines if a capture of the counter value can actually be done into the input
capture/compare register 1 (TIMx_CCR1) or not.
0: Capture disabled.
1: Capture enabled.

RM0090 Rev 17

681/1747
695

General-purpose timers (TIM9 to TIM14)

RM0090

Table 102. Output control bit for standard OCx channels
CCxE bit

OCx output state

0

Output disabled (OCx=’0’, OCx_EN=’0’)

1

OCx=OCxREF + Polarity, OCx_EN=’1’

Note:

The states of the external I/O pins connected to the standard OCx channels depend on the
state of the OCx channel and on the GPIO registers.

19.4.8

TIM9/12 counter (TIMx_CNT)
Address offset: 0x24
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

CNT[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

Bits 15:0 CNT[15:0]: Counter value

19.4.9

TIM9/12 prescaler (TIMx_PSC)
Address offset: 0x28
Reset value: 0x0000

15

14

13

12

11

10

9

8

PSC[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 15:0 PSC[15:0]: Prescaler value
The counter clock frequency CK_CNT is equal to fCK_PSC / (PSC[15:0] + 1).
PSC contains the value to be loaded in the active prescaler register at each update event
(including when the counter is cleared through UG bit of TIMx_EGR register or through
trigger controller when configured in “reset mode”).

19.4.10

TIM9/12 auto-reload register (TIMx_ARR)
Address offset: 0x2C
Reset value: 0xFFFF

15

14

13

12

11

10

9

8

rw

rw

rw

rw

rw

rw

rw

rw

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

ARR[15:0]
rw

Bits 15:0 ARR[15:0]: Auto-reload value
ARR is the value to be loaded into the actual auto-reload register.
Refer to the Section 19.3.1: Time-base unit for more details about ARR update and
behavior.
The counter is blocked while the auto-reload value is null.

682/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM9 to TIM14)

19.4.11

TIM9/12 capture/compare register 1 (TIMx_CCR1)
Address offset: 0x34
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

CCR1[15:0]
rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

Bits 15:0 CCR1[15:0]: Capture/Compare 1 value
If channel CC1 is configured as output:
CCR1 is the value to be loaded into the actual capture/compare 1 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR1 register
(OC1PE bit). Else the preload value is copied into the active capture/compare 1 register
when an update event occurs.
The active capture/compare register contains the value to be compared to the TIMx_CNT
counter and signaled on the OC1 output.
If channel CC1is configured as input:
CCR1 is the counter value transferred by the last input capture 1 event (IC1). The
TIMx_CCR1 register is read-only and cannot be programmed.

19.4.12

TIM9/12 capture/compare register 2 (TIMx_CCR2)
Address offset: 0x38
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

CCR2[15:0]
rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

Bits 15:0 CCR2[15:0]: Capture/Compare 2 value
If channel CC2 is configured as output:
CCR2 is the value to be loaded into the actual capture/compare 2 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR2 register
(OC2PE bit). Else the preload value is copied into the active capture/compare 2 register
when an update event occurs.
The active capture/compare register contains the value to be compared to the TIMx_CNT
counter and signalled on the OC2 output.
If channel CC2 is configured as input:
CCR2 is the counter value transferred by the last input capture 2 event (IC2). The
TIMx_CCR2 register is read-only and cannot be programmed.

RM0090 Rev 17

683/1747
695

General-purpose timers (TIM9 to TIM14)

19.4.13

RM0090

TIM9/12 register map
TIM9/12 registers are mapped as 16-bit addressable registers as described below. The
reserved memory areas are highlighted in gray in the table.

0

0

0

0

0

CC2S
[1:0]

0

0

0

0

4

3

2

1

0

URS

UDIS

CEN

OC1M
[2:0]
0

0

0

IC1F[3:0]

UIE
UIF

0

0

0
UG

CC1IE

0

CC1IF

0

CC1G

0
CC2IF

CC2IE

0

0

0

0

OC1FE

0

0

0

CC1
S
[1:0]
0

0

IC1
PSC
[1:0]

CC1
S
[1:0]

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CC1E

0

CC1P

TIMx_ARR

0

Reserved

TIMx_PSC

CC1NP

TIMx_CNT

0

CC2E

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CNT[15:0]

Reserved
0

0

0

0

0

0

0

0

0

0

PSC[15:0]

Reserved
0

0

0

0

0

0

0

0

0

0

ARR[15:0]

Reserved

Reset value

0

0

0

0

0

0

0

0

0

0

Reserved

0x30

TIMx_CCR1

CCR1[15:0]

Reserved

Reset value

684/1747

Reserved

CC2P

TIMx_CCER

Reset value

0x34

Reserved

0

Reserved

Reset value

0x2C

Reserved

0

OC1PE

0

IC2
PSC
[1:0]

Reset value

0x28

0

Reserved

0

0

CC2S
[1:0]

Reserved

0

IC2F[3:0]

Reserved

Reset value

0x24

0

SMS[2:0]

CC2NP

0

OC2M
[2:0]

OC2FE

Reserved

0x1C

0x20

0

0

Reset value
TIMx_CCMR1
Input Capture
mode

TS[2:0]

0

Reserved

OC2PE

0x18

0

TIF

0

Reserved

0

Reset value
TIMx_CCMR1
Output Compare
mode

0

TG

TIMx_EGR

CC1OF

Reserved

Reset value
0x14

0

0
CC2OF

TIMx_SR

OPM

6

Reserved

Reset value

0x10

0

TIE

0

TIMx_DIER

5

7

8

Reserved

Reset value

0x0C

Reserved

0

CC2G

TIMx_SMCR

0

Reserved

0

MSM

Reset value
0x08

9

11

10

12

13

14

15

16

17

18

20

19

CKD
[1:0]

Reserved

ARPE

TIMx_CR1

0x00

21

22

23

24

25

26

27

28

29

Register

30

Offset

31

Table 103. TIM9/12 register map and reset values

0

RM0090 Rev 17

0

0

0

0

0

0

0

0

0

RM0090

General-purpose timers (TIM9 to TIM14)

3

2

1

0

6

7

8

9

11

10

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

0

0

0

0

0

0

CCR2[15:0]

Reserved

Reset value
0x3C to
0x4C

4

TIMx_CCR2

5

0x38

Register

30

Offset

31

Table 103. TIM9/12 register map and reset values (continued)

0

0

0

0

0

0

0

0

0

0

Reserved

Refer to Section 2.3: Memory map for the register boundary addresses.

RM0090 Rev 17

685/1747
695

General-purpose timers (TIM9 to TIM14)

19.5

RM0090

TIM10/11/13/14 registers
The peripheral registers have to be written by half-words (16 bits) or words (32 bits). Read
accesses can be done by bytes (8 bits), half-words (16 bits) or words (32 bits).

19.5.1

TIM10/11/13/14 control register 1 (TIMx_CR1)
Address offset: 0x00
Reset value: 0x0000

15

14

13

12

11

10

Reserved

9

8

CKD[1:0]
rw

7
ARPE

rw

rw

6

5
Reserved

4

3

2

1

0

OPM

URS

UDIS

CEN

rw

rw

rw

rw

Bits 15:10 Reserved, must be kept at reset value.
Bits 9:8 CKD: Clock division
This bit-field indicates the division ratio between the timer clock (CK_INT) frequency and
sampling clock used by the digital filters (TIx),
00: tDTS = tCK_INT
01: tDTS = 2 × tCK_INT
10: tDTS = 4 × tCK_INT
11: Reserved
Bit 7 ARPE: Auto-reload preload enable
0: TIMx_ARR register is not buffered
1: TIMx_ARR register is buffered
Bits 6:4 Reserved, must be kept at reset value.
Bit 3 OPM: One-pulse mode
0: Counter is not stopped on the update event
1: Counter stops counting on the next update event (clearing the CEN bit).
Bit 2 URS: Update request source
This bit is set and cleared by software to select the update interrupt (UEV) sources.
0: Any of the following events generate an UEV if enabled:
–
Counter overflow
–
Setting the UG bit
1: Only counter overflow generates an UEV if enabled.
Bit 1 UDIS: Update disable
This bit is set and cleared by software to enable/disable update interrupt (UEV) event
generation.
0: UEV enabled. An UEV is generated by one of the following events:
–
Counter overflow
–
Setting the UG bit.
Buffered registers are then loaded with their preload values.
1: UEV disabled. No UEV is generated, shadow registers keep their value (ARR, PSC,
CCRx). The counter and the prescaler are reinitialized if the UG bit is set.
Bit 0 CEN: Counter enable
0: Counter disabled
1: Counter enabled

686/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM9 to TIM14)

19.5.2

TIM10/11/13/14 Interrupt enable register (TIMx_DIER)
Address offset: 0x0C
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

Reserved

Bits 15:2

1

0

CC1IE

UIE

rw

rw

Reserved, must be kept at reset value.

Bit 1 CC1IE: Capture/Compare 1 interrupt enable
0: CC1 interrupt disabled
1: CC1 interrupt enabled
Bit 0 UIE: Update interrupt enable
0: Update interrupt disabled
1: Update interrupt enabled

19.5.3

TIM10/11/13/14 status register (TIMx_SR)
Address offset: 0x10
Reset value: 0x0000

15

14

13

12

Reserved

Bits 15:10

11

10

9

8

7

CC1OF
rc_w0

6

5
Reserved

4

3

2

1

0

CC1IF

UIF

rc_w0

rc_w0

Reserved, must be kept at reset value.

Bit 9 CC1OF: Capture/Compare 1 overcapture flag
This flag is set by hardware only when the corresponding channel is configured in input
capture mode. It is cleared by software by writing it to ‘0’.
0: No overcapture has been detected.
1: The counter value has been captured in TIMx_CCR1 register while CC1IF flag was
already set
Bits 8:2

Reserved, must be kept at reset value.

Bit 1 CC1IF: Capture/compare 1 interrupt flag
If channel CC1 is configured as output:
This flag is set by hardware when the counter matches the compare value. It is cleared by
software.
0: No match.
1: The content of the counter TIMx_CNT matches the content of the TIMx_CCR1 register.
When the contents of TIMx_CCR1 are greater than the contents of TIMx_ARR, the CC1IF bit
goes high on the counter overflow.
If channel CC1 is configured as input:
This bit is set by hardware on a capture. It is cleared by software or by reading the
TIMx_CCR1 register.
0: No input capture occurred.
1: The counter value has been captured in TIMx_CCR1 register (an edge has been detected
on IC1 which matches the selected polarity).

RM0090 Rev 17

687/1747
695

General-purpose timers (TIM9 to TIM14)

RM0090

Bit 0 UIF: Update interrupt flag
This bit is set by hardware on an update event. It is cleared by software.
0: No update occurred.
1: Update interrupt pending. This bit is set by hardware when the registers are updated:
–
At overflow and if UDIS=’0’ in the TIMx_CR1 register.
–
When CNT is reinitialized by software using the UG bit in TIMx_EGR register, if
URS=’0’ and UDIS=’0’ in the TIMx_CR1 register.

19.5.4

TIM10/11/13/14 event generation register (TIMx_EGR)
Address offset: 0x14
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

Reserved

1

0

CC1G

UG

w

w

Bits 15:2 Reserved, must be kept at reset value.
Bit 1 CC1G: Capture/compare 1 generation
This bit is set by software in order to generate an event, it is automatically cleared by
hardware.
0: No action
1: A capture/compare event is generated on channel 1:
If channel CC1 is configured as output:
CC1IF flag is set, Corresponding interrupt or is sent if enabled.
If channel CC1 is configured as input:
The current value of the counter is captured in TIMx_CCR1 register. The CC1IF flag is set,
the corresponding interrupt is sent if enabled. The CC1OF flag is set if the CC1IF flag was
already high.
Bit 0 UG: Update generation
This bit can be set by software, it is automatically cleared by hardware.
0: No action
1: Re-initialize the counter and generates an update of the registers. Note that the prescaler
counter is cleared too (anyway the prescaler ratio is not affected). The counter is cleared.

19.5.5

TIM10/11/13/14 capture/compare mode register 1 (TIMx_CCMR1)
Address offset: 0x18
Reset value: 0x0000
The channels can be used in input (capture mode) or in output (compare mode). The
direction of a channel is defined by configuring the corresponding CCxS bits. All the other
bits of this register have a different function in input and in output mode. For a given bit,
OCxx describes its function when the channel is configured in output, ICxx describes its
function when the channel is configured in input. So take care that the same bit can have a
different meaning for the input stage and for the output stage.

15

14

13

12

11

10

9

8

7

6

Reserved
Reserved

688/1747

5

4

OC1M[2:0]

RM0090 Rev 17

rw

rw

2

OC1PE OC1FE

IC1F[3:0]
rw

3

IC1PSC[1:0]
rw

rw

rw

1

0

CC1S[1:0]
rw

rw

RM0090

General-purpose timers (TIM9 to TIM14)

Output compare mode
Bits 15:7

Reserved, must be kept at reset value.

Bits 6:4 OC1M: Output compare 1 mode
These bits define the behavior of the output reference signal OC1REF from which OC1 is
derived. OC1REF is active high whereas OC1 active level depends on CC1P bit.
000: Frozen. The comparison between the output compare register TIMx_CCR1 and the
counter TIMx_CNT has no effect on the outputs.
001: Set channel 1 to active level on match. OC1REF signal is forced high when the counter
TIMx_CNT matches the capture/compare register 1 (TIMx_CCR1).
010: Set channel 1 to inactive level on match. OC1REF signal is forced low when the
counter TIMx_CNT matches the capture/compare register 1 (TIMx_CCR1).
011: Toggle - OC1REF toggles when TIMx_CNT = TIMx_CCR1.
100: Force inactive level - OC1REF is forced low.
101: Force active level - OC1REF is forced high.
110: PWM mode 1 - Channel 1 is active as long as TIMx_CNT < TIMx_CCR1 else inactive.
111: PWM mode 2 - Channel 1 is inactive as long as TIMx_CNT < TIMx_CCR1 else active.
Note: In PWM mode 1 or 2, the OCREF level changes when the result of the comparison
changes or when the output compare mode switches from frozen to PWM mode.
Bit 3 OC1PE: Output compare 1 preload enable
0: Preload register on TIMx_CCR1 disabled. TIMx_CCR1 can be written at anytime, the
new value is taken in account immediately.
1: Preload register on TIMx_CCR1 enabled. Read/Write operations access the preload
register. TIMx_CCR1 preload value is loaded in the active register at each update event.
Note: The PWM mode can be used without validating the preload register only in one pulse
mode (OPM bit set in TIMx_CR1 register). Else the behavior is not guaranteed.
Bit 2 OC1FE: Output compare 1 fast enable
This bit is used to accelerate the effect of an event on the trigger in input on the CC output.
0: CC1 behaves normally depending on counter and CCR1 values even when the trigger is
ON. The minimum delay to activate CC1 output when an edge occurs on the trigger input is
5 clock cycles.
1: An active edge on the trigger input acts like a compare match on CC1 output. OC is then
set to the compare level independently of the result of the comparison. Delay to sample the
trigger input and to activate CC1 output is reduced to 3 clock cycles. OC1FE acts only if the
channel is configured in PWM1 or PWM2 mode.
Bits 1:0 CC1S: Capture/Compare 1 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC1 channel is configured as output.
01: CC1 channel is configured as input, IC1 is mapped on TI1.
10:
11:
Note: CC1S bits are writable only when the channel is OFF (CC1E = 0 in TIMx_CCER).

RM0090 Rev 17

689/1747
695

General-purpose timers (TIM9 to TIM14)

RM0090

Input capture mode
Bits 15:8

Reserved, must be kept at reset value.

Bits 7:4 IC1F: Input capture 1 filter
This bit-field defines the frequency used to sample TI1 input and the length of the digital filter
applied to TI1. The digital filter is made of an event counter in which N consecutive events
are needed to validate a transition on the output:
0000: No filter, sampling is done at fDTS
0001: fSAMPLING=fCK_INT, N=2
0010: fSAMPLING=fCK_INT, N=4
0011: fSAMPLING=fCK_INT, N=8
0100: fSAMPLING=fDTS/2, N=6
0101: fSAMPLING=fDTS/2, N=8
0110: fSAMPLING=fDTS/4, N=6
0111: fSAMPLING=fDTS/4, N=8
1000: fSAMPLING=fDTS/8, N=6
1001: fSAMPLING=fDTS/8, N=8
1010: fSAMPLING=fDTS/16, N=5
1011: fSAMPLING=fDTS/16, N=6
1100: fSAMPLING=fDTS/16, N=8
1101: fSAMPLING=fDTS/32, N=5
1110: fSAMPLING=fDTS/32, N=6
1111: fSAMPLING=fDTS/32, N=8
Bits 3:2 IC1PSC: Input capture 1 prescaler
This bit-field defines the ratio of the prescaler acting on CC1 input (IC1).
The prescaler is reset as soon as CC1E=’0’ (TIMx_CCER register).
00: no prescaler, capture is done each time an edge is detected on the capture input
01: capture is done once every 2 events
10: capture is done once every 4 events
11: capture is done once every 8 events
Bits 1:0 CC1S: Capture/Compare 1 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC1 channel is configured as output
01: CC1 channel is configured as input, IC1 is mapped on TI1
10: Reserved
11: Reserved
Note: CC1S bits are writable only when the channel is OFF (CC1E = 0 in TIMx_CCER).

690/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM9 to TIM14)

19.5.6

TIM10/11/13/14 capture/compare enable register (TIMx_CCER)
Address offset: 0x20
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3
CC1NP

Reserved

rw

2
Res.

1

0

CC1P

CC1E

rw

rw

Bits 15:4 Reserved, must be kept at reset value.
Bit 3 CC1NP: Capture/Compare 1 complementary output Polarity.
CC1 channel configured as output: CC1NP must be kept cleared.
CC1 channel configured as input: CC1NP bit is used in conjunction with CC1P to define
TI1FP1 polarity (refer to CC1P description).
Bit 2 Reserved, must be kept at reset value.
Bit 1 CC1P: Capture/Compare 1 output Polarity.
CC1 channel configured as output:
0: OC1 active high
1: OC1 active low
CC1 channel configured as input:
The CC1P bit selects TI1FP1 and TI2FP1 polarity for trigger or capture operations.
00: noninverted/rising edge
Circuit is sensitive to TI1FP1 rising edge (capture mode), TI1FP1 is not inverted.
01: inverted/falling edge
Circuit is sensitive to TI1FP1 falling edge (capture mode), TI1FP1 is inverted.
10: reserved, do not use this configuration.
11: noninverted/both edges
Circuit is sensitive to both TI1FP1 rising and falling edges (capture mode), TI1FP1 is not
inverted.
Bit 0 CC1E: Capture/Compare 1 output enable.
CC1 channel configured as output:
0: Off - OC1 is not active
1: On - OC1 signal is output on the corresponding output pin
CC1 channel configured as input:
This bit determines if a capture of the counter value can actually be done into the input
capture/compare register 1 (TIMx_CCR1) or not.
0: Capture disabled
1: Capture enabled

Table 104. Output control bit for standard OCx channels
CCxE bit

Note:

OCx output state

0

Output Disabled (OCx=’0’, OCx_EN=’0’)

1

OCx=OCxREF + Polarity, OCx_EN=’1’

The state of the external I/O pins connected to the standard OCx channels depends on the
OCx channel state and the GPIO registers.

RM0090 Rev 17

691/1747
695

General-purpose timers (TIM9 to TIM14)

19.5.7

RM0090

TIM10/11/13/14 counter (TIMx_CNT)
Address offset: 0x24
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

CNT[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 15:0 CNT[15:0]: Counter value

19.5.8

TIM10/11/13/14 prescaler (TIMx_PSC)
Address offset: 0x28
Reset value: 0x0000

15

14

13

12

11

10

9

8

rw

rw

rw

rw

rw

rw

rw

rw

7

PSC[15:0]
rw

Bits 15:0 PSC[15:0]: Prescaler value
The counter clock frequency CK_CNT is equal to fCK_PSC / (PSC[15:0] + 1).
PSC contains the value to be loaded in the active prescaler register at each update event
(including when the counter is cleared through UG bit of TIMx_EGR register or through
trigger controller when configured in “reset mode”).

19.5.9

TIM10/11/13/14 auto-reload register (TIMx_ARR)
Address offset: 0x2C
Reset value: 0xFFFF

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

ARR[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 15:0 ARR[15:0]: Auto-reload value
ARR is the value to be loaded in the actual auto-reload register.
Refer to Section 19.3.1: Time-base unit for more details about ARR update and behavior.
The counter is blocked while the auto-reload value is null.

692/1747

RM0090 Rev 17

RM0090

General-purpose timers (TIM9 to TIM14)

19.5.10

TIM10/11/13/14 capture/compare register 1 (TIMx_CCR1)
Address offset: 0x34
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

CCR1[15:0]
rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

rw/ro

Bits 15:0 CCR1[15:0]: Capture/Compare 1 value
If channel CC1 is configured as output:
CCR1 is the value to be loaded in the actual capture/compare 1 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR1 register (bit
OC1PE). Else the preload value is copied in the active capture/compare 1 register when an
update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signaled on OC1 output.
If channel CC1is configured as input:
CCR1 is the counter value transferred by the last input capture 1 event (IC1). The TIMx_CCR1
register is read-only and cannot be programmed.

19.5.11

TIM11 option register 1 (TIM11_OR)
Address offset: 0x50
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

Reserved

Bits 15:2
Bits 1:0

6

5

4

3

2

1

0

TI1_RMP[1:0]
rw

Reserved, must be kept at reset value.
TI1_RMP[1:0]: TIM11 Input 1 remapping capability
Set and cleared by software.
00,01,11: TIM11 Channel1 is connected to the GPIO (refer to the Alternate function mapping
table in the datasheets).
10: HSE_RTC clock (HSE divided by programmable prescaler) is connected to the
TIM11_CH1 input for measurement purposes

RM0090 Rev 17

693/1747
695

General-purpose timers (TIM9 to TIM14)

19.5.12

RM0090

TIM10/11/13/14 register map
TIMx registers are mapped as 16-bit addressable registers as described in the table below.

0

TIMx_SMCR

0x08

0

0

0

0

0

CEN

Reset value

Reserve
d

URS

CKD
[1:0]

Reserved

UDIS

TIMx_CR1

0x00

OPM

Register

ARPE

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 105. TIM10/11/13/14 register map and reset values

0

Reserved

Reset value

0

TIMx_EGR

0x14

Reserved

Reserved

0

0

IC1F[3:0]

Reserved

Reset value

0

0

0

0

TIMx_PSC

0

TIMx_ARR

UIE

CC1S
[1:0]

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

PSC[15:0]
0

0

0

0

0

0

0

0

0

0

ARR[15:0]

Reserved
0

0

0

0

0

0

0

0

0

0

Reserved
TIMx_CCR1

CCR1[15:0]

Reserved

Reset value

694/1747

0

Reserved

0x30

0x38 to
0x4C

0

CNT[15:0]

Reserved

Reset value

0x34

0

IC1
PSC
[1:0]

CC1E

TIMx_CNT

Reset value
0x2C

CC1S
[1:0]

0

Reserved

Reset value
0x28

0

0

CC1P

TIMx_CCER
Reset value

0x24

0

0

Reserved

0x1C

0x20

0

OC1FE

Reset value
TIMx_CCMR1
Input capture
mode

0

Reserved

0x18

OC1M
[2:0]

Reserved

0

CC1NP

TIMx_CCMR1
Output compare
mode

OC1PE

Reset value

UIF

Reserved

0

UG

CC1OF

TIMx_SR

0x10

0
CC1IF

Reserved

Reset value

CC1G

TIMx_DIER

0x0C

CC1IE

Reset value

0
Reserved

RM0090 Rev 17

0

0

0

0

0

0

0

0

0

RM0090

General-purpose timers (TIM9 to TIM14)

0x50

Register

TIMx_OR

Reserved

Reset value

TI1_RMP

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 105. TIM10/11/13/14 register map and reset values (continued)

0

0

Refer to Section 3.3: Memory map for the register boundary addresses.

RM0090 Rev 17

695/1747
695

Basic timers (TIM6 and TIM7)

20

RM0090

Basic timers (TIM6 and TIM7)
This section applies to the whole STM32F4xx family, unless otherwise specified.

20.1

TIM6 and TIM7 introduction
The basic timers TIM6 and TIM7 consist of a 16-bit auto-reload counter driven by a
programmable prescaler.
They may be used as generic timers for time-base generation but they are also specifically
used to drive the digital-to-analog converter (DAC). In fact, the timers are internally
connected to the DAC and are able to drive it through their trigger outputs.
The timers are completely independent, and do not share any resources.

20.2

TIM6 and TIM7 main features
Basic timer (TIM6 and TIM7) features include:
•

16-bit auto-reload upcounter

•

16-bit programmable prescaler used to divide (also “on the fly”) the counter clock
frequency by any factor between 1 and 65536

•

Synchronization circuit to trigger the DAC

•

Interrupt/DMA generation on the update event: counter overflow
Figure 203. Basic timer block diagram

7,0[&/.IURP5&&

7ULJJHU 75*2
FRQWUROOHU

,QWHUQDOFORFN &.B,17

5HVHW(QDEOH&RXQW

&RQWUROOHU

8

$XWRUHORDG5HJLVWHU

6WRS&OHDURUXS
&.B36&

)ODJ

&.B&17
36&
“
3UHVFDOHU

WR'$&

8,
8

&17
&2817(5

3UHORDGUHJLVWHUVWUDQVIHUUHG
WRDFWLYHUHJLVWHUVRQ8HYHQWDFFRUGLQJWRFRQWUROELW
HYHQW
LQWHUUXSW '0$RXWSXW

696/1747

DLE

RM0090 Rev 17

RM0090

Basic timers (TIM6 and TIM7)

20.3

TIM6 and TIM7 functional description

20.3.1

Time-base unit
The main block of the programmable timer is a 16-bit upcounter with its related auto-reload
register. The counter clock can be divided by a prescaler.
The counter, the auto-reload register and the prescaler register can be written or read by
software. This is true even when the counter is running.
The time-base unit includes:
•

Counter Register (TIMx_CNT)

•

Prescaler Register (TIMx_PSC)

•

Auto-Reload Register (TIMx_ARR)

The auto-reload register is preloaded. The preload register is accessed each time an
attempt is made to write or read the auto-reload register. The contents of the preload
register are transferred into the shadow register permanently or at each update event UEV,
depending on the auto-reload preload enable bit (ARPE) in the TIMx_CR1 register. The
update event is sent when the counter reaches the overflow value and if the UDIS bit equals
0 in the TIMx_CR1 register. It can also be generated by software. The generation of the
update event is described in detail for each configuration.
The counter is clocked by the prescaler output CK_CNT, which is enabled only when the
counter enable bit (CEN) in the TIMx_CR1 register is set.
Note that the actual counter enable signal CNT_EN is set 1 clock cycle after CEN.

Prescaler description
The prescaler can divide the counter clock frequency by any factor between 1 and 65536. It
is based on a 16-bit counter controlled through a 16-bit register (in the TIMx_PSC register).
It can be changed on the fly as the TIMx_PSC control register is buffered. The new
prescaler ratio is taken into account at the next update event.
Figure 204 and Figure 205 give some examples of the counter behavior when the prescaler
ratio is changed on the fly.

RM0090 Rev 17

697/1747
707

Basic timers (TIM6 and TIM7)

RM0090

Figure 204. Counter timing diagram with prescaler division change from 1 to 2
&.B36&

&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU

)

)

)

)$ )% )&









8SGDWHHYHQW 8(9
3UHVFDOHUFRQWUROUHJLVWHU





:ULWHDQHZYDOXHLQ7,0[B36&
3UHVFDOHUEXIIHU



3UHVFDOHUFRXQWHU





















069

Figure 205. Counter timing diagram with prescaler division change from 1 to 4
&.B36&
&17B(1

7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU

)

)

)

)$ )% )&





8SGDWHHYHQW 8(9
3UHVFDOHUFRQWUROUHJLVWHU





:ULWHDQHZYDOXHLQ7,0[B36&
3UHVFDOHUEXIIHU



3UHVFDOHUFRXQWHU




















069

698/1747

RM0090 Rev 17

RM0090

20.3.2

Basic timers (TIM6 and TIM7)

Counting mode
The counter counts from 0 to the auto-reload value (contents of the TIMx_ARR register),
then restarts from 0 and generates a counter overflow event.
An update event can be generate at each counter overflow or by setting the UG bit in the
TIMx_EGR register (by software or by using the slave mode controller).
The UEV event can be disabled by software by setting the UDIS bit in the TIMx_CR1
register. This avoids updating the shadow registers while writing new values into the preload
registers. In this way, no update event occurs until the UDIS bit has been written to 0,
however, the counter and the prescaler counter both restart from 0 (but the prescale rate
does not change). In addition, if the URS (update request selection) bit in the TIMx_CR1
register is set, setting the UG bit generates an update event UEV, but the UIF flag is not set
(so no interrupt or DMA request is sent).
When an update event occurs, all the registers are updated and the update flag (UIF bit in
the TIMx_SR register) is set (depending on the URS bit):
•

The buffer of the prescaler is reloaded with the preload value (contents of the
TIMx_PSC register)

•

The auto-reload shadow register is updated with the preload value (TIMx_ARR)

The following figures show some examples of the counter behavior for different clock
frequencies when TIMx_ARR = 0x36.
Figure 206. Counter timing diagram, internal clock divided by 1
&.B,17

&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU









  















&RXQWHURYHUIORZ

8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

RM0090 Rev 17

699/1747
707

Basic timers (TIM6 and TIM7)

RM0090

Figure 207. Counter timing diagram, internal clock divided by 2
&.B,17
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU















&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
069

Figure 208. Counter timing diagram, internal clock divided by 4
&.B,17
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU









&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
06Y9

Figure 209. Counter timing diagram, internal clock divided by N
&.B,17

7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU

)





&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
06Y9

700/1747

RM0090 Rev 17

RM0090

Basic timers (TIM6 and TIM7)
Figure 210. Counter timing diagram, update event when ARPE=0 (TIMx_ARR not
preloaded)
&.B,17
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU



            

&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
$XWRUHORDGUHJLVWHU

))



:ULWHDQHZYDOXHLQ7,0[B$55
06Y9

Figure 211. Counter timing diagram, update event when ARPE=1 (TIMx_ARR
preloaded)
&.B36&
&17B(1
7LPHUFORFN &.B&17
&RXQWHUUHJLVWHU

)

) ) ) ) )        

&RXQWHURYHUIORZ
8SGDWHHYHQW 8(9
8SGDWHLQWHUUXSWIODJ 8,)
$XWRUHORDGSUHORDGUHJLVWHU

$XWRUHORDGVKDGRZUHJLVWHU

)



)



:ULWHDQHZYDOXHLQ7,0[B$55
06Y9

20.3.3

Clock source
The counter clock is provided by the Internal clock (CK_INT) source.
The CEN (in the TIMx_CR1 register) and UG bits (in the TIMx_EGR register) are actual
control bits and can be changed only by software (except for UG that remains cleared
automatically). As soon as the CEN bit is written to 1, the prescaler is clocked by the internal
clock CK_INT.
Figure 212 shows the behavior of the control circuit and the upcounter in normal mode,
without prescaler.

RM0090 Rev 17

701/1747
707

Basic timers (TIM6 and TIM7)

RM0090

Figure 212. Control circuit in normal mode, internal clock divided by 1

,QWHUQDOFORFN
&(1 &17B(1
8*

&17B,1,7

&RXQWHUFORFN &.B&17 &.B36&
&RXQWHUUHJLVWHU









 



 

  




069

20.3.4

Debug mode
When the microcontroller enters the debug mode (Cortex®-M4 with FPU core - halted), the
TIMx counter either continues to work normally or stops, depending on the
DBG_TIMx_STOP configuration bit in the DBG module. For more details, refer to
Section 38.16.2: Debug support for timers, watchdog, bxCAN and I2C.

20.4

TIM6 and TIM7 registers
Refer to Section 2.2 for a list of abbreviations used in register descriptions.
The peripheral registers have to be written by half-words (16 bits) or words (32 bits). Read
accesses can be done by bytes (8 bits), half-words (16 bits) or words (32 bits).

20.4.1

TIM6 and TIM7 control register 1 (TIMx_CR1)
Address offset: 0x00
Reset value: 0x0000

15

14

13

12

11

Reserved

10

9

8

7
ARPE
rw

6

5
Reserved

4

3

2

1

0

OPM

URS

UDIS

CEN

rw

rw

rw

rw

Bits 15:8 Reserved, must be kept at reset value.
Bit 7 ARPE: Auto-reload preload enable
0: TIMx_ARR register is not buffered.
1: TIMx_ARR register is buffered.
Bits 6:4 Reserved, must be kept at reset value.
Bit 3 OPM: One-pulse mode
0: Counter is not stopped at update event
1: Counter stops counting at the next update event (clearing the CEN bit).

702/1747

RM0090 Rev 17

RM0090

Basic timers (TIM6 and TIM7)

Bit 2 URS: Update request source
This bit is set and cleared by software to select the UEV event sources.
0: Any of the following events generates an update interrupt or DMA request if enabled.
These events can be:
–
Counter overflow/underflow
–
Setting the UG bit
–
Update generation through the slave mode controller
1: Only counter overflow/underflow generates an update interrupt or DMA request if enabled.
Bit 1 UDIS: Update disable
This bit is set and cleared by software to enable/disable UEV event generation.
0: UEV enabled. The Update (UEV) event is generated by one of the following events:
–
Counter overflow/underflow
–
Setting the UG bit
–
Update generation through the slave mode controller
Buffered registers are then loaded with their preload values.
1: UEV disabled. The Update event is not generated, shadow registers keep their value
(ARR, PSC). However the counter and the prescaler are reinitialized if the UG bit is set or if
a hardware reset is received from the slave mode controller.
Bit 0 CEN: Counter enable
0: Counter disabled
1: Counter enabled
Note: Gated mode can work only if the CEN bit has been previously set by software. However
trigger mode can set the CEN bit automatically by hardware.
CEN is cleared automatically in one-pulse mode, when an update event occurs.

RM0090 Rev 17

703/1747
707

Basic timers (TIM6 and TIM7)

20.4.2

RM0090

TIM6 and TIM7 control register 2 (TIMx_CR2)
Address offset: 0x04
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

MMS[2:0]

Reserved

rw

rw

2

1

0

Reserved

rw

Bits 15:7 Reserved, must be kept at reset value.
Bits 6:4 MMS[2:0]: Master mode selection
These bits are used to select the information to be sent in master mode to slave timers for
synchronization (TRGO). The combination is as follows:
000: Reset - the UG bit from the TIMx_EGR register is used as a trigger output (TRGO). If
reset is generated by the trigger input (slave mode controller configured in reset mode) then
the signal on TRGO is delayed compared to the actual reset.
001: Enable - the Counter enable signal, CNT_EN, is used as a trigger output (TRGO). It is
useful to start several timers at the same time or to control a window in which a slave timer
is enabled. The Counter Enable signal is generated by a logic OR between CEN control bit
and the trigger input when configured in gated mode.
When the Counter Enable signal is controlled by the trigger input, there is a delay on TRGO,
except if the master/slave mode is selected (see the MSM bit description in the TIMx_SMCR
register).
010: Update - The update event is selected as a trigger output (TRGO). For instance a
master timer can then be used as a prescaler for a slave timer.
Note: The clock of the slave timer and ADC must be enabled prior to receiving events from
the master timer, and must not be changed on-the-fly while triggers are received from
the master timer.
Bits 3:0 Reserved, must be kept at reset value.

20.4.3

TIM6 and TIM7 DMA/Interrupt enable register (TIMx_DIER)
Address offset: 0x0C
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

UDE

Reserved

rw

Bits 15:9 Reserved, must be kept at reset value.
Bit 8 UDE: Update DMA request enable
0: Update DMA request disabled.
1: Update DMA request enabled.
Bits 7:1 Reserved, must be kept at reset value.
Bit 0 UIE: Update interrupt enable
0: Update interrupt disabled.
1: Update interrupt enabled.

704/1747

RM0090 Rev 17

6

5

4
Reserved

3

2

1

0
UIE
rw

RM0090

Basic timers (TIM6 and TIM7)

20.4.4

TIM6 and TIM7 status register (TIMx_SR)
Address offset: 0x10
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0
UIF

Reserved

rc_w0

Bits 15:1 Reserved, must be kept at reset value.
Bit 0 UIF: Update interrupt flag
This bit is set by hardware on an update event. It is cleared by software.
0: No update occurred.
1: Update interrupt pending. This bit is set by hardware when the registers are updated:
– At overflow or underflow and if UDIS = 0 in the TIMx_CR1 register.
– When CNT is reinitialized by software using the UG bit in the TIMx_EGR register, if URS = 0
and UDIS = 0 in the TIMx_CR1 register.

20.4.5

TIM6 and TIM7 event generation register (TIMx_EGR)
Address offset: 0x14
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0
UG

Reserved

w

Bits 15:1 Reserved, must be kept at reset value.
Bit 0 UG: Update generation
This bit can be set by software, it is automatically cleared by hardware.
0: No action.
1: Re-initializes the timer counter and generates an update of the registers. Note that the
prescaler counter is cleared too (but the prescaler ratio is not affected).

20.4.6

TIM6 and TIM7 counter (TIMx_CNT)
Address offset: 0x24
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

CNT[15:0]
rw

rw

Bits 15:0

rw

rw

rw

rw

rw

rw

rw

CNT[15:0]: Counter value

RM0090 Rev 17

705/1747
707

Basic timers (TIM6 and TIM7)

20.4.7

RM0090

TIM6 and TIM7 prescaler (TIMx_PSC)
Address offset: 0x28
Reset value: 0x0000

15

14

13

12

11

10

9

8

rw

rw

rw

rw

rw

rw

rw

rw

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

PSC[15:0]
rw

Bits 15:0 PSC[15:0]: Prescaler value
The counter clock frequency CK_CNT is equal to fCK_PSC / (PSC[15:0] + 1).
PSC contains the value to be loaded in the active prescaler register at each update event
(including when the counter is cleared through UG bit of TIMx_EGR register or through
trigger controller when configured in “reset mode”).

20.4.8

TIM6 and TIM7 auto-reload register (TIMx_ARR)
Address offset: 0x2C
Reset value: 0xFFFF

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

ARR[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 15:0 ARR[15:0]: Auto-reload value
ARR is the value to be loaded into the actual auto-reload register.
Refer to Section 20.3.1: Time-base unit for more details about ARR update and behavior.
The counter is blocked while the auto-reload value is null.

706/1747

RM0090 Rev 17

RM0090

20.4.9

Basic timers (TIM6 and TIM7)

TIM6 and TIM7 register map
TIMx registers are mapped as 16-bit addressable registers as described in the table below.

TIMx_CR2

MMS[2:0]

Reserved

Reset value

0

URS

UDIS

CEN

0

0

0

Reserved

UDE

TIMx_DIER

Reserved

0

TIMx_SR

0
UIF

Reset value
0x10

0

0

Reserved

0x08

0x0C

0

UIE

0x04

0

Reserved

Reserved

Reset value

OPM

TIMx_CR1

Reserved

0x00

Register

ARPE

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 106. TIM6 and TIM7 register map and reset values

Reserved

0x14

0

TIMx_EGR

UG

Reset value
Reserved

Reset value

0

0x18

Reserved

0x1C

Reserved

0x20

Reserved

0x24

TIMx_CNT

CNT[15:0]

Reserved

Reset value
0x28

TIMx_PSC

0

TIMx_ARR

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

1

1

1

1

PSC[15:0]

Reserved

Reset value
0x2C

0

0

0

0

0

0

0

0

0

0

0

ARR[15:0]

Reserved

Reset value

1

1

1

1

1

1

1

1

1

1

Refer to Section 2.3: Memory map for the register boundary addresses.

RM0090 Rev 17

707/1747
707

Independent watchdog (IWDG)

21

RM0090

Independent watchdog (IWDG)
This section applies to the whole STM32F4xx family, unless otherwise specified.

21.1

IWDG introduction
The devices have two embedded watchdog peripherals which offer a combination of high
safety level, timing accuracy and flexibility of use. Both watchdog peripherals (Independent
and Window) serve to detect and resolve malfunctions due to software failure, and to trigger
system reset or an interrupt (window watchdog only) when the counter reaches a given
timeout value.
The independent watchdog (IWDG) is clocked by its own dedicated low-speed clock (LSI)
and thus stays active even if the main clock fails. The window watchdog (WWDG) clock is
prescaled from the APB1 clock and has a configurable time-window that can be
programmed to detect abnormally late or early application behavior.
The IWDG is best suited to applications which require the watchdog to run as a totally
independent process outside the main application, but have lower timing accuracy
constraints. The WWDG is best suited to applications which require the watchdog to react
within an accurate timing window. For further information on the window watchdog, refer to
Section 22 on page 713.

21.2

21.3

IWDG main features
•

Free-running downcounter

•

clocked from an independent RC oscillator (can operate in Standby and Stop modes)

•

Reset (if watchdog activated) when the downcounter value of 0x000 is reached

IWDG functional description
Figure 213 shows the functional blocks of the independent watchdog module.
When the independent watchdog is started by writing the value 0xCCCC in the Key register
(IWDG_KR), the counter starts counting down from the reset value of 0xFFF. When it
reaches the end of count value (0x000) a reset signal is generated (IWDG reset).
Whenever the key value 0xAAAA is written in the IWDG_KR register, the IWDG_RLR value
is reloaded in the counter and the watchdog reset is prevented.

21.3.1

Hardware watchdog
If the “Hardware watchdog” feature is enabled through the device option bits, the watchdog
is automatically enabled at power-on, and will generate a reset unless the Key register is
written by the software before the counter reaches end of count.

21.3.2

Register access protection
Write access to the IWDG_PR and IWDG_RLR registers is protected. To modify them, first
write the code 0x5555 in the IWDG_KR register. A write access to this register with a

708/1747

RM0090 Rev 17

RM0090

Independent watchdog (IWDG)
different value will break the sequence and register access will be protected again. This
implies that it is the case of the reload operation (writing 0xAAAA).
A status register is available to indicate that an update of the prescaler or the down-counter
reload value is on going.

21.3.3

Debug mode
When the microcontroller enters debug mode (Cortex®-M4 with FPU core halted), the IWDG
counter either continues to work normally or stops, depending on DBG_IWDG_STOP
configuration bit in DBG module. For more details, refer to Section 38.16.2: Debug support
for timers, watchdog, bxCAN and I2C.
Figure 213. Independent watchdog block diagram
#/2%
0RESCALER REGISTER
)7$'?02

3TATUS REGISTER
)7$'?32

2ELOAD REGISTER
)7$'?2,2

+EY REGISTER
)7$'?+2

 BIT RELOAD VALUE

 BIT
,3)
 K(Z PRESCALER

 BIT DOWNCOUNTER

)7$' RESET

6$$ VOLTAGE DOMAIN
-36

Note:

The watchdog function is implemented in the VDD voltage domain, still functional in Stop and
Standby modes.
Table 107. Min/max IWDG timeout period (in ms) at 32 kHz (LSI)(1)
Prescaler divider

PR[2:0] bits

Min timeout RL[11:0]= 0x000 Max timeout RL[11:0]= 0xFFF

/4

0

0.125

512

/8

1

0.25

1024

/16

2

0.5

2048

/32

3

1

4096

/64

4

2

8192

/128

5

4

16384

/256

6

8

32768

1. These timings are given for a 32 kHz clock but the microcontroller internal RC frequency can vary. Refer to
the LSI oscillator characteristics table in the device datasheet for maximum and minimum values.

RM0090 Rev 17

709/1747
712

Independent watchdog (IWDG)

21.4

RM0090

IWDG registers
Refer to Section 2.2 on page 45 for a list of abbreviations used in register descriptions.
The peripheral registers have to be accessed by half-words (16 bits) or words (32 bits).

21.4.1

Key register (IWDG_KR)
Address offset: 0x00
Reset value: 0x0000 0000 (reset by Standby mode)

31 30 29 28 27 26 25

24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

w

w

w

w

w

w

w

KEY[15:0]

Reserved

w

w

w

w

w

w

w

w

w

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 KEY[15:0]: Key value (write only, read 0000h)
These bits must be written by software at regular intervals with the key value AAAAh,
otherwise the watchdog generates a reset when the counter reaches 0.
Writing the key value 5555h to enable access to the IWDG_PR and IWDG_RLR registers
(see Section 21.3.2)
Writing the key value CCCCh starts the watchdog (except if the hardware watchdog option is
selected)

21.4.2

Prescaler register (IWDG_PR)
Address offset: 0x04
Reset value: 0x0000 0000

31 30 29 28 27 26 25

24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

7

6

5

4

3

2

1

0

PR[2:0]
rw

rw

rw

Bits 31:3 Reserved, must be kept at reset value.
Bits 2:0 PR[2:0]: Prescaler divider
These bits are write access protected seeSection 21.3.2. They are written by software to
select the prescaler divider feeding the counter clock. PVU bit of IWDG_SR must be reset in
order to be able to change the prescaler divider.
000: divider /4
001: divider /8
010: divider /16
011: divider /32
100: divider /64
101: divider /128
110: divider /256
111: divider /256
Note: Reading this register returns the prescaler value from the VDD voltage domain. This
value may not be up to date/valid if a write operation to this register is ongoing. For this
reason the value read from this register is valid only when the PVU bit in the IWDG_SR
register is reset.

710/1747

RM0090 Rev 17

RM0090

21.4.3

Independent watchdog (IWDG)

Reload register (IWDG_RLR)
Address offset: 0x08
Reset value: 0x0000 0FFF (reset by Standby mode)

31 30 29 28 27 26 25

24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

RL[11:0]

Reserved

rw rw

rw

rw

rw

rw

rw

Bits 31:12 Reserved, must be kept at reset value.
Bits11:0 RL[11:0]: Watchdog counter reload value
These bits are write access protected see Section 21.3.2. They are written by software to
define the value to be loaded in the watchdog counter each time the value AAAAh is written
in the IWDG_KR register. The watchdog counter counts down from this value. The timeout
period is a function of this value and the clock prescaler. Refer to Table 107.
The RVU bit in the IWDG_SR register must be reset in order to be able to change the reload
value.
Note: Reading this register returns the reload value from the VDD voltage domain. This value
may not be up to date/valid if a write operation to this register is ongoing on this
register. For this reason the value read from this register is valid only when the RVU bit
in the IWDG_SR register is reset.

21.4.4

Status register (IWDG_SR)
Address offset: 0x0C
Reset value: 0x0000 0000 (not reset by Standby mode)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

7

6

5

4

3

2

1

0

RVU PVU
r

r

Bits 31:2 Reserved, must be kept at reset value.
Bit 1 RVU: Watchdog counter reload value update
This bit is set by hardware to indicate that an update of the reload value is ongoing. It is reset
by hardware when the reload value update operation is completed in the VDD voltage domain
(takes up to 5 RC 40 kHz cycles).
Reload value can be updated only when RVU bit is reset.
Bit 0 PVU: Watchdog prescaler value update
This bit is set by hardware to indicate that an update of the prescaler value is ongoing. It is
reset by hardware when the prescaler update operation is completed in the VDD voltage
domain (takes up to 5 RC 40 kHz cycles).
Prescaler value can be updated only when PVU bit is reset.

Note:

If several reload values or prescaler values are used by application, it is mandatory to wait
until RVU bit is reset before changing the reload value and to wait until PVU bit is reset
before changing the prescaler value. However, after updating the prescaler and/or the
reload value it is not necessary to wait until RVU or PVU is reset before continuing code
execution (even in case of low-power mode entry, the write operation is taken into account
and will complete)

RM0090 Rev 17

711/1747
712

Independent watchdog (IWDG)

21.4.5

RM0090

IWDG register map
The following table gives the IWDG register map and reset values.

0x04
0x08

0x0C

IWDG_KR
Reset value
IWDG_PR

Reset value
IWDG_SR

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

PR[2:0]
0

0

0

1

1

1

0

0

RL[11:0]

Reserved

1

1

1

1

1

Reserved

Reset value

Refer to Section 2.3: Memory map for the register boundary addresses.

712/1747

0

Reserved

Reset value
IWDG_RLR

KEY[15:0]

Reserved

PVU

0x00

Register

RVU

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 108. IWDG register map and reset values

RM0090 Rev 17

1

1

1

1

RM0090

22

Window watchdog (WWDG)

Window watchdog (WWDG)
This section applies to the whole STM32F4xx family, unless otherwise specified.

22.1

WWDG introduction
The window watchdog is used to detect the occurrence of a software fault, usually
generated by external interference or by unforeseen logical conditions, which causes the
application program to abandon its normal sequence. The watchdog circuit generates an
MCU reset on expiry of a programmed time period, unless the program refreshes the
contents of the downcounter before the T6 bit becomes cleared. An MCU reset is also
generated if the 7-bit downcounter value (in the control register) is refreshed before the
downcounter has reached the window register value. This implies that the counter must be
refreshed in a limited window.

22.2

WWDG main features
•

Programmable free-running downcounter

•

Conditional reset

•

22.3

–

Reset (if watchdog activated) when the downcounter value becomes less than
0x40

–

Reset (if watchdog activated) if the downcounter is reloaded outside the window
(see Figure 215)

Early wakeup interrupt (EWI): triggered (if enabled and the watchdog activated) when
the downcounter is equal to 0x40.

WWDG functional description
If the watchdog is activated (the WDGA bit is set in the WWDG_CR register) and when the
7-bit downcounter (T[6:0] bits) rolls over from 0x40 to 0x3F (T6 becomes cleared), it initiates
a reset. If the software reloads the counter while the counter is greater than the value stored
in the window register, then a reset is generated.

RM0090 Rev 17

713/1747
719

Window watchdog (WWDG)

RM0090
Figure 214. Watchdog block diagram

5(6(7

:DWFKGRJFRQILJXUDWLRQUHJLVWHU ::'*B&)5
FRPSDUDWRU
ZKHQ
7!:



: : : : : : :

:ULWH::'*B&5
:DWFKGRJFRQWUROUHJLVWHU ::'*B&5
:'*$
3&/.
IURP5&&FORFNFRQWUROOHU


7

7 7 7

7 7 7

ELWGRZQFRXQWHU &17

:'*SUHVFDOHU
:'*7%
06Y9

The application program must write in the WWDG_CR register at regular intervals during
normal operation to prevent an MCU reset. This operation must occur only when the counter
value is lower than the window register value. The value to be stored in the WWDG_CR
register must be between 0xFF and 0xC0.

Enabling the watchdog
The watchdog is always disabled after a reset. It is enabled by setting the WDGA bit in the
WWDG_CR register, then it cannot be disabled again except by a reset.

Controlling the downcounter
This downcounter is free-running, counting down even if the watchdog is disabled. When
the watchdog is enabled, the T6 bit must be set to prevent generating an immediate reset.
The T[5:0] bits contain the number of increments which represents the time delay before the
watchdog produces a reset. The timing varies between a minimum and a maximum value
due to the unknown status of the prescaler when writing to the WWDG_CR register (see
Figure 215). The Configuration register (WWDG_CFR) contains the high limit of the window:
To prevent a reset, the downcounter must be reloaded when its value is lower than the
window register value and greater than 0x3F. Figure 215 describes the window watchdog
process.
Note:

The T6 bit can be used to generate a software reset (the WDGA bit is set and the T6 bit is
cleared).

Advanced watchdog interrupt feature
The Early Wakeup Interrupt (EWI) can be used if specific safety operations or data logging
must be performed before the actual reset is generated. The EWI interrupt is enabled by
setting the EWI bit in the WWDG_CFR register. When the downcounter reaches the value
0x40, an EWI interrupt is generated and the corresponding interrupt service routine (ISR)
can be used to trigger specific actions (such as communications or data logging), before
resetting the device.

714/1747

RM0090 Rev 17

RM0090

Window watchdog (WWDG)
In some applications, the EWI interrupt can be used to manage a software system check
and/or system recovery/graceful degradation, without generating a WWDG reset. In this
case, the corresponding interrupt service routine (ISR) should reload the WWDG counter to
avoid the WWDG reset, then trigger the required actions.
The EWI interrupt is cleared by writing '0' to the EWIF bit in the WWDG_SR register.

Note:

When the EWI interrupt cannot be served, e.g. due to a system lock in a higher priority task,
the WWDG reset will eventually be generated.

22.4

How to program the watchdog timeout
Warning:

When writing to the WWDG_CR register, always write 1 in the
T6 bit to avoid generating an immediate reset.

Figure 215. Window watchdog timing diagram
4;= #.4 DOWNCOUNTER

7;=

X&

2EFRESH NOT ALLOWED 2EFRESH ALLOWED

4IME

4 BIT
2%3%4
AIC

The formula to calculate the WWDG timeout value is given by:
WDGTB[1:0]
t WWDG = t PCLK1 × 4096 × 2
× ( T[5:0] + 1 )

( ms )

where:
tWWDG: WWDG timeout
tPCLK1: APB1 clock period measured in ms
4096: value corresponding to internal divider

RM0090 Rev 17

715/1747
719

Window watchdog (WWDG)

RM0090

As an example, let us assume APB1 frequency is equal to 24 MHz, WDGTB[1:0] is set to 3
and T[5:0] is set to 63:
3
t WWDG = 1 ⁄ 24000 × 4096 × 2 × ( 63 + 1 ) = 21.85ms
Refer to Table 109 for the minimum and maximum values of the tWWDG.
Table 109. Minimum and maximum timeout values at 30 MHz (fPCLK1)

22.5

Prescaler

WDGTB

Min timeout (µs)
T[5:0] = 0x00

Max timeout (ms)
T[5:0] = 0x3F

1

0

136.53

8.74

2

1

273.07

17.48

4

2

546.13

34.95

8

3

1092.27

69.91

Debug mode
When the microcontroller enters debug mode (Cortex®-M4 with FPU core halted), the
WWDG counter either continues to work normally or stops, depending on
DBG_WWDG_STOP configuration bit in DBG module. For more details, refer to
Section 38.16.2: Debug support for timers, watchdog, bxCAN and I2C.

716/1747

RM0090 Rev 17

RM0090

22.6

Window watchdog (WWDG)

WWDG registers
Refer to Section 2.2 on page 45 for a list of abbreviations used in register descriptions.
The peripheral registers have to be accessed by half-words (16 bits) or words (32 bits).

22.6.1

Control register (WWDG_CR)
Address offset: 0x00
Reset value: 0x0000 007F

31

30

29

28

27

26

25

24

15

14

13

12

11

10

9

8

23

22

21

20

6

5

4

19

18

17

16

3

2

1

0

Reserved

Reserved

7
WDGA

T[6:0]

rs

rw

Bits 31:8 Reserved, must be kept at reset value.
Bit 7 WDGA: Activation bit
This bit is set by software and only cleared by hardware after a reset. When WDGA = 1, the
watchdog can generate a reset.
0: Watchdog disabled
1: Watchdog enabled
Bits 6:0 T[6:0]: 7-bit counter (MSB to LSB)
These bits contain the value of the watchdog counter. It is decremented every (4096 x
2WDGTB[1:0]) PCLK1 cycles. A reset is produced when it rolls over from 0x40 to 0x3F (T6
becomes cleared).

RM0090 Rev 17

717/1747
719

Window watchdog (WWDG)

22.6.2

RM0090

Configuration register (WWDG_CFR)
Address offset: 0x04
Reset value: 0x0000 007F

31

30

29

28

27

26

25

24

23

22

21

20

6

5

4

19

18

17

16

3

2

1

0

Reserved
15

14

13

12

11

10

Reserved

9

8

7

EWI

WDGTB[1:0]

W[6:0]

rs

rw

rw

Bit 31:10 Reserved, must be kept at reset value.
Bit 9 EWI: Early wakeup interrupt
When set, an interrupt occurs whenever the counter reaches the value 0x40. This interrupt is
only cleared by hardware after a reset.
Bits 8:7 WDGTB[1:0]: Timer base
The time base of the prescaler can be modified as follows:
00: CK Counter Clock (PCLK1 div 4096) div 1
01: CK Counter Clock (PCLK1 div 4096) div 2
10: CK Counter Clock (PCLK1 div 4096) div 4
11: CK Counter Clock (PCLK1 div 4096) div 8
Bits 6:0 W[6:0]: 7-bit window value
These bits contain the window value to be compared to the downcounter.

22.6.3

Status register (WWDG_SR)
Address offset: 0x08
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

Reserved
15

14

13

12

11

10

9

8

7

Reserved

EWIF
rc_w0

Bits 31:1 Reserved, must be kept at reset value.
Bit 0 EWIF: Early wakeup interrupt flag
This bit is set by hardware when the counter has reached the value 0x40. It must be cleared
by software by writing ‘0’. A write of ‘1’ has no effect. This bit is also set if the interrupt is not
enabled.

718/1747

RM0090 Rev 17

RM0090

22.6.4

Window watchdog (WWDG)

WWDG register map
The following table gives the WWDG register map and reset values.

WWDG_CR

Reserved

Reserved

Reset value

0x08

WWDG_SR

WDGTB0

WWDG_CFR

EWI

0x04

0
WDGTB1

Reset value

0

0

0

Reserved

Reset value

T[6:0]
1

1

1

1

1

1

1

1

1

W[6:0]
1

1

1

1

1

EWIF

0x00

Register

WDGA

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 110. WWDG register map and reset values

0

Refer to Section 2.3: Memory map for the register boundary addresses.

RM0090 Rev 17

719/1747
719

Cryptographic processor (CRYP)

23

RM0090

Cryptographic processor (CRYP)
This section applies to STM32F415/417xx and STM32F43xxx devices.

23.1

CRYP introduction
The cryptographic processor can be used to both encipher and decipher data using the
DES, Triple-DES or AES (128, 192, or 256) algorithms. It is a fully compliant implementation
of the following standards:
•

The data encryption standard (DES) and Triple-DES (TDES) as defined by Federal
Information Processing Standards Publication (FIPS PUB 46-3, 1999 October 25). It
follows the American National Standards Institute (ANSI) X9.52 standard.

•

The advanced encryption standard (AES) as defined by Federal Information
Processing Standards Publication (FIPS PUB 197, 2001 November 26)

The CRYP processor performs data encryption and decryption using DES and TDES
algorithms in Electronic codebook (ECB) or Cipher block chaining (CBC) mode.
The CRYP peripheral is a 32-bit AHB2 peripheral. It supports DMA transfer for incoming and
processed data, and has input and output FIFOs (each 8 words deep).

23.2

CRYP main features
•

Suitable for AES, DES and TDES enciphering and deciphering operations

•

AES
–

Supports the ECB, CBC, CTR, CCM and GCM chaining algorithms (CCM and
GCM are available on STM32F42xxx and STM32F43xxx only)

–

Supports 128-, 192- and 256-bit keys

–

4 × 32-bit initialization vectors (IV) used in the CBC, CTR, CCM and GCM modes

Table 111. Number of cycles required to process each 128-bit block
(STM32F415/417xx)
Algorithm / Key size

ECB

CBC

CTR

128b

14

14

14

192b

16

16

16

256b

18

18

18

Table 112. Number of cycles required to process each 128-bit block
(STM32F43xxx)
GCM

Algorithm /
Key size

ECB

128b

14

720/1747

CBC

Init
14

CCM

CTR
14

24

Header Payload

Tag

Init

10

14

12

14

RM0090 Rev 17

Header Payload
14

25

Tag
14

RM0090

Cryptographic processor (CRYP)
Table 112. Number of cycles required to process each 128-bit block
(STM32F43xxx)

192b

16

16

16

28

10

16

16

14

16

29

16

256b

18

18

18

32

10

18

18

16

18

33

18

•

•

DES/TDES
–

Direct implementation of simple DES algorithms (a single key, K1, is used)

–

Supports the ECB and CBC chaining algorithms

–

Supports 64-, 128- and 192-bit keys (including parity)

–

2 × 32-bit initialization vectors (IV) used in the CBC mode

–

16 HCLK cycles to process one 64-bit block in DES

–

48 HCLK cycles to process one 64-bit block in TDES

Common to DES/TDES and AES
–

IN and OUT FIFO (each with an 8-word depth, a 32-bit width, corresponding to 4
DES blocks or 2 AES blocks)

–

Automatic data flow control with support of direct memory access (DMA) (using 2
channels, one for incoming data the other for processed data)

–

Data swapping logic to support 1-, 8-, 16- or 32-bit data

RM0090 Rev 17

721/1747
766

Cryptographic processor (CRYP)

23.3

RM0090

CRYP functional description
The cryptographic processor implements a Triple-DES (TDES, that also supports DES) core
and an AES cryptographic core. Section 23.3.1 and Section 23.3.2 provide details on these
cores.
Since the TDES and the AES algorithms use block ciphers, incomplete input data blocks
have to be padded prior to encryption (extra bits should be appended to the trailing end of
the data string). After decryption, the padding has to be discarded. The hardware does not
manage the padding operation, the software has to handle it.
Figure 216 shows the block diagram of the cryptographic processor.
Figure 216. Block diagram (STM32F415/417xx)

ELW$+%EXV
6WDWXV
&5<3B65

&5<3B',1

&5<3B'287

îELW
,1),)2

îELW
287),)2

'0$FRQWUROUHJLVWHU
&5<3B'0$&5
,QWHUUXSWUHJLVWHUV
&5<3B,06&5
&5<3B5,6
&5<3B0,65
&RQWUROUHJLVWHU
VZDSSL QJ

&5<3B&5
,QLWLDOL]DWLRQYHFWRUV
&5<3B,9,9
.H\
&5<3B. .

VZDSSLQ J

,9,9
'(67'(6$(6
NN

3URFHVVRU FRUH
DLE

722/1747

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)
Figure 217. Block diagram (STM32F43xxx)
ELW$+%EXV
6WDWXV
&5<3B65

&5<3B',1

&5<3B'287

'0$FRQWUROUHJLVWHU
&5<3B'0$&5
,QWHUUXSWUHJLVWHUV

îELW
,1),)2

&5<3B,06&5

îELW
287),)2

&5<3B5,6
&5<3B0,65
&RQWUROUHJLVWHU
VZDSSL QJ

&5<3B&5
,QLWLDOL]DWLRQYHFWRUV
&5<3B,9,9
.H\
&5<3B. .

VZDSSLQ J

&RQWH[WVZDSSLQJ

,9,9
'(67'(6$(6

&5<3B&6*&0&&0

3URFHVVRU FRUH

NN

&5<3B&6*&0

069

23.3.1

DES/TDES cryptographic core
The DES/Triple-DES cryptographic core consists of three components:
•

The DES algorithm (DEA)

•

Multiple keys (1 for the DES algorithm, 1 to 3 for the TDES algorithm)

•

The initialization vector (used in the CBC mode)

The basic processing involved in the TDES is as follows: an input block is read in the DEA
and encrypted using the first key, K1 (K0 is not used in TDES mode). The output is then
decrypted using the second key, K2, and encrypted using the third key, K3. The key
depends on the algorithm which is used:
•

DES mode: Key = [K1]

•

TDES mode: Key = [K3 K2 K1]

where Kx=[KxR KxL], R = right, L = left
According to the mode implemented, the resultant output block is used to calculate the
ciphertext.
Note that the outputs of the intermediate DEA stages is never revealed outside the
cryptographic boundary.

RM0090 Rev 17

723/1747
766

Cryptographic processor (CRYP)

RM0090

The TDES allows three different keying options:
•

Three independent keys
The first option specifies that all the keys are independent, that is, K1, K2 and K3 are
independent. FIPS PUB 46-3 – 1999 (and ANSI X9.52 – 1998) refers to this option as
the Keying Option 1 and, to the TDES as 3-key TDES.

•

Two independent keys
The second option specifies that K1 and K2 are independent and K3 is equal to K1,
that is, K1 and K2 are independent, K3 = K1. FIPS PUB 46-3 – 1999 (and ANSI X9.52
– 1998) refers to this second option as the Keying Option 2 and, to the TDES as 2-key
TDES.

•

Three equal keys
The third option specifies that K1, K2 and K3 are equal, that is, K1 = K2 = K3. FIPS
PUB 46-3 – 1999 (and ANSI X9.52 – 1998) refers to the third option as the Keying
Option 3. This “1-key” TDES is equivalent to single DES.

FIPS PUB 46-3 – 1999 (and ANSI X9.52-1998) provides a thorough explanation of the
processing involved in the four operation modes supplied by the TDEA (TDES algorithm):
TDES-ECB encryption, TDES-ECB decryption, TDES-CBC encryption and TDES-CBC
decryption.
This reference manual only gives a brief explanation of each mode.

DES and TDES Electronic codebook (DES/TDES-ECB) mode
•

DES/TDES-ECB mode encryption
Figure 218 illustrates the encryption in DES and TDES Electronic codebook
(DES/TDES-ECB) mode. A 64-bit plaintext data block (P) is used after bit/byte/halfword swapping (refer to Section 23.3.3: Data type on page 739) as the input block (I).
The input block is processed through the DEA in the encrypt state using K1. The output
of this process is fed back directly to the input of the DEA where the DES is performed
in the decrypt state using K2. The output of this process is fed back directly to the input
of the DEA where the DES is performed in the encrypt state using K3. The resultant 64bit output block (O) is used, after bit/byte/half-word swapping, as ciphertext (C) and it is
pushed into the OUT FIFO.

•

DES/TDES-ECB mode decryption
Figure 219 illustrates the DES/TDES-ECB decryption. A 64-bit ciphertext block (C) is
used, after bit/byte/half-word swapping, as the input block (I). The keying sequence is
reversed compared to that used in the encryption process. The input block is
processed through the DEA in the decrypt state using K3. The output of this process is
fed back directly to the input of the DEA where the DES is performed in the encrypt
state using K2. The new result is directly fed to the input of the DEA where the DES is
performed in the decrypt state using K1. The resultant 64-bit output block (O), after
bit/byte/half-word swapping, produces the plaintext (P).

724/1747

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)
Figure 218. DES/TDES-ECB mode encryption
). &)&/
PLAINTEXT 0
0  BITS
$!4!490%

SWAPPING


$%! ENCRYPT

+


$%! DECRYPT

+


$%! ENCRYPT

+

/  BITS
$!4!490%

SWAPPING
#  BITS
/54 &)&/
CIPHERTEXT #
AIB

1. K: key; C: cipher text; I: input block; O: output block; P: plain text.

Figure 219. DES/TDES-ECB mode decryption
). &)&/
CIPHERTEXT #
#  BITS
$!4!490%

SWAPPING
)  BITS



$%! DECRYPT

+


$%! ENCRYPT

+


+

$%! DECRYPT
/  BITS

$!4!490%

SWAPPING
0  BITS
/54 &)&/
PLAINTEXT 0
-36

1. K: key; C: cipher text; I: input block; O: output block; P: plain text.

RM0090 Rev 17

725/1747
766

Cryptographic processor (CRYP)

RM0090

DES and TDES Cipher block chaining (DES/TDES-CBC) mode
•

DES/TDES-CBC mode encryption
Figure 220 illustrates the DES and Triple-DES Cipher block chaining (DES/TDESCBC) mode encryption. This mode begins by dividing a plaintext message into 64-bit
data blocks. In TCBC encryption, the first input block (I1), obtained after bit/byte/halfword swapping (refer to Section 23.3.3: Data type on page 739), is formed by
exclusive-ORing the first plaintext data block (P1) with a 64-bit initialization vector IV (I1
= IV ⊕ P1). The input block is processed through the DEA in the encrypt state using K1.
The output of this process is fed back directly to the input of the DEA, which performs
the DES in the decrypt state using K2. The output of this process is fed directly to the
input of the DEA, which performs the DES in the encrypt state using K3. The resultant
64-bit output block (O1) is used directly as the ciphertext (C1), that is, C1 = O1. This first
ciphertext block is then exclusive-ORed with the second plaintext data block to produce
the second input block, (I2) = (C1 ⊕ P2). Note that I2 and P2 now refer to the second
block. The second input block is processed through the TDEA to produce the second
ciphertext block. This encryption process continues to “chain” successive cipher and
plaintext blocks together until the last plaintext block in the message is encrypted. If the
message does not consist of an integral number of data blocks, then the final partial
data block should be encrypted in a manner specified for the application.

•

DES/TDES-CBC mode decryption
In DES/TDES-CBC decryption (see Figure 221), the first ciphertext block (C1) is used
directly as the input block (I1). The keying sequence is reversed compared to that used
for the encrypt process. The input block is processed through the DEA in the decrypt
state using K3. The output of this process is fed directly to the input of the DEA where
the DES is processed in the encrypt state using K2. This resulting value is directly fed
to the input of the DEA where the DES is processed in the decrypt state using K1. The
resulting output block is exclusive-ORed with the IV (which must be the same as that
used during encryption) to produce the first plaintext block (P1 = O1 ⊕ IV). The second
ciphertext block is then used as the next input block and is processed through the
TDEA. The resulting output block is exclusive-ORed with the first ciphertext block to
produce the second plaintext data block (P2 = O2 ⊕ C1). (Note that P2 and O2 refer to
the second block of data.) The TCBC decryption process continues in this manner until
the last complete ciphertext block has been decrypted. Ciphertext representing a
partial data block must be decrypted in a manner specified for the application.

726/1747

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)
Figure 220. DES/TDES-CBC mode encryption

). &)&/
PLAINTEXT 0
0  BITS
$!4!490%

!(" DATA WRITE
BEFORE #290
IS ENABLED

SWAPPING



0S  BITS

)6,2
)  BITS



$%! ENCRYPT

+
/ IS WRITTEN BACK
INTO )6 AT THE
SAME TIME AS IT
IS PUSHED INTO
THE /54 &)&/


$%! DECRYPT

+


+

$%! ENCRYPT
/  BITS

$!4!490%

SWAPPING
#  BITS
/54 &)&/
CIPHERTEXT #
AIB

1. K: key; C: cipher text; I: input block; O: output block; Ps: plain text before swapping (when decoding) or
after swapping (when encoding); P: plain text; IV: initialization vectors.

RM0090 Rev 17

727/1747
766

Cryptographic processor (CRYP)

RM0090

Figure 221. DES/TDES-CBC mode decryption

). &)&/
CIPHERTEXT #
#  BITS
$!4!490%

SWAPPING
)  BITS


) IS WRITTEN BACK
INTO )6 AT THE
SAME TIME AS 0
IS PUSHED INTO
THE /54 &)&/

$%! DECRYPT

+


$%! ENCRYPT

+


+
!(" DATA WRITE
BEFORE #290
IS ENABLED

$%! DECRYPT



/  BITS

)6,2
0S  BITS
$!4!490%

SWAPPING
0  BITS
/54 &)&/
PLAINTEXT 0
-36

1. K: key; C: cipher text; I: input block; O: output block; Ps: plain text before swapping (when decoding) or
after swapping (when encoding); P: plain text; IV: initialization vectors.

23.3.2

AES cryptographic core
The AES cryptographic core consists of three components:
•

The AES algorithm (AEA: advanced encryption algorithm)

•

Multiple keys

•

Initialization vector(s) or Nonce

The AES utilizes keys of 3 possible lengths: 128, 192 or 256 bits and, depending on the
operation mode used, zero or one 128-bit initialization vector (IV).
The basic processing involved in the AES is as follows: an input block of 128 bits is read
from the input FIFO and sent to the AEA to be encrypted using the key (K0...3). The key
format depends on the key size:
•

If Key size = 128: Key = [K3 K2]

•

If Key size = 192: Key = [K3 K2 K1]

•

If Key size = 256: Key = [K3 K2 K1 K0]

where Kx=[KxR KxL],R=right, L=left
According to the mode implemented, the resultant output block is used to calculate the
ciphertext.
FIPS PUB 197 (November 26, 2001) provides a thorough explanation of the processing
involved in the four operation modes supplied by the AES core: AES-ECB encryption, AES-

728/1747

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)
ECB decryption, AES-CBC encryption and AES-CBC decryption.This reference manual
only gives a brief explanation of each mode.

AES Electronic codebook (AES-ECB) mode
•

AES-ECB mode encryption
Figure 222 illustrates the AES Electronic codebook (AES-ECB) mode encryption.
In AES-ECB encryption, a 128- bit plaintext data block (P) is used after bit/byte/halfword swapping (refer to Section 23.3.3: Data type on page 739) as the input block (I).
The input block is processed through the AEA in the encrypt state using the 128, 192 or
256-bit key. The resultant 128-bit output block (O) is used after bit/byte/half-word
swapping as ciphertext (C). It is then pushed into the OUT FIFO.

•

AES-ECB mode decryption
Figure 223 illustrates the AES Electronic codebook (AES-ECB) mode encryption.
To perform an AES decryption in the ECB mode, the secret key has to be prepared (it is
necessary to execute the complete key schedule for encryption) by collecting the last
round key, and using it as the first round key for the decryption of the ciphertext. This
preparation function is computed by the AES core. Refer to Section 23.3.6: Procedure
to perform an encryption or a decryption for more details on how to prepare the key.
In AES-ECB decryption, a 128-bit ciphertext block (C) is used after bit/byte/half-word
swapping as the input block (I). The keying sequence is reversed compared to that of
the encryption process. The resultant 128-bit output block (O), after bit/byte or halfword swapping, produces the plaintext (P).
Figure 222. AES-ECB mode encryption
). &)&/
PLAINTEXT 0
0  BITS
$!4!490%

SWAPPING
)  BITS


OR 

+ 

!%! ENCRYPT

$!4!490%

SWAPPING
#  BITS

/54 &)&/
CIPHERTEXT #
AIB

1. K: key; C: cipher text; I: input block; O: output block; P: plain text.
2. If Key size = 128: Key = [K3 K2].
If Key size = 192: Key = [K3 K2 K1]
If Key size = 256: Key = [K3 K2 K1 K0].

RM0090 Rev 17

729/1747
766

Cryptographic processor (CRYP)

RM0090
Figure 223. AES-ECB mode decryption
). &)&/
CIPHERTEXT #
#  BITS
$!4!490%


OR 

+ 

SWAPPING
)  BITS

!%! DECRYPT
/  BITS

$!4!490%

SWAPPING
0  BITS

/54 &)&/
PLAINTEXT 0
-36

1. K: key; C: cipher text; I: input block; O: output block; P: plain text.
2. If Key size = 128 => Key = [K3 K2].
If Key size = 192 => Key = [K3 K2 K1]
If Key size = 256 => Key = [K3 K2 K1 K0].

AES Cipher block chaining (AES-CBC) mode
•

AES-CBC mode encryption
The AES Cipher block chaining (AES-CBC) mode decryption is shown on Figure 224.
In AES-CBC encryption, the first input block (I1) obtained after bit/byte/half-word
swapping (refer to Section 23.3.3: Data type on page 739) is formed by exclusiveORing the first plaintext data block (P1) with a 128-bit initialization vector IV (I1 = IV ⊕
P1). The input block is processed through the AEA in the encrypt state using the 128-,
192- or 256-bit key (K0...K3). The resultant 128-bit output block (O1) is used directly as
ciphertext (C1), that is, C1 = O1. This first ciphertext block is then exclusive-ORed with
the second plaintext data block to produce the second input block, (I2) = (C1 ⊕ P2).
Note that I2 and P2 now refer to the second block. The second input block is processed
through the AEA to produce the second ciphertext block. This encryption process
continues to “chain” successive cipher and plaintext blocks together until the last
plaintext block in the message is encrypted. If the message does not consist of an
integral number of data blocks, then the final partial data block should be encrypted in a
manner specified for the application.
In the CBC mode, like in the ECB mode, the secret key must be prepared to perform an
AES decryption. Refer to Section 23.3.6: Procedure to perform an encryption or a
decryption on page 744 for more details on how to prepare the key.

•

AES-CBC mode decryption
In AES-CBC decryption (see Figure 225), the first 128-bit ciphertext block (C1) is used
directly as the input block (I1). The input block is processed through the AEA in the
decrypt state using the 128-, 192- or 256-bit key. The resulting output block is
exclusive-ORed with the 128-bit initialization vector IV (which must be the same as that
used during encryption) to produce the first plaintext block (P1 = O1 ⊕ IV). The second
ciphertext block is then used as the next input block and is processed through the AEA.
The resulting output block is exclusive-ORed with the first ciphertext block to produce
the second plaintext data block (P2 = O2 ⊕ C1). (Note that P2 and O2 refer to the second

730/1747

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)
block of data.) The AES-CBC decryption process continues in this manner until the last
complete ciphertext block has been decrypted. Ciphertext representing a partial data
block must be decrypted in a manner specified for the application.
Figure 224. AES-CBC mode encryption
). &)&/
PLAINTEXT 0
0  BITS
$!4!490%
!(" DATA WRITE
BEFORE #290
IS ENABLED

SWAPPING


0S  BITS

)6;)6 )6=

+ 

 
OR 

)  BITS

!%! ENCRYPT

/ IS WRITTEN
BACK INTO )6
AT THE SAME TIME
AS IT IS PUSHED
INTO THE /54 &)&/

/  BITS
$!4!490%

SWAPPING
#  BITS
/54 &)&/
CIPHERTEXT #
AIB

1. K: key; C: cipher text; I: input block; O: output block; Ps: plain text before swapping (when decoding) or
after swapping (when encoding); P: plain text; IV: Initialization vectors.
2. IVx=[IVxR IVxL], R=right, L=left.
3. If Key size = 128 => Key = [K3 K2].
If Key size = 192 => Key = [K3 K2 K1]
If Key size = 256 => Key = [K3 K2 K1 K0].

RM0090 Rev 17

731/1747
766

Cryptographic processor (CRYP)

RM0090
Figure 225. AES-CBC mode decryption
). &)&/
CIPHERTEXT #
#  BITS
$!4!490%

+ 

SWAPPING
)  BITS

 
OR 

!%! DECRYPT

) IS WRITTEN
BACK INTO )6
AT THE SAME TIME
AS 0 IS PUSHED
INTO THE /54 &)&/
!(" DATA WRITE
BEFORE #290
IS ENABLED



)6;)6 )6=
$!4!490%

/  BITS

0S  BITS

SWAPPING
0  BITS

/54 &)&/
PLAINTEXT 0
-36

1. K: key; C: cipher text; I: input block; O: output block; Ps: plain text before swapping (when decoding) or
after swapping (when encoding); P: plain text; IV: Initialization vectors.
2. IVx=[IVxR IVxL], R=right, L=left.
3. If Key size = 128 => Key = [K3 K2].
If Key size = 192 => Key = [K3 K2 K1]
If Key size = 256 => Key = [K3 K2 K1 K0].

AES counter mode (AES-CTR) mode
The AES counter mode uses the AES block as a key stream generator. The generated keys
are then XORed with the plaintext to obtain the cipher. For this reason, it makes no sense to
speak of different CTR encryption/decryption, since the two operations are exactly the
same.
In fact, given:
•

Plaintext: P[0], P[1], ..., P[n] (128 bits each)

•

A key K to be used (the size does not matter)

•

An initial counter block (call it ICB but it has the same functionality as the IV of CBC)

The cipher is computed as follows:
C[i] = enck(iv[i]) xor P[i], where:
iv[0] = ICB and iv[i+1] = func(iv[i]), where func is an update function
applied to the previous iv block; func is basically an increment of one of the fields
composing the iv block.
Given that the ICB for decryption is the same as the one for encryption, the key stream
generated during decryption is the same as the one generated during encryption. Then, the
ciphertext is XORed with the key stream in order to retrieve the original plaintext. The
decryption operation therefore acts exactly in the same way as the encryption operation.

732/1747

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)
Figure 226 and Figure 227 illustrate AES-CTR encryption and decryption, respectively.
Figure 226. AES-CTR mode encryption
). &)&/
PLAINTEXT 0
3ELWV
$+%GDWDZULWH
EHIRUH&5<3
LVHQDEOHG

'$7$7<3(

SWAPPING

)6,2
3VELWV



+

,ELWV

 
OR 

!%! ENCRYPT

, LVZULWWHQ
EDFNLQWR,9
DWVDPHWLPH
WKDQ&LVSXVKHG
LQ287),)2

2ELWV

&VELW
'$7$7<3(

SWAPPING
&ELWV

/54 &)&/
CIPHERTEXT #
AIB

1. K: key; C: cipher text; I: input Block; o: output block; Ps: plain text before swapping (when decoding) or
after swapping (when encoding); Cs: cipher text after swapping (when decoding) or before swapping (when
encoding); P: plain text; IV: Initialization vectors.

RM0090 Rev 17

733/1747
766

Cryptographic processor (CRYP)

RM0090
Figure 227. AES-CTR mode decryption
). &)&/
CIPHERTEXT 0
&ELWV
'$7$7<3(

$+%GDWDZULWH
EHIRUH&5<3
LVHQDEOHG

SWAPPING

)6,2
&VELWV



+

,ELWV

RU

, LVZULWWHQ
EDFNLQWR,9
DWVDPHWLPH
WKDQ3LVSXVKHG
LQ287),)2

!%! ENCRYPT

2ELWV

3VELWV
'$7$7<3(

SWAPPING
3ELWV

/54 &)&/
PLAINTEXT #
-36

1. K: key; C: cipher text; I: input Block; o: output block; Ps: plain text before swapping (when decoding) or
after swapping (when encoding); Cs: cipher text after swapping (when decoding) or before swapping (when
encoding); P: plain text; IV: Initialization vectors.

Figure 228 shows the structure of the IV block as defined by the standard [2]. It is composed
of three distinct fields.
Figure 228. Initial counter block structure for the Counter mode
.ONCE

 BITS

)NITIALIZATION VECTOR )6

 BITS

#OUNTER

 BITS
AI

•

Nonce is a 32-bit, single-use value. A new nonce should be assigned to each different
communication.

•

The initialization vector (IV) is a 64-bit value and the standard specifies that the
encryptor must choose IV so as to ensure that a given value is used only once for a
given key

•

The counter is a 32-bit big-endian integer that is incremented each time a block has
been encrypted. The initial value of the counter should be set to ‘1’.

The block increments the least significant 32 bits, while it leaves the other (most significant)
96 bits unchanged.

734/1747

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)

AES Galois/counter mode (GCM)
The AES Galois/counter mode (GCM) allows encrypting and authenticating the plaintext,
and generating the correspondent ciphertext and tag (also known as message
authentication code or message integrity check). This algorithm is based on AES counter
mode to ensure confidentiality. It uses a multiplier over a fixed finite field to generate the tag.
An initialization vector is required at the beginning of the algorithm.
The message to be processed is split into 2 parts:

Note:

•

The header (also knows as additional authentication data): data which is authenticated
but no protected (such as information for routing the packet)

•

The payload (also knows as plaintext or ciphertext): the message itself which is
authenticated and encrypted.

The header must precede the payload and the two parts cannot be mixed together.
The GCM standard requires to pass, at the end of the message, a specific 128-bit block
composed of the size of the header (64 bits) and the size of the payload (64 bits). During the
computation, the header blocks must be distinguished from the payload blocks.
In GCM mode, four steps are required to perform an encryption/decryption:

RM0090 Rev 17

735/1747
766

Cryptographic processor (CRYP)
1.

RM0090

GCM init phase
During this first step, the HASH key is calculated and saved internally to be used for
processing all the blocks. It is recommended to follow the sequence below:

2.

a)

Make sure that the cryptographic processor is disabled by clearing the CRYPEN
bit in the CRYP_CR register.

b)

Select the GCM chaining mode by programming ALGOMODE bits to ‘01000’ in
CRYP_CR.

c)

Configure GCM_CCMPH bits to ‘00’ in CRYP_CR to start the GCM Init phase.

d)

Initialize the key registers (128,192 and 256 bits) in CRYP_KEYRx as well as the
initialization vector (IV).

e)

Set CRYPEN bit to ‘1’ to start the calculation of the HASH key.

f)

Wait for the CRYPEN bit to be cleared to ‘0’ before moving on to the next phase.

g)

Set the CRYPEN bit to ‘1’.

GCM header phase
This step must be performed after the GCM Init phase:

3.

h)

Set the GCM_CCMPH bits to ‘01’ in CRYP_CR to indicate that the header phase
has started.

i)

Write the header data. Three methods can be used:

–

Program the data by blocks of 32 bits into the CRYP_DIN register, and use the
IFNF flag to determine if the input FIFO can receive data. The size of the header
must be a multiple of 128 bits (4 words).

–

Program the data into the CRYP_DIN register by blocks of 8 words, and use the
IFEM flag to determine if the input FIFO can receive data (IFEM=’1’). The size of
the header must be a multiple of 128 bits (4 words).

–

Use the DMA.

j)

Once all header data have been supplied, wait until the BUSY bit is cleared in the
CRYP_SR register.

GCM payload phase (encryption/decryption)
This step must be performed after the GCM header phase:
k)

Configure GCM_CCMPH to ‘10’ in the CRYP_CR register.

l)

Select the algorithm direction (encryption or decryption) by using the ALGODIR bit
in CRYP_CR.

m) Program the payload message into the CRYP_DIN register, and use the IFNF flag
to determine if the input FIFO can receive data. Alternatively, the data could be
programmed into the CRYP_DIN register by blocks of 8 words and the IFEM flag
used to determine if the input FIFO can receive data (IFEM=’1’). In parallel, the

736/1747

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)
OFNE/OFFU flag of the CRYP_DOUT register can be monitored to check if the
output FIFO is not empty.
n)
4.

Repeat the previous step until all payload blocks have been encrypted or
decrypted. Alternatively, DMA could be used.

GCM final phase
This step generates the authentication tag:

Note:

o)

Configure GCM_CCMPH[1:0] to ‘11’ in CRYP_CR.

p)

Write the input into the CRYP_DIN register 4 times. The input must contain the
number of bits in the header (64 bits) concatenated with the number of bits in the
payload (64 bits).

q)

Wait till the OFNE flag (FIFO output not empty) is set to ‘1’ in the CRYP_SR
register.

r)

Read the CRYP_DOUT register 4 times: the output corresponds to the
authentication tag.

s)

Disable the cryptographic processor (CRYPEN bit in CRYP_CR = ‘0’)

When a decryption is performed, it is not required to compute the key at the beginning. At
the end of the decryption, the generated tag should be compared with the expected tag
passed with the message. In addition, the ALGODIR bit (algorithm direction) must be set to ‘1’.
No need to disable/enable CRYP processor when moving from header phase to tag phase.

AES Galois message authentication code (GMAC)
The cryptographic processor also supports GMAC to authenticate the plaintext. It uses the
GCM algorithm and a multiplier over a fixed finite field to generate the corresponding tag.
An initialization vector is required at the beginning of the algorithm.
Actually, the GMAC algorithm corresponds to the GCM algorithm applied on a message
composed of the header only. As a consequence, the payload phase is not required.

AES combined cipher machine (CCM)
The CCM algorithm allows encrypting and authenticating the plaintext, as well as generating
the corresponding ciphertext and tag (also known as message authentication code or
message integrity check). This algorithm is based on AES counter mode to ensure
confidentiality. It uses the AES CBC mode to generate a 128-bit tag.
The CCM standard (RFC 3610 Counter with CBC-MAC (CCM) dated September 2003)
defines particular encoding rules for the first authentication block (called B0 in the standard).
In particular, the first block includes flags, a nonce and the payload length expressed in
bytes. The CCM standard specifies another format, called A or counter, for
encryption/decryption. The counter is incremented during the payload phase and its 32 LSB
bits are initialized to ‘1’ during the tag generation (called A0 packet in the CCM standard).
Note:

The hardware does not perform the formatting operation of the B0 packet. It should be
handled by the software.
As for the GCM algorithm, the message to be processed is split into 2 parts:
•

The header (also knows as additional authentication data): data which is authenticated
but no protected (such as information for routing the packet)

•

The payload (also knows as plaintext or ciphertext): the message itself which is
authenticated and encrypted.

RM0090 Rev 17

737/1747
766

Cryptographic processor (CRYP)
Note:

RM0090

The header part must precede the payload and the two parts cannot be mixed together.
In CCM mode, 4 steps are required to perform and encryption or decryption:
1.

CCM init phase
In this first step, the B0 packet of the CCM message (1st packet) is programmed into
the CRYP_DIN register. During this phase, the CRYP_DOUT register does not contain
any output data.
The following sequence must be followed:

2.

a)

Make sure that the cryptographic processor is disabled by clearing the CRYPEN
bit in the CRYP_CR register.

b)

Select the CCM chaining mode by programming the ALGOMODE bits to ‘01001’
in the CRYP_CR register.

c)

Configure the GCM_CCMPH bits to ‘00’ in CRYP_CR to start the CCM Init phase.

d)

Initialize the key registers (128,192 and 256 bits) in CRYP_KEYRx as well as the
initialization vector (IV).

e)

Set the CRYPEN bit to ‘1’ in CRYP_CR.

f)

Program the B0 packet into the input data register.

g)

Wait for the CRYPEN bit to be cleared before moving on to the next phase.

h)

Set CRYPEN to ‘1’.

CCM header phase
This step must be performed after the CCM Init phase. The sequence is identical for
encryption and decryption.
During this phase, the CRYP_DOUT register does not contain any output data.
This phase can be skipped if there is no additional authenticated data.
The following sequence must be followed:

738/1747

i)

Set the GCM_CCMPH bit to ’01’ in CRYP_CR to indicate that the header phase
has started.

j)

Three methods can be used:

–

Program the header data by blocks of 32 bits into the CRYP_DIN register, and use
the IFNF flag to determine if the input FIFO can receive data. The size of the
header must be a multiple of 128 bits (4 words).

–

Program the header data into the CRYP_DIN register by blocks of 8 words, and
use the IFEM flag to determine if the input FIFO can receive data (IFEM=’1’). The
size of the header must be a multiple of 128 bits (4 words).

–

Use the DMA.

RM0090 Rev 17

RM0090
Note:

Cryptographic processor (CRYP)
The first block B1 must be formatted with the header length. This task should be handled by
software.
k)
3.

Once all header data have been supplied, wait until the BUSY flag is cleared.

CCM payload phase (encryption/decryption)
This step must be performed after the CCM header phase. During this phase, the
encrypted/decrypted payload is stored in the CRYP_DOUT register.
The following sequence must be followed:
l)

Configure GCM_CCMPH bits to ‘10’ in CRYP_CR.

m) Select the algorithm direction (encryption or decryption) by using the ALGODIR bit
in CRYP_CR.

4.

n)

Program the payload message into the CRYP_DIN register, and use the IFNF flag
to determine if the input FIFO can receive data. Alternatively, the data could be
programmed into the CRYP_DIN register by blocks of 8 words and the IFEM flag
used to determine if the input FIFO can receive data (IFEM=’1’). In parallel, the
OFNE/OFFU flag of the CRYP_DOUT register can be monitored to check if the
output FIFO is not empty.

o)

Repeat the previous step until all payload blocks have been encrypted or
decrypted. Alternatively, DMA could be used.

CCM final phase
This step generates the authentication tag. During this phase, the authentication tag of
the message is generated and stored in the CRYP_DOUT register.

Note:

p)

Configure GCM_CCMPH[1:0] bits to ‘11’ in CRYP_CR.

q)

Load the A0 initialized counter, and program the 128-bit A0 value by writing 4
times 32 bits into the CRYP_DIN register.

r)

Wait till the OFNE flag (FIFO output not empty) is set to ‘1’ in the CRYP_SR
register.

s)

Read the CRYP_DOUT register 4 times: the output corresponds to the encrypted
authentication tag.

t)

Disable the cryptographic processor (CRYPEN bit in CRYP_CR = ‘0’)

The hardware does not perform the formatting of the original B0 and B1 packets and the tag
comparison between encryption and decryption. They have to be handled by software.
The cryptographic processor does not need to be disabled/enabled when moving from the
header phase to the tag phase.
AES cipher message authentication code (CMAC)
The CMAC algorithm allows authenticating the plaintext, and generating the corresponding
tag. The CMAC sequence is identical to the CCM one, except that the payload phase is
skipped.

23.3.3

Data type
Data enter the CRYP processor 32 bits (word) at a time as they are written into the
CRYP_DIN register. The principle of the DES is that streams of data are processed 64 bits
by 64 bits and, for each 64-bit block, the bits are numbered from M1 to M64, with M1 the leftmost bit and M64 the right-most bit of the block. The same principle is used for the AES, but
with a 128-bit block size.

RM0090 Rev 17

739/1747
766

Cryptographic processor (CRYP)

RM0090

The system memory organization is little-endian: whatever the data type (bit, byte, 16-bit
half-word, 32-bit word) used, the least-significant data occupy the lowest address locations.
A bit, byte, or half-word swapping operation (depending on the kind of data to be encrypted)
therefore has to be performed on the data read from the IN FIFO before they enter the
CRYP processor. The same swapping operation should be performed on the CRYP data
before they are written into the OUT FIFO. For example, the operation would be byte
swapping for an ASCII text stream.
The kind of data to be processed is configured with the DATATYPE bitfield in the CRYP
control register (CRYP_CR).
Table 113. Data types
DATATYPE in
CRYP_CR

System memory data
(plaintext or cypher)

Swapping performed

Example: TDES block value 0xABCD77206973FE01 is
represented in system memory as:
00b

No swapping

V\VWHPPHPRU\

7'(6EORFNVL]H ELW [ELW

[$%&'
[)(

[$%&')(

#
#



Example: TDES block value 0xABCD77206973FE01 is
represented in system memory as:
01b

Half-word (16-bit)
swapping

V\VWHPPHPRU\
7'(6EORFNVL]H ELW [ELW

[$%&'
[)(

[$%&')(

#
#

Example: TDES block value 0xABCD77206973FE01 is
represented in system memory as:
10b

Byte (8-bit) swapping

V\VWHPPHPRU\
7'(6EORFNVL]H ELW [ELW

[$%&')(

[&'$%
[)(

#
#



TDES block value 0x4E6F772069732074 is represented in system
memory as:

7'(6%ORFVL]H ELW [ELW

11b

Bit swapping

V\VWHPPHPRU\

[()









[(()

#

[(&(

#

  #



 #


Figure 229 shows how the 64-bit data block M1...64 is constructed from two consecutive 32bit words popped off the IN FIFO by the CRYP processor, according to the DATATYPE
value. The same schematic can easily be extended to form the 128-bit block for the AES

740/1747

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)
cryptographic algorithm (for the AES, the block length is four 32-bit words, but swapping
only takes place at word level, so it is identical to the one described here for the TDES).

Note:

The same swapping is performed between the IN FIFO and the CRYP data block, and
between the CRYP data block and the OUT FIFO.
Figure 229. 64-bit block construction according to DATATYPE
$!4!490%  B

BIT SWAPPING OPERATION
BIT 

BIT 

BIT 

BIT 

BIT 

SECOND WORD WRITTEN INTO THE #290?$). REGISTER

). &)&/

BIT 

BIT 

BIT 

BIT 

BIT 

FIRST WORD WRITTEN INTO THE #290?$). REGISTER

BIT STRING

-

-

-

-

-

-

-

-

-

$!4!490%  B

BYTE SWAPPING OPERATION

). &)&/

-

"YTE   "YTE   "YTE  
"YTE  
BITS  BITS  BITS  BITS 
"YTE   "YTE   "YTE  
"YTE  
BITS  BITS  BITS  BITS 

SECOND WORD WRITTEN INTO THE #290?$). REGISTER
FIRST WORD WRITTEN INTO THE #290?$). REGISTER
BIT ORDERING WITHIN BYTE IS UNCHANGED

BIT STRING

"YTE   "YTE   "YTE  
"YTE   "YTE   "YTE   "YTE  
"YTE  
-
-
- - - -
- -

HALD WORD SWAPPING OPERATION

). &)&/

BIT STRING

$!4!490%  B

(ALF WORD  
BITS 
(ALF WORD  
BITS 

(ALF WORD  
BITS 
(ALF WORD  
BITS 

(ALF WORD  
-

(ALF WORD  
-

.O SWAPPING OPERATION

). &)&/

BIT STRING

SECOND WORD WRITTEN INTO THE #290?$). REGISTER
FIRST WORD WRITTEN INTO THE #290?$). REGISTER

(ALF WORD  
-

(ALF WORD  
-
$!4!490%  B

7ORD 
BITS 
7ORD 
BITS 

7ORD 
-

SECOND WORD WRITTEN INTO THE #290?$). REGISTER
FIRST WORD WRITTEN INTO THE #290?$). REGISTER

7ORD 
-
AI

RM0090 Rev 17

741/1747
766

Cryptographic processor (CRYP)

23.3.4

RM0090

Initialization vectors - CRYP_IV0...1(L/R)
Initialization vectors are considered as two 64-bit data items. They therefore do not have the
same data format and representation in system memory as plaintext or cypher data, and
they are not affected by the DATATYPE value.
Initialization vectors are defined by two consecutive 32-bit words, CRYP_IVL (left part,
noted as bits IV1...32) and CRYP_IVR (right part, noted as bits IV33...64).
During the DES or TDES CBC encryption, the CRYP_IV0(L/R) bits are XORed with the 64bit data block popped off the IN FIFO after swapping (according to the DATATYPE value),
that is, with the M1...64 bits of the data block. When the output of the DEA3 block is
available, it is copied back into the CRYP_IV0(L/R) vector, and this new content is XORed
with the next 64-bit data block popped off the IN FIFO, and so on.
During the DES or TDES CBC decryption, the CRYP_IV0(L/R) bits are XORed with the 64bit data block (that is, with the M1...64 bits) delivered by the TDEA1 block before swapping
(according to the DATATYPE value), and pushed into the OUT FIFO. When the XORed
result is swapped and pushed into the OUT FIFO, the CRYP_IV0(L/R) value is replaced by
the output of the IN FIFO, then the IN FIFO is popped, and a new 64-bit data block can be
processed.
During the AES CBC encryption, the CRYP_IV0...1(L/R) bits are XORed with the 128-bit
data block popped off the IN FIFO after swapping (according to the DATATYPE value).
When the output of the AES core is available, it is copied back into the CRYP_IV0...1(L/R)
vector, and this new content is XORed with the next 128-bit data block popped off the IN
FIFO, and so on.
During the AES CBC decryption, the CRYP_IV0...1(L/R) bits are XORed with the 128-bit
data block delivered by the AES core before swapping (according to the DATATYPE value)
and pushed into the OUT FIFO. When the XORed result is swapped and pushed into the
OUT FIFO, the CRYP_IV0...1(L/R) value is replaced by the output of the IN FIFO, then the
IN FIFO is popped, and a new 128-bit data block can be processed.
During the AES CTR encryption or decryption, the CRYP_IV0...1(L/R) bits are encrypted by
the AES core. Then the result of the encryption is XORed with the 128-bit data block popped
off the IN FIFO after swapping (according to the DATATYPE value). When the XORed result
is swapped and pushed into the OUT FIFO, the counter part of the CRYP_IV0...1(L/R) value
(32 LSB) is incremented.
Any write operation to the CRYP_IV0...1(L/R) registers when bit BUSY = 1b in the
CRYP_SR register is disregarded (CRYP_IV0...1(L/R) register content not modified). Thus,
you must check that bit BUSY = 0b before modifying initialization vectors.

742/1747

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)
Figure 230. Initialization vectors use in the TDES-CBC encryption
4$%3 #"# EN CRYPTION EX AMPLE $!4!490%  B

BIT STRING

BIT  BIT 

BIT 

BIT 

BIT  BIT 

BIT 

BIT  BIT 

#290?)6,





)6 )6

)

SECOND WORD WRITTEN INTO THE #290?$). REGISTER
FIRST WORD WRITTEN INTO THE #290?$). REGISTER

- - - - -

- -



BIT 







)



#290?)62



)6 )6 )6 )6 )6

)

)

) )

$%! %NCRYPT



- - -

+

$%! %NCRYPT

+

)6 )6











)6 )6 )6

)

) )

+

$%! $ECRYPT

#290?)6,


)



#290?)62


)6 )6 )6 )6 )6







)6 )6 )6

#290 RESULT IS COPIED
BACK TO THE #290?)6,2
REGISTERS AFTER CYPHERING

/54 &)&/
&IRST WORD FROM THE /54 &)&/ CONTAINS THE LEFT PART OF THE CYPHERTEXT BLOCK /
3ECOND WORD FROM /54 &)&/ CONTAINS THE RIGHT PART OF CYPHERTEXT BLOCK /
AI

23.3.5

CRYP busy state
When there is enough data in the input FIFO (at least 2 words for the DES or TDES
algorithm mode, 4 words for the AES algorithm mode) and enough free-space in the output
FIFO (at least 2 (DES/TDES) or 4 (AES) word locations), and when the bit CRYPEN = 1 in
the CRYP_CR register, then the cryptographic processor automatically starts an encryption
or decryption process (according to the value of the ALGODIR bit in the CRYP_CR
register).
This process takes 48 AHB2 clock cycles for the Triple-DES algorithm, 16 AHB2 clock
cycles for the simple DES algorithm, and 14, 16 or 18 AHB2 clock cycles for the AES with
key lengths of 128, 192 or 256 bits, respectively. During the whole process, the BUSY bit in
the CRYP_SR register is set to ‘1’. At the end of the process, two (DES/TDES) or four (AES)
words are written by the CRYP Core into the output FIFO, and the BUSY bit is cleared. In

RM0090 Rev 17

743/1747
766

Cryptographic processor (CRYP)

RM0090

the CBC, CTR mode, the initialization vectors CRYP_IVx(L/R)R (x = 0..3) are updated as
well.
A write operation to the key registers (CRYP_Kx(L/R)R, x = 0..3), the initialization registers
(CRYP_IVx(L/R)R, x = 0..3), or to bits [9:2] in the CRYP_CR register are ignored when the
cryptographic processor is busy (bit BUSY = 1b in the CRYP_SR register), and the registers
are not modified. It is thus not possible to modify the configuration of the cryptographic
processor while it is processing a block of data. It is however possible to clear the CRYPEN
bit while BUSY = 1, in which case the ongoing DES, TDES or AES processing is completed
and the two/four word results are written into the output FIFO, and then, only then, the
BUSY bit is cleared.
Note:

When a block is being processed in the DES or TDES mode, if the output FIFO becomes full
and if the input FIFO contains at least one new block, then the new block is popped off the
input FIFO and the BUSY bit remains high until there is enough space to store this new
block into the output FIFO.

23.3.6

Procedure to perform an encryption or a decryption
Initialization
1.

2.

Initialize the peripheral (the order of operations is not important except for the key
preparation for AES-ECB or AES-CBC decryption. The key size and the key value
must be entered before preparing the key and the algorithm must be configured once
the key has been prepared):
a)

Configure the key size (128-, 192- or 256-bit, in the AES only) with the KEYSIZE
bits in the CRYP_CR register

b)

Write the symmetric key into the CRYP_KxL/R registers (2 to 8 registers to be
written depending on the algorithm)

c)

Configure the data type (1-, 8-, 16- or 32-bit), with the DATATYPE bits in the
CRYP_CR register

d)

In case of decryption in AES-ECB or AES-CBC, you must prepare the key:
configure the key preparation mode by setting the ALGOMODE bits to ‘111’ in the
CRYP_CR register. Then write the CRYPEN bit to ‘1’: the BUSY bit is set. Wait
until BUSY returns to 0 (CRYPEN is automatically cleared as well): the key is
prepared for decryption

e)

Configure the algorithm and chaining (the DES/TDES in ECB/CBC, the AES in
ECB/CBC/CTR/GCM/CCM) with the ALGOMODE bits in the CRYP_CR register

f)

Configure the direction (encryption/decryption), with the ALGODIR bit in the
CRYP_CR register

g)

Write the initialization vectors into the CRYP_IVxL/R register (in CBC or CTR
modes only)

Flush the IN and OUT FIFOs by writing the FFLUSH bit to 1 in the CRYP_CR register

Processing when the DMA is used to transfer the data from/to the memory
1.

744/1747

Configure the DMA controller to transfer the input data from the memory. The transfer
length is the length of the message. As message padding is not managed by the
peripheral, the message length must be an entire number of blocks. The data are
transferred in burst mode. The burst length is 4 words in the AES and 2 or 4 words in

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)
the DES/TDES. The DMA should be configured to set an interrupt on transfer
completion of the output data to indicate that the processing is finished.
2.

Enable the cryptographic processor by writing the CRYPEN bit to 1. Enable the DMA
requests by setting the DIEN and DOEN bits in the CRYP_DMACR register.

3.

All the transfers and processing are managed by the DMA and the cryptographic
processor. The DMA interrupt indicates that the processing is complete. Both FIFOs
are normally empty and BUSY = 0.

Processing when the data are transferred by the CPU during interrupts
1.

Enable the interrupts by setting the INIM and OUTIM bits in the CRYP_IMSCR register.

2.

Enable the cryptographic processor by setting the CRYPEN bit in the CRYP_CR
register.

3.

In the interrupt managing the input data: load the input message into the IN FIFO. You
can load 2 or 4 words at a time, or load data until the FIFO is full. When the last word of
the message has been entered into the FIFO, disable the interrupt by clearing the INIM
bit.

4.

In the interrupt managing the output data: read the output message from the OUT
FIFO. You can read 1 block (2 or 4 words) at a time or read data until the FIFO is
empty. When the last word has been read, INIM=0, BUSY=0 and both FIFOs are empty
(IFEM=1 and OFNE=0). You can disable the interrupt by clearing the OUTIM bit and,
the peripheral by clearing the CRYPEN bit.

Processing without using the DMA nor interrupts
1.

Enable the cryptographic processor by setting the CRYPEN bit in the CRYP_CR
register.

2.

Write the first blocks in the input FIFO (2 to 8 words).

3.

4.

23.3.7

Repeat the following sequence until the complete message has been processed:
a)

Wait for OFNE=1, then read the OUT-FIFO (1 block or until the FIFO is empty)

b)

Wait for IFNF=1, then write the IN FIFO (1 block or until the FIFO is full)

At the end of the processing, BUSY=0 and both FIFOs are empty (IFEM=1 and
OFNE=0). You can disable the peripheral by clearing the CRYPEN bit.

Context swapping
If a context switching is needed because a new task launched by the OS requires this
resource, the following tasks have to be performed for full context restoration (example
when the DMA is used):

RM0090 Rev 17

745/1747
766

Cryptographic processor (CRYP)

RM0090

Case of the AES and DES
1.

Context saving
a)

Stop DMA transfers on the IN FIFO by clearing the DIEN bit in the CRYP_DMACR
register.

b)

Wait until both the IN and OUT FIFOs are empty (IFEM=1 and OFNE=0 in the
CRYP_SR register) and the BUSY bit is cleared.

c)

Stop DMA transfers on the OUT FIFO by writing the DOEN bit to 0 in the
CRYP_DMACR register and clear the CRYPEN bit.

d)

Save the current configuration (bits [9:2] and bits 19 in the CRYP_CR register)
and, if not in ECB mode, the initialization vectors. The key value must already be
available in the memory. When needed, save the DMA status (pointers for IN and
OUT messages, number of remaining bytes, etc.).
Additional bits should be saved when GCM/GMAC or CCM/CMAC algorithms are
used:

–

bits [17:16] in the CRYP_CR register

–

context swap registers:
CRYP_CSGCMCCM0..7 for GCM/GMAC or CCM/CMAC algorithm
CRYP_CSGCM0..7 for GCM/GMAC algorithm.

2.
3.

Configure and execute the other processing.
Context restoration
a)

Configure the processor as in Section 23.3.6: Procedure to perform an encryption
or a decryption on page 744, Initialization with the saved configuration. For the
AES-ECB or AES-CBC decryption, the key must be prepared again.

b)

If needed, reconfigure the DMA controller to transfer the rest of the message.

c)

Enable the processor by setting the CRYPEN bit and, the DMA requests by setting
the DIEN and DOEN bits.

Case of the TDES
Context swapping can be done in the TDES in the same way as in the AES. But as the input
FIFO can contain up to 4 unprocessed blocks and as the processing duration per block is
higher, it can be faster in certain cases to interrupt the processing without waiting for the IN
FIFO to be empty.
1.

746/1747

Context saving
a)

Stop DMA transfers on the IN FIFO by clearing the DIEN bit in the CRYP_DMACR
register.

b)

Disable the processor by clearing the CRYPEN bit (the processing will stop at the
end of the current block).

c)

Wait until the OUT FIFO is empty (OFNE=0 in the CRYP_SR register) and the
BUSY bit is cleared.

d)

Stop DMA transfers on the OUT FIFO by writing the DOEN bit to 0 in the
CRYP_DMACR register.

e)

Save the current configuration (bits [9:2] and bits 19 in the CRYP_CR register)
and, if not in ECB mode, the initialization vectors. The key value must already be
available in the memory. When needed, save the DMA status (pointers for IN and
OUT messages, number of remaining bytes, etc.). Read back the data loaded in

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)
the IN FIFO that have not been processed and save them in the memory until the
FIFO is empty.

Note:

23.4

In GCM/GMAC or CCM/CMAC mode, bits [17:16] of the CRYP_CR register should also be
saved.
2.

Configure and execute the other processing.

3.

Context restoration
a)

Configure the processor as in Section 23.3.6: Procedure to perform an encryption
or a decryption on page 744, Initialization with the saved configuration. For the
AES-ECB or AES-CBC decryption, the key must be prepared again.

b)

Write the data that were saved during context saving into the IN FIFO.

c)

If needed, reconfigure the DMA controller to transfer the rest of the message.

d)

Enable the processor by setting the CRYPEN bit and, the DMA requests by setting
the DIEN and DOEN bits.

CRYP interrupts
There are two individual maskable interrupt sources generated by the CRYP. These two
sources are combined into a single interrupt signal, which is the only interrupt signal from
the CRYP that drives the NVIC (nested vectored interrupt controller). This combined
interrupt, which is an OR function of the individual masked sources, is asserted if any of the
individual interrupts listed below is asserted and enabled.
You can enable or disable the interrupt sources individually by changing the mask bits in the
CRYP_IMSCR register. Setting the appropriate mask bit to ‘1’ enables the interrupt.
The status of the individual interrupt sources can be read either from the CRYP_RISR
register, for raw interrupt status, or from the CRYP_MISR register, for the masked interrupt
status.

Output FIFO service interrupt - OUTMIS
The output FIFO service interrupt is asserted when there is one or more (32-bit word) data
items in the output FIFO. This interrupt is cleared by reading data from the output FIFO until
there is no valid (32-bit) word left (that is, the interrupt follows the state of the OFNE (output
FIFO not empty) flag).
The output FIFO service interrupt OUTMIS is NOT enabled with the CRYP enable bit.
Consequently, disabling the CRYP will not force the OUTMIS signal low if the output FIFO is
not empty.

Input FIFO service interrupt - INMIS
The input FIFO service interrupt is asserted when there are less than four words in the input
FIFO. It is cleared by performing write operations to the input FIFO until it holds four or more
words.
The input FIFO service interrupt INMIS is enabled with the CRYP enable bit. Consequently,
when CRYP is disabled, the INMIS signal is low even if the input FIFO is empty.

RM0090 Rev 17

747/1747
766

Cryptographic processor (CRYP)

RM0090

Figure 231. CRYP interrupt mapping diagram
#290%.

). &)&/ )NTERRUPT ).-)3

).2)3
).)-

).-)3
'LOBAL )NTERRUPT

/542)3
/54)-

23.5

/54-)3

/54 &)&/ )NTERRUPT /54-)3
AI

CRYP DMA interface
The cryptographic processor provides an interface to connect to the DMA controller. The
DMA operation is controlled through the CRYP DMA control register, CRYP_DMACR.
The burst and single transfer request signals are not mutually exclusive. They can both be
asserted at the same time. For example, when there are 6 words available in the OUT FIFO,
the burst transfer request and the single transfer request are asserted. After a burst transfer
of 4 words, the single transfer request only is asserted to transfer the last 2 available words.
This is useful for situations where the number of words left to be received in the stream is
less than a burst.
Each request signal remains asserted until the relevant DMA clear signal is asserted. After
the request clear signal is deasserted, a request signal can become active again, depending
on the above described conditions. All request signals are deasserted if the CRYP
peripheral is disabled or the DMA enable bit is cleared (DIEN bit for the IN FIFO and DOEN
bit for the OUT FIFO in the CRYP_DMACR register).

Note:

The DMA controller must be configured to perform burst of 4 words or less. Otherwise some
data could be lost.
In order to let the DMA controller empty the OUT FIFO before filling up the IN FIFO, the
OUTDMA channel should have a higher priority than the INDMA channel.

23.6

CRYP registers
The cryptographic core is associated with several control and status registers, eight key
registers and four initialization vectors registers.

23.6.1

CRYP control register (CRYP_CR) for STM32F415/417xx
Address offset: 0x00
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

6

5

4

3

18

17

16

Reserved
15

14

CRYPEN FFLUSH
rw

748/1747

w

13

12

11

Reserved

10

9

8

KEYSIZE
rw

rw

7

DATATYPE
rw

RM0090 Rev 17

rw

ALGOMODE[2:0]
rw

rw

rw

2

1

0

ALGODIR

Res.

Res.

rw

RM0090

Cryptographic processor (CRYP)

Bits 31:18 Reserved, must be kept at reset value
Bits 17:16 Reserved, must be kept at reset value
Bit 15 CRYPEN: Cryptographic processor enable
0: CRYP processor is disabled
1: CRYP processor is enabled
Note: The CRYPEN bit is automatically cleared by hardware when the key
preparation process ends (ALGOMODE=111b) or GCM_CCM init Phase
Bit 14 FFLUSH: FIFO flush
When CRYPEN = 0, writing this bit to 1 flushes the IN and OUT FIFOs (that is
read and write pointers of the FIFOs are reset. Writing this bit to 0 has no effect.
When CRYPEN = 1, writing this bit to 0 or 1 has no effect.
Reading this bit always returns 0.
Bits 13:10 Reserved, must be kept at reset value
Bits 9:8 KEYSIZE[1:0]: Key size selection (AES mode only)
This bitfield defines the bit-length of the key used for the AES cryptographic core.
This bitfield is ‘don’t care’ in the DES or TDES modes.
00: 128 bit key length
01: 192 bit key length
10: 256 bit key length
11: Reserved, do not use this value
Bits 7:6 DATATYPE[1:0]: Data type selection
This bitfield defines the format of data entered in the CRYP_DIN register (refer to
Section 23.3.3: Data type).
00: 32-bit data. No swapping of each word. First word pushed into the IN FIFO
(or popped off the OUT FIFO) forms bits 1...32 of the data block, the second
word forms bits 33...64.
01: 16-bit data, or half-word. Each word pushed into the IN FIFO (or popped off
the OUT FIFO) is considered as 2 half-words, which are swapped with each
other.
10: 8-bit data, or bytes. Each word pushed into the IN FIFO (or popped off the
OUT FIFO) is considered as 4 bytes, which are swapped with each other.
11: bit data, or bit-string. Each word pushed into the IN FIFO (or popped off the
OUT FIFO) is considered as 32 bits (1st bit of the string at position 0), which are
swapped with each other.

RM0090 Rev 17

749/1747
766

Cryptographic processor (CRYP)

RM0090

Bits 5:3 ALGOMODE[2:0]: Algorithm mode
000: TDES-ECB (triple-DES Electronic codebook): no feedback between blocks
of data. Initialization vectors (CRYP_IV0(L/R)) are not used, three key vectors
(K1, K2, and K3) are used (K0 is not used).
001: TDES-CBC (triple-DES Cipher block chaining): output block is XORed with
the subsequent input block before its entry into the algorithm. Initialization
vectors (CRYP_IV0L/R) must be initialized, three key vectors (K1, K2, and K3)
are used (K0 is not used).
010: DES-ECB (simple DES Electronic codebook): no feedback between blocks
of data. Initialization vectors (CRYP_IV0L/R) are not used, only one key vector
(K1) is used (K0, K2, K3 are not used).
011: DES-CBC (simple DES Cipher block chaining): output block is XORed with
the subsequent input block before its entry into the algorithm. Initialization
vectors (CRYP_IV0L/R) must be initialized. Only one key vector (K1) is used
(K0, K2, K3 are not used).
100: AES-ECB (AES Electronic codebook): no feedback between blocks of data.
Initialization vectors (CRYP_IV0L/R...1L/R) are not used. All four key vectors
(K0...K3) are used.
101: AES-CBC (AES Cipher block chaining): output block is XORed with the
subsequent input block before its entry into the algorithm. Initialization vectors
(CRYP_IV0L/R...1L/R) must be initialized. All four key vectors (K0...K3) are
used.
110: AES-CTR (AES counter mode): output block is XORed with the subsequent
input block before its entry into the algorithm. Initialization vectors
(CRYP_IV0L/R...1L/R) must be initialized. All four key vectors (K0...K3) are
used. CTR decryption does not differ from CTR encryption, since the core
always encrypts the current counter block to produce the key stream that will be
XORed with the plaintext or cipher in input. Thus, ALGODIR is don’t care when
ALGOMODE = 110b, and the key must NOT be unrolled (prepared) for
decryption.
111: AES key preparation for decryption mode. Writing this value when CRYPEN
= 1 immediately starts an AES round for key preparation. The secret key must
have previously been loaded into the K0...K3 registers. The BUSY bit in the
CRYP_SR register is set during the key preparation. After key processing, the
resulting key is copied back into the K0...K3 registers, and the BUSY bit is
cleared.
Bit 2 ALGODIR: Algorithm direction
0: Encrypt
1: Decrypt
Bits 1:0 Reserved, must be kept at reset value

Note:

Writing to the KEYSIZE, DATATYPE, ALGOMODE and ALGODIR bits while BUSY=1 has
no effect. These bits can only be configured when BUSY=0.
The FFLUSH bit has to be set only when BUSY=0. If not, the FIFO is flushed, but the block
being processed may be pushed into the output FIFO just after the flush operation, resulting
in a nonempty FIFO condition.

23.6.2

CRYP control register (CRYP_CR) for STM32F415/417xx
Address offset: 0x00
Reset value: 0x0000 0000

750/1747

RM0090 Rev 17

RM0090

31

Cryptographic processor (CRYP)

30

29

28

27

26

25

24

23

22

21

20

Reserved

19

18

ALGO
MODE
[3]

Res.

rw
15

14

CRYPEN FFLUSH
rw

w

13

12

11

Reserved

10

9

8

KEYSIZE
rw

rw

7

6

DATATYPE
rw

rw

5

4

3

ALGOMODE[2:0]
rw

rw

rw

2
ALGODIR
rw

17

16

GCM_CCMPH
rw

rw

1

0

Reserved

Bits 31:20 Reserved, forced by hardware to 0.
Bit 18 Reserved, forced by hardware to 0.
Bits 17:16 GCM_CCMPH[1:0]: no effect if “GCM or CCM algorithm” is not set
00: GCM_CCM init Phase
01: GCM_CCM header phase
10: GCM_CCM payload phase
11: GCM_CCM final phase
Bit 15 CRYPEN: Cryptographic processor enable
0: CRYP processor is disabled
1: CRYP processor is enabled
Note: The CRYPEN bit is automatically cleared by hardware when the key
preparation process ends (ALGOMODE=111b) or GCM_CCM init Phase
Bit 14 FFLUSH: FIFO flush
When CRYPEN = 0, writing this bit to 1 flushes the IN and OUT FIFOs (that is
read and write pointers of the FIFOs are reset. Writing this bit to 0 has no effect.
When CRYPEN = 1, writing this bit to 0 or 1 has no effect.
Reading this bit always returns 0.
Bits 13:10 Reserved, forced by hardware to 0.
Bits 9:8 KEYSIZE[1:0]: Key size selection (AES mode only)
This bitfield defines the bit-length of the key used for the AES cryptographic core.
This bitfield is ‘don’t care’ in the DES or TDES modes.
00: 128 bit key length
01: 192 bit key length
10: 256 bit key length
11: Reserved, do not use this value
Bits 7:6 DATATYPE[1:0]: Data type selection
This bitfield defines the format of data entered in the CRYP_DIN register (refer to
Section 23.3.3: Data type).
00: 32-bit data. No swapping of each word. First word pushed into the IN FIFO
(or popped off the OUT FIFO) forms bits 1...32 of the data block, the second
word forms bits 33...64.
01: 16-bit data, or half-word. Each word pushed into the IN FIFO (or popped off
the OUT FIFO) is considered as 2 half-words, which are swapped with each
other.
10: 8-bit data, or bytes. Each word pushed into the IN FIFO (or popped off the
OUT FIFO) is considered as 4 bytes, which are swapped with each other.
11: bit data, or bit-string. Each word pushed into the IN FIFO (or popped off the
OUT FIFO) is considered as 32 bits (1st bit of the string at position 0), which are
swapped with each other.

RM0090 Rev 17

751/1747
766

Cryptographic processor (CRYP)

RM0090

Bits 19 and 5:3 ALGOMODE[3:0]: Algorithm mode
0000: TDES-ECB (triple-DES Electronic codebook): no feedback between
blocks of data. Initialization vectors (CRYP_IV0(L/R)) are not used, three key
vectors (K1, K2, and K3) are used (K0 is not used).
0001: TDES-CBC (triple-DES Cipher block chaining): output block is XORed
with the subsequent input block before its entry into the algorithm. Initialization
vectors (CRYP_IV0L/R) must be initialized, three key vectors (K1, K2, and K3)
are used (K0 is not used).
0010: DES-ECB (simple DES Electronic codebook): no feedback between
blocks of data. Initialization vectors (CRYP_IV0L/R) are not used, only one key
vector (K1) is used (K0, K2, K3 are not used).
0011: DES-CBC (simple DES Cipher block chaining): output block is XORed with
the subsequent input block before its entry into the algorithm. Initialization
vectors (CRYP_IV0L/R) must be initialized. Only one key vector (K1) is used
(K0, K2, K3 are not used).
0100: AES-ECB (AES Electronic codebook): no feedback between blocks of
data. Initialization vectors (CRYP_IV0L/R...1L/R) are not used. All four key
vectors (K0...K3) are used.
0101: AES-CBC (AES Cipher block chaining): output block is XORed with the
subsequent input block before its entry into the algorithm. Initialization vectors
(CRYP_IV0L/R...1L/R) must be initialized. All four key vectors (K0...K3) are
used.
0110: AES-CTR (AES Counter mode): output block is XORed with the
subsequent input block before its entry into the algorithm. Initialization vectors
(CRYP_IV0L/R...1L/R) must be initialized. All four key vectors (K0...K3) are
used. CTR decryption does not differ from CTR encryption, since the core
always encrypts the current counter block to produce the key stream that will be
XORed with the plaintext or cipher in input. Thus, ALGODIR is don’t care when
ALGOMODE = 110b, and the key must NOT be unrolled (prepared) for
decryption.
0111: AES key preparation for decryption mode. Writing this value when
CRYPEN = 1 immediately starts an AES round for key preparation. The secret
key must have previously been loaded into the K0...K3 registers. The BUSY bit
in the CRYP_SR register is set during the key preparation. After key processing,
the resulting key is copied back into the K0...K3 registers, and the BUSY bit is
cleared.
1000: Galois Counter Mode (GCM). This algorithm mode is also used for the
GMAC algorithm.
1001: Counter with CBC-MAC (CCM). This algorithm mode is also used for the
CMAC algorithm.
Bit 2 ALGODIR: Algorithm direction
0: Encrypt
1: Decrypt
Bits 1:0 Reserved, must be kept to 0.

Note:

752/1747

Writing to the KEYSIZE, DATATYPE, ALGOMODE and ALGODIR bits while BUSY=1 has
no effect. These bits can only be configured when BUSY=0.
The FFLUSH bit has to be set only when BUSY=0. If not, the FIFO is flushed, but the block
being processed may be pushed into the output FIFO just after the flush operation, resulting
in a nonempty FIFO condition.

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)

23.6.3

CRYP status register (CRYP_SR)
Address offset: 0x04
Reset value: 0x0000 0003

31

30

29

28

27

26

25

24

23

22

21

6

5

20

19

18

17

16

Reserved
15

14

13

12

11

10

9

8

7

Reserved

4

3

2

1

0

BUSY

OFFU

OFNE

IFNF

IFEM

r

r

r

r

r

Bits 31:5 Reserved, must be kept at reset value
Bit 4 BUSY: Busy bit
0: The CRYP Core is not processing any data. The reason is either that:
–
the CRYP core is disabled (CRYPEN=0 in the CRYP_CR register) and
the last processing has completed, or
–
The CRYP core is waiting for enough data in the input FIFO or enough
free space in the output FIFO (that is in each case at least 2 words in
the DES, 4 words in the AES).
1: The CRYP core is currently processing a block of data or a key preparation
(for AES decryption).
Bit 3 OFFU: Output FIFO full
0: Output FIFO is not full
1: Output FIFO is full
Bit 2 OFNE: Output FIFO not empty
0: Output FIFO is empty
1: Output FIFO is not empty
Bit 1 IFNF: Input FIFO not full
0: Input FIFO is full
1: Input FIFO is not full
Bit 0 IFEM: Input FIFO empty
0: Input FIFO is not empty
1: Input FIFO is empty

RM0090 Rev 17

753/1747
766

Cryptographic processor (CRYP)

23.6.4

RM0090

CRYP data input register (CRYP_DIN)
Address offset: 0x08
Reset value: 0x0000 0000
The CRYP_DIN register is the data input register. It is 32-bit wide. It is used to enter up to
four 64-bit (TDES) or two 128-bit (AES) plaintext (when encrypting) or ciphertext (when
decrypting) blocks into the input FIFO, one 32-bit word at a time.
The first word written into the FIFO is the MSB of the input block. The LSB of the input block
is written at the end. Disregarding the data swapping, this gives:
•

In the DES/TDES modes: a block is a sequence of bits numbered from bit 1 (leftmost
bit) to bit 64 (rightmost bit). Bit 1 corresponds to the MSB (bit 31) of the first word
entered into the FIFO, bit 64 corresponds to the LSB (bit 0) of the second word entered
into the FIFO.

•

In the AES mode: a block is a sequence of bits numbered from 0 (leftmost bit) to 127
(rightmost bit). Bit 0 corresponds to the MSB (bit 31) of the first word written into the
FIFO, bit 127 corresponds to the LSB (bit 0) of the 4th word written into the FIFO.

To fit different data sizes, the data written in the CRYP_DIN register can be swapped before
being processed by configuring the DATATYPE bits in the CRYP_CR register. Refer to
Section 23.3.3: Data type on page 739 for more details.
When CRYP_DIN register is written to, the data are pushed into the input FIFO. When at
least two 32-bit words in the DES/TDES mode (or four 32-bit words in the AES mode) have
been pushed into the input FIFO, and when at least 2 words are free in the output FIFO, the
CRYP engine starts an encrypting or decrypting process. This process takes two 32-bit
words in the DES/TDES mode (or four 32-bit words in the AES mode) from the input FIFO
and delivers two 32-bit words (or 4, respectively) to the output FIFO per process round.
When CRYP_DIN register is read:
•

If CRYPEN = 0, the FIFO is popped, and then the data present in the Input FIFO are
returned, from the oldest one (first reading) to the newest one (last reading). The IFEM
flag must be checked before each read operation to make sure that the FIFO is not
empty.

•

if CRYPEN = 1, an undefined value is returned.

After the CRYP_DIN register has been read once or several times, the FIFO must be
flushed by setting the FFLUSH bit prior to processing new data.
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

DATAIN
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

DATAIN

Bits 31:0 DATAIN: Data input
Read = returns Input FIFO content if CRYPEN = 0, else returns an undefined
value.
Write = Input FIFO is written.

754/1747

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)

23.6.5

CRYP data output register (CRYP_DOUT)
Address offset: 0x0C
Reset value: 0x0000 0000
The CRYP_DOUT register is the data output register. It is read-only and 32-bit wide. It is
used to retrieve up to four 64-bit (TDES mode) or two 128-bit (AES mode) ciphertext (when
encrypting) or plaintext (when decrypting) blocks from the output FIFO, one 32-bit word at a
time.
Like for the input data, the MSB of the output block is the first word read from the output
FIFO. The LSB of the output block is read at the end. Disregarding data swapping, this
gives:
•

In the DES/TDES modes: Bit 1 (leftmost bit) corresponds to the MSB (bit 31) of the first
word read from the FIFO, bit 64 (rightmost bit) corresponds to the LSB (bit 0) of the
second word read from the FIFO.

•

In the AES mode: Bit 0 (leftmost bit) corresponds to the MSB (bit 31) of the first word
read from the FIFO, bit 127 (rightmost bit) corresponds to the LSB (bit 0) of the 4th
word read from the FIFO.

To fit different data sizes, the data can be swapped after processing by configuring the
DATATYPE bits in the CRYP_CR register. Refer to Section 23.3.3: Data type on page 739
for more details.
When CRYP_DOUT register is read, the last data entered into the output FIFO (pointed to
by the read pointer) is returned.
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

DATAOUT
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

r

r

r

r

DATAOUT
r

r

Bits 31:0 DATAOUT: Data output
Read = returns output FIFO content.
Write = No effect.

RM0090 Rev 17

755/1747
766

Cryptographic processor (CRYP)

23.6.6

RM0090

CRYP DMA control register (CRYP_DMACR)
Address offset: 0x10
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

6

5

4

3

2

17

16

Reserved
15

14

13

12

11

10

9

8

7

Reserved

1

0

DOEN

DIEN

rw

rw

Bits 31:2 Reserved, must be kept at reset value
Bit 1 DOEN: DMA output enable
0: DMA for outgoing data transfer is disabled
1: DMA for outgoing data transfer is enabled
Bit 0 DIEN: DMA input enable
0: DMA for incoming data transfer is disabled
1: DMA for incoming data transfer is enabled

23.6.7

CRYP interrupt mask set/clear register (CRYP_IMSCR)
Address offset: 0x14
Reset value: 0x0000 0000
The CRYP_IMSCR register is the interrupt mask set or clear register. It is a read/write
register. On a read operation, this register gives the current value of the mask on the
relevant interrupt. Writing 1 to the particular bit sets the mask, enabling the interrupt to be
read. Writing 0 to this bit clears the corresponding mask. All the bits are cleared to 0 when
the peripheral is reset.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

OUTIM

INIM

rw

rw

Reserved
15

14

13

12

11

10

9

8

7

Reserved

Bits 31:2 Reserved, must be kept at reset value
Bit 1 OUTIM: Output FIFO service interrupt mask
0: Output FIFO service interrupt is masked
1: Output FIFO service interrupt is not masked
Bit 0 INIM: Input FIFO service interrupt mask
0: Input FIFO service interrupt is masked
1: Input FIFO service interrupt is not masked

756/1747

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)

23.6.8

CRYP raw interrupt status register (CRYP_RISR)
Address offset: 0x18
Reset value: 0x0000 0001
The CRYP_RISR register is the raw interrupt status register. It is a read-only register. On a
read, this register gives the current raw status of the corresponding interrupt prior to
masking. A write has no effect.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

6

5

4

3

2

17

16

Reserved
15

14

13

12

11

10

9

8

7

Reserved

1

0

OUTRIS

INRIS

r

r

Bits 31:2 Reserved, must be kept at reset value
Bit 1 OUTRIS: Output FIFO service raw interrupt status
Gives the raw interrupt state prior to masking of the output FIFO service
interrupt.
0: Raw interrupt not pending
1: Raw interrupt pending
Bit 0 INRIS: Input FIFO service raw interrupt status
Gives the raw interrupt state prior to masking of the Input FIFO service interrupt.
0: Raw interrupt not pending
1: Raw interrupt pending

23.6.9

CRYP masked interrupt status register (CRYP_MISR)
Address offset: 0x1C
Reset value: 0x0000 0000
The CRYP_MISR register is the masked interrupt status register. It is a read-only register.
On a read, this register gives the current masked status of the corresponding interrupt prior
to masking. A write has no effect.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

6

5

4

3

2

17

16

Reserved
15

14

13

12

11

10

9

8

7

Reserved

RM0090 Rev 17

1

0

OUTMIS

INMIS

r

r

757/1747
766

Cryptographic processor (CRYP)

RM0090

Bits 31:2 Reserved, must be kept at reset value
Bit 1 OUTMIS: Output FIFO service masked interrupt status
Gives the interrupt state after masking of the output FIFO service interrupt.
0: Interrupt not pending
1: Interrupt pending
Bit 0 INMIS: Input FIFO service masked interrupt status
Gives the interrupt state after masking of the input FIFO service interrupt.
0: Interrupt not pending
1: Interrupt pending when CRYPEN = 1

23.6.10

CRYP key registers (CRYP_K0...3(L/R)R)
Address offset: 0x20 to 0x3C
Reset value: 0x0000 0000
These registers contain the cryptographic keys.
In the TDES mode, keys are 64-bit binary values (number from left to right, that is the
leftmost bit is bit 1), named K1, K2 and K3 (K0 is not used), each key consists of 56
information bits and 8 parity bits. The parity bits are reserved for error detection purposes
and are not used by the current block. Thus, bits 8, 16, 24, 32, 40, 48, 56 and 64 of each 64bit key value Kx[1:64] are not used.
In the AES mode, the key is considered as a single 128-, 192- or 256-bit long bit sequence,
k0k1k2...k127/191/255 (k0 being the leftmost bit). The AES key is entered into the registers as
follows:
•

for AES-128: k0..k127 corresponds to b127..b0 (b255..b128 are not used),

•

for AES-192: k0..k191 corresponds to b191..b0 (b255..b192 are not used),

•

for AES-256: k0..k255 corresponds to b255..b0.

In any case b0 is the rightmost bit.

CRYP_K0LR (address offset: 0x20)
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

b255

b254

b253

b252

b251

b250

b249

b248

b247

b246

b245

b244

b243

b242

b241

b240
w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

b239

b238

b237

b236

b235

b234

b233

b232

b231

b230

b229

b228

b227

b226

b225

b224

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

CRYP_K0RR (address offset: 0x24)
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

b223

b222

b221

b220

b219

b218

b217

b216

b215

b214

b213

b212

b211

b210

b209

b208

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

b207

b206

b205

b204

b203

b202

b201

b200

b199

b198

b197

b196

b195

b194

b193

b192

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

758/1747

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)

CRYP_K1LR (address offset: 0x28)
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

k1.1
b191

k1.2
b190

k1.3
b189

k1.4
b188

k1.5
b187

k1.6
b186

k1.7
b185

k1.8
b184

k1.9
b183

k1.10
b182

k1.11
b181

k1.12
b180

k1.13
b179

k1.14
b178

k1.15
b177

k1.16
b176
w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

k1.17
b175

k1.18
b174

k1.19
b173

k1.20
b172

k1.21
b171

k1.22
b170

k1.23
b169

k1.24
b168

k1.25
b167

k1.26
b166

k1.27
b165

k1.28
b164

k1.29
b163

k1.30
b162

k1.31
b161

k1.32
b160

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

CRYP_K1RR (address offset: 0x2C)
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

k1.33
b159

k1.34
b158

k1.35
b157

k1.36
b156

k1.37
b155

k1.38
b154

k1.39
b153

k1.40
b152

k1.41
b151

k1.42
b150

k1.43
b149

k1.44
b148

k1.45
b147

k1.46
b146

k1.47
b145

k1.48
b144
w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

k1.49
b143

k1.50
b142

k1.51
b141

k1.52
b140

k1.53
b139

k1.54
b138

k1.55
b137

k1.56
b136

k1.57
b135

k1.58
b134

k1.59
b133

k1.60
b132

k1.61
b131

k1.62
b130

k1.63
b129

k1.64
b128

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

CRYP_K2LR (address offset: 0x30)
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

k2.1
b127

k2.2
b126

k2.3
b125

k2.4
b124

k2.5
b123

k2.6
b122

k2.7
b121

k2.8
b120

k2.9
b119

k2.10
b118

k2.11
b117

k2.12
b116

k2.13
b115

k2.14
b114

k2.15
b113

k2.16
b112
w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

k2.17
b111

k2.18
b110

k2.19
b109

k2.20
b108

k2.21
b107

k2.22
b106

k2.23
b105

k2.24
b104

k2.25
b103

k2.26
b102

k2.27
b101

k2.28
b100

k2.29
b99

k2.30
b98

k2.31
b97

k2.32
b96

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

CRYP_K2RR (address offset: 0x34)
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

k2.33
b95

k2.34
b94

k2.35
b93

k2.36
b92

k2.37
b91

k2.38
b90

k2.39
b89

k2.40
b88

k2.41
b87

k2.42
b86

k2.43
b85

k2.44
b84

k2.45
b83

k2.46
b82

k2.47
b81

k2.48
b80
w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

k2.49
b79

k2.50
b78

k2.51
b77

k2.52
b76

k2.53
b75

k2.54
b74

k2.55
b73

k2.56
b72

k2.57
b71

k2.58
b70

k2.59
b69

k2.60
b68

k2.61
b67

k2.62
b66

k2.63
b65

k2.64
b64

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

CRYP_K3LR (address offset: 0x38)
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

k3.1
b63

k3.2
b62

k3.3
b61

k3.4
b60

k3.5
b59

k3.6
b58

k3.7
b57

k3.8
b56

k3.9
b55

k3.10
b54

k3.11
b53

k3.12
b52

k3.13
b51

k3.14
b50

k3.15
b49

k3.16
b48
w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

k3.17
b47

k3.18
b46

k3.19
b45

k3.20
b44

k3.21
b43

k3.22
b42

k3.23
b41

k3.24
b40

k3.25
b39

k3.26
b38

k3.27
b37

k3.28
b36

k3.29
b35

k3.30
b34

k3.31
b33

k3.32
b32

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

RM0090 Rev 17

759/1747
766

Cryptographic processor (CRYP)

RM0090

CRYP_K3RR (address offset: 0x3C)
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

k3.33
b31

k3.34
b30

k3.35
b29

k3.36
b28

k3.37
b27

k3.38
b26

k3.39
b25

k3.40
b24

k3.41
b23

k3.42
b22

k3.43
b21

k3.44
b20

k3.45
b19

k3.46
b18

k3.47
b17

k3.48
b16
w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

k3.49
b15

k3.50
b14

k3.51
b13

k3.52
b12

k3.53
b11

k3.54
b10

k3.55
b9

k3.56
b8

k3.57
b7

k3.58
b6

k3.59
b5

k3.60
b4

k3.61
b3

k3.62
b2

k3.63
b1

k3.64
b0

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

w

Note:

Write accesses to these registers are disregarded when the cryptographic processor is busy
(bit BUSY = 1 in the CRYP_SR register).

23.6.11

CRYP initialization vector registers (CRYP_IV0...1(L/R)R)
Address offset: 0x40 to 0x4C
Reset value: 0x0000 0000
The CRYP_IV0...1(L/R)R are the left-word and right-word registers for the initialization
vector (64 bits for DES/TDES and 128 bits for AES) and are used in the CBC (Cipher block
chaining) and Counter (CTR) modes. After each computation round of the TDES or AES
Core, the CRYP_IV0...1(L/R)R registers are updated as described in Section : DES and
TDES Cipher block chaining (DES/TDES-CBC) mode on page 726, Section : AES Cipher
block chaining (AES-CBC) mode on page 730 and Section : AES counter mode (AES-CTR)
mode on page 732.
IV0 is the leftmost bit whereas IV63 (DES, TDES) or IV127 (AES) are the rightmost bits of
the initialization vector. IV1(L/R)R is used only in the AES.

CRYP_IV0LR (address offset: 0x40)
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

IV0

IV1

IV2

IV3

IV4

IV5

IV6

IV7

IV8

IV9

IV10

IV11

IV12

IV13

IV14

IV15
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

IV16

IV17

IV18

IV19

IV20

IV21

IV22

IV23

IV24

IV25

IV26

IV27

IV28

IV29

IV30

IV31

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

CRYP_IV0RR (address offset: 0x44)
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

IV32

IV33

IV34

IV35

IV36

IV37

IV38

IV39

IV40

IV41

IV42

IV43

IV44

IV45

IV46

IV47

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

IV48

IV49

IV50

IV51

IV52

IV53

IV54

IV55

IV56

IV57

IV58

IV59

IV60

IV61

IV62

IV63

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

760/1747

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)

CRYP_IV1LR (address offset: 0x48)
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

IV64

IV65

IV66

IV67

IV68

IV69

IV70

IV71

IV72

IV73

IV74

IV75

IV76

IV77

IV78

IV79

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

IV80

IV81

IV82

IV83

IV84

IV85

IV86

IV87

IV88

IV89

IV90

IV91

IV92

IV93

IV94

IV95

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

CRYP_IV1RR (address offset: 0x4C)
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

IV96

IV97

IV98

IV99

IV100

IV101

IV102

IV103

IV104

IV105

IV106

IV107

IV108

IV109

IV110

IV111

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

IV112

IV113

IV114

IV115

IV116

IV117

IV118

IV119

IV120

IV121

IV122

IV123

IV124

IV125

IV126

IV127

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Note:

In DES/3DES modes, only CRYP_IV0(L/R) is used.
Write access to these registers are disregarded when the cryptographic processor is busy
(bit BUSY = 1 in the CRYP_SR register).

RM0090 Rev 17

761/1747
766

Cryptographic processor (CRYP)

23.6.12

RM0090

CRYP context swap registers (CRYP_CSGCMCCM0..7R and
CRYP_CSGCM0..7R) for STM32F42xxx and STM32F43xxx
Address offset:
•

CRYP_CSGCMCCM0..7: 0x050 to 0x06C: used for GCM/GMAC or CCM/CMAC
alogrithm only

•

CRYP_CSGCM0..7: 0x070 to 0x08C: used for GCM/GMAC algorithm only

Reset value: 0x0000 0000
These registers contain the complete internal register states of the CRYP processor when
the GCM/GMAC or CCM/CMAC algorithm is selected. They are useful when a context swap
has to be performed because a high-priority task needs the cryptographic processor while it
is already in use by another task.
When such an event occurs, the CRYP_CSGCMCCM0..7R and CRYP_CSGCM0..7R (in
GCM/GMAC mode) or CRYP_CSGCMCCM0..7R (in CCM/CMAC mode) registers have to
be read and the values retrieved have to be saved in the system memory space. The
cryptographic processor can then be used by the preemptive task, and when the
cryptographic computation is complete, the saved context can be read from memory and
written back into the corresponding context swap registers.
Note:

These registers are used only when GCM/GMAC or CCM/CMAC algorithm mode is
selected.

CRYP_CSGCMCCMxR: where x=[7:0]
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

CRYP_CSGCMCCMxR
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

23

22

21

20

19

18

17

16

CRYP_CSGCMCCMxR
rw

CRYP_CSGCMxR: where x=[7:0]
31

30

29

28

27

26

25

24

CRYP_CSGCMxR
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

CRYP_CSGCMxR
rw

762/1747

rw

rw

rw

rw

rw

rw

rw

rw

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)

23.6.13

CRYP register map

Reset value
Reserved

Reset value

0x10

Reset value
Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_DMAC
R

Reserved

0 0

CRYP_IMSC
R

Reserved

0 0

CRYP_RISR

Reserved

Reset value

0x1C

0 1

CRYP_MISR

Reserved

Reset value
0x20
0x24

Res.

DATAOUT

Reset value

0x18

ALGODIR

0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_DOUT

Reset value
0x14

ALOMODE[2:0]

DATAIN

DOEN
DIEN

0x0C

0 0 0 1 1

CRYP_DIN

OUTIM
INIM

0x08

0 0 0 0 0 0 0 0

OUTRIS
INRIS

CRYP_SR

0 0

OUTMIS
IN%IS

0x04

Reserved

BUSY
OFFU
OFNE
IFNF
IFEM

Reserved

DATATYPE

CRYP_CR

KEYSIZE

0x00
0x00

Register
name and
reset value

CRYPEN
FFLUSH

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 114. CRYP register map and reset values for STM32F415/417xx

0 0

CRYP_K0LR

CRYP_K0LR

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_K0RR

CRYP_K0RR

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
...

0x38
0x3C
0x40

CRYP_K3LR

CRYP_K3LR

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_K3RR

CRYP_K3RR

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_IV0LR

CRYP_IV0LR

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

RM0090 Rev 17

763/1747
766

Cryptographic processor (CRYP)

RM0090

Register
name and
reset value

Offset

0x44
0x48
0x4C

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 114. CRYP register map and reset values for STM32F415/417xx (continued)

CRYP_IV0RR

CRYP_IV0RR

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_IV1LR

CRYP_IV1LR

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_IV1RR

CRYP_IV1RR

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Register
name

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 115. CRYP register map and reset values for STM32F43xxx

Reset value
Reset value
CRYP_DOUT
Reset value
CRYP_DMAC
R

CRYP_IMSC
R

Reserved

CRYP_RISR

Reserved

Res..
0 0

Reserved

Reset value

764/1747

ALGODIR

0 0

Reset value

0x18

ALOMODE[2:0]

0 0 0 0 0 0 0 0 0 0 0 0 0

Reset value
0x14

DATATYPE

0 0 0 0 0 0 0 0 0 0 0 0 0
DATAOUT
DOEN
DIEN

0x10

DATAIN

OUTIM
INIM

0x0C

0 0 0 1 1

CRYP_DIN

0x08

KEYSIZE

CRYPEN
FFLUSH

Reserved

0 0 0 0 0 0 0 0

OUTRIS
INRIS

CRYP_SR

0x04

0 0 0 0

Reserved

BUSY
OFFU
OFNE
IFNF
IFEM

Reset value

Reserved

GCM_CCMPH

CRYP_CR

0x00
0x00

ALGOMODE[3]
Res.

reset value

0 1

RM0090 Rev 17

RM0090

Cryptographic processor (CRYP)

Offset

Register
name

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 115. CRYP register map and reset values for STM32F43xxx (continued)

0x1C

CRYP_MISR

Reserved

Reset value
0x20
0x24

OUTMIS
IN%IS

reset value

0 0

CRYP_K0LR

CRYP_K0LR

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_K0RR

CRYP_K0RR

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
...
...

0x38
0x3C
0x40
0x44
0x48
0x4C

0x50

0x54

0x58

0x5C

CRYP_K3LR

CRYP_K3LR

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_K3RR

CRYP_K3RR

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_IV0LR

CRYP_IV0LR

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_IV0RR

CRYP_IV0RR

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_IV1LR

CRYP_IV1LR

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_IV1RR

CRYP_IV1RR

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_
CSGCMCCM
R

CRYP_CSGCMCCM0R

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_
CSGCMCCM
1R

CRYP_CSGCMCCM1R

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_
CSGCMCCM
2R

CRYP_CSGCMCCM2R

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_
CSGCMCCM
3R

CRYP_CSGCMCCM3R

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

RM0090 Rev 17

765/1747
766

Cryptographic processor (CRYP)

RM0090

Register
name

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 115. CRYP register map and reset values for STM32F43xxx (continued)

reset value
CRYP_
CSGCMCCM
4R

CRYP_CSGCMCCM4R

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_
CSGCMCCM
5R

CRYP_CSGCMCCM5R

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_
CSGCMCCM
6R

CRYP_CSGCMCCM6R

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_
CSGCMCCM
7R

CRYP_CSGCMCCM7R

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0x70

CRYP_CSGC
M0R

CRYP_CSGCM0R

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0x74

CRYP_CSGC
M1R

CRYP_CSGCM1R

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0x78

CRYP_CSGC
M2R

CRYP_CSGCM2R

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_CSGC
M3R

CRYP_CSGCM3R

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0x80

CRYP_CSGC
M4R

CRYP_CSGCM4R

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0x84

CRYP_CSGC
M5R

CRYP_CSGCM5R

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0x88

CRYP_CSGC
M6R

CRYP_CSGCM6R

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CRYP_CSGC
M7R

CRYP_CSGCM7R

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0x60

0x64

0x68

0x6C

0x7C

0x8C

Refer to Section 2.3: Memory map for the register boundary addresses.

766/1747

RM0090 Rev 17

RM0090

24

Random number generator (RNG)

Random number generator (RNG)
This section applies to the whole STM32F4xx family, unless otherwise specified.

24.1

RNG introduction
The RNG processor is a random number generator, based on a continuous analog noise,
that provides a random 32-bit value to the host when read.
The RNG passed the FIPS PUB 140-2 (2001 October 10) tests with a success ratio of 99%.

24.2

24.3

RNG main features
•

It delivers 32-bit random numbers, produced by an analog generator

•

40 periods of the RNG_CLK clock signal between two consecutive random numbers

•

Monitoring of the RNG entropy to flag abnormal behavior (generation of stable values,
or of a stable sequence of values)

•

It can be disabled to reduce power consumption

RNG functional description
Figure 232 shows the RNG block diagram.
Figure 232. Block diagram
 BIT !(" BUS
DATA REGISTER
#ONTROL REGISTER

2.'?$2

2.' ?#2

,&32

2.'?#,+
3TATUS REGISTER
2.' ?32

#LOCK CHECKER 
FAULT DETECTOR

FEED A ,INEAR &EEDBACK
3HIFT 2EGISTER

!NALOG SEED
AI

1. For more details about RNG Clock (RNG_CLK) source, please refer to Section 6: Reset and clock control
for STM32F42xxx and STM32F43xxx (RCC) and Section 7: Reset and clock control for
STM32F405xx/07xx and STM32F415xx/17xx(RCC).

The random number generator implements an analog circuit. This circuit generates seeds
that feed a linear feedback shift register (RNG_LFSR) in order to produce 32-bit random
numbers.

RM0090 Rev 17

767/1747
771

Random number generator (RNG)

RM0090

The analog circuit is made of several ring oscillators whose outputs are XORed to generate
the seeds. The RNG_LFSR is clocked by a dedicated clock (RNG_CLK) at a constant
frequency, so that the quality of the random number is independent of the HCLK frequency.
The contents of the RNG_LFSR are transferred into the data register (RNG_DR) when a
significant number of seeds have been introduced into the RNG_LFSR.
In parallel, the analog seed and the dedicated RNG_CLK clock are monitored. Status bits (in
the RNG_SR register) indicate when an abnormal sequence occurs on the seed or when
the frequency of the RNG_CLK clock is too low. An interrupt can be generated when an
error is detected.

24.3.1

Operation
To run the RNG, follow the steps below:
1.

Enable the interrupt if needed (to do so, set the IE bit in the RNG_CR register). An
interrupt is generated when a random number is ready or when an error occurs.

2.

Enable the random number generation by setting the RNGEN bit in the RNG_CR
register. This activates the analog part, the RNG_LFSR and the error detector.

3.

At each interrupt, check that no error occurred (the SEIS and CEIS bits should be ‘0’ in
the RNG_SR register) and that a random number is ready (the DRDY bit is ‘1’ in the
RNG_SR register). The contents of the RNG_DR register can then be read.

As required by the FIPS PUB (Federal Information Processing Standard Publication) 140-2,
the first random number generated after setting the RNGEN bit should not be used, but
saved for comparison with the next generated random number. Each subsequent generated
random number has to be compared with the previously generated number. The test fails if
any two compared numbers are equal (continuous random number generator test).

24.3.2

Error management
If the CEIS bit is read as ‘1’ (clock error)
In the case of a clock, the RNG is no more able to generate random numbers because the
RNG_CLK clock is not correct. Check that the clock controller is correctly configured to
provide the RNG clock and clear the CEIS bit. The RNG can work when the CECS bit is ‘0’.
The clock error has no impact on the previously generated random numbers, and the
RNG_DR register contents can be used.

If the SEIS bit is read as ‘1’ (seed error)
In the case of a seed error, the generation of random numbers is interrupted for as long as
the SECS bit is ‘1’. If a number is available in the RNG_DR register, it must not be used
because it may not have enough entropy.
What you should do is clear the SEIS bit, then clear and set the RNGEN bit to reinitialize
and restart the RNG.

24.4

RNG registers
The RNG is associated with a control register, a data register and a status register. They
have to be accessed by words (32 bits).

768/1747

RM0090 Rev 17

RM0090

Random number generator (RNG)

24.4.1

RNG control register (RNG_CR)
Address offset: 0x00
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

6

5

4

19

18

17

16

1

0

Reserved
15

14

13

12

11

10

9

8

7

Reserved

3

2

IE

RNGEN

rw

rw

Reserved

Bits 31:4 Reserved, must be kept at reset value
Bit 3 IE: Interrupt enable
0: RNG Interrupt is disabled
1: RNG Interrupt is enabled. An interrupt is pending as soon as DRDY=1 or SEIS=1 or
CEIS=1 in the RNG_SR register.
Bit 2 RNGEN: Random number generator enable
0: Random number generator is disabled
1: random Number Generator is enabled.
Bits 1:0 Reserved, must be kept at reset value

24.4.2

RNG status register (RNG_SR)
Address offset: 0x04
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

15

14

13

12

11

10

9

8

23

22

21

20

19

4

3

18

17

16

Reserved
7

Reserved

6

5

SEIS

CEIS

rc_w0

rc_w0

Reserved

2

1

0

SECS

CECS

DRDY

r

r

r

Bits 31:3 Reserved, must be kept at reset value
Bit 6 SEIS: Seed error interrupt status
This bit is set at the same time as SECS, it is cleared by writing it to 0.
0: No faulty sequence detected
1: One of the following faulty sequences has been detected:
–
More than 64 consecutive bits at the same value (0 or 1)
–
More than 32 consecutive alternances of 0 and 1 (0101010101...01)
An interrupt is pending if IE = 1 in the RNG_CR register.
Bit 5 CEIS: Clock error interrupt status
This bit is set at the same time as CECS, it is cleared by writing it to 0.
0: The RNG_CLK clock was correctly detected
1: The RNG_CLK was not correctly detected (fRNG_CLK< fHCLK/16)
An interrupt is pending if IE = 1 in the RNG_CR register.
Bits 4:3 Reserved, must be kept at reset value

RM0090 Rev 17

769/1747
771

Random number generator (RNG)

RM0090

Bit 2 SECS: Seed error current status
0: No faulty sequence has currently been detected. If the SEIS bit is set, this means that a
faulty sequence was detected and the situation has been recovered.
1: One of the following faulty sequences has been detected:
–
More than 64 consecutive bits at the same value (0 or 1)
–
More than 32 consecutive alternances of 0 and 1 (0101010101...01)
Bit 1 CECS: Clock error current status
0: The RNG_CLK clock has been correctly detected. If the CEIS bit is set, this means that a
clock error was detected and the situation has been recovered
1: The RNG_CLK was not correctly detected (fRNG_CLK< fHCLK/16).
Bit 0 DRDY: Data ready
0: The RNG_DR register is not yet valid, no random data is available
1: The RNG_DR register contains valid random data
Note: An interrupt is pending if IE = 1 in the RNG_CR register.
Once the RNG_DR register has been read, this bit returns to 0 until a new valid value is
computed.

24.4.3

RNG data register (RNG_DR)
Address offset: 0x08
Reset value: 0x0000 0000
The RNG_DR register is a read-only register that delivers a 32-bit random value when read.
After being read, this register delivers a new random value after a maximum time of 40
periods of the RNG_CLK clock. The software must check that the DRDY bit is set before
reading the RNDATA value.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

RNDATA

RNDATA
r

r

r

r

r

r

r

r

r

Bits 31:0 RNDATA: Random data
32-bit random data.

770/1747

RM0090 Rev 17

RM0090

24.4.4

Random number generator (RNG)

RNG register map
Table 116 gives the RNG register map and reset values.
Table 116. RNG register map and reset map

Reserved
CECS

DRDY

RNG_DR
0x0000000

IE

0x08

Reserved

RNGEN

RNG_SR
0x0000000

SECS

0x04

Reserved

Reserved

RNG_CR
0x0000000

SEIS

0x00

CEIS

reset value

Register size
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Offset

Register
name

RNDATA[31:0]

RM0090 Rev 17

771/1747
771

Hash processor (HASH)

25

RM0090

Hash processor (HASH)
This section applies to STM32F415/417xx and STM32F43xxx devices.

25.1

HASH introduction
The hash processor is a fully compliant implementation of the secure hash algorithm
(SHA-1, SHA-224, SHA-256), the MD5 (message-digest algorithm 5) hash algorithm and
the HMAC (keyed-hash message authentication code) algorithm suitable for a variety of
applications. It computes a message digest (160 bits for the SHA-1 algorithm, 256 bits for
the SHA-256 algorithm and 224 bits for the SHA-224 algorithm,128 bits for the MD5
algorithm) for messages of up to (264 – 1) bits, while HMAC algorithms provide a way of
authenticating messages by means of hash functions. HMAC algorithms consist in calling
the SHA-1, SHA-224, SHA-256 or MD5 hash function twice.

25.2

HASH main features
•

Note:

772/1747

Suitable for data authentication applications, compliant with:
–

FIPS PUB 180-2 (Federal Information Processing Standards Publication 180-2)

–

Secure Hash Standard specifications (SHA-1, SHA-224 and SHA-256)

–

IETF RFC 1321 (Internet Engineering Task Force Request For Comments number
1321) specifications (MD5)

•

Fast computation of SHA-1, SHA-224 and SHA-256, and MD5 (SHA-224 and SHA-256
are available on STM32F43xxx only)

•

AHB slave peripheral

•

32-bit data words for input data, supporting word, half-word, byte and bit bit-string
representations, with little-endian data representation only.

•

Automatic swapping to comply with the big-endian SHA1, SHA-224 and SHA-256
computation standard with little-endian input bit-string representation

•

Automatic padding to complete the input bit string to fit modulo 512 (16 × 32 bits)
message digest computing

•

5× 32-bit words (H0 to H5) on STM32F415/417xx and 8 × 32-bit words (H0 to H7) on
STM32F43xxx for output message digest, reload able to continue interrupted message
digest computation.

•

Corresponding 32-bit words of the digest from consecutive message blocks are added
to each other to form the digest of the whole message

•

Automatic data flow control with support for direct memory access (DMA)

Padding, as defined in the SHA-1, SHA-224 and SHA-256 algorithm, consists in adding a bit
at bx1 followed by N bits at bx0 to get a total length congruent to 448 modulo 512. After this,
the message is completed with a 64-bit integer which is the binary representation of the
original message length.
For this hash processor, the quanta for entering the message is a 32-bit word, so an
additional information must be provided at the end of the message entry, which is the
number of valid bits in the last 32-bit word entered.

RM0090 Rev 17

RM0090

25.3

Hash processor (HASH)

HASH functional description
Figure 1 shows the block diagram of the hash processor.
Figure 233. Block diagram for STM32F415/417xx
 BIT !(" BUS
). BUFFER
$ATA
REGISTER
(!3(?$).

#ONTROL AND STATUS
REGISTERS

WRITE INTO (!3(?$).
OR WRITE $#!, BIT TO 
OR  COMPLETE BLOCK
TRANSFERRED BY THE $-!

)NTERRUPT REGISTERS
(!3(?)-2
 §  BIT
). &)&/

(!3(?32
#ONTROL REGISTER
(!3(?#2
3TART REGISTER

SWA PPIN G

(!3(?34 2

#ONTEXT SWAPPING
(!3(?#32
-ESSAGE DIGEST
(!3(?((

). &)&/ FULL
OR $#!, WRITTEN TO 
CONTEXT
DIGEST

3(!   -$
(ASH  (-!#
PROCESSOR CORE

AI

RM0090 Rev 17

773/1747
797

Hash processor (HASH)

RM0090
Figure 234. Block diagram for STM32F43xxx
 BIT !(" BUS
). BUFFER
$ATA
REGISTER
(!3(?$).

#ONTROL AND STATUS
REGISTERS

WRITE INTO (!3(?$).
OR WRITE $#!, BIT TO 
OR  COMPLETE BLOCK
TRANSFERRED BY THE $-!

)NTERRUPT REGISTERS
(!3(?)-2
 §  BIT
). &)&/

(!3(?32
#ONTROL REGISTER
(!3(?#2
3TART REGISTER

SWA PPIN G

(!3(?34 2

). &)&/ FULL
OR $#!, WRITTEN TO 

#ONTEXT SWAPPING
(!3(?#32
-ESSAGE DIGEST
(!3(?((

CONTEXT
DIGEST

3(!  3(!  3(!  AND -$
(ASH  (-!# PROCESSOR CORE

-36

The FIPS PUB 180-2 standard and the IETF RFC 1321 publication specify the SHA-1, SHA224 and SHA-256 and MD5 secure hash algorithms, respectively, for computing a
condensed representation of a message or data file. When a message of any length below
264 bits is provided on input, the SHA-1, SHA-224 and SHA-256 and MD5 produce
respective a 160-bit, 224 bit, 256 bit and 128-bit output string, respectively, called a
message digest. The message digest can then be processed with a digital signature
algorithm in order to generate or verify the signature for the message. Signing the message
digest rather than the message often improves the efficiency of the process because the
message digest is usually much smaller in size than the message. The verifier of a digital
signature has to use the same hash algorithm as the one used by the creator of the digital
signature.
The SHA-1, SHA-224 and SHA-256 and MD5 are qualified as “secure” because it is
computationally infeasible to find a message that corresponds to a given message digest, or
to find two different messages that produce the same message digest. Any change to a
message in transit will, with very high probability, result in a different message digest, and
the signature will fail to verify. For more detail on the SHA-1 or SHA-224 and SHA-256
algorithm, please refer to the FIPS PUB 180-2 (Federal Information Processing Standards
Publication 180-2), 2002 august 1.
The current implementation of this standard works with little-endian input data convention.
For example, the C string “abc” must be represented in memory as the 24-bit hexadecimal
value 0x434241.
A message or data file to be processed by the hash processor should be considered a bit
string. The length of the message is the number of bits in the message (the empty message

774/1747

RM0090 Rev 17

RM0090

Hash processor (HASH)
has length 0). You can consider that 32 bits of this bit string forms a 32-bit word. Note that
the FIPS PUB 180-1 standard uses the convention that bit strings grow from left to right, and
bits can be grouped as bytes (8 bits) or words (32 bits) (but some implementations also use
half-words (16 bits), and implicitly, uses the big-endian byte (half-word) ordering. This
convention is mainly important for padding (see Section 1.3.4: Message padding on
page 12).

25.3.1

Duration of the processing
The computation of an intermediate block of a message takes:
•

66 HCLK clock cycles in SHA-1

•

50 HCLK clock cycles in SHA-224

•

50 HCLK clock cycles in SHA-256

•

50 HCLK clock cycles in MD5

to which you must add the time needed to load the 16 words of the block into the processor
(at least 16 clock cycles for a 512-bit block).
The time needed to process the last block of a message (or of a key in HMAC) can be
longer. This time depends on the length of the last block and the size of the key (in HMAC
mode). Compared to the processing of an intermediate block, it can be increased by a factor
of:

25.3.2

•

1 to 2.5 for a hash message

•

around 2.5 for an HMAC input-key

•

1 to 2.5 for an HMAC message

•

around 2.5 for an HMAC output key in case of a short key

•

3.5 to 5 for an HMAC output key in case of a long key

Data type
Data are entered into the hash processor 32 bits (word) at a time, by writing them into the
HASH_DIN register. But the original bit-string can be organized in bytes, half-words or
words, or even be represented as bits. As the system memory organization is little-endian
and SHA1, SHA-224 and SHA-256 computation is big-endian, depending on the way the
original bit string is grouped, a bit, byte, or half-word swapping operation is performed
automatically by the hash processor.
The kind of data to be processed is configured with the DATATYPE bitfield in the HASH
control register (HASH_CR).

RM0090 Rev 17

775/1747
797

Hash processor (HASH)

RM0090
Figure 235. Bit, byte and half-word swapping

! )N CASE OF BINARY DATA HASH ALL BITS SHOULD BE SWAPPED AS BELOW
$!4 !
 4 90%BX

"ITSW
 APP ING OPERA TION
(!3
 (?$). BIT BIT





BIT 



BIT 

BIT 

"ITS ENTRED WITH LITTLE %NDIAN FORMAT
PA DDING IS PERFORMED
ON THIS SIDE OF THE
BITSTRIN G

BIT STRING ORGANIZATION IN
(ASH PROCESSOR "IG %NDIAN

 BIT

BIT 

BIT BIT BIT

hv hSv

BIT STRING GROWS
IN TH ISDIRECTION
ASDE FIN ED BY
 5"  STD 
&)030

" )N CASE OF BYTE DATA HASH ALL BYTES SHOULD BE SWAPPED AS BELOW
$!
 4! 4 9 0%B X

"YTESW
 APP INGOPERA TION
"YTE 
"YTE 
  BITS


BITS


BITS

"YTE 
"YTE 
"IT STRING ORGANIZATION IN
(ASH PROCESSOR "IG %NDIAN BITS BITS

BITS 

"YTE 
  

BITS 

(!3
 (?$).

BITS 

"YTE 

"YTE 

"YTE 
 

"YTES ENTRED WITH LITTLE %NDIAN FORMAT

BIT STRING GROWS
IN TH ISDIRECTION
ASDE FIN ED BY
&)030
 5"  STD 

# )N CASE OF HALF WORD HASH ALL HALF WORD SHOULD BE SWAPPED AS BELOW
(ALFW
 ORDSWAPPINGOPE RATION
(!3
 ( ?$ ).

"IT STRING ORGANIZATION IN
(ASH PROCESSOR "IG %NDIAN


BITS

(ALF WORD 


BITS

(ALF WORD 

 4! 4 9 0%B X
$!

BITS

(ALF WORD 


BITS

(ALF WORD 

HALF WORD ENTRED WITH LITTLE %NDIAN FORMAT

BIT STRINGGROW
S
IN THISDIRECTIO N
AS DEFINE D BY
&)0
 305"  STD 
AIB

776/1747

RM0090 Rev 17

RM0090

Hash processor (HASH)
The least significant bit of the message has to be at position 0 (right) in the first word
entered into the hash processor, the 32nd bit of the bit string has to be at position 0 in the
second word entered into the hash processor and so on.

25.3.3

Message digest computing
The HASH sequentially processes blocks of 512 bits when computing the message digest.
Thus, each time 16 × 32-bit words (= 512 bits) have been written by the DMA or the CPU,
into the hash processor, the HASH automatically starts computing the message digest. This
operation is known as a partial digest computation.
The message to be processed is entered into the peripheral by 32-bit words written into the
HASH_DIN register. The current contents of the HASH_DIN register are transferred to the
input FIFO (IN FIFO) each time the register is written with new data. HASH_DIN and the
input FIFO form a FIFO of a 17-word length (named the IN buffer).
The processing of a block can start only once the last value of the block has entered the IN
FIFO. The peripheral must get the information as to whether the HASH_DIN register
contains the last bits of the message or not. Two cases may occur:
•

•

When the DMA is not used:
–

In case of a partial digest computation, this is done by writing an additional word
into the HASH_DIN register (actually the first word of the next block). Then the
software must wait until the processor is ready again (when DINIS=1) before
writing new data into HASH_DIN.

–

In case of a final digest computation (last block entered), this is done by writing the
DCAL bit to 1.

When the DMA is used:
The contents of the HASH_DIN register are interpreted automatically with the
information sent by the DMA controller.
–

In case of a single DMA transfer: Multiple DMA transfer (MDMAT) bit should be
cleared on STM32F43xxx. When the last block has been transferred to the
HASH_DIN register via DMA channel, DCAL bit will be set to automatically to 1 in
the HASH_STR register in order to launch the final digest calculation.

–

In case of a multiple DMA transfer (available only on STM32F43xxx): Multiple
DMA transfer (MDMAT) bit should be set to 1 by software so DCAL bit does not
get set automatically by HW, in this case the final digest calculation for hash and
for each phases for HMAC (for more details about HMAC phases please refer to
HMAC operation section) will not be launched a the end of the DMA transfer
request, allowing the processor to receive a new DMA transfer. During the last
DMA transfer, Multiple DMA transfer (MDMAT) bit should be cleared by software in
order to set automatically DCAL bit at the end of the last bloc and lunch the final
digest.

–

The contents of the HASH_DIN register are interpreted automatically with the
information sent by the DMA controller.

This process —data entering + partial digest computation— continues until the last bits of
the original message are written to the HASH_DIN register. As the length (number of bits) of
a message can be any integer value, the last word written into the HASH processor may
have a valid number of bits between 1 and 32. This number of valid bits in the last word,
NBLW, has to be written into the HASH_STR register, so that message padding is correctly
performed before the final message digest computation.

RM0090 Rev 17

777/1747
797

Hash processor (HASH)

RM0090

Once this is done, writing into HASH_STR with bit DCAL = 1 starts the processing of the last
entered block of message by the hash processor. This processing consists in:
•

Automatically performing the message padding operation: the purpose of this operation
is to make the total length of a padded message a multiple of 512. The HASH
sequentially processes blocks of 512 bits when computing the message digest

•

Computing the final message digest

When the DMA is enabled, it provides the information to the hash processor when it is
transferring the last data word. Then the padding and digest computation are performed
automatically as if DCAL had been written to 1.

25.3.4

Message padding
Message padding consists in appending a “1” followed by m “0”s followed by a 64-bit integer
to the end of the original message to produce a padded message block of length 512. The
“1” is added to the last word written into the HASH_DIN register at the bit position defined by
the NBLW bitfield, and the remaining upper bits are cleared (“0”s).
Example: let us assume that the original message is the ASCII binary-coded form of “abc”,
of length L = 24:
byte 0
byte 1
byte 2
byte 3
01100001 01100010 01100011 UUUUUUUU
<-- 1st word written to HASH_DIN -->
NBLW has to be loaded with the value 24: a “1” is appended at bit location 24 in the bit string
(starting counting from left to right in the above bit string), which corresponds to bit 31 in the
HASH_DIN register (little-endian convention):
01100001 01100010 01100011 1UUUUUUU
Since L = 24, the number of bits in the above bit string is 25, and 423 “0”s are appended,
making now 448. This gives (in hexadecimal, big-endian format):
61626380
00000000
00000000
00000000

00000000 00000000 00000000
00000000 00000000 00000000
00000000 00000000 00000000
00000000

The L value, in two-word representation (that is 00000000 00000018) is appended. Hence
the final padded message in hexadecimal:
61626380
00000000
00000000
00000000

00000000
00000000
00000000
00000000

00000000
00000000
00000000
00000000

00000000
00000000
00000000
00000028

If the HASH is programmed to use the little-endian byte input format, the above message
has to be entered by doing the following steps:

778/1747

1.

0xUU636261 is written into the HASH_DIN register (where ‘U’ means don’t care)

2.

0x18 is written into the HASH_STR register (the number of valid bits in the last word
written into the HASH_DIN register is 24, as the original message length is 24 bits)

3.

0x10 is written into the HASH_STR register to start the message padding and digest
computation. When NBLW ≠ 0x00, the message padding puts a “1” into the HASH_DIN
register at the bit position defined by the NBLW value, and inserts “0”s at bit locations
[31:(NBLW+1)]. When NBLW == 0x00, the message padding inserts one new word with

RM0090 Rev 17

RM0090

Hash processor (HASH)
value 0x0000 0001. Then an all zero word (0x0000 0000) is added and the message
length in a two-word representation, to get a block of 16 x 32-bit words.
4.

The HASH computing is performed, and the message digest is then available in the
HASH_Hx registers (x = 0...4) for the SHA-1 algorithm. For example:
H0
H1
H2
H3
H4

25.3.5

=
=
=
=
=

0xA9993E36
0x4706816A
0xBA3E2571
0x7850C26C
0x9CD0D89D

Hash operation
The hash function (SHA-1, SHA-224, SHA-256 and MD5) is selected when the INIT bit is
written to ‘1’ in the HASH_CR register while the MODE bit is at ‘0’ in HASH_CR. The
algorithm (SHA-1, SHA-224,SHA-256 or MD5) is selected at the same time (that is when the
INIT bit is set) using the ALGO bits.
The message can then be sent by writing it word by word into the HASH_DIN register. When
a block of 512 bits —that is 16 words— has been written, a partial digest computation starts
upon writing the first data of the next block. The hash processor remains busy for 66 cycles
for the SHA-1 algorithm, or 50 cycles for the MD5 algorithm, SHA-224 algorithm and SHA256 algorithm.
The process can then be repeated until the last word of the message. If DMA transfers are
used, refer to the Procedure where the data are loaded by DMA section. Otherwise, if the
message length is not an exact multiple of 512 bits, then the HASH_STR register has to be
written to launch the computation of the final digest.
Once computed, the digest can be read from the HASH_H0...HASH_H4 registers (for the
MD5 algorithm, HASH_H4 is not relevant) on STM32F415/417xx, and from the
HASH_H0...HASH_H7 registers on STM32F43xxx where:
HASH_H4..HASH_H7 are not relevant when the MD5 algorithm is selected,
HASH_H5.. HASH_H7 are not relevant when the SHA-1algorithm is selected,
HASH_H7 is not relevant when the SHA-224 algorithm is selected.

25.3.6

HMAC operation
The HMAC algorithm is used for message authentication, by irreversibly binding the
message being processed to a key chosen by the user. For HMAC specifications, refer to
“HMAC: keyed-hashing for message authentication, H. Krawczyk, M. Bellare, R. Canetti,
February 1997.
Basically, the algorithm consists of two nested hash operations:
HMAC(message) = Hash[((key | pad) XOR 0x5C)
| Hash(((key | pad) XOR 0x36) | message)]
where:
•

pad is a sequence of zeroes needed to extend the key to the length of the underlying
hash function data block (that is 512 bits for both the SHA-1, SHA224, SHA-256 and
MD5 hash algorithms)

•

| represents the concatenation operator

To compute the HMAC, four different phases are required:

RM0090 Rev 17

779/1747
797

Hash processor (HASH)

Note:

1

25.3.7

RM0090

1.

The block is initialized by writing the INIT bit to ‘1’ with the MODE bit at ‘1’ and the
ALGO bits set to the value corresponding to the desired algorithm. The LKEY bit must
also be set during this phase if the key being used is longer than 64 bytes (in this case,
the HMAC specifications specify that the hash of the key should be used in place of the
real key).

2.

The key (to be used for the inner hash function) is then given to the core. This
operation follows the same mechanism as the one used to send the message in the
hash operation (that is, by writing into HASH_DIN and, finally, into HASH_STR).

3.

Once the last word has been entered and computation has started, the hash processor
elaborates the key. It is then ready to accept the message text using the same
mechanism as the one used to send the message in the hash operation.

4.

After the first hash round, the hash processor returns “ready” to indicate that it is ready
to receive the key to be used for the outer hash function (normally, this key is the same
as the one used for the inner hash function). When the last word of the key is entered
and computation starts, the HMAC result is made available in the
HASH_H0...HASH_H4 registers on STM32F415/417xx and on HASH_H0...HASH_H7
registers on STM32F43xxx.

The computation latency of the HMAC primitive depends on the lengths of the keys and
message. You could the HMAC as two nested underlying hash functions with the same key
length (long or short).

Context swapping
It is possible to interrupt a hash/HMAC process to perform another processing with a higher
priority, and to complete the interrupted process later on, when the higher-priority task is
complete. To do so, the context of the interrupted task must be saved from the hash
registers to memory, and then be restored from memory to the hash registers.
The procedures where the data flow is controlled by software or by DMA are described
below.

780/1747

RM0090 Rev 17

RM0090

Hash processor (HASH)

Procedure where the data are loaded by software
The context can be saved only when no block is currently being processed. That is, you
must wait for DINIS = 1 (the last block has been processed and the input FIFO is empty) or
NBW ≠ 0 (the FIFO is not full and no processing is ongoing).
•

Context saving:
Store the contents of the following registers into memory:

•

–

HASH_IMR

–

HASH_STR

–

HASH_CR

–

HASH_CSR0 to HASH_CSR50 on STM32F415/417xx, and HASH_CSR0 to
HASH_CSR53 on STM32F43xxx.

Context restoring:
The context can be restored when the high-priority task is complete. Please follow the
order of the sequence below.
a)

Write the following registers with the values saved in memory: HASH_IMR,
HASH_STR and HASH_CR

b)

Initialize the hash processor by setting the INIT bit in the HASH_CR register

c)

Write the HASH_CSR0 to HASH_CSR50 (STM32F415/417xx), and HASH_CSR0
to HASH_CSR53 (STM32F43xxx) registers with the values saved in memory

You can now restart the processing from the point where it has been interrupted.

Procedure where the data are loaded by DMA
In this case it is not possible to predict if a DMA transfer is in progress or if the process is
ongoing. Thus, you must stop the DMA transfers, then wait until the HASH is ready in order
to interrupt the processing of a message.
•

•

Interrupting a processing:
–

Clear the DMAE bit to disable the DMA interface

–

Wait until the current DMA transfer is complete (wait for DMAES = 0 in the
HASH_SR register). Note that the block may or not have been totally transferred
to the HASH.

–

Disable the corresponding channel in the DMA controller

–

Wait until the hash processor is ready (no block is being processed), that is wait
for DINIS = 1

The context saving and context restoring phases are the same as above (see
Procedure where the data are loaded by software).

Reconfigure the DMA controller so that it transfers the end of the message. You can now
restart the processing from the point where it was interrupted by setting the DMAE bit.
Note:

If context swapping does not involve HMAC operations, the HASH_CSR38 to
HASH_CSR50 (STM32F415/417xx) and HASH_CSR38 to HASH_CSR53 (STM32F43xxx)
registers do not have to be saved and restored.
If context swapping occurs between two blocks (the last block was completely processed
and the next block has not yet been pushed into the IN FIFO, NBW = 000 in the HASH_CR
register), the HASH_CSR22 to HASH_CSR37 registers do not have to be saved and
restored.

RM0090 Rev 17

781/1747
797

Hash processor (HASH)

25.3.8

RM0090

HASH interrupt
There are two individual maskable interrupt sources generated by the HASH processor.
They are connected to the same interrupt vector.
You can enable or disable the interrupt sources individually by changing the mask bits in the
HASH_IMR register. Setting the appropriate mask bit to 1 enables the interrupt.
The status of the individual interrupt sources can be read from the HASH_SR register.
Figure 236. HASH interrupt mapping diagram
$#)3
$#)(!3( INTERRUPT TO .6)#
$).)3
$).)AI

25.4

HASH registers
The HASH core is associated with several control and status registers and five message
digest registers.
All these registers are accessible through word accesses only, else an AHB error is
generated.

25.4.1

HASH control register (HASH_CR) for STM32F415/417xx
Address offset: 0x00
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

Reserved
15

14
Reserved

13

12

11

10

DINNE
r

9

8

NBW
r

r

rw
7

6

ALGO[0] MODE
r

r

16
LKEY

rw

rw

5

4

DATATYPE
rw

rw

3

2

DMAE

INIT

rw

w

1

0

Reserved

Bits 31:17 Reserved, forced by hardware to 0.
Bit 16 LKEY: Long key selection
This bit selects between short key (≤64 bytes) or long key (> 64 bytes) in HMAC
mode
0: Short key (≤64 bytes)
1: Long key (> 64 bytes)
Note: This selection is only taken into account when the INIT bit is set and MODE
= 1. Changing this bit during a computation has no effect.
Bits 15:13 Reserved, forced by hardware to 0.

782/1747

RM0090 Rev 17

RM0090

Hash processor (HASH)

Bit 12 DINNE: DIN not empty
This bit is set when the HASH_DIN register holds valid data (that is after being
written at least once). It is cleared when either the INIT bit (initialization) or the
DCAL bit (completion of the previous message processing) is written to 1.
0: No data are present in the data input buffer
1: The input buffer contains at least one word of data
Bits 11:8 NBW: Number of words already pushed
This bitfield reflects the number of words in the message that have already been
pushed into the IN FIFO.
NBW increments (+1) when a write access is performed to the HASH_DIN
register while DINNE = 1.
It goes to 0000 when the INIT bit is written to 1 or when a digest calculation
starts (DCAL written to 1 or DMA end of transfer).
″
If the DMA is not used:
0000 and DINNE=0: no word has been pushed into the DIN buffer (the buffer is
empty, both the HASH_DIN register and the IN FIFO are empty)
0000 and DINNE=1: 1 word has been pushed into the DIN buffer (The
HASH_DIN register contains 1 word, the IN FIFO is empty)
0001: 2 words have been pushed into the DIN buffer (the HASH_DIN register
and the IN FIFO contain 1 word each)
...
1111: 16 words have been pushed into the DIN buffer
″
If the DMA is used, NBW is the exact number of words that have been
pushed into the IN FIFO.
Bit 7 ALGO[1:0]: Algorithm selection
These bits selects the SHA-1 or the MD5 algorithm:
0: SHA-1 algorithm selected
1: MD5 algorithm selected
Note: This selection is only taken into account when the INIT bit is set. Changing
this bit during a computation has no effect.
Bit 6 MODE: Mode selection
This bit selects the HASH or HMAC mode for the selected algorithm:
0: Hash mode selected
1: HMAC mode selected. LKEY must be set if the key being used is longer than
64 bytes.
Note: This selection is only taken into account when the INIT bit is set. Changing
this bit during a computation has no effect.
Bits 5:4 DATATYPE: Data type selection
Defines the format of the data entered into the HASH_DIN register:
00: 32-bit data. The data written into HASH_DIN are directly used by the HASH
processing, without reordering.
01: 16-bit data, or half-word. The data written into HASH_DIN are considered as
2 half-words, and are swapped before being used by the HASH processing.
10: 8-bit data, or bytes. The data written into HASH_DIN are considered as 4
bytes, and are swapped before being used by the HASH processing.
11: bit data, or bit-string. The data written into HASH_DIN are considered as 32
bits (1st bit of the sting at position 0), and are swapped before being used by the
HASH processing (1st bit of the string at position 31).

RM0090 Rev 17

783/1747
797

Hash processor (HASH)

RM0090

Bit 3 DMAE: DMA enable
0: DMA transfers disabled
1: DMA transfers enabled. A DMA request is sent as soon as the HASH core is
ready to receive data.
Note: 1: This bit is cleared by hardware when the DMA asserts the DMA terminal
count signal (while transferring the last data of the message). This bit is not
cleared when the INIT bit is written to 1.
2: If this bit is written to 0 while a DMA transfer has already been requested
to the DMA, DMAE is cleared but the current transfer is not aborted.
Instead, the DMA interface remains internally enabled until the transfer is
complete or INIT is written to 1.
Bit 2 INIT: Initialize message digest calculation
Writing this bit to 1 resets the hash processor core, so that the HASH is ready to
compute the message digest of a new message.
Writing this bit to 0 has no effect.
Reading this bit always return 0.
Bits 1:0 Reserved, must be kept cleared.

784/1747

RM0090 Rev 17

RM0090

25.4.2

Hash processor (HASH)

HASH control register (HASH_CR) for STM32F43xxx
Address offset: 0x00

31

30

29

28

27

26

25

24

23

22

21

20

19

Reserved
15

14

Reserved

13

12

11

10

r

r

MDMAT DINNE
rw

r

18

17

16

ALGO[1]

Reserved

Reset value: 0x0000 0000

LKEY

1

0

rw

9

8

r

r

NBW

7

6

ALGO[0] MODE
rw

rw

5

4

DATATYPE
rw

rw

3

2

DMAE

INIT

rw

w

rw

Reserved

Bits 31:19 Reserved, forced by hardware to 0.
Bit 17 Reserved, forced by hardware to 0.
Bit 16 LKEY: Long key selection
This bit selects between short key (≤64 bytes) or long key (> 64 bytes) in HMAC
mode
0: Short key (≤64 bytes)
1: Long key (> 64 bytes)
Note: This selection is only taken into account when the INIT bit is set and MODE
= 1. Changing this bit during a computation has no effect.
Bits 15:14 Reserved, forced by hardware to 0.
Bit 13 MDMAT: Multiple DMA Transfers
This bit is set when hashing large files when multiple DMA transfers are needed.
0: DCAL is automatically set at the end of a DMA transfer.
1: DCAL is not automatically set at the end of a DMA transfer.
Bit 12 DINNE: DIN not empty
This bit is set when the HASH_DIN register holds valid data (that is after being
written at least once). It is cleared when either the INIT bit (initialization) or the
DCAL bit (completion of the previous message processing) is written to 1.
0: No data are present in the data input buffer
1: The input buffer contains at least one word of data

RM0090 Rev 17

785/1747
797

Hash processor (HASH)

RM0090

Bits 11:8 NBW: Number of words already pushed
This bitfield reflects the number of words in the message that have already been
pushed into the IN FIFO.
NBW increments (+1) when a write access is performed to the HASH_DIN
register while DINNE = 1.
It goes to 0000 when the INIT bit is written to 1 or when a digest calculation
starts (DCAL written to 1 or DMA end of transfer).
″
If the DMA is not used:
0000 and DINNE=0: no word has been pushed into the DIN buffer (the buffer is
empty, both the HASH_DIN register and the IN FIFO are empty)
0000 and DINNE=1: 1 word has been pushed into the DIN buffer (The
HASH_DIN register contains 1 word, the IN FIFO is empty)
0001: 2 words have been pushed into the DIN buffer (the HASH_DIN register
and the IN FIFO contain 1 word each)
...
1111: 16 words have been pushed into the DIN buffer
″
If the DMA is used, NBW is the exact number of words that have been
pushed into the IN FIFO.
Bit 18 and bit 7 ALGO[1:0]: Algorithm selection
These bits selects the SHA-1, SHA-224, SHA256 or the MD5 algorithm:
00: SHA-1 algorithm selected
01: MD5 algorithm selected
10: SHA224 algorithm selected
11: SHA256 algorithm selected
Note: This selection is only taken into account when the INIT bit is set. Changing
this bit during a computation has no effect.
Bit 6 MODE: Mode selection
This bit selects the HASH or HMAC mode for the selected algorithm:
0: Hash mode selected
1: HMAC mode selected. LKEY must be set if the key being used is longer than
64 bytes.
Note: This selection is only taken into account when the INIT bit is set. Changing
this bit during a computation has no effect.
Bits 5:4 DATATYPE: Data type selection
Defines the format of the data entered into the HASH_DIN register:
00: 32-bit data. The data written into HASH_DIN are directly used by the HASH
processing, without reordering.
01: 16-bit data, or half-word. The data written into HASH_DIN are considered as
2 half-words, and are swapped before being used by the HASH processing.
10: 8-bit data, or bytes. The data written into HASH_DIN are considered as 4
bytes, and are swapped before being used by the HASH processing.
11: bit data, or bit-string. The data written into HASH_DIN are considered as 32
bits (1st bit of the sting at position 0), and are swapped before being used by the
HASH processing (1st bit of the string at position 31).

786/1747

RM0090 Rev 17

RM0090

Hash processor (HASH)

Bit 3 DMAE: DMA enable
0: DMA transfers disabled
1: DMA transfers enabled. A DMA request is sent as soon as the HASH core is
ready to receive data.
Note: 1: This bit is cleared by hardware when the DMA asserts the DMA terminal
count signal (while transferring the last data of the message). This bit is not
cleared when the INIT bit is written to 1.
2: If this bit is written to 0 while a DMA transfer has already been requested
to the DMA, DMAE is cleared but the current transfer is not aborted.
Instead, the DMA interface remains internally enabled until the transfer is
complete or INIT is written to 1.
Bit 2 INIT: Initialize message digest calculation
Writing this bit to 1 resets the hash processor core, so that the HASH is ready to
compute the message digest of a new message.
Writing this bit to 0 has no effect.
Reading this bit always return 0.
Bits 1:0 Reserved, must be kept cleared.

RM0090 Rev 17

787/1747
797

Hash processor (HASH)

25.4.3

RM0090

HASH data input register (HASH_DIN)
Address offset: 0x04
Reset value: 0x0000 0000
HASH_DIN is the data input register. It is 32-bit wide. It is used to enter the message by
blocks of 512 bits. When the HASH_DIN register is written to, the value presented on the
AHB databus is ‘pushed’ into the HASH core and the register takes the new value presented
on the AHB databus. The DATATYPE bits must previously have been configured in the
HASH_CR register to get a correct message representation.
When a block of 16 words has been written to the HASH_DIN register, an intermediate
digest calculation is launched:
•

by writing new data into the HASH_DIN register (the first word of the next block) if the
DMA is not used (intermediate digest calculation)

•

automatically if the DMA is used

When the last block has been written to the HASH_DIN register, the final digest calculation
(including padding) is launched:
•

by writing the DCAL bit to 1 in the HASH_STR register (final digest calculation)

•

automatically if the DMA is used and MDMAT bit is set to ‘0’.

When a digest calculation (intermediate or final) is in progress, any new write access to the
HASH_DIN register is extended (by wait-state insertion on the AHB bus) until the HASH
calculation completes.
When the HASH_DIN register is read, the last word written in this location is accessed (zero
after reset).
.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

DATAIN
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

DATAIN
rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 DATAIN: Data input
Read = returns the current register content.
Write = the current register content is pushed into the IN FIFO, and the register
takes the new value presented on the AHB databus.

788/1747

RM0090 Rev 17

RM0090

Hash processor (HASH)

25.4.4

HASH start register (HASH_STR)
Address offset: 0x08
Reset value: 0x0000 0000
The HASH_STR register has two functions:

31

30

•

It is used to define the number of valid bits in the last word of the message entered in
the hash processor (that is the number of valid least significant bits in the last data
written into the HASH_DIN register)

•

It is used to start the processing of the last block in the message by writing the DCAL
bit to 1
29

28

27

26

25

24

23

22

21

20

19

6

5

4

3

18

17

16

2

1

0

rw

rw

Reserved
15

14

13

12

11

Reserved

10

9

8
DCAL
w

7

Reserved

NBLW
rw

rw

rw

Bits 31:9 Reserved, forced by hardware to 0.
Bit 8 DCAL: Digest calculation
Writing this bit to 1 starts the message padding, using the previously written
value of NBLW, and starts the calculation of the final message digest with all data
words written to the IN FIFO since the INIT bit was last written to 1.
Reading this bit returns 0.
Note
Bits 7:5 Reserved, forced by hardware to 0.
Bits 4:0 NBLW: Number of valid bits in the last word of the message in the bit string
organization of hash processor
When these bits are written and DCAL is at ‘0’, they take the value on the AHB
databus:
0x00: All 32 bits of the last data written in the bit string organization of hash
processor (after data swapping) are valid.
0x01: Only bit [31] of the last data written in the bit string organization of hash
processor (after data swapping) are valid
0x02: Only bits [31:30] of the last data written in the bit string organization of
hash processor (after data swapping) are valid
0x03: Only bits [31:29] of the last data written in the bit string organization of
hash processor (after data swapping) are valid
...
0x1F: Only bits [0] of the last data written in the bit string organization of hash
processor (after data swapping) are valid
When these bits are written and DCAL is at ‘1’, the bitfield is not changed.
Reading them returns the last value written to NBLW.
Note: These bits must be configured before setting the DCAL bit, else they are not
taken into account. Especially, it is not possible to configure NBLW and set
DCAL at the same time.

RM0090 Rev 17

789/1747
797

Hash processor (HASH)

25.4.5

RM0090

HASH digest registers (HASH_HR0..4/5/6/7)
Address offset: 0x0C to 0x1C (STM32F415/417xx), plus 0x310 to 0x32C (STM32F43xxx)
Reset value: 0x0000 0000
These registers contain the message digest result named as:
1.

H0, H1, H2, H3 and H4, respectively, in the SHA1 algorithm description
Note that in this case, the HASH_H5 to HASH_H7 register is not used, and is read as
zero.

2.

A, B, C and D, respectively, in the MD5 algorithm description
Note that in this case, the HASH_H4 to HASH_H7 register is not used, and is read as
zero.

3.

H0 to H6, respectively, in the SHA224 algorithm description,
Note that in this case, the HASH_H7 register is not used, and is read as zero.

4.

H0 to H7, respectively, in the SHA256 algorithm description,

If a read access to one of these registers occurs while the HASH core is calculating an
intermediate digest or a final message digest (that is when the DCAL bit has been written to
1), then the read is stalled until the completion of the HASH calculation.
Note:

H0, H1, H2, H3 and H4 mapping are duplicated in two region.

HASH_HR0
Address offset: 0x0C and 0x310
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

r

r

r

r

r

r

r

r

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

23

22

21

20

19

18

17

16

H0
r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8
H0

r

r

r

r

r

r

r

HASH_HR1
Address offset: 0x10 and 0x314
31

30

29

28

27

26

25

24

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

23

22

21

20

19

18

17

16

r

r

r

r

r

r

r

r

H1

H1
r

r

r

r

r

r

r

r

HASH_HR2
Address offset: 0x14 and 0x318
31

30

29

28

27

26

25

24
H2

r

790/1747

r

r

r

r

r

r

r

RM0090 Rev 17

RM0090

15

Hash processor (HASH)

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

23

22

21

20

19

18

17

16

H2
r

r

r

r

r

r

r

r

HASH_HR3
Address offset: 0x18 and 0x31C
31

30

29

28

27

26

25

24
H3

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

23

22

21

20

19

18

17

16

H3

HASH_HR4
Address offset: 0x1C and 0x320
31

30

29

28

27

26

25

24
H4

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

26

25

24

23

22

21

20

19

18

17

16

H4

HASH_HR5
Address offset: 0x324
31

30

29

28

27

H5
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

26

25

24

23

22

21

20

19

18

17

16

H5

HASH_HR6
Address offset: 0x328
31

30

29

28

27

H6
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

H6

HASH_HR7
Address offset: 0x32C

RM0090 Rev 17

791/1747
797

Hash processor (HASH)

RM0090

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

H7

H7
r

r

r

r

r

r

r

r

Note:

When starting a digest computation for a new bit stream (by writing the INIT bit to 1), these
registers assume their reset values.

25.4.6

HASH interrupt enable register (HASH_IMR)
Address offset: 0x20
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

6

5

4

3

2

17

16

Reserved
15

14

13

12

11

10

9

8

7

Reserved

Bits 31:2 Reserved, forced by hardware to 0.
Bit 1 DCIE: Digest calculation completion interrupt enable
0: Digest calculation completion interrupt disabled
1: Digest calculation completion interrupt enabled.
Bit 0 DINIE: Data input interrupt enable
0: Data input interrupt disabled
1: Data input interrupt enabled

792/1747

RM0090 Rev 17

1

0

DCIE

DINIE

rw

rw

RM0090

Hash processor (HASH)

25.4.7

HASH status register (HASH_SR)
Address offset: 0x24
Reset value: 0x0000 0001

31

30

29

28

27

26

25

24

23

22

21

20

6

5

4

19

18

17

16

Reserved
15

14

13

12

11

10

9

8

7

Reserved

3

2

1

0

BUSY

DMAS

DCIS

DINIS

r

r

rc_w0

rc_w0

Bits 31:4 Reserved, forced by hardware to 0.
Bit 3 BUSY: Busy bit
0: No block is currently being processed
1: The hash core is processing a block of data
Bit 2 DMAS: DMA Status
This bit provides information on the DMA interface activity. It is set with DMAE
and cleared when DMAE=0 and no DMA transfer is ongoing. No interrupt is
associated with this bit.
0: DMA interface is disabled (DMAE=0) and no transfer is ongoing
1: DMA interface is enabled (DMAE=1) or a transfer is ongoing
Bit 1 DCIS: Digest calculation completion interrupt status
This bit is set by hardware when a digest becomes ready (the whole message
has been processed). It is cleared by writing it to 0 or by writing the INIT bit to 1
in the HASH_CR register.
0: No digest available in the HASH_Hx registers
1: Digest calculation complete, a digest is available in the HASH_Hx registers.
An interrupt is generated if the DCIE bit is set in the HASH_IMR register.
Bit 0 DINIS: Data input interrupt status
This bit is set by hardware when the input buffer is ready to get a new block (16
locations are free). It is cleared by writing it to 0 or by writing the HASH_DIN
register.
0: Less than 16 locations are free in the input buffer
1: A new block can be entered into the input buffer. An interrupt is generated if
the DINIE bit is set in the HASH_IMR register.

RM0090 Rev 17

793/1747
797

Hash processor (HASH)

25.4.8

RM0090

HASH context swap registers (HASH_CSRx)
Address offset: 0x0F8 to 0x1C0
•

For HASH_CSR0 register: Reset value is 0x0000 0002.

•

For others registers: Reset value is 0x0000 0000 , except for STM32F43xxx devices
where the HASH_CSR2 register reset value is 0x2000 0000

Additional registers are available from 0x1C1 to 0x1CC on STM32F43xxx:
•

Reset value: 0x0000 0000.

These registers contain the complete internal register states of the hash processor, and are
useful when a context swap has to be done because a high-priority task has to use the hash
processor while it is already in use by another task.
When such an event occurs, the HASH_CSRx registers have to be read and the read
values have to be saved somewhere in the system memory space. Then the hash
processor can be used by the preemptive task, and when hash computation is finished, the
saved context can be read from memory and written back into these HASH_CSRx registers.

HASH_CSRx
Address offset: 0x0F8 to 0x1C0 on STM32F415/417xx
Address offset: 0x0F8 to 0x1CC on STM32F43xxx
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

rw

rw

rw

rw

rw

rw

rw

rw

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

CSx
rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8
CSx

rw

794/1747

rw

rw

rw

rw

rw

rw

rw

RM0090 Rev 17

RM0090

25.4.9

Hash processor (HASH)

HASH register map
Table 9 gives the summary HASH register map and reset values.
Table 117. HASH register map and reset values on STM32F415/417xx

HASH_CR

Reset value

0

DATAIN

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_STR

Reserved

Reset value

0x10
0x14
0x18
0x1C

0x20

H0

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
H1

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR2

H2

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR3

H3

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR4

H4

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_IMR

Reserved

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_SR

Reserved

Reset value
0xF8

0 0 0 0 0

HASH_HR1

Reset value
0x24

0

HASH_HR0

NBLW

DCIE
DINIE

0x0C

0 0 0 0 0 0 0 0 0 0 0

HASH_DIN

BUSY
DMAS
DCIS
DINIS

0x08

NBW

Reserved

Reserved

0x04

Register size

DCAL

0x00

Register name
reset value

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
LKEY
16
15
Reserved
14
13
DINNE
12
11
10
9
8
ALGO[0]
7
MODE
6
DATATYPE 5
4
DMAE
3
INIT
2
1
Reserved
0

Offset

0 0 0 1

HASH_CSR0

CSR0

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0
...
...

0x1C0

HASH_CSR50

CSR50

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Reserved

0x310
0x314

HASH_HR0

H0

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR1

H1

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

RM0090 Rev 17

795/1747
797

Hash processor (HASH)

RM0090

Table 117. HASH register map and reset values on STM32F415/417xx (continued)
Register name
reset value

0x318
0x31C
0x320

Register size
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Offset

HASH_HR2

H2

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR3

H3

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR4

H4

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Table 118. HASH register map and reset values on STM32F43xxx

HASH_CR

Register size

NBW

Reserved

Reset value
HASH_DIN

DATAIN

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_STR

Reserved

Reset value

0x10

0x14

0x18

0x1C

0x20

H0

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR1

H1

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR2

H2

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR3

H3

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR4

H4

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_IMR

HASH_SR

Reserved

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Reserved

Reset value

796/1747

0 0 0 0 0

HASH_HR0

Reset value

0x24

0

NBLW

DCIE
DINIE

0x0C

0 0 0 0 0 0 0 0 0 0 0 0

BUSY
DMAS
DCIS
DINIS

0x08

0

Reserved

0x04

0

DCAL

0x00

Register name
reset value

31
30
29
28
27
26
25
24
23
22
21
20
19
ALGO[1] 18
17
Reserved
16
LKEY
Reserved 15
14
MDMAT 13
12
DINNE
11
10
9
8
7
ALGO[0]
6
MODE
DATATYPE 5
4
3
DMAE
2
INIT
1
Reserved
0

Offset

0 0 0 1

RM0090 Rev 17

RM0090

Hash processor (HASH)
Table 118. HASH register map and reset values on STM32F43xxx (continued)

0xF8

Register size

Register name
reset value

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Offset

HASH_CSR0

CSR0

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0
...
...

0x1CC

HASH_CSR53

CSR53

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Reserved

0x310

0x314

0x318

0x31C

0x320

0x324

0x328

0x32C

HASH_HR0

H0

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR1

H1

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR2

H2

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR3

H3

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR4

H4

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR5

H5

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR6

H6

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

HASH_HR7

H7

Reset value

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

RM0090 Rev 17

797/1747
797

Real-time clock (RTC)

26

RM0090

Real-time clock (RTC)
This section applies to the whole STM32F4xx family, unless otherwise specified.

26.1

Introduction
The real-time clock (RTC) is an independent BCD timer/counter. The RTC provides a timeof-day clock/calendar, two programmable alarm interrupts, and a periodic programmable
wakeup flag with interrupt capability. The RTC also includes an automatic wakeup unit to
manage low-power modes.
Two 32-bit registers contain the seconds, minutes, hours (12- or 24-hour format), day (day
of week), date (day of month), month, and year, expressed in binary coded decimal format
(BCD). The sub-seconds value is also available in binary format.
Compensations for 28-, 29- (leap year), 30-, and 31-day months are performed
automatically. Daylight saving time compensation can also be performed.
Additional 32-bit registers contain the programmable alarm subseconds, seconds, minutes,
hours, day, and date.
A digital calibration feature is available to compensate for any deviation in crystal oscillator
accuracy.
After backup domain reset, all RTC registers are protected against possible parasitic write
accesses.
As long as the supply voltage remains in the operating range, the RTC never stops,
regardless of the device status (Run mode, low-power mode or under reset).

798/1747

RM0090 Rev 17

RM0090

26.2

Real-time clock (RTC)

RTC main features
The RTC unit main features are the following (see Figure 237: RTC block diagram):
•

Calendar with subseconds, seconds, minutes, hours (12 or 24 format), day (day of
week), date (day of month), month, and year.

•

Daylight saving compensation programmable by software.

•

Two programmable alarms with interrupt function. The alarms can be triggered by any
combination of the calendar fields.

•

Automatic wakeup unit generating a periodic flag that triggers an automatic wakeup
interrupt.

•

Reference clock detection: a more precise second source clock (50 or 60 Hz) can be
used to enhance the calendar precision.

•

Accurate synchronization with an external clock using the subsecond shift feature.

•

Maskable interrupts/events:

•

–

Alarm A

–

Alarm B

–

Wakeup interrupt

–

Timestamp

–

Tamper detection

Digital calibration circuit (periodic counter correction)
–

5 ppm accuracy

–

0.95 ppm accuracy, obtained in a calibration window of several seconds

•

Timestamp function for event saving (1 event)

•

Tamper detection:
–

2 tamper events with configurable filter and internal pull-up.

•

20 backup registers (80 bytes). The backup registers are reset when a tamper
detection event occurs.

•

Alternate function output (RTC_OUT) which selects one of the following two outputs:
–

RTC_CALIB: 512 Hz or 1 Hz clock output (with an LSE frequency of 32.768 kHz).
This output is enabled by setting the COE bit in the RTC_CR register. It is routed
to the device RTC_AF1 function.

–

RTC_ALARM (Alarm A, Alarm B or wakeup).
This output is selected by configuring the OSEL[1:0] bits in the RTC_CR register.
It is routed to the device RTC_AF1 function.

•

RTC alternate function inputs:
–

RTC_TS: timestamp event detection. It is routed to the device RTC_AF1 and
RTC_AF2 functions.

–

RTC_TAMP1: TAMPER1 event detection. It is routed to the device RTC_AF1 and
RTC_AF2 functions.

–

RTC_TAMP2: TAMPER2 event detection.

–

RTC_REFIN: reference clock input (usually the mains, 50 or 60 Hz).

Refer to Section 8.3.15: Selection of RTC_AF1 and RTC_AF2 alternate functions.

RM0090 Rev 17

799/1747
838

Real-time clock (RTC)

RM0090
Figure 237. RTC block diagram

57&B7$03

%DFNXSUHJLVWHUV
DQG57&WDPSHU
FRQWUROUHJLVWHU

57&B7$03
57&B7$03

7$03[)

7LPHVWDPS
UHJLVWHUV

57&B76
57&B5(),1

76)

/6( +]
+6(
57&&/.

/6,

57&B&$/5
6PRRWK
FDOLEUDWLRQ

57&B35(5
$V\QFKURQRXV
ELWSUHVFDOHU
GHIDXOW 

FNBDSUH
GHIDXOW+]

FNBVSUH
GHIDXOW+]

57&B35(5
6\QFKURQRXV
ELWSUHVFDOHU
GHIDXOW 

&DOHQGDU

6KDGRZUHJLVWHU
57&B665
:8&.6(/>@

6KDGRZUHJLVWHUV
57&B75
57&B'5

3UHVFDOHU

57&B&$/,%
57&B$/$50

2XWSXW
FRQWURO

57&B287

57&B:875
:87)

ELWZDNHXS
DXWRUHORDGWLPHU
26(/>@
$ODUP$
57&B$/50$5
57&B$/50$665

$/5$)

$ODUP%
57&B$/50%5
57&B$/50%665

$/5%)

069

1. On STM32F4xx devices, the RTC_AF1 and RTC_AF2 alternate functions are connected to PC13 and PI8,
respectively.

800/1747

RM0090 Rev 17

RM0090

Real-time clock (RTC)

26.3

RTC functional description

26.3.1

Clock and prescalers
The RTC clock source (RTCCLK) is selected through the clock controller among the LSE
clock, the LSI oscillator clock, and the HSE clock. For more information on the RTC clock
source configuration, refer to Section 7: Reset and clock control for STM32F405xx/07xx
and STM32F415xx/17xx(RCC) and Section 6: Reset and clock control for STM32F42xxx
and STM32F43xxx (RCC).
A programmable prescaler stage generates a 1 Hz clock which is used to update the
calendar. To minimize power consumption, the prescaler is split into 2 programmable
prescalers (see Figure 237: RTC block diagram):

Note:

•

A 7-bit asynchronous prescaler configured through the PREDIV_A bits of the
RTC_PRER register.

•

A 15-bit synchronous prescaler configured through the PREDIV_S bits of the
RTC_PRER register.

When both prescalers are used, it is recommended to configure the asynchronous prescaler
to a high value to minimize consumption.
The asynchronous prescaler division factor is set to 128, and the synchronous division
factor to 256, to obtain an internal clock frequency of 1 Hz (ck_spre) with an LSE frequency
of 32.768 kHz.
The minimum division factor is 1 and the maximum division factor is 222.
This corresponds to a maximum input frequency of around 4 MHz.
fck_apre is given by the following formula:
f RTCCLK
f CK_APRE = --------------------------------------PREDIV_A + 1

The ck_apre clock is used to clock the binary RTC_SSR subseconds downcounter. When it
reaches 0, RTC_SSR is reloaded with the content of PREDIV_S.
fck_spre is given by the following formula:
f RTCCLK
f CK_SPRE = ----------------------------------------------------------------------------------------------( PREDIV_S + 1 ) × ( PREDIV_A + 1 )

The ck_spre clock can be used either to update the calendar or as timebase for the 16-bit
wakeup auto-reload timer. To obtain short timeout periods, the 16-bit wakeup auto-reload
timer can also run with the RTCCLK divided by the programmable 4-bit asynchronous
prescaler (see Section 26.3.4: Periodic auto-wakeup for details).

RM0090 Rev 17

801/1747
838

Real-time clock (RTC)

26.3.2

RM0090

Real-time clock and calendar
The RTC calendar time and date registers are accessed through shadow registers which
are synchronized with PCLK1 (APB1 clock). They can also be accessed directly in order to
avoid waiting for the synchronization duration.
•

RTC_SSR for the subseconds

•

RTC_TR for the time

•

RTC_DR for the date

Every two RTCCLK periods, the current calendar value is copied into the shadow registers,
and the RSF bit of RTC_ISR register is set (see Section 26.6.4). The copy is not performed
in Stop and Standby mode. When exiting these modes, the shadow registers are updated
after up to 2 RTCCLK periods.
When the application reads the calendar registers, it accesses the content of the shadow
registers.t is possible to make a direct access to the calendar registers by setting the
BYPSHAD control bit in the RTC_CR register. By default, this bit is cleared, and the user
accesses the shadow registers.
When reading the RTC_SSR, RTC_TR or RTC_DR registers in BYPSHAD=0 mode, the
frequency of the APB clock (fAPB) must be at least 7 times the frequency of the RTC clock
(fRTCCLK).
The shadow registers are reset by system reset.

26.3.3

Programmable alarms
The RTC unit provides two programmable alarms, Alarm A and Alarm B.
The programmable alarm functions are enabled through the ALRAIE and ALRBIE bits in the
RTC_CR register. The ALRAF and ALRBF flags are set to 1 if the calendar subseconds,
seconds, minutes, hours, date or day match the values programmed in the alarm registers
RTC_ALRMASSR/RTC_ALRMAR and RTC_ALRMBSSR/RTC_ALRMBR, respectively.
Each calendar field can be independently selected through the MSKx bits of the
RTC_ALRMAR and RTC_ALRMBR registers, and through the MASKSSx bits of the
RTC_ALRMASSR and RTC_ALRMBSSR registers. The alarm interrupts are enabled
through the ALRAIE and ALRBIE bits in the RTC_CR register.
Alarm A and Alarm B (if enabled by bits OSEL[1:0] in RTC_CR register) can be routed to the
RTC_ALARM output. RTC_ALARM polarity can be configured through bit POL in the
RTC_CR register.

Caution:

If the seconds field is selected (MSK0 bit reset in RTC_ALRMAR or RTC_ALRMBR), the
synchronous prescaler division factor set in the RTC_PRER register must be at least 3 to
ensure correct behavior.

26.3.4

Periodic auto-wakeup
The periodic wakeup flag is generated by a 16-bit programmable auto-reload down-counter.
The wakeup timer range can be extended to 17 bits.
The wakeup function is enabled through the WUTE bit in the RTC_CR register.

802/1747

RM0090 Rev 17

RM0090

Real-time clock (RTC)
The wakeup timer clock input can be:
•

RTC clock (RTCCLK) divided by 2, 4, 8, or 16.
When RTCCLK is LSE(32.768kHz), this allows configuring the wakeup interrupt period
from 122 µs to 32 s, with a resolution down to 61µs.

•

ck_spre (usually 1 Hz internal clock)
When ck_spre frequency is 1Hz, this allows achieving a wakeup time from 1 s to
around 36 hours with one-second resolution. This large programmable time range is
divided in 2 parts:
–

from 1s to 18 hours when WUCKSEL [2:1] = 10

–

and from around 18h to 36h when WUCKSEL[2:1] = 11. In this last case 216 is
added to the 16-bit counter current value.When the initialization sequence is
complete (see Programming the wakeup timer on page 805), the timer starts
counting down.When the wakeup function is enabled, the down-counting remains
active in low-power modes. In addition, when it reaches 0, the WUTF flag is set in
the RTC_ISR register, and the wakeup counter is automatically reloaded with its
reload value (RTC_WUTR register value).

The WUTF flag must then be cleared by software.
When the periodic wakeup interrupt is enabled by setting the WUTIE bit in the RTC_CR2
register, it can exit the device from low-power modes.
The periodic wakeup flag can be routed to the RTC_ALARM output provided it has been
enabled through bits OSEL[1:0] of RTC_CR register. RTC_ALARM polarity can be
configured through the POL bit in the RTC_CR register.
System reset, as well as low-power modes (Sleep, Stop and Standby) have no influence on
the wakeup timer.

26.3.5

RTC initialization and configuration
RTC register access
The RTC registers are 32-bit registers. The APB interface introduces 2 wait-states in RTC
register accesses except on read accesses to calendar shadow registers when
BYPSHAD=0.

RTC register write protection
After system reset, the RTC registers are protected against parasitic write access with the
DBP bit of the PWR power control register (PWR_CR). The DBP bit must be set to enable
RTC registers write access.
After backup domain reset, all the RTC registers are write-protected. Writing to the RTC
registers is enabled by writing a key into the Write Protection register, RTC_WPR.
The following steps are required to unlock the write protection on all the RTC registers
except for RTC_ISR[13:8], RTC_TAFCR, and RTC_BKPxR.
1.

Write ‘0xCA’ into the RTC_WPR register.

2.

Write ‘0x53’ into the RTC_WPR register.

Writing a wrong key reactivates the write protection.
The protection mechanism is not affected by system reset.

RM0090 Rev 17

803/1747
838

Real-time clock (RTC)

RM0090

Calendar initialization and configuration
To program the initial time and date calendar values, including the time format and the
prescaler configuration, the following sequence is required:
1.

Set INIT bit to 1 in the RTC_ISR register to enter initialization mode. In this mode, the
calendar counter is stopped and its value can be updated.

2.

Poll INITF bit of in the RTC_ISR register. The initialization phase mode is entered when
INITF is set to 1. It takes from 1 to 2 RTCCLK clock cycles (due to clock
synchronization).

3.

To generate a 1 Hz clock for the calendar counter, program first the synchronous
prescaler factor in RTC_PRER register, and then program the asynchronous prescaler
factor. Even if only one of the two fields needs to be changed, 2 separate write
accesses must be performed to the RTC_PRER register.

4.

Load the initial time and date values in the shadow registers (RTC_TR and RTC_DR),
and configure the time format (12 or 24 hours) through the FMT bit in the RTC_CR
register.

5.

Exit the initialization mode by clearing the INIT bit. The actual calendar counter value is
then automatically loaded and the counting restarts after 4 RTCCLK clock cycles.

When the initialization sequence is complete, the calendar starts counting.
Note:

After a system reset, the application can read the INITS flag in the RTC_ISR register to
check if the calendar has been initialized or not. If this flag equals 0, the calendar has not
been initialized since the year field is set at its backup domain reset default value (0x00).
To read the calendar after initialization, the software must first check that the RSF flag is set
in the RTC_ISR register.

Daylight saving time
The daylight saving time management is performed through bits SUB1H, ADD1H, and BKP
of the RTC_CR register.
Using SUB1H or ADD1H, the software can subtract or add one hour to the calendar in one
single operation without going through the initialization procedure.
In addition, the software can use the BKP bit to memorize this operation.

Programming the alarm
A similar procedure must be followed to program or update the programmable alarm (Alarm
A or Alarm B):

Note:

804/1747

1.

Clear ALRAE or ALRBIE in RTC_CR to disable Alarm A or Alarm B.

2.

Poll ALRAWF or ALRBWF in RTC_ISR until it is set to make sure the access to alarm
registers is allowed. This takes 1 to 2 RTCCLK clock cycles (due to clock
synchronization).

3.

Program the Alarm A or Alarm B registers (RTC_ALRMASSR/RTC_ALRMAR or
RTC_ALRMBSSR/RTC_ALRMBR).

4.

Set ALRAE or ALRBIE in the RTC_CR register to enable Alarm A or Alarm B again.

Each change of the RTC_CR register is taken into account after 1 to 2 RTCCLK clock cycles
due to clock synchronization.

RM0090 Rev 17

RM0090

Real-time clock (RTC)

Programming the wakeup timer
The following sequence is required to configure or change the wakeup timer auto-reload
value (WUT[15:0] in RTC_WUTR):

26.3.6

1.

Clear WUTE in RTC_CR to disable the wakeup timer.

2.

Poll WUTWF until it is set in RTC_ISR to make sure the access to wakeup auto-reload
counter and to WUCKSEL[2:0] bits is allowed. It takes 1 to 2 RTCCLK clock cycles
(due to clock synchronization).

3.

Program the wakeup auto-reload value WUT[15:0], and the wakeup clock selection
(WUCKSEL[2:0] bits in RTC_CR).Set WUTE in RTC_CR to enable the timer again.
The wakeup timer restarts down-counting. The WUTWF bit is cleared up to 2 RTCCLK
clocks cycles after WUTE is cleared, due to clock synchronization.

Reading the calendar
When BYPSHAD control bit is cleared in the RTC_CR register
To read the RTC calendar registers (RTC_SSR, RTC_TR and RTC_DR) properly, the APB1
clock frequency (fPCLK1) must be equal to or greater than seven times the fRTCCLK RTC
clock frequency. This ensures a secure behavior of the synchronization mechanism.
If the APB1 clock frequency is less than seven times the RTC clock frequency, the software
must read the calendar time and date registers twice. If the second read of the RTC_TR
gives the same result as the first read, this ensures that the data is correct. Otherwise a third
read access must be done. In any case the APB1 clock frequency must never be lower than
the RTC clock frequency.
The RSF bit is set in RTC_ISR register each time the calendar registers are copied into the
RTC_SSR, RTC_TR and RTC_DR shadow registers. The copy is performed every two
RTCCLK cycles. To ensure consistency between the 3 values, reading either RTC_SSR or
RTC_TR locks the values in the higher-order calendar shadow registers until RTC_DR is
read. In case the software makes read accesses to the calendar in a time interval smaller
than 2 RTCCLK periods: RSF must be cleared by software after the first calendar read, and
then the software must wait until RSF is set before reading again the RTC_SSR, RTC_TR
and RTC_DR registers.
After waking up from low-power mode (Stop or Standby), RSF must be cleared by software.
The software must then wait until it is set again before reading the RTC_SSR, RTC_TR and
RTC_DR registers.
The RSF bit must be cleared after wakeup and not before entering low-power mode.

Note:

After a system reset, the software must wait until RSF is set before reading the RTC_SSR,
RTC_TR and RTC_DR registers. Indeed, a system reset resets the shadow registers to
their default values.
After an initialization (refer to Calendar initialization and configuration on page 804): the
software must wait until RSF is set before reading the RTC_SSR, RTC_TR and RTC_DR
registers.
After synchronization (refer to Section 26.3.8: RTC synchronization): the software must wait
until RSF is set before reading the RTC_SSR, RTC_TR and RTC_DR registers.

RM0090 Rev 17

805/1747
838

Real-time clock (RTC)

RM0090

When the BYPSHAD control bit is set in the RTC_CR register (bypass shadow
registers)
Reading the calendar registers gives the values from the calendar counters directly, thus
eliminating the need to wait for the RSF bit to be set. This is especially useful after exiting
from low-power modes (STOP or Standby), since the shadow registers are not updated
during these modes.
When the BYPSHAD bit is set to 1, the results of the different registers might not be
coherent with each other if an RTCCLK edge occurs between two read accesses to the
registers. Additionally, the value of one of the registers may be incorrect if an RTCCLK edge
occurs during the read operation. The software must read all the registers twice, and then
compare the results to confirm that the data is coherent and correct. Alternatively, the
software can just compare the two results of the least-significant calendar register.
Note:

While BYPSHAD=1, instructions which read the calendar registers require one extra APB
cycle to complete.

26.3.7

Resetting the RTC
The calendar shadow registers (RTC_SSR, RTC_TR and RTC_DR) and some bits of the
RTC status register (RTC_ISR) are reset to their default values by all available system reset
sources.
On the contrary, the following registers are resetted to their default values by a backup
domain reset and are not affected by a system reset: the RTC current calendar registers,
the RTC control register (RTC_CR), the prescaler register (RTC_PRER), the RTC
calibration registers (RTC_CALIBR or RTC_CALR), the RTC shift register (RTC_SHIFTR),
the RTC timestamp registers (RTC_TSSSR, RTC_TSTR and RTC_TSDR), the RTC tamper
and alternate function configuration register (RTC_TAFCR), the RTC backup registers
(RTC_BKPxR), the wakeup timer register (RTC_WUTR), the Alarm A and Alarm B registers
(RTC_ALRMASSR/RTC_ALRMAR and RTC_ALRMBSSR/RTC_ALRMBR).
In addition, when the RTC is clocked by the LSE, it goes on running under system reset if
the reset source is different from a backup domain reset one. Refer to section RTC clock of
the Reset and clock controller for details about the list of the RTC clock sources that are not
affected by system reset.
When a backup domain reset occurs, the RTC is stopped and all the RTC registers are set
to their reset values.

26.3.8

RTC synchronization
The RTC can be synchronized to a remote clock with a high degree of precision. After
reading the sub-second field (RTC_SSR or RTC_TSSSR), a calculation can be made of the
precise offset between the times being maintained by the remote clock and the RTC. The
RTC can then be adjusted to eliminate this offset by “shifting” its clock by a fraction of a
second using RTC_SHIFTR.
RTC_SSR contains the value of the synchronous prescaler’s counter. This allows one to
calculate the exact time being maintained by the RTC down to a resolution of
1 / (PREDIV_S + 1) seconds. As a consequence, the resolution can be improved by
increasing the synchronous prescaler value (PREDIV_S[14:0]. The maximum resolution
allowed (30.52 μs with a 32768 Hz clock) is obtained with PREDIV_S set to 0x7FFF.

806/1747

RM0090 Rev 17

RM0090

Real-time clock (RTC)
However, increasing PREDIV_S means that PREDIV_A must be decreased in order to
maintain the synchronous prescaler’s output at 1 Hz. In this way, the frequency of the
asynchronous prescaler’s output increases, which may increase the RTC dynamic
consumption.
The RTC can be finely adjusted using the RTC shift control register (RTC_SHIFTR). Writing
to RTC_SHIFTR can shift (either delay or advance) the clock by up to a second with a
resolution of 1 / (PREDIV_S + 1) seconds. The shift operation consists of adding the
SUBFS[14:0] value to the synchronous prescaler counter SS[15:0]: this will delay the clock.
If at the same time the ADD1S bit is set, this results in adding one second and at the same
time subtracting a fraction of second, so this will advance the clock.

Caution:

Before initiating a shift operation, the user must check that SS[15] = 0 in order to ensure that
no overflow will occur.
As soon as a shift operation is initiated by a write to the RTC_SHIFTR register, the SHPF
flag is set by hardware to indicate that a shift operation is pending. This bit is cleared by
hardware as soon as the shift operation has completed.

Caution:

This synchronization feature is not compatible with the reference clock detection feature:
firmware must not write to RTC_SHIFTR when REFCKON=1.

26.3.9

RTC reference clock detection
The RTC calendar update can be synchronized to a reference clock RTC_REFIN, usually
the mains (50 or 60 Hz). The RTC_REFIN reference clock should have a higher precision
than the 32.768 kHz LSE clock. When the RTC_REFIN detection is enabled (REFCKON bit
of RTC_CR set to 1), the calendar is still clocked by the LSE, and RTC_REFIN is used to
compensate for the imprecision of the calendar update frequency (1 Hz).
Each 1 Hz clock edge is compared to the nearest reference clock edge (if one is found
within a given time window). In most cases, the two clock edges are properly aligned. When
the 1 Hz clock becomes misaligned due to the imprecision of the LSE clock, the RTC shifts
the 1 Hz clock a bit so that future 1 Hz clock edges are aligned. Thanks to this mechanism,
the calendar becomes as precise as the reference clock.
The RTC detects if the reference clock source is present by using the 256 Hz clock
(ck_apre) generated from the 32.768 kHz quartz. The detection is performed during a time
window around each of the calendar updates (every 1 s). The window equals 7 ck_apre
periods when detecting the first reference clock edge. A smaller window of 3 ck_apre
periods is used for subsequent calendar updates.
Each time the reference clock is detected in the window, the synchronous prescaler which
outputs the ck_spre clock is forced to reload. This has no effect when the reference clock
and the 1 Hz clock are aligned because the prescaler is being reloaded at the same
moment. When the clocks are not aligned, the reload shifts future 1 Hz clock edges a little
for them to be aligned with the reference clock.
If the reference clock halts (no reference clock edge occurred during the 3 ck_apre window),
the calendar is updated continuously based solely on the LSE clock. The RTC then waits for
the reference clock using a large 7 ck_apre period detection window centered on the
ck_spre edge.

RM0090 Rev 17

807/1747
838

Real-time clock (RTC)

RM0090

When the reference clock detection is enabled, PREDIV_A and PREDIV_S must be set to
their default values:
•

PREDIV_A = 0x007F

•

PREDIV_S = 0x00FF

Note:

The reference clock detection is not available in Standby mode.

Caution:

The reference clock detection feature cannot be used in conjunction with the coarse digital
calibration: RTC_CALIBR must be kept at 0x0000 0000 when REFCKON=1.

26.3.10

RTC coarse digital calibration
Two digital calibration methods are available: coarse and smooth calibration. To perform
coarse calibration refer to Section 26.6.7: RTC calibration register (RTC_CALIBR).
The two calibration methods are not intended to be used together, the application must
select one of the two methods. Coarse calibration is provided for compatibly reasons. To
perform smooth calibration refer to Section 26.3.11: RTC smooth digital calibration and the
Section 26.6.16: RTC calibration register (RTC_CALR)
The coarse digital calibration can be used to compensate crystal inaccuracy by adding
(positive calibration) or masking (negative calibration) clock cycles at the output of the
asynchronous prescaler (ck_apre).
Positive and negative calibration are selected by setting the DCS bit in RTC_CALIBR
register to ‘0’ and ‘1’, respectively.
When positive calibration is enabled (DCS = ‘0’), 2 ck_apre cycles are added every minute
(around 15360 ck_apre cycles) for 2xDC minutes. This causes the calendar to be updated
sooner, thereby adjusting the effective RTC frequency to be a bit higher.
When negative calibration is enabled (DCS = ‘1’), 1 ck_apre cycle is removed every minute
(around 15360 ck_apre cycles) for 2xDC minutes. This causes the calendar to be updated
later, thereby adjusting the effective RTC frequency to be a bit lower.
DC is configured through bits DC[4:0] of RTC_CALIBR register. This number ranges from 0
to 31 corresponding to a time interval (2xDC) ranging from 0 to 62.
The coarse digital calibration can be configured only in initialization mode, and starts when
the INIT bit is cleared. The full calibration cycle lasts 64 minutes. The first 2xDC minutes of
the 64 -minute cycle are modified as just described.
Negative calibration can be performed with a resolution of about 2 ppm while positive
calibration can be performed with a resolution of about 4 ppm. The maximum calibration
ranges from −63 ppm to 126 ppm.
The calibration can be performed either on the LSE or on the HSE clock.

Caution:

Digital calibration may not work correctly if PREDIV_A < 6.

Case of RTCCLK=32.768 kHz and PREDIV_A+1=128
The following description assumes that ck_apre frequency is 256 Hz obtained with an LSE
clock nominal frequency of 32.768 kHz, and PREDIV_A set to 127 (default value).
The ck_spre clock frequency is only modified during the first 2xDC minutes of the 64-minute
cycle. For example, when DC equals 1, only the first 2 minutes are modified. This means
that the first 2xDC minutes of each 64-minute cycle have, once per minute, one second

808/1747

RM0090 Rev 17

RM0090

Real-time clock (RTC)
either shortened by 256 or lengthened by 128 RTCCLK cycles, given that each ck_apre
cycle represents 128 RTCCLK cycles (with PREDIV_A+1=128).
Therefore each calibration step has the effect of adding 512 or subtracting 256 oscillator
cycles for every 125829120 RTCCLK cycles (64min x 60 s/min x 32768 cycles/s). This is
equivalent to +4.069 ppm or-2.035 ppm per calibration step. As a result, the calibration
resolution is +10.5 or −5.27 seconds per month, and the total calibration ranges from +5.45
to −2.72 minutes per month.
In order to measure the clock deviation, a 512 Hz clock is output for calibration.Refer to
Section 26.3.14: Calibration clock output.

26.3.11

RTC smooth digital calibration
RTC frequency can be digitally calibrated with a resolution of about 0.954 ppm with a range
from -487.1 ppm to +488.5 ppm. The correction of the frequency is performed using series
of small adjustments (adding and/or subtracting individual RTCCLK pulses). These
adjustments are fairly well distributed so that the RTC is well calibrated even when observed
over short durations of time.
The smooth digital calibration is performed during a cycle of about 220 RTCCLK pulses, or
32 seconds when the input frequency is 32768 Hz. This cycle is maintained by a 20-bit
counter, cal_cnt[19:0], clocked by RTCCLK.
The smooth calibration register (RTC_CALR) specifies the number of RTCCLK clock cycles
to be masked during the 32-second cycle:

Note:

•

Setting the bit CALM[0] to 1 causes exactly one pulse to be masked during the 32second cycle.

•

Setting CALM[1] to 1 causes two additional cycles to be masked

•

Setting CALM[2] to 1 causes four additional cycles to be masked

•

and so on up to CALM[8] set to 1 which causes 256 clocks to be masked.

CALM[8:0] (RTC_CALRx) specifies the number of RTCCLK pulses to be masked during the
32-second cycle. Setting the bit CALM[0] to ‘1’ causes exactly one pulse to be masked
during the 32-second cycle at the moment when cal_cnt[19:0] is 0x80000; CALM[1]=1
causes two other cycles to be masked (when cal_cnt is 0x40000 and 0xC0000); CALM[2]=1
causes four other cycles to be masked (cal_cnt = 0x20000/0x60000/0xA0000/ 0xE0000);
and so on up to CALM[8]=1 which causes 256 clocks to be masked (cal_cnt = 0xXX800).
While CALM allows the RTC frequency to be reduced by up to 487.1 ppm with fine
resolution, the bit CALP can be used to increase the frequency by 488.5 ppm. Setting CALP
to ‘1’ effectively inserts an extra RTCCLK pulse every 211 RTCCLK cycles, which means
that 512 clocks are added during every 32-second cycle.
Using CALM together with CALP, an offset ranging from -511 to +512 RTCCLK cycles can
be added during the 32-second cycle, which translates to a calibration range of -487.1 ppm
to +488.5 ppm with a resolution of about 0.954 ppm.
The formula to calculate the effective calibrated frequency (FCAL) given the input frequency
(FRTCCLK) is as follows:
FCAL = FRTCCLK x [1 + (CALP x 512 - CALM) / (220 + CALM - CALP x 512)]

Calibration when PREDIV_A<3
The CALP bit can not be set to 1 when the asynchronous prescaler value (PREDIV_A bits in
RTC_PRER register) is less than 3. If CALP was already set to 1 and PREDIV_A bits are
RM0090 Rev 17

809/1747
838

Real-time clock (RTC)

RM0090

set to a value less than 3, CALP is ignored and the calibration operates as if CALP was
equal to 0.
To perform a calibration with PREDIV_A less than 3, the synchronous prescaler value
(PREDIV_S) should be reduced so that each second is accelerated by 8 RTCCLK clock
cycles, which is equivalent to adding 256 clock cycles every 32 seconds. As a result,
between 255 and 256 clock pulses (corresponding to a calibration range from 243.3 to
244.1 ppm) can effectively be added during each 32-second cycle using only the CALM bits.
With a nominal RTCCLK frequency of 32768 Hz, when PREDIV_A equals 1 (division factor
of 2), PREDIV_S should be set to 16379 rather than 16383 (4 less). The only other
interesting case is when PREDIV_A equals 0, PREDIV_S should be set to 32759 rather
than 32767 (8 less).
If PREDIV_S is reduced in this way, the formula given the effective frequency of the
calibrated input clock is as follows:
FCAL = FRTCCLK x [1 + (256 - CALM) / (220 + CALM - 256)]
In this case, CALM[7:0] equals 0x100 (the midpoint of the CALM range) is the correct
setting if RTCCLK is exactly 32768.00 Hz.

Verifying the RTC calibration
RTC precision is performed by measuring the precise frequency of RTCCLK and calculating
the correct CALM value and CALP values. An optional 1 Hz output is provided to allow
applications to measure and verify the RTC precision.
Measuring the precise frequency of the RTC over a limited interval can result in a
measurement error of up to 2 RTCCLK clock cycles over the measurement period,
depending on how the digital calibration cycle is aligned with the measurement period.
However, this measurement error can be eliminated if the measurement period is the same
length as the calibration cycle period. In this case, the only error observed is the error due to
the resolution of the digital calibration.
•

By default, the calibration cycle period is 32 seconds.
Using this mode and measuring the accuracy of the 1 Hz output over exactly 32
seconds guarantees that the measure is within 0.477 ppm (0.5 RTCCLK cycles over 32
seconds, due to the limitation of the calibration resolution).

•

CALW16 bit of the RTC_CALR register can be set to 1 to force a 16- second calibration
cycle period.
In this case, the RTC precision can be measured during 16 seconds with a maximum
error of 0.954 ppm (0.5 RTCCLK cycles over 16 seconds). However, since the
calibration resolution is reduced, the long term RTC precision is also reduced to 0.954
ppm: CALM[0] bit is stuck at 0 when CALW16 is set to 1.

•

CALW8 bit of the RTC_CALR register can be set to 1 to force a 8- second calibration
cycle period.
In this case, the RTC precision can be measured during 8 seconds with a maximum
error of 1.907 ppm (0.5 RTCCLK cycles over 8s). The long term RTC precision is also
reduced to 1.907 ppm: CALM[1:0] bits are stuck at 00 when CALW8 is set to 1.

Re-calibration on-the-fly
The calibration register (RTC_CALR) can be updated on-the-fly while RTC_ISR/INITF=0, by
using the follow process:
810/1747

RM0090 Rev 17

RM0090

26.3.12

Real-time clock (RTC)
1.

Poll the RTC_ISR/RECALPF (re-calibration pending flag).

2.

If it is set to 0, write a new value to RTC_CALR, if necessary. RECALPF is then
automatically set to 1

3.

Within three ck_apre cycles after the write operation to RTC_CALR, the new calibration
settings take effect.

Timestamp function
Timestamp is enabled by setting the TSE bit of RTC_CR register to 1.
The calendar is saved in the timestamp registers (RTC_TSSSR, RTC_TSTR, RTC_TSDR)
when a timestamp event is detected on the pin to which the TIMESTAMP alternate function
is mapped. When a timestamp event occurs, the timestamp flag bit (TSF) in RTC_ISR
register is set.
By setting the TSIE bit in the RTC_CR register, an interrupt is generated when a timestamp
event occurs.
If a new timestamp event is detected while the timestamp flag (TSF) is already set, the
timestamp overflow flag (TSOVF) flag is set and the timestamp registers (RTC_TSTR and
RTC_TSDR) maintain the results of the previous event.

Note:

TSF is set 2 ck_apre cycles after the timestamp event occurs due to synchronization
process.
There is no delay in the setting of TSOVF. This means that if two timestamp events are
close together, TSOVF can be seen as '1' while TSF is still '0'. As a consequence, it is
recommended to poll TSOVF only after TSF has been set.

Caution:

If a timestamp event occurs immediately after the TSF bit is supposed to be cleared, then
both TSF and TSOVF bits are set. To avoid masking a timestamp event occurring at the
same moment, the application must not write ‘0’ into TSF bit unless it has already read it to
‘1’.
Optionally, a tamper event can cause a timestamp to be recorded. See the description of the
TAMPTS control bit in Section 26.6.17: RTC tamper and alternate function configuration
register (RTC_TAFCR). If the timestamp event is on the same pin as a tamper event
configured in filtered mode (TAMPFLT set to a non-zero value), the timestamp on tamper
detection event mode must be selected by setting TAMPTS='1' in RTC_TAFCR register.

TIMESTAMP alternate function
The TIMESTAMP alternate function (RTC_TS) can be mapped either to RTC_AF1 or to
RTC_AF2 depending on the value of the TSINSEL bit in the RTC_TAFCR register (see
Section 26.6.17: RTC tamper and alternate function configuration register (RTC_TAFCR)).
Mapping the timestamp event on RTC_AF2 is not allowed if RTC_AF1 is used as TAMPER
in filtered mode (TAMPFLT set to a non-zero value).

26.3.13

Tamper detection
Two tamper detection inputs are available. They can be configured either for edge detection,
or for level detection with filtering.

RTC backup registers
The backup registers (RTC_BKPxR) are twenty 32-bit registers for storing 80 bytes of user
application data. They are implemented in the backup domain that remains powered-on by
RM0090 Rev 17

811/1747
838

Real-time clock (RTC)

RM0090

VBAT when the VDD power is switched off. They are not reset by system reset or when the
device wakes up from Standby mode. They are reset by a backup domain reset
The backup registers are reset when a tamper detection event occurs (see Section 26.6.20:
RTC backup registers (RTC_BKPxR) and Tamper detection initialization on page 812.

Tamper detection initialization
Each tamper detection input is associated with the TAMP1F/TAMP2F flags in the RTC_ISR2
register. Each input can be enabled by setting the corresponding TAMP1E/TAMP2E bits to 1
in the RTC_TAFCR register.
A tamper detection event resets all backup registers (RTC_BKPxR).
By setting the TAMPIE bit in the RTC_TAFCR register, an interrupt is generated when a
tamper detection event occurs.

Timestamp on tamper event
With TAMPTS set to ‘1 , any tamper event causes a timestamp to occur. In this case, either
the TSF bit or the TSOVF bit are set in RTC_ISR, in the same manner as if a normal
timestamp event occurs. The affected tamper flag register (TAMP1F, TAMP2F) is set at the
same time that TSF or TSOVF is set.

Edge detection on tamper inputs
If the TAMPFLT bits are “00”, the TAMPER pins generate tamper detection events
(RTC_TAMP[2:1]) when either a rising edge is observed or an falling edge is observed
depending on the corresponding TAMPxTRG bit. The internal pull-up resistors on the
TAMPER inputs are deactivated when edge detection is selected.
Caution:

To avoid losing tamper detection events, the signal used for edge detection is logically
ANDed with TAMPxE in order to detect a tamper detection event in case it occurs before the
TAMPERx pin is enabled.
•

When TAMPxTRG = 0: if the TAMPERx alternate function is already high before
tamper detection is enabled (TAMPxE bit set to 1), a tamper event is detected as soon
as TAMPERx is enabled, even if there was no rising edge on TAMPERx after TAMPxE
was set.

•

When TAMPxTRG = 1: if the TAMPERx alternate function is already low before tamper
detection is enabled, a tamper event is detected as soon as TAMPERx is enabled
(even if there was no falling edge on TAMPERx after TAMPxE was set.

After a tamper event has been detected and cleared, the TAMPERx alternate function
should be disabled and then re-enabled (TAMPxE set to 1) before re-programming the
backup registers (RTC_BKPxR). This prevents the application from writing to the backup
registers while the TAMPERx value still indicates a tamper detection. This is equivalent to a
level detection on the TAMPERx alternate function.
Note:

Tamper detection is still active when VDD power is switched off. To avoid unwanted resetting
of the backup registers, the pin to which the TAMPER alternate function is mapped should
be externally tied to the correct level.

Level detection with filtering on tamper inputs
Level detection with filtering is performed by setting TAMPFLT to a non-zero value. A tamper
detection event is generated when either 2, 4, or 8 (depending on TAMPFLT) consecutive

812/1747

RM0090 Rev 17

RM0090

Real-time clock (RTC)
samples are observed at the level designated by the TAMPxTRG bits
(TAMP1TRG/TAMP2TRG).
The TAMPER inputs are pre-charged through the I/O internal pull-up resistance before its
state is sampled, unless disabled by setting TAMPPUDIS to 1,The duration of the precharge
is determined by the TAMPPRCH bits, allowing for larger capacitances on the tamper
inputs.
The trade-off between tamper detection latency and power consumption through the pull-up
can be optimized by using TAMPFREQ to determine the frequency of the sampling for level
detection.

Note:

Refer to the datasheets for the electrical characteristics of the pull-up resistors.

TAMPER alternate function detection
The TAMPER1 alternate function (RTC_TAMP1) can be mapped either to RTC_AF1(PC13)
or RTC_AF2 (PI8) depending on the value of TAMP1INSEL bit in RTC_TAFCR register (see
Section 26.6.17: RTC tamper and alternate function configuration register (RTC_TAFCR)).
TAMPE bit must be cleared when TAMP1INSEL is modified to avoid unwanted setting of
TAMPF.
The TAMPER 2 alternate function corresponds to RTC_TAMP2 pin.

26.3.14

Calibration clock output
When the COE bit is set to 1 in the RTC_CR register, a reference clock is provided on the
RTC_CALIB device output. If the COSEL bit in the RTC_CR register is reset and
PREDIV_A = 0x7F, the RTC_CALIB frequency is fRTCCLK/64. This corresponds to a
calibration output at 512 Hz for an RTCCLK frequency at 32.768 kHz.
The RTC_CALIB output is not impacted by the calibration value programmed in
RTC_CALIBR register. The RTC_CALIB duty cycle is irregular: there is a light jitter on falling
edges. It is therefore recommended to use rising edges.
If COSEL is set and (PREDIV_S+1) is a non-zero multiple of 256 (i.e: PREDIV_S[7:0] =
0xFF), the RTC_CALIB frequency is fRTCCLK/(256 * (PREDIV_A+1)). This corresponds to a
calibration output at 1 Hz for prescaler default values (PREDIV_A = 0x7F, PREDIV_S =
0xFF), with an RTCCLK frequency at 32.768 kHz. The 1 Hz output is affected when a shift
operation is on going and may toggle during the shift operation (SHPF=1).

Calibration alternate function output
When the COE bit in the RTC_CR register is set to 1, the calibration alternate function
(RTC_CALIB) is enabled on RTC_AF1.
Note:

When RTC_CALIB or RTC_ALARM is selected, RTC_AF1 is automatically configured in
output alternate function.

26.3.15

Alarm output
Three functions can be selected on Alarm output: ALRAF, ALRBF and WUTF. These
functions reflect the contents of the corresponding flags in the RTC_ISR register.
The OSEL[1:0] control bits in the RTC_CR register are used to activate the alarm alternate
function output (RTC_ALARM) in RTC_AF1, and to select the function which is output on
RTC_ALARM.

RM0090 Rev 17

813/1747
838

Real-time clock (RTC)

RM0090

The polarity of the output is determined by the POL control bit in RTC_CR so that the
opposite of the selected flag bit is output when POL is set to 1.

Alarm alternate function output
RTC_ALARM can be configured in output open drain or output push-pull using the control
bit ALARMOUTTYPE in the RTC_TAFCR register.
Note:

Once RTC_ALARM is enabled, it has priority over RTC_CALIB (COE bit is don't care on
RTC_AF1).
When RTC_CALIB or RTC_ALARM is selected, RTC_AF1 is automatically configured in
output alternate function.

26.4

RTC and low-power modes
Table 119. Effect of low-power modes on RTC
Mode

Description

Sleep

No effect
RTC interrupts cause the device to exit the Sleep mode.

Stop

The RTC remains active when the RTC clock source is LSE or LSI. RTC alarm, RTC
tamper event, RTC time stamp event, and RTC Wakeup cause the device to exit the Stop
mode.

The RTC remains active when the RTC clock source is LSE or LSI. RTC alarm, RTC
Standby tamper event, RTC time stamp event, and RTC Wakeup cause the device to exit the
Standby mode.

26.5

RTC interrupts
All RTC interrupts are connected to the EXTI controller.
To enable the RTC Alarm interrupt, the following sequence is required:
1.

Configure and enable the EXTI Line 17 in interrupt mode and select the rising edge
sensitivity.

2.

Configure and enable the RTC_Alarm IRQ channel in the NVIC.

3.

Configure the RTC to generate RTC alarms (Alarm A or Alarm B).

To enable the RTC Wakeup interrupt, the following sequence is required:
1.

Configure and enable the EXTI Line 22 in interrupt mode and select the rising edge
sensitivity.

2.

Configure and enable the RTC_WKUP IRQ channel in the NVIC.

3.

Configure the RTC to generate the RTC wakeup timer event.

To enable the RTC Tamper interrupt, the following sequence is required:
1.

Configure and enable the EXTI Line 21 in interrupt mode and select the rising edge
sensitivity.

2.

Configure and Enable the TAMP_STAMP IRQ channel in the NVIC.

3.

Configure the RTC to detect the RTC tamper event.

To enable the RTC TimeStamp interrupt, the following sequence is required:
814/1747

RM0090 Rev 17

RM0090

Real-time clock (RTC)
1.

Configure and enable the EXTI Line 21 in interrupt mode and select the rising edge
sensitivity.

2.

Configure and Enable the TAMP_STAMP IRQ channel in the NVIC.

3.

Configure the RTC to detect the RTC timestamp event.
Table 120. Interrupt control bits
Event flag

Enable
control
bit

Exit the
Sleep
mode

Exit the
Stop
mode

Exit the
Standby
mode

Alarm A

ALRAF

ALRAIE

yes

yes(1)

yes(1)

Alarm B

ALRBF

ALRBIE

yes

yes(1)

yes(1)

Wakeup

WUTF

WUTIE

yes

yes(1)

yes(1)

TSF

TSIE

yes

yes(1)

yes(1)

Tamper1 detection

TAMP1F

TAMPIE

yes

yes(1)

yes(1)

Tamper2 detection(2)

TAMP2F

TAMPIE

yes

yes(1)

yes(1)

Interrupt event

TimeStamp

1. Wakeup from STOP and Standby modes is possible only when the RTC clock source is LSE or LSI.
2. If RTC_TAMPER2 pin is present. Refer to device datasheet pinout.

RM0090 Rev 17

815/1747
838

Real-time clock (RTC)

26.6

RM0090

RTC registers
Refer to Section 1.1: List of abbreviations for registers for registers for a list of abbreviations
used in register descriptions.
The peripheral registers have to be accessed by words (32 bits).

26.6.1

RTC time register (RTC_TR)
The RTC_TR is the calendar time shadow register. This register must be written in
initialization mode only. Refer to Calendar initialization and configuration on page 804 and
Reading the calendar on page 805.
Address offset: 0x00
Backup domain reset value: 0x0000 0000
System reset: 0x0000 0000 when BYPSHAD = 0. Not affected when BYPSHAD = 1.

31

30

29

28

27

26

25

24

23

Res.

14

13

12

11

MNT[2:0]
rw

rw

10

9

8

MNU[3:0]
rw

rw

21

PM

Reserved
15

22

rw

rw

rw

7
Res.

20

19

18

HT[1:0]

17

16

HU[3:0]

rw

rw

rw

rw

rw

rw

rw

6

5

4

3

2

1

0

rw

rw

ST[2:0]
rw

rw

SU[3:0]
rw

rw

rw

Bits 31-24 Reserved
Bit 23 Reserved, must be kept at reset value.
Bit 22 PM: AM/PM notation
0: AM or 24-hour format
1: PM
Bits 21:20 HT[1:0]: Hour tens in BCD format
Bits 19:16 HU[3:0]: Hour units in BCD format
Bit 15 Reserved, must be kept at reset value.
Bits 14:12 MNT[2:0]: Minute tens in BCD format
Bit 11:8 MNU[3:0]: Minute units in BCD format
Bit 7 Reserved, must be kept at reset value.
Bits 6:4 ST[2:0]: Second tens in BCD format
Bits 3:0 SU[3:0]: Second units in BCD format

Note:

816/1747

This register is write protected. The write access procedure is described in RTC register
write protection on page 803.

RM0090 Rev 17

RM0090

Real-time clock (RTC)

26.6.2

RTC date register (RTC_DR)
The RTC_DR is the calendar date shadow register. This register must be written in
initialization mode only. Refer to Calendar initialization and configuration on page 804 and
Reading the calendar on page 805.
Address offset: 0x04
Backup domain reset value: 0x0000 2101
System reset: 0x0000 2101 when BYPSHAD = 0. Not affected when BYPSHAD = 1.

31

30

29

28

27

26

25

24

14

13

WDU[2:0]
rw

rw

12

11

10

rw

rw

MT
rw

rw

22

21

20

19

18

YT[3:0]

Reserved
15

23

9

8

rw

rw

MU[3:0]

17

16

YU[3:0]

rw

rw

rw

rw

rw

rw

rw

rw

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

Reserved

DT[1:0]
rw

DU[3:0]

Bits 31-24 Reserved
Bits 23:20 YT[3:0]: Year tens in BCD format
Bits 19:16 YU[3:0]: Year units in BCD format
Bits 15:13 WDU[2:0]: Week day units
000: forbidden
001: Monday
...
111: Sunday
Bit 12 MT: Month tens in BCD format
Bits 11:8 MU: Month units in BCD format
Bits 7:6 Reserved, must be kept at reset value.
Bits 5:4 DT[1:0]: Date tens in BCD format
Bits 3:0 DU[3:0]: Date units in BCD format

Note:

This register is write protected. The write access procedure is described in RTC register
write protection on page 803.

RM0090 Rev 17

817/1747
838

Real-time clock (RTC)

26.6.3

RM0090

RTC control register (RTC_CR)
Address offset: 0x08
Backup domain reset value: 0x0000 0000
System reset: not affected

31

30

29

28

27

26

25

24

Reserved
15
TSIE

14

13

12

WUTIE ALRBIE ALRAIE

rw

rw

rw

rw

11

23
COE

10

TSE

WUTE

rw

rw

9

8

ALRBE ALRAE
rw

22

20

19

18

POL

COSEL

BKP

17

16

SUB1H ADD1H

rw

rw

rw

rw

rw

rw

w

w

7

6

5

4

3

2

1

0

DCE

FMT

rw

rw

rw

21

OSEL[1:0]

BYPS
REFCKON TSEDGE
HAD
rw

rw

rw

WUCKSEL[2:0]
rw

rw

rw

Bits 31:24 Reserved, must be kept at reset value.
Bit 23 COE: Calibration output enable
This bit enables the RTC_CALIB output
0: Calibration output disabled
1: Calibration output enabled
Bits 22:21 OSEL[1:0]: Output selection
These bits are used to select the flag to be routed to RTC_ALARM output
00: Output disabled
01: Alarm A output enabled
10:Alarm B output enabled
11: Wakeup output enabled
Bit 20 POL: Output polarity
This bit is used to configure the polarity of RTC_ALARM output
0: The pin is high when ALRAF/ALRBF/WUTF is asserted (depending on OSEL[1:0])
1: The pin is low when ALRAF/ALRBF/WUTF is asserted (depending on OSEL[1:0]).
Bit 19 COSEL: Calibration output selection
When COE=1, this bit selects which signal is output on RTC_CALIB.
0: Calibration output is 512 Hz
1: Calibration output is 1 Hz
These frequencies are valid for RTCCLK at 32.768 kHz and prescalers at their default
values (PREDIV_A=127 and PREDIV_S=255). Refer to Section 26.3.14: Calibration clock
output
Bit 18 BKP: Backup
This bit can be written by the user to memorize whether the daylight saving time change has
been performed or not.
Bit 17 SUB1H: Subtract 1 hour (winter time change)
When this bit is set outside initialization mode, 1 hour is subtracted to the calendar time if the
current hour is not 0. This bit is always read as 0.
Setting this bit has no effect when current hour is 0.
0: No effect
1: Subtracts 1 hour to the current time. This can be used for winter time change.

818/1747

RM0090 Rev 17

RM0090

Real-time clock (RTC)

Bit 16 ADD1H: Add 1 hour (summer time change)
When this bit is set, 1 hour is added to the calendar time. This bit is always read as 0.
0: No effect
1: Adds 1 hour to the current time. This can be used for summer time change outside
initialization mode.
Bit 15 TSIE: Timestamp interrupt enable
0: Timestamp Interrupt disable
1: Timestamp Interrupt enable
Bit 14 WUTIE: Wakeup timer interrupt enable
0: Wakeup timer interrupt disabled
1: Wakeup timer interrupt enabled
Bit 13 ALRBIE: Alarm B interrupt enable
0: Alarm B Interrupt disable
1: Alarm B Interrupt enable
Bit 12 ALRAIE: Alarm A interrupt enable
0: Alarm A interrupt disabled
1: Alarm A interrupt enabled
Bit 11 TSE: Time stamp enable
0: Time stamp disable
1: Time stamp enable
Bit 10 WUTE: Wakeup timer enable
0: Wakeup timer disabled
1: Wakeup timer enabled
Note: When the wakeup timer is disabled, wait for WUTWF=1 before enabling it again.
Bit 9 ALRBE: Alarm B enable
0: Alarm B disabled
1: Alarm B enabled
Bit 8 ALRAE: Alarm A enable
0: Alarm A disabled
1: Alarm A enabled
Bit 7 DCE: Coarse digital calibration enable
0: Digital calibration disabled
1: Digital calibration enabled
PREDIV_A must be 6 or greater
Bit 6 FMT: Hour format
0: 24 hour/day format
1: AM/PM hour format
Bit 5 BYPSHAD: Bypass the shadow registers
0: Calendar values (when reading from RTC_SSR, RTC_TR, and RTC_DR) are taken from
the shadow registers, which are updated once every two RTCCLK cycles.
1: Calendar values (when reading from RTC_SSR, RTC_TR, and RTC_DR) are taken
directly from the calendar counters.
Note: If the frequency of the APB1 clock is less than seven times the frequency of RTCCLK,
BYPSHAD must be set to ‘1’.

RM0090 Rev 17

819/1747
838

Real-time clock (RTC)

RM0090

Bit 4 REFCKON: Reference clock detection enable (50 or 60 Hz)
0: Reference clock detection disabled
1: Reference clock detection enabled
Note: PREDIV_S must be 0x00FF.
Bit 3 TSEDGE: Timestamp event active edge
0: TIMESTAMP rising edge generates a timestamp event
1: TIMESTAMP falling edge generates a timestamp event
TSE must be reset when TSEDGE is changed to avoid unwanted TSF setting
Bits 2:0 WUCKSEL[2:0]: Wakeup clock selection
000: RTC/16 clock is selected
001: RTC/8 clock is selected
010: RTC/4 clock is selected
011: RTC/2 clock is selected
10x: ck_spre (usually 1 Hz) clock is selected
11x: ck_spre (usually 1 Hz) clock is selected and 216 is added to the WUT counter value
(see note below)

Note:

WUT = Wakeup unit counter value. WUT = (0x0000 to 0xFFFF) + 0x10000 added when
WUCKSEL[2:1 = 11].
Bits 7, 6 and 4 of this register can be written in initialization mode only (RTC_ISR/INITF = 1).
Bits 2 to 0 of this register can be written only when RTC_CR WUTE bit = 0 and RTC_ISR
WUTWF bit = 1.
It is recommended not to change the hour during the calendar hour increment as it could
mask the incrementation of the calendar hour.
ADD1H and SUB1H changes are effective in the next second.
To avoid spuriously setting of TSF, TSE must be reset when TSEDGE is changed.
This register is write protected. The write access procedure is described in RTC register
write protection on page 803.

26.6.4

RTC initialization and status register (RTC_ISR)
Address offset: 0x0C
Backup domain reset value: 0x0000 0007
System reset value: Not affected except INIT, INITF and RSF which are cleared to 0.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16
RECAL
PF

Reserved
15

14

13

12

11

10

Res.

TAMP
2F

TAMP
1F

TSOVF

TSF

WUTF

ALRBF ALRAF

rc_w0

rc_w0

rc_w0

rc_w0

rc_w0

rc_w0

820/1747

9

8

rc_w0

r
7

6

5

4

3

2

1

0

INIT

INITF

RSF

INITS

SHPF

WUT
WF

ALRB
WF

ALRA
WF

rw

r

rc_w0

r

r

r

r

r

RM0090 Rev 17

RM0090

Real-time clock (RTC)

Bits 31:17 Reserved
Bit 16 RECALPF: Recalibration pending Flag
The RECALPF status flag is automatically set to ‘1’ when software writes to the RTC_CALR
register, indicating that the RTC_CALR register is blocked. When the new calibration
settings are taken into account, this bit returns to ‘0’. Refer to Section : Re-calibration on-thefly.
Bit 15 Reserved, must be kept at reset value.
Bit 14 TAMP2F: TAMPER2 detection flag
This flag is set by hardware when a tamper detection event is detected on tamper input 2.
It is cleared by software writing 0.
Bit 13 TAMP1F: Tamper detection flag
This flag is set by hardware when a tamper detection event is detected.
It is cleared by software writing 0.
Bit 12 TSOVF: Timestamp overflow flag
This flag is set by hardware when a timestamp event occurs while TSF is already set.
This flag is cleared by software by writing 0. It is recommended to check and then clear
TSOVF only after clearing the TSF bit. Otherwise, an overflow might not be noticed if a
timestamp event occurs immediately before the TSF bit is cleared.
Bit 11 TSF: Timestamp flag
This flag is set by hardware when a timestamp event occurs.
This flag is cleared by software by writing 0.
Bit 10 WUTF: Wakeup timer flag
This flag is set by hardware when the wakeup auto-reload counter reaches 0.
This flag is cleared by software by writing 0.
This flag must be cleared by software at least 1.5 RTCCLK periods before WUTF is set to 1
again.
Bit 9 ALRBF: Alarm B flag
This flag is set by hardware when the time/date registers (RTC_TR and RTC_DR) match the
Alarm B register (RTC_ALRMBR).
This flag is cleared by software by writing 0.
Bit 8 ALRAF: Alarm A flag
This flag is set by hardware when the time/date registers (RTC_TR and RTC_DR) match the
Alarm A register (RTC_ALRMAR).
This flag is cleared by software by writing 0.
Bit 7 INIT: Initialization mode
0: Free running mode
1: Initialization mode used to program time and date register (RTC_TR and RTC_DR), and
prescaler register (RTC_PRER). Counters are stopped and start counting from the new
value when INIT is reset.
Bit 6 INITF: Initialization flag
When this bit is set to 1, the RTC is in initialization state, and the time, date and prescaler
registers can be updated.
0: Calendar registers update is not allowed
1: Calendar registers update is allowed.

RM0090 Rev 17

821/1747
838

Real-time clock (RTC)

RM0090

Bit 5 RSF: Registers synchronization flag
This bit is set by hardware each time the calendar registers are copied into the shadow
registers (RTC_SSRx, RTC_TRx and RTC_DRx). This bit is cleared by hardware in
initialization mode, while a shift operation is pending (SHPF=1), or when in bypass shadow
register mode (BYPSHAD=1). This bit can also be cleared by software.
0: Calendar shadow registers not yet synchronized
1: Calendar shadow registers synchronized
Bit 4 INITS: Initialization status flag
This bit is set by hardware when the calendar year field is different from 0 (backup domain
reset value state).
0: Calendar has not been initialized
1: Calendar has been initialized
Bit 3 SHPF: Shift operation pending
0: No shift operation is pending
1: A shift operation is pending
This flag is set by hardware as soon as a shift operation is initiated by a write to the
RTC_SHIFTR. It is cleared by hardware when the corresponding shift operation has been
executed. Writing to SHPF has no effect.
Bit 2 WUTWF: Wakeup timer write flag
This bit is set by hardware up to 2 RTCCLK cycles after the WUTE bit has been set to 0 in
RTC_CR, and is cleared up to 2 RTCCLK cycles after the WUTE bit has been set to 1. The
wakeup timer values can be changed when WUTE bit is cleared and WUTWF is set.
0: Wakeup timer configuration update not allowed
1: Wakeup timer configuration update allowed
Bit 1 ALRBWF: Alarm B write flag
This bit is set by hardware when Alarm B values can be changed, after the ALRBIE bit has
been set to 0 in RTC_CR.
It is cleared by hardware in initialization mode.
0: Alarm B update not allowed
1: Alarm B update allowed.
Bit 0 ALRAWF: Alarm A write flag
This bit is set by hardware when Alarm A values can be changed, after the ALRAE bit has
been set to 0 in RTC_CR.
It is cleared by hardware in initialization mode.
0: Alarm A update not allowed
1: Alarm A update allowed

Note:

The ALRAF, ALRBF, WUTF and TSF bits are cleared 2 APB clock cycles after programming
them to 0.
This register is write protected (except for RTC_ISR[13:8] bits). The write access procedure
is described in RTC register write protection on page 803.

822/1747

RM0090 Rev 17

RM0090

Real-time clock (RTC)

26.6.5

RTC prescaler register (RTC_PRER)
Address offset: 0x10
Backup domain reset value: 0x007F 00FF
System reset: not affected

31

30

29

28

27

26

25

24

23

Res.

14

13

12

11

21

20

19

18

17

16

PREDIV_A[6:0]

Reserved
15

22

10

9

8

rw

rw

rw

rw

rw

rw

rw

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

7
PREDIV_S[14:0]

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:24 Reserved
Bit 23 Reserved, must be kept at reset value.
Bits 22:16 PREDIV_A[6:0]: Asynchronous prescaler factor
This is the asynchronous division factor:
ck_apre frequency = RTCCLK frequency/(PREDIV_A+1)
Bit 15 Reserved, must be kept at reset value.
Bits 14:0 PREDIV_S[14:0]: Synchronous prescaler factor
This is the synchronous division factor:
ck_spre frequency = ck_apre frequency/(PREDIV_S+1)

Note:

This register must be written in initialization mode only. The initialization must be performed
in two separate write accesses. Refer to Calendar initialization and configuration on
page 804
This register is write protected. The write access procedure is described in RTC register
write protection on page 803.

26.6.6

RTC wakeup timer register (RTC_WUTR)
Address offset: 0x14
Backup domain reset value: 0x0000 FFFF
System reset: not affected

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

Reserved
15

14

13

12

11

10

9

8

rw

rw

rw

rw

rw

rw

rw

rw

7

WUT[15:0]
rw

RM0090 Rev 17

823/1747
838

Real-time clock (RTC)

RM0090

Bits 31:16 Reserved
Bits 15:0 WUT[15:0]: Wakeup auto-reload value bits
When the wakeup timer is enabled (WUTE set to 1), the WUTF flag is set every (WUT[15:0]
+ 1) ck_wut cycles. The ck_wut period is selected through WUCKSEL[2:0] bits of the
RTC_CR register
When WUCKSEL[2] = 1, the wakeup timer becomes 17-bits and WUCKSEL[1] effectively
becomes WUT[16] the most-significant bit to be reloaded into the timer.
Note: The first assertion of WUTF occurs (WUT+1) ck_wut cycles after WUTE is set. Setting
WUT[15:0] to 0x0000 with WUCKSEL[2:0] =011 (RTCCLK/2) is forbidden.

Note:

This register can be written only when WUTWF is set to 1 in RTC_ISR.
This register is write protected. The write access procedure is described in RTC register
write protection on page 803.

26.6.7

RTC calibration register (RTC_CALIBR)
Address offset: 0x18
Backup domain reset value: 0x0000 0000
System reset: not affected

31

30

29

28

27

26

25

24

23

22

21

20

6

5

4

19

18

17

16

Reserved
15

14

13

12

11

Reserved

10

9

8

7
DCS
rw

Reserved

3

2

1

0

rw

rw

DC[4:0]
rw

rw

rw

Bits 31:8 Reserved
Bit 7 DCS: Digital calibration sign
0: Positive calibration: calendar update frequency is increased
1: Negative calibration: calendar update frequency is decreased
Bits 6:5 Reserved, must be kept at reset value.
Bits 4:0 DC[4:0]: Digital calibration
DCS = 0 (positive calibration)
00000: + 0 ppm
00001: + 4 ppm (rounded value)
00010: + 8 ppm (rounded value)
..
11111: + 126 ppm (rounded value)
DCS = 1 (negative calibration)
00000: − 0 ppm
00001: − 2 ppm (rounded value)
00010: − 4 ppm (rounded value)
..
11111: − 63 ppm (rounded value)
Refer to Case of RTCCLK=32.768 kHz and PREDIV_A+1=128 on page 808 for the exact
step value.

824/1747

RM0090 Rev 17

RM0090

Real-time clock (RTC)

Note:

This register can be written in initialization mode only (RTC_ISR/INITF = ‘1’).
This register is write protected. The write access procedure is described in RTC register
write protection on page 803.

26.6.8

RTC alarm A register (RTC_ALRMAR)
Address offset: 0x1C
Backup domain reset value: 0x0000 0000
System reset: not affected

31

30

MSK4

WDSEL

29

28

27

DT[1:0]

26

25

24

DU[3:0]

23

22

MSK3

PM

21

20

19

18

HT[1:0]

17

16

HU[3:0]

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

MSK2
rw

MNT[2:0]
rw

rw

MNU[3:0]
rw

MSK1
rw

rw

ST[2:0]
rw

rw

SU[3:0]

Bit 31 MSK4: Alarm A date mask
0: Alarm A set if the date/day match
1: Date/day don’t care in Alarm A comparison
Bit 30 WDSEL: Week day selection
0: DU[3:0] represents the date units
1: DU[3:0] represents the week day. DT[1:0] is don’t care.
Bits 29:28 DT[1:0]: Date tens in BCD format.
Bits 27:24 DU[3:0]: Date units or day in BCD format.
Bit 23 MSK3: Alarm A hours mask
0: Alarm A set if the hours match
1: Hours don’t care in Alarm A comparison
Bit 22 PM: AM/PM notation
0: AM or 24-hour format
1: PM
Bits 21:20 HT[1:0]: Hour tens in BCD format.
Bits 19:16 HU[3:0]: Hour units in BCD format.
Bit 15 MSK2: Alarm A minutes mask
0: Alarm A set if the minutes match
1: Minutes don’t care in Alarm A comparison
Bits 14:12 MNT[2:0]: Minute tens in BCD format.
Bits 11:8 MNU[3:0]: Minute units in BCD format.
Bit 7 MSK1: Alarm A seconds mask
0: Alarm A set if the seconds match
1: Seconds don’t care in Alarm A comparison
Bits 6:4 ST[2:0]: Second tens in BCD format.
Bits 3:0 SU[3:0]: Second units in BCD format.

RM0090 Rev 17

825/1747
838

Real-time clock (RTC)
Note:

RM0090

This register can be written only when ALRAWF is set to 1 in RTC_ISR, or in initialization
mode.
This register is write protected. The write access procedure is described in RTC register
write protection on page 803.

26.6.9

RTC alarm B register (RTC_ALRMBR)
Address offset: 0x20
Backup domain reset value: 0x0000 0000
System reset: not affected

31

30

MSK4

WDSEL

rw

rw

15

14

MSK2
rw

29

28

27

rw
13

rw
12

rw

25

24

DU[3:0]
rw
11

MNT[2:0]
rw

26

DT[1:0]

rw
10

rw

rw

9

8

MNU[3:0]
rw

rw

rw

rw

23

22

MSK3

PM

rw

rw

7

6

MSK1
rw

rw

21

rw
5

rw

rw

Bit 30 WDSEL: Week day selection
0: DU[3:0] represents the date units
1: DU[3:0] represents the week day. DT[1:0] is don’t care.
Bits 29:28 DT[1:0]: Date tens in BCD format
Bits 27:24 DU[3:0]: Date units or day in BCD format
Bit 23 MSK3: Alarm B hours mask
0: Alarm B set if the hours match
1: Hours don’t care in Alarm B comparison
Bit 22 PM: AM/PM notation
0: AM or 24-hour format
1: PM
Bits 21:20 HT[1:0]: Hour tens in BCD format
Bits 19:16 HU[3:0]: Hour units in BCD format
Bit 15 MSK2: Alarm B minutes mask
0: Alarm B set if the minutes match
1: Minutes don’t care in Alarm B comparison
Bits 14:12 MNT[2:0]: Minute tens in BCD format
Bits 11:8 MNU[3:0]: Minute units in BCD format
Bit 7 MSK1: Alarm B seconds mask
0: Alarm B set if the seconds match
1: Seconds don’t care in Alarm B comparison

Bits 3:0 SU[3:0]: Second units in BCD format

826/1747

RM0090 Rev 17

19

18

rw
4

17

16

HU[3:0]
rw
3

rw
2

ST[2:0]

Bit 31 MSK4: Alarm B date mask
0: Alarm B set if the date and day match
1: Date and day don’t care in Alarm B comparison

Bits 6:4 ST[2:0]: Second tens in BCD format

20

HT[1:0]

rw
1

rw
0

SU[3:0]
rw

rw

rw

rw

rw

RM0090

Real-time clock (RTC)

Note:

This register can be written only when ALRBWF is set to 1 in RTC_ISR, or in initialization
mode.
This register is write protected. The write access procedure is described in RTC register
write protection on page 803.

26.6.10

RTC write protection register (RTC_WPR)
Address offset: 0x24
Backup domain reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

6

5

4

19

18

17

16

3

2

1

0

w

w

w

w

Reserved
15

14

13

12

11

10

9

8

7

KEY

Reserved

w

w

w

w

Bits 31:8 Reserved, must be kept at reset value.
Bits 7:0 KEY: Write protection key
This byte is written by software.
Reading this byte always returns 0x00.
Refer to RTC register write protection for a description of how to unlock RTC register write
protection.

26.6.11

RTC sub second register (RTC_SSR)
Address offset: 0x28
Backup domain reset value: 0x0000 0000
System reset: 0x0000 0000 when BYPSHAD = 0. Not affected when BYPSHAD = 1.

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

r

r

r

r

Reserved
r

r

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

SS[15:0]
r

Bits 31:16 Reserved
Bits 15:0 SS: Sub second value
SS[15:0] is the value in the synchronous prescaler’s counter. The fraction of a second is
given by the formula below:
Second fraction = ( PREDIV_S - SS ) / ( PREDIV_S + 1 )
Note: SS can be larger than PREDIV_S only after a shift operation. In that case, the correct
time/date is one second less than as indicated by RTC_TR/RTC_DR.

RM0090 Rev 17

827/1747
838

Real-time clock (RTC)

26.6.12

RM0090

RTC shift control register (RTC_SHIFTR)
Address offset: 0x2C
Backup domain reset value: 0x0000 0000
System reset: not affected

31

30

29

28

27

26

25

24

ADD1S

23

22

21

20

19

18

17

16

r

r

r

r

Reserved

w

r

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

w

w

w

w

w

w

w

w

w

w

w

w

w

w

Res.

SUBFS[14:0]

r

w

Bit 31 ADD1S: Add one second
0: No effect
1: Add one second to the clock/calendar
This bit is write only and is always read as zero. Writing to this bit has no effect when a shift
operation is pending (when SHPF=1, in RTC_ISR).
This function is intended to be used with SUBFS (see description below) in order to
effectively add a fraction of a second to the clock in an atomic operation.
Bits 30:15 Reserved
Bits 14:0 SUBFS: Subtract a fraction of a second
These bits are write only and is always read as zero. Writing to this bit has no effect when a
shift operation is pending (when SHPF=1, in RTC_ISR).
The value which is written to SUBFS is added to the synchronous prescaler’s counter. Since
this counter counts down, this operation effectively subtracts from (delays) the clock by:
Delay (seconds) = SUBFS / ( PREDIV_S + 1 )
A fraction of a second can effectively be added to the clock (advancing the clock) when the
ADD1S function is used in conjunction with SUBFS, effectively advancing the clock by:
Advance (seconds) = ( 1 - ( SUBFS / ( PREDIV_S + 1 ) ) ) .
Note: Writing to SUBFS causes RSF to be cleared. Software can then wait until RSF=1 to be
sure that the shadow registers have been updated with the shifted time.
Refer to Section 26.3.8: RTC synchronization.

Note:

This register is write protected. The write access procedure is described in RTC register
write protection on page 803

26.6.13

RTC time stamp time register (RTC_TSTR)
Address offset: 0x30
Backup domain reset value: 0x0000 0000
System reset: not affected

31

30

29

28

27

26

25

24

23

Reserved

828/1747

22

21

PM
r

RM0090 Rev 17

20

19

18

HT[1:0]
r

17

16

HU[3:0]
r

r

r

r

r

RM0090

Real-time clock (RTC)

15

14

13

12

11

MNT[2:0]

Res.

r

r

10

9

8

MNU[3:0]
r

r

r

r

r

7
Res.

6

5

4

3

2

ST[2:0]
r

r

1

0

SU[3:0]
r

r

r

r

r

Bits 31:23 Reserved, must be kept at reset value.
Bit 22 PM: AM/PM notation
0: AM or 24-hour format
1: PM
Bits 21:20 HT[1:0]: Hour tens in BCD format.
Bits 19:16 HU[3:0]: Hour units in BCD format.
Bit 15

Reserved, must be kept at reset value.

Bits 14:12 MNT[2:0]: Minute tens in BCD format.
Bits 11:8 MNU[3:0]: Minute units in BCD format.
Bit 7 Reserved, must be kept at reset value.
Bits 6:4 ST[2:0]: Second tens in BCD format.
Bits 3:0 SU[3:0]: Second units in BCD format.

Note:

The content of this register is valid only when TSF is set to 1 in RTC_ISR. It is cleared when
TSF bit is reset.

26.6.14

RTC time stamp date register (RTC_TSDR)
Address offset: 0x34
Backup domain reset value: 0x0000 0000
System reset: not affected

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

r

r

r

r

r

Reserved
15

14

13

WDU[1:0]
r

r

12

11

10

r

r

MT
r

r

9

8

r

r

MU[3:0]

7

Reserved

DT[1:0]
r

DU[3:0]

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:13 WDU[1:0]: Week day units
Bit 12 MT: Month tens in BCD format
Bits 11:8 MU[3:0]: Month units in BCD format
Bits 7:6 Reserved, must be kept at reset value.
Bits 5:4 DT[1:0]: Date tens in BCD format
Bit 3:0 DU[3:0]: Date units in BCD format

Note:

The content of this register is valid only when TSF is set to 1 in RTC_ISR. It is cleared when
TSF bit is reset.

RM0090 Rev 17

829/1747
838

Real-time clock (RTC)

26.6.15

RM0090

RTC timestamp sub second register (RTC_TSSSR)
Address offset: 0x38
Backup domain reset value: 0x0000 0000
System reset: not affected

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

r

r

r

r

Reserved
r

r

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

SS[15:0]
r

Bits 31:16 Reserved
Bits 15:0 SS: Sub second value
SS[15:0] is the value of the synchronous prescaler’s counter when the timestamp event
occurred.

Note:

The content of this register is valid only when RTC_ISR/TSF is set. It is cleared when the
RTC_ISR/TSF bit is reset.

26.6.16

RTC calibration register (RTC_CALR)
Address offset: 0x3C
Backup domain reset value: 0x0000 0000
System reset: not affected

31

30

29

28

27

26

25

24

r

r

r

r

r

r

r

r

14

13

10

9

8

23

22

21

20

19

18

17

16

r

r

r

r

r

r

r

r

7

6

5

Reserved

15
CALP

12

CALW8 CALW16

rw

830/1747

rw

rw

11

Reserved
r

r

r

4

3

2

1

0

rw

rw

rw

rw

CALM[8:0]
r

rw

rw

RM0090 Rev 17

rw

rw

rw

RM0090

Real-time clock (RTC)

Bits 31:16 Reserved
Bit 15 CALP: Increase frequency of RTC by 488.5 ppm
0: No RTCCLK pulses are added.
1: One RTCCLK pulse is effectively inserted every 211 pulses (frequency increased by
488.5 ppm).
This feature is intended to be used in conjunction with CALM, which lowers the frequency of
the calendar with a fine resolution. if the input frequency is 32768 Hz, the number of
RTCCLK pulses added during a 32-second window is calculated as follows: (512 * CALP) CALM.
Refer to Section 26.3.11: RTC smooth digital calibration.
Bit 14 CALW8: Use an 8-second calibration cycle period
When CALW8 is set to ‘1’, the 8-second calibration cycle period is selected.
CALM[1:0] are stuck at “00” when CALW8=’1’.
Refer to Section 26.3.11: RTC smooth digital calibration.
Bit 13 CALW16: Use a 16-second calibration cycle period
When CALW16 is set to ‘1’, the 16-second calibration cycle period is selected. This bit must
not be set to ‘1’ if CALW8=1.
Note: CALM[0] is stuck at ‘0’ when CALW16=’1’.
Refer to Section 26.3.11: RTC smooth digital calibration.
Bits 12:9 Reserved
Bits 8:0 CALM[8:0]: Calibration minus
The frequency of the calendar is reduced by masking CALM out of 220 RTCCLK pulses (32
seconds if the input frequency is 32768 Hz). This decreases the frequency of the calendar
with a resolution of 0.9537 ppm.
To increase the frequency of the calendar, this feature should be used in conjunction with
CALP.
See Section 26.3.11: RTC smooth digital calibration on page 809.

Note:

This register is write protected. The write access procedure is described in RTC register
write protection on page 803

RM0090 Rev 17

831/1747
838

Real-time clock (RTC)

26.6.17

RM0090

RTC tamper and alternate function configuration register
(RTC_TAFCR)
Address offset: 0x40
Backup domain reset value: 0x0000 0000
System reset: not affected

31

30

29

28

27

26

25

24

23

22

21

20

19

Reserved
15
TAMPPUDIS
rw

14

13

TAMPPRCH[1:0]
rw

rw

12

11

TAMPFLT[1:0]
rw

rw

10

9

8

TAMPFREQ[2:0]
rw

rw

7
TAMPT
S

rw

6

5

Reserved

rw

4

3

TAMP2 TAMP2
-TRG
E
rw

rw

18

17

16

ALARMOUT
TYPE

TSIN
SEL

TAMP1
INSEL

rw

rw

rw

2

1

0

TAMPIE
rw

TAMP1 TAMP1
TRG
E
rw

rw

Bits 31:19 Reserved. Always read as 0.
Bit 18 ALARMOUTTYPE: RTC_ALARM output type
0: RTC_ALARM is an open-drain output
1: RTC_ALARM is a push-pull output
Bit 17 TSINSEL: TIMESTAMP mapping
0: RTC_AF1 used as TIMESTAMP
1: RTC_AF2 used as TIMESTAMP
Bit 16 TAMP1INSEL: TAMPER1 mapping
0: RTC_AF1 used as TAMPER1
1: RTC_AF2 used as TAMPER1
Note: TAMP1E must be reset when TAMP1INSEL is changed to avoid unwanted setting of
TAMP1F.
Bit 15 TAMPPUDIS: TAMPER pull-up disable
This bit determines if each of the tamper pins are pre-charged before each sample.
0: Precharge tamper pins before sampling (enable internal pull-up)
1: Disable precharge of tamper pins
Note:
Bits 14:13 TAMPPRCH[1:0]: Tamper precharge duration
These bit determines the duration of time during which the pull-up/is activated before each
sample. TAMPPRCH is valid for each of the tamper inputs.
0x0: 1 RTCCLK cycle
0x1: 2 RTCCLK cycles
0x2: 4 RTCCLK cycles
0x3: 8 RTCCLK cycles

832/1747

RM0090 Rev 17

RM0090

Real-time clock (RTC)

Bits 12:11 TAMPFLT[1:0]: Tamper filter count
These bits determines the number of consecutive samples at the specified level
(TAMP*TRG) necessary to activate a Tamper event. TAMPFLT is valid for each of the tamper
inputs.
0x0: Tamper is activated on edge of tamper input transitions to the active level (no internal
pull-up on tamper input).
0x1: Tamper is activated after 2 consecutive samples at the active level.
0x2: Tamper is activated after 4 consecutive samples at the active level.
0x3: Tamper is activated after 8 consecutive samples at the active level.
Bits 10:8 TAMPFREQ[2:0]: Tamper sampling frequency
Determines the frequency at which each of the tamper inputs are sampled.
0x0: RTCCLK / 32768 (1 Hz when RTCCLK = 32768 Hz)
0x1: RTCCLK / 16384 (2 Hz when RTCCLK = 32768 Hz)
0x2: RTCCLK / 8192 (4 Hz when RTCCLK = 32768 Hz)
0x3: RTCCLK / 4096 (8 Hz when RTCCLK = 32768 Hz)
0x4: RTCCLK / 2048 (16 Hz when RTCCLK = 32768 Hz)
0x5: RTCCLK / 1024 (32 Hz when RTCCLK = 32768 Hz)
0x6: RTCCLK / 512 (64 Hz when RTCCLK = 32768 Hz)
0x7: RTCCLK / 256 (128 Hz when RTCCLK = 32768 Hz)
Bit 7 TAMPTS: Activate timestamp on tamper detection event
0: Tamper detection event does not cause a timestamp to be saved
1: Save timestamp on tamper detection event
TAMPTS is valid even if TSE=0 in the RTC_CR register.
Bits 6:5 Reserved. Always read as 0.
Bit 4 TAMP2TRG: Active level for tamper 2
if TAMPFLT != 00
0: TAMPER2 staying low triggers a tamper detection event.
1: TAMPER2 staying high triggers a tamper detection event.
if TAMPFLT = 00:
0: TAMPER2 rising edge triggers a tamper detection event.
1: TAMPER2 falling edge triggers a tamper detection event.
Bit 3 TAMP2E: Tamper 2 detection enable
0: Tamper 2 detection disabled
1: Tamper 2 detection enabled
Bit 2 TAMPIE: Tamper interrupt enable
0: Tamper interrupt disabled
1: Tamper interrupt enabled

RM0090 Rev 17

833/1747
838

Real-time clock (RTC)

RM0090

Bit 1 TAMP1TRG: Active level for tamper 1
if TAMPFLT != 00
0: TAMPER1 staying low triggers a tamper detection event.
1: TAMPER1 staying high triggers a tamper detection event.
if TAMPFLT = 00:
0: TAMPER1 rising edge triggers a tamper detection event.
1: TAMPER1 falling edge triggers a tamper detection event.
Caution: When TAMPFLT = 0, TAMPxE must be reset when TAMPxTRG is changed to avoid
spuriously setting TAMPxF.
Bit 0 TAMP1E: Tamper 1 detection enable
0: Tamper 1 detection disabled
1: Tamper 1 detection enabled

26.6.18

RTC alarm A sub second register (RTC_ALRMASSR)
Address offset: 0x44
Backup domain reset value: 0x0000 0000
System reset: not affected

31

30

29

28

27

Reserved
r

r

15

26

25

24

22

21

r

r

rw

14

13

12

11

rw

rw

rw

rw

rw

20

19

18

17

16

r

r

r

Reserved

rw

rw

10

9

8

rw

rw

rw

Reserved
r

23

MASKSS[3:0]
r

r

r

r

r

7

6

5

4

3

2

1

0

rw

rw

rw

rw

w

rw

rw

SS[14:0]
rw

Bits 31:28 Reserved
Bits 27:24 MASKSS[3:0]: Mask the most-significant bits starting at this bit
0: No comparison on sub seconds for Alarm A. The alarm is set when the seconds unit is
incremented (assuming that the rest of the fields match).
1: SS[14:1] are don’t care in Alarm A comparison. Only SS[0] is compared.
2: SS[14:2] are don’t care in Alarm A comparison. Only SS[1:0] are compared.
3: SS[14:3] are don’t care in Alarm A comparison. Only SS[2:0] are compared.
...
12: SS[14:12] are don’t care in Alarm A comparison. SS[11:0] are compared.
13: SS[14:13] are don’t care in Alarm A comparison. SS[12:0] are compared.
14: SS[14] is don’t care in Alarm A comparison. SS[13:0] are compared.
15: All 15 SS bits are compared and must match to activate alarm.
The overflow bits of the synchronous counter (bits 15) is never compared. This bit can be
different from 0 only after a shift operation.
Bits 23:15 Reserved
Bits 14:0 SS[14:0]: Sub seconds value
This value is compared with the contents of the synchronous prescaler’s counter to
determine if Alarm A is to be activated. Only bits 0 up MASKSS-1 are compared.

834/1747

RM0090 Rev 17

RM0090

Real-time clock (RTC)

Note:

This register can be written only when ALRAE is reset in RTC_CR register, or in initialization
mode.
This register is write protected. The write access procedure is described in RTC register
write protection on page 803

26.6.19

RTC alarm B sub second register (RTC_ALRMBSSR)
Address offset: 0x48
Backup domain reset value: 0x0000 0000
System reset: not affected

31

30

29

28

27

Reserved
r

r

15

14

26

25

24

22

21

r

r

13

12

rw
11

rw
10

20

19

18

17

16

r

r

r

Reserved

rw

rw

9

8

Reserved
r

23

MASKSS[3:0]
r

r

r

r

r

7

6

5

4

3

2

1

0

rw

rw

rw

rw

w

rw

rw

SS[14:0]
rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:28 Reserved
Bits 27:24 MASKSS[3:0]: Mask the most-significant bits starting at this bit
0x0: No comparison on sub seconds for Alarm B. The alarm is set when the seconds unit is
incremented (assuming that the rest of the fields match).
0x1: SS[14:1] are don’t care in Alarm B comparison. Only SS[0] is compared.
0x2: SS[14:2] are don’t care in Alarm B comparison. Only SS[1:0] are compared.
0x3: SS[14:3] are don’t care in Alarm B comparison. Only SS[2:0] are compared.
...
0xC: SS[14:12] are don’t care in Alarm B comparison. SS[11:0] are compared.
0xD: SS[14:13] are don’t care in Alarm B comparison. SS[12:0] are compared.
0xE: SS[14] is don’t care in Alarm B comparison. SS[13:0] are compared.
0xF: All 15 SS bits are compared and must match to activate alarm.
The overflow bits of the synchronous counter (bits 15) is never compared. This bit can be
different from 0 only after a shift operation.
Bits 23:15 Reserved
Bits 14:0 SS[14:0]: Sub seconds value
This value is compared with the contents of the synchronous prescaler’s counter to
determine if Alarm B is to be activated. Only bits 0 up to MASKSS-1 are compared.

Note:

This register can be written only when ALRBIE is reset in RTC_CR register, or in
initialization mode.
This register is write protected.The write access procedure is described in Section : RTC
register write protection

RM0090 Rev 17

835/1747
838

Real-time clock (RTC)

26.6.20

RM0090

RTC backup registers (RTC_BKPxR)
Address offset: 0x50 to 0x9C
Backup domain reset value: 0x0000 0000
System reset: not affected

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

BKP[31:16]
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

w

rw

rw

BKP[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 BKP[31:0]
The application can write or read data to and from these registers.
They are powered-on by VBAT when VDD is switched off, so that they are not reset by
System reset, and their contents remain valid when the device operates in low-power mode.
This register is reset on a tamper detection event, as long as TAMPxF=1.

26.6.21

RTC register map

WDU[2:0]

RTC_CALIBR

ALRBE

ALRAE

DCE

FMT

BYPSHAD

REFCKON

TSEDGE

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

ALRAWF

0

ALRBWF

0

WUTWF

0

SHPF

0

INITS

1

WCKSEL
[2:0]

RSF

1

1

INITF

WUTE

0

0

INIT

TSE

0

0

ALRAF

ALRAIE

0

1

0

ALRBF

ALRBIE

0

1

2

3

4

5

6

7

8

9

11

10

WUTIE

0

1

0

WUTF

TSIE

0

0

0

0

0

1

0

1

1

1

1

1

1

1

1

1

1

1

1

1

PREDIV_S[14:0]

0 0

0

0

0

0

0

1

1

WUT[15:0]

Reserved
1
Reserved

Reset value

836/1747

0

0

Reset value
0x18

0

0

1

1

1

1

1

1

1

1

0

RM0090 Rev 17

1
Reserved

RTC_WUTR

1

0

DU[3:0]

0

DCS

0x14

1

1

DT
[1:0]

0

PREDIV_A[6:0]

Reset value

0

0

0

Reserved

Reserved

0

0

0

Reserved

RTC_PRER

0

0

ADD1H

0

Reset value

0x10

MU[3:0]

0

SUB1H

0

1

0

TSF

RTC_ISR

0

0

0

TSOVF

0x0C

0

0

0

BKP

Reset value

OSEL
[1:0]

0

COSEL

Reserved

0

SU[3:0]

POL

RTC_CR

COE

Reset value

0x08

0

ST[2:0]

Reserved

YU[3:0]

0

12

0

Reserved

0

MT

0

13

14

17

15

0

YT[3:0]

Reserved

18

19

20
0

MNU[3:0]

TAMP1F

RTC_DR

0

MNT[2:0]

HU[3:0]

TAMP2F

0x04

0

HT
[1:0]

16

Reset value

21

22

23

24

25

26

27

28

29

Reserved

Reserved

RTC_TR

PM

0x00

Register

30

Offset

31

Table 121. RTC register map and reset values

DC[4:0]
0

0

0

0

0

RM0090

Real-time clock (RTC)

0

DU[3:0]
0

0

0

0

0

0

RTC_WPR

0

0

HT
[1:0]
0

0

0

0

0

HU[3:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

MNU[3:0]
0

0

0

0

0

RTC_SSR

RTC_SHIFTR

ADD1S

0

Reset value

0

0

0

0

0

0

0

0

MNT[2:0]
0

0

0

1

2

3

0
0

SU[3:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

MNU[3:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CALM[8:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

SS[14:0]

0

0

0

0

0

0

0

0

0

0

SS[14:0]

Reserved

0

RTC_BKP0R

0

TAMP1E

0

0

Reserved

0

0

TAMPFREQ[2:0]

0

0

TAMPFLT[1:0]

0

0

Reserved

CALW16
0

TAMPPRCH[1:0]

CALW8
0

TAMPPUDIS

0
TAMP1INSEL

0

TSINSEL

0

Reserved

MASKSS[3:0]

Reserved

0

TAMP1ETRG

0

SU[3:0]

TAMPIE

0

0

ST[2:0]

TAMP2E

0

RTC_
ALRMBSSR

4

5

6
0

0

TAMP2TRG

MASKSS[3:0]

Reserved

0

ALARMOUTTYPE

Reserved

RTC_
ALRMASSR

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

BKP[31:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

to
RTC_BKP19R
Reset value

ST[2:0]

0

SS[15:0]

Reserved

RTC_TAFCR

Reset value

0

CALP

RTC_ CALR

Reset value

0x50
to 0x9C

0

Reserved

Reset value
0x48

0

TAMPTS

RTC_TSSSR

0

HU[3:0]

Reserved

Reserved

HT[1:0]

RTC_TSTR

Reset value
0x44

0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Reset value

0x40

0

SUBFS[14:0]

Reset value

0x3C

0

Reserved

Reset value
0x38

0

SS[15:0]

Reserved

PM

0x30

0

0

SU[3:0]

KEY[7:0]

Reset value

0x2C

0

ST[2:0]

Reserved

Reset value
0x28

7

8

MSK1

9

11

MNU[3:0]

MNT[2:0]

0

10

12

14

MNT[2:0]

0

13

15

16

17

18

19

20

0

MSK2

0

21

22

23
0

HU[3:0]

MSK2

0

HT
[1:0]

MSK2

0

PM

0

PM

DT
[1:0]
0

24

0

0

MSK3

0

DU[3:0]

MSK3

Reset value

0

25

RTC_ALRMBR

0

26

0

27

30

0

28

31
MSK4

Reset value

DT
[1:0]

Reserved

0x24

WDSEL

0x20

RTC_ALRMAR

MSK4

0x1C

Register

WDSEL

Offset

29

Table 121. RTC register map and reset values (continued)

0

0

0

BKP[31:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

RM0090 Rev 17

0

0

837/1747
838

Real-time clock (RTC)

RM0090

Refer to Section 2.3: Memory map for the register boundary addresses.
Caution:

838/1747

In Table 121, the reset value is the value after a backup domain reset. The majority of the
registers are not affected by a system reset. For more information, please refer to
Section 26.3.7: Resetting the RTC.

RM0090 Rev 17

RM0090

27

Inter-integrated circuit (I2C) interface

Inter-integrated circuit (I2C) interface
This section applies to the whole STM32F4xx family, unless otherwise specified.

27.1

I2C introduction
I2C (inter-integrated circuit) bus Interface serves as an interface between the microcontroller
and the serial I2C bus. It provides multimaster capability, and controls all I2C bus-specific
sequencing, protocol, arbitration and timing. It supports the standard mode (Sm, up to
100 kHz) and Fm mode (Fm, up to 400 kHz).
It may be used for a variety of purposes, including CRC generation and verification, SMBus
(system management bus) and PMBus (power management bus).
Depending on specific device implementation DMA capability can be available for reduced
CPU overload.

27.2

I2C main features
•

Parallel-bus/I2C protocol converter

•

Multimaster capability: the same interface can act as Master or Slave

•

I2C Master features:

•

–

Clock generation

–

Start and Stop generation

I2C

Slave features:

–

Programmable I2C Address detection

–

Dual Addressing Capability to acknowledge 2 slave addresses

–

Stop bit detection

•

Generation and detection of 7-bit/10-bit addressing and General Call

•

Supports different communication speeds:
–

Standard Speed (up to 100 kHz)

–

Fast Speed (up to 400 kHz)

•

Analog noise filter

•

Programmable digital noise filter for STM32F42xxx and STM32F43xxx

•

Status flags:

•

•

–

Transmitter/Receiver mode flag

–

End-of-Byte transmission flag

–

I2C busy flag

Error flags:
–

Arbitration lost condition for master mode

–

Acknowledgment failure after address/ data transmission

–

Detection of misplaced start or stop condition

–

Overrun/Underrun if clock stretching is disabled

2 Interrupt vectors:

RM0090 Rev 17

839/1747
872

Inter-integrated circuit (I2C) interface

RM0090

–

1 Interrupt for successful address/ data communication

–

1 Interrupt for error condition

•

Optional clock stretching

•

1-byte buffer with DMA capability

•

Configurable PEC (packet error checking) generation or verification:

•

•

–

PEC value can be transmitted as last byte in Tx mode

–

PEC error checking for last received byte

SMBus 2.0 Compatibility:
–

25 ms clock low timeout delay

–

10 ms master cumulative clock low extend time

–

25 ms slave cumulative clock low extend time

–

Hardware PEC generation/verification with ACK control

–

Address Resolution Protocol (ARP) supported

PMBus Compatibility

Note:

Some of the above features may not be available in certain products. The user should refer
to the product data sheet, to identify the specific features supported by the I2C interface
implementation.

27.3

I2C functional description
In addition to receiving and transmitting data, this interface converts it from serial to parallel
format and vice versa. The interrupts are enabled or disabled by software. The interface is
connected to the I2C bus by a data pin (SDA) and by a clock pin (SCL). It can be connected
with a standard (up to 100 kHz) or fast (up to 400 kHz) I2C bus.

27.3.1

Mode selection
The interface can operate in one of the four following modes:
•

Slave transmitter

•

Slave receiver

•

Master transmitter

•

Master receiver

By default, it operates in slave mode. The interface automatically switches from slave to
master, after it generates a START condition and from master to slave, if an arbitration loss
or a Stop generation occurs, allowing multimaster capability.

Communication flow
In Master mode, the I2C interface initiates a data transfer and generates the clock signal. A
serial data transfer always begins with a start condition and ends with a stop condition. Both
start and stop conditions are generated in master mode by software.
In Slave mode, the interface is capable of recognizing its own addresses (7 or 10-bit), and
the General Call address. The General Call address detection may be enabled or disabled
by software.

840/1747

RM0090 Rev 17

RM0090

Inter-integrated circuit (I2C) interface
Data and addresses are transferred as 8-bit bytes, MSB first. The first byte(s) following the
start condition contain the address (one in 7-bit mode, two in 10-bit mode). The address is
always transmitted in Master mode.
A 9th clock pulse follows the 8 clock cycles of a byte transfer, during which the receiver must
send an acknowledge bit to the transmitter. Refer to Figure 238.
Figure 238. I2C bus protocol
6'$
06%

$&.

6&/






6WDUW
FRQGLWLRQ



6WRS
FRQGLWLRQ
069

Acknowledge may be enabled or disabled by software. The I2C interface addresses (dual
addressing 7-bit/ 10-bit and/or general call address) can be selected by software.
The block diagram of the I2C interface is shown in Figure 239.

RM0090 Rev 17

841/1747
872

Inter-integrated circuit (I2C) interface

RM0090

Figure 239. I2C block diagram for STM32F40x/41x

$ATA REGISTER

3$!

.OISE
FILTER

$ATA
CONTROL

$ATA SHIFT REGISTER

0%# CALCULATION

#OMPARATOR

/WN ADDRESS REGISTER
$UAL ADDRESS REGISTER
3#,

.OISE
FILTER

#LOCK
CONTROL

0%# REGISTER

#LOCK CONTROL
2EGISTER ##2
#ONTROL REGISTERS
#2#2
#ONTROL
LOGIC

3TATUS REGISTERS
3232
3-"!
)NTERRUPTS

$-! REQUESTS  !#+
-36

842/1747

RM0090 Rev 17

RM0090

Inter-integrated circuit (I2C) interface
Figure 240. I2C block diagram for STM32F42x/43x

$ATA REGISTER

3$!

.OISE
FILTER

$ATA
CONTROL

$ATA SHIFT REGISTER

0%# CALCULATION

#OMPARATOR

/WN ADDRESS REGISTER
$UAL ADDRESS REGISTER
3#,

.OISE
FILTER

#LOCK
CONTROL

0%# REGISTER

#LOCK CONTROL
2EGISTER ##2
#ONTROL REGISTERS
#2#2
#ONTROL
LOGIC

3TATUS REGISTERS
3232
3-"!
)NTERRUPTS

$-! REQUESTS  !#+
-36

1. SMBA is an optional signal in SMBus mode. This signal is not applicable if SMBus is disabled.

27.3.2

I2C slave mode
By default the I2C interface operates in Slave mode. To switch from default Slave mode to
Master mode a Start condition generation is needed.
The peripheral input clock must be programmed in the I2C_CR2 register in order to
generate correct timings. The peripheral input clock frequency must be at least:
•

2 MHz in Sm mode

•

4 MHz in Fm mode

As soon as a start condition is detected, the address is received from the SDA line and sent
to the shift register. Then it is compared with the address of the interface (OAR1) and with
OAR2 (if ENDUAL=1) or the General Call address (if ENGC = 1).
Note:

In 10-bit addressing mode, the comparison includes the header sequence (11110xx0),
where xx denotes the two most significant bits of the address.

RM0090 Rev 17

843/1747
872

Inter-integrated circuit (I2C) interface

RM0090

Header or address not matched: the interface ignores it and waits for another Start
condition.
Header matched (10-bit mode only): the interface generates an acknowledge pulse if the
ACK bit is set and waits for the 8-bit slave address.
Address matched: the interface generates in sequence:
•

An acknowledge pulse if the ACK bit is set

•

The ADDR bit is set by hardware and an interrupt is generated if the ITEVFEN bit is
set.

•

If ENDUAL=1, the software has to read the DUALF bit to check which slave address
has been acknowledged.

In 10-bit mode, after receiving the address sequence the slave is always in Receiver mode.
It will enter Transmitter mode on receiving a repeated Start condition followed by the header
sequence with matching address bits and the least significant bit set (11110xx1).
The TRA bit indicates whether the slave is in Receiver or Transmitter mode.

Slave transmitter
Following the address reception and after clearing ADDR, the slave sends bytes from the
DR register to the SDA line via the internal shift register.
The slave stretches SCL low until ADDR is cleared and DR filled with the data to be sent
(see Figure 241 Transfer sequencing EV1 EV3).
When the acknowledge pulse is received:
•

The TxE bit is set by hardware with an interrupt if the ITEVFEN and the ITBUFEN bits
are set.

If TxE is set and some data were not written in the I2C_DR register before the end of the
next data transmission, the BTF bit is set and the interface waits until BTF is cleared by a
read to I2C_SR1 followed by a write to the I2C_DR register, stretching SCL low.

844/1747

RM0090 Rev 17

RM0090

Inter-integrated circuit (I2C) interface
Figure 241. Transfer sequence diagram for slave transmitter
ELWVODYHWUDQVPLWWHU
6 $GGUHVV

$

'DWD
(9 (9 (9

$

'DWD

$

(9

(9



'DWD1

1$

3
(9

ELWVODYHWUDQVPLWWHU
6 +HDGHU

$

$GGUHVV

$
(9
6U +HDGHU $

'DWD
(9 (9B

(9

$

 'DWD1

1$ 3

(9

(9

/HJHQG6 6WDUW6U 5HSHDWHG6WDUW3 6WRS$ $FNQRZOHGJH1$ 1RQDFNQRZOHGJH
(9[ (YHQW ZLWKLQWHUUXSWLI,7(9)(1 
(9 $''5 FOHDUHGE\UHDGLQJ65IROORZHGE\UHDGLQJ65
(97[( VKLIWUHJLVWHUHPSW\GDWDUHJLVWHUHPSW\ZULWH'DWDLQ'5
(97[( VKLIWUHJLVWHUQRWHPSW\GDWDUHJLVWHUHPSW\FOHDUHGE\ZULWLQJ'5

(9$) $)LVFOHDUHGE\ZULWLQJµ¶LQ$)ELWRI65UHJLVWHU
DL

1. The EV1 and EV3_1 events stretch SCL low until the end of the corresponding software sequence.
2. The EV3 event stretches SCL low if the software sequence is not completed before the end of the next byte
transmission.

Slave receiver
Following the address reception and after clearing ADDR, the slave receives bytes from the
SDA line into the DR register via the internal shift register. After each byte the interface
generates in sequence:
•

An acknowledge pulse if the ACK bit is set

•

The RxNE bit is set by hardware and an interrupt is generated if the ITEVFEN and
ITBUFEN bit is set.

If RxNE is set and the data in the DR register is not read before the end of the next data
reception, the BTF bit is set and the interface waits until BTF is cleared by a read from the
I2C_DR register, stretching SCL low (see Figure 242 Transfer sequencing).

RM0090 Rev 17

845/1747
872

Inter-integrated circuit (I2C) interface

RM0090

Figure 242. Transfer sequence diagram for slave receiver
ELWVODYH UHFHLYHU
6 $GGUHVV

$

'DWD

$

(9

'DWD

$

(9

(9



'DWD1

$

3
(9

(9

ELWVODY HUHFHLYHU
6 +HDGHU

$

$GGUHVV

$

'DWD

$

(9

(9



'DWD1

$

3
(9

(9

/HJHQG6 6WDUW6U 5HSHDWHG6WDUW3 6WRS$ $FNQRZOHGJH
(9[ (YHQW ZLWKLQWHUUXSWLI,7(9)(1 
(9 $''5 FOHDUHGE\UHDGLQJ65IROORZHGE\UHDGLQJ65
(95[1( FOHDUHGE\UHDGLQJ'5UHJLVWHU
(9 6723) FOHDUHGE\UHDGLQJ65UHJLVWHUIROORZHGE\ZULWLQJWRWKH&5UHJLVWHU
DL

1. The EV1 event stretches SCL low until the end of the corresponding software sequence.
2. The EV2 event stretches SCL low if the software sequence is not completed before the end of the next byte
reception.
3. After checking the SR1 register content, the user should perform the complete clearing sequence for each
flag found set.
Thus, for ADDR and STOPF flags, the following sequence is required inside the I2C interrupt routine:
READ SR1
if (ADDR == 1) {READ SR1; READ SR2}
if (STOPF == 1) {READ SR1; WRITE CR1}
The purpose is to make sure that both ADDR and STOPF flags are cleared if both are found set.

Closing slave communication
After the last data byte is transferred a Stop Condition is generated by the master. The
interface detects this condition and sets:
•

The STOPF bit and generates an interrupt if the ITEVFEN bit is set.

The STOPF bit is cleared by a read of the SR1 register followed by a write to the CR1
register (see EV4 in Figure 242).

27.3.3

I2C master mode
In Master mode, the I2C interface initiates a data transfer and generates the clock signal. A
serial data transfer always begins with a Start condition and ends with a Stop condition.
Master mode is selected as soon as the Start condition is generated on the bus with a
START bit.
The following is the required sequence in master mode.
•

Program the peripheral input clock in I2C_CR2 Register in order to generate correct
timings

•

Configure the clock control registers

•

Configure the rise time register

•

Program the I2C_CR1 register to enable the peripheral

•

Set the START bit in the I2C_CR1 register to generate a Start condition

The peripheral input clock frequency must be at least:

846/1747

•

2 MHz in Sm mode

•

4 MHz in Fm mode
RM0090 Rev 17

RM0090

Inter-integrated circuit (I2C) interface

SCL master clock generation
The CCR bits are used to generate the high and low level of the SCL clock, starting from the
generation of the rising and falling edge (respectively). As a slave may stretch the SCL line,
the peripheral checks the SCL input from the bus at the end of the time programmed in
TRISE bits after rising edge generation.
•

If the SCL line is low, it means that a slave is stretching the bus, and the high level
counter stops until the SCL line is detected high. This allows to guarantee the minimum
HIGH period of the SCL clock parameter.

•

If the SCL line is high, the high level counter keeps on counting.

Indeed, the feedback loop from the SCL rising edge generation by the peripheral to the SCL
rising edge detection by the peripheral takes time even if no slave stretches the clock. This
loopback duration is linked to the SCL rising time (impacting SCL VIH input detection), plus
delay due to the noise filter present on the SCL input path, plus delay due to internal SCL
input synchronization with APB clock. The maximum time used by the feedback loop is
programmed in the TRISE bits, so that the SCL frequency remains stable whatever the SCL
rising time.

Start condition
Setting the START bit causes the interface to generate a Start condition and to switch to
Master mode (MSL bit set) when the BUSY bit is cleared.
Note:

In master mode, setting the START bit causes the interface to generate a ReStart condition
at the end of the current byte transfer.
Once the Start condition is sent:
•

The SB bit is set by hardware and an interrupt is generated if the ITEVFEN bit is set.

Then the master waits for a read of the SR1 register followed by a write in the DR register
with the Slave address (see Figure 243 and Figure 244 Transfer sequencing EV5).

Slave address transmission
Then the slave address is sent to the SDA line via the internal shift register.
•

In 10-bit addressing mode, sending the header sequence causes the following event:
–

The ADD10 bit is set by hardware and an interrupt is generated if the ITEVFEN bit
is set.

Then the master waits for a read of the SR1 register followed by a write in the DR
register with the second address byte (see Figure 243 and Figure 244 Transfer
sequencing).
–

The ADDR bit is set by hardware and an interrupt is generated if the ITEVFEN bit
is set.

Then the master waits for a read of the SR1 register followed by a read of the SR2
register (see Figure 243 and Figure 244 Transfer sequencing).
•

In 7-bit addressing mode, one address byte is sent.
As soon as the address byte is sent,
–

The ADDR bit is set by hardware and an interrupt is generated if the ITEVFEN bit
is set.

Then the master waits for a read of the SR1 register followed by a read of the SR2
register (see Figure 243 and Figure 244 Transfer sequencing).

RM0090 Rev 17

847/1747
872

Inter-integrated circuit (I2C) interface

RM0090

The master can decide to enter Transmitter or Receiver mode depending on the LSB of the
slave address sent.
•

•

In 7-bit addressing mode,
–

To enter Transmitter mode, a master sends the slave address with LSB reset.

–

To enter Receiver mode, a master sends the slave address with LSB set.

In 10-bit addressing mode,
–

To enter Transmitter mode, a master sends the header (11110xx0) and then the
slave address, (where xx denotes the two most significant bits of the address).

–

To enter Receiver mode, a master sends the header (11110xx0) and then the
slave address. Then it should send a repeated Start condition followed by the
header (11110xx1), (where xx denotes the two most significant bits of the
address).

The TRA bit indicates whether the master is in Receiver or Transmitter mode.

Master transmitter
Following the address transmission and after clearing ADDR, the master sends bytes from
the DR register to the SDA line via the internal shift register.
The master waits until the first data byte is written into I2C_DR (see Figure 243 Transfer
sequencing EV8_1).
When the acknowledge pulse is received, the TxE bit is set by hardware and an interrupt is
generated if the ITEVFEN and ITBUFEN bits are set.
If TxE is set and a data byte was not written in the DR register before the end of the last data
transmission, BTF is set and the interface waits until BTF is cleared by a write to I2C_DR,
stretching SCL low.
Closing the communication
After the last byte is written to the DR register, the STOP bit is set by software to generate a
Stop condition (see Figure 243 Transfer sequencing EV8_2). The interface automatically
goes back to slave mode (MSL bit cleared).
Note:

848/1747

Stop condition should be programmed during EV8_2 event, when either TxE or BTF is set.

RM0090 Rev 17

RM0090

Inter-integrated circuit (I2C) interface
Figure 243. Transfer sequence diagram for master transmitter

ELWPDVWHUWUDQVPLWWHU
6

$GGUHVV

$

(9

'DWD
(9 (9B

$

'DWD

(9

$

(9

(9



'DWD1

$

3
(9B

ELW PDVWHUWUDQVPLWWHU
6

+HDGHU
(9

$

$GGUHVV
(9

$

'DWD
(9

(9B

$

(9

(9



'DWD1

$

3
(9B

/HJHQG6 6WDUW6U 5HSHDWHG6WDUW3 6WRS$ $FNQRZOHGJH
(9[ (YHQW ZLWKLQWHUUXSWLI,7(9)(1 
(9 6% FOHDUHGE\UHDGLQJ65UHJLVWHUIROORZHGE\ZULWLQJ'5UHJLVWHUZLWK$GGUHVV
(9 $''5 FOHDUHGE\UHDGLQJ65UHJLVWHUIROORZHGE\UHDGLQJ65
(9B 7[( VKLIWUHJLVWHUHPSW\GDWDUHJLVWHUHPSW\ZULWH'DWDLQ'5
(9 7[( VKLIWUHJLVWHUQRWHPSW\GDWDUHJLVWHUHPSW\FOHDUHGE\ZULWLQJ'5UHJLVWHU

(9B 7[( %7) 3URJUDP6WRSUHTXHVW7[(DQG%7)DUHFOHDUHGE\KDUGZDUHE\WKH6WRSFRQGLWLRQ
(9 $'' FOHDUHGE\UHDGLQJ65UHJLVWHUIROORZHGE\ZULWLQJ'5UHJLVWHU
DL

1. The EV5, EV6, EV9, EV8_1 and EV8_2 events stretch SCL low until the end of the corresponding software sequence.
2. The EV8 event stretches SCL low if the software sequence is not complete before the end of the next byte transmission.

Master receiver
Following the address transmission and after clearing ADDR, the I2C interface enters
Master Receiver mode. In this mode the interface receives bytes from the SDA line into the
DR register via the internal shift register. After each byte the interface generates in
sequence:
1.

An acknowledge pulse if the ACK bit is set

2.

The RxNE bit is set and an interrupt is generated if the ITEVFEN and ITBUFEN bits are
set (see Figure 244 Transfer sequencing EV7).

If the RxNE bit is set and the data in the DR register is not read before the end of the last
data reception, the BTF bit is set by hardware and the interface waits until BTF is cleared by
a read in the DR register, stretching SCL low.
Closing the communication
The master sends a NACK for the last byte received from the slave. After receiving this
NACK, the slave releases the control of the SCL and SDA lines. Then the master can send
a Stop/Restart condition.
1.

To generate the nonacknowledge pulse after the last received data byte, the ACK bit
must be cleared just after reading the second last data byte (after second last RxNE
event).

2.

In order to generate the Stop/Restart condition, software must set the STOP/START bit
after reading the second last data byte (after the second last RxNE event).

3.

In case a single byte has to be received, the Acknowledge disable is made during EV6
(before ADDR flag is cleared) and the STOP condition generation is made after EV6.

After the Stop condition generation, the interface goes automatically back to slave mode
(MSL bit cleared).

RM0090 Rev 17

849/1747
872

Inter-integrated circuit (I2C) interface

RM0090

Figure 244. Transfer sequence diagram for master receiver
 BIT MASTER RECEIVER
3

!DDRESS

!

%6

$ATA

!

%6

$ATA

!

%6

%6

$ATA.



.!

%6?

0
%6

 BIT MASTER RECEIVER
3

(EADER
%6

!

!DDRESS

!

%6

%6

(EADER

3R

$ATA

!

%6

%6

! $ATA
%6

!
%6



$ATA.

.!

%6?

0
%6

,EGEND 3 3TART 3R  2EPEATED 3TART 0 3TOP ! !CKNOWLEDGE .! .ON ACKNOWLEDGE
%6X %VENT WITH INTERRUPT IF )4%6&%.
%6 3" CLEARED BY READING 32 REGISTER FOLLOWED BY WRITING $2 REGISTER
%6 !$$2 CLEARED BY READING 32 REGISTER FOLLOWED BY READING 32 )N  BIT MASTER RECEIVER MODE THIS SE
QUENCE SHOULD BE FOLLOWED BY WRITING #2 WITH 34!24  
)N CASE OF THE RECEPTION OF  BYTE THE !CKNOWLEDGE DISABLE MUST BE PERFORMED DURING %6 EVENT IE BEFORE CLEARING !$$2 FLAG
%6 2X.% CLEARED BY READING $2 REGISTER
%6? 2X.% CLEARED BY READING $2 REGISTER PROGRAM !# + AND 34/0 REQUEST
%6 !$$ CLEARED BY READING 32 REGISTER FOLLOWED BY WRITING $2 REGISTER
AID

1. If a single byte is received, it is NA.
2. The EV5, EV6 and EV9 events stretch SCL low until the end of the corresponding software sequence.
3. The EV7 event stretches SCL low if the software sequence is not completed before the end of the next byte
reception.
4. The EV7_1 software sequence must be completed before the ACK pulse of the current byte transfer.

The procedures described below are recommended if the EV7-1 software sequence is not
completed before the ACK pulse of the current byte transfer.
These procedures must be followed to make sure:
•

The ACK bit is set low on time before the end of the last data reception

•

The STOP bit is set high after the last data reception without reception of
supplementary data.

For 2-byte reception:

850/1747

•

Wait until ADDR = 1 (SCL stretched low until the ADDR flag is cleared)

•

Set ACK low, set POS high

•

Clear ADDR flag

•

Wait until BTF = 1 (Data 1 in DR, Data2 in shift register, SCL stretched low until a data
1 is read)

•

Set STOP high

•

Read data 1 and 2

RM0090 Rev 17

RM0090

Inter-integrated circuit (I2C) interface
For N >2 -byte reception, from N-2 data reception

27.3.4

•

Wait until BTF = 1 (data N-2 in DR, data N-1 in shift register, SCL stretched low until
data N-2 is read)

•

Set ACK low

•

Read data N-2

•

Wait until BTF = 1 (data N-1 in DR, data N in shift register, SCL stretched low until a
data N-1 is read)

•

Set STOP high

•

Read data N-1 and N

Error conditions
The following are the error conditions which may cause communication to fail.

Bus error (BERR)
This error occurs when the I2C interface detects an external Stop or Start condition during
an address or a data transfer. In this case:
•

the BERR bit is set and an interrupt is generated if the ITERREN bit is set

•

in Slave mode: data are discarded and the lines are released by hardware:

•

–

in case of a misplaced Start, the slave considers it is a restart and waits for an
address, or a Stop condition

–

in case of a misplaced Stop, the slave behaves like for a Stop condition and the
lines are released by hardware

In Master mode: the lines are not released and the state of the current transmission is
not affected. It is up to the software to abort or not the current transmission

Acknowledge failure (AF)
This error occurs when the interface detects a nonacknowledge bit. In this case:
•

the AF bit is set and an interrupt is generated if the ITERREN bit is set

•

a transmitter which receives a NACK must reset the communication:
–

If Slave: lines are released by hardware

–

If Master: a Stop or repeated Start condition must be generated by software

Arbitration lost (ARLO)
This error occurs when the I2C interface detects an arbitration lost condition. In this case
•

the ARLO bit is set by hardware (and an interrupt is generated if the ITERREN bit is
set)

•

the I2C Interface goes automatically back to slave mode (the MSL bit is cleared). When
the I2C loses the arbitration, it is not able to acknowledge its slave address in the same
transfer, but it can acknowledge it after a repeated Start from the winning master.

•

lines are released by hardware

RM0090 Rev 17

851/1747
872

Inter-integrated circuit (I2C) interface

RM0090

Overrun/underrun error (OVR)
An overrun error can occur in slave mode when clock stretching is disabled and the I2C
interface is receiving data. The interface has received a byte (RxNE=1) and the data in DR
has not been read, before the next byte is received by the interface. In this case,
•

The last received byte is lost.

•

In case of Overrun error, software should clear the RxNE bit and the transmitter should
re-transmit the last received byte.

Underrun error can occur in slave mode when clock stretching is disabled and the I2C
interface is transmitting data. The interface has not updated the DR with the next byte
(TxE=1), before the clock comes for the next byte. In this case,
•

The same byte in the DR register will be sent again

•

The user should make sure that data received on the receiver side during an underrun
error are discarded and that the next bytes are written within the clock low time
specified in the I2C bus standard.

For the first byte to be transmitted, the DR must be written after ADDR is cleared and before
the first SCL rising edge. If not possible, the receiver must discard the first data.

27.3.5

Programmable noise filter
The programmable noise filter is available on STM32F42xxx and STM32F43xxx devices
only.
In Fm mode, the I2C standard requires that spikes are suppressed to a length of 50 ns on
SDA and SCL lines.
An analog noise filter is implemented in the SDA and SCL I/Os. This filter is enabled by
default and can be disabled by setting the ANOFF bit in the I2C_FLTR register.
A digital noise filter can be enabled by configuring the DNF[3:0] bits to a non-zero value.
This suppresses the spikes on SDA and SCL inputs with a length of up to DNF[3:0] *
TPCLK1.
Enabling the digital noise filter increases the SDA hold time by (DNF[3:0] +1)* TPCLK.
To be compliant with the maximum hold time of the I2C-bus specification version 2.1
(Thd:dat), the DNF bits must be programmed using the constraints shown in Table 122, and
assuming that the analog filter is disabled.

Note:

DNF[3:0] must only be configured when the I2C is disabled (PE = 0). If the analog filter is
also enabled, the digital filter is added to the analog filter.
Table 122. Maximum DNF[3:0] value to be compliant with Thd:dat(max)
Maximum DNF value
PCLK1 frequency

852/1747

Sm mode

Fm mode

2 <= FPCLK1 <= 5

2

0

5 < FPCLK1 <= 10

12

0

10 < FPCLK1 <= 20

15

1

20 < FPCLK1 <= 30

15

7

RM0090 Rev 17

RM0090

Inter-integrated circuit (I2C) interface
Table 122. Maximum DNF[3:0] value to be compliant with Thd:dat(max) (continued)
Maximum DNF value
PCLK1 frequency
Sm mode

Fm mode

30 < FPCLK1 <= 40

15

13

40 < FPCLK1 <= 50

15

15

Note:

For each frequency range, the constraint is given based on the worst case which is the
minimum frequency of the range. Greater DNF values can be used if the system can
support maximum hold time violation.

27.3.6

SDA/SCL line control
•

•

27.3.7

If clock stretching is enabled:
–

Transmitter mode: If TxE=1 and BTF=1: the interface holds the clock line low
before transmission to wait for the microcontroller to write the byte in the Data
Register (both buffer and shift register are empty).

–

Receiver mode: If RxNE=1 and BTF=1: the interface holds the clock line low after
reception to wait for the microcontroller to read the byte in the Data Register (both
buffer and shift register are full).

If clock stretching is disabled in Slave mode:
–

Overrun Error in case of RxNE=1 and no read of DR has been done before the
next byte is received. The last received byte is lost.

–

Underrun Error in case TxE=1 and no write into DR has been done before the next
byte must be transmitted. The same byte will be sent again.

–

Write Collision not managed.

SMBus
Introduction
The System Management Bus (SMBus) is a two-wire interface through which various
devices can communicate with each other and with the rest of the system. It is based on I2C
principles of operation. SMBus provides a control bus for system and power management
related tasks. A system may use SMBus to pass messages to and from devices instead of
toggling individual control lines.
The System Management Bus Specification refers to three types of devices. A slave is a
device that is receiving or responding to a command. A master is a device that issues
commands, generates the clocks, and terminates the transfer. A host is a specialized
master that provides the main interface to the system's CPU. A host must be a master-slave
and must support the SMBus host notify protocol. Only one host is allowed in a system.

Similarities between SMBus and I2C
•

2 wire bus protocol (1 Clk, 1 Data) + SMBus Alert line optional

•

Master-slave communication, Master provides clock

•

Multi master capability

•

SMBus data format similar to I2C 7-bit addressing format (Figure 238).

RM0090 Rev 17

853/1747
872

Inter-integrated circuit (I2C) interface

RM0090

Differences between SMBus and I2C
The following table describes the differences between SMBus and I2C.
Table 123. SMBus vs. I2C
I2C

SMBus
Max. speed 100 kHz

Max. speed 400 kHz

Min. clock speed 10 kHz

No minimum clock speed

35 ms clock low timeout

No timeout

Logic levels are fixed

Logic levels are VDD dependent

Different address types (reserved, dynamic etc.)

7-bit, 10-bit and general call slave address types

Different bus protocols (quick command, process
No bus protocols
call etc.)

SMBus application usage
With System Management Bus, a device can provide manufacturer information, tell the
system what its model/part number is, save its state for a suspend event, report different
types of errors, accept control parameters, and return its status. SMBus provides a control
bus for system and power management related tasks.

Device identification
Any device that exists on the System Management Bus as a slave has a unique address
called the Slave Address. For the list of reserved slave addresses, refer to the SMBus
specification version. 2.0 (http://smbus.org/).

Bus protocols
The SMBus specification supports up to nine bus protocols. For more details of these
protocols and SMBus address types, refer to SMBus specification version. 2.0. These
protocols should be implemented by the user software.

Address resolution protocol (ARP)
SMBus slave address conflicts can be resolved by dynamically assigning a new unique
address to each slave device. The Address Resolution Protocol (ARP) has the following
attributes:
•

Address assignment uses the standard SMBus physical layer arbitration mechanism

•

Assigned addresses remain constant while device power is applied; address retention
through device power loss is also allowed

•

No additional SMBus packet overhead is incurred after address assignment. (i.e.
subsequent accesses to assigned slave addresses have the same overhead as
accesses to fixed address devices.)

•

Any SMBus master can enumerate the bus

Unique device identifier (UDID)
In order to provide a mechanism to isolate each device for the purpose of address
assignment, each device must implement a unique device identifier (UDID).

854/1747

RM0090 Rev 17

RM0090

Inter-integrated circuit (I2C) interface
For the details on 128-bit UDID and more information on ARP, refer to SMBus specification
version 2.0.

SMBus alert mode
SMBus Alert is an optional signal with an interrupt line for devices that want to trade their
ability to master for a pin. SMBA is a wired-AND signal just as the SCL and SDA signals are.
SMBA is used in conjunction with the SMBus General Call Address. Messages invoked with
the SMBus are two bytes long.
A slave-only device can signal the host through SMBA that it wants to talk by setting ALERT
bit in I2C_CR1 register. The host processes the interrupt and simultaneously accesses all
SMBA devices through the Alert Response Address (known as ARA having a value 0001
100X). Only the device(s) which pulled SMBA low will acknowledge the Alert Response
Address. This status is identified using SMBALERT Status flag in I2C_SR1 register. The
host performs a modified Receive Byte operation. The 7 bit device address provided by the
slave transmit device is placed in the 7 most significant bits of the byte. The eighth bit can
be a zero or one.
If more than one device pulls SMBA low, the highest priority (lowest address) device will win
communication rights via standard arbitration during the slave address transfer. After
acknowledging the slave address the device must disengage its SMBA pull-down. If the
host still sees SMBA low when the message transfer is complete, it knows to read the ARA
again.
A host which does not implement the SMBA signal may periodically access the ARA.
For more details on SMBus Alert mode, refer to SMBus specification version 2.0.

Timeout error
There are differences in the timing specifications between I2C and SMBus.
SMBus defines a clock low timeout, TIMEOUT of 35 ms. Also SMBus specifies TLOW:
SEXT as the cumulative clock low extend time for a slave device. SMBus specifies TLOW:
MEXT as the cumulative clock low extend time for a master device. For more details on
these timeouts, refer to SMBus specification version 2.0.
The status flag Timeout or Tlow Error in I2C_SR1 shows the status of this feature.

How to use the interface in SMBus mode
To switch from I2C mode to SMBus mode, the following sequence should be performed.
•

Set the SMBus bit in the I2C_CR1 register

•

Configure the SMBTYPE and ENARP bits in the I2C_CR1 register as required for the
application

If you want to configure the device as a master, follow the Start condition generation
procedure in Section 27.3.3: I2C master mode. Otherwise, follow the sequence in
Section 27.3.2: I2C slave mode.
The application has to control the various SMBus protocols by software.
•

SMB Device Default Address acknowledged if ENARP=1 and SMBTYPE=0

•

SMB Host Header acknowledged if ENARP=1 and SMBTYPE=1

•

SMB Alert Response Address acknowledged if SMBALERT=1

RM0090 Rev 17

855/1747
872

Inter-integrated circuit (I2C) interface

27.3.8

RM0090

DMA requests
DMA requests (when enabled) are generated only for data transfer. DMA requests are
generated by Data Register becoming empty in transmission and Data Register becoming
full in reception. The DMA must be initialized and enabled before the I2C data transfer. The
DMAEN bit must be set in the I2C_CR2 register before the ADDR event. In master mode or
in slave mode when clock stretching is enabled, the DMAEN bit can also be set during the
ADDR event, before clearing the ADDR flag. The DMA request must be served before the
end of the current byte transfer. When the number of data transfers which has been
programmed for the corresponding DMA stream is reached, the DMA controller sends an
End of Transfer EOT signal to the I2C interface and generates a Transfer Complete interrupt
if enabled:
•

Master transmitter: In the interrupt routine after the EOT interrupt, disable DMA
requests then wait for a BTF event before programming the Stop condition.

•

Master receiver
–

When the number of bytes to be received is equal to or greater than two, the DMA
controller sends a hardware signal, EOT_1, corresponding to the last but one data
byte (number_of_bytes – 1). If, in the I2C_CR2 register, the LAST bit is set, I2C
automatically sends a NACK after the next byte following EOT_1. The user can
generate a Stop condition in the DMA Transfer Complete interrupt routine if
enabled.

–

When a single byte must be received: the NACK must be programmed during EV6
event, i.e. program ACK=0 when ADDR=1, before clearing ADDR flag. Then the
user can program the STOP condition either after clearing ADDR flag, or in the
DMA Transfer Complete interrupt routine.

Transmission using DMA
DMA mode can be enabled for transmission by setting the DMAEN bit in the I2C_CR2
register. Data will be loaded from a Memory area configured using the DMA peripheral (refer
to the DMA specification) to the I2C_DR register whenever the TxE bit is set. To map a DMA
stream x for I2C transmission (where x is the stream number), perform the following
sequence:
1.

Set the I2C_DR register address in the DMA_SxPAR register. The data will be moved
to this address from the memory after each TxE event.

2.

Set the memory address in the DMA_SxMA0R register (and in DMA_SxMA1R register
in the case of a bouble buffer mode). The data will be loaded into I2C_DR from this
memory after each TxE event.

3.

Configure the total number of bytes to be transferred in the DMA_SxNDTR register.
After each TxE event, this value will be decremented.

4.

Configure the DMA stream priority using the PL[0:1] bits in the DMA_SxCR register

5.

Set the DIR bit in the DMA_SxCR register and configure interrupts after half transfer or
full transfer depending on application requirements.

6.

Activate the stream by setting the EN bit in the DMA_SxCR register.

When the number of data transfers which has been programmed in the DMA Controller
registers is reached, the DMA controller sends an End of Transfer EOT/ EOT_1 signal to the
I2C interface and the DMA generates an interrupt, if enabled, on the DMA stream interrupt
vector.
Note:

856/1747

Do not enable the ITBUFEN bit in the I2C_CR2 register if DMA is used for transmission.

RM0090 Rev 17

RM0090

Inter-integrated circuit (I2C) interface

Reception using DMA
DMA mode can be enabled for reception by setting the DMAEN bit in the I2C_CR2 register.
Data will be loaded from the I2C_DR register to a Memory area configured using the DMA
peripheral (refer to the DMA specification) whenever a data byte is received. To map a DMA
stream x for I2C reception (where x is the stream number), perform the following sequence:
1.

Set the I2C_DR register address in DMA_SxPAR register. The data will be moved from
this address to the memory after each RxNE event.

2.

Set the memory address in the DMA_SxMA0R register (and in DMA_SxMA1R register
in the case of a bouble buffer mode). The data will be loaded from the I2C_DR register
to this memory area after each RxNE event.

3.

Configure the total number of bytes to be transferred in the DMA_SxNDTR register.
After each RxNE event, this value will be decremented.

4.

Configure the stream priority using the PL[0:1] bits in the DMA_SxCR register

5.

Reset the DIR bit and configure interrupts in the DMA_SxCR register after half transfer
or full transfer depending on application requirements.

6.

Activate the stream by setting the EN bit in the DMA_SxCR register.

When the number of data transfers which has been programmed in the DMA Controller
registers is reached, the DMA controller sends an End of Transfer EOT/ EOT_1 signal to the
I2C interface and DMA generates an interrupt, if enabled, on the DMA stream interrupt
vector.
Note:

Do not enable the ITBUFEN bit in the I2C_CR2 register if DMA is used for reception.

27.3.9

Packet error checking
A PEC calculator has been implemented to improve the reliability of communication. The
PEC is calculated by using the C(x) = x8 + x2 + x + 1 CRC-8 polynomial serially on each bit.
•

PEC calculation is enabled by setting the ENPEC bit in the I2C_CR1 register. PEC is a
CRC-8 calculated on all message bytes including addresses and R/W bits.
–

In transmission: set the PEC transfer bit in the I2C_CR1 register after the TxE
event corresponding to the last byte. The PEC will be transferred after the last
transmitted byte.

–

In reception: set the PEC bit in the I2C_CR1 register after the RxNE event
corresponding to the last byte so that the receiver sends a NACK if the next
received byte is not equal to the internally calculated PEC. In case of MasterReceiver, a NACK must follow the PEC whatever the check result. The PEC must

RM0090 Rev 17

857/1747
872

Inter-integrated circuit (I2C) interface

RM0090

be set before the ACK of the CRC reception in slave mode. It must be set when
the ACK is set low in master mode.

27.4

•

A PECERR error flag/interrupt is also available in the I2C_SR1 register.

•

If DMA and PEC calculation are both enabled:–

In transmission: when the I2C interface receives an EOT signal from the DMA
controller, it automatically sends a PEC after the last byte.

–

In reception: when the I2C interface receives an EOT_1 signal from the DMA
controller, it will automatically consider the next byte as a PEC and will check it. A
DMA request is generated after PEC reception.

•

To allow intermediate PEC transfers, a control bit is available in the I2C_CR2 register
(LAST bit) to determine if it is really the last DMA transfer or not. If it is the last DMA
request for a master receiver, a NACK is automatically sent after the last received byte.

•

PEC calculation is corrupted by an arbitration loss.

I2C interrupts
The table below gives the list of I2C interrupt requests.
Table 124. I2C Interrupt requests
Interrupt event

Event flag

Start bit sent (Master)

SB

Address sent (Master) or Address matched (Slave)

ADDR

10-bit header sent (Master)

ADD10

Stop received (Slave)

STOPF

Data byte transfer finished

BTF

Receive buffer not empty

RxNE

Transmit buffer empty

TxE

Bus error

BERR

Arbitration loss (Master)

ARLO

Acknowledge failure

ITEVFEN

ITEVFEN and ITBUFEN

AF

Overrun/Underrun

OVR

PEC error

PECERR

Timeout/Tlow error

TIMEOUT

SMBus Alert

Note:

Enable control bit

ITERREN

SMBALERT

SB, ADDR, ADD10, STOPF, BTF, RxNE and TxE are logically OR-ed on the same interrupt
channel.
BERR, ARLO, AF, OVR, PECERR, TIMEOUT and SMBALERT are logically OR-ed on the
same interrupt channel.

858/1747

RM0090 Rev 17

RM0090

Inter-integrated circuit (I2C) interface
Figure 245. I2C interrupt mapping diagram
SB

ITEVFEN

ADDR
ADD10
STOPF

it_event
BTF
TxE

ITBUFEN

RxNE
ITERREN

BERR
ARLO

it_error

AF
OVR
PECERR
TIMEOUT
SMBALERT

MS42082V1

RM0090 Rev 17

859/1747
872

Inter-integrated circuit (I2C) interface

RM0090

I2C debug mode

27.5

When the microcontroller enters the debug mode (Cortex®-M4 with FPU core halted), the
SMBUS timeout either continues to work normally or stops, depending on the
DBG_I2Cx_SMBUS_TIMEOUT configuration bits in the DBG module. For more details,
refer to Section 38.16.2: Debug support for timers, watchdog, bxCAN and I2C on
page 1701.

I2C registers

27.6

Refer to Section 1.1 for a list of abbreviations used in register descriptions.
The peripheral registers have to be accessed by half-words (16 bits) or words (32 bits).

I2C Control register 1 (I2C_CR1)

27.6.1

Address offset: 0x00
Reset value: 0x0000
15
SWRST

14
Res.

rw

13

12

11

10

ALERT

PEC

POS

ACK

rw

rw

rw

rw

9

8

STOP START
rw

rw

7
NO
STRETCH
rw

6

5

4

ENGC ENPEC ENARP
rw

rw

rw

3

2

1

0

SMB
TYPE

Res.

SMBU
S

PE

rw

rw

rw

Bit 15 SWRST: Software reset
When set, the I2C is under reset state. Before resetting this bit, make sure the I2C lines are
released and the bus is free.
0: I2C Peripheral not under reset
1: I2C Peripheral under reset state
Note: This bit can be used to reinitialize the peripheral after an error or a locked state. As an
example, if the BUSY bit is set and remains locked due to a glitch on the bus, the
SWRST bit can be used to exit from this state.
Bit 14 Reserved, must be kept at reset value
Bit 13 ALERT: SMBus alert
This bit is set and cleared by software, and cleared by hardware when PE=0.
0: Releases SMBA pin high. Alert Response Address Header followed by NACK.
1: Drives SMBA pin low. Alert Response Address Header followed by ACK.
Bit 12 PEC: Packet error checking
This bit is set and cleared by software, and cleared by hardware when PEC is transferred or
by a START or Stop condition or when PE=0.
0: No PEC transfer
1: PEC transfer (in Tx or Rx mode)
Note: PEC calculation is corrupted by an arbitration loss.

860/1747

RM0090 Rev 17

RM0090

Inter-integrated circuit (I2C) interface

Bit 11 POS: Acknowledge/PEC Position (for data reception)
This bit is set and cleared by software and cleared by hardware when PE=0.
0: ACK bit controls the (N)ACK of the current byte being received in the shift register. The
PEC bit indicates that current byte in shift register is a PEC.
1: ACK bit controls the (N)ACK of the next byte which will be received in the shift register.
The PEC bit indicates that the next byte in the shift register is a PEC
Note: The POS bit must be used only in 2-byte reception configuration in master mode. It
must be configured before data reception starts, as described in the 2-byte reception
procedure recommended in Section : Master receiver on page 849.
Bit 10 ACK: Acknowledge enable
This bit is set and cleared by software and cleared by hardware when PE=0.
0: No acknowledge returned
1: Acknowledge returned after a byte is received (matched address or data)
Bit 9 STOP: Stop generation
The bit is set and cleared by software, cleared by hardware when a Stop condition is
detected, set by hardware when a timeout error is detected.
In Master Mode:
0: No Stop generation.
1: Stop generation after the current byte transfer or after the current Start condition is sent.
In Slave mode:
0: No Stop generation.
1: Release the SCL and SDA lines after the current byte transfer.
Bit 8 START: Start generation
This bit is set and cleared by software and cleared by hardware when start is sent or PE=0.
In Master Mode:
0: No Start generation
1: Repeated start generation
In Slave mode:
0: No Start generation
1: Start generation when the bus is free
Bit 7 NOSTRETCH: Clock stretching disable (Slave mode)
This bit is used to disable clock stretching in slave mode when ADDR or BTF flag is set, until
it is reset by software.
0: Clock stretching enabled
1: Clock stretching disabled
Bit 6 ENGC: General call enable
0: General call disabled. Address 00h is NACKed.
1: General call enabled. Address 00h is ACKed.
Bit 5 ENPEC: PEC enable
0: PEC calculation disabled
1: PEC calculation enabled
Bit 4 ENARP: ARP enable
0: ARP disable
1: ARP enable
SMBus Device default address recognized if SMBTYPE=0
SMBus Host address recognized if SMBTYPE=1
Bit 3 SMBTYPE: SMBus type
0: SMBus Device
1: SMBus Host

RM0090 Rev 17

861/1747
872

Inter-integrated circuit (I2C) interface

RM0090

Bit 2 Reserved, must be kept at reset value
Bit 1 SMBUS: SMBus mode
0: I2C mode
1: SMBus mode
Bit 0 PE: Peripheral enable
0: Peripheral disable
1: Peripheral enable
Note: If this bit is reset while a communication is on going, the peripheral is disabled at the
end of the current communication, when back to IDLE state.
All bit resets due to PE=0 occur at the end of the communication.
In master mode, this bit must not be reset before the end of the communication.

Note:

When the STOP, START or PEC bit is set, the software must not perform any write access
to I2C_CR1 before this bit is cleared by hardware. Otherwise there is a risk of setting a
second STOP, START or PEC request.

27.6.2

I2C Control register 2 (I2C_CR2)
Address offset: 0x04
Reset value: 0x0000

15

14
Reserved

13

12
LAST
rw

11

10

9

8

DMAEN ITBUFEN ITEVTEN ITERREN
rw

rw

rw

7

6

5

4

2

1

0

rw

rw

FREQ[5:0]

Reserved

rw

3

rw

rw

rw

rw

Bits 15:13 Reserved, must be kept at reset value
Bit 12 LAST: DMA last transfer
0: Next DMA EOT is not the last transfer
1: Next DMA EOT is the last transfer
Note: This bit is used in master receiver mode to permit the generation of a NACK on the last
received data.
Bit 11 DMAEN: DMA requests enable
0: DMA requests disabled
1: DMA request enabled when TxE=1 or RxNE =1
Bit 10 ITBUFEN: Buffer interrupt enable
0: TxE = 1 or RxNE = 1 does not generate any interrupt.
1: TxE = 1 or RxNE = 1 generates Event Interrupt (whatever the state of DMAEN)

862/1747

RM0090 Rev 17

RM0090

Inter-integrated circuit (I2C) interface

Bit 9 ITEVTEN: Event interrupt enable
0: Event interrupt disabled
1: Event interrupt enabled
This interrupt is generated when:
–
SB = 1 (Master)
–
ADDR = 1 (Master/Slave)
–
ADD10= 1 (Master)
–
STOPF = 1 (Slave)
–
BTF = 1 with no TxE or RxNE event
–
TxE event to 1 if ITBUFEN = 1
–
RxNE event to 1if ITBUFEN = 1
Bit 8 ITERREN: Error interrupt enable
0: Error interrupt disabled
1: Error interrupt enabled
This interrupt is generated when:
–
BERR = 1
–
ARLO = 1
–
AF = 1
–
OVR = 1
–
PECERR = 1
–
TIMEOUT = 1
–
SMBALERT = 1
Bits 7:6 Reserved, must be kept at reset value
Bits 5:0 FREQ[5:0]: Peripheral clock frequency
The FREQ bits must be configured with the APB clock frequency value (I2C peripheral
connected to APB). The FREQ field is used by the peripheral to generate data setup and
hold times compliant with the I2C specifications. The minimum allowed frequency is 2 MHz,
the maximum frequency is limited by the maximum APB frequency and cannot exceed
50 MHz (peripheral intrinsic maximum limit).
0b000000: Not allowed
0b000001: Not allowed
0b000010: 2 MHz
...
0b110010: 50 MHz
Higher than 0b100100: Not allowed

RM0090 Rev 17

863/1747
872

Inter-integrated circuit (I2C) interface

RM0090

I2C Own address register 1 (I2C_OAR1)

27.6.3

Address offset: 0x08
Reset value: 0x0000
15

14

13

ADD
MODE

12

11

10

9

8

7

6

5

ADD[9:8]

Reserved

rw

rw

rw

4

3

2

1

ADD[7:1]
rw

rw

rw

rw

0
ADD0

rw

rw

rw

rw

Bit 15 ADDMODE Addressing mode (slave mode)
0: 7-bit slave address (10-bit address not acknowledged)
1: 10-bit slave address (7-bit address not acknowledged)
Bit 14
Bits 13:10

Should always be kept at 1 by software.
Reserved, must be kept at reset value

Bits 9:8 ADD[9:8]: Interface address
7-bit addressing mode: don’t care
10-bit addressing mode: bits9:8 of address
Bits 7:1 ADD[7:1]: Interface address
bits 7:1 of address
Bit 0 ADD0: Interface address
7-bit addressing mode: don’t care
10-bit addressing mode: bit 0 of address

I2C Own address register 2 (I2C_OAR2)

27.6.4

Address offset: 0x0C
Reset value: 0x0000
15

14

13

12

11

Reserved

10

9

8

7

6

5

rw

rw

rw

4

3

2

1

rw

rw

rw

ADD2[7:1]
rw

ENDUAL

Bits 15:8 Reserved, must be kept at reset value
Bits 7:1 ADD2[7:1]: Interface address
bits 7:1 of address in dual addressing mode
Bit 0 ENDUAL: Dual addressing mode enable
0: Only OAR1 is recognized in 7-bit addressing mode
1: Both OAR1 and OAR2 are recognized in 7-bit addressing mode

864/1747

RM0090 Rev 17

0

rw

RM0090

Inter-integrated circuit (I2C) interface

I2C Data register (I2C_DR)

27.6.5

Address offset: 0x10
Reset value: 0x0000
15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

DR[7:0]

Reserved

rw

rw

rw

rw

rw

Bits 15:8 Reserved, must be kept at reset value
Bits 7:0 DR[7:0] 8-bit data register
Byte received or to be transmitted to the bus.
– Transmitter mode: Byte transmission starts automatically when a byte is written in the DR
register. A continuous transmit stream can be maintained if the next data to be transmitted is
put in DR once the transmission is started (TxE=1)
– Receiver mode: Received byte is copied into DR (RxNE=1). A continuous transmit stream
can be maintained if DR is read before the next data byte is received (RxNE=1).
Note: In slave mode, the address is not copied into DR.
Write collision is not managed (DR can be written if TxE=0).
If an ARLO event occurs on ACK pulse, the received byte is not copied into DR and so
cannot be read.

I2C Status register 1 (I2C_SR1)

27.6.6

Address offset: 0x14
Reset value: 0x0000
15

14

SMB
ALERT

TIME
OUT

rc_w0

rc_w0

13

12

11

10

9

8

7

6

Res.

PEC
ERR

OVR

AF

ARLO

BERR

TxE

RxNE

rc_w0

rc_w0

rc_w0

rc_w0

rc_w0

r

r

5
Res.

4

3

STOPF ADD10
r

r

2

1

0

BTF

ADDR

SB

r

r

r

Bit 15 SMBALERT: SMBus alert
In SMBus host mode:
0: no SMBALERT
1: SMBALERT event occurred on pin
In SMBus slave mode:
0: no SMBALERT response address header
1: SMBALERT response address header to SMBALERT LOW received
– Cleared by software writing 0, or by hardware when PE=0.

RM0090 Rev 17

865/1747
872

Inter-integrated circuit (I2C) interface

RM0090

Bit 14 TIMEOUT: Timeout or Tlow error
0: No timeout error
1: SCL remained LOW for 25 ms (Timeout)
or
Master cumulative clock low extend time more than 10 ms (Tlow:mext)
or
Slave cumulative clock low extend time more than 25 ms (Tlow:sext)
– When set in slave mode: slave resets the communication and lines are released by
hardware
– When set in master mode: Stop condition sent by hardware
– Cleared by software writing 0, or by hardware when PE=0.
Note: This functionality is available only in SMBus mode.
Bit 13 Reserved, must be kept at reset value
Bit 12 PECERR: PEC Error in reception
0: no PEC error: receiver returns ACK after PEC reception (if ACK=1)
1: PEC error: receiver returns NACK after PEC reception (whatever ACK)
– Cleared by software writing 0, or by hardware when PE=0.
Note: When the received CRC is wrong, PECERR is not set in slave mode if the PEC control
bit is not set before the end of the CRC reception. Nevertheless, reading the PEC value
determines whether the received CRC is right or wrong.
Bit 11 OVR: Overrun/Underrun
0: No overrun/underrun
1: Overrun or underrun
– Set by hardware in slave mode when NOSTRETCH=1 and:
– In reception when a new byte is received (including ACK pulse) and the DR register has not
been read yet. New received byte is lost.
– In transmission when a new byte should be sent and the DR register has not been written
yet. The same byte is sent twice.
– Cleared by software writing 0, or by hardware when PE=0.
Note: If the DR write occurs very close to SCL rising edge, the sent data is unspecified and a
hold timing error occurs
Bit 10 AF: Acknowledge failure
0: No acknowledge failure
1: Acknowledge failure
– Set by hardware when no acknowledge is returned.
– Cleared by software writing 0, or by hardware when PE=0.
Bit 9 ARLO: Arbitration lost (master mode)
0: No Arbitration Lost detected
1: Arbitration Lost detected
Set by hardware when the interface loses the arbitration of the bus to another master
– Cleared by software writing 0, or by hardware when PE=0.
After an ARLO event the interface switches back automatically to Slave mode (MSL=0).
Note: In SMBUS, the arbitration on the data in slave mode occurs only during the data phase,
or the acknowledge transmission (not on the address acknowledge).

866/1747

RM0090 Rev 17

RM0090

Inter-integrated circuit (I2C) interface

Bit 8 BERR: Bus error
0: No misplaced Start or Stop condition
1: Misplaced Start or Stop condition
– Set by hardware when the interface detects an SDA rising or falling edge while SCL is high,
occurring in a non-valid position during a byte transfer.
– Cleared by software writing 0, or by hardware when PE=0.
Bit 7 TxE: Data register empty (transmitters)
0: Data register not empty
1: Data register empty
– Set when DR is empty in transmission. TxE is not set during address phase.
– Cleared by software writing to the DR register or by hardware after a start or a stop condition
or when PE=0.
TxE is not set if either a NACK is received, or if next byte to be transmitted is PEC (PEC=1)
Note: TxE is not cleared by writing the first data being transmitted, or by writing data when
BTF is set, as in both cases the data register is still empty.
Bit 6 RxNE: Data register not empty (receivers)
0: Data register empty
1: Data register not empty
– Set when data register is not empty in receiver mode. RxNE is not set during address phase.
– Cleared by software reading or writing the DR register or by hardware when PE=0.
RxNE is not set in case of ARLO event.
Note: RxNE is not cleared by reading data when BTF is set, as the data register is still full.
Bit 5 Reserved, must be kept at reset value
Bit 4 STOPF: Stop detection (slave mode)
0: No Stop condition detected
1: Stop condition detected
– Set by hardware when a Stop condition is detected on the bus by the slave after an
acknowledge (if ACK=1).
– Cleared by software reading the SR1 register followed by a write in the CR1 register, or by
hardware when PE=0
Note: The STOPF bit is not set after a NACK reception.
It is recommended to perform the complete clearing sequence (READ SR1 then
WRITE CR1) after the STOPF is set. Refer to Figure 242.
Bit 3 ADD10: 10-bit header sent (Master mode)
0: No ADD10 event occurred.
1: Master has sent first address byte (header).
– Set by hardware when the master has sent the first byte in 10-bit address mode.
– Cleared by software reading the SR1 register followed by a write in the DR register of the
second address byte, or by hardware when PE=0.
Note: ADD10 bit is not set after a NACK reception

RM0090 Rev 17

867/1747
872

Inter-integrated circuit (I2C) interface

RM0090

Bit 2 BTF: Byte transfer finished
0: Data byte transfer not done
1: Data byte transfer succeeded
– Set by hardware when NOSTRETCH=0 and:
– In reception when a new byte is received (including ACK pulse) and DR has not been read
yet (RxNE=1).
– In transmission when a new byte should be sent and DR has not been written yet (TxE=1).
– Cleared by software by either a read or write in the DR register or by hardware after a start
or a stop condition in transmission or when PE=0.
Note: The BTF bit is not set after a NACK reception
The BTF bit is not set if next byte to be transmitted is the PEC (TRA=1 in I2C_SR2
register and PEC=1 in I2C_CR1 register)
Bit 1 ADDR: Address sent (master mode)/matched (slave mode)
This bit is cleared by software reading SR1 register followed reading SR2, or by hardware
when PE=0.
Address matched (Slave)
0: Address mismatched or not received.
1: Received address matched.
– Set by hardware as soon as the received slave address matched with the OAR registers
content or a general call or a SMBus Device Default Address or SMBus Host or SMBus Alert
is recognized. (when enabled depending on configuration).
Note: In slave mode, it is recommended to perform the complete clearing sequence (READ
SR1 then READ SR2) after ADDR is set. Refer to Figure 242: Transfer sequence
diagram for slave receiver on page 846.
Address sent (Master)
0: No end of address transmission
1: End of address transmission
– For 10-bit addressing, the bit is set after the ACK of the 2nd byte.
– For 7-bit addressing, the bit is set after the ACK of the byte.
Note: ADDR is not set after a NACK reception
Bit 0 SB: Start bit (Master mode)
0: No Start condition
1: Start condition generated.
– Set when a Start condition generated.
– Cleared by software by reading the SR1 register followed by writing the DR register, or by
hardware when PE=0

I2C Status register 2 (I2C_SR2)

27.6.7

Address offset: 0x18
Reset value: 0x0000
Note:

15

Reading I2C_SR2 after reading I2C_SR1 clears the ADDR flag, even if the ADDR flag was
set after reading I2C_SR1. Consequently, I2C_SR2 must be read only when ADDR is found
set in I2C_SR1 or when the STOPF bit is cleared.
14

13

12

11

10

9

8

PEC[7:0]
r

868/1747

r

r

r

r

r

r

r

7

6

5

4

3

DUALF

SMB
HOST

SMBDE
FAULT

GEN
CALL

Res.

r

r

r

r

RM0090 Rev 17

2

1

0

TRA

BUSY

MSL

r

r

r

RM0090

Inter-integrated circuit (I2C) interface

Bits 15:8 PEC[7:0] Packet error checking register
This register contains the internal PEC when ENPEC=1.
Bit 7 DUALF: Dual flag (Slave mode)
0: Received address matched with OAR1
1: Received address matched with OAR2
– Cleared by hardware after a Stop condition or repeated Start condition, or when PE=0.
Bit 6 SMBHOST: SMBus host header (Slave mode)
0: No SMBus Host address
1: SMBus Host address received when SMBTYPE=1 and ENARP=1.
– Cleared by hardware after a Stop condition or repeated Start condition, or when PE=0.
Bit 5 SMBDEFAULT: SMBus device default address (Slave mode)
0: No SMBus Device Default address
1: SMBus Device Default address received when ENARP=1
– Cleared by hardware after a Stop condition or repeated Start condition, or when PE=0.
Bit 4 GENCALL: General call address (Slave mode)
0: No General Call
1: General Call Address received when ENGC=1
– Cleared by hardware after a Stop condition or repeated Start condition, or when PE=0.
Bit 3 Reserved, must be kept at reset value
Bit 2 TRA: Transmitter/receiver
0: Data bytes received
1: Data bytes transmitted
This bit is set depending on the R/W bit of the address byte, at the end of total address
phase.
It is also cleared by hardware after detection of Stop condition (STOPF=1), repeated Start
condition, loss of bus arbitration (ARLO=1), or when PE=0.
Bit 1 BUSY: Bus busy
0: No communication on the bus
1: Communication ongoing on the bus
– Set by hardware on detection of SDA or SCL low
– cleared by hardware on detection of a Stop condition.
It indicates a communication in progress on the bus. This information is still updated when
the interface is disabled (PE=0).
Bit 0 MSL: Master/slave
0: Slave Mode
1: Master Mode
– Set by hardware as soon as the interface is in Master mode (SB=1).
– Cleared by hardware after detecting a Stop condition on the bus or a loss of arbitration
(ARLO=1), or by hardware when PE=0.

Note:

Reading I2C_SR2 after reading I2C_SR1 clears the ADDR flag, even if the ADDR flag was
set after reading I2C_SR1. Consequently, I2C_SR2 must be read only when ADDR is found
set in I2C_SR1 or when the STOPF bit is cleared.

RM0090 Rev 17

869/1747
872

Inter-integrated circuit (I2C) interface

RM0090

I2C Clock control register (I2C_CCR)

27.6.8

Address offset: 0x1C
Reset value: 0x0000
Note:

fPCLK1 must be at least 2 MHz to achieve Sm mode I²C frequencies. It must be at least 4
MHz to achieve Fm mode I²C frequencies. It must be a multiple of 10MHz to reach the
400 kHz maximum I²C Fm mode clock.
The CCR register must be configured only when the I2C is disabled (PE = 0).

15

14

13

F/S

DUTY

rw

rw

12

Reserved

11

10

9

8

7

6

rw

rw

rw

rw

rw

rw

5

4

3

2

1

0

rw

rw

rw

rw

rw

CCR[11:0]
rw

Bit 15 F/S: I2C master mode selection
0: Sm mode I2C
1: Fm mode I2C
Bit 14 DUTY: Fm mode duty cycle
0: Fm mode tlow/thigh = 2
1: Fm mode tlow/thigh = 16/9 (see CCR)
Bits 13:12 Reserved, must be kept at reset value
Bits 11:0 CCR[11:0]: Clock control register in Fm/Sm mode (Master mode)
Controls the SCL clock in master mode.
Sm mode or SMBus:
Thigh = CCR * TPCLK1
Tlow = CCR * TPCLK1
Fm mode:
If DUTY = 0:
Thigh = CCR * TPCLK1
Tlow = 2 * CCR * TPCLK1
If DUTY = 1: (to reach 400 kHz)
Thigh = 9 * CCR * TPCLK1
Tlow = 16 * CCR * TPCLK1
For instance: in Sm mode, to generate a 100 kHz SCL frequency:
If FREQR = 08, TPCLK1 = 125 ns so CCR must be programmed with 0x28
(0x28 <=> 40d x 125 ns = 5000 ns.)
Note: The minimum allowed value is 0x04, except in FAST DUTY mode where the minimum
allowed value is 0x01
thigh = tr(SCL) + tw(SCLH). See device datasheet for the definitions of parameters.
tlow = tf(SCL) + tw(SCLL). See device datasheet for the definitions of parameters.
I2C communication speed, fSCL ~ 1/(thigh + tlow). The real frequency may differ due to
the analog noise filter input delay.
The CCR register must be configured only when the I2C is disabled (PE = 0).

870/1747

RM0090 Rev 17

RM0090

Inter-integrated circuit (I2C) interface

I2C TRISE register (I2C_TRISE)

27.6.9

Address offset: 0x20
Reset value: 0x0002
15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

TRISE[5:0]

Reserved

rw

rw

rw

rw

Bits 15:6 Reserved, must be kept at reset value
Bits 5:0 TRISE[5:0]: Maximum rise time in Fm/Sm mode (Master mode)
These bits should provide the maximum duration of the SCL feedback loop in master mode.
The purpose is to keep a stable SCL frequency whatever the SCL rising edge duration.
These bits must be programmed with the maximum SCL rise time given in the I2C bus
specification, incremented by 1.
For instance: in Sm mode, the maximum allowed SCL rise time is 1000 ns.
If, in the I2C_CR2 register, the value of FREQ[5:0] bits is equal to 0x08 and TPCLK1 = 125 ns
therefore the TRISE[5:0] bits must be programmed with 09h.
(1000 ns / 125 ns = 8 + 1)
The filter value can also be added to TRISE[5:0].
If the result is not an integer, TRISE[5:0] must be programmed with the integer part, in order
to respect the tHIGH parameter.
Note: TRISE[5:0] must be configured only when the I2C is disabled (PE = 0).

27.6.10

I2C FLTR register (I2C_FLTR)
Address offset: 0x24
Reset value: 0x0000
The I2C_FLTR is available on STM32F42xxx and STM32F43xxx only.

15

14

13

12

11

10

9

8

7

6

5

4

3

Reserved

rw

2

1

0

DNF[3:0]

ANOFF
rw

rw

rw

rw

Bits 15:5 Reserved, must be kept at reset value
Bit 4 ANOFF: Analog noise filter OFF
0: Analog noise filter enable
1: Analog noise filter disable
Note: ANOFF must be configured only when the I2C is disabled (PE = 0).
Bits 3:0 DNF[3:0]: Digital noise filter
These bits are used to configure the digital noise filter on SDA and SCL inputs. The digital filter
will suppress the spikes with a length of up to DNF[3:0] * TPCLK1.
0000: Digital noise filter disable
0001: Digital noise filter enabled and filtering capability up to 1* TPCLK1.
...
1111: Digital noise filter enabled and filtering capability up to 15* TPCLK1.
Note: DNF[3:0] must be configured only when the I2C is disabled (PE = 0). If the analog filter
is also enabled, the digital filter is added to the analog filter.

RM0090 Rev 17

871/1747
872

Inter-integrated circuit (I2C) interface

27.6.11

RM0090

I2C register map
The table below provides the I2C register map and reset values.

NOSTRETCH

ENGC

ENPEC

ENARP

SMBTYPE

0

0

0

0

0

0

0

0

ITBUFEN

ITEVTEN

ITERREN

0

0

0

0

0

0

0

I2C_TRISE

0

I2C_FLTR

0

0

0

ADD0

0

0

0

0

0

0

0

0

SB

TxE

RxNE

0

0

0

0

0

0

0

0

0

0

MSL

BERR
0

BTF

ARLO
0

ADDR

AF

Reserved

OVR

0

TRA

0

0

BUSY

0

0

GENCALL

0

0

PECERR

Reserved

TIMEOUT

0

0

0

0

0

0

0

0

CCR[11:0]
0

0

0

0

0

0

0

0

Reserved

0

0

0

Refer to Section 2.3: Memory map for the register boundary addresses table.

0

TRISE[5:0]

Reserved

RM0090 Rev 17

0

0

0

Reset value

872/1747

0

0

Reset value

0x24

0

0

0

Reserved

Reserved

0

0

0

PEC[7:0]

Reset value

0x20

0

F/S

I2C_CCR

0

Reserved

Reset value

0x1C

SMBALERT

Reserved

0

DR[7:0]

DUTY

I2C_SR2

0

Reserved

Reset value

0x18

0

0

ADD10

I2C_DR

I2C_SR1

0

ADD2[7:1]

Reset value

0x14

0

Reserved

Reset value

0x10

0

Reserved

I2C_OAR2

0

0

ADD[7:1]

SMBDEFAULT

0x0C

0

0

STOPF

Reset value

ADD[9:8]

Reserved

0

0

ANOFF

Reserved

0

DUALF

I2C_OAR1

0

FREQ[5:0]

SMBHOST

0x08

ADDMODE

Reset value

0

ENDUAL

Reserved

PE

START

0

SMBUS

STOP

0

Reserved

7
6
5
4
3
2
1
0

ACK

0

Reserved

8

PEC

I2C_CR2

POS

0x04

0

DMAEN

Reset value

ALERT

Reserved

LAST

I2C_CR1

SWRST

0x00

Register

Reserved

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9

Table 125. I2C register map and reset values

0

1

0

DNF[3:0]
0

0

0

0

RM0090

28

Serial peripheral interface (SPI)

Serial peripheral interface (SPI)
This section applies to the whole STM32F4xx family, unless otherwise specified.

28.1

SPI introduction
The SPI interface provides two main functions, supporting either the SPI protocol or the I2S
audio protocol. By default, it is the SPI function that is selected. It is possible to switch the
interface from SPI to I2S by software.
The serial peripheral interface (SPI) allows half/ full-duplex, synchronous, serial
communication with external devices. The interface can be configured as the master and in
this case it provides the communication clock (SCK) to the external slave device. The
interface is also capable of operating in multimaster configuration.
It may be used for a variety of purposes, including simplex synchronous transfers on two
lines with a possible bidirectional data line or reliable communication using CRC checking.
The I2S is also a synchronous serial communication interface. It can address four different
audio standards including the I2S Philips standard, the MSB- and LSB-justified standards,
and the PCM standard. It can operate as a slave or a master device in full-duplex mode
(using 4 pins) or in half-duplex mode (using 3 pins). Master clock can be provided by the
interface to an external slave component when the I2S is configured as the communication
master.

Warning:

Since some SPI1 and SPI3/I2S3 pins may be mapped onto
some pins used by the JTAG interface (SPI1_NSS onto JTDI,
SPI3_NSS/I2S3_WS onto JTDI and SPI3_SCK/I2S3_CK onto
JTDO), you may either:
– map SPI/I2S onto other pins
– disable the JTAG and use the SWD interface prior to
configuring the pins listed as SPI I/Os (when debugging the
application) or
– disable both JTAG/SWD interfaces (for standalone
applications).
For more information on the configuration of the JTAG/SWD
interface pins, please refer to Section 8.3.2: I/O pin
multiplexer and mapping.

RM0090 Rev 17

873/1747
925

Serial peripheral interface (SPI)

RM0090

28.2

SPI and I2S main features

28.2.1

SPI features

874/1747

•

Full-duplex synchronous transfers on three lines

•

Simplex synchronous transfers on two lines with or without a bidirectional data line

•

8- or 16-bit transfer frame format selection

•

Master or slave operation

•

Multimaster mode capability

•

8 master mode baud rate prescalers (fPCLK/2 max.)

•

Slave mode frequency (fPCLK/2 max)

•

Faster communication for both master and slave

•

NSS management by hardware or software for both master and slave: dynamic change
of master/slave operations

•

Programmable clock polarity and phase

•

Programmable data order with MSB-first or LSB-first shifting

•

Dedicated transmission and reception flags with interrupt capability

•

SPI bus busy status flag

•

SPI TI mode

•

Hardware CRC feature for reliable communication:
–

CRC value can be transmitted as last byte in Tx mode

–

Automatic CRC error checking for last received byte

•

Master mode fault, overrun and CRC error flags with interrupt capability

•

1-byte transmission and reception buffer with DMA capability: Tx and Rx requests

RM0090 Rev 17

RM0090

28.2.2

Serial peripheral interface (SPI)

I2S features
•

Full duplex communication

•

Half-duplex communication (only transmitter or receiver)

•

Master or slave operations

•

8-bit programmable linear prescaler to reach accurate audio sample frequencies (from
8 kHz to 192 kHz)

•

Data format may be 16-bit, 24-bit or 32-bit

•

Packet frame is fixed to 16-bit (16-bit data frame) or 32-bit (16-bit, 24-bit, 32-bit data
frame) by audio channel

•

Programmable clock polarity (steady state)

•

Underrun flag in slave transmission mode, overrun flag in reception mode (master and
slave), and Frame Error flag in reception and transmission mode (slave only)

•

16-bit register for transmission and reception with one data register for both channel
sides

•

Supported I2S protocols:
–

•

I2S Phillps standard

–

MSB-justified standard (left-justified)

–

LSB-justified standard (right-justified)

–

PCM standard (with short and long frame synchronization on 16-bit channel frame
or 16-bit data frame extended to 32-bit channel frame)

Data direction is always MSB first

•

DMA capability for transmission and reception (16-bit wide)

•

Master clock may be output to drive an external audio component. Ratio is fixed at
256 × FS (where FS is the audio sampling frequency)

•

Both I2S (I2S2 and I2S3) have a dedicated PLL (PLLI2S) to generate an even more
accurate clock.

•

I2S (I2S2 and I2S3) clock can be derived from an external clock mapped on the
I2S_CKIN pin.

RM0090 Rev 17

875/1747
925

Serial peripheral interface (SPI)

RM0090

28.3

SPI functional description

28.3.1

General description
The block diagram of the SPI is shown in Figure 246.
Figure 246. SPI block diagram
!DDRESS AND DATA BUS

2EAD
2X BUFFER
30)?#2

-/3)

48%
)%

3HIFT REGISTER
-)3/

,3" FIRST





48$- 28$33/% !%. !%.

30)?32
"39

4X BUFFER

28.% %22
)%
)%

-/$ #2#
%22
&

/62





48% 28.%

7RITE


#OMMUNICATION
CONTROL
3#+

"AUD RATE GENERATOR



"2;=
,3"
&)234

30% "2

"2 "2

-342 #0/, #0(!

30)?#2
-ASTER CONTROL LOGIC

")$) ")$)
-/$% /%

#2# #2#
28
%.
.EXT $&& /.,9 33-

33)

.33
AI

Usually, the SPI is connected to external devices through four pins:

876/1747

•

MISO: Master In / Slave Out data. This pin can be used to transmit data in slave mode
and receive data in master mode.

•

MOSI: Master Out / Slave In data. This pin can be used to transmit data in master
mode and receive data in slave mode.

•

SCK: Serial Clock output for SPI masters and input for SPI slaves.

•

NSS: Slave select. This is an optional pin to select a slave device. This pin acts as a
‘chip select’ to let the SPI master communicate with slaves individually and to avoid
contention on the data lines. Slave NSS inputs can be driven by standard IO ports on
the master device. The NSS pin may also be used as an output if enabled (SSOE bit)
and driven low if the SPI is in master configuration. In this manner, all NSS pins from
devices connected to the Master NSS pin see a low level and become slaves when
they are configured in NSS hardware mode. When configured in master mode with
NSS configured as an input (MSTR=1 and SSOE=0) and if NSS is pulled low, the SPI
enters the master mode fault state: the MSTR bit is automatically cleared and the
device is configured in slave mode (refer to Section 28.3.10: Error flags on page 897).

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)
A basic example of interconnections between a single master and a single slave is
illustrated in Figure 247.
Figure 247. Single master/ single slave application
0DVWHU
06%LW

6ODYH
/6%LW

06%LW

ELWVKLIWUHJLVWHU

63,FORFN
JHQHUDWRU

0,62

0,62

026,

026,

6&.

6&.

166 

9''

/6%LW

ELWVKLIWUHJLVWHU

166 
1RWXVHGLI166LVPDQDJHG
E\VRIWZDUH
DL

1. Here, the NSS pin is configured as an input.

The MOSI pins are connected together and the MISO pins are connected together. In this
way data is transferred serially between master and slave (most significant bit first).
The communication is always initiated by the master. When the master device transmits
data to a slave device via the MOSI pin, the slave device responds via the MISO pin. This
implies full-duplex communication with both data out and data in synchronized with the
same clock signal (which is provided by the master device via the SCK pin).

Slave select (NSS) pin management
Hardware or software slave select management can be set using the SSM bit in the
SPI_CR1 register.
•

Software NSS management (SSM = 1)
The slave select information is driven internally by the value of the SSI bit in the
SPI_CR1 register. The external NSS pin remains free for other application uses.

•

Hardware NSS management (SSM = 0)
Two configurations are possible depending on the NSS output configuration (SSOE bit
in register SPI_CR2).
–

NSS output enabled (SSM = 0, SSOE = 1)
This configuration is used only when the device operates in master mode. The
NSS signal is driven low when the master starts the communication and is kept
low until the SPI is disabled.

–

NSS output disabled (SSM = 0, SSOE = 0)
This configuration allows multimaster capability for devices operating in master
mode. For devices set as slave, the NSS pin acts as a classical NSS input: the
slave is selected when NSS is low and deselected when NSS high.

RM0090 Rev 17

877/1747
925

Serial peripheral interface (SPI)

RM0090

Clock phase and clock polarity
Four possible timing relationships may be chosen by software, using the CPOL and CPHA
bits in the SPI_CR1 register. The CPOL (clock polarity) bit controls the steady state value of
the clock when no data is being transferred. This bit affects both master and slave modes. If
CPOL is reset, the SCK pin has a low-level idle state. If CPOL is set, the SCK pin has a
high-level idle state.
If the CPHA (clock phase) bit is set, the second edge on the SCK pin (falling edge if the
CPOL bit is reset, rising edge if the CPOL bit is set) is the MSBit capture strobe. Data are
latched on the occurrence of the second clock transition. If the CPHA bit is reset, the first
edge on the SCK pin (falling edge if CPOL bit is set, rising edge if CPOL bit is reset) is the
MSBit capture strobe. Data are latched on the occurrence of the first clock transition.
The combination of the CPOL (clock polarity) and CPHA (clock phase) bits selects the data
capture clock edge.
Figure 248, shows an SPI transfer with the four combinations of the CPHA and CPOL bits.
The diagram may be interpreted as a master or slave timing diagram where the SCK pin,
the MISO pin, the MOSI pin are directly connected between the master and the slave
device.
Note:

Prior to changing the CPOL/CPHA bits the SPI must be disabled by resetting the SPE bit.
Master and slave must be programmed with the same timing mode.
The idle state of SCK must correspond to the polarity selected in the SPI_CR1 register (by
pulling up SCK if CPOL=1 or pulling down SCK if CPOL=0).
The Data Frame Format (8- or 16-bit) is selected through the DFF bit in SPI_CR1 register,
and determines the data length during transmission/reception.

878/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)
Figure 248. Data clock timing diagram
&3+$ 
&32/ 

&32/ 

026,

/6%LW

06%LW


0,62


















/6%LW

06%LW

166
WRVODYH
&DSWXUHVWUREH

&3+$ 
&32/ 

&32/ 

026,

 
0,62

/6%LW

06%LW










06%LW

/6%LW

166
WRVODYH

&DSWXUHVWUREH
DLG

1. These timings are shown with the LSBFIRST bit reset in the SPI_CR1 register.

Data frame format
Data can be shifted out either MSB-first or LSB-first depending on the value of the
LSBFIRST bit in the SPI_CR1 Register.
Each data frame is 8 or 16 bits long depending on the size of the data programmed using
the DFF bit in the SPI_CR1 register. The selected data frame format is applicable for
transmission and/or reception.

28.3.2

Configuring the SPI in slave mode
In the slave configuration, the serial clock is received on the SCK pin from the master
device. The value set in the BR[2:0] bits in the SPI_CR1 register, does not affect the data
transfer rate.

RM0090 Rev 17

879/1747
925

Serial peripheral interface (SPI)
Note:

RM0090

It is recommended to enable the SPI slave before the master sends the clock. If not,
undesired data transmission might occur. The data register of the slave needs to be ready
before the first edge of the communication clock or before the end of the ongoing
communication. It is mandatory to have the polarity of the communication clock set to the
steady state value before the slave and the master are enabled.
Follow the procedure below to configure the SPI in slave mode:

Procedure
1.

Set the DFF bit to define 8- or 16-bit data frame format

2.

Select the CPOL and CPHA bits to define one of the four relationships between the
data transfer and the serial clock (see Figure 248). For correct data transfer, the CPOL
and CPHA bits must be configured in the same way in the slave device and the master
device. This step is not required when the TI mode is selected through the FRF bit in
the SPI_CR2 register.

3.

The frame format (MSB-first or LSB-first depending on the value of the LSBFIRST bit in
the SPI_CR1 register) must be the same as the master device. This step is not
required when TI mode is selected.

4.

In Hardware mode (refer to Slave select (NSS) pin management), the NSS pin must be
connected to a low level signal during the complete byte transmit sequence. In NSS
software mode, set the SSM bit and clear the SSI bit in the SPI_CR1 register. This step
is not required when TI mode is selected.

5.

Set the FRF bit in the SPI_CR2 register to select the TI mode protocol for serial
communications.

6.

Clear the MSTR bit and set the SPE bit (both in the SPI_CR1 register) to assign the
pins to alternate functions.

In this configuration the MOSI pin is a data input and the MISO pin is a data output.

Transmit sequence
The data byte is parallel-loaded into the Tx buffer during a write cycle.
The transmit sequence begins when the slave device receives the clock signal and the most
significant bit of the data on its MOSI pin. The remaining bits (the 7 bits in 8-bit data frame
format, and the 15 bits in 16-bit data frame format) are loaded into the shift-register. The
TXE flag in the SPI_SR register is set on the transfer of data from the Tx Buffer to the shift
register and an interrupt is generated if the TXEIE bit in the SPI_CR2 register is set.

Receive sequence
For the receiver, when data transfer is complete:
•

The Data in shift register is transferred to Rx Buffer and the RXNE flag (SPI_SR
register) is set

•

An Interrupt is generated if the RXNEIE bit is set in the SPI_CR2 register.

After the last sampling clock edge the RXNE bit is set, a copy of the data byte received in
the shift register is moved to the Rx buffer. When the SPI_DR register is read, the SPI
peripheral returns this buffered value.
Clearing of the RXNE bit is performed by reading the SPI_DR register.

880/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)

SPI TI protocol in slave mode
In slave mode, the SPI interface is compatible with the TI protocol. The FRF bit of the
SPI_CR2 register can be used to configure the slave SPI serial communications to be
compliant with this protocol.
The clock polarity and phase are forced to conform to the TI protocol requirements whatever
the values set in the SPI_CR1 register. NSS management is also specific to the TI protocol
which makes the configuration of NSS management through the SPI_CR1 and SPI_CR2
registers (such as SSM, SSI, SSOE) transparent for the user.
In Slave mode (Figure 249: TI mode - Slave mode, single transfer and Figure 250: TI mode
- Slave mode, continuous transfer), the SPI baud rate prescaler is used to control the
moment when the MISO pin state changes to HI-Z. Any baud rate can be used thus allowing
to determine this moment with optimal flexibility. However, the baud rate is generally set to
the external master clock baud rate. The time for the MISO signal to become HI-Z (trelease)
depends on internal resynchronizations and on the baud rate value set in through BR[2:0] of
SPI_CR1 register. It is given by the formula:
t baud_rate
t baud_rate
---------------------------- + 4 × t
---------------------------- + 6 × t
<
t
<
pclk release
pclk
2
2
Note:

This feature is not available for Motorola SPI communications (FRF bit set to 0).
To detect TI frame errors in Slave transmitter only mode by using the Error interrupt (ERRIE
= 1), the SPI must be configured in 2-line unidirectional mode by setting BIDIMODE and
BIDIOE to 1 in the SPI_CR1 register. When BIDIMODE is set to 0, OVR is set to 1 because
the data register is never read and error interrupt are always generated, while when
BIDIMODE is set to 1, data are not received and OVR is never set.
Figure 249. TI mode - Slave mode, single transfer

.33
INPUT

TRIGGER
EDGE

SAMPLING TRIGGER
EDGE
EDGE

3#+
INPUT
-/3)
INPUT
-)3/
OUTPUT

SAMPLING
EDGE

TRIGGER
EDGE

T

$/.4#!2%

 OR 

SAMPLING
EDGE

2ELEASE

-3").

,3").

-3"/54

,3"/54

$/.4#!2%

AI

RM0090 Rev 17

881/1747
925

Serial peripheral interface (SPI)

RM0090

Figure 250. TI mode - Slave mode, continuous transfer

166
LQSXW

WULJJHU

VDPSOLQJ WULJJHU

VDPSOLQJ WULJJHU

VDPSOLQJ

6&.
LQSXW
026,
LQSXW

'217&$5(

0,62
RXWSXW

RU

06%,1

/6%,1

06%287

/6%287

06%,1

/6%,1

06%287

/6%287

)5$0(

'217&$5(

)5$0(

DL

28.3.3

Configuring the SPI in master mode
In the master configuration, the serial clock is generated on the SCK pin.

Procedure
1.

Select the BR[2:0] bits to define the serial clock baud rate (see SPI_CR1 register).

2.

Select the CPOL and CPHA bits to define one of the four relationships between the
data transfer and the serial clock (see Figure 248). This step is not required when the
TI mode is selected.

3.

Set the DFF bit to define 8- or 16-bit data frame format

4.

Configure the LSBFIRST bit in the SPI_CR1 register to define the frame format. This
step is not required when the TI mode is selected.

5.

If the NSS pin is required in input mode, in hardware mode, connect the NSS pin to a
high-level signal during the complete byte transmit sequence. In NSS software mode,
set the SSM and SSI bits in the SPI_CR1 register. If the NSS pin is required in output
mode, the SSOE bit only should be set. This step is not required when the TI mode is
selected.

6.

Set the FRF bit in SPI_CR2 to select the TI protocol for serial communications.

7.

The MSTR and SPE bits must be set (they remain set only if the NSS pin is connected
to a high-level signal).

In this configuration the MOSI pin is a data output and the MISO pin is a data input.

Transmit sequence
The transmit sequence begins when a byte is written in the Tx Buffer.
The data byte is parallel-loaded into the shift register (from the internal bus) during the first
bit transmission and then shifted out serially to the MOSI pin MSB first or LSB first
depending on the LSBFIRST bit in the SPI_CR1 register. The TXE flag is set on the transfer
of data from the Tx Buffer to the shift register and an interrupt is generated if the TXEIE bit in
the SPI_CR2 register is set.

882/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)

Receive sequence
For the receiver, when data transfer is complete:
•

The data in the shift register is transferred to the RX Buffer and the RXNE flag is set

•

An interrupt is generated if the RXNEIE bit is set in the SPI_CR2 register

At the last sampling clock edge the RXNE bit is set, a copy of the data byte received in the
shift register is moved to the Rx buffer. When the SPI_DR register is read, the SPI
peripheral returns this buffered value.
Clearing the RXNE bit is performed by reading the SPI_DR register.
A continuous transmit stream can be maintained if the next data to be transmitted is put in
the Tx buffer once the transmission is started. Note that TXE flag should be ‘1 before any
attempt to write the Tx buffer is made.
Note:

When a master is communicating with SPI slaves which need to be de-selected between
transmissions, the NSS pin must be configured as GPIO or another GPIO must be used and
toggled by software.

SPI TI protocol in master mode
In master mode, the SPI interface is compatible with the TI protocol. The FRF bit of the
SPI_CR2 register can be used to configure the master SPI serial communications to be
compliant with this protocol.
The clock polarity and phase are forced to conform to the TI protocol requirements whatever
the values set in the SPI_CR1 register. NSS management is also specific to the TI protocol
which makes the configuration of NSS management through the SPI_CR1 and SPI_CR2
registers (SSM, SSI, SSOE) transparent for the user.
Figure 251: TI mode - master mode, single transfer and Figure 252: TI mode - master mode,
continuous transfer) show the SPI master communication waveforms when the TI mode is
selected in master mode.
Figure 251. TI mode - master mode, single transfer
.33
OUTPUT

TRIGGER
EDGE

SAMPLING TRIGGER
EDGE
EDGE

SAMPLING
EDGE

TRIGGER
EDGE

SAMPLING
EDGE

3#+
OUTPUT
-/3)
INPUT

$/.4#!2%

-3").

,3").

-)3/
OUTPUT

 OR 

-3"/54

,3"/54

$/.4#!2%

AI

RM0090 Rev 17

883/1747
925

Serial peripheral interface (SPI)

RM0090

Figure 252. TI mode - master mode, continuous transfer

166
RXWSXW

WULJJHU

VDPSOLQJ WULJJHU VDPSOLQJ WULJJHU

VDPSOLQJ

6&.
RXWSXW
026,
RXWSXW

'217&$5(

0,62
LQWSXW

RU

06%287

/6%287

06%287

06%,1

/6%,1

06%,1

)5$0(

/6%287

/6%,1

'217&$5(

)5$0(

DL

28.3.4

Configuring the SPI for half-duplex communication
The SPI is capable of operating in half-duplex mode in 2 configurations.
•

1 clock and 1 bidirectional data wire

•

1 clock and 1 data wire (receive-only or transmit-only)

1 clock and 1 bidirectional data wire (BIDIMODE = 1)
This mode is enabled by setting the BIDIMODE bit in the SPI_CR1 register. In this mode
SCK is used for the clock and MOSI in master or MISO in slave mode is used for data
communication. The transfer direction (Input/Output) is selected by the BIDIOE bit in the
SPI_CR1 register. When this bit is 1, the data line is output otherwise it is input.

1 clock and 1 unidirectional data wire (BIDIMODE = 0)
In this mode, the application can use the SPI either in transmit-only mode or in receive-only
mode.
•

Transmit-only mode is similar to full-duplex mode (BIDIMODE=0, RXONLY=0): the
data are transmitted on the transmit pin (MOSI in master mode or MISO in slave mode)
and the receive pin (MISO in master mode or MOSI in slave mode) can be used as a
general-purpose IO. In this case, the application just needs to ignore the Rx buffer (if
the data register is read, it does not contain the received value).

•

In receive-only mode, the application can disable the SPI output function by setting the
RXONLY bit in the SPI_CR1 register. In this case, it frees the transmit IO pin (MOSI in
master mode or MISO in slave mode), so it can be used for other purposes.

To start the communication in receive-only mode, configure and enable the SPI:

884/1747

•

In master mode, the communication starts immediately and stops when the SPE bit is
cleared and the current reception stops. There is no need to read the BSY flag in this
mode. It is always set when an SPI communication is ongoing.

•

In slave mode, the SPI continues to receive as long as the NSS is pulled down (or the
SSI bit is cleared in NSS software mode) and the SCK is running.

RM0090 Rev 17

RM0090

28.3.5

Serial peripheral interface (SPI)

Data transmission and reception procedures
Rx and Tx buffers
In reception, data are received and then stored into an internal Rx buffer while In
transmission, data are first stored into an internal Tx buffer before being transmitted.
A read access of the SPI_DR register returns the Rx buffered value whereas a write access
to the SPI_DR stores the written data into the Tx buffer.

Start sequence in master mode
•

•

•

•

In full-duplex (BIDIMODE=0 and RXONLY=0)
–

The sequence begins when data are written into the SPI_DR register (Tx buffer).

–

The data are then parallel loaded from the Tx buffer into the 8-bit shift register
during the first bit transmission and then shifted out serially to the MOSI pin.

–

At the same time, the received data on the MISO pin is shifted in serially to the 8bit shift register and then parallel loaded into the SPI_DR register (Rx buffer).

In unidirectional receive-only mode (BIDIMODE=0 and RXONLY=1)
–

The sequence begins as soon as SPE=1

–

Only the receiver is activated and the received data on the MISO pin are shifted in
serially to the 8-bit shift register and then parallel loaded into the SPI_DR register
(Rx buffer).

In bidirectional mode, when transmitting (BIDIMODE=1 and BIDIOE=1)
–

The sequence begins when data are written into the SPI_DR register (Tx buffer).

–

The data are then parallel loaded from the Tx buffer into the 8-bit shift register
during the first bit transmission and then shifted out serially to the MOSI pin.

–

No data are received.

In bidirectional mode, when receiving (BIDIMODE=1 and BIDIOE=0)
–

The sequence begins as soon as SPE=1 and BIDIOE=0.

–

The received data on the MOSI pin are shifted in serially to the 8-bit shift register
and then parallel loaded into the SPI_DR register (Rx buffer).

–

The transmitter is not activated and no data are shifted out serially to the MOSI
pin.

Start sequence in slave mode
•

In full-duplex mode (BIDIMODE=0 and RXONLY=0)
–

The sequence begins when the slave device receives the clock signal and the first
bit of the data on its MOSI pin. The 7 remaining bits are loaded into the shift
register.

–

At the same time, the data are parallel loaded from the Tx buffer into the 8-bit shift
register during the first bit transmission, and then shifted out serially to the MISO

RM0090 Rev 17

885/1747
925

Serial peripheral interface (SPI)

RM0090

pin. The software must have written the data to be sent before the SPI master
device initiates the transfer.
•

•

•

In unidirectional receive-only mode (BIDIMODE=0 and RXONLY=1)
–

The sequence begins when the slave device receives the clock signal and the first
bit of the data on its MOSI pin. The 7 remaining bits are loaded into the shift
register.

–

The transmitter is not activated and no data are shifted out serially to the MISO
pin.

In bidirectional mode, when transmitting (BIDIMODE=1 and BIDIOE=1)
–

The sequence begins when the slave device receives the clock signal and the first
bit in the Tx buffer is transmitted on the MISO pin.

–

The data are then parallel loaded from the Tx buffer into the 8-bit shift register
during the first bit transmission and then shifted out serially to the MISO pin. The
software must have written the data to be sent before the SPI master device
initiates the transfer.

–

No data are received.

In bidirectional mode, when receiving (BIDIMODE=1 and BIDIOE=0)
–

The sequence begins when the slave device receives the clock signal and the first
bit of the data on its MISO pin.

–

The received data on the MISO pin are shifted in serially to the 8-bit shift register
and then parallel loaded into the SPI_DR register (Rx buffer).

–

The transmitter is not activated and no data are shifted out serially to the MISO
pin.

Handling data transmission and reception
The TXE flag (Tx buffer empty) is set when the data are transferred from the Tx buffer to the
shift register. It indicates that the internal Tx buffer is ready to be loaded with the next data.
An interrupt can be generated if the TXEIE bit in the SPI_CR2 register is set. Clearing the
TXE bit is performed by writing to the SPI_DR register.
Note:

The software must ensure that the TXE flag is set to 1 before attempting to write to the Tx
buffer. Otherwise, it overwrites the data previously written to the Tx buffer.
The RXNE flag (Rx buffer not empty) is set on the last sampling clock edge, when the data
are transferred from the shift register to the Rx buffer. It indicates that data are ready to be
read from the SPI_DR register. An interrupt can be generated if the RXNEIE bit in the
SPI_CR2 register is set. Clearing the RXNE bit is performed by reading the SPI_DR
register.
For some configurations, the BSY flag can be used during the last data transfer to wait until
the completion of the transfer.
Full-duplex transmit and receive procedure in master or slave mode (BIDIMODE=0 and
RXONLY=0)
The software has to follow this procedure to transmit and receive data (see Figure 253 and
Figure 254):

886/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)
1.

Enable the SPI by setting the SPE bit to 1.

2.

Write the first data item to be transmitted into the SPI_DR register (this clears the TXE
flag).

3.

Wait until TXE=1 and write the second data item to be transmitted. Then wait until
RXNE=1 and read the SPI_DR to get the first received data item (this clears the RXNE
bit). Repeat this operation for each data item to be transmitted/received until the n–1
received data.

4.

Wait until RXNE=1 and read the last received data.

5.

Wait until TXE=1 and then wait until BSY=0 before disabling the SPI.

This procedure can also be implemented using dedicated interrupt subroutines launched at
each rising edges of the RXNE or TXE flag.
Figure 253. TXE/RXNE/BSY behavior in Master / full-duplex mode (BIDIMODE=0 and
RXONLY=0) in case of continuous transfers
([DPSOHLQ0DVWHUPRGHZLWK&32/ &3+$ 
6&.
'$7$ [)
0,62026, RXW

VHWE\KDUGZDUH
FOHDUHGE\VRIWZDUH

7;(IODJ
7[EXIIHU
ZULWH63,B'5

%6<IODJ

'$7$  [)

[)

VHWE\KDUGZDUH
FOHDUHGE\VRIWZDUH

[)

VHWE\KDUGZDUH

UHVHWE\KDUGZDUH
'$7$ [$

'$7$ [$

E E E E E E E E E E E E E E E E E E E E E E E E

5;1(IODJ

VHWE\KDUGZDUH

5[EXIIHU
UHDG63,B'5

VRIWZDUH
ZULWHV[)
LQWR63,B'5

VHWE\KDUGZDUH

[)

'$7$  [$
0,62026, LQ

'$7$ [)

E E E E E E E E E E E E E E E E E E E E E E E E

VRIWZDUHZDLWV
XQWLO7;( DQG
ZULWHV[)LQWR
63,B'5

FOHDUHGE\VRIWZDUH

[$

VRIWZDUHZDLWV
XQWLO5;1( 
DQGUHDGV[$
IURP63,B'5

VRIWZDUHZDLWV
XQWLO7;( DQG
ZULWHV[)LQWR
63,B'5

[$

VRIWZDUHZDLWV
XQWLO5;1( 
DQGUHDGV[$
IURP63,B'5

[$

VRIWZDUHZDLWV
XQWLO5;1( 
DQGUHDGV[$
IURP63,B'5
DL

RM0090 Rev 17

887/1747
925

Serial peripheral interface (SPI)

RM0090

Figure 254. TXE/RXNE/BSY behavior in Slave / full-duplex mode (BIDIMODE=0,
RXONLY=0) in case of continuous transfers
([DPSOHLQ6ODYHPRGHZLWK&32/ &3+$ 
6&.
'$7$ [)

'$7$ [)
0,62026, RXW

VHWE\KDUGZDUH
FOHDUHGE\VRIWZDUH

7;(IODJ

7[EXIIHU
ZULWHWR63,B'5

[)

VHWE\KDUGZDUH
FOHDUHGE\VRIWZDUH

[)

VHWE\KDUGZDUH

[)

VHWE\FOHDUHGE\VRIWZDUH

%6<IODJ

UHVHWE\KDUGZDUH
'$7$ [$

'$7$  [$
0,62026, LQ

'$7$ [$

E E E E E E E E E E E E E E E E E E E E E E E E

VHWE\KDUGZDUH

5;1(IODJ
5[EXIIHU
UHDGIURP63,B'5

VRIWZDUH
ZULWHV[)
LQWR63,B'5

'$7$ [)

E E E E E E E E E E E E E E E E E E E E E E E E

FOHDUHGE\VRIWZDUH

[$

VRIWZDUHZDLWV
XQWLO7;( DQG
ZULWHV[)LQWR
63,B'5

VRIWZDUHZDLWV
XQWLO5;1( 
DQGUHDGV[$
IURP63,B'5

VRIWZDUHZDLWV
XQWLO7;( DQG
ZULWHV[)LQWR
63,B'5

[$

VRIWZDUHZDLWV
XQWLO5;1( 
DQGUHDGV[$
IURP63,B'5

[$

VRIWZDUHZDLWV
XQWLO5;1( 
DQGUHDGV[$
IURP63,B'5
DL

Transmit-only procedure (BIDIMODE=0 RXONLY=0)
In this mode, the procedure can be reduced as described below and the BSY bit can be
used to wait until the completion of the transmission (see Figure 255 and Figure 256).
1.

Enable the SPI by setting the SPE bit to 1.

2.

Write the first data item to send into the SPI_DR register (this clears the TXE bit).

3.

Wait until TXE=1 and write the next data item to be transmitted. Repeat this step for
each data item to be transmitted.

4.

After writing the last data item into the SPI_DR register, wait until TXE=1, then wait until
BSY=0, this indicates that the transmission of the last data is complete.

This procedure can be also implemented using dedicated interrupt subroutines launched at
each rising edge of the TXE flag.
Note:

During discontinuous communications, there is a 2 APB clock period delay between the
write operation to SPI_DR and the BSY bit setting. As a consequence, in transmit-only
mode, it is mandatory to wait first until TXE is set and then until BSY is cleared after writing
the last data.
After transmitting two data items in transmit-only mode, the OVR flag is set in the SPI_SR
register since the received data are never read.

888/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)

Figure 255. TXE/BSY behavior in Master transmit-only mode (BIDIMODE=0 and RXONLY=0)
in case of continuous transfers
([DPSOHLQ0DVWHUPRGHZLWK&32/ &3+$ 
6&.
'$7$ [)

'$7$ [)

0,62026, RXW

VHWE\KDUGZDUH
FOHDUHGE\VRIWZDUH

7;(IODJ
7[EXIIHU
ZULWHWR63,B'5
%6<IODJ

'$7$ [)

E E E E E E E E E E E E E E E E E E E E E E E E

[)

VHWE\KDUGZDUH
FOHDUHGE\VRIWZDUH

[)

VHWE\KDUGZDUH

[)

UHVHWE\KDUGZDUH

VHWE\KDUGZDUH

VRIWZDUHZULWHV VRIWZDUHZDLWV VRIWZDUHZDLWV
[)LQWR
XQWLO7;( DQG XQWLO7;( DQG
63,B'5
ZULWHV[)LQWR ZULWHV[)LQWR
63,B'5
63,B'5

VRIWZDUHZDLWVXQWLO7;( 

VRIWZDUHZDLWVXQWLO%6< 
DL

Figure 256. TXE/BSY in Slave transmit-only mode (BIDIMODE=0 and RXONLY=0) in case of
continuous transfers
([DPSOHLQVODYHPRGHZLWK&32/ &3+$ 
6&.
'$7$ [)

0,62026, RXW

VHWE\KDUGZDUH
FOHDUHGE\VRIWZDUH

7;(IODJ
7[EXIIHU
ZULWHWR63,B'5
%6<IODJ

'$7$ [)

'$7$ [)

E E E E E E E E E E E E E E E E E E E E E E E E

[)

VHWE\KDUGZDUH
FOHDUHGE\VRIWZDUH

[)

VHWE\KDUGZDUH

[)

VHWE\KDUGZDUH

VRIWZDUHZULWHV VRIWZDUHZDLWV
[)LQWR
XQWLO7;( DQG
63,B'5
ZULWHV[)LQWR
63,B'5

UHVHWE\KDUGZDUH
VRIWZDUHZDLWV
XQWLO7;( DQG
ZULWHV[)LQWR
63,B'5

VRIWZDUHZDLWVXQWLO7;( 

VRIWZDUHZDLWVXQWLO%6< 
DL

Bidirectional transmit procedure (BIDIMODE=1 and BIDIOE=1)
In this mode, the procedure is similar to the procedure in Transmit-only mode except that
the BIDIMODE and BIDIOE bits both have to be set in the SPI_CR2 register before enabling
the SPI.

Unidirectional receive-only procedure (BIDIMODE=0 and RXONLY=1)
In this mode, the procedure can be reduced as described below (see Figure 257):

RM0090 Rev 17

889/1747
925

Serial peripheral interface (SPI)

RM0090

1.

Set the RXONLY bit in the SPI_CR1 register.

2.

Enable the SPI by setting the SPE bit to 1:

3.

a)

In master mode, this immediately activates the generation of the SCK clock, and
data are serially received until the SPI is disabled (SPE=0).

b)

In slave mode, data are received when the SPI master device drives NSS low and
generates the SCK clock.

Wait until RXNE=1 and read the SPI_DR register to get the received data (this clears
the RXNE bit). Repeat this operation for each data item to be received.

This procedure can also be implemented using dedicated interrupt subroutines launched at
each rising edge of the RXNE flag.
Note:

If it is required to disable the SPI after the last transfer, follow the recommendation
described in Section 28.3.8: Disabling the SPI.
Figure 257. RXNE behavior in receive-only mode (BIDIRMODE=0 and RXONLY=1)
in case of continuous transfers

([DPSOHZLWK&32/ &3+$ 5;21/< 
6&.
'$7$ [$

0,62026, LQ

5;1(IODJ

'$7$ [$

'$7$ [$

E E E E E E E E E E E E E E E E E E E E E E E E
VHWE\KDUGZDUH

5[EXIIHU
UHDGIURP63,B'5

FOHDUHGE\VRIWZDUH

[$

VRIWZDUHZDLWVXQWLO5;1( 
DQGUHDGV[$IURP63,B'5

VRIWZDUHZDLWVXQWLO5;1( 
DQGUHDGV[$IURP63,B'5

[$

[$

VRIWZDUHZDLWVXQWLO5;1( 
DQGUHDGV[$IURP63,B'5
DL

Bidirectional receive procedure (BIDIMODE=1 and BIDIOE=0)
In this mode, the procedure is similar to the Receive-only mode procedure except that the
BIDIMODE bit has to be set and the BIDIOE bit cleared in the SPI_CR2 register before
enabling the SPI.

Continuous and discontinuous transfers
When transmitting data in master mode, if the software is fast enough to detect each rising
edge of TXE (or TXE interrupt) and to immediately write to the SPI_DR register before the
ongoing data transfer is complete, the communication is said to be continuous. In this case,
there is no discontinuity in the generation of the SPI clock between each data item and the
BSY bit is never cleared between each data transfer.
On the contrary, if the software is not fast enough, this can lead to some discontinuities in
the communication. In this case, the BSY bit is cleared between each data transmission
(see Figure 258).
In Master receive-only mode (RXONLY=1), the communication is always continuous and
the BSY flag is always read at 1.

890/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)
In slave mode, the continuity of the communication is decided by the SPI master device. In
any case, even if the communication is continuous, the BSY flag goes low between each
transfer for a minimum duration of one SPI clock cycle (see Figure 256).
Figure 258. TXE/BSY behavior when transmitting (BIDIRMODE=0 and RXONLY=0)
in case of discontinuous transfers

([DPSOHZLWK&32/ &3+$ 
6&.
'$7$ [)

026, RXW

E E E E E E E E

'$7$ [)

'$7$ [)

E E E E E E E E

E E E E E E E E

7;(IODJ
7[EXIIHU
ZULWHWR63,B'5

[)

[)

[)

%6<IODJ

VRIWZDUHZULWHV[) VRIWZDUHZDLWVXQWLO7;( EXWLV VRIWZDUHZDLWVXQWLO7;( EXW
LQWR63,B'5
ODWHWRZULWH[)LQWR63,B'5
LVODWHWRZULWH[)LQWR
63,B'5

VRIWZDUHZDLWV
XQWLO7;( 

VRIWZDUHZDLWVXQWLO%6< 

DL

28.3.6

CRC calculation
A CRC calculator has been implemented for communication reliability. Separate CRC
calculators are implemented for transmitted data and received data. The CRC is calculated
using a programmable polynomial serially on each bit. It is calculated on the sampling clock
edge defined by the CPHA and CPOL bits in the SPI_CR1 register.

Note:

This SPI offers two kinds of CRC calculation standard which depend directly on the data
frame format selected for the transmission and/or reception: 8-bit data (CR8) and 16-bit data
(CRC16).
CRC calculation is enabled by setting the CRCEN bit in the SPI_CR1 register. This action
resets the CRC registers (SPI_RXCRCR and SPI_TXCRCR). In full duplex or transmitter
only mode, when the transfers are managed by the software (CPU mode), it is necessary to
write the bit CRCNEXT immediately after the last data to be transferred is written to the
SPI_DR. At the end of this last data transfer, the SPI_TXCRCR value is transmitted.
In receive only mode and when the transfers are managed by software (CPU mode), it is
necessary to write the CRCNEXT bit after the second last data has been received. The CRC
is received just after the last data reception and the CRC check is then performed.
At the end of data and CRC transfers, the CRCERR flag in the SPI_SR register is set if
corruption occurs during the transfer.
If data are present in the TX buffer, the CRC value is transmitted only after the transmission
of the data byte. During CRC transmission, the CRC calculator is switched off and the
register value remains unchanged.
SPI communication using the CRC is possible through the following procedure:

RM0090 Rev 17

891/1747
925

Serial peripheral interface (SPI)
1.

Program the CPOL, CPHA, LSBFirst, BR, SSM, SSI and MSTR values.

2.

Program the polynomial in the SPI_CRCPR register.

3.

Enable the CRC calculation by setting the CRCEN bit in the SPI_CR1 register. This
also clears the SPI_RXCRCR and SPI_TXCRCR registers.

4.

Enable the SPI by setting the SPE bit in the SPI_CR1 register.

5.

Start the communication and sustain the communication until all but one byte or halfword have been transmitted or received.

6.

Note:

RM0090

–

In full duplex or transmitter-only mode, when the transfers are managed by
software, when writing the last byte or half word to the Tx buffer, set the
CRCNEXT bit in the SPI_CR1 register to indicate that the CRC will be transmitted
after the transmission of the last byte.

–

In receiver only mode, set the bit CRCNEXT just after the reception of the second
to last data to prepare the SPI to enter in CRC Phase at the end of the reception of
the last data. CRC calculation is frozen during the CRC transfer.

After the transfer of the last byte or half word, the SPI enters the CRC transfer and
check phase. In full duplex mode or receiver-only mode, the received CRC is
compared to the SPI_RXCRCR value. If the value does not match, the CRCERR flag in
SPI_SR is set and an interrupt can be generated when the ERRIE bit in the SPI_CR2
register is set.

When the SPI is in slave mode, be careful to enable CRC calculation only when the clock is
stable, that is, when the clock is in the steady state. If not, a wrong CRC calculation may be
done. In fact, the CRC is sensitive to the SCK slave input clock as soon as CRCEN is set,
and this, whatever the value of the SPE bit.
With high bitrate frequencies, be careful when transmitting the CRC. As the number of used
CPU cycles has to be as low as possible in the CRC transfer phase, it is forbidden to call
software functions in the CRC transmission sequence to avoid errors in the last data and
CRC reception. In fact, CRCNEXT bit has to be written before the end of the
transmission/reception of the last data.
For high bit rate frequencies, it is advised to use the DMA mode to avoid the degradation of
the SPI speed performance due to CPU accesses impacting the SPI bandwidth.
When the devices are configured as slaves and the NSS hardware mode is used, the NSS
pin needs to be kept low between the data phase and the CRC phase.
When the SPI is configured in slave mode with the CRC feature enabled, CRC calculation
takes place even if a high level is applied on the NSS pin. This may happen for example in
case of a multislave environment where the communication master addresses slaves
alternately.
Between a slave deselection (high level on NSS) and a new slave selection (low level on
NSS), the CRC value should be cleared on both master and slave sides in order to
resynchronize the master and slave for their respective CRC calculation.
To clear the CRC, follow the procedure below:

892/1747

1.

Disable SPI (SPE = 0)

2.

Clear the CRCEN bit

3.

Set the CRCEN bit

4.

Enable the SPI (SPE = 1)

RM0090 Rev 17

RM0090

28.3.7

Serial peripheral interface (SPI)

Status flags
Four status flags are provided for the application to completely monitor the state of the SPI
bus.

Tx buffer empty flag (TXE)
When it is set, this flag indicates that the Tx buffer is empty and the next data to be
transmitted can be loaded into the buffer. The TXE flag is cleared when writing to the
SPI_DR register.

Rx buffer not empty (RXNE)
When set, this flag indicates that there are valid received data in the Rx buffer. It is cleared
when SPI_DR is read.

BUSY flag
This BSY flag is set and cleared by hardware (writing to this flag has no effect). The BSY
flag indicates the state of the communication layer of the SPI.
When BSY is set, it indicates that the SPI is busy communicating. There is one exception in
master mode / bidirectional receive mode (MSTR=1 and BDM=1 and BDOE=0) where the
BSY flag is kept low during reception.
The BSY flag is useful to detect the end of a transfer if the software wants to disable the SPI
and enter Halt mode (or disable the peripheral clock). This avoids corrupting the last
transfer. For this, the procedure described below must be strictly respected.
The BSY flag is also useful to avoid write collisions in a multimaster system.
The BSY flag is set when a transfer starts, with the exception of master mode / bidirectional
receive mode (MSTR=1 and BDM=1 and BDOE=0).
It is cleared:
•

when a transfer is finished (except in master mode if the communication is continuous)

•

when the SPI is disabled

•

when a master mode fault occurs (MODF=1)

When communication is not continuous, the BSY flag is low between each communication.
When communication is continuous:

Note:

•

in master mode, the BSY flag is kept high during all the transfers

•

in slave mode, the BSY flag goes low for one SPI clock cycle between each transfer

Do not use the BSY flag to handle each data transmission or reception. It is better to use the
TXE and RXNE flags instead.

RM0090 Rev 17

893/1747
925

Serial peripheral interface (SPI)

28.3.8

RM0090

Disabling the SPI
When a transfer is terminated, the application can stop the communication by disabling the
SPI peripheral. This is done by clearing the SPE bit.
For some configurations, disabling the SPI and entering the Halt mode while a transfer is
ongoing can cause the current transfer to be corrupted and/or the BSY flag might become
unreliable.
To avoid any of those effects, it is recommended to respect the following procedure when
disabling the SPI:

In master or slave full-duplex mode (BIDIMODE=0, RXONLY=0)
1.

Wait until RXNE=1 to receive the last data

2.

Wait until TXE=1

3.

Then wait until BSY=0

4.

Disable the SPI (SPE=0) and, eventually, enter the Halt mode (or disable the peripheral
clock)

In master or slave unidirectional transmit-only mode (BIDIMODE=0,
RXONLY=0) or bidirectional transmit mode (BIDIMODE=1, BIDIOE=1)
After the last data is written into the SPI_DR register:
1.

Wait until TXE=1

2.

Then wait until BSY=0

3.

Disable the SPI (SPE=0) and, eventually, enter the Halt mode (or disable the peripheral
clock)

In master unidirectional receive-only mode (MSTR=1, BIDIMODE=0,
RXONLY=1) or bidirectional receive mode (MSTR=1, BIDIMODE=1, BIDIOE=0)
This case must be managed in a particular way to ensure that the SPI does not initiate a
new transfer. The sequence below is valid only for SPI Motorola configuration (FRF bit set to
0):
1.

Wait for the second to last occurrence of RXNE=1 (n–1)

2.

Then wait for one SPI clock cycle (using a software loop) before disabling the SPI
(SPE=0)

3.

Then wait for the last RXNE=1 before entering the Halt mode (or disabling the
peripheral clock)

When the SPI is configured in TI mode (Bit FRF set to 1), the following procedure has to be
respected to avoid generating an undesired pulse on NSS when the SPI is disabled:

Note:

894/1747

1.

Wait for the second to last occurrence of RXNE = 1 (n-1).

2.

Disable the SPI (SPE = 0) in the following window frame using a software loop:
–

After at least one SPI clock cycle,

–

Before the beginning of the LSB data transfer.

In master bidirectional receive mode (MSTR=1 and BDM=1 and BDOE=0), the BSY flag is
kept low during transfers.

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)

In slave receive-only mode (MSTR=0, BIDIMODE=0, RXONLY=1) or
bidirectional receive mode (MSTR=0, BIDIMODE=1, BIDOE=0)

28.3.9

1.

You can disable the SPI (write SPE=1) at any time: the current transfer will complete
before the SPI is effectively disabled

2.

Then, if you want to enter the Halt mode, you must first wait until BSY = 0 before
entering the Halt mode (or disabling the peripheral clock).

SPI communication using DMA (direct memory addressing)
To operate at its maximum speed, the SPI needs to be fed with the data for transmission
and the data received on the Rx buffer should be read to avoid overrun. To facilitate the
transfers, the SPI features a DMA capability implementing a simple request/acknowledge
protocol.
A DMA access is requested when the enable bit in the SPI_CR2 register is enabled.
Separate requests must be issued to the Tx and Rx buffers (see Figure 259 and
Figure 260):
•

In transmission, a DMA request is issued each time TXE is set to 1. The DMA then
writes to the SPI_DR register (this clears the TXE flag).

•

In reception, a DMA request is issued each time RXNE is set to 1. The DMA then reads
the SPI_DR register (this clears the RXNE flag).

When the SPI is used only to transmit data, it is possible to enable only the SPI Tx DMA
channel. In this case, the OVR flag is set because the data received are not read.
When the SPI is used only to receive data, it is possible to enable only the SPI Rx DMA
channel.
In transmission mode, when the DMA has written all the data to be transmitted (flag TCIF is
set in the DMA_ISR register), the BSY flag can be monitored to ensure that the SPI
communication is complete. This is required to avoid corrupting the last transmission before
disabling the SPI or entering the Stop mode. The software must first wait until TXE=1 and
then until BSY=0.
Note:

During discontinuous communications, there is a 2 APB clock period delay between the
write operation to SPI_DR and the BSY bit setting. As a consequence, it is mandatory to
wait first until TXE=1 and then until BSY=0 after writing the last data.

RM0090 Rev 17

895/1747
925

Serial peripheral interface (SPI)

RM0090

Figure 259. Transmission using DMA
([DPSOHZLWK&32/ &3+$ 
6&.
'$7$ [)

'$7$  [)
0,62026, RXW

'$7$  [)

E E E E E E E E E E E E E E E E E E E E E E E E
VHWE\KDUGZDUH

VHWE\KDUGZDUH
FOHDUHGE\'0$ZULWH

7;(IODJ

FOHDUE\'0$ZULWH

VHWE\KDUGZDUH

UHVHW
E\KDUGZDUH

VHWE\KDUGZDUH

%6<IODJ

LJQRUHGE\WKH'0$EHFDXVH
'0$WUDQVIHULVFRPSOHWH

'0$UHTXHVW

7[EXIIHU
ZULWHWR63,B'5

[)

[)

[)

'0$ZULWHVWR63,B'5
VHWE\KDUGZDUH

'0$7&,)IODJ
'0$WUDQVIHUFRPSOHWH

VRIWZDUHFRQILJXUHVWKH
'0$63,7[FKDQQHO
WRVHQGGDWDLWHPV
DQGHQDEOHVWKH63,

'0$ZULWHV
'$7$LQWR
63,B'5

'0$ZULWHV
'$7$LQWR
63,B'5

'0$ZULWHV
'$7$LQWR
63,B'5

FOHDUE\VRIWZDUH

'0$WUDQVIHULV
FRPSOHWH 7&,) LQ
'0$B,65

VRIWZDUHZDLWV
XQWLO7;( 

VRIWZDUHZDLWVXQWLO%6< 

DL

Figure 260. Reception using DMA
([DPSOHZLWK&32/ &3+$ 
6&.
'$7$  [$
0,62026, LQ

5;1(IODJ

'$7$ [$

'$7$ [$

E E E E E E E E E E E E E E E E E E E E E E E E

VHWE\KDUGZDUH

FOHDUE\'0$UHDG

'0$UHTXHVW

5[EXIIHU
UHDGIURP63,B'5

[$

[$

[$

'0$UHDGIURP63,B'5
VHWE\KDUGZDUH

IODJ'0$7&,)
'0$WUDQVIHUFRPSOHWH
VRIWZDUHFRQILJXUHVWKH
'0$63,5[FKDQQHO
WRUHFHLYHGDWDLWHPV
DQGHQDEOHVWKH63,

'0$UHDGV
'$7$IURP
63,B'5

'0$UHDGV
'$7$IURP
63,B'5

'0$UHDGV
'$7$IURP
63,B'5

FOHDU
E\VRIWZDUH

7KH'0$WUDQVIHULV
FRPSOHWH 7&,) LQ
'0$B,65
DL

896/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)

DMA capability with CRC
When SPI communication is enabled with CRC communication and DMA mode, the
transmission and reception of the CRC at the end of communication are automatic that is
without using the bit CRCNEXT. After the CRC reception, the CRC must be read in the
SPI_DR register to clear the RXNE flag.
At the end of data and CRC transfers, the CRCERR flag in SPI_SR is set if corruption
occurs during the transfer.

28.3.10

Error flags
Master mode fault (MODF)
Master mode fault occurs when the master device has its NSS pin pulled low (in NSS
hardware mode) or SSI bit low (in NSS software mode), this automatically sets the MODF
bit. Master mode fault affects the SPI peripheral in the following ways:
•

The MODF bit is set and an SPI interrupt is generated if the ERRIE bit is set.

•

The SPE bit is cleared. This blocks all output from the device and disables the SPI
interface.

•

The MSTR bit is cleared, thus forcing the device into slave mode.

Use the following software sequence to clear the MODF bit:
1.

Make a read or write access to the SPI_SR register while the MODF bit is set.

2.

Then write to the SPI_CR1 register.

To avoid any multiple slave conflicts in a system comprising several MCUs, the NSS pin
must be pulled high during the MODF bit clearing sequence. The SPE and MSTR bits can
be restored to their original state after this clearing sequence.
As a security, hardware does not allow the setting of the SPE and MSTR bits while the
MODF bit is set.
In a slave device the MODF bit cannot be set. However, in a multimaster configuration, the
device can be in slave mode with this MODF bit set. In this case, the MODF bit indicates
that there might have been a multimaster conflict for system control. An interrupt routine can
be used to recover cleanly from this state by performing a reset or returning to a default
state.

Overrun condition
An overrun condition occurs when the master device has sent data bytes and the slave
device has not cleared the RXNE bit resulting from the previous data byte transmitted.
When an overrun condition occurs:
•

the OVR bit is set and an interrupt is generated if the ERRIE bit is set.

In this case, the receiver buffer contents will not be updated with the newly received data
from the master device. A read from the SPI_DR register returns this byte. All other
subsequently transmitted bytes are lost.
Clearing the OVR bit is done by a read from the SPI_DR register followed by a read access
to the SPI_SR register.

RM0090 Rev 17

897/1747
925

Serial peripheral interface (SPI)

RM0090

CRC error
This flag is used to verify the validity of the value received when the CRCEN bit in the
SPI_CR1 register is set. The CRCERR flag in the SPI_SR register is set if the value
received in the shift register does not match the receiver SPI_RXCRCR value.

TI mode frame format error
A TI mode frame format error is detected when an NSS pulse occurs during an ongoing
communication when the SPI is acting in slave mode and configured to conform to the TI
mode protocol. When this error occurs, the FRE flag is set in the SPI_SR register. The SPI
is not disabled when an error occurs, the NSS pulse is ignored, and the SPI waits for the
next NSS pulse before starting a new transfer. The data may be corrupted since the error
detection may result in the lost of two data bytes.
The FRE flag is cleared when SPI_SR register is read. If the bit ERRIE is set, an interrupt is
generated on the NSS error detection. In this case, the SPI should be disabled because
data consistency is no more guaranteed and communications should be reinitiated by the
master when the slave SPI is re-enabled.
Figure 261. TI mode frame format error detection
.33
OUTPUT

TRIGGER

SAMPLING TRIGGER SAMPLING TRIGGER

SAMPLING TRIGGER

SAMPLING TRIGGER SAMPLING TRIGGER

SAMPLING TRIGGER

SAMPLING

3#+
OUTPUT
-/3)
INPUT
-)3/
OUTPUT

$/.4#!2%

 OR 

-3").

,3").

$/.4#!2%

,3"/54

-3"/54

-3").

-3"/54

,3").

,3"/54

4)&2&%

AI

28.3.11

SPI interrupts
Table 126. SPI interrupt requests
Interrupt event

Event flag

Enable Control bit

TXE

TXEIE

Receive buffer not empty flag

RXNE

RXNEIE

Master Mode fault event

MODF

Transmit buffer empty flag

Overrun error

OVR

CRC error flag

CRCERR

TI frame format error

898/1747

FRE

RM0090 Rev 17

ERRIE

ERRIE

RM0090

Serial peripheral interface (SPI)

28.4

I2S functional description

28.4.1

I2S general description
The block diagram of the I2S is shown in Figure 262.
Figure 262. I2S block diagram

!DDRESS AND DATA BUS

4X BUFFER
#(
"39 /62 -/$& #2# 5$2
4X% 2X.% &2%
3)$%
%22

 BIT
-/3) 3$
-)3/
)3EXT?3$
)3EXT?3$

3HIFT REGISTER
 BIT

,3" FIRST

#OMMUNICATION
CONTROL

2X BUFFER

.3373

)3#&'
;=

)334$
;=

#+
$!4,%. #(
,%.
0/, ;=
)3
-/$ )3%

-ASTER CONTROL LOGIC
"IDI
"IDI #2# #2#
$&&
MODE /% %. .EXT

30)
BAUD RATE GENERATOR

2X
33- 33)
ONLY

,3"
30% "2 "2 "2 -342 #0/, #0(!
&IRST

#+
)3 CLOCK GENERATOR
)3?#+
)3-/$
-#+

-#+/%/$$

)3$)6;=

)3X#,+
-36

1. I2S2ext_SD and I2S3ext_SD are the extended SD pins that control the I2S full duplex mode.
The SPI could function as an audio I2S interface when the I2S capability is enabled

(by
setting the I2SMOD bit in the SPI_I2SCFGR register). This interface uses almost the same
pins, flags and interrupts as the SPI.
RM0090 Rev 17

899/1747
925

Serial peripheral interface (SPI)

RM0090

The I2S shares three common pins with the SPI:
•

SD: Serial Data (mapped on the MOSI pin) to transmit or receive the two timemultiplexed data channels (in half-duplex mode only).

•

WS: Word Select (mapped on the NSS pin) is the data control signal output in master
mode and input in slave mode.

•

CK: Serial Clock (mapped on the SCK pin) is the serial clock output in master mode
and serial clock input in slave mode.

•

I2S2ext_SD and I2S3ext_SD: additional pins (mapped on the MISO pin) to control the
I2S full duplex mode.

An additional pin could be used when a master clock output is needed for some external
audio devices:
•

MCK: Master Clock (mapped separately) is used, when the I2S is configured in master
mode (and when the MCKOE bit in the SPI_I2SPR register is set), to output this
additional clock generated at a preconfigured frequency rate equal to 256 × FS, where
FS is the audio sampling frequency.

The I2S uses its own clock generator to produce the communication clock when it is set in
master mode. This clock generator is also the source of the master clock output. Two
additional registers are available in I2S mode. One is linked to the clock generator
configuration SPI_I2SPR and the other one is a generic I2S configuration register
SPI_I2SCFGR (audio standard, slave/master mode, data format, packet frame, clock
polarity, etc.).
The SPI_CR1 register and all CRC registers are not used in the I2S mode. Likewise, the
SSOE bit in the SPI_CR2 register and the MODF and CRCERR bits in the SPI_SR are not
used.
The I2S uses the same SPI register for data transfer (SPI_DR) in 16-bit wide mode.

28.4.2

I2S full duplex
To support I2S full duplex mode, two extra I2S instances called extended I2Ss (I2S2_ext,
I2S3_ext) are available in addition to I2S2 and I2S3 (see Figure 263). The first I2S fullduplex interface is consequently based on I2S2 and I2S2_ext, and the second one on I2S3
and I2S3_ext.

Note:

I2S2_ext an I2S3_ext are used only in full-duplex mode.
Figure 263. I2S full duplex block diagram

)3X?3#+

30))3X

30)X?-/3))3X?3$INOUT

)3? 73

)3X?EXT

)3X?EXT3$INOUT
-36

1. Where x can be 2 or 3.

900/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)
I2Sx can operate in master mode. As a result:
•

Only I2Sx can output SCK and WS in half duplex mode

•

Only I2Sx can deliver SCK and WS to I2S2_ext and I2S3_ext in full duplex mode.

The extended I2Ss (I2Sx_ext) can be used only in full duplex mode. The I2Sx_ext operate
always in slave mode.
Both I2Sx and I2Sx_ext can be configured as transmitters or receivers.

28.4.3

Supported audio protocols
The four-line bus has to handle only audio data generally time-multiplexed on two channels:
the right channel and the left channel. However there is only one 16-bit register for the
transmission and the reception. So, it is up to the software to write into the data register the
adequate value corresponding to the considered channel side, or to read the data from the
data register and to identify the corresponding channel by checking the CHSIDE bit in the
SPI_SR register. Channel Left is always sent first followed by the channel right (CHSIDE
has no meaning for the PCM protocol).
Four data and packet frames are available. Data may be sent with a format of:
•

16-bit data packed in 16-bit frame

•

16-bit data packed in 32-bit frame

•

24-bit data packed in 32-bit frame

•

32-bit data packed in 32-bit frame

When using 16-bit data extended on 32-bit packet, the first 16 bits (MSB) are the significant
bits, the 16-bit LSB is forced to 0 without any need for software action or DMA request (only
one read/write operation).
The 24-bit and 32-bit data frames need two CPU read or write operations to/from the
SPI_DR or two DMA operations if the DMA is preferred for the application. For 24-bit data
frame specifically, the 8 nonsignificant bits are extended to 32 bits with 0-bits (by hardware).
For all data formats and communication standards, the most significant bit is always sent
first (MSB first).
The I2S interface supports four audio standards, configurable using the I2SSTD[1:0] and
PCMSYNC bits in the SPI_I2SCFGR register.

I2S Philips standard
For this standard, the WS signal is used to indicate which channel is being transmitted. It is
activated one CK clock cycle before the first bit (MSB) is available.

RM0090 Rev 17

901/1747
925

Serial peripheral interface (SPI)

RM0090

Figure 264. I2S Philips protocol waveforms (16/32-bit full accuracy, CPOL = 0)
&.

:6

WUDQVPLVVLRQ

UHFHSWLRQ

&DQEHELWRUELW

6'

06%

/6%

06%

&KDQQHOOHIW
&KDQQHO
ULJKW
069

Data are latched on the falling edge of CK (for the transmitter) and are read on the rising
edge (for the receiver). The WS signal is also latched on the falling edge of CK.
Figure 265. I2S Philips standard waveforms (24-bit frame with CPOL = 0)
&.
:6

7UDQVPLVVLRQ

5HFHSWLRQ

ELWGDWD

6'

ELWUHPDLQLQJIRUFHG

06%

/6%

&KDQQHOOHIWELW
&KDQQHOULJKW
069

This mode needs two write or read operations to/from the SPI_DR.
•

In transmission mode:
if 0x8EAA33 has to be sent (24-bit):
Figure 266. Transmitting 0x8EAA33
)LUVWZULWHWR'DWDUHJLVWHU

6HFRQGZULWHWR'DWDUHJLVWHU

[($$

[;;
2QO\WKH06%DUHVHQW
WRFRPSDUHWKHELWV
/6%VKDYHQRPHDQLQJ
DQGFDQEHDQ\WKLQJ
069

•

In reception mode:
if data 0x8EAA33 is received:

902/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)
Figure 267. Receiving 0x8EAA33
)LUVWUHDGWR'DWDUHJLVWHU

6HFRQGUHDGWR'DWDUHJLVWHU

[($$

[;;
2QO\WKH06%DUHVHQW
WRFRPSDUHWKHELWV
/6%VKDYHQRPHDQLQJ
DQGFDQEHDQ\WKLQJ
069

Figure 268. I2S Philips standard (16-bit extended to 32-bit packet frame with
CPOL = 0)
&.
:6

7UDQVPLVVLRQ

5HFHSWLRQ

ELWGDWD

6'

06%

ELWUHPDLQLQJIRUFHG
/6%

&KDQQHOOHIWELW

&KDQQHOULJKW

069

When 16-bit data frame extended to 32-bit channel frame is selected during the I2S
configuration phase, only one access to SPI_DR is required. The 16 remaining bits are
forced by hardware to 0x0000 to extend the data to 32-bit format.
If the data to transmit or the received data are 0x76A3 (0x76A30000 extended to 32-bit), the
operation shown in Figure 269 is required.
Figure 269. Example
2QO\RQHDFFHVVWR63,[B'5
[$
069

For transmission, each time an MSB is written to SPI_DR, the TXE flag is set and its
interrupt, if allowed, is generated to load SPI_DR with the new value to send. This takes
place even if 0x0000 have not yet been sent because it is done by hardware.
For reception, the RXNE flag is set and its interrupt, if allowed, is generated when the first
16 MSB half-word is received.
In this way, more time is provided between two write or read operations, which prevents
underrun or overrun conditions (depending on the direction of the data transfer).

MSB justified standard
For this standard, the WS signal is generated at the same time as the first data bit, which is
the MSBit.

RM0090 Rev 17

903/1747
925

Serial peripheral interface (SPI)

RM0090

Figure 270. MSB justified 16-bit or 32-bit full-accuracy length with CPOL = 0
&.
7UDQVPLVVLRQ

:6

5HFHSWLRQ
RUELWGDWD

6'
/6% 06%

06%
&KDQQHOOHIW

&KDQQHOULJKW
069

Data are latched on the falling edge of CK (for transmitter) and are read on the rising edge
(for the receiver).
Figure 271. MSB justified 24-bit frame length with CPOL = 0
&.
7UDQVPLVVLRQ

:6

5HFHSWLRQ

ELWGDWD

ELWUHPDLQLQJ
IRUFHG

6'
/6%

06%

&KDQQHOOHIWELW
&KDQQHOULJKW
069

Figure 272. MSB justified 16-bit extended to 32-bit packet frame with CPOL = 0
&.
7UDQVPLVVLRQ

:6

5HFHSWLRQ

ELWGDWD

ELWUHPDLQLQJ
IRUFHG

6'
06%

/6%
&KDQQHOOHIWELW
&KDQQHOULJKW
069

LSB justified standard
This standard is similar to the MSB justified standard (no difference for the 16-bit and 32-bit
full-accuracy frame formats).

904/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)
Figure 273. LSB justified 16-bit or 32-bit full-accuracy with CPOL = 0
&.
:6
5HFHSWLRQ

7UDQVPLVVLRQ
RUELWGDWD

6'
/6% 06%

06%
&KDQQHOOHIW

&KDQQHOULJKW
069

Figure 274. LSB justified 24-bit frame length with CPOL = 0
&.
:6

5HFHSWLRQ
7UDQVPLVVLRQ

6'

ELWGDWD
IRUFHG

ELWUHPDLQLQJ
06%

/6%

&KDQQHOOHIWELW
&KDQQHOULJKW
069

•

In transmission mode:
If data 0x3478AE have to be transmitted, two write operations to the SPI_DR register
are required from software or by DMA. The operations are shown below.
Figure 275. Operations required to transmit 0x3478AE
)LUVWZULWHWR'DWDUHJLVWHU
FRQGLWLRQHGE\7;( 
[;;

2QO\WKH/6%RIWKH
KDOIZRUGDUHVLJQLILFDQW
$ILHOGRI[LVIRUFHG
LQVWHDGRIWKH06%V

•

6HFRQGZULWHWR'DWDUHJLVWHU
FRQGLWLRQHGE\7;( 
[$(

069

In reception mode:
If data 0x3478AE are received, two successive read operations from SPI_DR are
required on each RXNE event.

RM0090 Rev 17

905/1747
925

Serial peripheral interface (SPI)

RM0090

Figure 276. Operations required to receive 0x3478AE
)LUVWUHDGIURP'DWDUHJLVWHU
FRQGLWLRQHGE\5;1( 

6HFRQGUHDGIURP'DWDUHJLVWHU
FRQGLWLRQHGE\5;1( 

[;;

[$(

2QO\WKH/6%RIWKH
KDOIZRUGDUHVLJQLILFDQW
$ILHOGRI[LVIRUFHG
LQVWHDGRIWKH06%V
069

Figure 277. LSB justified 16-bit extended to 32-bit packet frame with CPOL = 0
&.
5HFHSWLRQ

:6
7UDQVPLVVLRQ

6'

ELWUHPDLQLQJ

ELWGDWD
IRUFHG
06%

/6%

&KDQQHOOHIWELW

&KDQQHOULJKW

069

When 16-bit data frame extended to 32-bit channel frame is selected during the I2S
configuration phase, Only one access to SPI_DR is required. The 16 remaining bits are
forced by hardware to 0x0000 to extend the data to 32-bit format. In this case it corresponds
to the half-word MSB.
If the data to transmit or the received data are 0x76A3 (0x0000 76A3 extended to 32-bit),
the operation shown in Figure 278 is required.
Figure 278. Example of LSB justified 16-bit extended to 32-bit packet frame
2QO\RQHDFFHVVWRWKH63,['5UHJLVWHU

[$
069

In transmission mode, when TXE is asserted, the application has to write the data to be
transmitted (in this case 0x76A3). The 0x000 field is transmitted first (extension on 32-bit).
TXE is asserted again as soon as the effective data (0x76A3) is sent on SD.
In reception mode, RXNE is asserted as soon as the significant half-word is received (and
not the 0x0000 field).
In this way, more time is provided between two write or read operations to prevent underrun
or overrun conditions.

906/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)

PCM standard
For the PCM standard, there is no need to use channel-side information. The two PCM
modes (short and long frame) are available and configurable using the PCMSYNC bit in
SPI_I2SCFGR.
Figure 279. PCM standard waveforms (16-bit)
&.
:6
VKRUWIUDPH
ELWV

:6
ORQJIUDPH

6'

06%

/6% 06%
069

For long frame synchronization, the WS signal assertion time is fixed 13 bits in master
mode.
For short frame synchronization, the WS synchronization signal is only one cycle long.
Figure 280. PCM standard waveforms (16-bit extended to 32-bit packet frame)
&.
:6
VKRUWIUDPH
:6
ORQJIUDPH

8SWRELWV

ELWV

6'

06%

/6%

069

Note:

For both modes (master and slave) and for both synchronizations (short and long), the
number of bits between two consecutive pieces of data (and so two synchronization signals)
needs to be specified (DATLEN and CHLEN bits in the SPI_I2SCFGR register) even in
slave mode.

28.4.4

Clock generator
The I2S bitrate determines the dataflow on the I2S data line and the I2S clock signal
frequency.
I2S bitrate = number of bits per channel × number of channels × sampling audio frequency
For a 16-bit audio, left and right channel, the I2S bitrate is calculated as follows:
I2S bitrate = 16 × 2 × FS
It will be: I2S bitrate = 32 x 2 x FS if the packet length is 32-bit wide.
RM0090 Rev 17

907/1747
925

Serial peripheral interface (SPI)

RM0090

Figure 281. Audio sampling frequency definition
RUELW
ULJKWFKDQQHO

RUELWOHIW
FKDQQHO
RUELWV
)6

VDPSOLQJSRLQW

VDPSOLQJSRLQW

)6DXGLRVDPSOLQJIUHTXHQF\
069

When the master mode is configured, a specific action needs to be taken to properly
program the linear divider in order to communicate with the desired audio frequency.
Figure 282. I2S clock generator architecture

0&.
,ð6[&/.

ELWOLQHDUGLYLGHU
UHVKDSLQJVWDJH



'LY

'LYLGHUE\







&.

,ð6',9>@

&+/(1

0&.2( 2''

,ð602'

0&.2(

069

1. Where x could be 2 or 3.

Figure 281 presents the communication clock architecture. To achieve high-quality audio
performance, the I2SxCLK clock source can be either the PLLI2S output (through R division
factor) or an external clock (mapped to I2S_CKIN pin).
The audio sampling frequency can be 192 kHz, 96 kHz, or 48 kHz. In order to reach the
desired frequency, the linear divider needs to be programmed according to the formulas
below:
When the master clock is generated (MCKOE in the SPI_I2SPR register is set):
FS = I2SxCLK / [(16*2)*((2*I2SDIV)+ODD)*8)] when the channel frame is 16-bit wide
FS = I2SxCLK / [(32*2)*((2*I2SDIV)+ODD)*4)] when the channel frame is 32-bit wide
When the master clock is disabled (MCKOE bit cleared):
FS = I2SxCLK / [(16*2)*((2*I2SDIV)+ODD))] when the channel frame is 16-bit wide
FS = I2SxCLK / [(32*2)*((2*I2SDIV)+ODD))] when the channel frame is 32-bit wide
Table 127 provides example precision values for different clock configurations.
Note:

908/1747

Other configurations are possible that allow optimum clock precision.

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)
Table 127. Audio frequency precision (for PLLM VCO = 1 MHz or 2 MHz)(1)

Master
clock

Target fS
(Hz)

Data
format

PLLI2SN

PLLI2SR

I2SDIV

I2SODD

Real fS (Hz)

Error

16-bit

192

2

187

1

8000

0.0000%

32-bit

192

3

62

1

8000

0.0000%

16-bit

192

3

62

1

16000

0.0000%

32-bit

256

2

62

1

16000

0.0000%

16-bit

256

2

62

1

32000

0.0000%

32-bit

256

5

12

1

32000

0.0000%

16-bit

192

5

12

1

48000

0.0000%

32-bit

384

5

12

1

48000

0.0000%

16-bit

384

5

12

1

96000

0.0000%

32-bit

424

3

11

1

96014.49219

0.0151%

16-bit

290

3

68

1

22049.87695

0.0006%

32-bit

302

2

53

1

22050.23438

0.0011%

16-bit

302

2

53

1

44100.46875

0.0011%

32-bit

429

4

19

0

44099.50781

0.0011%

16-bit

424

3

11

1

192028.9844

0.0151%

32-bit

258

3

3

1

191964.2813

0.0186%

8000

don't care

256

5

12

1

8000

0.0000%

16000

don't care

213

2

13

0

16000.60059

0.0038%

32000

don't care

213

2

6

1

32001.20117

0.0038%

48000

don't care

258

3

3

1

47991.07031

0.0186%

96000

don't care

344

2

3

1

95982.14063

0.0186%

22050

don't care

429

4

9

1

22049.75391

0.0011%

44100

don't care

271

2

6

0

44108.07422

0.0183%

8000

16000

32000

48000
Disabled
96000

22050

44100

192000

Enabled

1. This table gives only example values for different clock configurations. Other configurations allowing optimum clock
precision are possible.

28.4.5

I2S master mode
The I2S can be configured as follows:
•

In master mode for transmission or reception (half-duplex mode using I2Sx)

•

In master mode transmission and reception (full duplex mode using I2Sx and
I2Sx_ext).

This means that the serial clock is generated on the CK pin as well as the Word Select
signal WS. Master clock (MCK) may be output or not, thanks to the MCKOE bit in the
SPI_I2SPR register.

RM0090 Rev 17

909/1747
925

Serial peripheral interface (SPI)

RM0090

Procedure
1.

Select the I2SDIV[7:0] bits in the SPI_I2SPR register to define the serial clock baud
rate to reach the proper audio sample frequency. The ODD bit in the SPI_I2SPR
register also has to be defined.

2.

Select the CKPOL bit to define the steady level for the communication clock. Set the
MCKOE bit in the SPI_I2SPR register if the master clock MCK needs to be provided to
the external DAC/ADC audio component (the I2SDIV and ODD values should be
computed depending on the state of the MCK output, for more details refer to
Section 28.4.4: Clock generator).

3.

Set the I2SMOD bit in SPI_I2SCFGR to activate the I2S functionalities and choose the
I2S standard through the I2SSTD[1:0] and PCMSYNC bits, the data length through the
DATLEN[1:0] bits and the number of bits per channel by configuring the CHLEN bit.
Select also the I2S master mode and direction (Transmitter or Receiver) through the
I2SCFG[1:0] bits in the SPI_I2SCFGR register.

4.

If needed, select all the potential interruption sources and the DMA capabilities by
writing the SPI_CR2 register.

5.

The I2SE bit in SPI_I2SCFGR register must be set.

WS and CK are configured in output mode. MCK is also an output, if the MCKOE bit in
SPI_I2SPR is set.

Transmission sequence
The transmission sequence begins when a half-word is written into the Tx buffer.
Assumedly, the first data written into the Tx buffer correspond to the channel Left data.
When data are transferred from the Tx buffer to the shift register, TXE is set and data
corresponding to the channel Right have to be written into the Tx buffer. The CHSIDE flag
indicates which channel is to be transmitted. It has a meaning when the TXE flag is set
because the CHSIDE flag is updated when TXE goes high.
A full frame has to be considered as a Left channel data transmission followed by a Right
channel data transmission. It is not possible to have a partial frame where only the left
channel is sent.
The data half-word is parallel loaded into the 16-bit shift register during the first bit
transmission, and then shifted out, serially, to the MOSI/SD pin, MSB first. The TXE flag is
set after each transfer from the Tx buffer to the shift register and an interrupt is generated if
the TXEIE bit in the SPI_CR2 register is set.
For more details about the write operations depending on the I2S standard mode selected,
refer to Section 28.4.3: Supported audio protocols).
To ensure a continuous audio data transmission, it is mandatory to write the SPI_DR with
the next data to transmit before the end of the current transmission.
To switch off the I2S, by clearing I2SE, it is mandatory to wait for TXE = 1 and BSY = 0.

Reception sequence
The operating mode is the same as for the transmission mode except for the point 3 (refer to
the procedure described in Section 28.4.5: I2S master mode), where the configuration
should set the master reception mode through the I2SCFG[1:0] bits.
Whatever the data or channel length, the audio data are received by 16-bit packets. This
means that each time the Rx buffer is full, the RXNE flag is set and an interrupt is generated
910/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)
if the RXNEIE bit is set in SPI_CR2 register. Depending on the data and channel length
configuration, the audio value received for a right or left channel may result from one or two
receptions into the Rx buffer.
Clearing the RXNE bit is performed by reading the SPI_DR register.
CHSIDE is updated after each reception. It is sensitive to the WS signal generated by the
I2S cell.
For more details about the read operations depending on the I2S standard mode selected,
refer to Section 28.4.3: Supported audio protocols.
If data are received while the previously received data have not been read yet, an overrun is
generated and the OVR flag is set. If the ERRIE bit is set in the SPI_CR2 register, an
interrupt is generated to indicate the error.
To switch off the I2S, specific actions are required to ensure that the I2S completes the
transfer cycle properly without initiating a new data transfer. The sequence depends on the
configuration of the data and channel lengths, and on the audio protocol mode selected. In
the case of:
•

•

•

16-bit data length extended on 32-bit channel length (DATLEN = 00 and CHLEN = 1)
using the LSB justified mode (I2SSTD = 10)
a)

Wait for the second to last RXNE = 1 (n – 1)

b)

Then wait 17 I2S clock cycles (using a software loop)

c)

Disable the I2S (I2SE = 0)

16-bit data length extended on 32-bit channel length (DATLEN = 00 and CHLEN = 1) in
MSB justified, I2S or PCM modes (I2SSTD = 00, I2SSTD = 01 or I2SSTD = 11,
respectively)
a)

Wait for the last RXNE

b)

Then wait 1 I2S clock cycle (using a software loop)

c)

Disable the I2S (I2SE = 0)

For all other combinations of DATLEN and CHLEN, whatever the audio mode selected
through the I2SSTD bits, carry out the following sequence to switch off the I2S:
a)

Wait for the second to last RXNE = 1 (n – 1)

b)

Then wait one I2S clock cycle (using a software loop)

c)

Disable the I2S (I2SE = 0)

Note:

The BSY flag is kept low during transfers.

28.4.6

I2S slave mode
The I2S can be configured as follows:
•

In slave mode for transmission or reception (half-duplex mode using I2Sx)

•

In slave mode transmission and reception (full duplex mode using I2Sx and I2Sx_ext).

The operating mode is following mainly the same rules as described for the I2S master
configuration. In slave mode, there is no clock to be generated by the I2S interface. The
clock and WS signals are input from the external master connected to the I2S interface.
There is then no need, for the user, to configure the clock.
The configuration steps to follow are listed below:

RM0090 Rev 17

911/1747
925

Serial peripheral interface (SPI)

RM0090

1.

Set the I2SMOD bit in the SPI_I2SCFGR register to reach the I2S functionalities and
choose the I2S standard through the I2SSTD[1:0] bits, the data length through the
DATLEN[1:0] bits and the number of bits per channel for the frame configuring the
CHLEN bit. Select also the mode (transmission or reception) for the slave through the
I2SCFG[1:0] bits in SPI_I2SCFGR register.

2.

If needed, select all the potential interrupt sources and the DMA capabilities by writing
the SPI_CR2 register.

3.

The I2SE bit in SPI_I2SCFGR register must be set.

Transmission sequence
The transmission sequence begins when the external master device sends the clock and
when the NSS_WS signal requests the transfer of data. The slave has to be enabled before
the external master starts the communication. The I2S data register has to be loaded before
the master initiates the communication.
For the I2S, MSB justified and LSB justified modes, the first data item to be written into the
data register corresponds to the data for the left channel. When the communication starts,
the data are transferred from the Tx buffer to the shift register. The TXE flag is then set in
order to request the right channel data to be written into the I2S data register.
The CHSIDE flag indicates which channel is to be transmitted. Compared to the master
transmission mode, in slave mode, CHSIDE is sensitive to the WS signal coming from the
external master. This means that the slave needs to be ready to transmit the first data
before the clock is generated by the master. WS assertion corresponds to left channel
transmitted first.
Note:

The I2SE has to be written at least two PCLK cycles before the first clock of the master
comes on the CK line.
The data half-word is parallel-loaded into the 16-bit shift register (from the internal bus)
during the first bit transmission, and then shifted out serially to the MOSI/SD pin MSB first.
The TXE flag is set after each transfer from the Tx buffer to the shift register and an interrupt
is generated if the TXEIE bit in the SPI_CR2 register is set.
Note that the TXE flag should be checked to be at 1 before attempting to write the Tx buffer.
For more details about the write operations depending on the I2S standard mode selected,
refer to Section 28.4.3: Supported audio protocols.
To secure a continuous audio data transmission, it is mandatory to write the SPI_DR
register with the next data to transmit before the end of the current transmission. An
underrun flag is set and an interrupt may be generated if the data are not written into the
SPI_DR register before the first clock edge of the next data communication. This indicates
to the software that the transferred data are wrong. If the ERRIE bit is set into the SPI_CR2
register, an interrupt is generated when the UDR flag in the SPI_SR register goes high. In
this case, it is mandatory to switch off the I2S and to restart a data transfer starting from the
left channel.
To switch off the I2S, by clearing the I2SE bit, it is mandatory to wait for TXE = 1 and BSY =
0.

Reception sequence
The operating mode is the same as for the transmission mode except for the point 1 (refer to
the procedure described in Section 28.4.6: I2S slave mode), where the configuration should
set the master reception mode using the I2SCFG[1:0] bits in the SPI_I2SCFGR register.
912/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)
Whatever the data length or the channel length, the audio data are received by 16-bit
packets. This means that each time the RX buffer is full, the RXNE flag in the SPI_SR
register is set and an interrupt is generated if the RXNEIE bit is set in the SPI_CR2 register.
Depending on the data length and channel length configuration, the audio value received for
a right or left channel may result from one or two receptions into the RX buffer.
The CHSIDE flag is updated each time data are received to be read from SPI_DR. It is
sensitive to the external WS line managed by the external master component.
Clearing the RXNE bit is performed by reading the SPI_DR register.
For more details about the read operations depending the I2S standard mode selected, refer
to Section 28.4.3: Supported audio protocols.
If data are received while the precedent received data have not yet been read, an overrun is
generated and the OVR flag is set. If the bit ERRIE is set in the SPI_CR2 register, an
interrupt is generated to indicate the error.
To switch off the I2S in reception mode, I2SE has to be cleared immediately after receiving
the last RXNE = 1.

Note:

The external master components should have the capability of sending/receiving data in 16bit or 32-bit packets via an audio channel.

28.4.7

Status flags
Three status flags are provided for the application to fully monitor the state of the I2S bus.

Busy flag (BSY)
The BSY flag is set and cleared by hardware (writing to this flag has no effect). It indicates
the state of the communication layer of the I2S.
When BSY is set, it indicates that the I2S is busy communicating. There is one exception in
master receive mode (I2SCFG = 11) where the BSY flag is kept low during reception.
The BSY flag is useful to detect the end of a transfer if the software needs to disable the I2S.
This avoids corrupting the last transfer. For this, the procedure described below must be
strictly respected.
The BSY flag is set when a transfer starts, except when the I2S is in master receiver mode.
The BSY flag is cleared:
•

when a transfer completes (except in master transmit mode, in which the
communication is supposed to be continuous)

•

when the I2S is disabled

When communication is continuous:

Note:

•

In master transmit mode, the BSY flag is kept high during all the transfers

•

In slave mode, the BSY flag goes low for one I2S clock cycle between each transfer

Do not use the BSY flag to handle each data transmission or reception. It is better to use the
TXE and RXNE flags instead.

RM0090 Rev 17

913/1747
925

Serial peripheral interface (SPI)

RM0090

Tx buffer empty flag (TXE)
When set, this flag indicates that the Tx buffer is empty and the next data to be transmitted
can then be loaded into it. The TXE flag is reset when the Tx buffer already contains data to
be transmitted. It is also reset when the I2S is disabled (I2SE bit is reset).

RX buffer not empty (RXNE)
When set, this flag indicates that there are valid received data in the RX Buffer. It is reset
when SPI_DR register is read.

Channel Side flag (CHSIDE)
In transmission mode, this flag is refreshed when TXE goes high. It indicates the channel
side to which the data to transfer on SD has to belong. In case of an underrun error event in
slave transmission mode, this flag is not reliable and I2S needs to be switched off and
switched on before resuming the communication.
In reception mode, this flag is refreshed when data are received into SPI_DR. It indicates
from which channel side data have been received. Note that in case of error (like OVR) this
flag becomes meaningless and the I2S should be reset by disabling and then enabling it
(with configuration if it needs changing).
This flag has no meaning in the PCM standard (for both Short and Long frame modes).
When the OVR or UDR flag in the SPI_SR is set and the ERRIE bit in SPI_CR2 is also set,
an interrupt is generated. This interrupt can be cleared by reading the SPI_SR status
register (once the interrupt source has been cleared).

28.4.8

Error flags
There are three error flags for the I2S cell.

Underrun flag (UDR)
In slave transmission mode this flag is set when the first clock for data transmission appears
while the software has not yet loaded any value into SPI_DR. It is available when the
I2SMOD bit in SPI_I2SCFGR is set. An interrupt may be generated if the ERRIE bit in
SPI_CR2 is set.
The UDR bit is cleared by a read operation on the SPI_SR register.

Overrun flag (OVR)
This flag is set when data are received and the previous data have not yet been read from
SPI_DR. As a result, the incoming data are lost. An interrupt may be generated if the ERRIE
bit is set in SPI_CR2.
In this case, the receive buffer contents are not updated with the newly received data from
the transmitter device. A read operation to the SPI_DR register returns the previous
correctly received data. All other subsequently transmitted half-words are lost.
Clearing the OVR bit is done by a read operation on the SPI_DR register followed by a read
access to the SPI_SR register.

Frame error flag (FRE)
This flag can be set by hardware only if the I2S is configured in Slave mode. It is set if the
external master is changing the WS line at a moment when the slave is not expected this
914/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)
change. If the synchronization is lost, to recover from this state and resynchronize the
external master device with the I2S slave device, follow the steps below:
1.

Disable the I2S

2.

Re-enable it when the correct level is detected on the WS line (WS line is high in I2S
mode, or low for MSB- or LSB-justified or PCM modes).

Desynchronization between the master and slave device may be due to noisy environment
on the SCK communication clock or on the WS frame synchronization line. An error interrupt
can be generated if the ERRIE bit is set. The desynchronization flag (FRE) is cleared by
software when the status register is read.

28.4.9

I2S interrupts
Table 128 provides the list of I2S interrupts.
Table 128. I2S interrupt requests
Interrupt event

28.4.10

Event flag

Enable Control bit

Transmit buffer empty flag

TXE

TXEIE

Receive buffer not empty flag

RXNE

RXNEIE

Overrun error

OVR

Underrun error

UDR

Frame error flag

FRE

ERRIE
ERRIE

DMA features
DMA is working in exactly the same way as for the SPI mode. There is no difference on the
I2S. Only the CRC feature is not available in I2S mode since there is no data transfer
protection system.

RM0090 Rev 17

915/1747
925

Serial peripheral interface (SPI)

RM0090

SPI and I2S registers

28.5

The peripheral registers have to be accessed by half-words (16 bits) or words (32 bits).

SPI control register 1 (SPI_CR1) (not used in I2S mode)

28.5.1

Address offset: 0x00
Reset value: 0x0000
15

14

13

12

11

10

BIDI
MODE

BIDI
OE

CRC
EN

CRC
NEXT

DFF

RX
ONLY

rw

rw

rw

rw

rw

rw

9

8

7

6

SSM

SSI

LSB
FIRST

SPE

rw

rw

rw

rw

5

4

3

BR [2:0]
rw

rw

rw

2

1

0

MSTR

CPOL

CPHA

rw

rw

rw

Bit 15 BIDIMODE: Bidirectional data mode enable
0: 2-line unidirectional data mode selected
1: 1-line bidirectional data mode selected
Note: This bit is not used in I2S mode
Bit 14 BIDIOE: Output enable in bidirectional mode
This bit combined with the BIDImode bit selects the direction of transfer in bidirectional mode
0: Output disabled (receive-only mode)
1: Output enabled (transmit-only mode)
Note: This bit is not used in I2S mode.
In master mode, the MOSI pin is used while the MISO pin is used in slave mode.
Bit 13 CRCEN: Hardware CRC calculation enable
0: CRC calculation disabled
1: CRC calculation enabled
Note: This bit should be written only when SPI is disabled (SPE = ‘0’) for correct operation.
It is not used in I2S mode.
Bit 12 CRCNEXT: CRC transfer next
0: Data phase (no CRC phase)
1: Next transfer is CRC (CRC phase)
Note: When the SPI is configured in full duplex or transmitter only modes, CRCNEXT must be
written as soon as the last data is written to the SPI_DR register.
When the SPI is configured in receiver only mode, CRCNEXT must be set after the
second last data reception.
This bit should be kept cleared when the transfers are managed by DMA.
It is not used in I2S mode.
Bit 11 DFF: Data frame format
0: 8-bit data frame format is selected for transmission/reception
1: 16-bit data frame format is selected for transmission/reception
Note: This bit should be written only when SPI is disabled (SPE = ‘0’) for correct operation.
It is not used in I2S mode.

916/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)

Bit 10 RXONLY: Receive only
This bit combined with the BIDImode bit selects the direction of transfer in 2-line
unidirectional mode. This bit is also useful in a multislave system in which this particular
slave is not accessed, the output from the accessed slave is not corrupted.
0: Full duplex (Transmit and receive)
1: Output disabled (Receive-only mode)
Note: This bit is not used in I2S mode
Bit 9 SSM: Software slave management
When the SSM bit is set, the NSS pin input is replaced with the value from the SSI bit.
0: Software slave management disabled
1: Software slave management enabled
Note: This bit is not used in I2S mode and SPI TI mode
Bit 8 SSI: Internal slave select
This bit has an effect only when the SSM bit is set. The value of this bit is forced onto the
NSS pin and the IO value of the NSS pin is ignored.
Note: This bit is not used in I2S mode and SPI TI mode
Bit 7 LSBFIRST: Frame format
0: MSB transmitted first
1: LSB transmitted first
Note: This bit should not be changed when communication is ongoing.
It is not used in I2S mode and SPI TI mode
Bit 6 SPE: SPI enable
0: Peripheral disabled
1: Peripheral enabled
Note: This bit is not used in I2S mode.
When disabling the SPI, follow the procedure described in Section 28.3.8: Disabling the
SPI.
Bits 5:3 BR[2:0]: Baud rate control
000: fPCLK/2
001: fPCLK/4
010: fPCLK/8
011: fPCLK/16
100: fPCLK/32
101: fPCLK/64
110: fPCLK/128
111: fPCLK/256
Note: These bits should not be changed when communication is ongoing.
They are not used in I2S mode.
Bit 2 MSTR: Master selection
0: Slave configuration
1: Master configuration
Note: This bit should not be changed when communication is ongoing.
It is not used in I2S mode.

RM0090 Rev 17

917/1747
925

Serial peripheral interface (SPI)

RM0090

Bit1 CPOL: Clock polarity
0: CK to 0 when idle
1: CK to 1 when idle
Note: This bit should not be changed when communication is ongoing.
It is not used in I2S mode and SPI TI mode.
Bit 0 CPHA: Clock phase
0: The first clock transition is the first data capture edge
1: The second clock transition is the first data capture edge
Note: This bit should not be changed when communication is ongoing.
It is not used in I2S mode and SPI TI mode.

28.5.2

SPI control register 2 (SPI_CR2)
Address offset: 0x04
Reset value: 0x0000

15

14

13

12

11

10

9

Reserved

8

7

6

5

TXEIE RXNEIE ERRIE
rw

rw

rw

4
FRF
rw

3
Res.

2
SSOE
rw

1

0

TXDMAEN RXDMAEN
rw

rw

Bits 15:8 Reserved, must be kept at reset value.
Bit 7 TXEIE: Tx buffer empty interrupt enable
0: TXE interrupt masked
1: TXE interrupt not masked. Used to generate an interrupt request when the TXE flag is set.
Bit 6 RXNEIE: RX buffer not empty interrupt enable
0: RXNE interrupt masked
1: RXNE interrupt not masked. Used to generate an interrupt request when the RXNE flag is
set.
Bit 5 ERRIE: Error interrupt enable
This bit controls the generation of an interrupt when an error condition occurs )(CRCERR,
OVR, MODF in SPI mode, FRE in TI mode and UDR, OVR, and FRE in I2S mode).
0: Error interrupt is masked
1: Error interrupt is enabled
Bit 4 FRF: Frame format
0: SPI Motorola mode
1 SPI TI mode
Note: This bit is not used in I2S mode.
Bit 3 Reserved. Forced to 0 by hardware.
Bit 2 SSOE: SS output enable
0: SS output is disabled in master mode and the cell can work in multimaster configuration
1: SS output is enabled in master mode and when the cell is enabled. The cell cannot work
in a multimaster environment.
Note: This bit is not used in I2S mode and SPI TI mode.

918/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)

Bit 1 TXDMAEN: Tx buffer DMA enable
When this bit is set, the DMA request is made whenever the TXE flag is set.
0: Tx buffer DMA disabled
1: Tx buffer DMA enabled
Bit 0 RXDMAEN: Rx buffer DMA enable
When this bit is set, the DMA request is made whenever the RXNE flag is set.
0: Rx buffer DMA disabled
1: Rx buffer DMA enabled

28.5.3

SPI status register (SPI_SR)
Address offset: 0x08
Reset value: 0x0002

15

14

13

12
Reserved

11

10

9

8

7

6

5

4

3

2

1

0

UDR

CHSIDE

TXE

RXNE

r

r

r

r

FRE

BSY

OVR

MODF

CRC
ERR

r

r

r

r

rc_w0

Bits 15:9 Reserved. Forced to 0 by hardware.
Bit 8 FRE: Frame format error
0: No frame format error
1: A frame format error occurred
This flag is set by hardware and cleared by software when the SPIx_SR register is read.
Note: This flag is used when the SPI operates in TI slave mode or I2S slave mode (refer to
Section 28.3.10).
Bit 7 BSY: Busy flag
0: SPI (or I2S) not busy
1: SPI (or I2S) is busy in communication or Tx buffer is not empty
This flag is set and cleared by hardware.
Note: BSY flag must be used with caution: refer to Section 28.3.7: Status flags and
Section 28.3.8: Disabling the SPI.
Bit 6 OVR: Overrun flag
0: No overrun occurred
1: Overrun occurred
This flag is set by hardware and reset by a software sequence. Refer to Section 28.4.8: Error
flags for the software sequence.
Bit 5 MODF: Mode fault
0: No mode fault occurred
1: Mode fault occurred
This flag is set by hardware and reset by a software sequence. Refer to Section 28.3.10:
Error flags for the software sequence.
Note: This bit is not used in I2S mode
Bit 4 CRCERR: CRC error flag
0: CRC value received matches the SPI_RXCRCR value
1: CRC value received does not match the SPI_RXCRCR value
This flag is set by hardware and cleared by software writing 0.
Note: This bit is not used in I2S mode.

RM0090 Rev 17

919/1747
925

Serial peripheral interface (SPI)

RM0090

Bit 3 UDR: Underrun flag
0: No underrun occurred
1: Underrun occurred
This flag is set by hardware and reset by a software sequence. Refer to Section 28.4.8: Error
flags for the software sequence.
Note: This bit is not used in SPI mode.
Bit 2 CHSIDE: Channel side
0: Channel Left has to be transmitted or has been received
1: Channel Right has to be transmitted or has been received
Note: This bit is not used for SPI mode and is meaningless in PCM mode.
Bit 1 TXE: Transmit buffer empty
0: Tx buffer not empty
1: Tx buffer empty
Bit 0 RXNE: Receive buffer not empty
0: Rx buffer empty
1: Rx buffer not empty

28.5.4

SPI data register (SPI_DR)
Address offset: 0x0C
Reset value: 0x0000

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

DR[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 15:0 DR[15:0]: Data register
Data received or to be transmitted.
The data register is split into 2 buffers - one for writing (Transmit Buffer) and another one for
reading (Receive buffer). A write to the data register will write into the Tx buffer and a read
from the data register will return the value held in the Rx buffer.
Note: These notes apply to SPI mode:
Depending on the data frame format selection bit (DFF in SPI_CR1 register), the data
sent or received is either 8-bit or 16-bit. This selection has to be made before enabling
the SPI to ensure correct operation.
For an 8-bit data frame, the buffers are 8-bit and only the LSB of the register
(SPI_DR[7:0]) is used for transmission/reception. When in reception mode, the MSB of
the register (SPI_DR[15:8]) is forced to 0.
For a 16-bit data frame, the buffers are 16-bit and the entire register, SPI_DR[15:0] is
used for transmission/reception.

28.5.5

SPI CRC polynomial register (SPI_CRCPR) (not used in I2S
mode)
Address offset: 0x10

920/1747

RM0090 Rev 17

RM0090

Serial peripheral interface (SPI)
Reset value: 0x0007

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

CRCPOLY[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 15:0 CRCPOLY[15:0]: CRC polynomial register
This register contains the polynomial for the CRC calculation.
The CRC polynomial (0007h) is the reset value of this register. Another polynomial can be
configured as required.
Note: These bits are not used for the I2S mode.

SPI RX CRC register (SPI_RXCRCR) (not used in I2S mode)

28.5.6

Address offset: 0x14
Reset value: 0x0000
15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

RXCRC[15:0]
r

r

r

r

r

r

r

r

r

Bits 15:0 RXCRC[15:0]: Rx CRC register
When CRC calculation is enabled, the RxCRC[15:0] bits contain the computed CRC value of
the subsequently received bytes. This register is reset when the CRCEN bit in SPI_CR1
register is written to 1. The CRC is calculated serially using the polynomial programmed in
the SPI_CRCPR register.
Only the 8 LSB bits are considered when the data frame format is set to be 8-bit data (DFF
bit of SPI_CR1 is cleared). CRC calculation is done based on any CRC8 standard.
The entire 16-bits of this register are considered when a 16-bit data frame format is selected
(DFF bit of the SPI_CR1 register is set). CRC calculation is done based on any CRC16
standard.
Note: A read to this register when the BSY Flag is set could return an incorrect value.
These bits are not used for I2S mode.

SPI TX CRC register (SPI_TXCRCR) (not used in I2S mode)

28.5.7

Address offset: 0x18
Reset value: 0x0000
15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

TXCRC[15:0]
r

r

r

r

r

r

r

r

r

RM0090 Rev 17

921/1747
925

Serial peripheral interface (SPI)

RM0090

Bits 15:0 TXCRC[15:0]: Tx CRC register
When CRC calculation is enabled, the TxCRC[7:0] bits contain the computed CRC value of
the subsequently transmitted bytes. This register is reset when the CRCEN bit of SPI_CR1
is written to 1. The CRC is calculated serially using the polynomial programmed in the
SPI_CRCPR register.
Only the 8 LSB bits are considered when the data frame format is set to be 8-bit data (DFF
bit of SPI_CR1 is cleared). CRC calculation is done based on any CRC8 standard.
The entire 16-bits of this register are considered when a 16-bit data frame format is selected
(DFF bit of the SPI_CR1 register is set). CRC calculation is done based on any CRC16
standard.
Note: A read to this register when the BSY flag is set could return an incorrect value.
These bits are not used for I2S mode.

SPI_I2S configuration register (SPI_I2SCFGR)

28.5.8

Address offset: 0x1C
Reset value: 0x0000
15

14

13

Reserved

12

11

10

I2SMOD

I2SE

rw

rw

9

8

I2SCFG
rw

rw

7
PCMSY
NC
rw

6

5

4

I2SSTD

Res.

rw

3
CKPOL

rw

Bits 15:12 Reserved, must be kept at reset value.
Bit 11 I2SMOD: I2S mode selection
0: SPI mode is selected
1: I2S mode is selected
Note: This bit should be configured when the SPI or I2S is disabled
Bit 10 I2SE: I2S Enable
0: I2S peripheral is disabled
1: I2S peripheral is enabled
Note: This bit is not used in SPI mode.
Bits 9:8 I2SCFG: I2S configuration mode
00: Slave - transmit
01: Slave - receive
10: Master - transmit
11: Master - receive
Note: This bit should be configured when the I2S is disabled.
It is not used in SPI mode.
Bit 7 PCMSYNC: PCM frame synchronization
0: Short frame synchronization
1: Long frame synchronization
Note: This bit has a meaning only if I2SSTD = 11 (PCM standard is used)
It is not used in SPI mode.
Bit 6 Reserved: forced at 0 by hardware

922/1747

RM0090 Rev 17

rw

2

1

DATLEN
rw

rw

0
CHLEN
rw

RM0090

Serial peripheral interface (SPI)

Bits 5:4 I2SSTD: I2S standard selection
00: I2S Philips standard.
01: MSB justified standard (left justified)
10: LSB justified standard (right justified)
11: PCM standard
For more details on I2S standards, refer to Section 28.4.3: Supported audio protocols. Not used in
SPI mode.
Note: For correct operation, these bits should be configured when the I2S is disabled.
Bit 3 CKPOL: Steady state clock polarity
0: I2S clock steady state is low level
1: I2S clock steady state is high level
Note: For correct operation, this bit should be configured when the I2S is disabled.
This bit is not used in SPI mode
Bits 2:1 DATLEN: Data length to be transferred
00: 16-bit data length
01: 24-bit data length
10: 32-bit data length
11: Not allowed
Note: For correct operation, these bits should be configured when the I2S is disabled.
This bit is not used in SPI mode.
Bit 0 CHLEN: Channel length (number of bits per audio channel)
0: 16-bit wide
1: 32-bit wide
The bit write operation has a meaning only if DATLEN = 00 otherwise the channel length is fixed to
32-bit by hardware whatever the value filled in. Not used in SPI mode.
Note: For correct operation, this bit should be configured when the I2S is disabled.

SPI_I2S prescaler register (SPI_I2SPR)

28.5.9

Address offset: 0x20
Reset value: 0000 0010 (0x0002)
15

14

13

12

Reserved

11

10

9

8

MCKOE

ODD

7

I2SDIV

rw

rw

rw

RM0090 Rev 17

6

5

4

3

2

1

0

923/1747
925

Serial peripheral interface (SPI)

RM0090

Bits 15:10 Reserved, must be kept at reset value.
Bit 9 MCKOE: Master clock output enable
0: Master clock output is disabled
1: Master clock output is enabled
Note: This bit should be configured when the I2S is disabled. It is used only when the I2S is in master
mode.
This bit is not used in SPI mode.
Bit 8 ODD: Odd factor for the prescaler
0: real divider value is = I2SDIV *2
1: real divider value is = (I2SDIV * 2)+1
Refer to Section 28.4.4: Clock generator. Not used in SPI mode.
Note: This bit should be configured when the I2S is disabled. It is used only when the I2S is in master
mode.
Bits 7:0 I2SDIV: I2S Linear prescaler
I2SDIV [7:0] = 0 or I2SDIV [7:0] = 1 are forbidden values.
Refer to Section 28.4.4: Clock generator. Not used in SPI mode.
Note: These bits should be configured when the I2S is disabled. It is used only when the I2S is in
master mode.

924/1747

RM0090 Rev 17

RM0090

28.5.10

Serial peripheral interface (SPI)

SPI register map
The table provides shows the SPI register map and reset values.

0

SPI_I2SPR

UDR

CHSIDE

TXE

RXNE

RXDMAEN

FRF
CRCERR

TXDMAEN

ERRIE
MODF

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CRCPOLY[15:0]

Reserved

0

0

0

0

0

0

0

0

0

0

RxCRC[15:0]

Reserved

0

0

0

0

0

0

0

0

0

0

0

Reserved

Reserved

Reset value

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CHLEN

Reserved

DATLEN

TxCRC[15:0]

Reset value

0x20

0

CKPOL

SPI_I2SCFGR

0

0

I2SSTD

Reset value

0

OVR

SPI_TXCRCR

0

0

Reserved

Reset value

0

I2SCFG

0x1C

SPI_RXCRCR

SSOE

RXNEIE

0

ODD

0x18

Reset value

0

0

MCKOE

0x14

SPI_CRCPR

Reserved

I2SE

0x10

Reset value

0

DR[15:0]

I2SMOD

0x0C

0

0

Reset value
SPI_DR

0

BSY

Reserved

0

0

PCMSYNC

SPI_SR

0

0

Reserved

Reset value

0x08

0

5

0

1

6
SPE

0

CPHA

7
LSBFIRST

0

2

8
SSI

0

CPOL

9
SSM

0

3

11

10
RXONLY

0

MSTR

12

DFF

0

4

13

CRCNEXT

0

Reserved

14

CRCEN

0

TXEIE

SPI_CR2

BR [2:0]

FRE

0x04

15

Reset value

16

17

18

19

20

21

22

23

24

26

27

28

29

25

Reserved

BIDIOE

SPI_CR1

BIDIMODE

0x00

Register

30

Offset

31

Table 129. SPI register map and reset values

0

0

0

0

1

0

I2SDIV
0

0

0

0

0

Refer to Section 2.3: Memory map for the register boundary addresses.

RM0090 Rev 17

925/1747
925

Serial audio interface (SAI)

29

RM0090

Serial audio interface (SAI)
This section applies to the STM32F42xxx and STM32F43xxx family.

29.1

Introduction
The SAI interface (serial audio interface) offers a wide set of audio protocols due to its
flexibility and wide range of configurations. Many stereo or mono audio applications may be
targeted. I2S standards, LSB or MSB-justified, PCM/DSP, TDM, and AC’97 protocols may
be addressed for example.
To bring this level of flexibility and configurability, the SAI contains two audio sub-blocks that
are fully independent of each other. Each audio sub-block is connected to up to 4 pins (SD,
SCK, FS, MCLK). Some of these pins can be shared if the two sub-blocks are declared as
synchronous to leave some free to be used as general purpose I/Os. The MCLK pin can be
output, or not, depending on the application, the decoder requirement and whether the
audio block is configured as the master.
The SAI can work in master or slave configuration. The audio sub-blocks can be either
receiver or transmitter and can work synchronously or not (with respect to the other one).

926/1747

RM0090 Rev 17

RM0090

29.2

Serial audio interface (SAI)

Main features
•

Two independent audio sub-blocks which can be transmitters or receivers with their
respective FIFO.

•

8-word integrated FIFOs for each audio sub-block.

•

Synchronous or asynchronous mode between the audio sub-blocks.

•

Master or slave configuration independent for both audio sub-blocks.

•

Clock generator for each audio block to target independent audio frequency sampling
when both audio sub-blocks are configured in master mode.

•

Data size configurable: 8-, 10-, 16-, 20-, 24-, 32-bit.

•

Peripheral with large configurability and flexibility allowing to target as example the
following audio protocol: I2S, LSB or MSB-justified, PCM/DSP, TDM, AC’97

•

Up to 16 slots available with configurable size and with the possibility to select which
ones are active in the audio frame.

•

Number of bits by frame may be configurable.

•

Frame synchronization active level configurable (offset, bit length, level).

•

First active bit position in the slot is configurable.

•

LSB first or MSB first for data transfer.

•

Mute mode.

•

Stereo/Mono audio frame capability.

•

Communication clock strobing edge configurable (SCK).

•

Error flags with associated interrupts if enabled respectively.

•

•

–

Overrun and underrun detection,

–

Anticipated frame synchronization signal detection in slave mode,

–

Late frame synchronization signal detection in slave mode,

–

Codec not ready for the AC’97 mode in reception.

Interruption sources when enabled:
–

Errors,

–

FIFO requests.

DMA interface with 2 dedicated channels to handle access to the dedicated integrated
FIFO of each SAI audio sub-block.

RM0090 Rev 17

927/1747
964

Serial audio interface (SAI)

29.3

RM0090

Functional block diagram
The block diagram of the SAI is shown in Figure 283.
Figure 283. Functional block diagram
$3%

6HULDO$XGLR,QWHUIDFH
6$,

$3%LQWHUIDFH

6$,B;&5

$XGLREORFN$
),)2FWUO

),)2
&ORFNJHQHUDWRU
$XGLREORFN$

&RQILJXUDWLRQ
UHJLVWHUVDQG
6WDWXVUHJLVWHU

)60

ELWVKLIWUHJLVWHU

$XGLREORFN%
6$,B&.B%

&ORFNJHQHUDWRU
$XGLREORFN%

),)2FWUO

),)2

)60

6$,B;&5

&RQILJXUDWLRQ
UHJLVWHUVDQG
6WDWXVUHJLVWHU

,2OLQH0DQDJHPHQW

6$,B&.B$

V\QFKUR
FWUORXW


LQWBVFN
LQWB)6
)6B$
6&.B$
6'B$
0&/.B$

)6B%
6&.B%
6'B%
0&/.B%

ELWVKLIWUHJLVWHU

$3%LQWHUIDFH

$3%

069

The SAI is mainly composed of two audio sub-blocks with their own clock generator. Each
audio block integrates a 32-bit shift register controlled by their own functional state machine.
Data are stored or read from the dedicated FIFO. FIFO may be accessed by the CPU, or by
DMA in order to leave the CPU free during the communication. Each audio block is
independent. They can be synchronous with each other.
An I/O line controller manages each dedicated pins for a given audio block in the SAI. If the
two blocks are synchronized, this controller reduces the number of I/Os used, freeing up an
FS pin, an SCK pin and eventually an MCLK pin, making them general purpose I/Os.
The functional state machine can be configured to address a wide range of audio protocols.
Some registers are present to set-up the desired protocols (audio frame waveform
generator).
The audio block can be a transmitter or receiver, in master or slave mode. The master mode
means the bit clock SCK and the frame synchronization signal are generated from the SAI,
whereas in slave mode, they come from another external or internal master. There is a
particular case for which the FS signal direction is not directly linked to the master or slave

928/1747

RM0090 Rev 17

RM0090

Serial audio interface (SAI)
mode definition. In AC’97 protocol, it will be an SAI output even if the SAI (link controller) is
set-up to consume the SCK clock (and so to be in Slave mode).

29.4

Main SAI modes
Each audio sub-block of the SAI can be configured to be master or slave via bit MODE[0] in
the SAI_xCR1 register of the selected audio block.
In master mode:
•

The bit clock is generated by the SAI using the clock generator on pin SCK_A or
SCK_B (depending which audio block is declared as a master in the SAI).

•

The dedicated pin SCK_x is considered as an output.

In slave mode:
•

The slave must be enabled before the master is enabled.

•

The slave audio block’s SCK clock I/O pin is considered input if it is configured in
asynchronous mode.

•

If the audio block is declared synchronous with the second audio block in the SAI, its
SCK I/O pin is released to leave it free to be used as a general purpose I/O and is
connected internally to the SCK pin of the device with which it will be synchronized.

Each audio sub-block can be independently defined as a transmitter or receiver by bit
MODE[1] in the SAI_xCR1 register of the relevant audio block. The I/O pin SD will be
defined respectively as an output or an input.
It is possible to declare two master audio blocks in the same SAI with two different MCLK
and SCK clock frequencies (they have to be declared asynchronous).
Each of the audio blocks in the SAI are enabled by bit SAIxEN in the SAI_xCR1 register. As
soon as this bit is active, the transmitter or the receiver is sensitive to the activity on the
clock line, data line and synchronization line in slave mode.
In master TX mode, enabling the audio block immediately generates the bit clock for the
external slaves even if there is no data in the FIFO, However FS signal generation is
conditioned by the presence of data in the FIFO. After the FIFO receives the first data to
transmit, this data is output to external slaves. If there is no data to transmit in the FIFO, 0
values are then sent in the audio frame with an underrun flag generation.
In slave mode, the audio frame starts when the audio block is enabled and when a start of
frame is detected.
In Slave TX mode, no underrun event is possible on the first frame after the audio block is
enabled, because the mandatory operating sequence in this case is:
1.

Write into the SAI_xDR (by software or by DMA).

2.

Wait until the FIFO threshold (FLH) flag is different from 000b (FIFO empty).

3.

Enable the audio block in slave transmitter mode.

RM0090 Rev 17

929/1747
964

Serial audio interface (SAI)

29.5

RM0090

SAI synchronization mode
Internal synchronization
An audio block can be declared synchronous with the second audio block. In this case, the
bit clock and the frame synchronization signals are shared to reduce the number of external
pins used for the communication. The audio block declared as synchronous with the other
one will see its own SCK_x, FS_x, and MCLK_x pins released to bring them back as GPIOs.
The one declared asynchronous is the one for which the I/O pins FS_x and SCK_x ad
MCLK_x (if the audio block is considered as master) are considered.
Typically, the audio block synchronous mode may be used to configure the SAI in full duplex
mode. One of the two audio blocks can be configured as master and the other as slave, or
both can be slaves; with one block declared as asynchronous (respective bit SYNCEN[1:0]
= 00 in SAI_xCR1) and the other one declared as synchronous with the other audio block
(respective bit SYNCEN[1:0] = 01 in the SAI_xCR1).

Note:

APB frequency PCLK must be greater or equal to twice the bit rate clock frequency (due to
internal resynchronization stages).

29.6

Audio data size
The audio frame can target different data sizes by configuring bit DS[2:0] in the SAI_xCR1
register. The data sizes may be 8-, 10-, 16-, 20-, 24- or 32-bit. During the transfer, either the
MSB or the LSB of the data are sent first, depending on the configuration of bit LSBFIRST in
the SAI_xCR1 register.

29.7

Frame synchronization
The FS signal acts as the Frame synchronization signal in the audio frame (start of frame).
The shape of this signal is completely configurable in order to target the different audio
protocols with their own specificities concerning this Frame synchronization behavior. This
configurability is done using register SAI_xFRCR. Figure 284 gives a view of this flexibility.
Figure 284. Audio frame
&S LENGTH    BITS
&S ACTIVE LENGTH    BITS
&S OFFSET
&S
SCK
SD

3LOT 

3LOT 

3LOT 

3LOT 

x

x

3LOT N

-36

930/1747

RM0090 Rev 17

RM0090

Serial audio interface (SAI)
In AC’97 mode (bit PRTCFG[1:0] = 10 in the SAI_xCR1 register), the frame synchronization
shape is forced to be configured to target these protocols. The SAI_xFRCR register value is
ignored.
Each audio block is independent and so each requires a specific configuration.

29.7.1

Frame length
•

Master mode: The audio frame length can be configured up to 256 bit clock, setting bit
FRL[7:0] in the SAI_xFRCR register. If the frame length is greater than the number of
declared slots for the frame, the remaining bits to transmit will be extended to 0 or the
SD line will be released to HI-z depending the state of bit TRIS in the SAI_xCR2
register (refer to Section 29.12.4). In reception mode, the remaining bit is ignored.

•

Slave mode: The audio frame length is mainly used in order to specify to the slave the
number of bit clocks per audio frame sent by the external master. It is used mainly to
detect from the master, any anticipated or late occurrence of the Frame
synchronization signal during an on-going audio frame. An error will be generated in
such case. For more details please refer to the Section 29.13.

The number of bits in the frame is equal to FRL[7:0] + 1.
The minimum number of bits to transfer in an audio frame is 8. This is the case when the
data size is 8-bit and only one slot is defined in NBSLOT[3:0] in the SAI_xSLOTR register
(NBSLOT[3:0] = 0000 for slot 0).
In master mode:
•

If bit NODIV in the SAI_xCR1 register is cleared, the frame length should be aligned to
a number equal to a power of 2, from 8 to 256. This is to ensure that an audio frame
contains an integer number of MCLK pulses per bit clock, which ensures correct
operation of the external DAC/ADC inside the decoders. If the value set in FRL[7:0]
does not respect this rule, flag WCKCFG is set when the audio block is enabled and an
interrupt is generated if bit WCKCFGIE is set in the SAI_xIM register. The SAI is
automatically disabled.

•

If bit NODIV in the SAI_xCR1 register is set, the FRL[7:0] bit can take any of the values
without constraint since the input clock of the audio block should be equal to the bit
clock. There is no MCLK_x clock which can be output. MCLK_x output pad is
automatically disabled.

In slave mode, there are no constraints for the FRL[7:0] bit configuration in the SAI_xFRCR
register.

29.7.2

Frame synchronization polarity
Bit FSPOL in the SAI_xFRCR register sets the active polarity of the FS pin from which a
frame is started. The start of frame is edge sensitive.
In slave mode, the audio block waits for a valid frame to start to transmit or to receive. Start
of frame is synchronized to this signal. It is effective only if the start of frame is not detected
during an on-going communication and assimilated to an anticipated start of frame (refer to
Section 29.13).
In master mode, the frame synchronization is sent continuously each time an audio frame is
complete until the SAIxEN bit in the SAI_xCR1 register is cleared. If no data is present in the
FIFO at the end of the previous audio frame, an underrun condition will be managed as
described in Section 29.13), but there will be no interruption in the audio communication

RM0090 Rev 17

931/1747
964

Serial audio interface (SAI)

RM0090

flow.

29.7.3

Frame synchronization active level length
Bit FSALL[6:0] in the SAI_xFRCR register configures the length of the active level of the
Frame synchronization signal. The length can be set from 1 to 128 bit clock SCK.
The active length may be half of the frame length in I2S, LSB or MSB-justified modes for
instance, or one-bit wide for PCM/DSP or TDM mode, or even 16-bit length in AC’97.

29.7.4

Frame synchronization offset
Depending on the audio protocol targeted in the application, the Frame synchronization
signal can be asserted when transmitting the last bit or the first bit of the audio frame (as for
instance, respectively, in I2S standard protocol and in MSB-justified protocol). Bit FSOFF in
the SAI_xFRCR register makes the choice.

29.7.5

FS signal role
The FS signal may have a different meaning depending on the FS function. Bit FSDEF in
the SAI_xFRCR register selects which meaning it will have. It may be either:
•

0: a start of frame, like for instance the PCM/DSP, TDM, AC’97, audio protocols,

•

1: a start of frame and a channel side identification within the audio frame like for the
I2S, the MSB or LSB-justified protocols.

When the FS signal is considered as a start of frame and channel side identification within
the frame, the number of declared slots must be considered to be half of the number for the
left channel and half of the number for the right channel. If the number of bit clock on half
audio frame is greater than the number of slots dedicated to a channel side, if TRIS = 0, 0 is
sent for transmission for the remaining bit clock in the SAI_xCR2 register, otherwise if TRIS
= 1, the SD line is released to HI-Z. In reception, the remaining bit clock are not considered
until the channel side changes.

932/1747

RM0090 Rev 17

RM0090

Serial audio interface (SAI)

Figure 285. FS role is start of frame + channel side identification (FSDEF = TRIS = 1)
1XPEHURIVORWVQRWDOLJQHGZLWKWKHDXGLRIUDPH
$XGLRIUDPH
+DOIRIIUDPH
)6
VFN
VORW

3LOT  /.

3LOT  /. 3LOT  /&& 3LOT  /.

3LOT  /&& 3LOT  /.

1XPEHURIVORWVDOLJQHGZLWKWKHDXGLRIUDPH
$XGLRIUDPH
+DOIRIIUDPH
)6
VFN
VORW

3LOT 

3LOT 

3LOT 

3LOT 

3LOT 

3LOT 
-36

1. The frame length should be even.

If bit FSDEF in SAI_xFRCR is kept clear, so FS signal is equivalent to a start of frame, and
if the number of slots defined in bit NBSLOT[3:0] in SAI_xSLOTR multiplied by
the number of bits by slot configured in bit SLOTSZ[1:0] in SAI_xSLOTR is less than the
frame size (bit FRL[7:0] in the SAI_xFRCR register), then,
•

if TRIS = 0 in the SAI_xCR2 register, the remaining bit after the last slot will be forced to
0 until the end of frame in case of transmission,

•

if TRIS = 1, the line will be released to HI-Z during the transfer of these remaining bits.
In reception mode, these bits are discarded.
Figure 286. FS role is start of frame (FSDEF = 0)
$XGLRIUDPH

VFN

VORW

6ORW

6ORW

6ORW



6ORWQ

'DWD DIWHUVORWQLI75,6 
6'RXWSXWUHOHDVHG +,= DIWHUVORWQLI75,6 
069

29.8

Slot configuration
The slot is the basic element in the audio frame. The number of slots in the audio frame is
equal to the configured setting of bit NBSLOT[3:0] in the SAI_xSLOTR register +1.
The maximum number of slots per audio frame is fixed at 16.
For AC’97 protocol (when bit PRTCFG[1:0] = 10), the number of slots is automatically set to
target the protocol specification, and the value of NBSLOT[3:0] is ignored.

RM0090 Rev 17

933/1747
964

Serial audio interface (SAI)

RM0090

Each slot can be defined as a valid slot, or not, by setting bit SLOTEN[15:0] in the
SAI_xSLOTR register. In an audio frame, during the transfer of a non-valid slot, 0 value will
be forced on the data line or the SD data line will be released to HI-z (refer to
Section 29.12.4) if the audio block is transmitter, or the received value from the end of this
slot will be ignored. Consequently, there will be no FIFO access and so no request to read
or write the FIFO linked to this inactive slot status.
The slot size is also configurable as shown in the Figure 287. The size of the slots is
selected by setting bit SLOTSZ[1:0] in the SAI_xSLOTR register. The size is applied
identically for each slot in an audio frame.
Figure 287. Slot size configuration with FBOFF = 0 in SAI_xSLOTR
$XGLREORFNLVUHFHLYHU

$XGLREORFNLVWUDQVPLWWHU

6ORWVL]H GDWDVL]H

6ORWVL]H GDWDVL]H
VORW[

VORW[

GDWDVL]H

GDWDVL]H
GDWDVL]H

GDWDVL]H
VORW[

GDWDVL]H



VORW[

GDWDVL]H
ELW

ELW
VORW[

;;



GDWDVL]H

VORW[

GDWDVL]H

;;;;

ELW

ELW
;GRQ¶WFDUH

069

It is possible to choose the position of the first data bit to transfer within the slots, this offset
is configured by bit FBOFF[5:0] in the SAI_xSLOTR register. 0 values will be injected in
transmitter mode from the beginning of the slot until this offset position is reached. In
reception, the bit in the offset phase is ignored. This feature targets the LSB justified
protocol (if the offset is equal to the slot size minus the data size).
Figure 288. First bit offset
$XGLREORFNLVUHFHLYHU

$XGLREORFNLVWUDQVPLWWHU

6ORWVL]H GDWDVL]H

6ORWVL]H GDWDVL]H
VORW[

VORW[

GDWDVL]H

GDWDVL]H
GDWDVL]H

GDWDVL]H

)%2))

)%2))

VORW[

GDWDVL]H

;;


VORW[

ELW

GDWDVL]H



VORW[

ELW

;;;;
ELW
;GRQ¶WFDUH

ELW

934/1747

;;

)%2)) 6/276='6

)%2)) 6/276='6
VORW[

GDWDVL]H

RM0090 Rev 17

GDWDVL]H

069

RM0090

Serial audio interface (SAI)
It is mandatory to respect the following conditions in order to avoid bad SAI behavior:
FBOFF ≤(SLOTSZ - DS),
DS ≤SLOTSZ,
NBSLOT x SLOTSZ ≤FRL (frame length),
The number of slots should be even when bit FSDEF in the SAI_xFRCR register is set.
In AC’97 (bit PRTCFG[1:0] = 10), the slot size is automatically set as defined in
Section 29.11.

29.9

SAI clock generator
Each audio block has its own clock generator to make these two blocks completely
independent. There is no difference in terms of functionality between these two clock
generators. They are exactly the same.
When the audio block is defined as Master, the clock generator generates the
communication clock (the bit clock) and the master clock for external decoders.
When the audio block is defined as slave, the clock generator is OFF.
Figure 289 illustrates the architecture of the audio block clock generator.
Figure 289. Audio block clock generator overview

12',9

0&/.B[

0&.',9>@

6$,B&.B[



0DVWHUFORFN
GLYLGHU



12',9



)5/>@

%LWFORFNGLYLGHU

12',9


6&.B[



069

Note:

If NoDiv is set to 1, the MCLK_x signal will be set at 0 level if this pin is configured as the
SAI pin in GPIO peripherals.
The clock source for the clock generator comes from the product clock controller. The
SAI_CK_x clock is equivalent to the master clock which may be divided for the external
decoders using bit MCKDIV[3:0]:
MCLK_x = SAI_CK_x / (MCKDIV[3:0] * 2), if MCKDIV[3:0] is not equal to 0000.
MCLK_x = SAI_CK_x, if MCKDIV[3:0] is equal to 0000.
MCLK_x signal is used only in TDM.
The division must be even in order to keep 50% on the Duty cycle on the MCLK output and
on the SCK_x clock. If bit MCKDIV[3:0] = 0000, division by one is applied to have MCLK_x
= SAI_CK_x.
In the SAI, the single ratio MCLK/FS = 256 is considered. Mostly, three frequency ranges
will be encountered as illustrated in the Table 130.

RM0090 Rev 17

935/1747
964

Serial audio interface (SAI)

RM0090

Table 130. Example of possible audio frequency sampling range
Input SAI_CK_x clock
frequency

192 kHz x 256

44.1 kHz x 256
SAI_CK_x = MCLK(1)

Most usual audio frequency
sampling achievable

MCKDIV[3:0]

192 kHz

MCKDIV[3:0] = 0000

96 kHz

MCKDIV[3:0] = 0001

48 kHz

MCKDIV[3:0] = 0010

16 kHz

MCKDIV[3:0] = 0100

8 kHz

MCKDIV[3:0] = 1000

44.1 kHz

MCKDIV[3:0] = 0000

22.05 kHz

MCKDIV[3:0] = 0001

11.025 kHz

MCKDIV[3:0] = 0010

MCLK

MCKDIV[3:0] = 0000

1. This may happen when the product clock controller selects an external clock source, instead of PLL clock.

The master clock may be generated externally on an I/O pad for external decoders if the
corresponding audio block is declared as master with bit NODIV = 0 in the SAI_xCR1
register. In slave, the value set in this last bit is ignored since the clock generator is OFF,
and the MCLK_x I/O pin is released for use as a general purpose I/O.
The bit clock is derived from the master clock. The bit clock divider sets the divider factor
between the bit clock SCK_x and the master clock MCLK_x following the formula:
SCK_x = MCLK x (FRL[7:0] +1) / 256
where:
256 is the fixed ratio between MCLK and the audio frequency sampling.
FRL[7:0] is the number of bit clock - 1 in the audio frame, configured in the SAI_xFRCR
register.
It is mandatory in master mode that (FRL[7:0] +1) should be equal to a number with a power
of 2 (refer to Section 29.7) in order to have an even integer number of MCLK_x pulses by bit
clock. The 50% duty cycle is guaranteed on the bit clock SCK_x.
The SAI_CK_x clock can be also equal to the bit clock frequency. In this case, bit NODIV in
the SAI_xCR1 register should be set and the value inside the MCKDIV divider and the bit
clock divider will be ignored. In this case, the number of bits per frame is fully configurable
without the need to be equal to a power of two.
The bit clock strobing edge on SCK can be configured by bit CKSTR in the SAI_xCR1
register.

29.10

Internal FIFOs
Each audio block in the SAI has its own FIFO. Depending if the block is defined to be a
transmitter or a receiver, the FIFO will be written or read, respectively. There is therefore
only one FIFO request linked to FREQ bit in the SAI_xSR register.

936/1747

RM0090 Rev 17

RM0090

Serial audio interface (SAI)
An interrupt is generated if FREQIE bit is enabled in the SAI_xIM register. This depends on:
•

FIFO threshold setting (FLTH bits in SAI_CR2)

•

Communication direction transmitter or receiver (see Section : Interrupt generation in
transmitter mode and Section : Interrupt generation in reception mode)

Interrupt generation in transmitter mode
The interrupt generation depends on the FIFO configuration in transmitter mode:
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO empty
(FTH[2:0] set to 000b), an interrupt is generated (FREQ bit set by hardware to 1 in
SAI_XSR register) if no data are available in SAI_xDR register (FLTH[2:0] bits in SAI_xSR
is less than 001b). This Interrupt (FREQ bit in SAI_XSR register) is cleared by hardware
when the FIFO became not empty (FLTH[2:0] bits in SAI_xSR are different from 000b) i.e
one or more data are stored in the FIFO.
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO quarter full
(FTH[2:0] set to 001b), an interrupt is generated (FREQ bit set by hardware to 1 in
SAI_XSR register) if less than a quarter of the FIFO contains data (FLTH[2:0] bits in
SAI_xSR are less than 010b). This Interrupt (FREQ bit in SAI_XSR register) is cleared by
hardware when at least a quarter of the FIFO contains data (FLTH[2:0] bits in SAI_xSR
are higher or equal to 010b).
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO half full
(FTH[2:0] set to 010b), an interrupt is generated (FREQ bit set by hardware to 1 in
SAI_XSR register) if less than half of the FIFO contains data (FLTH[2:0] bits in SAI_xSR
are less than 011b). This Interrupt (FREQ bit in SAI_XSR register) is cleared by hardware
when at least half of the FIFO contains data (FLTH[2:0] bits in SAI_xSR are higher or
equal to 011b).
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO three quarter
(FTH[2:0] set to 011b), an interrupt is generated (FREQ bit is set by hardware to 1 in
SAI_XSR register) if less than three quarters of the FIFO contain data (FLTH[2:0] bits in
SAI_xSR are less than 100b). This Interrupt (FREQ bit in SAI_XSR register) is cleared by
hardware when at least three quarters of the FIFO contain data (FLTH[2:0] bits in
SAI_xSR are higher or equal to 100b).
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO full (FTH[2:0]
set to 100b), an interrupt is generated (FREQ bit is set by hardware to 1 in SAI_XSR
register) if the FIFO is not full (FLTH[2:0] bits in SAI_xSR is less than 101b). This Interrupt
(FREQ bit in SAI_XSR register) is cleared by hardware when the FIFO is full (FLTH[2:0]
bits in SAI_xSR is equal to 101b value).

Interrupt generation in reception mode
The interrupt generation depends on the FIFO configuration in reception mode:
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO empty
(FTH[2:0] set to 000b), an interrupt is generated (FREQ bit is set by hardware to 1 in
SAI_XSR register) if at least one data is available in SAI_xDR register(FLTH[2:0] bits in
SAI_xSR is higher or equal to 001b). This Interrupt (FREQ bit in SAI_XSR register) is
cleared by hardware when the FIFO became empty (FLTH[2:0] bits in SAI_xSR is equal
to 000b) i.e no data is stored in FIFO.
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO quarter fully
(FTH[2:0] set to 001b), an interrupt is generated (FREQ bit is set by hardware to 1 in
SAI_XSR register) if at less one quarter of the FIFO data locations are available

RM0090 Rev 17

937/1747
964

Serial audio interface (SAI)

RM0090

(FLTH[2:0] bits in SAI_xSR is higher or equal to 010b). This Interrupt (FREQ bit in
SAI_XSR register) is cleared by hardware when less than a quarter of the FIFO data
locations become available (FLTH[2:0] bits in SAI_xSR is less than 010b).
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO half fully
(FTH[2:0] set to 010b value), an interrupt is generated (FREQ bit is set by hardware to 1
in SAI_XSR register) if at least half of the FIFO data locations are available (FLTH[2:0]
bits in SAI_xSR is higher or equal to 011b). This Interrupt (FREQ bit in SAI_XSR register)
is cleared by hardware when less than half of the FIFO data locations become available
(FLTH[2:0] bits in SAI_xSR is less than 011b).
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO three quarter
full(FTH[2:0] set to 011b value), an interrupt is generated (FREQ bit is set by hardware to
1 in SAI_XSR register) if at least three quarters of the FIFO data locations are available
(FLTH[2:0] bits in SAI_xSR is higher or equal to 100b). This Interrupt (FREQ bit in
SAI_XSR register) is cleared by hardware when the FIFO has less than three quarters of
the FIFO data locations avalable(FLTH[2:0] bits in SAI_xSR is less than 100b).
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO full(FTH[2:0]
set to 100b), an interrupt is generated (FREQ bit is set by hardware to 1 in SAI_XSR
register) if the FIFO is full (FLTH[2:0] bits in SAI_xSR is equal to 101b). This Interrupt
(FREQ bit in SAI_XSR register) is cleared by hardware when the FIFO is not full
(FLTH[2:0] bits in SAI_xSR is less than 101b).
Like interrupt generation, the SAI can use the DMA if DMAEN bit in the SAI_xCR1 register is
set. The FREQ bit assertion mechanism is the same as the interruption generation
mechanism described above for FREQIE.
Each FIFO is an 8-word FIFO. Each read or write operation from/to the FIFO targets one
word FIFO allocation whatever the access size. Each FIFO word contains one audio frame.
FIFO pointers are incremented by one word after each access to the SAI_xDR register.
Data should be right aligned when it is written in the SAI_xDR.
Data received will be right aligned in the SAI_xDR.
The FIFO pointers can be reinitialized when the SAI is disabled by setting bit FFLUSH in the
SAI_xCR2 register. If FFLUSH is set when the SAI is enabled the data present in the FIFO
will be lost automatically.

938/1747

RM0090 Rev 17

RM0090

29.11

Serial audio interface (SAI)

AC’97 link controller
The SAI is able to work as an AC’97 link controller. In this protocol:
•

The slot number and the slot size are fixed.

•

The frame synchronization signal is perfectly defined and has a fixed shape.

To select this protocol, set bit PRTCFG[1:0] in the SAI_xCR1 register to 10. When AC’97
mode is selected the data sizes that can be used are 16-bit or 20-bit only, else SAI behavior
is not guaranteed.
•

Bits NBSLOT[3:0] and SLOTSZ[1:0] are consequently ignored.

•

The number of slots is fixed at 13 slots. The first one is 16 bits wide and all the others
are 20 bits wide (data slots).

•

Bit FBOFF[5:0] in the SAI_xSLOTR register is ignored

•

The SAI_xFRCR register is ignored.

The FS signal from the block defined as asynchronous is configured automatically as an
output, since the AC’97 controller link drives the FS signal whatever the master or slave
configuration.
Figure 290 presents an AC’97 audio frame structure.
Figure 290. AC’97 audio frame
)6


6',

7DJ

6'2

7DJ















&0' &0'
3&0
3&0
/,1(
3&0
3&0
3&0
$''5 '$7$ /)5217 5)5217 '$& &(17(5 /6855 56855

67$786 67$786 3&0 3&0
$''5 '$7$ /()7 5,*+7

/,1(
$'&

3&0
0,&

565
9'

565
9'









3&0
/)(

/,1(
'$&

+6(7
'$&

,2
&75/

565
/9'

/,1(
$'&

+6(7

,2
67$786

069

Note:

In AC’97 protocol, bit 2 of the tag is reserved (always 0), so whatever the value written in the
SAI FIFO, bit 2 of the TAG is forced to 0 level.
For more details about TAG representation, please refer to the AC’97 protocol standard.
One SAI can be used to target an AC’97 point-to-point communication.
In receiver mode, the SAI acting as an AC’97 link controller will require no FIFO request and
so no data storage in the FIFO when the codec ready bit in the slot 0 is decoded low. If bit
CNRDYIE is enabled in the SAI_xIM register, flag CNRDY will be set in the SAI_xSR
register and an interrupt is generated. This flag is dedicated to the AC’97 protocol.

RM0090 Rev 17

939/1747
964

Serial audio interface (SAI)

29.12

RM0090

Specific features
The SAI has some specific functions which can be useful depending on the audio protocol
selected. These functions are accessible through specific bits in the SAI_xCR2 register.

29.12.1

Mute mode
Mute mode may be used when the audio block is a transmitter or receiver.

Transmitter
In transmitter mode, Mute mode can be selected at anytime. Mute mode is active for entire
audio frames. The bit MUTE in the SAI_xCR2 register requests Mute mode when it is set
during an on-going frame.
The mute mode bit is strobed only at the end of the frame. If set at this time, the mute mode
is active at the beginning of the new audio frame, for a complete frame, until the next end of
frame, it then strobes the bit to determine if the next frame will still be a mute frame.
If the number of slots set in bit NBSLOT[3:0] in the SAI_xSLOTR register is lower than or
equal to two, it is possible to specify if the value sent during the Mute mode is 0 or if it is the
last value of each slot. The selection is done via bit MUTEVAL in the SAI_xCR2 register.
If the number of slots set in bit NBSLOT[3:0] in the SAI_xSLOTR register is greater than
two, MUTEVAL bit in the SAI_xCR2 has no meaning as 0 values are sent on each bit on
each slot.
During Mute mode, the FIFO pointers are still incremented, meaning that data which was
present in the FIFO and for which the Mute mode is requested is discarded.

Receiver
In receiver mode, it is possible to detect a Mute mode sent from the external transmitter
when all the declared and valid slots of the audio frame receive 0 for a given consecutive
number of audio frames (bit MUTECNT[5:0] in the SAI_xCR2 register).
When the number of MUTE frames is detected, flag MUTEDET in the SAI_xSR register is
set and an interrupt can be generated if bit MUTEDETIE is set in the SAI_xCR2.
The mute frame counter is cleared when the audio block is disabled or when a valid slot
receives at least one data in an audio frame. The interrupt is generated just once, when the
counter reaches the specified value in bit MUTECNT[5:0]. Then the interrupt event is rearmed when the counter is cleared.

29.12.2

MONO/STEREO function
In transmission mode, it is possible to address the Mono mode without any data preprocessing in memory when the number of slot is equal to 2 (NBSLOT[3:0] = 0001 in the
SAI_xSLOTR). In such a case, the access to and from the FIFO will be reduced by two
since in transmission, the data for slot 0 is duplicated into data slot 1.
To select the Mono feature, set bit MONO in the SAI_xCR1 register.
In reception mode, bit MONO can be set and has a meaning only if the number of slots is
equal to 2 like for the transmission mode. When it is set, only the data of slot 0 will be stored
in the FIFO. The data belonging to slot 1 will be discarded since in this case, it is supposed
to be the same as the previous slot. If the data flux in reception is a real stereo audio flow

940/1747

RM0090 Rev 17

RM0090

Serial audio interface (SAI)
with a distinct and different left and right data, bit MONO has no meaning. The conversion
from the output stereo file to the equivalent mono file is done by software.

Note:

To enable Mono mode, NBSLOT and SLOTEN must equal two and MONO bit set to 1.

29.12.3

Companding mode
Telecommunication applications may require to process the data to transmit or to receive
with a data companding algorithm.
Depending on the COMP[1:0] bit in the SAI_xCR2 register (used only when TDM mode is
selected), the software may choose to process or not the data before sending it on SD serial
output line (compression) or to expand the data after the reception on SD serial input line
(expansion) as illustrated in Figure 291,.The two companding modes supported are the µLaw and the A-Law log which are a part of the CCITT G.711 recommendation.
The companding standard employed in the United States and Japan is the µ-Law and allows
14 bits of dynamic range (COMP[1:0] = 10 in the SAI_xCR2 register).
The European companding standard is A-Law and allows 13 bits of dynamic range
(COMP[1:0] = 11 in the SAI_xCR2 register).
Companding standard (µ-Law or A-Law) can be computed based on 1’s complement or 2’s
complement representation depending on the CPL bit setting in the SAI_xCR2 register.
The µ-Law and A-Law formats encode data into 8-bit code elements with MSB alignment.
Companded data is always 8 bits wide. For this reason, bit DS[2:0] in the SAI_xCR1 register
will be forced to 010 when the SAI audio block is enabled (bit SAIxEN = 1 in the SAI_xCR1
register) and when the COMP[1:0] bit selects one of these two companding modes.
If no companding processing is required, COMP[1:0] bit in the SAI_xCR2 register should be
kept cleared.
Figure 291. Data companding hardware in an audio block in the SAI
5HFHLYHUPRGH ELW02'(>@ LQ6$,B[&5
&203>@

),)2



6'

H[SDQG
ELWVKLIWUHJLVWHU



7UDQVPLWWHUPRGH ELW02'(>@ LQ6$,B[&5

),)2


FRPSUHVV

6'
ELWVKLIWUHJLVWHU



&203>@
069

Note:

Not applicable when AC’97 selected.

RM0090 Rev 17

941/1747
964

Serial audio interface (SAI)

RM0090

Expansion or compression mode is automatically selected by the SAI configuration.

29.12.4

•

If the SAI audio block is configured to be a transmitter, and if the COMP[1] bit is set in
the SAI_xCR2 register, the compression mode will be applied.

•

If the SAI audio block is declared as a receiver, the expansion algorithm will be applied.

Output data line management on an inactive slot
In transmitter mode, it is possible to choose the behavior of the SD line in output when an
inactive slot is sent on the data line (via bit TRIS in the SAI_xCR2 register when the SAI is
disabled).
•

Either the SAI forces 0 on the SD output line when an inactive slot is transmitted, or

•

The line is released in HI-z state at the end of the last bit of data transferred, to release
the line for other transmitters connected to this node.

It is important to note that the two transmitters do not attempt to drive the same SD output
pin simultaneously, which could result in a short circuit. In order to ensure a gap between
transmissions, if the data is lower than 32-bit, the data can be extended to 32-bit by setting
bit SLOTSZ[1:0] = 10 in the SAI_xSLOTR register. Then, the SD output pin will be tristated
at the end of the LSB of the active slot (during the padding to 0 phase to extend the data to
32-bit) if the following slot is declared inactive.
In addition, if the number of slots multiplied by the slot size is lower than the frame length,
the SD output line will be tristated when the padding to 0 is done to complete the audio
frame.
Figure 292 illustrates these behaviors.

942/1747

RM0090 Rev 17

RM0090

Serial audio interface (SAI)
Figure 292. Tristate strategy on SD output line on an inactive slot
%LW75,6 LQWKH6$,B[&5DQGIUDPHOHQJWK QXPEHURIVORWV
$XGLRIUDPH
SCK
6ORWVL]H GDWDVL]H
3LOT  /.

VORW
6' RXWSXW

3LOT  /&&

3LOT  /&&

$ATA 

3LOT  /.
$ATA 

 /.


 /.


3LOT N /.
$ATA M

6ORWVL]H!GDWDVL]H
VORW

3LOT  /.

6' RXWSXW
VORW

3LOT  /&&

3LOT  /&&

$ATA 

$ATA 
3LOT  /.

3LOT  /.

3LOT  /&&

3LOT  /&&

6' RXWSXW

3LOT  /.
$ATA 

 /.

 /.


 /.

 /.


3LOT N /.
$ATA M
3LOT N /.
$ATA M

%LW75,6 LQWKH6$,B[&5DQGIUDPHOHQJWK!QXPEHURIVORWV
$XGLRIUDPH
SCK
6ORWVL]H GDWDVL]H
VORW
6' RXWSXW

3LOT  /.

3LOT  /&&

3LOT  /&&

$ATA 

 /.


3LOT N /.
$ATA M

6ORWVL]H!GDWDVL]H
VORW
6' RXWSXW

VORW
6' RXWSXW

3LOT  /.

3LOT  /&&

3LOT  /&&



$ATA 

3LOT  /.

 /.

3LOT  /&&

3LOT  /&&

 /.


3LOT N /.
$ATA M

 /.
$ATA M
069

When the selected audio protocol uses the FS signal as a start of frame and a channel side
identification (bit FSDEF = 1 in the SAI_xFRCR register), the tristate mode is managed
according to Figure 293 (where bit TRIS in the SAI_xCR1 register = 1, and FSDEF=1, and
half frame length > number of slots/2, and NBSLOT=6).

RM0090 Rev 17

943/1747
964

Serial audio interface (SAI)

RM0090

Figure 293. Tristate on output data line in a protocol like I2S

SCK

6ORWVL]H GDWDVL]H
VORW
6' RXWSXW

3LOT  /.

3LOT  /&&

$ATA 

3LOT  /.

3LOT  /.

$ATA 

$ATA 

3LOT  /&&

3LOT  /.
$ATA 

6ORWVL]H!GDWDVL]H
VORW
6' RXWSXW

VORW
6' RXWSXW

3LOT  /.

3LOT  /&&

$ATA 

3LOT  /.
$ATA 

3LOT  /.

3LOT  /&&

$ATA 

3LOT  /.

3LOT  /&&

$ATA 

3LOT  /.

3LOT  /.

$ATA 

3LOT  /.
$ATA 

3LOT  /&&

3LOT  /.
$ATA M
069

If the TRIS bit in the SAI_xCR2 register is cleared, all the High impedance states on the SD
output line on Figure 292 and Figure 293 are replaced by a drive with a value of 0.

29.13

Error flags
The SAI embeds some error flags:

29.13.1

–

FIFO overrun/underrun,

–

Anticipated frame synchronization detection,

–

Late frame synchronization detection,

–

Codec not ready (AC’97 exclusively),

–

Wrong clock configuration in master mode.

FIFO overrun/underrun (OVRUDR)
The FIFO Overrun/Underrun bit is called OVRUDR in the SAI_xSR register.
The overrun or underrun errors occupy the same bit since an audio block can be either
receiver or transmitter and each audio block in an SAI has its own SAI_xSR register.

Overrun
When the audio block is configured as receiver, an overrun condition may appear if data is
received in an audio frame when the FIFO is full and is not able to store the received data.
In this case, the received data is lost, the flag OVRUDR in the SAI_xSR register is set and
an interrupt is generated if bit OVRUDRIE is set in the SAI_xIM register. The slot number
from which the overrun occurs, is stored internally. No more data will be stored into the FIFO
until it becomes free to store new data. When the FIFO has at least one data free, the SAI
audio block receiver will store new data (from new audio frame) from the slot number which

944/1747

RM0090 Rev 17

RM0090

Serial audio interface (SAI)
was stored internally when the overrun condition was detected, and this, to avoid data slot
de-alignment in the destination memory (refer to Figure 294).
The OVRUDR flag is cleared when bit COVRUDR is set in the SAI_xCLRFR register.
Figure 294. Overrun detection error
([DPSOH),)2RYHUUXQRQ6ORW
$XGLRIUDPH

$XGLRIUDPH

VFN
GDWD

6ORW21 6ORW21 3LOT  /. 6ORW21

6ORW21

21

6ORWQ21

),)2IXOO
5HFHLYHGGDWDGLVFDUGHG

'DWDVWRUHGDJDLQLQ),)2

2958'5
&2958'5 
069

Underrun
An underrun may occur when the audio block in the SAI is a transmitter and the FIFO is
empty when data needs to be transmitted (the audio block configuration (Master or Slave) is
not relevant). If an underrun is detected, the software must resynchronize data and slot.
Proceed as follows:
1.

Disable the SAI peripheral by resetting the SAIEN bit of the SAI_xCR1 register. Check
that the SAI has been disabled by reading back the SAIEN bit (SAIEN should be equal
to 0).

2.

Flush the Tx FIFO through the FFLUS bit of the SAI_xCR2 register.

3.

Re-assigned to the correct data to be transferred on the first active slot of the new
frame.

4.

Re-enabling the SAI peripheral (SAIEN bit set to 1).

The underrun event sets the OVRUDR flag in the SAI_xSR register and an interrupt is
generated if the OVRUDRIE bit is set in the SAI_xIM register. To clear this flag, set the
COVRUDR bit in the SAI_xCLRFR register.

RM0090 Rev 17

945/1747
964

Serial audio interface (SAI)

RM0090
Figure 295. FIFO underrun event
([DPSOH),)2XQGHUUXQRQ6ORW
$XGLRIUDPH

$XGLRIUDPH

VFN
3LOT SIZE  DATA SIZE
GDWD
6' RXWSXW

6ORW21

087(

087(

087(

6ORW21

21

6ORW21

),)2HPSW\

29581'
29581' 
-36

29.13.2

Anticipated frame synchronisation detection (AFSDET)
This flag AFSDET is used only in Slave mode. In master mode, it is never asserted. It
informs about the detection of a frame synchronisation (FS) earlier than expected since the
frame length, the frame polarity, the frame offset are defined and known.
Early detection sets flag AFSDET in the SAI_xSR register.
This detection has no effect on the current audio frame which is not sensitive to the
anticipated FS. This means that “parasitic” events on signal FS are flagged without any
perturbation of the current audio frame.
If bit AFSDETIE is set in the SAI_xIM register, an interrupt is generated. To clear the flag
AFSDET, bit CAFSDET in the SAI_xCLRFR register has to be set.
To resynchronize with the master after Anticipated frame detection error, four steps should
be respected:
1.

SAI block should be disabled by resetting SAIEN bit in SAI_xCR1 register, to be sure
that the SAI is disabled SAIEN bit is should be equal to 0 (reading back this bit).

2.

FIFO should be flushed via FFLUS bit in SAI_xCR2 register.

3.

Re-enabling the SAI peripheral (SAIEN bit set to 1) then the SAI.

4.

SAI block will wait for the assertion on FS to restart the synchronization with master.

Note:

This flag is not asserted in AC’97 since the SAI audio block acts as a link controller and
generates the FS signal even when declared as slave.

29.13.3

Late frame synchronization detection
Flag LFSDET in the SAI_xSR register can be set only when the SAI audio block is defined
as slave. The frame length, the frame polarity and the frame offset configuration are known
in register SAI_xFRCR.
If the external master does not send the FS signal at the expecting time (generating the
signal too late), the flag LFSDET in the SAI_xSR register will be set and an interrupt is
generated if bit LFSDETIE in the SAI_xIM register is set.
The flag is cleared when bit CLFSDET is set in the SAI_xCLRFR register.

946/1747

RM0090 Rev 17

RM0090

Serial audio interface (SAI)
The late frame synchronisation detection flag is set when the error is detected, SAI needs to
be resynchronized with the master (the four steps described above should be respected).
This detection and flag assertion can detect glitches on the SCK clock in a noisy
environment, detected by the state machine of the audio block. It could incorrectly shift the
SAI audio block state machine from one state in the current audio frame, thus corrupting the
frame.
There is no corruption if the external master is not managing the audio data frame transfer in
a continuous mode, which should not be the case for most application purposes. In this
case, flag LFSDET will be set.

Note:

This flag is not asserted in AC’97 mode since the SAI audio block acts as a link controller
and generates the FS signal even when declared as slave.

29.13.4

Codec not ready (CNRDY AC’97)
The flag CNRDY in the SAI_xSR register is relevant only if the SAI audio block is configured
to work in AC’97 mode (bit PRTCFG[1:0] = 10 in the SAI_xCR1 register). If bit CNRDYIE is
set in the SAI_xIM register, an interrupt will be generated when the flag CNRDY is set.
It is asserted when the codec is not ready to communicate during the reception of the TAG 0
(slot0) of the AC’97 audio frame. In this case, there will be no data automatically stored into
the FIFO since the codec is not ready, until the TAG 0 indicates that the codec is ready. All
the active slots defined in the SAI_xSLOTR register will be captured when the codec is
ready.
To clear the flag, bit CCNRDY in the SAI_xCLRFR register has to be set.

29.13.5

Wrong clock configuration in master mode (with NODIV = 0)
When the audio block is master (MODE[1] = 0 in the SAI_xCR1 register) and if bit NODIV in
the SAI_xCR1 is clear, the flag WCKCFG will be set if bit FRL[7:0] in the SAI_xFRCR is not
set with a proper value when the SAIxEN bit in the SAI_xCR1 register is set, in order to
respect this following rule:
( FRL[7,0] ) + 1 = 2

n

where n is in the range from 3 to 8.
If bit WCKCFGIE is set, an interrupt is generated when flag WCKCFG is set in the SAI_xSR
register. To clear the flag, set bit CWCKCFG bit in the SAI_xCLRFR register.
When bit WCKCFG is set, the audio block is automatically disabled, clearing bit SAIxEN in
the SAI_xCR1 register via hardware.
The above formula is intended to guarantee that the number of MCLK pulses by bit clock is
an even integer in the audio frame with a 50% duty cycle bit clock generation to guarantee
the good quality of the audio sounds or acquisitions.

RM0090 Rev 17

947/1747
964

Serial audio interface (SAI)

29.14

RM0090

Interrupt sources
The SAI has 7 possible interrupt sources as illustrated by Table .
Table 131. Interrupt sources
Interrupt
source

Interru
pt
group

Audio block mode

Interrupt enable

Interrupt clear

Depend on:

FREQ

FREQ

Master or Slave
Receiver or transmitter

FREQIE in
SAI_xIM register

- FIFO threshold setting
(FLTH bits in SAI_CR2)
- Communication direction
transmitter or receiver
for more details please refer
to Internal FIFOs section

OVRUDR ERROR

Master or Slave
Receiver or transmitter

OVRUDRIE in
SAI_xIM register

COVRUDR = 1 in
SAI_xCLRFR register

AFSDET

Slave
ERROR (Not used in AC’97
mode)

AFSDETIE in
SAI_xIM register

CAFSDET = 1 in
SAI_xCLRFR register

LFSDET

Slave
ERROR (Not used in AC’97
mode)

LFSDETIE in
SAI_xIM register

CLFSDET = 1 in
SAI_xCLRFR register

CNRDY

ERROR

Slave
(Only in AC’97 mode)

CNRDYIE in
SAI_xIM register

CCNRDY = 1 in
SAI_xCLRFR register

Master or slave
Receiver mode only

MUTEDETIE in
SAI_xIM register

CMUTEDET = 1 in
SAI_xCLRFR register

WCKCFGIE in
SAI_xIM register

CWCKCFG = 1 in
SAI_xCLRFR register

MUTEDE
MUTE
T

Master with NODIV = 0
WCKCFG ERROR in the SAI_xCR1
register

Below are the SAI configuration steps to follow when an interrupt occurs:

29.15

1.

Disable SAI interrupt.

2.

Configure SAI.

3.

Configure SAI interrupt source.

4.

Enable SAI.

Disabling the SAI
The audio block in the SAI can be disabled at any moment by clearing bit SAIxEN in the
SAI_xCR1 register. All the frames that have already started will be automatically completed
before the total extinction of the SAI. Bit SAIxEN in the SAI_xCR1 register will stay high until
the SAI is completely switched-off at the end of the current audio frame transfer.

948/1747

RM0090 Rev 17

RM0090

Serial audio interface (SAI)
If there is an audio block in the SAI synchronous with the other one, the one which is the
master must be disabled first.

29.16

SAI DMA interface
In order to free the CPU and to optimize the bus bandwidth, each SAI audio block has an
independent DMA interface in order to read or to write into the SAI_xDR register (to hit the
internal FIFO). There is one DMA channel per audio block following basic DMA
request/acknowledge protocol.
To configure the audio block to transfer through the DMA interface, set bit DMAEN in the
SAI_xCR1 register. The DMA request is managed directly by the FIFO controller depend of
FIFO threshold level (for more details please refer to Internal FIFOs section). DMA direction
is linked to the SAI audio block configuration:
•

If the audio block is a transmitter, the audio block’s FIFO controller outputs a DMA
request to load the FIFO with data written in the SAI_xDR register.

•

If the audio block is a receiver, the DMA request will concern read operations from the
SAI_xDR register.

Below are the SAI configuration steps followed when DMA is used:

Note:

1.

Configure SAI and FIFO Threshold level (in order to specify when the DMA request to
be launched)

2.

Configure SAI DMA channel

3.

Enable DMA

4.

Enable SAI

Before configuring the SAI block, the SAI DMA channel must be disabled.

RM0090 Rev 17

949/1747
964

Serial audio interface (SAI)

RM0090

29.17

SAI registers

29.17.1

SAI xConfiguration register 1 (SAI_xCR1) where x is A or B
Address offset: Block A: 0x004
Address offset: Block B: 0x024
Reset value: 0x0000 0040

31

30

29

28

27

26

25

24

23

14

Reserved

13

12

OutDri
MONO
v
rw

rw

11

10

SYNCEN[1:0]
rw

21

20

MCKDIV[3:0]

Reserved
15

22

9

8

CKSTR

LSBFIR
ST

rw

rw

rw

19
NODIV

rw

rw

rw

rw

rw

7

6

5

4

3

DS[2:0]
rw

rw

Res.
rw

18
Res.
2

17

16

DMAEN SAIxEN
rw

rw

1

0

PRTCFG[1:0]

MODE[1:0]

rw

rw

rw

rw

Bits 31:24 Reserved, always read as 0.
Bit 23:20 MCKDIV[3:0]: Master clock divider. These bits are set and cleared by software.
0000: Divides by 1 the master clock input.
Otherwise, The Master clock frequency is calculated accordingly to the following formula:
MCLK_x = SAI_CK_x / (MCKDIV[3:0] * 2)
These bits have no meaning when the audio block is slave.
They have to be configured when the audio block is disabled.
Bit 19 NODIV: No divider. This bit is set and cleared by software.
0: Master Clock divider is enabled
1: No divider used in the clock generator (in this case Master Clock Divider bit has no effect)
Bit 18 Reserved, always read as 0.
Bit 17 DMAEN: DMA enable. This bit is set and cleared by software.
0: DMA is disabled
1: DMA is enabled
Note: In receiver mode, the bits MODE must be configured before setting bit DMAEN to avoid a DMA
request since the audio block is transmitter after reset (default setting)
Bit 16 SAIxEN: Audio block enable where x is A or B. This bit is set by software. It is cleared by hardware,
after disabling it by software (writing the bit low), the audio is completely disabled (waiting for the end
of the current frame).
0: Audio block is disabled
1: Audio block is enabled: this bit can be set only if it is at 0 during the write operation (means the
SAI is completely disabled before being re-enabled).
This bit allows to control the state of the audio block. If it is disabled somewhere in an audio frame,
the on-going transfer will be completed and the cell will be totally disabled at the end of this audio
frame transfer.
Note: When SAIx block is configured as master mode, clock must be present on the input of the SAI
before setting SAIxEN bit.
Bits 15:14 Reserved, always read as 0.
Bit 13 OUTDRIV: Output drive. This bit is set and cleared by software.
0: Audio block output driven when SAIEN is set
1: Audio block output driven immediately after the setting of this bit.
Note: This bit has to be set before enabling the audio block but after the audio block configuration.

950/1747

RM0090 Rev 17

RM0090

Serial audio interface (SAI)

Bit 12 MONO: Mono mode. This bit is set and cleared by software.
0: Stereo mode
1: Mono mode.
This bit has a meaning only when the number of slots is equal to 2.
When the Mono mode is selected, the data of the slot 0 data is duplicated on the slot 1 when the
audio block is a transmitter. In reception mode, the slot1 is discarded and only the data received
from the slot 0 will be stored.
Refer to Section 29.12.2 for more details.
Bits 11:10 SYNCEN[1:0]: Synchronization enable. This bit is set and cleared by software.
00: audio block is asynchronous.
01: audio block is synchronous with the other internal audio block. In this case audio block should be
configured in Slave mode
10: Reserved.
11: Not used
These bits have to be configured when the audio block is disabled.
Bit 9 CKSTR: Clock strobing edge. This bit is set and cleared by software.
0: data strobing edge is falling edge of SCK
1: data strobing edge is rising edge of SCK
This bit has to be configured when the audio block is disabled.
Bit 8 LSBFIRST: Least significant bit first. This bit is set and cleared by software.
0: data is transferred with the MSB of the data first
1: data is transferred with the LSB of the data first
This bit has to be configured when the audio block is disabled.
This bit has no meaning in AC’97 audio protocol since in AC’97 data is transferred with the MSB of
the data first.
Bits 7:5 DS[2:0]: Data size. These bits are set and cleared by software.
000: Not used
001: Not used
010: 8-bit
011: 10-bit
100: 16-bit
101: 20-bit
110: 24-bit
111: 32-bit
When the companding mode is selected (bit COMP[1:0]), these DS[1:0] are ignored since the data
size is fixed to 8-bit mode by the algorithm itself.
These bits must be configured when the audio block is disabled.
Note: When AC’97 mode is selected the data sizes that can be used are: 16-bit or 20-bit only, else
SAI behavior is not guaranteed.

RM0090 Rev 17

951/1747
964

Serial audio interface (SAI)

RM0090

Bit 4 Reserved, always read as 0.
Bits 3:2 PRTCFG[1:0]: Protocol configuration. These bits are set and cleared by software.
00: Free protocol
01: Not used
10: AC’97 protocol
11: Not used
Free protocol selection allows to use the powerful configuration of the audio block to address a
specific audio protocol (like I2S, LSB/MSB justified, TDM, PCM/DSP...) setting most of the
configuration register bits as well as frame configuration register.
These bits have to be configured when the audio block is disabled.
Bits 1:0 MODE[1:0]: Audio block mode. These bits are set and cleared by software.
00: Master transmitter
01: Master receiver
10: Slave transmitter
11: Slave receiver
These bits have to be configured when the audio block is disabled.
Note: In Master transmitter mode the audio block will start to generate the FS and clocks

952/1747

RM0090 Rev 17

RM0090

Serial audio interface (SAI)

29.17.2

SAI xConfiguration register 2 (SAI_xCR2) where x is A or B
Address offset: Block A: 0x008
Address offset: Block B: 0x028
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

2

1

0

Reserved
15

14

COMP[1:0]
rw

rw

13

12

11

CPL
rw

10

9

8

7

MUTECNT[5:0]
rw

rw

rw

rw

rw

rw

6

5

4

3

MUTE
VAL

Mute

TRIS

FFLUS

rw

rw

rw

rw

FTH
rw

rw

rw

Bits 31:16 Reserved, always read as 0
Bits 15:14 COMP[1:0]: Companding mode. These bits are set and cleared by software.
00: No companding algorithm
01: Reserved.
10: µ-Law algorithm
11: A-Law algorithm
The µ-Law and the A-Law log are a part of the CCITT G.711 recommendation, the type of
complement that will be used depends on ComPLement bit.
The data expansion or data compression are determined by the state of bit MODE[0].
The data compression is applied if the audio block is configured as a transmitter.
The data expansion is automatically applied when the audio block is configured as a receiver.
Refer to Section 29.12.3 for more details.
Note: Companding mode is applicable only when TDM is selected.
Bit 13 CPL: Complement bit. This bit is set and cleared by software.
It defines the type of complement to be used for companding mode
0: 1’s complement representation.
1: 2’s complement representation.
Note: This bit has effect only when the companding mode is µ-Law algorithm or A-Law algorithm.
Bits 12:7 MUTECNT[5:0]: Mute counter. These bits are set and cleared by software.
These bits are used only in reception mode.
The value set in these bits is compared to the number of consecutive mute frames detected in
reception. When the number of mute frames is equal to this value, the flag MUTEDET will be set and
an interrupt will be generated if bit MUTEDETIE is set.
Refer to Section 29.12.1 for more details.

RM0090 Rev 17

953/1747
964

Serial audio interface (SAI)

RM0090

Bit 6 MUTEVAL: Mute value. This bit is set and cleared by software.This bit has to be written before
enabling the audio block: SAIxEN.
0: Bit value 0 is sent during the MUTE mode.
1: Last values are sent during the MUTE mode.
This bit has a meaning only when the audio block is a transmitter and when the number of slots is
lower or equal to 2 and if the MUTE bit is set.
If more slots are declared, the bit value sent during the transmission in mute mode will be equal to 0,
whatever the value of this MUTEVAL bit.
if the number of slot is lower or equal to 2 and MUTEVAL = 1, the mute value transmitted for each
slot will be the ones sent during the previous frame.
Refer to Section 29.12.1 for more details.
Bit 5 MUTE: Mute. This bit is set and cleared by software.
0: No Mute mode.
1: Mute mode enabled.
This bit has a meaning only when the audio block is a transmitter. The MUTE value is linked to the
MUTEVAL value if the number of slots is lower or equal to 2, or equal to 0 if it is greater than 2.
Refer to Section 29.12.1 for more details.
Bit 4 TRIS: Tristate management on data line. This bit is set and cleared by software.
0: SD output line is still driven by the SAI when a slot is inactive.
1: SD output line is released (HI-Z) at the end of the last data bit of the last active slot if the next one
is inactive.
This bit has a meaning only if the audio block is configured to be a transmitter.
This bit should be configured when SAI is disabled.
Refer to Section 29.12.4 for more details.
Bit 3 FFLUSH: FIFO flush. This bit is set by software. It is always read low.
0: No FIFO flush.
1: FIFO flush.
Writing 1 to the bit triggers the FIFO Flush. All the internal FIFO pointers (read and write) are
cleared.
Data still present in the FIFO will be lost in such case (no more transmission or received data lost).
This bit should be configured when SAI is disabled.
Before flushing SAI, DMA stream/interruption must be disabled
Bits 2:0 FTH: FIFO threshold. This bit is set and cleared by software.
000: FIFO empty
001: ¼ FIFO
010: ½ FIFO
011: ¾ FIFO
100: FIFO full
101: Reserved
110: Reserved
111: Reserved

954/1747

RM0090 Rev 17

RM0090

Serial audio interface (SAI)

29.17.3

SAI xFrame configuration register (SAI_XFRCR) where x is A or B
Address offset: Block A: 0x00C
Address offset: Block B: 0x02C
Reset value: 0x0000 0007

Note:
31

This register has no meaning in AC’97 audio protocol
30

29

28

27

26

25

24

23

22

21

20

19

14

13

12

11

10

9

8

7

6

5

FSALL[6:0]

Res.
rw

rw

rw

rw

rw

17

16

FSOFF FSPOL FSDEF

Reserved
15

18

4

rw

rw

r

3

2

1

0

rw

rw

rw

rw

FRL[7:0]
rw

rw

rw

rw

rw

rw

Bits 31:19 Reserved, always read as 0.
Bit 18 FSOFF: Frame synchronization offset. This bit is set and cleared by software.
0: FS is asserted on the first bit of the slot 0.
1: FS is asserted one bit before the first bit of the slot 0.
This bit has no meaning and is not used in AC’97 audio block configuration.
This bit must be configured when the audio block is disabled.
Bit 17 FSPOL: Frame synchronization polarity. This bit is set and cleared by software
0: FS is active low (falling edge)
1: FS is active high (rising edge)
This bit is used to configure the level of the start of frame on the FS signal.
This bit has no meaning and is not used in AC’97 audio block configuration.
This bit must be configured when the audio block is disabled.
Bit 16 FSDEF: Frame synchronization definition. This bit is set and cleared by software.
0: FS signal is a start frame signal
1: FS signal is a start of frame signal + channel side identification
When the bit is set, the number of slots defined in the SAI_ASLOTR register has to be even. It
means that there will be half of this number of slots dedicated for the left channel and the other slots
for the right channel (e.g: this bit has to be set for I2S or MSB/LSB-justified protocols...)
This bit has no meaning and is not used in AC’97 audio block configuration.
This bit must be configured when the audio block is disabled.

RM0090 Rev 17

955/1747
964

Serial audio interface (SAI)

RM0090

Bit 15 Reserved, always read as 0.
Bits 14:8 FSALL[6:0]: Frame synchronization active level length. These bits are set and cleared by software
The value set in these bits specifies the length in number of bit clock (SCK) + 1 (FSALL[6:0] + 1) of
the active level of the FS signal in the audio frame
These bits have no meaning and are not used in AC’97 audio block configuration.
These bits must be configured when the audio block is disabled.
Bits 7:0 FRL[7:0]: Frame length. These bits are set and cleared by software.
They define the length of the audio frame. More precisely, these bits define the number of SCK
clocks for each audio frame.
The number of bits in the frame is equal to FRL[7:0] + 1.
The minimum number of bits to transfer in an audio frame has to be equal to 8 or else the audio
block will have unexpected behavior. This is the case when the data size is 8-bit and only one slot 0
is defined in NBSLOT[4:0] in the SAI_ASLOTR register (NBSLOT[3:0] = 0000).
In master mode, if the master clock MCLK_x pin is declared as an output, the frame length should
be aligned to a number equal to a power of 2, from 8 to 256 in order to keep in an audio frame, an
integer number of MCLK pulses by bit clock for correct operation for external DAC/ADC inside the
decoders.
The Frame length should be even.
These bits have no meaning and are not used in AC’97 audio block configuration.

956/1747

RM0090 Rev 17

RM0090

Serial audio interface (SAI)

29.17.4

SAI xSlot register (SAI_xSLOTR) where x is A or B
Address offset: Block A: 0x010
Address offset: Block B: 0x030
Reset value: 0x0000 0000

Note:
31

This register has no meaning in AC’97 audio protocol
30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

SLOTEN[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

NBSLOT[3:0]

Reserved
rw

rw

rw

SLOTSZ[1:0]
rw

rw

rw

FBOFF[4:0]

Res
rw

rw

rw

Bits 31:16 SLOTEN[15:0]: Slot enable. These bits are set and cleared by software.
Each bit of the SLOTEN bits identify a slot position from 0 to 15 (maximum 16 slots)
0: Inactive slot.
1: Active slot.
These bits must be set when the audio block is disabled.
They are ignored in AC’97 mode.
Bits 15:12 Reserved, always read as 0.
Bits 11:8 NBSLOT[3:0]: Number of slots in an audio frame. These bits are set and cleared by software.
The value set in these bits register represents the number of slots + 1 in the audio frame (including
the number of inactive slots). The maximum number of slots is 16.
The number of slots should be even if bit FSDEF in the SAI_AFRCR register is set.
If the size is greater than the data size, the remaining bits will be forced to 0 if bit TRIS in the
SAI_xCR1 register is clear, otherwise they will be forced to 0 if the next slot is active or the SD line
will be forced to HI-Z if the next slot is inactive and bit TRIS = 1.
These bits must be set when the audio block is disabled.
They are ignored in AC’97 omode.
Bits 7:6 SLOTSZ[1:0]: Slot size
This bits is set and cleared by software.
00: The slot size is equivalent to the data size (specified in DS[3:0] in the SAI_ACR1 register).
01: 16-bit
10: 32-bit
11: Reserved
The slot size must be greater or equal to the data size. If this condition is not respected, the behavior
of the SAI will be undetermined.
These bits must be set when the audio block is disabled.
They are ignored in AC’97 mode.
Bit 1 Reserved, always read as 0.
Bits 4:0 FBOFF[4:0]: First bit offset
These bits are set and cleared by software.
The value set in these bits represents the position of the first data transfer bit in the slot. It represents
an offset value. During this offset phase 0 value are sent on the data line for transmission mode. For
reception mode, the received bit are discarded during the offset phase.
These bits must be set when the audio block is disabled.
They are ignored in AC’97 mode.

RM0090 Rev 17

957/1747
964

Serial audio interface (SAI)

29.17.5

RM0090

SAI xInterrupt mask register (SAI_xIM) where x is A or B
Address offset: blockA: 0x014
Address offset: block B: 0x034
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

Reserved
15

14

13

12

11

10

Reserved

9

8

7

MUT
LFSDETI AFSDET CNRDY FREQI WCKC
OVRU
EDET
E
IE
IE
E
FGIE
DRIE
IE
rw

rw

rw

rw

rw

rw

rw

Bits 31:7 Reserved, always read as 0.
Bit 6 LFSDETIE: Late frame synchronization detection interrupt enable. This bit is set and cleared by
software.
0: Interrupt is disabled
1: Interrupt is enabled
When this bit is set, an interrupt will be generated if the LFSDET bit is set in the SAI_ASR register.
This bit has no meaning in AC’97 mode. It has no meaning also if the audio block is master.
Bit 5 AFSDETIE: Anticipated frame synchronization detection interrupt enable. This bit is set and cleared by
software.
0: Interrupt is disabled
1: Interrupt is enabled
When this bit is set, an interrupt will be generated if the AFSDET bit in the SAI_ASR register is set.
This bit has no meaning in AC’97 mode. It has no meaning also if the audio block is master.
Bit 4 CNRDYIE: Codec not ready interrupt enable (ac’97). This bit is set and cleared by software.
0: Interrupt is disabled
1: Interrupt is enabled
When the interrupt is enabled, the audio block will detect in the slot 0 (tag0) of the AC’97 frame if the
codec connected on this line is ready or not. If not, the flag CNRDY in the SAI_ASR register will be
set and an interruption will be generated.
This bit has a meaning only if the AC97 mode is selected (bit PRTCFG[1:0]) and the audio block is a
receiver.
Bit 3 FREQIE: FIFO request interrupt enable. This bit is set and cleared by software.
0: Interrupt is disabled
1: Interrupt is enabled
When this bit is set, an interrupt will be generated if the FREQ bit in the SAI_ASR register is set.
In receiver mode, the bit MODE must be configured before setting bit FREQIE to avoid a parasitic
interruption since the audio block is a transmitter (default setting).

958/1747

RM0090 Rev 17

RM0090

Serial audio interface (SAI)

Bit 2 WCKCFGIE: Wrong clock configuration interrupt enable. This bit is set and cleared by software.
0: Interrupt is disabled
1: Interrupt is enabled
This bit is considered only if the audio block is configured as master (MODE[1] = 0 in the SAI_ACR1
register) and bit NODIV = 0 in the SAI_xCR1 register.
It generates an interrupt if the flag WCKCFG in the SAI_ASR register is set.
Note: This bit is used only in TDM mode and has no meaning for other modes.
Bit 1 MUTEDETIE: Mute detection interrupt enable. This bit is set and cleared by software.
0: Interrupt is disabled
1: Interrupt is enabled
When this bit is set, an interrupt will be generated if the MUTEDET bit in the SAI_ASR register is set.
This bit has a meaning only if the audio block is configured in receiver mode.
Bit 0 OVRUDRIE: Overrun/underrun interrupt enable. This bit is set and cleared by software.
0: Interrupt is disabled
1: Interrupt is enabled
When this bit is set, an interrupt will be generated if the OVRUDR bit in the SAI_ASR register is set.

RM0090 Rev 17

959/1747
964

Serial audio interface (SAI)

29.17.6

RM0090

SAI xStatus register (SAI_xSR) where x is A or B
Address offset: block A: 0x018
Address offset: block B: 0x038
Reset value: 0x0000 0008

31

30

29

28

27

26

25

24

23

22

21

20

19

18

14

13

12

11

10

9

Reserved

8

16

FLTH

Reserved
15

17

7

6

5

4

LFSDET AFSDET CNRDY
r

r

r

r

r

r

3

2

1

0

FREQ

WCKCFG

r

r

MUTED
OVRUDR
ET
r

r

Bits 31:19 Reserved, always read as 0.
Bits 18:16 FLTH: FIFO level threshold. This bit is read only. The FIFO level threshold flag is managed only by
hardware and its setting depends on SAI block configuration (transmitter or receiver mode).
If SAI block is configured as transmitter:
000: FIFO_empty
001: FIFO <= ¼ but not empty
010: ¼ < FIFO <= ½
011: ½ < FIFO <= ¾
100: ¾ < FIFO but not full
101: FIFO full
If SAI block is configured as receiver:
000: FIFO_empty
001: FIFO < ¼ but not empty
010: ¼ <= FIFO < ½
011: ½ =< FIFO < ¾
100: ¾ =< FIFO but not full
101: FIFO full
Bits 15:7 Reserved, always read as 0.
Bit 6 LFSDET: Late frame synchronization detection. This bit is read only.
0: No error.
1: Frame synchronization signal is not present at the right time.
This flag can be set only if the audio block is configured in Slave mode.
It is not used in AC’97 mode.
It may generate an interrupt if bit LFSDETIE in the SAI_xIM register is set.
This flag is cleared when the software sets bit CLFSDET in the SAI_xCLRFR register
Bit 5 AFSDET: Anticipated frame synchronization detection. This bit is read only.
0: No error.
1: Frame synchronization signal is detected earlier than expected.
This flag can be set only if the audio block is configured in Slave mode.
It is not used in AC’97.
It may generate an interrupt if bit AFSDETIE in the SAI_xIM register is set.
This flag is cleared when the software sets bit CAFSDET in the SAI_xCLRFR register

960/1747

RM0090 Rev 17

RM0090

Serial audio interface (SAI)

Bit 4 CNRDY: Codec not ready. This bit is read only.
0: The external AC’97 codec is ready
1: The external AC’97 codec is not ready
This bit is used only when the AC’97 audio protocol is selected in the SAI_xCR1 register and is
configured in receiver mode.
It may generate an interrupt if bit CNRDYIE in the SAI_xIM register is set.
This flag is cleared when the software sets bit CCNRDY in the SAI_xCLRFR register
Bit 3 FREQ: FIFO request. This bit is read only.
0: No FIFO request.
1: FIFO request to read or to write the SAI_xDR.
The request depends on the audio block configuration.
If configured in transmission, the FIFO request concerns a write request operation in the SAI_xDR.
If configured in reception, the FIFO request concerns a read request operation from the SAI_xDR.
This flag can generate an interrupt if bit FREQIE in the SAI_xIM register is set.
Bit 2 WCKCFG: Wrong clock configuration flag. This bit is read only.
0: The clock configuration is correct
1: The clock configuration does not respect the rule concerning the frame length specification
defined in Section 29.7 (configuration of FRL[7:0] bit in the SAI_x FRCR register)
This bit is used only when the audio block is master (MODE[1] = 0 in the SAI_xCR1 register) and
when NODIV = 0 in the SAI_xCR1 register.
It may generate an interrupt if bit WCKCFGIE in the SAI_xIM register is set.
This flag is cleared when the software sets bit CWCKCFG in the SAI_xCLRFR register
Bit 1 MUTEDET: Mute detection. This bit is read only.
0: No MUTE detection on the SD input line
1: MUTE value detected on the SD input line (0 value) for a specified number of consecutive audio
frame
This flag is set if consecutive 0 values are received in each slot of an audio frame and for a
consecutive number of audio frames (set in the MUTECNT bit in the SAI_xCR2 register).
It may generate an interrupt if bit MUTEDETIE in the SAI_xIM register is set.
This flag is cleared when the software sets bit CMUTEDET in the SAI_xCLRFR register.
Bit 0 OVRUDR: Overrun / underrun. This bit is read only.
0: No overrun/underrun error.
1: Overrun/underrun error detection.
The overrun condition can occur only when the audio block is configured in reception.
The underrun condition can occur only when the audio block is configured in transmission.
It may generate an interrupt if bit OVRUDRIE in the SAI_xIM register is set.
This flag is cleared when the software set bit COVRUDR bit in the SAI_xCLRFR register.

RM0090 Rev 17

961/1747
964

Serial audio interface (SAI)

29.17.7

RM0090

SAI xClear flag register (SAI_xCLRFR) where X is A or B
Address offset: block A: 0x01C
Address offset: block B: 0x03C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

5

4

3

2

1

0

Reserved
15

14

13

12

11

10

9

8

Reserved

7

6

CAFSDE
CLFSDET
T
rw

rw

CCNRDY

Reserved

CMUTE COVRUD
CWCKCFG
DET
R

rw

rw

rw

rw

Bits 31:7 Reserved, always read as 0.
Bit 6 CLFSDET: Clear late frame synchronization detection flag. This bit is write only.
Writing 1 in this bit clears the flag LFSDET in the SAI_xSR register.
It is not used in AC’97.
Reading this bit always returns the value 0.
Bit 5 .CAFSDET: Clear anticipated frame synchronization detection flag. This bit is write only.
Writing 1 in this bit clears the flag AFSDET in the SAI_xSR register.
It is not used in AC’97.
Reading this bit always returns the value 0.
Bit 4 CCNRDY: Clear codec not ready flag. This bit is write only.
Writing 1 in this bit clears the flag CNRDY in the SAI_xSR register.
This bit is used only when the AC’97 audio protocol is selected in the SAI_xCR1 register.
Reading this bit always returns the value 0.
Bit 3 Reserved, always read as 0.
Bit 2 CWCKCFG: Clear wrong clock configuration flag. This bit is write only.
Writing 1 in this bit clears the flag WCKCFG in the SAI_xSR register.
This bit is used only when the audio block is set as master (MODE[1] = 0 in the SAI_ACR1 register)
and bit NODIV = 0 in the SAI_xCR1 register.
Reading this bit always returns the value 0.
Bit 1 CMUTEDET: Mute detection flag. This bit is write only.
Writing 1 in this bit clears the flag MUTEDET in the SAI_xSR register.
Reading this bit always returns the value 0.
Bit 0 COVRUDR: Clear overrun / underrun. This bit is write only.
Writing 1 in this bit clears the flag OVRUDR in the SAI_xSR register.
Reading this bit always returns the value 0.

962/1747

RM0090 Rev 17

RM0090

Serial audio interface (SAI)

29.17.8

SAI xData register (SAI_xDR) where x is A or B
Address offset: block A: 0x020
Address offset: block B: 0x040
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

DATA[31:16]
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

DATA[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 DATA[31:0]: Data
A write into this register has the effect of loading the FIFO if the FIFO is not full.
A read from this register has to effect of draining-up the FIFO if the FIFO is not empty.

29.17.9

SAI register map
The following table summarizes the SAI registers.

MUTE

TRIS

FFLUS
0

FSALL[6:0]
0

0

0

0

0

0

0

0

0

0 0 0
SLOTS
Z[1:0}
0

0

0

RM0090 Rev 17

FTH
0

0

0

1

1

1

0

0

0

0

0

0

0

0

0

1

0

0

0

0

FRL[7:0]

NBSLOT[3:0]
0

0

FBOFF[4:0]
0

0

0

0

0
OVRUDRIE

0

Res.

MUTE VAL

CKSTR

LSBFIRST

OutDri

MONO

CPL

0

0

0

0

OVRUDR

Reserved

0

0

WCKCFG

0

0

MUTEDET

0

0

WCKCFG

0

0

0

MUTEDET

0

0

0

FREQIE

0

0

0

FREQ

0

0

0

CNRDYIE

0

MUTECN[5:0]

1

CNRDY

0

0

LFSDET

0

0

AFSDETIE Reserved

0

0

PRTCF MODE[
G[1:0] 1:0]

LFSDET

0

0

0

DS[2:0]

AFSDET

0

0

0

Reserved

0

0

Reserved

0

Reserved

0

FSDEF

0

FLVL[2:0]

0

Reserved.

SAIxEN

Res.

0

0

SYNCE
N[1:0]

Reserved

0

Reset value

Reset value

0

SLOTEN[15:0]

SAI_xIM

SAI_xSR

0

COMP[
1:0]

FSOFF

0x000Cor SAI_xFRCR
0x002C
Reset value

0x0018
or
0x0038

0

Reserved

Reset value

0x0014
or 0x0034

0

FSPOL

SAI_xCR2

0x0010 or SAI_xSLOTR
0x0030 Reset value 0

0

Reserved

0x0008
or 0x0028

0

DMAEN

0

Reset value

NODIV

SAI_xCR1

MCJDIV[3:0]

0x0004
or
0x0024

Reserved

Register
Offset and reset
value

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 132. SAI register map and reset values

0

963/1747
964

Serial audio interface (SAI)

RM0090

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Refer to Section 2.3: Memory map for the register boundary addresses.

964/1747

RM0090 Rev 17

0

0

0

Res.

0

WCKCFG

0

0
0

0

OVRUDR

0

0

MUTEDET

0

DATA[31:0]
0 0 0 0

CNRDY

Reset value
0x0020
SAI_xDR
or 0x0040 Reset value

LFSDET

0x001C SAI_xCLRFR
or 0x003C

CAFSDET

Reserved

Register
Offset and reset
value

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 132. SAI register map and reset values (continued)

0

0

0

0

RM0090

30

Universal synchronous asynchronous receiver transmitter (USART)

Universal synchronous asynchronous receiver
transmitter (USART)
This section applies to the whole STM32F4xx family, unless otherwise specified.

30.1

USART introduction
The universal synchronous asynchronous receiver transmitter (USART) offers a flexible
means of full-duplex data exchange with external equipment requiring an industry standard
NRZ asynchronous serial data format. The USART offers a very wide range of baud rates
using a fractional baud rate generator.
It supports synchronous one-way communication and half-duplex single wire
communication. It also supports the LIN (local interconnection network), Smartcard Protocol
and IrDA (infrared data association) SIR ENDEC specifications, and modem operations
(CTS/RTS). It allows multiprocessor communication.
High speed data communication is possible by using the DMA for multibuffer configuration.

30.2

USART main features
•

Full duplex, asynchronous communications

•

NRZ standard format (Mark/Space)

•

Configurable oversampling method by 16 or by 8 to give flexibility between speed and
clock tolerance

•

Fractional baud rate generator systems
–

Common programmable transmit and receive baud rate (refer to the datasheets
for the value of the baud rate at the maximum APB frequency.

•

Programmable data word length (8 or 9 bits)

•

Configurable stop bits - support for 1 or 2 stop bits

•

LIN Master Synchronous Break send capability and LIN slave break detection
capability
–

13-bit break generation and 10/11 bit break detection when USART is hardware
configured for LIN

•

Transmitter clock output for synchronous transmission

•

IrDA SIR encoder decoder
–

•

Support for 3/16 bit duration for normal mode

Smartcard emulation capability
–

The Smartcard interface supports the asynchronous protocol Smartcards as
defined in the ISO 7816-3 standards

–

0.5, 1.5 stop bits for Smartcard operation

•

Single-wire half-duplex communication

•

Configurable multibuffer communication using DMA (direct memory access)
–

•

Buffering of received/transmitted bytes in reserved SRAM using centralized DMA

Separate enable bits for transmitter and receiver

RM0090 Rev 17

965/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)
•

•

•

•

30.3

RM0090

Transfer detection flags:
–

Receive buffer full

–

Transmit buffer empty

–

End of transmission flags

Parity control:
–

Transmits parity bit

–

Checks parity of received data byte

Four error detection flags:
–

Overrun error

–

Noise detection

–

Frame error

–

Parity error

Ten interrupt sources with flags:
–

CTS changes

–

LIN break detection

–

Transmit data register empty

–

Transmission complete

–

Receive data register full

–

Idle line received

–

Overrun error

–

Framing error

–

Noise error

–

Parity error

•

Multiprocessor communication - enter into mute mode if address match does not occur

•

Wake up from mute mode (by idle line detection or address mark detection)

•

Two receiver wakeup modes: Address bit (MSB, 9th bit), Idle line

USART functional description
The interface is externally connected to another device by three pins (see Figure 296). Any
USART bidirectional communication requires a minimum of two pins: Receive Data In (RX)
and Transmit Data Out (TX):
RX: Receive Data Input is the serial data input. Oversampling techniques are used for data
recovery by discriminating between valid incoming data and noise.
TX: Transmit Data Output. When the transmitter is disabled, the output pin returns to its I/O
port configuration. When the transmitter is enabled and nothing is to be transmitted, the TX
pin is at high level. In single-wire and smartcard modes, this I/O is used to transmit and
receive the data (at USART level, data are then received on SW_RX).

966/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)
Through these pins, serial data is transmitted and received in normal USART mode as
frames comprising:
•

An Idle Line prior to transmission or reception

•

A start bit

•

A data word (8 or 9 bits) least significant bit first

•

0.5,1, 1.5, 2 Stop bits indicating that the frame is complete

•

This interface uses a fractional baud rate generator - with a 12-bit mantissa and 4-bit
fraction

•

A status register (USART_SR)

•

Data Register (USART_DR)

•

A baud rate register (USART_BRR) - 12-bit mantissa and 4-bit fraction.

•

A Guardtime Register (USART_GTPR) in case of Smartcard mode.

Refer to Section 30.6: USART registers on page 1007 for the definitions of each bit.
The following pin is required to interface in synchronous mode:
•

CK: Transmitter clock output. This pin outputs the transmitter data clock for
synchronous transmission corresponding to SPI master mode (no clock pulses on start
bit and stop bit, and a software option to send a clock pulse on the last data bit). In
parallel data can be received synchronously on RX. This can be used to control
peripherals that have shift registers (e.g. LCD drivers). The clock phase and polarity
are software programmable. In smartcard mode, CK can provide the clock to the
smartcard.

The following pins are required in Hardware flow control mode:
•

CTS: Clear To Send blocks the data transmission at the end of the current transfer
when high

•

RTS: Request to send indicates that the USART is ready to receive a data (when low).

RM0090 Rev 17

967/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

Figure 296. USART block diagram
35'$7$

3:'$7$
:ULWH

5HDG
&38RU'0$

&38RU'0$
7;
5;
6:B5;

'DWDUHJLVWHU '5

7UDQVPLWGDWDUHJLVWHU 7'5

5HFHLYHGDWDUHJLVWHU 5'5

7UDQVPLW6KLIW5HJLVWHU

5HFHLYH6KLIW5HJLVWHU

,U'$
6,5
(1'(&
EORFN

,5'$B287
,5'$B,1

*735
*7

36&

&5

6&/.FRQWURO

'0$7 '0$5 6&(1 1$&. +' ,5/3 ,5(1

/,1(

6723>@ &.(1 &32/ &3+$ /%&/

&5

&5
8(

86$57$GGUHVV

576
&76

&.

&5

0

:$.( 3&(

36

3(,(

+DUGZDUH
IORZ
FRQWUROOHU
:DNHXS
XQLW

7UDQVPLW
FRQWURO

5HFHLYHU
FORFN

5HFHLYHU
FRQWURO

65

&5
,'/( 7(
7;(,( 7&,( 5;1(
,(
,(

&76 /%'

5( 5:8 6%.

7;( 7& 5;1( ,'/( 25( 1) )( 3(

86$57
LQWHUUXSW
FRQWURO

86$57B%55

&5
29(5

7UDQVPLWWHU
FORFN

>[ 29(5 @

7UDQVPLWWHUUDWH
FRQWURO

7(



86$57',9

6$03/,1*
',9,'(5

',9B0DQWLVVD


I3&/.[ [ 

',9B)UDFWLRQ




5HFHLYHUUDWH
FRQWURO

5(

&RQYHQWLRQDOEDXGUDWHJHQHUDWRU

86$57',9 ',9B0DQWLVVD ',9B)UDFWLRQî ±29(5

968/1747

RM0090 Rev 17

DLE

RM0090

30.3.1

Universal synchronous asynchronous receiver transmitter (USART)

USART character description
Word length may be selected as being either 8 or 9 bits by programming the M bit in the
USART_CR1 register (see Figure 297).
The TX pin is in low state during the start bit. It is in high state during the stop bit.
An Idle character is interpreted as an entire frame of “1”s followed by the start bit of the
next frame which contains data (The number of “1” ‘s will include the number of stop bits).
A Break character is interpreted on receiving “0”s for a frame period. At the end of the
break frame the transmitter inserts either 1 or 2 stop bits (logic “1” bit) to acknowledge the
start bit.
Transmission and reception are driven by a common baud rate generator, the clock for each
is generated when the enable bit is set respectively for the transmitter and receiver.
The details of each block is given below.
Figure 297. Word length programming
ELWZRUGOHQJWK 0ELWLVVHW 6WRSELW
3RVVLEOH
3DULW\
ELW

'DWDIUDPH
6WDUW
ELW

%LW

%LW

%LW

%LW

%LW

%LW

%LW

%LW

%LW

1H[WGDWDIUDPH
6WRS
ELW

1H[W
6WDUW
ELW

&ORFN
6WDUW
ELW

,GOHIUDPH

6WRS
ELW

%UHDNIUDPH

6WDUW
ELW

/%&/ELWFRQWUROVODVWGDWDFORFNSXOVH
ELWZRUGOHQJWK 0ELWLVUHVHW 6WRSELW
3RVVLEOH
3DULW\
ELW

'DWDIUDPH
6WDUW
ELW

%LW

%LW

%LW

%LW

%LW

%LW

%LW

%LW

1H[WGDWDIUDPH
6WRS
ELW

1H[W
6WDUW
ELW

&ORFN
6WDUW
ELW

,GOHIUDPH
%UHDNIUDPH

6WRS 6WDUW
ELW
ELW

/%&/ELWFRQWUROVODVWGDWDFORFNSXOVH
069

RM0090 Rev 17

969/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

30.3.2

RM0090

Transmitter
The transmitter can send data words of either 8 or 9 bits depending on the M bit status.
When the transmit enable bit (TE) is set, the data in the transmit shift register is output on
the TX pin and the corresponding clock pulses are output on the CK pin.

Character transmission
During an USART transmission, data shifts out least significant bit first on the TX pin. In this
mode, the USART_DR register consists of a buffer (TDR) between the internal bus and the
transmit shift register (see Figure 296).
Every character is preceded by a start bit which is a logic level low for one bit period. The
character is terminated by a configurable number of stop bits.
The following stop bits are supported by USART: 0.5, 1, 1.5 and 2 stop bits.
Note:

The TE bit should not be reset during transmission of data. Resetting the TE bit during the
transmission will corrupt the data on the TX pin as the baud rate counters will get frozen.
The current data being transmitted will be lost.
An idle frame will be sent after the TE bit is enabled.

Configurable stop bits
The number of stop bits to be transmitted with every character can be programmed in
Control register 2, bits 13,12.
•

1 stop bit: This is the default value of number of stop bits.

•

2 Stop bits: This will be supported by normal USART, single-wire and modem modes.

•

0.5 stop bit: To be used when receiving data in Smartcard mode.

•

1.5 stop bits: To be used when transmitting and receiving data in Smartcard mode.

An idle frame transmission will include the stop bits.
A break transmission will be 10 low bits followed by the configured number of stop bits
(when m = 0) and 11 low bits followed by the configured number of stop bits (when m = 1). It
is not possible to transmit long breaks (break of length greater than 10/11 low bits).

970/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)
Figure 298. Configurable stop bits
ELW:RUGOHQJWK 0ELWLVUHVHW

3RVVLEOH
SDULW\
ELW

'DWDIUDPH
6WDUW
%LW

%LW

%LW

%LW

%LW

%LW

%LW

%LW

1H[WGDWDIUDPH
6WRS
ELW

%LW

1H[W
VWDUW

&/2&.
/%&/ELWFRQWUROVODVWGDWDFORFNSXOVH

D 6WRS%LW
3RVVLEOH
3DULW\
%LW

'DWDIUDPH
6WDUW
%LW

%LW

%LW

%LW

%LW

%LW

%LW

%LW

3RVVLEOH
SDULW\
ELW

'DWDIUDPH
%LW

%LW

%LW

%LW

%LW

%LW

%LW

'DWDIUDPH
%LW

%LW

%LW

%LW
3RVVLEOH
SDULW\
ELW

F 6WRS%LWV
6WDUW
%LW

%LW

VWRSELWV

E VWRS%LWV

6WDUW
%LW

1H[WGDWDIUDPH
1H[W
VWDUW
ELW

%LW

%LW

%LW

%LW

G 6WRS%LW

%LW

1H[WGDWDIUDPH
1H[W
6WDUW
%LW

6WRS%LWV

1H[W
6WDUW
%LW

1H[WGDWDIUDPH

6WRS%LW
06Y9

Procedure:
1.

Enable the USART by writing the UE bit in USART_CR1 register to 1.

2.

Program the M bit in USART_CR1 to define the word length.

3.

Program the number of stop bits in USART_CR2.

4.

Select DMA enable (DMAT) in USART_CR3 if Multi buffer Communication is to take
place. Configure the DMA register as explained in multibuffer communication.

5.

Select the desired baud rate using the USART_BRR register.

6.

Set the TE bit in USART_CR1 to send an idle frame as first transmission.

7.

Write the data to send in the USART_DR register (this clears the TXE bit). Repeat this
for each data to be transmitted in case of single buffer.

8.

After writing the last data into the USART_DR register, wait until TC=1. This indicates
that the transmission of the last frame is complete. This is required for instance when
the USART is disabled or enters the Halt mode to avoid corrupting the last
transmission.

Single byte communication
Clearing the TXE bit is always performed by a write to the data register.
The TXE bit is set by hardware and it indicates:
•

The data has been moved from TDR to the shift register and the data transmission has
started.

•

The TDR register is empty.

•

The next data can be written in the USART_DR register without overwriting the
previous data.

This flag generates an interrupt if the TXEIE bit is set.

RM0090 Rev 17

971/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

When a transmission is taking place, a write instruction to the USART_DR register stores
the data in the TDR register and which is copied in the shift register at the end of the current
transmission.
When no transmission is taking place, a write instruction to the USART_DR register places
the data directly in the shift register, the data transmission starts, and the TXE bit is
immediately set.
If a frame is transmitted (after the stop bit) and the TXE bit is set, the TC bit goes high. An
interrupt is generated if the TCIE bit is set in the USART_CR1 register.
After writing the last data into the USART_DR register, it is mandatory to wait for TC=1
before disabling the USART or causing the microcontroller to enter the low-power mode
(see Figure 299: TC/TXE behavior when transmitting).
The TC bit is cleared by the following software sequence:

Note:

1.

A read from the USART_SR register

2.

A write to the USART_DR register

The TC bit can also be cleared by writing a ‘0 to it. This clearing sequence is recommended
only for Multibuffer communication.
Figure 299. TC/TXE behavior when transmitting
,GOHSUHDPEOH

)UDPH

)UDPH

)UDPH

7;OLQH
VHWE\KDUGZDUH
FOHDUHGE\VRIWZDUH

7;(IODJ

86$57B'5

)

VHWE\KDUGZDUH
FOHDUHGE\VRIWZDUH

)

6HWE\KDUGZDUH

)

VHW
E\KDUGZDUH

7&IODJ
6RIWZDUH
HQDEOHVWKH
86$57

6RIWZDUHZDLWVXQWLO7;( 
DQGZULWHV)LQWR'5

6RIWZDUHZDLWVXQWLO7;( 
DQGZULWHV)LQWR'5

7&LVQRWVHW
EHFDXVH7;( 

6RIWZDUHZDLWVXQWLO7;( 
DQGZULWHV)LQWR'5

7&LVQRWVHW
EHFDXVH7;( 

7&LVQRWVHW
EHFDXVH7;( 

6RIWZDUHZDLWXQWLO7& 
069

Break characters
Setting the SBK bit transmits a break character. The break frame length depends on the M
bit (see Figure 297).
If the SBK bit is set to ‘1 a break character is sent on the TX line after completing the current
character transmission. This bit is reset by hardware when the break character is completed
(during the stop bit of the break character). The USART inserts a logic 1 bit at the end of the
last break frame to guarantee the recognition of the start bit of the next frame.
Note:

972/1747

If the software resets the SBK bit before the commencement of break transmission, the
break character will not be transmitted. For two consecutive breaks, the SBK bit should be
set after the stop bit of the previous break.

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)

Idle characters
Setting the TE bit drives the USART to send an idle frame before the first data frame.

30.3.3

Receiver
The USART can receive data words of either 8 or 9 bits depending on the M bit in the
USART_CR1 register.

Start bit detection
The start bit detection sequence is the same when oversampling by 16 or by 8.
In the USART, the start bit is detected when a specific sequence of samples is recognized.
This sequence is: 1 1 1 0 X 0 X 0 X 0 0 0 0.
Figure 300. Start bit detection when oversampling by 16 or 8

Note:

If the sequence is not complete, the start bit detection aborts and the receiver returns to the
idle state (no flag is set) where it waits for a falling edge.
The start bit is confirmed (RXNE flag set, interrupt generated if RXNEIE=1) if the 3 sampled
bits are at 0 (first sampling on the 3rd, 5th and 7th bits finds the 3 bits at 0 and second
sampling on the 8th, 9th and 10th bits also finds the 3 bits at 0).
The start bit is validated (RXNE flag set, interrupt generated if RXNEIE=1) but the NE noise
flag is set if, for both samplings, at least 2 out of the 3 sampled bits are at 0 (sampling on the

RM0090 Rev 17

973/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

3rd, 5th and 7th bits and sampling on the 8th, 9th and 10th bits). If this condition is not met,
the start detection aborts and the receiver returns to the idle state (no flag is set).
If, for one of the samplings (sampling on the 3rd, 5th and 7th bits or sampling on the 8th, 9th
and 10th bits), 2 out of the 3 bits are found at 0, the start bit is validated but the NE noise
flag bit is set.

Character reception
During an USART reception, data shifts in least significant bit first through the RX pin. In this
mode, the USART_DR register consists of a buffer (RDR) between the internal bus and the
received shift register.
Procedure:
1.

Enable the USART by writing the UE bit in USART_CR1 register to 1.

2.

Program the M bit in USART_CR1 to define the word length.

3.

Program the number of stop bits in USART_CR2.

4.

Select DMA enable (DMAR) in USART_CR3 if multibuffer communication is to take
place. Configure the DMA register as explained in multibuffer communication. STEP 3

5.

Select the desired baud rate using the baud rate register USART_BRR

6.

Set the RE bit USART_CR1. This enables the receiver which begins searching for a
start bit.

When a character is received

Note:

•

The RXNE bit is set. It indicates that the content of the shift register is transferred to the
RDR. In other words, data has been received and can be read (as well as its
associated error flags).

•

An interrupt is generated if the RXNEIE bit is set.

•

The error flags can be set if a frame error, noise or an overrun error has been detected
during reception.

•

In multibuffer, RXNE is set after every byte received and is cleared by the DMA read to
the Data Register.

•

In single buffer mode, clearing the RXNE bit is performed by a software read to the
USART_DR register. The RXNE flag can also be cleared by writing a zero to it. The
RXNE bit must be cleared before the end of the reception of the next character to avoid
an overrun error.

The RE bit should not be reset while receiving data. If the RE bit is disabled during
reception, the reception of the current byte will be aborted.

Break character
When a break character is received, the USART handles it as a framing error.

Idle character
When an idle frame is detected, there is the same procedure as a data received character
plus an interrupt if the IDLEIE bit is set.

974/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)

Overrun error
An overrun error occurs when a character is received when RXNE has not been reset. Data
can not be transferred from the shift register to the RDR register until the RXNE bit is
cleared.
The RXNE flag is set after every byte received. An overrun error occurs if RXNE flag is set
when the next data is received or the previous DMA request has not been serviced. When
an overrun error occurs:

Note:

•

The ORE bit is set.

•

The RDR content will not be lost. The previous data is available when a read to
USART_DR is performed.

•

The shift register will be overwritten. After that point, any data received during overrun
is lost.

•

An interrupt is generated if either the RXNEIE bit is set or both the EIE and DMAR bits
are set.

•

The ORE bit is reset by a read to the USART_SR register followed by a USART_DR
register read operation.

The ORE bit, when set, indicates that at least 1 data has been lost. There are two
possibilities:
•

if RXNE=1, then the last valid data is stored in the receive register RDR and can be
read,

•

if RXNE=0, then it means that the last valid data has already been read and thus there
is nothing to be read in the RDR. This case can occur when the last valid data is read in
the RDR at the same time as the new (and lost) data is received. It may also occur
when the new data is received during the reading sequence (between the USART_SR
register read access and the USART_DR read access).

Selecting the proper oversampling method
The receiver implements different user-configurable oversampling techniques (except in
synchronous mode) for data recovery by discriminating between valid incoming data and
noise.
The oversampling method can be selected by programming the OVER8 bit in the
USART_CR1 register and can be either 16 or 8 times the baud rate clock (Figure 301 and
Figure 302).
Depending on the application:
•

select oversampling by 8 (OVER8=1) to achieve higher speed (up to fPCLK/8). In this
case the maximum receiver tolerance to clock deviation is reduced (refer to
Section 30.3.5: USART receiver tolerance to clock deviation on page 988)

•

select oversampling by 16 (OVER8=0) to increase the tolerance of the receiver to clock
deviations. In this case, the maximum speed is limited to maximum fPCLK/16

RM0090 Rev 17

975/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

Programming the ONEBIT bit in the USART_CR3 register selects the method used to
evaluate the logic level. There are two options:
•

the majority vote of the three samples in the center of the received bit. In this case,
when the 3 samples used for the majority vote are not equal, the NF bit is set

•

a single sample in the center of the received bit
Depending on the application:
–

select the three samples’ majority vote method (ONEBIT=0) when operating in a
noisy environment and reject the data when a noise is detected (refer to
Figure 133) because this indicates that a glitch occurred during the sampling.

–

select the single sample method (ONEBIT=1) when the line is noise-free to
increase the receiver’s tolerance to clock deviations (see Section 30.3.5: USART
receiver tolerance to clock deviation on page 988). In this case the NF bit will
never be set.

When noise is detected in a frame:
•

The NF bit is set at the rising edge of the RXNE bit.

•

The invalid data is transferred from the Shift register to the USART_DR register.

•

No interrupt is generated in case of single byte communication. However this bit rises
at the same time as the RXNE bit which itself generates an interrupt. In case of
multibuffer communication an interrupt will be issued if the EIE bit is set in the
USART_CR3 register.

The NF bit is reset by a USART_SR register read operation followed by a USART_DR
register read operation.
Note:

Oversampling by 8 is not available in the Smartcard, IrDA and LIN modes. In those modes,
the OVER8 bit is forced to ‘0 by hardware.
Figure 301. Data sampling when oversampling by 16
5;OLQH
VDPSOHGYDOXHV
6DPSOHFORFN





































2QHELWWLPH

06Y9

976/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)
Figure 302. Data sampling when oversampling by 8

5;OLQH
VDPSOHGYDOXHV
6DPSOH
FORFN [





















2QHELWWLPH

06Y9

Table 133. Noise detection from sampled data
Sampled value

NE status

Received bit value

000

0

0

001

1

0

010

1

0

011

1

1

100

1

0

101

1

1

110

1

1

111

0

1

Framing error
A framing error is detected when:
The stop bit is not recognized on reception at the expected time, following either a desynchronization or excessive noise.
When the framing error is detected:
•

The FE bit is set by hardware

•

The invalid data is transferred from the Shift register to the USART_DR register.

•

No interrupt is generated in case of single byte communication. However this bit rises
at the same time as the RXNE bit which itself generates an interrupt. In case of
multibuffer communication an interrupt will be issued if the EIE bit is set in the
USART_CR3 register.

The FE bit is reset by a USART_SR register read operation followed by a USART_DR
register read operation.

RM0090 Rev 17

977/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

Configurable stop bits during reception
The number of stop bits to be received can be configured through the control bits of Control
Register 2 - it can be either 1 or 2 in normal mode and 0.5 or 1.5 in Smartcard mode.

30.3.4

1.

0.5 stop bit (reception in Smartcard mode): No sampling is done for 0.5 stop bit. As
a consequence, no framing error and no break frame can be detected when 0.5 stop bit
is selected.

2.

1 stop bit: Sampling for 1 stop Bit is done on the 8th, 9th and 10th samples.

3.

1.5 stop bits (Smartcard mode): When transmitting in smartcard mode, the device
must check that the data is correctly sent. Thus the receiver block must be enabled (RE
=1 in the USART_CR1 register) and the stop bit is checked to test if the smartcard has
detected a parity error. In the event of a parity error, the smartcard forces the data
signal low during the sampling - NACK signal-, which is flagged as a framing error.
Then, the FE flag is set with the RXNE at the end of the 1.5 stop bit. Sampling for 1.5
stop bits is done on the 16th, 17th and 18th samples (1 baud clock period after the
beginning of the stop bit). The 1.5 stop bit can be decomposed into 2 parts: one 0.5
baud clock period during which nothing happens, followed by 1 normal stop bit period
during which sampling occurs halfway through. Refer to Section 30.3.11: Smartcard on
page 997 for more details.

4.

2 stop bits: Sampling for 2 stop bits is done on the 8th, 9th and 10th samples of the
first stop bit. If a framing error is detected during the first stop bit the framing error flag
will be set. The second stop bit is not checked for framing error. The RXNE flag will be
set at the end of the first stop bit.

Fractional baud rate generation
The baud rate for the receiver and transmitter (Rx and Tx) are both set to the same value as
programmed in the Mantissa and Fraction values of USARTDIV.
Equation 1: Baud rate for standard USART (SPI mode included)
f CK
Tx/Rx baud = ------------------------------------------------------------------------------------8 × ( 2 – OVER8 ) × USARTDIV

Equation 2: Baud rate in Smartcard, LIN and IrDA modes
f CK
Tx/Rx baud = ---------------------------------------------16 × USARTDIV

USARTDIV is an unsigned fixed point number that is coded on the USART_BRR register.

Note:

978/1747

•

When OVER8=0, the fractional part is coded on 4 bits and programmed by the
DIV_fraction[3:0] bits in the USART_BRR register

•

When OVER8=1, the fractional part is coded on 3 bits and programmed by the
DIV_fraction[2:0] bits in the USART_BRR register, and bit DIV_fraction[3] must be kept
cleared.

The baud counters are updated to the new value in the baud registers after a write operation
to USART_BRR. Hence the baud rate register value should not be changed during
communication.

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)

How to derive USARTDIV from USART_BRR register values when OVER8=0
Example 1:
If DIV_Mantissa = 0d27 and DIV_Fraction = 0d12 (USART_BRR = 0x1BC), then
Mantissa (USARTDIV) = 0d27
Fraction (USARTDIV) = 12/16 = 0d0.75
Therefore USARTDIV = 0d27.75
Example 2:
To program USARTDIV = 0d25.62
This leads to:
DIV_Fraction = 16*0d0.62 = 0d9.92
The nearest real number is 0d10 = 0xA
DIV_Mantissa = mantissa (0d25.620) = 0d25 = 0x19
Then, USART_BRR = 0x19A hence USARTDIV = 0d25.625
Example 3:
To program USARTDIV = 0d50.99
This leads to:
DIV_Fraction = 16*0d0.99 = 0d15.84
The nearest real number is 0d16 = 0x10 => overflow of DIV_frac[3:0] => carry must be
added up to the mantissa
DIV_Mantissa = mantissa (0d50.990 + carry) = 0d51 = 0x33
Then, USART_BRR = 0x330 hence USARTDIV = 0d51.000

How to derive USARTDIV from USART_BRR register values when OVER8=1
Example 1:
If DIV_Mantissa = 0x27 and DIV_Fraction[2:0]= 0d6 (USART_BRR = 0x1B6), then
Mantissa (USARTDIV) = 0d27
Fraction (USARTDIV) = 6/8 = 0d0.75
Therefore USARTDIV = 0d27.75
Example 2:
To program USARTDIV = 0d25.62
This leads to:
DIV_Fraction = 8*0d0.62 = 0d4.96
The nearest real number is 0d5 = 0x5
DIV_Mantissa = mantissa (0d25.620) = 0d25 = 0x19

RM0090 Rev 17

979/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

Then, USART_BRR = 0x195 => USARTDIV = 0d25.625
Example 3:
To program USARTDIV = 0d50.99
This leads to:
DIV_Fraction = 8*0d0.99 = 0d7.92
The nearest real number is 0d8 = 0x8 => overflow of the DIV_frac[2:0] => carry must be
added up to the mantissa
DIV_Mantissa = mantissa (0d50.990 + carry) = 0d51 = 0x33
Then, USART_BRR = 0x0330 => USARTDIV = 0d51.000
Table 134. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 12 MHz,
oversampling by 16(1)
Oversampling by 16 (OVER8=0)
Baud rate7

fPCLK = 8 MHz

fPCLK = 12 MHz

Value
% Error =
programmed
(Calculated in the baud Desired) B.rate /
rate register
Desired B.rate

Actual

Value
programmed
in the baud
rate register

% Error

S.No

Desired

Actual

1

1.2 KBps

1.2 KBps

416.6875

0

1.2 KBps

625

0

2

2.4 KBps

2.4 KBps

208.3125

0.01

2.4 KBps

312.5

0

3

9.6 KBps

9.604 KBps

52.0625

0.04

9.6 KBps

78.125

0

4

19.2 KBps

19.185 KBps

26.0625

0.08

19.2 KBps

39.0625

0

5

38.4 KBps

38.462 KBps

13

0.16

38.339 KBps

19.5625

0.16

6

57.6 KBps

57.554 KBps

8.6875

0.08

57.692 KBps

13

0.16

7

115.2 KBps 115.942 KBps

4.3125

0.64

115.385 KBps

6.5

0.16

8

230.4 KBps 228.571 KBps

2.1875

0.79

230.769 KBps

3.25

0.16

9

460.8 KBps 470.588 KBps

1.0625

2.12

461.538 KBps

1.625

0.16

10

921.6 KBps

NA

NA

NA

NA

NA

NA

11

2 MBps

NA

NA

NA

NA

NA

NA

12

3 MBps

NA

NA

NA

NA

NA

NA

1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.

980/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)

Table 135. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK =12 MHz,
oversampling by 8(1)
Oversampling by 8 (OVER8 = 1)
Baud rate

fPCLK = 8 MHz

fPCLK = 12 MHz

% Error =
Value
(Calculated programmed
Desired)
in the baud
B.rate /
rate register
Desired
B.rate

Value
programmed
% Error
in the baud
rate register

S.No

Desired

Actual

1

1.2 KBps

1.2 KBps

833.375

0

1.2 KBps

1250

0

2

2.4 KBps

2.4 KBps

416.625

0.01

2.4 KBps

625

0

3

9.6 KBps

9.604 KBps

104.125

0.04

9.6 KBps

156.25

0

4

19.2 KBps

19.185 KBps

52.125

0.08

19.2 KBps

78.125

0

5

38.4 KBps

38.462 KBps

26

0.16

38.339 KBps

39.125

0.16

6

57.6 KBps

57.554 KBps

17.375

0.08

57.692 KBps

26

0.16

7

115.2 KBps

115.942 KBps

8.625

0.64

115.385 KBps

13

0.16

8

230.4 KBps

228.571 KBps

4.375

0.79

230.769 KBps

6.5

0.16

9

460.8 KBps

470.588 KBps

2.125

2.12

461.538 KBps

3.25

0.16

10

921.6 KBps

888.889 KBps

1.125

3.55

923.077 KBps

1.625

0.16

11

2 MBps

NA

NA

NA

NA

NA

NA

12

3 MBps

NA

NA

NA

NA

NA

NA

Actual

1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.

Table 136. Error calculation for programmed baud rates at fPCLK = 16 MHz or fPCLK = 24 MHz,
oversampling by 16(1)
Oversampling by 16 (OVER8 = 0)
Baud rate

fPCLK

= 16 MHz

fPCLK

Value
% Error =
programmed
(Calculated in the baud Desired) B.rate /
rate register Desired B.rate

= 24 MHz

Value
programmed
% Error
in the baud
rate register

S.No

Desired

Actual

1

1.2 KBps

1.2 KBps

833.3125

0

1.2

1250

0

2

2.4 KBps

2.4 KBps

416.6875

0

2.4

625

0

3

9.6 KBps

9.598 KBps

104.1875

0.02

9.6

156.25

0

4

19.2 KBps

19.208 KBps

52.0625

0.04

19.2

78.125

0

5

38.4 KBps

38.369 KBps

26.0625

0.08

38.4

39.0625

0

6

57.6 KBps

57.554 KBps

17.375

0.08

57.554

26.0625

0.08

RM0090 Rev 17

Actual

981/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

Table 136. Error calculation for programmed baud rates at fPCLK = 16 MHz or fPCLK = 24 MHz,
oversampling by 16(1) (continued)
Oversampling by 16 (OVER8 = 0)
Baud rate

fPCLK

= 16 MHz

fPCLK

Value
% Error =
programmed
(Calculated in the baud Desired) B.rate /
rate register Desired B.rate

= 24 MHz

Value
programmed
% Error
in the baud
rate register

S.No

Desired

Actual

7

115.2 KBps

115.108 KBps

8.6875

0.08

115.385

13

0.16

8

230.4 KBps

231.884 KBps

4.3125

0.64

230.769

6.5

0.16

9

460.8 KBps

457.143 KBps

2.1875

0.79

461.538

3.25

0.16

10

921.6 KBps

941.176 KBps

1.0625

2.12

923.077

1.625

0.16

11

2 MBps

NA

NA

NA

NA

NA

NA

12

3 MBps

NA

NA

NA

NA

NA

NA

Actual

1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.

Table 137. Error calculation for programmed baud rates at fPCLK = 16 MHz or fPCLK = 24 MHz,
oversampling by 8(1)
Oversampling by 8 (OVER8=1)
Baud rate

fPCLK = 16 MHz

fPCLK = 24 MHz

% Error =
Value
(Calculated programmed
in the baud Desired) B.rate /
rate register
Desired B.rate

Value
programmed
% Error
in the baud
rate register

S.No

Desired

Actual

1

1.2 KBps

1.2 KBps

1666.625

0

1.2 KBps

2500

0

2

2.4 KBps

2.4 KBps

833.375

0

2.4 KBps

1250

0

3

9.6 KBps

9.598 KBps

208.375

0.02

9.6 KBps

312.5

0

4

19.2 KBps

19.208 KBps

104.125

0.04

19.2 KBps

156.25

0

5

38.4 KBps

38.369 KBps

52.125

0.08

38.4 KBps

78.125

0

6

57.6 KBps

57.554 KBps

34.75

0.08

57.554 KBps

52.125

0.08

7

115.2 KBps

115.108 KBps

17.375

0.08

115.385 KBps

26

0.16

8

230.4 KBps

231.884 KBps

8.625

0.64

230.769 KBps

13

0.16

9

460.8 KBps

457.143 KBps

4.375

0.79

461.538 KBps

6.5

0.16

10

921.6 KBps

941.176 KBps

2.125

2.12

923.077 KBps

3.25

0.16

11

2 MBps

2000 KBps

1

0

2000 KBps

1.5

0

12

3 MBps

NA

NA

NA

3000 KBps

1

0

Actual

1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.

982/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)

Table 138. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 16 MHz,
oversampling by 16(1)
Oversampling by 16 (OVER8=0)
Baud rate

S.No

fPCLK = 8 MHz

Desired

Actual

Value
programme
d in the
baud rate
register

fPCLK = 16 MHz

% Error =
(Calculated Desired)B.Rate
/Desired B.Rate

Value
programmed
% Error
in the baud
rate register

Actual

1.

2.4 KBps

2.400 KBps

208.3125

0.00%

2.400 KBps

416.6875

0.00%

2.

9.6 KBps

9.604 KBps

52.0625

0.04%

9.598 KBps

104.1875

0.02%

3.

19.2 KBps

19.185 KBps

26.0625

0.08%

19.208 KBps

52.0625

0.04%

4.

57.6 KBps

57.554 KBps

8.6875

0.08%

57.554 KBps

17.3750

0.08%

5.

115.2 KBps

115.942 KBps 4.3125

0.64%

115.108 KBps

8.6875

0.08%

6.

230.4 KBps

228.571
KBps

2.1875

0.79%

231.884 KBps

4.3125

0.64%

7.

460.8 KBps

470.588
KBps

1.0625

2.12%

457.143 KBps

2.1875

0.79%

8.

896 KBps

NA

NA

NA

888.889 KBps

1.1250

0.79%

9.

921.6 KBps

NA

NA

NA

941.176 KBps

1.0625

2.12%

10.

1.792 MBps

NA

NA

NA

NA

NA

NA

11.

1.8432 MBps NA

NA

NA

NA

NA

NA

12.

3.584 MBps

NA

NA

NA

NA

NA

NA

13.

3.6864 MBps NA

NA

NA

NA

NA

NA

14.

7.168 MBps

NA

NA

NA

NA

NA

NA

15.

7.3728 MBps NA

NA

NA

NA

NA

NA

1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.

Table 139. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 16 MHz,
oversampling by 8(1)
Oversampling by 8 (OVER8=1)
Baud rate

S.No

Desired

fPCLK = 8 MHz

Actual

Value
% Error =
programmed
(Calculated in the baud Desired)B.Rate
rate register /Desired B.Rate

fPCLK = 16 MHz

Actual

Value
programmed
in the baud
rate register

%
Error

1.

2.4 KBps

2.400 KBps

416.625

0.01%

2.400 KBps

833.375

0.00%

2.

9.6 KBps

9.604 KBps

104.125

0.04%

9.598 KBps

208.375

0.02%

3.

19.2 KBps

19.185 KBps

52.125

0.08%

19.208 KBps

104.125

0.04%

RM0090 Rev 17

983/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

Table 139. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 16 MHz,
oversampling by 8(1) (continued)
Oversampling by 8 (OVER8=1)
Baud rate

S.No

fPCLK = 8 MHz

Desired

Actual

4.

57.6 KBps

57.557 KBps

5.

115.2 KBps

6.

fPCLK = 16 MHz

Value
% Error =
programmed
(Calculated in the baud Desired)B.Rate
rate register /Desired B.Rate
17.375

Value
programmed
in the baud
rate register

Actual

%
Error

0.08%

57.554 KBps

34.750

0.08%

115.942 KBps 8.625

0.64%

115.108 KBps 17.375

0.08%

230.4 KBps

228.571 KBps 4.375

0.79%

231.884 KBps 8.625

0.64%

7.

460.8 KBps

470.588 KBps 2.125

2.12%

457.143 KBps 4.375

0.79%

8.

896 KBps

888.889 KBps 1.125

0.79%

888.889 KBps 2.250

0.79%

9.

921.6 KBps

888.889 KBps 1.125

3.55%

941.176 KBps 2.125

2.12%

10.

1.792 MBps

NA

NA

NA

1.7777 MBps

1.125

0.79%

11.

1.8432 MBps

NA

NA

NA

1.7777 MBps

1.125

3.55%

12.

3.584 MBps

NA

NA

NA

NA

NA

NA

13.

3.6864 MBps

NA

NA

NA

NA

NA

NA

14.

7.168 MBps

NA

NA

NA

NA

NA

NA

15.

7.3728 MBps

NA

NA

NA

NA

NA

NA

1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.

Table 140. Error calculation for programmed baud rates at fPCLK = 30 MHz or fPCLK = 60 MHz,
oversampling by 16(1)(2)
Oversampling by 16 (OVER8=0)
Baud rate

S.No

Desired

fPCLK = 30 MHz

Actual

Value
% Error =
programmed
(Calculated in the baud Desired)B.Rate
rate register /Desired B.Rate

fPCLK = 60 MHz

Actual

Value
programmed
in the baud
rate register

%
Error

1.

2.4 KBps

2.400 KBps

781.2500

0.00%

2.400 KBps

1562.5000

0.00%

2.

9.6 KBps

9.600 KBps

195.3125

0.00%

9.600 KBps

390.6250

0.00%

3.

19.2 KBps

19.194 KBps

97.6875

0.03%

19.200 KBps

195.3125

0.00%

4.

57.6 KBps

57.582KBps

32.5625

0.03%

57.582 KBps

65.1250

0.03%

5.

115.2 KBps

115.385 KBps

16.2500

0.16%

115.163 KBps 32.5625

0.03%

6.

230.4 KBps

230.769 KBps

8.1250

0.16%

230.769KBps

0.16%

7.

460.8 KBps

461.538 KBps

4.0625

0.16%

461.538 KBps 8.1250

0.16%

8.

896 KBps

909.091 KBps

2.0625

1.46%

895.522 KBps 4.1875

0.05%

984/1747

RM0090 Rev 17

16.2500

RM0090

Universal synchronous asynchronous receiver transmitter (USART)

Table 140. Error calculation for programmed baud rates at fPCLK = 30 MHz or fPCLK = 60 MHz,
oversampling by 16(1)(2) (continued)
Oversampling by 16 (OVER8=0)
Baud rate

S.No

fPCLK = 30 MHz

fPCLK = 60 MHz

Value
% Error =
programmed
(Calculated in the baud Desired)B.Rate
rate register /Desired B.Rate

Value
programmed
in the baud
rate register

%
Error

Desired

Actual

9.

921.6 KBps

909.091 KBps

2.0625

1.36%

923.077 KBps 4.0625

0.16%

10.

1.792 MBps

1.1764 MBps

1.0625

1.52%

1.8182 MBps

2.0625

1.36%

11.

1.8432
MBps

1.8750 MBps

1.0000

1.73%

1.8182 MBps

2.0625

1.52%

12.

3.584 MBps

NA

NA

NA

3.2594 MBps

1.0625

1.52%

13.

3.6864
MBps

NA

NA

NA

3.7500 MBps

1.0000

1.73%

14.

7.168 MBps

NA

NA

NA

NA

NA

NA

15.

7.3728
MBps

NA

NA

NA

NA

NA

NA

Actual

1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.
2. Only USART1 and USART6 are clocked with PCLK2. Other USARTs are clocked with PCLK1. Refer to the device
datasheets for the maximum values for PCLK1 and PCLK2.

Table 141. Error calculation for programmed baud rates at fPCLK = 30 MHz or fPCLK = 60 MHz,
oversampling by 8(1) (2)
Oversampling by 8 (OVER8=1)
Baud rate

S.No

Desired

fPCLK = 30 MHz

Actual

Value
% Error =
programmed (Calculated in the baud Desired)B.Rate
rate register /Desired B.Rate

fPCLK =60 MHz

Actual

Value
programmed
in the baud
rate register

%
Error

1.

2.4 KBps

2.400 KBps

1562.5000

0.00%

2.400 KBps

3125.0000

0.00%

2.

9.6 KBps

9.600 KBps

390.6250

0.00%

9.600 KBps

781.2500

0.00%

3.

19.2 KBps

19.194 KBps

195.3750

0.03%

19.200 KBps

390.6250

0.00%

4.

57.6 KBps

57.582 KBps

65.1250

0.16%

57.582 KBps

130.2500

0.03%

5.

115.2 KBps

115.385 KBps

32.5000

0.16%

115.163 KBps 65.1250

0.03%

6.

230.4 KBps

230.769 KBps

16.2500

0.16%

230.769 KBps 32.5000

0.16%

7.

460.8 KBps

461.538 KBps

8.1250

0.16%

461.538 KBps 16.2500

0.16%

8.

896 KBps

909.091 KBps

4.1250

1.46%

895.522 KBps 8.3750

0.05%

9.

921.6 KBps

909.091 KBps

4.1250

1.36%

923.077 KBps 8.1250

0.16%

10.

1.792 MBps

1.7647 MBps

2.1250

1.52%

1.8182 MBps

1.46%

RM0090 Rev 17

4.1250

985/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

Table 141. Error calculation for programmed baud rates at fPCLK = 30 MHz or fPCLK = 60 MHz,
oversampling by 8(1) (2) (continued)
Oversampling by 8 (OVER8=1)
Baud rate

S.No

fPCLK = 30 MHz

Desired

Actual

Value
% Error =
programmed (Calculated in the baud Desired)B.Rate
rate register /Desired B.Rate

fPCLK =60 MHz
Value
programmed
in the baud
rate register

Actual

%
Error

11.

1.8432 MBps

1.8750 MBps

2.0000

1.73%

1.8182 MBps

4.1250

1.36%

12.

3.584 MBps

3.7500 MBps

1.0000

4.63%

3.5294 MBps

2.1250

1.52%

13.

3.6864 MBps

3.7500 MBps

1.0000

1.73%

3.7500 MBps

2.0000

1.73%

14.

7.168 MBps

NA

NA

NA

7.5000 MBps

1.0000

4.63%

15.

7.3728 MBps

NA

NA

NA

7.5000 MBps

1.0000

1.73%

1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.
2. Only USART1 and USART6 are clocked with PCLK2. Other USARTs are clocked with PCLK1. Refer to the device
datasheets for the maximum values for PCLK1 and PCLK2.

Table 142. Error calculation for programmed baud rates at fPCLK = 42 MHz or fPCLK = 84 Hz,
oversampling by 16(1)(2)
Oversampling by 16 (OVER8=0)
Baud rate

S.No

fPCLK = 42 MHz

Desired

Actual

Value
% Error =
programmed
(Calculated in the baud Desired)B.Rate
rate register /Desired B.Rate

fPCLK = 84 MHz

Actual

Value
programmed
in the baud
rate register

%
Error

1.

1.2 KBps

1.2 KBps

2187.5

0

1.2 KBps

NA

0

2.

2.4 KBps

2.4 KBps

1093.75

0

2.4 KBps

2187.5

0

3.

9.6 KBps

9.6 KBps

273.4375

0

9.6 KBps

546.875

0

4.

19.2 KBps

19.195 KBps

136.75

0.02

19.2 KBps

273.4375

0

5.

38.4 KBps

38.391 KBps

68.375

0.02

38.391 KBps

136.75

0.02

6.

57.6 KBps

57.613 KBps

45.5625

0.02

57.613 KBps

91.125

0.02

7.

115.2 KBps

115.068 KBps

22.8125

0.11

115.226 KBps 45.5625

0.02

8.

230.4 KBps

230.769 KBps

11.375

0.16

230.137 KBps 22.8125

0.11

9.

460.8 KBps

461.538 KBps

5.6875

0.16

461.538 KBps 11.375

0.16

10.

921.6 KBps

913.043 KBps

2.875

0.93

923.076 KBps 5.6875

0.93

11.

1.792 MBps

1.826 MBps

1.4375

1.9

1.787 MBps

2.9375

0.27

12.

1.8432 MBps 1.826 MBps

1.4375

0.93

1.826 MBps

2.875

0.93

13.

3.584 MBps

N.A

N.A

N.A

3.652 MBps

1.4375

1.9

14.

3.6864 MBps N.A

N.A

N.A

3.652 MBps

1.4375

0.93

986/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)

Table 142. Error calculation for programmed baud rates at fPCLK = 42 MHz or fPCLK = 84 Hz,
oversampling by 16(1)(2) (continued)
Oversampling by 16 (OVER8=0)
Baud rate

S.No

fPCLK = 42 MHz

Desired

Actual

fPCLK = 84 MHz

Value
% Error =
programmed
(Calculated in the baud Desired)B.Rate
rate register /Desired B.Rate

Value
programmed
in the baud
rate register

Actual

%
Error

15.

7.168 MBps

N.A

N.A

N.A

N.A

N.A

N.A

16.

7.3728 MBps N.A

N.A

N.A

N.A

N.A

N.A

17.

9 MBps

N.A

N.A

N.A

N.A

N.A

N.A

18.

10.5 MBps

N.A

N.A

N.A

N.A

N.A

N.A

1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can be
fixed with these data.
2. Only USART1 and USART6 are clocked with PCLK2. Other USARTs are clocked with PCLK1. Refer to the device
datasheets for the maximum values for PCLK1 and PCLK2.

Table 143. Error calculation for programmed baud rates at fPCLK = 42 MHz or fPCLK = 84 MHz,
oversampling by 8(1)(2)
Oversampling by 8 (OVER8=1)
Baud rate

S.No

fPCLK = 42 MHz

Desired

Actual

Value
programmed
in the baud
rate register

fPCLK = 84 MHz
Value
programmed
in the baud
rate register

Actual

Value
programmed
in the baud
rate register

%
Error

1.

2.4 KBps

2.4 KBps

2187.5

0

2.4 KBps

NA

0

2.

9.6 KBps

9.6 KBps

546.875

0

9.6 KBps

1093.75

0

3.

19.2 KBps

19.195 KBps

273.5

0.02

19.2 KBps

546.875

0

4.

38.4 KBps

38.391 KBps

136.75

0.02

38.391 KBps

273.5

0.02

5.

57.6 KBps

57.613 KBps

91.125

0.02

57.613 KBps

182.25

0.02

6.

115.2 KBps

115.068 KBps

45.625

0.11

115.226 KBps 91.125

0.02

7.

230.4 KBps

230.769 KBps

22.75

0.11

230.137 KBps 45.625

0.11

8.

460.8 KBps

461.538 KBps

11.375

0.16

461.538 KBps 22.75

0.16

9.

921.6 KBps

913.043 KBps

5.75

0.93

923.076 KBps 11.375

0.93

10.

1.792 MBps

1.826 MBps

2.875

1.9

1.787Mbps

5.875

0.27

11.

1.8432 MBps

1.826 MBps

2.875

0.93

1.826 MBps

5.75

0.93

12.

3.584 MBps

3.5 MBps

1.5

2.34

3.652 MBps

2.875

1.9

13.

3.6864 MBps

3.82 MBps

1.375

3.57

3.652 MBps

2.875

0.93

14.

7.168 MBps

N.A

N.A

N.A

7 MBps

1.5

2.34

15.

7.3728 MBps

N.A

N.A

N.A

7.636 MBps

1.375

3.57

RM0090 Rev 17

987/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

Table 143. Error calculation for programmed baud rates at fPCLK = 42 MHz or fPCLK = 84 MHz,
oversampling by 8(1)(2) (continued)
Oversampling by 8 (OVER8=1)
Baud rate

S.No

fPCLK = 42 MHz

Desired

Actual

Value
programmed
in the baud
rate register

fPCLK = 84 MHz
Value
programmed
in the baud
rate register

Value
programmed
in the baud
rate register

Actual

%
Error

16.

9 MBps

N.A

N.A

N.A

9.333 MBps

1.125

3.7

17.

10.5 MBps

N.A

N.A

N.A

10.5 MBps

1

0

1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.
2. Only USART1 and USART6 are clocked with PCLK2. Other USARTs are clocked with PCLK1. Refer to the device
datasheets for the maximum values for PCLK1 and PCLK2.

30.3.5

USART receiver tolerance to clock deviation
The USART asynchronous receiver works correctly only if the total clock system deviation is
smaller than the USART receiver’s tolerance. The causes which contribute to the total
deviation are:
•

DTRA: Deviation due to the transmitter error (which also includes the deviation of the
transmitter’s local oscillator)

•

DQUANT: Error due to the baud rate quantization of the receiver

•

DREC: Deviation of the receiver’s local oscillator

•

DTCL: Deviation due to the transmission line (generally due to the transceivers which
can introduce an asymmetry between the low-to-high transition timing and the high-tolow transition timing)

DTRA + DQUANT + DREC + DTCL < USART receiver’s tolerance
The USART receiver’s tolerance to properly receive data is equal to the maximum tolerated
deviation and depends on the following choices:
•

10- or 11-bit character length defined by the M bit in the USART_CR1 register

•

oversampling by 8 or 16 defined by the OVER8 bit in the USART_CR1 register

•

use of fractional baud rate or not

•

use of 1 bit or 3 bits to sample the data, depending on the value of the ONEBIT bit in
the USART_CR3 register
Table 144. USART receiver’s tolerance when DIV fraction is 0
M bit

988/1747

OVER8 bit = 0

OVER8 bit = 1

ONEBIT=0

ONEBIT=1

ONEBIT=0

ONEBIT=1

0

3.75%

4.375%

2.50%

3.75%

1

3.41%

3.97%

2.27%

3.41%

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)
Table 145. USART receiver tolerance when DIV_Fraction is different from 0
M bit

OVER8 bit = 0

OVER8 bit = 1

ONEBIT=0

ONEBIT=1

ONEBIT=0

ONEBIT=1

0

3.33%

3.88%

2%

3%

1

3.03%

3.53%

1.82%

2.73%

Note:

The figures specified in Table 144 and Table 145 may slightly differ in the special case when
the received frames contain some Idle frames of exactly 10-bit times when M=0 (11-bit times
when M=1).

30.3.6

Multiprocessor communication
There is a possibility of performing multiprocessor communication with the USART (several
USARTs connected in a network). For instance one of the USARTs can be the master, its TX
output is connected to the RX input of the other USART. The others are slaves, their
respective TX outputs are logically ANDed together and connected to the RX input of the
master.
In multiprocessor configurations it is often desirable that only the intended message
recipient should actively receive the full message contents, thus reducing redundant USART
service overhead for all non addressed receivers.
The non addressed devices may be placed in mute mode by means of the muting function.
In mute mode:
•

None of the reception status bits can be set.

•

All the receive interrupts are inhibited.

•

The RWU bit in USART_CR1 register is set to 1. RWU can be controlled automatically
by hardware or written by the software under certain conditions.

The USART can enter or exit from mute mode using one of two methods, depending on the
WAKE bit in the USART_CR1 register:
•

Idle Line detection if the WAKE bit is reset,

•

Address Mark detection if the WAKE bit is set.

Idle line detection (WAKE=0)
The USART enters mute mode when the RWU bit is written to 1.
It wakes up when an Idle frame is detected. Then the RWU bit is cleared by hardware but
the IDLE bit is not set in the USART_SR register. RWU can also be written to 0 by software.
An example of mute mode behavior using Idle line detection is given in Figure 303.

RM0090 Rev 17

989/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

Figure 303. Mute mode using Idle line detection
5;1(

'DWD 'DWD 'DWD 'DWD

5;

,'/(

0XWHPRGH

5:8
5:8ZULWWHQWR

5;1(

'DWD 'DWD

1RUPDOPRGH
,GOHIUDPHGHWHFWHG

06Y9

Address mark detection (WAKE=1)
In this mode, bytes are recognized as addresses if their MSB is a ‘1 else they are
considered as data. In an address byte, the address of the targeted receiver is put on the 4
LSB. This 4-bit word is compared by the receiver with its own address which is programmed
in the ADD bits in the USART_CR2 register.
The USART enters mute mode when an address character is received which does not
match its programmed address. In this case, the RWU bit is set by hardware. The RXNE
flag is not set for this address byte and no interrupt nor DMA request is issued as the
USART would have entered mute mode.
It exits from mute mode when an address character is received which matches the
programmed address. Then the RWU bit is cleared and subsequent bytes are received
normally. The RXNE bit is set for the address character since the RWU bit has been
cleared.
The RWU bit can be written to as 0 or 1 when the receiver buffer contains no data (RXNE=0
in the USART_SR register). Otherwise the write attempt is ignored.
An example of mute mode behavior using address mark detection is given in Figure 304.
Figure 304. Mute mode using address mark detection
,QWKLVH[DPSOHWKHFXUUHQWDGGUHVVRIWKHUHFHLYHULV
SURJUDPPHGLQWKH86$57B&5UHJLVWHU
5;1(

,'/(

5;

$GGU  'DWD 'DWD

5:8

,'/(

5;1(

$GGU  'DWD 'DWD $GGU  'DWD

0XWHPRGH

1RUPDOPRGH
0DWFKLQJDGGUHVV

5:8ZULWWHQWR
5;1(ZDVFOHDUHG

5;1(

0XWHPRGH

1RQPDWFKLQJDGGUHVV

1RQPDWFKLQJDGGUHVV
06Y9

990/1747

RM0090 Rev 17

RM0090

30.3.7

Universal synchronous asynchronous receiver transmitter (USART)

Parity control
Parity control (generation of parity bit in transmission and parity checking in reception) can
be enabled by setting the PCE bit in the USART_CR1 register. Depending on the frame
length defined by the M bit, the possible USART frame formats are as listed in Table 146.
Table 146. Frame formats
M bit

PCE bit

USART frame(1)

0

0

| SB | 8 bit data | STB |

0

1

| SB | 7-bit data | PB | STB |

1

0

| SB | 9-bit data | STB |

1

1

| SB | 8-bit data PB | STB |

1. Legends: SB: start bit, STB: stop bit, PB: parity bit.

Even parity
The parity bit is calculated to obtain an even number of “1s” inside the frame made of the 7
or 8 LSB bits (depending on whether M is equal to 0 or 1) and the parity bit.
E.g.: data=00110101; 4 bits set => parity bit will be 0 if even parity is selected (PS bit in
USART_CR1 = 0).

Odd parity
The parity bit is calculated to obtain an odd number of “1s” inside the frame made of the 7 or
8 LSB bits (depending on whether M is equal to 0 or 1) and the parity bit.
E.g.: data=00110101; 4 bits set => parity bit will be 1 if odd parity is selected (PS bit in
USART_CR1 = 1).

Parity checking in reception
If the parity check fails, the PE flag is set in the USART_SR register and an interrupt is
generated if PEIE is set in the USART_CR1 register. The PE flag is cleared by a software
sequence (a read from the status register followed by a read or write access to the
USART_DR data register).
Note:

In case of wakeup by an address mark: the MSB bit of the data is taken into account to
identify an address but not the parity bit. And the receiver does not check the parity of the
address data (PE is not set in case of a parity error).

Parity generation in transmission
If the PCE bit is set in USART_CR1, then the MSB bit of the data written in the data register
is transmitted but is changed by the parity bit (even number of “1s” if even parity is selected
(PS=0) or an odd number of “1s” if odd parity is selected (PS=1)).
Note:

The software routine that manages the transmission can activate the software sequence
which clears the PE flag (a read from the status register followed by a read or write access
to the data register). When operating in half-duplex mode, depending on the software, this
can cause the PE flag to be unexpectedly cleared.

RM0090 Rev 17

991/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

30.3.8

RM0090

LIN (local interconnection network) mode
The LIN mode is selected by setting the LINEN bit in the USART_CR2 register. In LIN
mode, the following bits must be kept cleared:
•

STOP[1:0] and CLKEN in the USART_CR2 register

•

SCEN, HDSEL and IREN in the USART_CR3 register.

LIN transmission
The same procedure explained in Section 30.3.2 has to be applied for LIN Master
transmission than for normal USART transmission with the following differences:
•

Clear the M bit to configure 8-bit word length.

•

Set the LINEN bit to enter LIN mode. In this case, setting the SBK bit sends 13 ‘0 bits
as a break character. Then a bit of value ‘1 is sent to allow the next start detection.

LIN reception
A break detection circuit is implemented on the USART interface. The detection is totally
independent from the normal USART receiver. A break can be detected whenever it occurs,
during Idle state or during a frame.
When the receiver is enabled (RE=1 in USART_CR1), the circuit looks at the RX input for a
start signal. The method for detecting start bits is the same when searching break
characters or data. After a start bit has been detected, the circuit samples the next bits
exactly like for the data (on the 8th, 9th and 10th samples). If 10 (when the LBDL = 0 in
USART_CR2) or 11 (when LBDL=1 in USART_CR2) consecutive bits are detected as ‘0,
and are followed by a delimiter character, the LBD flag is set in USART_SR. If the LBDIE
bit=1, an interrupt is generated. Before validating the break, the delimiter is checked for as it
signifies that the RX line has returned to a high level.
If a ‘1 is sampled before the 10 or 11 have occurred, the break detection circuit cancels the
current detection and searches for a start bit again.
If the LIN mode is disabled (LINEN=0), the receiver continues working as normal USART,
without taking into account the break detection.
If the LIN mode is enabled (LINEN=1), as soon as a framing error occurs (i.e. stop bit
detected at ‘0, which will be the case for any break frame), the receiver stops until the break
detection circuit receives either a ‘1, if the break word was not complete, or a delimiter
character if a break has been detected.
The behavior of the break detector state machine and the break flag is shown on the
Figure 305: Break detection in LIN mode (11-bit break length - LBDL bit is set) on page 993.
Examples of break frames are given on Figure 306: Break detection in LIN mode vs.
Framing error detection on page 994.

992/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)
Figure 305. Break detection in LIN mode (11-bit break length - LBDL bit is set)
&DVHEUHDNVLJQDOQRWORQJHQRXJK !EUHDNGLVFDUGHG/%'LVQRWVHW
%UHDNIUDPH

5;OLQH
&DSWXUHVWUREH
%UHDNVWDWH
PDFKLQH

,GOH

%LW %LW %LW %LW


5HDGVDPSOHV







%LW %LW %LW






%LW %LW %LW %LW






,GOH



&DVHEUHDNVLJQDOMXVWORQJHQRXJK !EUHDNGHWHFWHG/%'LVVHW
%UHDNIUDPH

5;OLQH

'HOLPLWHULVLPPHGLDWH
&DSWXUHVWUREH
%UHDNVWDWH
PDFKLQH

,GOH

%LW %LW %LW %LW


5HDGVDPSOHV







%LW %LW %LW






%LW %LW %LW %






,GOH



/%'

&DVHEUHDNVLJQDOORQJHQRXJK !EUHDNGHWHFWHG/%'LVVHW
%UHDNIUDPH

5;OLQH
&DSWXUHVWUREH
%UHDNVWDWH
,GOH
PDFKLQH
5HDGVDPSOHV

%LW %LW %LW %LW








%LW %LW %LW






%LW %LW %LW %LW ZDLWGHOLPLWHU ,GOH








/%'

06Y9

RM0090 Rev 17

993/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

Figure 306. Break detection in LIN mode vs. Framing error detection
&DVHEUHDNRFFXUULQJDIWHUDQ,GOH
5;OLQH

GDWD

,'/(

%5($.
GDWDWLPH

GDWD [

GDWD KHDGHU

GDWDWLPH

5;1()(
/%')

&DVHEUHDNRFFXUULQJZKLOHGDWDLVEHLQJUHFHLYHG
5;OLQH

GDWD

GDWD

%5($.

GDWDWLPH

GDWD [

GDWD KHDGHU

GDWDWLPH

5;1()(
/%')
06Y9

30.3.9

USART synchronous mode
The synchronous mode is selected by writing the CLKEN bit in the USART_CR2 register to
1. In synchronous mode, the following bits must be kept cleared:
•

LINEN bit in the USART_CR2 register,

•

SCEN, HDSEL and IREN bits in the USART_CR3 register.

The USART allows the user to control a bidirectional synchronous serial communications in
master mode. The CK pin is the output of the USART transmitter clock. No clock pulses are
sent to the CK pin during start bit and stop bit. Depending on the state of the LBCL bit in the
USART_CR2 register clock pulses will or will not be generated during the last valid data bit
(address mark). The CPOL bit in the USART_CR2 register allows the user to select the
clock polarity, and the CPHA bit in the USART_CR2 register allows the user to select the
phase of the external clock (see Figure 307, Figure 308 & Figure 309).
During the Idle state, preamble and send break, the external CK clock is not activated.
In synchronous mode the USART transmitter works exactly like in asynchronous mode. But
as CK is synchronized with TX (according to CPOL and CPHA), the data on TX is
synchronous.
In this mode the USART receiver works in a different manner compared to the
asynchronous mode. If RE=1, the data is sampled on CK (rising or falling edge, depending
on CPOL and CPHA), without any oversampling. A setup and a hold time must be
respected (which depends on the baud rate: 1/16 bit time).
Note:

994/1747

The CK pin works in conjunction with the TX pin. Thus, the clock is provided only if the
transmitter is enabled (TE=1) and a data is being transmitted (the data register USART_DR

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)
has been written). This means that it is not possible to receive a synchronous data without
transmitting data.
The LBCL, CPOL and CPHA bits have to be selected when both the transmitter and the
receiver are disabled (TE=RE=0) to ensure that the clock pulses function correctly. These
bits should not be changed while the transmitter or the receiver is enabled.
It is advised that TE and RE are set in the same instruction in order to minimize the setup
and the hold time of the receiver.
The USART supports master mode only: it cannot receive or send data related to an input
clock (CK is always an output).
Figure 307. USART example of synchronous transmission

5;
7;

'DWDRXW
'DWDLQ
6\QFKURQRXVGHYLFH
HJVODYH63,

86$57

&ORFN

&.

06Y9

Figure 308. USART data clock timing diagram (M=0)
,GOHRUSUHFHGLQJ
6WDUW
WUDQVPLVVLRQ

,GOHRUQH[W
6WRS WUDQVPLVVLRQ

0  GDWDELWV

&ORFN &32/ &3+$ 
&ORFN &32/ &3+$ 

&ORFN &32/ &3+$ 
&ORFN &32/ &3+$ 
'DWDRQ7;
IURPPDVWHU


6WDUW

'DWDRQ5;
IURPVODYH













/6%



06% 6WRS














06%

/6%
&DSWXUHVWUREH

/%&/ELWFRQWUROVODVWGDWDSXOVH
06Y9

RM0090 Rev 17

995/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

Figure 309. USART data clock timing diagram (M=1)
,GOHRU
SUHFHGLQJ 6WDUW
WUDQVPLVVLRQ

6WRS

0  GDWDELWV

,GOHRUQH[W
WUDQVPLVVLRQ

&ORFN &32/ 
&3+$ 
&ORFN &32/ 
&3+$ 
&ORFN &32/ 
&3+$ 
&ORFN &32/ 
&3+$ 
'DWDRQ7;
IURPPDVWHU

















6WDUW /6%

'DWDRQ5;
IURPVODYH




06%













/6%

&DSWXUH
VWUREH



6WRS


06%

/%&/ELWFRQWUROVODVWGDWDSXOVH
06Y9

Figure 310. RX data setup/hold time

&.
FDSWXUHVWUREHRQ&.ULVLQJ
HGJHLQWKLVH[DPSOH
'DWDRQ5; IURPVODYH

9DOLG'$7$ELW

W6(783

W+2/'

W6(783 W+2/'ELWWLPH
06Y9

Note:

The function of CK is different in Smartcard mode. Refer to the Smartcard mode chapter for
more details.

30.3.10

Single-wire half-duplex communication
The single-wire half-duplex mode is selected by setting the HDSEL bit in the USART_CR3
register. In this mode, the following bits must be kept cleared:
•

LINEN and CLKEN bits in the USART_CR2 register,

•

SCEN and IREN bits in the USART_CR3 register.

The USART can be configured to follow a single-wire half-duplex protocol where the TX and
RX lines are internally connected. The selection between half- and full-duplex
communication is made with a control bit ‘HALF DUPLEX SEL’ (HDSEL in USART_CR3).

996/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)
As soon as HDSEL is written to 1:
•

the TX and RX lines are internally connected

•

the RX pin is no longer used

•

the TX pin is always released when no data is transmitted. Thus, it acts as a standard
I/O in idle or in reception. It means that the I/O must be configured so that TX is
configured as floating input (or output high open-drain) when not driven by the USART.

Apart from this, the communications are similar to what is done in normal USART mode.
The conflicts on the line must be managed by the software (by the use of a centralized
arbiter, for instance). In particular, the transmission is never blocked by hardware and
continue to occur as soon as a data is written in the data register while the TE bit is set.

30.3.11

Smartcard
The Smartcard mode is selected by setting the SCEN bit in the USART_CR3 register. In
smartcard mode, the following bits must be kept cleared:
•

LINEN bit in the USART_CR2 register,

•

HDSEL and IREN bits in the USART_CR3 register.

Moreover, the CLKEN bit may be set in order to provide a clock to the smartcard.
The Smartcard interface is designed to support asynchronous protocol Smartcards as
defined in the ISO 7816-3 standard. The USART should be configured as:

Note:

•

8 bits plus parity: where M=1 and PCE=1 in the USART_CR1 register

•

1.5 stop bits when transmitting and receiving: where STOP=11 in the USART_CR2
register.

It is also possible to choose 0.5 stop bit for receiving but it is recommended to use 1.5 stop
bits for both transmitting and receiving to avoid switching between the two configurations.
Figure 311 shows examples of what can be seen on the data line with and without parity
error.
Figure 311. ISO 7816-3 asynchronous protocol
:LWKRXW3DULW\HUURU
6





*XDUGWLPH












S

6WDUWELW
:LWK3DULW\HUURU
6



*XDUGWLPH










6WDUWELW





S
/LQHSXOOHGORZE\UHFHLYHU
GXULQJVWRSLQFDVHRISDULW\HUURU
06Y9

When connected to a Smartcard, the TX output of the USART drives a bidirectional line that
is also driven by the Smartcard. The TX pin must be configured as open-drain.
Smartcard is a single wire half duplex communication protocol.
•

Transmission of data from the transmit shift register is guaranteed to be delayed by a
minimum of 1/2 baud clock. In normal operation a full transmit shift register will start
RM0090 Rev 17

997/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

shifting on the next baud clock edge. In Smartcard mode this transmission is further
delayed by a guaranteed 1/2 baud clock.

Note:

•

If a parity error is detected during reception of a frame programmed with a 0.5 or 1.5
stop bit period, the transmit line is pulled low for a baud clock period after the
completion of the receive frame. This is to indicate to the Smartcard that the data
transmitted to USART has not been correctly received. This NACK signal (pulling
transmit line low for 1 baud clock) will cause a framing error on the transmitter side
(configured with 1.5 stop bits). The application can handle re-sending of data according
to the protocol. A parity error is ‘NACK’ed by the receiver if the NACK control bit is set,
otherwise a NACK is not transmitted.

•

The assertion of the TC flag can be delayed by programming the Guard Time register.
In normal operation, TC is asserted when the transmit shift register is empty and no
further transmit requests are outstanding. In Smartcard mode an empty transmit shift
register triggers the guard time counter to count up to the programmed value in the
Guard Time register. TC is forced low during this time. When the guard time counter
reaches the programmed value TC is asserted high.

•

The de-assertion of TC flag is unaffected by Smartcard mode.

•

If a framing error is detected on the transmitter end (due to a NACK from the receiver),
the NACK will not be detected as a start bit by the receive block of the transmitter.
According to the ISO protocol, the duration of the received NACK can be 1 or 2 baud
clock periods.

•

On the receiver side, if a parity error is detected and a NACK is transmitted the receiver
will not detect the NACK as a start bit.

A break character is not significant in Smartcard mode. A 0x00 data with a framing error will
be treated as data and not as a break.
No Idle frame is transmitted when toggling the TE bit. The Idle frame (as defined for the
other configurations) is not defined by the ISO protocol.
Figure 312 details how the NACK signal is sampled by the USART. In this example the
USART is transmitting a data and is configured with 1.5 stop bits. The receiver part of the
USART is enabled in order to check the integrity of the data and the NACK signal.
Figure 312. Parity error detection using the 1.5 stop bits
%LW

3DULW\ELW

6WRSELW

ELWWLPH

ELWWLPH

6DPSOLQJDW

6DPSOLQJDW

WKWKWK

WKWKWK
ELWWLPH

6DPSOLQJDW

6DPSOLQJDW

WKWKWK

WKWKWK
06Y9

The USART can provide a clock to the smartcard through the CK output. In smartcard
mode, CK is not associated to the communication but is simply derived from the internal
peripheral input clock through a 5-bit prescaler. The division ratio is configured in the

998/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)
prescaler register USART_GTPR. CK frequency can be programmed from fCK/2 to fCK/62,
where fCK is the peripheral input clock.

30.3.12

IrDA SIR ENDEC block
The IrDA mode is selected by setting the IREN bit in the USART_CR3 register. In IrDA
mode, the following bits must be kept cleared:
•

LINEN, STOP and CLKEN bits in the USART_CR2 register,

•

SCEN and HDSEL bits in the USART_CR3 register.

The IrDA SIR physical layer specifies use of a Return to Zero, Inverted (RZI) modulation
scheme that represents logic 0 as an infrared light pulse (see Figure 313).
The SIR Transmit encoder modulates the Non Return to Zero (NRZ) transmit bit stream
output from USART. The output pulse stream is transmitted to an external output driver and
infrared LED. USART supports only bit rates up to 115.2Kbps for the SIR ENDEC. In normal
mode the transmitted pulse width is specified as 3/16 of a bit period.
The SIR receive decoder demodulates the return-to-zero bit stream from the infrared
detector and outputs the received NRZ serial bit stream to USART. The decoder input is
normally HIGH (marking state) in the Idle state. The transmit encoder output has the
opposite polarity to the decoder input. A start bit is detected when the decoder input is low.
•

IrDA is a half duplex communication protocol. If the Transmitter is busy (i.e. the USART
is sending data to the IrDA encoder), any data on the IrDA receive line will be ignored
by the IrDA decoder and if the Receiver is busy (USART is receiving decoded data
from the USART), data on the TX from the USART to IrDA will not be encoded by IrDA.
While receiving data, transmission should be avoided as the data to be transmitted
could be corrupted.

•

A ‘0 is transmitted as a high pulse and a ‘1 is transmitted as a ‘0. The width of the pulse
is specified as 3/16th of the selected bit period in normal mode (see Figure 314).

•

The SIR decoder converts the IrDA compliant receive signal into a bit stream for
USART.

•

The SIR receive logic interprets a high state as a logic one and low pulses as logic
zeros.

•

The transmit encoder output has the opposite polarity to the decoder input. The SIR
output is in low state when Idle.

•

The IrDA specification requires the acceptance of pulses greater than 1.41 us. The
acceptable pulse width is programmable. Glitch detection logic on the receiver end
filters out pulses of width less than 2 PSC periods (PSC is the prescaler value
programmed in the IrDA low-power Baud Register, USART_GTPR). Pulses of width
less than 1 PSC period are always rejected, but those of width greater than one and
less than two periods may be accepted or rejected, those greater than 2 periods will be
accepted as a pulse. The IrDA encoder/decoder doesn’t work when PSC=0.

•

The receiver can communicate with a low-power transmitter.

•

In IrDA mode, the STOP bits in the USART_CR2 register must be configured to “1 stop
bit”.

RM0090 Rev 17

999/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

IrDA low-power mode
Transmitter:
In low-power mode the pulse width is not maintained at 3/16 of the bit period. Instead, the
width of the pulse is 3 times the low-power baud rate which can be a minimum of 1.42 MHz.
Generally this value is 1.8432 MHz (1.42 MHz < PSC< 2.12 MHz). A low-power mode
programmable divisor divides the system clock to achieve this value.
Receiver:
Receiving in low-power mode is similar to receiving in normal mode. For glitch detection the
USART should discard pulses of duration shorter than 1/PSC. A valid low is accepted only if
its duration is greater than 2 periods of the IrDA low-power Baud clock (PSC value in
USART_GTPR).
Note:

A pulse of width less than two and greater than one PSC period(s) may or may not be
rejected.
The receiver set up time should be managed by software. The IrDA physical layer
specification specifies a minimum of 10 ms delay between transmission and reception (IrDA
is a half duplex protocol).
Figure 313. IrDA SIR ENDEC- block diagram

6,5(1

7;

25

86$57

5;

86$57B7;

6,5
7UDQVPLW
(QFRGHU

,U'$B287

6,5
5HFHLYH
'(FRGHU

,U'$B,1
86$57B5;

06Y9

Figure 314. IrDA data modulation (3/16) -Normal mode

7;

6WDUW
ELW


6WRS
ELW


















,U'$B287
%LWSHULRG

,U'$B,1

5;























06Y9

1000/1747

RM0090 Rev 17

RM0090

30.3.13

Universal synchronous asynchronous receiver transmitter (USART)

Continuous communication using DMA
The USART is capable of continuous communication using the DMA. The DMA requests for
Rx buffer and Tx buffer are generated independently.

Transmission using DMA
DMA mode can be enabled for transmission by setting DMAT bit in the USART_CR3
register. Data is loaded from a SRAM area configured using the DMA peripheral (refer to the
DMA specification) to the USART_DR register whenever the TXE bit is set. To map a DMA
channel for USART transmission, use the following procedure (x denotes the channel
number):
1.

Write the USART_DR register address in the DMA control register to configure it as the
destination of the transfer. The data will be moved to this address from memory after
each TXE event.

2.

Write the memory address in the DMA control register to configure it as the source of
the transfer. The data will be loaded into the USART_DR register from this memory
area after each TXE event.

3.

Configure the total number of bytes to be transferred to the DMA control register.

4.

Configure the channel priority in the DMA register

5.

Configure DMA interrupt generation after half/ full transfer as required by the
application.

6.

Clear the TC bit in the SR register by writing 0 to it.

7.

Activate the channel in the DMA register.

When the number of data transfers programmed in the DMA Controller is reached, the DMA
controller generates an interrupt on the DMA channel interrupt vector.
In transmission mode, once the DMA has written all the data to be transmitted (the TCIF flag
is set in the DMA_ISR register), the TC flag can be monitored to make sure that the USART
communication is complete. This is required to avoid corrupting the last transmission before
disabling the USART or entering the Stop mode. The software must wait until TC=1. The TC
flag remains cleared during all data transfers and it is set by hardware at the last frame’s
end of transmission.

RM0090 Rev 17

1001/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

Figure 315. Transmission using DMA
)DLE PREAMBLE

&RAME 

&RAME 

&RAME 

48 LINE
SET BY HARDWARE
CLEARED BY $-! READ

48% FLAG

SET BY HARDWARE
CLEARED BY $-! READ

IGNORED BY THE $-!
BECAUSE $-! TRANSFER IS COMPLETE

$-! REQUEST

53!24?$2

SET BY HARDWARE

&

&

&

SET
BY HARDWARE

4# FLAG
$-! WRITES
53!24?$2
FLAG $-! 4#)&
4RA NSFER COMPLETE

SET BY HARDWARE

VRIWZDUHFRQILJXUHV '0$ZULWHV) '0$ZULWHV) '0$ZULWHV)
WKH'0$WRVHQG
LQWR
LQWR
LQWR
GDWDDQGHQDEOHVWKH 86$57B'5
86$57B'5
86$57B'5
86$57

CLEAR
BY SOFTWARE

7KH'0$WUDQVIHU
LVFRPSOHWH
7&,) LQ
'0$B,65

VRIWZDUHZDLWVXQWLO7& 

AIB

Reception using DMA
DMA mode can be enabled for reception by setting the DMAR bit in USART_CR3 register.
Data is loaded from the USART_DR register to a SRAM area configured using the DMA
peripheral (refer to the DMA specification) whenever a data byte is received. To map a DMA
channel for USART reception, use the following procedure:
1.

Write the USART_DR register address in the DMA control register to configure it as the
source of the transfer. The data will be moved from this address to the memory after
each RXNE event.

2.

Write the memory address in the DMA control register to configure it as the destination
of the transfer. The data will be loaded from USART_DR to this memory area after each
RXNE event.

3.

Configure the total number of bytes to be transferred in the DMA control register.

4.

Configure the channel priority in the DMA control register

5.

Configure interrupt generation after half/ full transfer as required by the application.

6.

Activate the channel in the DMA control register.

When the number of data transfers programmed in the DMA Controller is reached, the DMA
controller generates an interrupt on the DMA channel interrupt vector. The DMAR bit should
be cleared by software in the USART_CR3 register during the interrupt subroutine.

1002/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)
Figure 316. Reception using DMA
&RAME 

&RAME 

&RAME 

48 LINE
SET BY HARDWARE
CLEARED BY $-! READ

28.% FLAG

$-! REQUEST
53!24?$2

&

&

&

$-! READS 53!24?$2

$-! 4#)& FLAG
4RA NSFER COMPLETE

VRIWZDUHFRQILJXUHVWKH
'0$WRUHFHLYHGDWD
EORFNVDQGHQDEOHV
WKH86$57

SET BY HARDWARE

'0$UHDGV)
IURP
86$57B'5

'0$UHDGV)
IURP
86$57B'5

CLEARED
BY SOFTWARE

'0$UHDGV) 7KH'0$WUDQVIHU
IURP
LVFRPSOHWH
86$57B'5
7&,) LQ
'0$B,65

AIB

Error flagging and interrupt generation in multibuffer communication
In case of multibuffer communication if any error occurs during the transaction the error flag
will be asserted after the current byte. An interrupt will be generated if the interrupt enable
flag is set. For framing error, overrun error and noise flag which are asserted with RXNE in
case of single byte reception, there will be separate error flag interrupt enable bit (EIE bit in
the USART_CR3 register), which if set will issue an interrupt after the current byte with
either of these errors.

30.3.14

Hardware flow control
It is possible to control the serial data flow between 2 devices by using the CTS input and
the RTS output. The Figure 317 shows how to connect 2 devices in this mode:
Figure 317. Hardware flow control between 2 USARTs

86$57

86$57
7;
7;FLUFXLW

&76
5;

5;
576

5;FLUFXLW

7;

5;FLUFXLW

7;FLUFXLW
576

&76

06Y9

RTS and CTS flow control can be enabled independently by writing respectively RTSE and
CTSE bits to 1 (in the USART_CR3 register).

RM0090 Rev 17

1003/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

RTS flow control
If the RTS flow control is enabled (RTSE=1), then RTS is asserted (tied low) as long as the
USART receiver is ready to receive a new data. When the receive register is full, RTS is
deasserted, indicating that the transmission is expected to stop at the end of the current
frame. Figure 318 shows an example of communication with RTS flow control enabled.
Figure 318. RTS flow control

5;

6WDUW
ELW

'DWD

6WRS
6WDUW
,GOH
ELW
ELW

'DWD

6WRS
ELW

576

5;1(

5;1(
'DWDUHDG
'DWDFDQQRZEHWUDQVPLWWHG

06Y9

CTS flow control
If the CTS flow control is enabled (CTSE=1), then the transmitter checks the CTS input
before transmitting the next frame. If CTS is asserted (tied low), then the next data is
transmitted (assuming that a data is to be transmitted, in other words, if TXE=0), else the
transmission does not occur. When CTS is deasserted during a transmission, the current
transmission is completed before the transmitter stops.
When CTSE=1, the CTSIF status bit is automatically set by hardware as soon as the CTS
input toggles. It indicates when the receiver becomes ready or not ready for communication.
An interrupt is generated if the CTSIE bit in the USART_CR3 register is set. The figure
below shows an example of communication with CTS flow control enabled.

1004/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)
Figure 319. CTS flow control
&76

&76

&76
7UDQVPLWGDWDUHJLVWHU
7'5

'DWD

7;

'DWD

HPSW\

6WRS 6WDUW
ELW ELW

HPSW\

'DWD

'DWD

:ULWLQJGDWDLQ7'5

6WRS
6WDUW
,GOH
ELW
ELW

'DWD

7UDQVPLVVLRQRI'DWDLV
GHOD\HGXQWLO&76 
06Y9

Note:

Special behavior of break frames: when the CTS flow is enabled, the transmitter does not
check the CTS input state to send a break.

RM0090 Rev 17

1005/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

30.4

RM0090

USART interrupts
Table 147. USART interrupt requests
Interrupt event

Event flag

Enable control
bit

Transmit Data Register Empty

TXE

TXEIE

CTS flag

CTS

CTSIE

Transmission Complete

TC

TCIE

Received Data Ready to be Read

RXNE

Overrun Error Detected

ORE

Idle Line Detected

IDLE

IDLEIE

Parity Error

PE

PEIE

Break Flag

LBD

LBDIE

Noise Flag, Overrun error and Framing Error in multibuffer
communication

NF or ORE or FE EIE

RXNEIE

The USART interrupt events are connected to the same interrupt vector (see Figure 320).
•

During transmission: Transmission Complete, Clear to Send or Transmit Data Register
empty interrupt.

•

While receiving: Idle Line detection, Overrun error, Receive Data register not empty,
Parity error, LIN break detection, Noise Flag (only in multi buffer communication) and
Framing Error (only in multi buffer communication).

These events generate an interrupt if the corresponding Enable Control Bit is set.
Figure 320. USART interrupt mapping diagram
7&
7&,(
7;(
7;(,(
&76,)
&76,(
,'/(
,'/(,(

86$57
LQWHUUXSW

5;1(,(
25(
5;1(,(
5;1(
3(
3(,(
/%'
/%',(
)(
1(
25(

(,(
'0$5

06Y9

1006/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)

30.5

USART mode configuration
Table 148. USART mode configuration(1)
USART
1

USART
2

USART
3

Asynchronous mode

X

X

X

X

X

X

Hardware flow control

X

X

X

NA

NA

X

Multibuffer communication (DMA)

X

X

X

X

X

X

Multiprocessor communication

X

X

X

X

X

X

Synchronous

X

X

X

NA

NA

X

Smartcard

X

X

X

NA

NA

X

Half-duplex (single-wire mode)

X

X

X

X

X

X

IrDA

X

X

X

X

X

X

LIN

X

X

X

X

X

X

USART modes

UART4

USART
6

UART5

1. X = supported; NA = not applicable.

30.6

USART registers
Refer to Section 1.1: List of abbreviations for registers for registers for a list of abbreviations
used in register descriptions.
The peripheral registers have to be accessed by half-words (16 bits) or words (32 bits).

30.6.1

Status register (USART_SR)
Address offset: 0x00
Reset value: 0x0000 00C0

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

Reserved
15

14

13

12

Reserved

11

10

9

8

7

6

5

4

3

2

1

0

CTS

LBD

TXE

TC

RXNE

IDLE

ORE

NF

FE

PE

rc_w0

rc_w0

r

rc_w0

rc_w0

r

r

r

r

r

RM0090 Rev 17

1007/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

Bits 31:10 Reserved, must be kept at reset value
Bit 9 CTS: CTS flag
This bit is set by hardware when the CTS input toggles, if the CTSE bit is set. It is cleared by
software (by writing it to 0). An interrupt is generated if CTSIE=1 in the USART_CR3
register.
0: No change occurred on the CTS status line
1: A change occurred on the CTS status line
Note: This bit is not available for UART4 & UART5.
Bit 8 LBD: LIN break detection flag
This bit is set by hardware when the LIN break is detected. It is cleared by software (by
writing it to 0). An interrupt is generated if LBDIE = 1 in the USART_CR2 register.
0: LIN Break not detected
1: LIN break detected
Note: An interrupt is generated when LBD=1 if LBDIE=1
Bit 7 TXE: Transmit data register empty
This bit is set by hardware when the content of the TDR register has been transferred into
the shift register. An interrupt is generated if the TXEIE bit =1 in the USART_CR1 register. It
is cleared by a write to the USART_DR register.
0: Data is not transferred to the shift register
1: Data is transferred to the shift register)
Note: This bit is used during single buffer transmission.
Bit 6 TC: Transmission complete
This bit is set by hardware if the transmission of a frame containing data is complete and if
TXE is set. An interrupt is generated if TCIE=1 in the USART_CR1 register. It is cleared by
a software sequence (a read from the USART_SR register followed by a write to the
USART_DR register). The TC bit can also be cleared by writing a '0' to it. This clearing
sequence is recommended only for multibuffer communication.
0: Transmission is not complete
1: Transmission is complete
Bit 5 RXNE: Read data register not empty
This bit is set by hardware when the content of the RDR shift register has been transferred
to the USART_DR register. An interrupt is generated if RXNEIE=1 in the USART_CR1
register. It is cleared by a read to the USART_DR register. The RXNE flag can also be
cleared by writing a zero to it. This clearing sequence is recommended only for multibuffer
communication.
0: Data is not received
1: Received data is ready to be read.
Bit 4 IDLE: IDLE line detected
This bit is set by hardware when an Idle Line is detected. An interrupt is generated if the
IDLEIE=1 in the USART_CR1 register. It is cleared by a software sequence (an read to the
USART_SR register followed by a read to the USART_DR register).
0: No Idle Line is detected
1: Idle Line is detected
Note: The IDLE bit will not be set again until the RXNE bit has been set itself (i.e. a new idle
line occurs).

1008/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)

Bit 3 ORE: Overrun error
This bit is set by hardware when the word currently being received in the shift register is
ready to be transferred into the RDR register while RXNE=1. An interrupt is generated if
RXNEIE=1 in the USART_CR1 register. It is cleared by a software sequence (an read to the
USART_SR register followed by a read to the USART_DR register).
0: No Overrun error
1: Overrun error is detected
Note: When this bit is set, the RDR register content will not be lost but the shift register will be
overwritten. An interrupt is generated on ORE flag in case of Multi Buffer
communication if the EIE bit is set.
Bit 2 NF: Noise detected flag
This bit is set by hardware when noise is detected on a received frame. It is cleared by a
software sequence (an read to the USART_SR register followed by a read to the
USART_DR register).
0: No noise is detected
1: Noise is detected
Note: This bit does not generate interrupt as it appears at the same time as the RXNE bit
which itself generates an interrupting interrupt is generated on NF flag in case of Multi
Buffer communication if the EIE bit is set.
Note: When the line is noise-free, the NF flag can be disabled by programming the ONEBIT
bit to 1 to increase the USART tolerance to deviations (Refer to Section 30.3.5: USART
receiver tolerance to clock deviation on page 988).
Bit 1 FE: Framing error
This bit is set by hardware when a de-synchronization, excessive noise or a break character
is detected. It is cleared by a software sequence (an read to the USART_SR register
followed by a read to the USART_DR register).
0: No Framing error is detected
1: Framing error or break character is detected
Note: This bit does not generate interrupt as it appears at the same time as the RXNE bit
which itself generates an interrupt. If the word currently being transferred causes both
frame error and overrun error, it will be transferred and only the ORE bit will be set.
An interrupt is generated on FE flag in case of Multi Buffer communication if the EIE bit
is set.
Bit 0 PE: Parity error
This bit is set by hardware when a parity error occurs in receiver mode. It is cleared by a
software sequence (a read from the status register followed by a read or write access to the
USART_DR data register). The software must wait for the RXNE flag to be set before
clearing the PE bit.
An interrupt is generated if PEIE = 1 in the USART_CR1 register.
0: No parity error
1: Parity error

RM0090 Rev 17

1009/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

30.6.2

RM0090

Data register (USART_DR)
Address offset: 0x04
Reset value: 0xXXXX XXXX
Bits 31:9 Reserved, must be kept at reset value
Bits 8:0 DR[8:0]: Data value
Contains the Received or Transmitted data character, depending on whether it is read from
or written to.
The Data register performs a double function (read and write) since it is composed of two
registers, one for transmission (TDR) and one for reception (RDR)
The TDR register provides the parallel interface between the internal bus and the output shift
register (see Figure 1).
The RDR register provides the parallel interface between the input shift register and the
internal bus.
When transmitting with the parity enabled (PCE bit set to 1 in the USART_CR1 register), the
value written in the MSB (bit 7 or bit 8 depending on the data length) has no effect because
it is replaced by the parity.
When receiving with the parity enabled, the value read in the MSB bit is the received parity
bit.

30.6.3

Baud rate register (USART_BRR)

Note:

The baud counters stop counting if the TE or RE bits are disabled respectively.
Address offset: 0x08
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

2

1

0

Reserved
15

14

13

12

11

rw

rw

rw

rw

rw

10

9

8

7

6

5

4

3

rw

rw

rw

rw

rw

rw

DIV_Mantissa[11:0]
rw

rw

DIV_Fraction[3:0]
rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value
Bits 15:4 DIV_Mantissa[11:0]: mantissa of USARTDIV
These 12 bits define the mantissa of the USART Divider (USARTDIV)
Bits 3:0 DIV_Fraction[3:0]: fraction of USARTDIV
These 4 bits define the fraction of the USART Divider (USARTDIV). When OVER8=1, the
DIV_Fraction3 bit is not considered and must be kept cleared.

30.6.4

Control register 1 (USART_CR1)
Address offset: 0x0C
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

5

4

3

2

1

0

TE

RE

RWU

SBK

rw

rw

rw

rw

Reserved
15

14

13

OVER8

Reserved

rw

Res.

1010/1747

12

11

10

9

8

7

6

UE

M

WAKE

PCE

PS

PEIE

TXEIE

TCIE

rw

rw

rw

rw

rw

rw

rw

rw

RM0090 Rev 17

RXNEIE IDLEIE
rw

rw

RM0090

Universal synchronous asynchronous receiver transmitter (USART)

Bits 31:16 Reserved, must be kept at reset value
Bit 15 OVER8: Oversampling mode
0: oversampling by 16
1: oversampling by 8
Note: Oversampling by 8 is not available in the Smartcard, IrDA and LIN modes: when
SCEN=1,IREN=1 or LINEN=1 then OVER8 is forced to ‘0 by hardware.
Bit 14 Reserved, must be kept at reset value
Bit 13 UE: USART enable
When this bit is cleared, the USART prescalers and outputs are stopped and the end of the
current byte transfer in order to reduce power consumption. This bit is set and cleared by
software.
0: USART prescaler and outputs disabled
1: USART enabled
Bit 12 M: Word length
This bit determines the word length. It is set or cleared by software.
0: 1 Start bit, 8 Data bits, n Stop bit
1: 1 Start bit, 9 Data bits, n Stop bit
Note: The M bit must not be modified during a data transfer (both transmission and reception)
Bit 11 WAKE: Wakeup method
This bit determines the USART wakeup method, it is set or cleared by software.
0: Idle Line
1: Address Mark
Bit 10 PCE: Parity control enable
This bit selects the hardware parity control (generation and detection). When the parity
control is enabled, the computed parity is inserted at the MSB position (9th bit if M=1; 8th bit
if M=0) and parity is checked on the received data. This bit is set and cleared by software.
Once it is set, PCE is active after the current byte (in reception and in transmission).
0: Parity control disabled
1: Parity control enabled
Bit 9 PS: Parity selection
This bit selects the odd or even parity when the parity generation/detection is enabled (PCE
bit set). It is set and cleared by software. The parity will be selected after the current byte.
0: Even parity
1: Odd parity
Bit 8 PEIE: PE interrupt enable
This bit is set and cleared by software.
0: Interrupt is inhibited
1: An USART interrupt is generated whenever PE=1 in the USART_SR register
Bit 7 TXEIE: TXE interrupt enable
This bit is set and cleared by software.
0: Interrupt is inhibited
1: An USART interrupt is generated whenever TXE=1 in the USART_SR register
Bit 6 TCIE: Transmission complete interrupt enable
This bit is set and cleared by software.
0: Interrupt is inhibited
1: An USART interrupt is generated whenever TC=1 in the USART_SR register

RM0090 Rev 17

1011/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

Bit 5 RXNEIE: RXNE interrupt enable
This bit is set and cleared by software.
0: Interrupt is inhibited
1: An USART interrupt is generated whenever ORE=1 or RXNE=1 in the USART_SR
register
Bit 4 IDLEIE: IDLE interrupt enable
This bit is set and cleared by software.
0: Interrupt is inhibited
1: An USART interrupt is generated whenever IDLE=1 in the USART_SR register
Bit 3 TE: Transmitter enable
This bit enables the transmitter. It is set and cleared by software.
0: Transmitter is disabled
1: Transmitter is enabled
Note: During transmission, a “0” pulse on the TE bit (“0” followed by “1”) sends a preamble
(idle line) after the current word, except in smartcard mode.
When TE is set, there is a 1 bit-time delay before the transmission starts.
Bit 2 RE: Receiver enable
This bit enables the receiver. It is set and cleared by software.
0: Receiver is disabled
1: Receiver is enabled and begins searching for a start bit
Bit 1 RWU: Receiver wakeup
This bit determines if the USART is in mute mode or not. It is set and cleared by software
and can be cleared by hardware when a wakeup sequence is recognized.
0: Receiver in active mode
1: Receiver in mute mode
Note: Before selecting Mute mode (by setting the RWU bit) the USART must first receive a
data byte, otherwise it cannot function in Mute mode with wakeup by Idle line detection.
In Address Mark Detection wakeup configuration (WAKE bit=1) the RWU bit cannot be
modified by software while the RXNE bit is set.
Bit 0 SBK: Send break
This bit set is used to send break characters. It can be set and cleared by software. It should
be set by software, and will be reset by hardware during the stop bit of break.
0: No break character is transmitted
1: Break character will be transmitted

1012/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)

30.6.5

Control register 2 (USART_CR2)
Address offset: 0x10
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

3

2

1

0

Reserved
15
Res.

14
LINEN
rw

13

12

STOP[1:0]
rw

rw

11

10

9

8

7

6

5

4

CLKEN

CPOL

CPHA

LBCL

Res.

LBDIE

LBDL

Res.

rw

rw

rw

rw

rw

rw

rw

ADD[3:0]
rw

rw

rw

rw

Bits 31:15 Reserved, must be kept at reset value
Bit 14 LINEN: LIN mode enable
This bit is set and cleared by software.
0: LIN mode disabled
1: LIN mode enabled
The LIN mode enables the capability to send LIN Synch Breaks (13 low bits) using the SBK bit in
the USART_CR1 register, and to detect LIN Sync breaks.
Bits 13:12 STOP: STOP bits
These bits are used for programming the stop bits.
00: 1 Stop bit
01: 0.5 Stop bit
10: 2 Stop bits
11: 1.5 Stop bit
Note: The 0.5 Stop bit and 1.5 Stop bit are not available for UART4 & UART5.
Bit 11 CLKEN: Clock enable
This bit allows the user to enable the CK pin.
0: CK pin disabled
1: CK pin enabled
This bit is not available for UART4 & UART5.
Bit 10 CPOL: Clock polarity
This bit allows the user to select the polarity of the clock output on the CK pin in synchronous mode.
It works in conjunction with the CPHA bit to produce the desired clock/data relationship
0: Steady low value on CK pin outside transmission window.
1: Steady high value on CK pin outside transmission window.
This bit is not available for UART4 & UART5.
Bit 9 CPHA: Clock phase
This bit allows the user to select the phase of the clock output on the CK pin in synchronous mode.
It works in conjunction with the CPOL bit to produce the desired clock/data relationship (see figures
308 to 309)
0: The first clock transition is the first data capture edge
1: The second clock transition is the first data capture edge
Note: This bit is not available for UART4 & UART5.

RM0090 Rev 17

1013/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

Bit 8 LBCL: Last bit clock pulse
This bit allows the user to select whether the clock pulse associated with the last data bit
transmitted (MSB) has to be output on the CK pin in synchronous mode.
0: The clock pulse of the last data bit is not output to the CK pin
1: The clock pulse of the last data bit is output to the CK pin
Note: 1: The last bit is the 8th or 9th data bit transmitted depending on the 8 or 9 bit format selected
by the M bit in the USART_CR1 register.
2: This bit is not available for UART4 & UART5.
Bit 7 Reserved, must be kept at reset value
Bit 6 LBDIE: LIN break detection interrupt enable
Break interrupt mask (break detection using break delimiter).
0: Interrupt is inhibited
1: An interrupt is generated whenever LBD=1 in the USART_SR register
Bit 5 LBDL: lin break detection length
This bit is for selection between 11 bit or 10 bit break detection.
0: 10-bit break detection
1: 11-bit break detection
Bit 4 Reserved, must be kept at reset value
Bits 3:0 ADD[3:0]: Address of the USART node
This bit-field gives the address of the USART node.
This is used in multiprocessor communication during mute mode, for wake up with address mark
detection.

Note:

These 3 bits (CPOL, CPHA, LBCL) should not be written while the transmitter is enabled.

30.6.6

Control register 3 (USART_CR3)
Address offset: 0x14
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

Reserved
15

14

13

Reserved

12

11

10

9

8

7

6

5

4

3

2

1

0

ONEBIT

CTSIE

CTSE

RTSE

DMAT

DMAR

SCEN

NACK

HDSEL

IRLP

IREN

EIE

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:12 Reserved, must be kept at reset value
Bit 11 ONEBIT: One sample bit method enable
This bit allows the user to select the sample method. When the one sample bit method is
selected the noise detection flag (NF) is disabled.
0: Three sample bit method
1: One sample bit method
Note: The ONEBIT feature applies only to data bits. It does not apply to START bit.
Bit 10 CTSIE: CTS interrupt enable
0: Interrupt is inhibited
1: An interrupt is generated whenever CTS=1 in the USART_SR register
Note: This bit is not available for UART4 & UART5.

1014/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)

Bit 9 CTSE: CTS enable
0: CTS hardware flow control disabled
1: CTS mode enabled, data is only transmitted when the CTS input is asserted (tied to
0). If the CTS input is deasserted while a data is being transmitted, then the
transmission is completed before stopping. If a data is written into the data register
while CTS is deasserted, the transmission is postponed until CTS is asserted.
Note: This bit is not available for UART4 & UART5.
Bit 8 RTSE: RTS enable
0: RTS hardware flow control disabled
1: RTS interrupt enabled, data is only requested when there is space in the receive
buffer. The transmission of data is expected to cease after the current character has
been transmitted. The RTS output is asserted (tied to 0) when a data can be received.
Note: This bit is not available for UART4 & UART5.
Bit 7 DMAT: DMA enable transmitter
This bit is set/reset by software
1: DMA mode is enabled for transmission.
0: DMA mode is disabled for transmission.
Bit 6 DMAR: DMA enable receiver
This bit is set/reset by software
1: DMA mode is enabled for reception
0: DMA mode is disabled for reception
Bit 5 SCEN: Smartcard mode enable
This bit is used for enabling Smartcard mode.
0: Smartcard Mode disabled
1: Smartcard Mode enabled
Note: This bit is not available for UART4 & UART5.
Bit 4 NACK: Smartcard NACK enable
0: NACK transmission in case of parity error is disabled
1: NACK transmission during parity error is enabled
Note: This bit is not available for UART4 & UART5.
Bit 3 HDSEL: Half-duplex selection
Selection of Single-wire Half-duplex mode
0: Half duplex mode is not selected
1: Half duplex mode is selected

RM0090 Rev 17

1015/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

RM0090

Bit 2 IRLP: IrDA low-power
This bit is used for selecting between normal and low-power IrDA modes
0: Normal mode
1: Low-power mode
Bit 1 IREN: IrDA mode enable
This bit is set and cleared by software.
0: IrDA disabled
1: IrDA enabled
Bit 0 EIE: Error interrupt enable
Error Interrupt Enable Bit is required to enable interrupt generation in case of a framing
error, overrun error or noise flag (FE=1 or ORE=1 or NF=1 in the USART_SR register) in
case of Multi Buffer Communication (DMAR=1 in the USART_CR3 register).
0: Interrupt is inhibited
1: An interrupt is generated whenever DMAR=1 in the USART_CR3 register and FE=1 or
ORE=1 or NF=1 in the USART_SR register.

1016/1747

RM0090 Rev 17

RM0090

Universal synchronous asynchronous receiver transmitter (USART)

30.6.7

Guard time and prescaler register (USART_GTPR)
Address offset: 0x18
Reset value: 0x0000 0000

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

rw

rw

rw

Reserved
15

14

13

12

11

10

9

8

7

GT[7:0]
rw

rw

rw

rw

rw

PSC[7:0]
rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value
Bits 15:8 GT[7:0]: Guard time value
This bit-field gives the Guard time value in terms of number of baud clocks.
This is used in Smartcard mode. The Transmission Complete flag is set after this guard
time value.
Note: This bit is not available for UART4 & UART5.
Bits 7:0 PSC[7:0]: Prescaler value
– In IrDA Low-power mode:
PSC[7:0] = IrDA Low-Power Baud Rate
Used for programming the prescaler for dividing the system clock to achieve the low-power
frequency:
The source clock is divided by the value given in the register (8 significant bits):
00000000: Reserved - do not program this value
00000001: divides the source clock by 1
00000010: divides the source clock by 2
...
– In normal IrDA mode: PSC must be set to 00000001.
– In smartcard mode:
PSC[4:0]: Prescaler value
Used for programming the prescaler for dividing the system clock to provide the smartcard
clock.
The value given in the register (5 significant bits) is multiplied by 2 to give the division factor
of the source clock frequency:
00000: Reserved - do not program this value
00001: divides the source clock by 2
00010: divides the source clock by 4
00011: divides the source clock by 6
...
Note: 1: Bits [7:5] have no effect if Smartcard mode is used.
2: This bit is not available for UART4 & UART5.

RM0090 Rev 17

1017/1747
1018

Universal synchronous asynchronous receiver transmitter (USART)

30.6.8

RM0090

USART register map
The table below gives the USART register map and reset values.

USART_DR

0

ORE

NF

FE

PE

0

DIV_Fraction
[3:0]

RWU

SBK

0

0

STOP
[1:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

EIE

RE
0

IREN

TE
0

IRLP

IDLEIE
0
Reserved

0

NACK

RXNEIE

0

LBDL

0

SCEN

TCIE

0

LBDIE

0

DMAR

0

LBCL

0

RTSE

0

0

CPOL

TXEIE

0

PS

0

PEIE

0

PCE

0

M

0

GT[7:0]

0

0

0

0

0

0

0

0

ADD[3:0]
0

0

0

PSC[7:0]
0

0

0

0

Refer to Section 2.3: Memory map for the register boundary addresses.

RM0090 Rev 17

0

0

Reserved
0

0

0

Reserved

0

0

0

CPHA

0

0

0

CTSE

Reserved

Reset value

1018/1747

0

HDSEL

USART_GTPR

0

0

Reset value

0x18

0

0

CTSIE

USART_CR3

0

WAKE

0

Reset value

0x14

0

0

CLKEN

USART_CR2

0

0

ONEBIT

0x10

0

0
UE

Reset value

0

OVER8

Reserved

0

Reserved

USART_CR1

0

DIV_Mantissa[15:4]

Reserved

Reset value

0x0C

0

DR[8:0]

LINEN

USART_BRR

1

Reserved

Reset value

0x08

IDLE

1

TC

0

RXNE

0

Reserved

0x04

TXE

Reserved

Reset value

LBD

USART_SR

DMAT

0x00

Register

CTS

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 149. USART register map and reset values

0

0

0

0

RM0090

31

Secure digital input/output interface (SDIO)

Secure digital input/output interface (SDIO)
This section applies to the whole STM32F4xx family, unless otherwise specified.

31.1

SDIO main features
The SD/SDIO MMC card host interface (SDIO) provides an interface between the APB2
peripheral bus and MultiMediaCards (MMCs), SD memory cards, SDIO cards and CE-ATA
devices.
The MultiMediaCard system specifications are available through the MultiMediaCard
Association website at http://www.jedec.org/, published by the MMCA technical committee.
SD memory card and SD I/O card system specifications are available through the SD card
Association website at http://www.sdcard.org.
CE-ATA system specifications are available through the CE-ATA workgroup website.
The SDIO features include the following:

Note:

•

Full compliance with MultiMediaCard System Specification Version 4.2. Card support
for three different databus modes: 1-bit (default), 4-bit and 8-bit

•

Full compatibility with previous versions of MultiMediaCards (forward compatibility)

•

Full compliance with SD Memory Card Specifications Version 2.0

•

Full compliance with SD I/O Card Specification Version 2.0: card support for two
different databus modes: 1-bit (default) and 4-bit

•

Full support of the CE-ATA features (full compliance with CE-ATA digital protocol
Rev1.1)

•

Data transfer up to 50 MHz for the 8 bit mode

•

Data and command output enable signals to control external bidirectional drivers.

The SDIO does not have an SPI-compatible communication mode.
The SD memory card protocol is a superset of the MultiMediaCard protocol as defined in the
MultiMediaCard system specification V2.11. Several commands required for SD memory
devices are not supported by either SD I/O-only cards or the I/O portion of combo cards.
Some of these commands have no use in SD I/O devices, such as erase commands, and
thus are not supported in the SDIO. In addition, several commands are different between
SD memory cards and SD I/O cards and thus are not supported in the SDIO. For details
refer to SD I/O card Specification Version 1.0. CE-ATA is supported over the MMC electrical
interface using a protocol that utilizes the existing MMC access primitives. The interface
electrical and signaling definition is as defined in the MMC reference.
The MultiMediaCard/SD bus connects cards to the controller.
The current version of the SDIO supports only one SD/SDIO/MMC4.2 card at any one time
and a stack of MMC4.1 or previous.

RM0090 Rev 17

1019/1747
1075

Secure digital input/output interface (SDIO)

31.2

RM0090

SDIO bus topology
Communication over the bus is based on command and data transfers.
The basic transaction on the MultiMediaCard/SD/SD I/O bus is the command/response
transaction. These types of bus transaction transfer their information directly within the
command or response structure. In addition, some operations have a data token.
Data transfers to/from SD/SDIO memory cards are done in data blocks. Data transfers
to/from MMC are done data blocks or streams. Data transfers to/from the CE-ATA Devices
are done in data blocks.
Figure 321. SDIO “no response” and “no data” operations
)URPKRVWWRFDUG V

6'00&B&0'

&RPPDQG

)URPKRVWWRFDUG

&RPPDQG

)URPFDUGWRKRVW

5HVSRQVH

6'00&B'

2SHUDWLRQ QRUHVSRQVH

2SHUDWLRQ QRGDWD
DLE

Figure 322. SDIO (multiple) block read operation
)URPKRVWWRFDUG )URPFDUGWRKRVW
GDWDIURPFDUGWRKRVW
6'00&B&0'

6'00&B'

&RPPDQG

5HVSRQVH

'DWDEORFN FUF

6WRSFRPPDQG
VWRSVGDWDWUDQVIHU
&RPPDQG

'DWDEORFN FUF

5HVSRQVH

'DWDEORFN FUF

%ORFNUHDGRSHUDWLRQ
0XOWLSOHEORFNUHDGRSHUDWLRQ

1020/1747

RM0090 Rev 17

'DWDVWRSRSHUDWLRQ
DLE

RM0090

Secure digital input/output interface (SDIO)
Figure 323. SDIO (multiple) block write operation
)URPKRVWWRFDUG )URPFDUGWRKRVW
GDWDIURPKRVWWRFDUG
6'00&B&0'

&RPPDQG

6WRSFRPPDQG
VWRSVGDWDWUDQVIHU

5HVSRQVH

6'00&B'

&RPPDQG

'DWDEORFN FUF

%XV\

5HVSRQVH

'DWDEORFN FUF

%XV\

'DWDVWRSRSHUDWLRQ

%ORFNZULWHRSHUDWLRQ
0XOWLSOHEORFNZULWHRSHUDWLRQ

DLE

Note:

The SDIO will not send any data as long as the Busy signal is asserted (SDIO_D0 pulled
low).
Figure 324. SDIO sequential read operation
)URPKRVWWR
FDUG V
)URPFDUGWRKRVW
'DWDIURPFDUGWRKRVW
6'00&B&0'

&RPPDQG

5HVSRQVH

6WRSFRPPDQG
VWRSVGDWDWUDQVIHU
&RPPDQG

5HVSRQVH

'DWDVWUHDP

6'00&B'

'DWDWUDQVIHURSHUDWLRQ

'DWDVWRSRSHUDWLRQ
DLE

Figure 325. SDIO sequential write operation
)URPKRVWWR
FDUG V
)URPFDUGWRKRVW
'DWDIURPKRVWWRFDUG
6'00&B&0'

&RPPDQG

5HVSRQVH

6WRSFRPPDQG
VWRSVGDWDWUDQVIHU
&RPPDQG

5HVSRQVH

'DWDVWUHDP

6'00&B'

'DWDWUDQVIHURSHUDWLRQ

'DWDVWRSRSHUDWLRQ
DLE

RM0090 Rev 17

1021/1747
1075

Secure digital input/output interface (SDIO)

31.3

RM0090

SDIO functional description
The SDIO consists of two parts:
•

The SDIO adapter block provides all functions specific to the MMC/SD/SD I/O card
such as the clock generation unit, command and data transfer.

•

The APB2 interface accesses the SDIO adapter registers, and generates interrupt and
DMA request signals.
Figure 326. SDIO block diagram
6'00&
6'00&B&.
,QWHUUXSWVDQG
'0$UHTXHVW

6'00&B&0'
$3%
LQWHUIDFH

6'00&
DGDSWHU

3&/.

6'00&&/.

6'00&B'>@

$3%EXV

DLE

By default SDIO_D0 is used for data transfer. After initialization, the host can change the
databus width.
If a MultiMediaCard is connected to the bus, SDIO_D0, SDIO_D[3:0] or SDIO_D[7:0] can be
used for data transfer. MMC V3.31 or previous, supports only 1 bit of data so only SDIO_D0
can be used.
If an SD or SD I/O card is connected to the bus, data transfer can be configured by the host
to use SDIO_D0 or SDIO_D[3:0]. All data lines are operating in push-pull mode.
SDIO_CMD has two operational modes:
•

Open-drain for initialization (only for MMCV3.31 or previous)

•

Push-pull for command transfer (SD/SD I/O card MMC4.2 use push-pull drivers also for
initialization)

SDIO_CK is the clock to the card: one bit is transferred on both command and data lines
with each clock cycle.
The SDIO uses two clock signals:
•

SDIO adapter clock SDIOCLK up to 50 MHz (48 MHz when in use with USB)

•

APB2 bus clock (PCLK2)

PCLK2 and SDIO_CK clock frequencies must respect the following condition:
Frequenc ( PCLK2 ) ≥ 3 ⁄ 8 × Frequency ( SDIO_CK )

The signals shown in Table 150 are used on the MultiMediaCard/SD/SD I/O card bus.

1022/1747

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)
Table 150. SDIO I/O definitions
Pin

31.3.1

Direction

Description

SDIO_CK

Output

MultiMediaCard/SD/SDIO card clock. This pin is the clock from
host to card.

SDIO_CMD

Bidirectional

MultiMediaCard/SD/SDIO card command. This pin is the
bidirectional command/response signal.

SDIO_D[7:0]

Bidirectional

MultiMediaCard/SD/SDIO card data. These pins are the
bidirectional databus.

SDIO adapter
Figure 327 shows a simplified block diagram of an SDIO adapter.
Figure 327. SDIO adapter
6'00&DGDSWHU

$GDSWHU
UHJLVWHUV
7R$3%
LQWHUIDFH

),)2

3&/.

6'00&B&.

&RPPDQG
SDWK

6'00&B&0'

'DWDSDWK

6'00&B'>@

&DUGEXV

&RQWUROXQLW

6'00&&/.
DLE

The SDIO adapter is a multimedia/secure digital memory card bus master that provides an
interface to a multimedia card stack or to a secure digital memory card. It consists of five
subunits:

Note:

•

Adapter register block

•

Control unit

•

Command path

•

Data path

•

Data FIFO

The adapter registers and FIFO use the APB2 bus clock domain (PCLK2). The control unit,
command path and data path use the SDIO adapter clock domain (SDIOCLK).

Adapter register block
The adapter register block contains all system registers. This block also generates the
signals that clear the static flags in the multimedia card. The clear signals are generated
when 1 is written into the corresponding bit location in the SDIO Clear register.

RM0090 Rev 17

1023/1747
1075

Secure digital input/output interface (SDIO)

RM0090

Control unit
The control unit contains the power management functions and the clock divider for the
memory card clock.
There are three power phases:
•

power-off

•

power-up

•

power-on
Figure 328. Control unit
&RQWUROXQLW

3RZHUPDQDJHPHQW

$GDSWHU
UHJLVWHUV

&ORFNPDQDJHPHQW

6'00&B&.

7RFRPPDQGDQGGDWDSDWK
DLE

The control unit is illustrated in Figure 328. It consists of a power management subunit and
a clock management subunit.
The power management subunit disables the card bus output signals during the power-off
and power-up phases.
The clock management subunit generates and controls the SDIO_CK signal. The SDIO_CK
output can use either the clock divide or the clock bypass mode. The clock output is
inactive:

1024/1747

•

after reset

•

during the power-off or power-up phases

•

if the power saving mode is enabled and the card bus is in the Idle state (eight clock
periods after both the command and data path subunits enter the Idle phase)

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)

Command path
The command path unit sends commands to and receives responses from the cards.
Figure 329. SDIO adapter command path

7RFRQWUROXQLW

6WDWXV
IODJ

&RQWURO
ORJLF

&RPPDQG
WLPHU

$GDSWHUUHJLVWHUV
6'00&B&0'LQ
&0'
$UJXPHQW
&5&
6KLIW
UHJLVWHU

&0'

7R$3%LQWHUIDFH

6'00&B&0'RXW

5HVSRQVH
UHJLVWHUV

DLE

•

Command path state machine (CPSM)
–

When the command register is written to and the enable bit is set, command
transfer starts. When the command has been sent, the command path state
machine (CPSM) sets the status flags and enters the Idle state if a response is not
required. If a response is required, it waits for the response (see Figure 330 on
page 1026). When the response is received, the received CRC code and the
internally generated code are compared, and the appropriate status flags are set.

RM0090 Rev 17

1025/1747
1075

Secure digital input/output interface (SDIO)

RM0090

Figure 330. Command path state machine (CPSM)

("#$

 ( (
!( 

 
( 
 !$ (#
"&( "
 (#( "
 ((



$

#! #("&( "
#( "( (
(


#

#! #(&(( 

 (( ($""%!$(
'$( "( 
( 
 !$ (#(


(
 (#$"$
#$( $

 (#( "(
 ("#! #

 ( #( "
 ($ %$

&

#! #
#$"$



$( "("#! #

$

#! #(&(( 
( (
 ($""%!$(('$( "( 

 ( !$ (#(#



When the Wait state is entered, the command timer starts running. If the timeout is reached
before the CPSM moves to the Receive state, the timeout flag is set and the Idle state is
entered.
Note:

The command timeout has a fixed value of 64 SDIO_CK clock periods.
If the interrupt bit is set in the command register, the timer is disabled and the CPSM waits
for an interrupt request from one of the cards. If a pending bit is set in the command register,
the CPSM enters the Pend state, and waits for a CmdPend signal from the data path
subunit. When CmdPend is detected, the CPSM moves to the Send state. This enables the
data counter to trigger the stop command transmission.

Note:

1026/1747

The CPSM remains in the Idle state for at least eight SDIO_CK periods to meet the NCC and
NRC timing constraints. NCC is the minimum delay between two host commands, and NRC is
the minimum delay between the host command and the card response.

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)
Figure 331. SDIO command transfer
DWOHDVW6'00&B&.
F\FOHV
6'00&B&.

&RPPDQG

&RPPDQG

5HVSRQVH

6WDWH

,GOH

6HQG

:DLW

5HFHLYH

,GOH

6HQG

6'00&B&0'

+L=

&RQWUROOHUGULYHV

+L=

&DUGGULYHV

+L=

&RQWUROOHUGULYHV
DLE

•

Command format
–

Command: a command is a token that starts an operation. Command are sent
from the host either to a single card (addressed command) or to all connected
cards (broadcast command are available for MMC V3.31 or previous). Commands
are transferred serially on the CMD line. All commands have a fixed length of 48
bits. The general format for a command token for MultiMediaCards, SD-Memory
cards and SDIO-Cards is shown in Table 151. CE-ATA commands are an
extension of MMC commands V4.2, and so have the same format.
The command path operates in a half-duplex mode, so that commands and
responses can either be sent or received. If the CPSM is not in the Send state, the
SDIO_CMD output is in the Hi-Z state, as shown in Figure 331 on page 1027.
Data on SDIO_CMD are synchronous with the rising edge of SDIO_CK. Table
shows the command format.
Table 151. Command format

Bit position

Width

Value

Description

47

1

0

Start bit

46

1

1

Transmission bit

[45:40]

6

-

Command index

[39:8]

32

-

Argument

[7:1]

7

-

CRC7

0

1

1

End bit

–

Response: a response is a token that is sent from an addressed card (or
synchronously from all connected cards for MMC V3.31 or previous), to the host
as an answer to a previously received command. Responses are transferred
serially on the CMD line.

The SDIO supports two response types. Both use CRC error checking:

Note:

•

48 bit short response

•

136 bit long response

If the response does not contain a CRC (CMD1 response), the device driver must ignore the
CRC failed status.

RM0090 Rev 17

1027/1747
1075

Secure digital input/output interface (SDIO)

RM0090

Table 152. Short response format
Bit position

Width

Value

Description

47

1

0

Start bit

46

1

0

Transmission bit

[45:40]

6

-

Command index

[39:8]

32

-

Argument

[7:1]

7

-

CRC7(or 1111111)

0

1

1

End bit

Table 153. Long response format
Bit position

Width

Value

Description

135

1

0

Start bit

134

1

0

Transmission bit

[133:128]

6

111111

Reserved

[127:1]

127

-

CID or CSD (including internal CRC7)

0

1

1

End bit

The command register contains the command index (six bits sent to a card) and the
command type. These determine whether the command requires a response, and whether
the response is 48 or 136 bits long (see Section 31.9.4 on page 1062). The command path
implements the status flags shown in Table 154:
Table 154. Command path status flags
Flag

Description

CMDREND

Set if response CRC is OK.

CCRCFAIL

Set if response CRC fails.

CMDSENT

Set when command (that does not require response) is sent

CTIMEOUT

Response timeout.

CMDACT

Command transfer in progress.

The CRC generator calculates the CRC checksum for all bits before the CRC code. This
includes the start bit, transmitter bit, command index, and command argument (or card
status). The CRC checksum is calculated for the first 120 bits of CID or CSD for the long
response format. Note that the start bit, transmitter bit and the six reserved bits are not used
in the CRC calculation.
The CRC checksum is a 7-bit value:
CRC[6:0] = Remainder [(M(x) * x7) / G(x)]
G(x) = x7 + x3 + 1
M(x) = (start bit) * x39 + ... + (last bit before CRC) * x0, or
M(x) = (start bit) * x119 + ... + (last bit before CRC) * x0

1028/1747

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)

Data path
The data path subunit transfers data to and from cards. Figure 332 shows a block diagram
of the data path.
Figure 332. Data path
'DWDSDWK

7RFRQWUROXQLW

6WDWXV
IODJ

&RQWURO
ORJLF

'DWD
WLPHU

'DWD),)2

6'00&B'LQ>@

7UDQVPLW
&5&

6'00&B'RXW>@

6KLIW
UHJLVWHU
5HFHLYH

DLE

The card databus width can be programmed using the clock control register. If the 4-bit wide
bus mode is enabled, data is transferred at four bits per clock cycle over all four data signals
(SDIO_D[3:0]). If the 8-bit wide bus mode is enabled, data is transferred at eight bits per
clock cycle over all eight data signals (SDIO_D[7:0]). If the wide bus mode is not enabled,
only one bit per clock cycle is transferred over SDIO_D0.
Depending on the transfer direction (send or receive), the data path state machine (DPSM)
moves to the Wait_S or Wait_R state when it is enabled:
•

Send: the DPSM moves to the Wait_S state. If there is data in the transmit FIFO, the
DPSM moves to the Send state, and the data path subunit starts sending data to a
card.

•

Receive: the DPSM moves to the Wait_R state and waits for a start bit. When it
receives a start bit, the DPSM moves to the Receive state, and the data path subunit
starts receiving data from a card.

Data path state machine (DPSM)
The DPSM operates at SDIO_CK frequency. Data on the card bus signals is synchronous to
the rising edge of SDIO_CK. The DPSM has six states, as shown in Figure 333: Data path
state machine (DPSM).

RM0090 Rev 17

1029/1747
1075

Secure digital input/output interface (SDIO)

RM0090

Figure 333. Data path state machine (DPSM)
/N RESET

$03- DISABLED

2EAD 7AIT

$03- ENABLED AND
2EAD 7AIT 3TARTED
AND 3$ )/ MODE ENABLED
$ISABLED OR &)&/ UNDERRUN OR
END OF DATA OR #2# FAIL

)DLE
$ISABLED OR #2# FAIL
OR TIMEOUT

%NABLE AND NOT SEND
2EAD7AIT 3TOP

$ISABLED OR
END OF DATA
$ISABLED OR
2X &)&/ EMPTY OR TIMEOUT OR
START BIT ERROR

"USY
.OT BUSY

%NABLE AND SEND

7AIT?2

$ATA RECEIVED AND
2EAD 7AIT 3TARTED AND
3$ )/ MODE ENABLED

%ND OF PACKET

7AIT?3

%ND OF PACKET OR
END OF DATA OR
&)&/ OVERRUN
$ISABLED OR #2# FAIL
3TART BIT

$ATA READY

3END

2ECEIVE
AIB

•

Idle: the data path is inactive, and the SDIO_D[7:0] outputs are in Hi-Z. When the data
control register is written and the enable bit is set, the DPSM loads the data counter
with a new value and, depending on the data direction bit, moves to either the Wait_S
or the Wait_R state.

•

Wait_R: if the data counter equals zero, the DPSM moves to the Idle state when the
receive FIFO is empty. If the data counter is not zero, the DPSM waits for a start bit on
SDIO_D. The DPSM moves to the Receive state if it receives a start bit before a
timeout, and loads the data block counter. If it reaches a timeout before it detects a
start bit, or a start bit error occurs, it moves to the Idle state and sets the timeout status
flag.

•

Receive: serial data received from a card is packed in bytes and written to the data
FIFO. Depending on the transfer mode bit in the data control register, the data transfer
mode can be either block or stream:
–

1030/1747

In block mode, when the data block counter reaches zero, the DPSM waits until it
receives the CRC code. If the received code matches the internally generated

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)
CRC code, the DPSM moves to the Wait_R state. If not, the CRC fail status flag is
set and the DPSM moves to the Idle state.
–

In stream mode, the DPSM receives data while the data counter is not zero. When
the counter is zero, the remaining data in the shift register is written to the data
FIFO, and the DPSM moves to the Wait_R state.

If a FIFO overrun error occurs, the DPSM sets the FIFO error flag and moves to the
Idle state:
•
Note:

Wait_S: the DPSM moves to the Idle state if the data counter is zero. If not, it waits until
the data FIFO empty flag is deasserted, and moves to the Send state.

The DPSM remains in the Wait_S state for at least two clock periods to meet the NWR timing
requirements, where NWR is the number of clock cycles between the reception of the card
response and the start of the data transfer from the host.
•

Send: the DPSM starts sending data to a card. Depending on the transfer mode bit in
the data control register, the data transfer mode can be either block or stream:
–

In block mode, when the data block counter reaches zero, the DPSM sends an
internally generated CRC code and end bit, and moves to the Busy state.

–

In stream mode, the DPSM sends data to a card while the enable bit is high and
the data counter is not zero. It then moves to the Idle state.

If a FIFO underrun error occurs, the DPSM sets the FIFO error flag and moves to the
Idle state.
•

Busy: the DPSM waits for the CRC status flag:
–

If it does not receive a positive CRC status, it moves to the Idle state and sets the
CRC fail status flag.

–

If it receives a positive CRC status, it moves to the Wait_S state if SDIO_D0 is not
low (the card is not busy).

If a timeout occurs while the DPSM is in the Busy state, it sets the data timeout flag and
moves to the Idle state.
The data timer is enabled when the DPSM is in the Wait_R or Busy state, and
generates the data timeout error:

•

–

When transmitting data, the timeout occurs if the DPSM stays in the Busy state for
longer than the programmed timeout period

–

When receiving data, the timeout occurs if the end of the data is not true, and if the
DPSM stays in the Wait_R state for longer than the programmed timeout period.

Data: data can be transferred from the card to the host or vice versa. Data is
transferred via the data lines. They are stored in a FIFO of 32 words, each word is 32
bits wide.
Table 155. Data token format
Description

Start bit

Data

CRC16

End bit

Block Data

0

-

yes

1

Stream Data

0

-

no

1

RM0090 Rev 17

1031/1747
1075

Secure digital input/output interface (SDIO)

RM0090

Data FIFO
The data FIFO (first-in-first-out) subunit is a data buffer with a transmit and receive unit.
The FIFO contains a 32-bit wide, 32-word deep data buffer, and transmit and receive logic.
Because the data FIFO operates in the APB2 clock domain (PCLK2), all signals from the
subunits in the SDIO clock domain (SDIOCLK) are resynchronized.
Depending on the TXACT and RXACT flags, the FIFO can be disabled, transmit enabled, or
receive enabled. TXACT and RXACT are driven by the data path subunit and are mutually
exclusive:

•

–

The transmit FIFO refers to the transmit logic and data buffer when TXACT is
asserted

–

The receive FIFO refers to the receive logic and data buffer when RXACT is
asserted

Transmit FIFO:
Data can be written to the transmit FIFO through the APB2 interface when the SDIO is
enabled for transmission.
The transmit FIFO is accessible via 32 sequential addresses. The transmit FIFO
contains a data output register that holds the data word pointed to by the read pointer.
When the data path subunit has loaded its shift register, it increments the read pointer
and drives new data out.
If the transmit FIFO is disabled, all status flags are deasserted. The data path subunit
asserts TXACT when it transmits data.
Table 156. Transmit FIFO status flags
Flag

Description

TXFIFOF

Set to high when all 32 transmit FIFO words contain valid data.

TXFIFOE

Set to high when the transmit FIFO does not contain valid data.

TXFIFOHE

Set to high when 8 or more transmit FIFO words are empty. This flag can be used
as a DMA request.

TXDAVL

Set to high when the transmit FIFO contains valid data. This flag is the inverse of
the TXFIFOE flag.

TXUNDERR

Set to high when an underrun error occurs. This flag is cleared by writing to the
SDIO Clear register.

•

Receive FIFO
When the data path subunit receives a word of data, it drives the data on the write
databus. The write pointer is incremented after the write operation completes. On the
read side, the contents of the FIFO word pointed to by the current value of the read
pointer is driven onto the read databus. If the receive FIFO is disabled, all status flags
are deasserted, and the read and write pointers are reset. The data path subunit
asserts RXACT when it receives data. Table 157 lists the receive FIFO status flags.
The receive FIFO is accessible via 32 sequential addresses.

1032/1747

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)
Table 157. Receive FIFO status flags
Flag

31.3.2

Description

RXFIFOF

Set to high when all 32 receive FIFO words contain valid data

RXFIFOE

Set to high when the receive FIFO does not contain valid data.

RXFIFOHF

Set to high when 8 or more receive FIFO words contain valid data. This flag can be
used as a DMA request.

RXDAVL

Set to high when the receive FIFO is not empty. This flag is the inverse of the
RXFIFOE flag.

RXOVERR

Set to high when an overrun error occurs. This flag is cleared by writing to the SDIO
Clear register.

SDIO APB2 interface
The APB2 interface generates the interrupt and DMA requests, and accesses the SDIO
adapter registers and the data FIFO. It consists of a data path, register decoder, and
interrupt/DMA logic.

SDIO interrupts
The interrupt logic generates an interrupt request signal that is asserted when at least one
of the selected status flags is high. A mask register is provided to allow selection of the
conditions that will generate an interrupt. A status flag generates the interrupt request if a
corresponding mask flag is set.

SDIO/DMA interface - procedure for data transfers between the SDIO and
memory
In the example shown, the transfer is from the SDIO host controller to an MMC (512 bytes
using CMD24 (WRITE_BLOCK). The SDIO FIFO is filled by data stored in a memory using
the DMA controller.
1.

Do the card identification process

2.

Increase the SDIO_CK frequency

3.

Select the card by sending CMD7

4.

Configure the DMA2 as follows:
a)

Enable DMA2 controller and clear any pending interrupts.

b)

Program the DMA2_Stream3 or DMA2_Stream6 Channel4 source address
register with the memory location’s base address and DMA2_Stream3 or

RM0090 Rev 17

1033/1747
1075

Secure digital input/output interface (SDIO)

RM0090

DMA2_Stream6 Channel4 destination address register with the SDIO_FIFO
register address.

5.

6.

c)

Program DMA2_Stream3 or DMA2_Stream6 Channel4 control register (memory
increment, not peripheral increment, peripheral and source width is word size).

d)

Program DMA2_Stream3 or DMA2_Stream6 Channel4 to select the peripheral as
flow controller (set PFCTRL bit in DMA_S3CR or DMA_S6CR configuration
register).

e)

Configure the incremental burst transfer to 4 beats (at least from peripheral side)
in DMA2_Stream3 or DMA2_Stream6 Channel4.

f)

Enable DMA2_Stream3 or DMA2_Stream6 Channel4

Send CMD24 (WRITE_BLOCK) as follows:
a)

Program the SDIO data length register (SDIO data timer register should be
already programmed before the card identification process).

b)

Program the SDIO argument register with the address location of the card where
data is to be transferred.

c)

Program the SDIO command register: CmdIndex with 24 (WRITE_BLOCK);
WaitResp with ‘1’ (SDIO card host waits for a response); CPSMEN with ‘1’ (SDIO
card host enabled to send a command). Other fields are at their reset value.

d)

Wait for SDIO_STA[6] = CMDREND interrupt, then program the SDIO data control
register: DTEN with ‘1’ (SDIO card host enabled to send data); DTDIR with ‘0’
(from controller to card); DTMODE with ‘0’ (block data transfer); DMAEN with ‘1’
(DMA enabled); DBLOCKSIZE with 0x9 (512 bytes). Other fields are don’t care.

e)

Wait for SDIO_STA[10] = DBCKEND.

Check that no channels are still enabled by polling the DMA Enabled Channel Status
register.

31.4

Card functional description

31.4.1

Card identification mode
While in card identification mode the host resets all cards, validates the operation voltage
range, identifies cards and sets a relative card address (RCA) for each card on the bus. All
data communications in the card identification mode use the command line (CMD) only.

31.4.2

Card reset
The GO_IDLE_STATE command (CMD0) is the software reset command and it puts the
MultiMediaCard and SD memory in the Idle state. The IO_RW_DIRECT command (CMD52)
resets the SD I/O card. After power-up or CMD0, all cards output bus drivers are in the highimpedance state and the cards are initialized with a default relative card address
(RCA=0x0001) and with a default driver stage register setting (lowest speed, highest driving
current capability).

31.4.3

Operating voltage range validation
All cards can communicate with the SDIO card host using any operating voltage within the
specification range. The supported minimum and maximum VDD values are defined in the
operation conditions register (OCR) on the card.

1034/1747

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)
Cards that store the card identification number (CID) and card specific data (CSD) in the
payload memory are able to communicate this information only under data-transfer VDD
conditions. When the SDIO card host module and the card have incompatible VDD ranges,
the card is not able to complete the identification cycle and cannot send CSD data. For this
purpose, the special commands, SEND_OP_COND (CMD1), SD_APP_OP_COND (ACMD41
for SD Memory), and IO_SEND_OP_COND (CMD5 for SD I/O), are designed to provide a
mechanism to identify and reject cards that do not match the VDD range desired by the
SDIO card host. The SDIO card host sends the required VDD voltage window as the
operand of these commands. Cards that cannot perform data transfer in the specified range
disconnect from the bus and go to the inactive state.
By using these commands without including the voltage range as the operand, the SDIO
card host can query each card and determine the common voltage range before placing outof-range cards in the inactive state. This query is used when the SDIO card host is able to
select a common voltage range or when the user requires notification that cards are not
usable.

31.4.4

Card identification process
The card identification process differs for MultiMediaCards and SD cards. For
MultiMediaCard cards, the identification process starts at clock rate Fod. The SDIO_CMD
line output drivers are open-drain and allow parallel card operation during this process. The
registration process is accomplished as follows:
1.

The bus is activated.

2.

The SDIO card host broadcasts SEND_OP_COND (CMD1) to receive operation
conditions.

3.

The response is the wired AND operation of the operation condition registers from all
cards.

4.

Incompatible cards are placed in the inactive state.

5.

The SDIO card host broadcasts ALL_SEND_CID (CMD2) to all active cards.

6.

The active cards simultaneously send their CID numbers serially. Cards with outgoing
CID bits that do not match the bits on the command line stop transmitting and must wait
for the next identification cycle. One card successfully transmits a full CID to the SDIO
card host and enters the Identification state.

7.

The SDIO card host issues SET_RELATIVE_ADDR (CMD3) to that card. This new
address is called the relative card address (RCA); it is shorter than the CID and
addresses the card. The assigned card changes to the Standby state, it does not react
to further identification cycles, and its output switches from open-drain to push-pull.

8.

The SDIO card host repeats steps 5 through 7 until it receives a timeout condition.

For the SD card, the identification process starts at clock rate Fod, and the SDIO_CMD line
output drives are push-pull drivers instead of open-drain. The registration process is
accomplished as follows:

RM0090 Rev 17

1035/1747
1075

Secure digital input/output interface (SDIO)

RM0090

1.

The bus is activated.

2.

The SDIO card host broadcasts SD_APP_OP_COND (ACMD41).

3.

The cards respond with the contents of their operation condition registers.

4.

The incompatible cards are placed in the inactive state.

5.

The SDIO card host broadcasts ALL_SEND_CID (CMD2) to all active cards.

6.

The cards send back their unique card identification numbers (CIDs) and enter the
Identification state.

7.

The SDIO card host issues SET_RELATIVE_ADDR (CMD3) to an active card with an
address. This new address is called the relative card address (RCA); it is shorter than
the CID and addresses the card. The assigned card changes to the Standby state. The
SDIO card host can reissue this command to change the RCA. The RCA of the card is
the last assigned value.

8.

The SDIO card host repeats steps 5 through 7 with all active cards.

For the SD I/O card, the registration process is accomplished as follows:
1.

31.4.5

The bus is activated.

2.

The SDIO card host sends IO_SEND_OP_COND (CMD5).

3.

The cards respond with the contents of their operation condition registers.

4.

The incompatible cards are set to the inactive state.

5.

The SDIO card host issues SET_RELATIVE_ADDR (CMD3) to an active card with an
address. This new address is called the relative card address (RCA); it is shorter than
the CID and addresses the card. The assigned card changes to the Standby state. The
SDIO card host can reissue this command to change the RCA. The RCA of the card is
the last assigned value.

Block write
During block write (CMD24 - 27) one or more blocks of data are transferred from the host to
the card with a CRC appended to the end of each block by the host. A card supporting block
write is always able to accept a block of data defined by WRITE_BL_LEN. If the CRC fails,
the card indicates the failure on the SDIO_D line and the transferred data are discarded and
not written, and all further transmitted blocks (in multiple block write mode) are ignored.
If the host uses partial blocks whose accumulated length is not block aligned and, block
misalignment is not allowed (CSD parameter WRITE_BLK_MISALIGN is not set), the card
will detect the block misalignment error before the beginning of the first misaligned block.
(ADDRESS_ERROR error bit is set in the status register). The write operation will also be
aborted if the host tries to write over a write-protected area. In this case, however, the card
will set the WP_VIOLATION bit.
Programming of the CID and CSD registers does not require a previous block length setting.
The transferred data is also CRC protected. If a part of the CSD or CID register is stored in
ROM, then this unchangeable part must match the corresponding part of the receive buffer.
If this match fails, then the card reports an error and does not change any register contents.
Some cards may require long and unpredictable times to write a block of data. After
receiving a block of data and completing the CRC check, the card begins writing and holds
the SDIO_D line low if its write buffer is full and unable to accept new data from a new
WRITE_BLOCK command. The host may poll the status of the card with a SEND_STATUS
command (CMD13) at any time, and the card will respond with its status. The
READY_FOR_DATA status bit indicates whether the card can accept new data or whether
the write process is still in progress. The host may deselect the card by issuing CMD7 (to

1036/1747

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)
select a different card), which will place the card in the Disconnect state and release the
SDIO_D line(s) without interrupting the write operation. When selecting the card again, it will
reactivate busy indication by pulling SDIO_D to low if programming is still in progress and
the write buffer is unavailable.

31.4.6

Block read
In Block read mode the basic unit of data transfer is a block whose maximum size is defined
in the CSD (READ_BL_LEN). If READ_BL_PARTIAL is set, smaller blocks whose start and
end addresses are entirely contained within one physical block (as defined by
READ_BL_LEN) may also be transmitted. A CRC is appended to the end of each block,
ensuring data transfer integrity. CMD17 (READ_SINGLE_BLOCK) initiates a block read and
after completing the transfer, the card returns to the Transfer state.
CMD18 (READ_MULTIPLE_BLOCK) starts a transfer of several consecutive blocks.
The host can abort reading at any time, within a multiple block operation, regardless of its
type. Transaction abort is done by sending the stop transmission command.
If the card detects an error (for example, out of range, address misalignment or internal
error) during a multiple block read operation (both types) it stops the data transmission and
remains in the data state. The host must than abort the operation by sending the stop
transmission command. The read error is reported in the response to the stop transmission
command.
If the host sends a stop transmission command after the card transmits the last block of a
multiple block operation with a predefined number of blocks, it is responded to as an illegal
command, since the card is no longer in the data state. If the host uses partial blocks whose
accumulated length is not block-aligned and block misalignment is not allowed, the card
detects a block misalignment error condition at the beginning of the first misaligned block
(ADDRESS_ERROR error bit is set in the status register).

31.4.7

Stream access, stream write and stream read
(MultiMediaCard only)
In stream mode, data is transferred in bytes and no CRC is appended at the end of each
block.

Stream write (MultiMediaCard only)
WRITE_DAT_UNTIL_STOP (CMD20) starts the data transfer from the SDIO card host to the
card, beginning at the specified address and continuing until the SDIO card host issues a
stop command. When partial blocks are allowed (CSD parameter WRITE_BL_PARTIAL is
set), the data stream can start and stop at any address within the card address space,
otherwise it can only start and stop at block boundaries. Because the amount of data to be
transferred is not determined in advance, a CRC cannot be used. When the end of the
memory range is reached while sending data and no stop command is sent by the SD card
host, any additional transferred data are discarded.

RM0090 Rev 17

1037/1747
1075

Secure digital input/output interface (SDIO)

RM0090

The maximum clock frequency for a stream write operation is given by the following
equation fields of the card-specific data register:
8 × 2 writebllen ) ( – NSAC ))
Maximumspeed = MIN (TRANSPEED,(------------------------------------------------------------------------TAAC × R2WFACTOR

•

Maximumspeed = maximum write frequency

•

TRANSPEED = maximum data transfer rate

•

writebllen = maximum write data block length

•

NSAC = data read access time 2 in CLK cycles

•

TAAC = data read access time 1

•

R2WFACTOR = write speed factor

If the host attempts to use a higher frequency, the card may not be able to process the data
and stop programming, set the OVERRUN error bit in the status register, and while ignoring
all further data transfer, wait (in the receive data state) for a stop command. The write
operation is also aborted if the host tries to write over a write-protected area. In this case,
however, the card sets the WP_VIOLATION bit.

Stream read (MultiMediaCard only)
READ_DAT_UNTIL_STOP (CMD11) controls a stream-oriented data transfer.
This command instructs the card to send its data, starting at a specified address, until the
SDIO card host sends STOP_TRANSMISSION (CMD12). The stop command has an
execution delay due to the serial command transmission and the data transfer stops after
the end bit of the stop command. When the end of the memory range is reached while
sending data and no stop command is sent by the SDIO card host, any subsequent data
sent are considered undefined.
The maximum clock frequency for a stream read operation is given by the following
equation and uses fields of the card specific data register.
( 8 × 2 readbllen ) ( – NSAC ))
Maximumspeed = MIN (TRANSPEED,-----------------------------------------------------------------------TAAC × R2WFACTOR

•

Maximumspeed = maximum read frequency

•

TRANSPEED = maximum data transfer rate

•

readbllen = maximum read data block length

•

writebllen = maximum write data block length

•

NSAC = data read access time 2 in CLK cycles

•

TAAC = data read access time 1

•

R2WFACTOR = write speed factor

If the host attempts to use a higher frequency, the card is not able to sustain data transfer. If
this happens, the card sets the UNDERRUN error bit in the status register, aborts the
transmission and waits in the data state for a stop command.

1038/1747

RM0090 Rev 17

RM0090

31.4.8

Secure digital input/output interface (SDIO)

Erase: group erase and sector erase
The erasable unit of the MultiMediaCard is the erase group. The erase group is measured in
write blocks, which are the basic writable units of the card. The size of the erase group is a
card-specific parameter and defined in the CSD.
The host can erase a contiguous range of Erase Groups. Starting the erase process is a
three-step sequence.
First the host defines the start address of the range using the ERASE_GROUP_START
(CMD35) command, next it defines the last address of the range using the
ERASE_GROUP_END (CMD36) command and, finally, it starts the erase process by issuing
the ERASE (CMD38) command. The address field in the erase commands is an Erase
Group address in byte units. The card ignores all LSBs below the Erase Group size,
effectively rounding the address down to the Erase Group boundary.
If an erase command is received out of sequence, the card sets the ERASE_SEQ_ERROR
bit in the status register and resets the whole sequence.
If an out-of-sequence (neither of the erase commands, except SEND_STATUS) command
received, the card sets the ERASE_RESET status bit in the status register, resets the erase
sequence and executes the last command.
If the erase range includes write protected blocks, they are left intact and only unprotected
blocks are erased. The WP_ERASE_SKIP status bit in the status register is set.
The card indicates that an erase is in progress by holding SDIO_D low. The actual erase
time may be quite long, and the host may issue CMD7 to deselect the card.

31.4.9

Wide bus selection or deselection
Wide bus (4-bit bus width) operation mode is selected or deselected using
SET_BUS_WIDTH (ACMD6). The default bus width after power-up or GO_IDLE_STATE
(CMD0) is 1 bit. SET_BUS_WIDTH (ACMD6) is only valid in a transfer state, which means
that the bus width can be changed only after a card is selected by
SELECT/DESELECT_CARD (CMD7).

31.4.10

Protection management
Three write protection methods for the cards are supported in the SDIO card host module:
1.

internal card write protection (card responsibility)

2.

mechanical write protection switch (SDIO card host module responsibility only)

3.

password-protected card lock operation

Internal card write protection
Card data can be protected against write and erase. By setting the permanent or temporary
write-protect bits in the CSD, the entire card can be permanently write-protected by the
manufacturer or content provider. For cards that support write protection of groups of
sectors by setting the WP_GRP_ENABLE bit in the CSD, portions of the data can be
protected, and the write protection can be changed by the application. The write protection
is in units of WP_GRP_SIZE sectors as specified in the CSD. The SET_WRITE_PROT and
CLR_WRITE_PROT commands control the protection of the addressed group. The
SEND_WRITE_PROT command is similar to a single block read command. The card sends
a data block containing 32 write protection bits (representing 32 write protect groups starting

RM0090 Rev 17

1039/1747
1075

Secure digital input/output interface (SDIO)

RM0090

at the specified address) followed by 16 CRC bits. The address field in the write protect
commands is a group address in byte units.
The card ignores all LSBs below the group size.

Mechanical write protect switch
A mechanical sliding tab on the side of the card allows the user to set or clear the write
protection on a card. When the sliding tab is positioned with the window open, the card is
write-protected, and when the window is closed, the card contents can be changed. A
matched switch on the socket side indicates to the SDIO card host module that the card is
write-protected. The SDIO card host module is responsible for protecting the card. The
position of the write protect switch is unknown to the internal circuitry of the card.

Password protect
The password protection feature enables the SDIO card host module to lock and unlock a
card with a password. The password is stored in the 128-bit PWD register and its size is set
in the 8-bit PWD_LEN register. These registers are nonvolatile so that a power cycle does
not erase them. Locked cards respond to and execute certain commands. This means that
the SDIO card host module is allowed to reset, initialize, select, and query for status,
however it is not allowed to access data on the card. When the password is set (as indicated
by a nonzero value of PWD_LEN), the card is locked automatically after power-up. As with
the CSD and CID register write commands, the lock/unlock commands are available in the
transfer state only. In this state, the command does not include an address argument and
the card must be selected before using it. The card lock/unlock commands have the
structure and bus transaction types of a regular single-block write command. The
transferred data block includes all of the required information for the command (the
password setting mode, the PWD itself, and card lock/unlock). The command data block
size is defined by the SDIO card host module before it sends the card lock/unlock
command, and has the structure shown in Table 171.
The bit settings are as follows:
•

ERASE: setting it forces an erase operation. All other bits must be zero, and only the
command byte is sent

•

LOCK_UNLOCK: setting it locks the card. LOCK_UNLOCK can be set simultaneously
with SET_PWD, however not with CLR_PWD

•

CLR_PWD: setting it clears the password data

•

SET_PWD: setting it saves the password data to memory

•

PWD_LEN: it defines the length of the password in bytes

•

PWD: the password (new or currently used, depending on the command)

The following sections list the command sequences to set/reset a password, lock/unlock the
card, and force an erase.

Setting the password

1040/1747

1.

Select a card (SELECT/DESELECT_CARD, CMD7), if none is already selected.

2.

Define the block length (SET_BLOCKLEN, CMD16) to send, given by the 8-bit card
lock/unlock mode, the 8-bit PWD_LEN, and the number of bytes of the new password.

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)
When a password replacement is done, the block size must take into account that both
the old and the new passwords are sent with the command.
3.

Send LOCK/UNLOCK (CMD42) with the appropriate data block size on the data line
including the 16-bit CRC. The data block indicates the mode (SET_PWD = 1), the
length (PWD_LEN), and the password (PWD) itself. When a password replacement is
done, the length value (PWD_LEN) includes the length of both passwords, the old and
the new one, and the PWD field includes the old password (currently used) followed by
the new password.

4.

When the password is matched, the new password and its size are saved into the PWD
and PWD_LEN fields, respectively. When the old password sent does not correspond
(in size and/or content) to the expected password, the LOCK_UNLOCK_FAILED error
bit is set in the card status register, and the password is not changed.

The password length field (PWD_LEN) indicates whether a password is currently set. When
this field is nonzero, there is a password set and the card locks itself after power-up. It is
possible to lock the card immediately in the current power session by setting the
LOCK_UNLOCK bit (while setting the password) or sending an additional command for card
locking.

Resetting the password
1.

Select a card (SELECT/DESELECT_CARD, CMD7), if none is already selected.

2.

Define the block length (SET_BLOCKLEN, CMD16) to send, given by the 8-bit card
lock/unlock mode, the 8-bit PWD_LEN, and the number of bytes in the currently used
password.

3.

Send LOCK/UNLOCK (CMD42) with the appropriate data block size on the data line
including the 16-bit CRC. The data block indicates the mode (CLR_PWD = 1), the
length (PWD_LEN) and the password (PWD) itself. The LOCK_UNLOCK bit is ignored.

4.

When the password is matched, the PWD field is cleared and PWD_LEN is set to 0.
When the password sent does not correspond (in size and/or content) to the expected
password, the LOCK_UNLOCK_FAILED error bit is set in the card status register, and
the password is not changed.

Locking a card
1.

Select a card (SELECT/DESELECT_CARD, CMD7), if none is already selected.

2.

Define the block length (SET_BLOCKLEN, CMD16) to send, given by the 8-bit card
lock/unlock mode (byte 0 in Table 171), the 8-bit PWD_LEN, and the number of bytes
of the current password.

3.

Send LOCK/UNLOCK (CMD42) with the appropriate data block size on the data line
including the 16-bit CRC. The data block indicates the mode (LOCK_UNLOCK = 1), the
length (PWD_LEN), and the password (PWD) itself.

4.

When the password is matched, the card is locked and the CARD_IS_LOCKED status
bit is set in the card status register. When the password sent does not correspond (in
size and/or content) to the expected password, the LOCK_UNLOCK_FAILED error bit
is set in the card status register, and the lock fails.

It is possible to set the password and to lock the card in the same sequence. In this case,
the SDIO card host module performs all the required steps for setting the password (see
Setting the password on page 1040), however it is necessary to set the LOCK_UNLOCK bit
in Step 3 when the new password command is sent.

RM0090 Rev 17

1041/1747
1075

Secure digital input/output interface (SDIO)

RM0090

When the password is previously set (PWD_LEN is not 0), the card is locked automatically
after power-on reset. An attempt to lock a locked card or to lock a card that does not have a
password fails and the LOCK_UNLOCK_FAILED error bit is set in the card status register.

Unlocking the card
1.

Select a card (SELECT/DESELECT_CARD, CMD7), if none is already selected.

2.

Define the block length (SET_BLOCKLEN, CMD16) to send, given by the 8-bit
cardlock/unlock mode (byte 0 in Table 171), the 8-bit PWD_LEN, and the number of
bytes of the current password.

3.

Send LOCK/UNLOCK (CMD42) with the appropriate data block size on the data line
including the 16-bit CRC. The data block indicates the mode (LOCK_UNLOCK = 0), the
length (PWD_LEN), and the password (PWD) itself.

4.

When the password is matched, the card is unlocked and the CARD_IS_LOCKED
status bit is cleared in the card status register. When the password sent is not correct in
size and/or content and does not correspond to the expected password, the
LOCK_UNLOCK_FAILED error bit is set in the card status register, and the card
remains locked.

The unlocking function is only valid for the current power session. When the PWD field is not
clear, the card is locked automatically on the next power-up.
An attempt to unlock an unlocked card fails and the LOCK_UNLOCK_FAILED error bit is set
in the card status register.

Forcing erase
If the user has forgotten the password (PWD content), it is possible to access the card after
clearing all the data on the card. This forced erase operation erases all card data and all
password data.
1.

Select a card (SELECT/DESELECT_CARD, CMD7), if none is already selected.

2.

Set the block length (SET_BLOCKLEN, CMD16) to 1 byte. Only the 8-bit card
lock/unlock byte (byte 0 in Table 171) is sent.

3.

Send LOCK/UNLOCK (CMD42) with the appropriate data byte on the data line including
the 16-bit CRC. The data block indicates the mode (ERASE = 1). All other bits must be
zero.

4.

When the ERASE bit is the only bit set in the data field, all card contents are erased,
including the PWD and PWD_LEN fields, and the card is no longer locked. When any
other bits are set, the LOCK_UNLOCK_FAILED error bit is set in the card status
register and the card retains all of its data, and remains locked.

An attempt to use a force erase on an unlocked card fails and the LOCK_UNLOCK_FAILED
error bit is set in the card status register.

31.4.11

Card status register
The response format R1 contains a 32-bit field named card status. This field is intended to
transmit the card status information (which may be stored in a local status register) to the
host. If not specified otherwise, the status entries are always related to the previously issued
command.
Table 158 defines the different entries of the status. The type and clear condition fields in
the table are abbreviated as follows:

1042/1747

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)
Type:
•

E: error bit

•

S: status bit

•

R: detected and set for the actual command response

•

X: detected and set during command execution. The SDIO card host must poll the card
by issuing the status command to read these bits.

Clear condition:
•

A: according to the card current state

•

B: always related to the previous command. Reception of a valid command clears it
(with a delay of one command)

•

C: clear by read
Table 158. Card status

Bits

31

30

Identifier

ADDRESS_
OUT_OF_RANGE

ADDRESS_MISALIGN

Type

ERX

-

Value

Description

Clear
condition

’0’= no error
’1’= error

The command address argument was out
of the allowed range for this card.
A multiple block or stream read/write
C
operation is (although started in a valid
address) attempting to read or write
beyond the card capacity.

’0’= no error
’1’= error

The commands address argument (in
accordance with the currently set block
length) positions the first data block
misaligned to the card physical blocks.
A multiple block read/write operation
(although started with a valid
address/block-length combination) is
attempting to read or write a data block
which is not aligned with the physical
blocks of the card.

C

29

BLOCK_LEN_ERROR

-

’0’= no error
’1’= error

Either the argument of a
SET_BLOCKLEN command exceeds the
maximum value allowed for the card, or
the previously defined block length is
illegal for the current command (e.g. the C
host issues a write command, the current
block length is smaller than the maximum
allowed value for the card and it is not
allowed to write partial blocks)

28

ERASE_SEQ_ERROR

-

’0’= no error
’1’= error

An error in the sequence of erase
commands occurred.

C

27

ERASE_PARAM

EX

’0’= no error
’1’= error

An invalid selection of erase groups for
erase occurred.

C

26

WP_VIOLATION

EX

’0’= no error
’1’= error

Attempt to program a write-protected
block.

RM0090 Rev 17

C

1043/1747
1075

Secure digital input/output interface (SDIO)

RM0090

Table 158. Card status (continued)
Bits

Identifier

Type

Value

Description

Clear
condition

25

CARD_IS_LOCKED

SR

‘0’ = card
unlocked
‘1’ = card locked

When set, signals that the card is locked
by the host

A

24

LOCK_UNLOCK_
FAILED

EX

’0’= no error
’1’= error

Set when a sequence or password error
has been detected in lock/unlock card
command

C

23

COM_CRC_ERROR

ER

’0’= no error
’1’= error

The CRC check of the previous command
B
failed.

22

ILLEGAL_COMMAND

ER

’0’= no error
’1’= error

Command not legal for the card state

B

21

CARD_ECC_FAILED

EX

’0’= success
’1’= failure

Card internal ECC was applied but failed
to correct the data.

C

20

CC_ERROR

ER

’0’= no error
’1’= error

(Undefined by the standard) A card error
occurred, which is not related to the host
command.

C

EX

’0’= no error
’1’= error

(Undefined by the standard) A generic
card error related to the (and detected
during) execution of the last host
command (e.g. read or write failures).

C

Can be either of the following errors:
– The CID register has already been
written and cannot be overwritten
– The read-only section of the CSD does C
not match the card contents
– An attempt to reverse the copy (set as
original) or permanent WP
(unprotected) bits was made

19

ERROR

18

Reserved

17

Reserved

16

CID/CSD_OVERWRITE

EX

’0’= no error ‘1’=
error

15

WP_ERASE_SKIP

EX

’0’= not protected Set when only partial address space
’1’= protected
was erased due to existing write

14

CARD_ECC_DISABLED S X

13

ERASE_RESET

1044/1747

-

C

’0’= enabled
’1’= disabled

The command has been executed without
A
using the internal ECC.

’0’= cleared
’1’= set

An erase sequence was cleared before
executing because an out of erase
sequence command was received
(commands other than CMD35, CMD36,
CMD38 or CMD13)

RM0090 Rev 17

C

RM0090

Secure digital input/output interface (SDIO)
Table 158. Card status (continued)

Bits

Identifier

Type

Value

Description

Clear
condition

12:9

CURRENT_STATE

SR

0 = Idle
1 = Ready
2 = Ident
3 = Stby
4 = Tran
5 = Data
6 = Rcv
7 = Prg
8 = Dis
9 = Btst
10-15 = reserved

8

READY_FOR_DATA

SR

’0’= not ready ‘1’
= ready

Corresponds to buffer empty signalling on
the bus

7

SWITCH_ERROR

EX

’0’= no error
’1’= switch error

If set, the card did not switch to the
expected mode as requested by the
SWITCH command

B

6

Reserved

5

APP_CMD

SR

‘0’ = Disabled
‘1’ = Enabled

The card will expect ACMD, or an
indication that the command has been
interpreted as ACMD

C

4

Reserved for SD I/O Card

3

AKE_SEQ_ERROR

ER

’0’= no error
’1’= error

Error in the sequence of the
authentication process

C

2

Reserved for application specific commands

1
0

The state of the card when receiving the
command. If the command execution
causes a state change, it will be visible to
B
the host in the response on the next
command. The four bits are interpreted as
a binary number between 0 and 15.

-

Reserved for manufacturer test mode

31.4.12

SD status register
The SD status contains status bits that are related to the SD memory card proprietary
features and may be used for future application-specific usage. The size of the SD Status is
one data block of 512 bits. The contents of this register are transmitted to the SDIO card
host if ACMD13 is sent (CMD55 followed with CMD13). ACMD13 can be sent to a card in
transfer state only (card is selected).
Table 159 defines the different entries of the SD status register. The type and clear condition
fields in the table are abbreviated as follows:
Type:
•

E: error bit

•

S: status bit

•

R: detected and set for the actual command response

•

X: detected and set during command execution. The SDIO card Host must poll the card
by issuing the status command to read these bits

RM0090 Rev 17

1045/1747
1075

Secure digital input/output interface (SDIO)

RM0090

Clear condition:
•

A: according to the card current state

•

B: always related to the previous command. Reception of a valid command clears it
(with a delay of one command)

•

C: clear by read
Table 159. SD status

Bits

Identifier

Type

Value

Description

Clear
condition

511: 510 DAT_BUS_WIDTH S R

’00’= 1 (default)
‘01’= reserved
‘10’= 4 bit width
‘11’= reserved

Shows the currently defined
databus width that was
defined by
SET_BUS_WIDTH
command

A

509

’0’= Not in the mode
’1’= In Secured Mode

Card is in Secured Mode of
operation (refer to the “SD
Security Specification”).

A

’00xxh’= SD Memory Cards as
defined in Physical Spec Ver1.012.00 (’x’= don’t care). The
following cards are currently
defined:
’0000’= Regular SD RD/WR Card.
’0001’= SD ROM Card

In the future, the 8 LSBs will
be used to define different
variations of an SD memory
card (each bit will define
different SD types). The 8
A
MSBs will be used to define
SD Cards that do not comply
with current SD physical
layer specification.

Size of protected area (See
below)

(See below)

A

Speed Class of the card (See
below)

(See below)

A

SECURED_MODE S R

508: 496 Reserved

495: 480 SD_CARD_TYPE

479: 448

SR

SIZE_OF_PROTE
SR
CT ED_AREA

447: 440 SPEED_CLASS

SR

439: 432

PERFORMANCE_
SR
MOVE

Performance of move indicated by
1 [MB/s] step.
(See below)
(See below)

A

431:428

AU_SIZE

SR

Size of AU
(See below)

(See below)

A

427:424

Reserved

423:408

ERASE_SIZE

SR

Number of AUs to be erased at a
time

(See below)

A

407:402

ERASE_TIMEOUT S R

Timeout value for erasing areas
specified by
UNIT_OF_ERASE_AU

(See below)

A

401:400

ERASE_OFFSET

Fixed offset value added to erase
(See below)
time.

A

399:312

Reserved

311:0

Reserved for Manufacturer

1046/1747

SR

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)

SIZE_OF_PROTECTED_AREA
Setting this field differs between standard- and high-capacity cards. In the case of a
standard-capacity card, the capacity of protected area is calculated as follows:
Protected area = SIZE_OF_PROTECTED_AREA_* MULT * BLOCK_LEN.
SIZE_OF_PROTECTED_AREA is specified by the unit in MULT*BLOCK_LEN.
In the case of a high-capacity card, the capacity of protected area is specified in this field:
Protected area = SIZE_OF_PROTECTED_AREA
SIZE_OF_PROTECTED_AREA is specified by the unit in bytes.

SPEED_CLASS
This 8-bit field indicates the speed class and the value can be calculated by PW/2 (where
PW is the write performance).
Table 160. Speed class code field
SPEED_CLASS

Value definition

00h

Class 0

01h

Class 2

02h

Class 4

03h

Class 6

04h – FFh

Reserved

PERFORMANCE_MOVE
This 8-bit field indicates Pm (performance move) and the value can be set by 1 [MB/sec]
steps. If the card does not move used RUs (recording units), Pm should be considered as
infinity. Setting the field to FFh means infinity.
Table 161. Performance move field
PERFORMANCE_MOVE

Value definition

00h

Not defined

01h

1 [MB/sec]

02h

02h 2 [MB/sec]

---------

---------

FEh

254 [MB/sec]

FFh

Infinity

RM0090 Rev 17

1047/1747
1075

Secure digital input/output interface (SDIO)

RM0090

AU_SIZE
This 4-bit field indicates the AU size and the value can be selected in the power of 2 base
from 16 KB.
Table 162. AU_SIZE field
AU_SIZE

Value definition

00h

Not defined

01h

16 KB

02h

32 KB

03h

64 KB

04h

128 KB

05h

256 KB

06h

512 KB

07h

1 MB

08h

2 MB

09h

4 MB

Ah – Fh

Reserved

The maximum AU size, which depends on the card capacity, is defined in Table 163. The
card can be set to any AU size between RU size and maximum AU size.
Table 163. Maximum AU size
Capacity
Maximum AU Size

16 MB-64 MB
512 KB

128 MB-256 MB
1 MB

512 MB
2 MB

1 GB-32 GB
4 MB

ERASE_SIZE
This 16-bit field indicates NERASE. When NERASE numbers of AUs are erased, the
timeout value is specified by ERASE_TIMEOUT (Refer to ERASE_TIMEOUT). The host
should determine the proper number of AUs to be erased in one operation so that the host
can show the progress of the erase operation. If this field is set to 0, the erase timeout
calculation is not supported.
Table 164. Erase size field
ERASE_SIZE

1048/1747

Value definition

0000h

Erase timeout calculation is not supported.

0001h

1 AU

0002h

2 AU

0003h

3 AU

---------

---------

FFFFh

65535 AU

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)

ERASE_TIMEOUT
This 6-bit field indicates TERASE and the value indicates the erase timeout from offset
when multiple AUs are being erased as specified by ERASE_SIZE. The range of
ERASE_TIMEOUT can be defined as up to 63 seconds and the card manufacturer can
choose any combination of ERASE_SIZE and ERASE_TIMEOUT depending on the
implementation. Determining ERASE_TIMEOUT determines the ERASE_SIZE.
Table 165. Erase timeout field
ERASE_TIMEOUT

Value definition

00

Erase timeout calculation is not supported.

01

1 [sec]

02

2 [sec]

03

3 [sec]

---------

---------

63

63 [sec]

ERASE_OFFSET
This 2-bit field indicates TOFFSET and one of four values can be selected. This field is
meaningless if the ERASE_SIZE and ERASE_TIMEOUT fields are set to 0.
Table 166. Erase offset field
ERASE_OFFSET

31.4.13

Value definition

0h

0 [sec]

1h

1 [sec]

2h

2 [sec]

3h

3 [sec]

SD I/O mode
SD I/O interrupts
To allow the SD I/O card to interrupt the MultiMediaCard/SD module, an interrupt function is
available on a pin on the SD interface. Pin 8, used as SDIO_D1 when operating in the 4-bit
SD mode, signals the cards interrupt to the MultiMediaCard/SD module. The use of the
interrupt is optional for each card or function within a card. The SD I/O interrupt is levelsensitive, which means that the interrupt line must be held active (low) until it is either
recognized and acted upon by the MultiMediaCard/SD module or deasserted due to the end
of the interrupt period. After the MultiMediaCard/SD module has serviced the interrupt, the
interrupt status bit is cleared via an I/O write to the appropriate bit in the SD I/O card’s
internal registers. The interrupt output of all SD I/O cards is active low and the application
must provide external pull-up resistors on all data lines (SDIO_D[3:0]). The
MultiMediaCard/SD module samples the level of pin 8 (SDIO_D/IRQ) into the interrupt
detector only during the interrupt period. At all other times, the MultiMediaCard/SD module
ignores this value.

RM0090 Rev 17

1049/1747
1075

Secure digital input/output interface (SDIO)

RM0090

The interrupt period is applicable for both memory and I/O operations. The definition of the
interrupt period for operations with single blocks is different from the definition for multipleblock data transfers.

SD I/O suspend and resume
Within a multifunction SD I/O or a card with both I/O and memory functions, there are
multiple devices (I/O and memory) that share access to the MMC/SD bus. To share access
to the MMC/SD module among multiple devices, SD I/O and combo cards optionally
implement the concept of suspend/resume. When a card supports suspend/resume, the
MMC/SD module can temporarily halt a data transfer operation to one function or memory
(suspend) to free the bus for a higher-priority transfer to a different function or memory. After
this higher-priority transfer is complete, the original transfer is resumed (restarted) where it
left off. Support of suspend/resume is optional on a per-card basis. To perform the
suspend/resume operation on the MMC/SD bus, the MMC/SD module performs the
following steps:
1.

Determines the function currently using the SDIO_D [3:0] line(s)

2.

Requests the lower-priority or slower transaction to suspend

3.

Waits for the transaction suspension to complete

4.

Begins the higher-priority transaction

5.

Waits for the completion of the higher priority transaction

6.

Restores the suspended transaction

SD I/O ReadWait
The optional ReadWait (RW) operation is defined only for the SD 1-bit and 4-bit modes. The
ReadWait operation allows the MMC/SD module to signal a card that it is reading multiple
registers (IO_RW_EXTENDED, CMD53) to temporarily stall the data transfer while allowing
the MMC/SD module to send commands to any function within the SD I/O device. To
determine when a card supports the ReadWait protocol, the MMC/SD module must test
capability bits in the internal card registers. The timing for ReadWait is based on the
interrupt period.

31.4.14

Commands and responses
Application-specific and general commands
The SD card host module system is designed to provide a standard interface for a variety of
applications types. In this environment, there is a need for specific customer/application
features. To implement these features, two types of generic commands are defined in the
standard: application-specific commands (ACMD) and general commands (GEN_CMD).
When the card receives the APP_CMD (CMD55) command, the card expects the next
command to be an application-specific command. ACMDs have the same structure as
regular MultiMediaCard commands and can have the same CMD number. The card
recognizes it as ACMD because it appears after APP_CMD (CMD55). When the command
immediately following the APP_CMD (CMD55) is not a defined application-specific
command, the standard command is used. For example, when the card has a definition for
SD_STATUS (ACMD13), and receives CMD13 immediately following APP_CMD (CMD55),
this is interpreted as SD_STATUS (ACMD13). However, when the card receives CMD7
immediately following APP_CMD (CMD55) and the card does not have a definition for
ACMD7, this is interpreted as the standard (SELECT/DESELECT_CARD) CMD7.

1050/1747

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)
To use one of the manufacturer-specific ACMDs the SD card Host must perform the
following steps:
1.

Send APP_CMD (CMD55)
The card responds to the MultiMediaCard/SD module, indicating that the APP_CMD bit
is set and an ACMD is now expected.

2.

Send the required ACMD
The card responds to the MultiMediaCard/SD module, indicating that the APP_CMD bit
is set and that the accepted command is interpreted as an ACMD. When a nonACMD
is sent, it is handled by the card as a normal MultiMediaCard command and the
APP_CMD bit in the card status register stays clear.

When an invalid command is sent (neither ACMD nor CMD) it is handled as a standard
MultiMediaCard illegal command error.
The bus transaction for a GEN_CMD is the same as the single-block read or write
commands (WRITE_BLOCK, CMD24 or READ_SINGLE_BLOCK,CMD17). In this case, the
argument denotes the direction of the data transfer rather than the address, and the data
block has vendor-specific format and meaning.
The card must be selected (in transfer state) before sending GEN_CMD (CMD56). The data
block size is defined by SET_BLOCKLEN (CMD16). The response to GEN_CMD (CMD56)
is in R1b format.

Command types
Both application-specific and general commands are divided into the four following types:
•

broadcast command (BC): sent to all cards; no responses returned.

•

broadcast command with response (BCR): sent to all cards; responses received
from all cards simultaneously.

•

addressed (point-to-point) command (AC): sent to the card that is selected; does
not include a data transfer on the SDIO_D line(s).

•

addressed (point-to-point) data transfer command (ADTC): sent to the card that is
selected; includes a data transfer on the SDIO_D line(s).

Command formats
See Table 151 on page 1027 for command formats.

Commands for the MultiMediaCard/SD module
Table 167. Block-oriented write commands
CMD
index

Type

Argument

Response
format

Abbreviation

Description

CMD23 ac

[31:16] set to 0
[15:0] number R1
of blocks

SET_BLOCK_COUNT

Defines the number of blocks which
are going to be transferred in the
multiple-block read or write command
that follows.

CMD24 adtc

[31:0] data
address

WRITE_BLOCK

Writes a block of the size selected by
the SET_BLOCKLEN command.

R1

RM0090 Rev 17

1051/1747
1075

Secure digital input/output interface (SDIO)

RM0090

Table 167. Block-oriented write commands (continued)
CMD
index

Type

CMD25 adtc

Argument

[31:0] data
address

Response
format

Abbreviation

Description

Continuously writes blocks of data
until a STOP_TRANSMISSION
WRITE_MULTIPLE_BLOCK
follows or the requested number of
blocks has been received.

R1

CMD26 adtc

[31:0] stuff bits R1

PROGRAM_CID

Programming of the card identification
register. This command must be
issued only once per card. The card
contains hardware to prevent this
operation after the first programming.
Normally this command is reserved
for manufacturer.

CMD27 adtc

[31:0] stuff bits R1

PROGRAM_CSD

Programming of the programmable
bits of the CSD.

Table 168. Block-oriented write protection commands
CMD
index

Type

Argument

Response
format

Abbreviation

Description

CMD28 ac

[31:0] data
address

R1b

SET_WRITE_PROT

If the card has write protection features,
this command sets the write protection bit
of the addressed group. The properties of
write protection are coded in the cardspecific data (WP_GRP_SIZE).

CMD29 ac

[31:0] data
address

R1b

CLR_WRITE_PROT

If the card provides write protection
features, this command clears the write
protection bit of the addressed group.

CMD30 adtc

[31:0] write
protect data
address

SEND_WRITE_PROT

If the card provides write protection
features, this command asks the card to
send the status of the write protection
bits.

R1

CMD31 Reserved

Table 169. Erase commands
CMD
index

Type

Argument

Response
format

Abbreviation

Description

CMD32
Reserved. These command indexes cannot be used in order to maintain backward compatibility with older
...
versions of the MultiMediaCard.
CMD34
CMD35 ac

[31:0] data address R1

Sets the address of the first erase
ERASE_GROUP_START group within a range to be selected
for erase.

CMD36 ac

[31:0] data address R1

ERASE_GROUP_END

1052/1747

RM0090 Rev 17

Sets the address of the last erase
group within a continuous range to be
selected for erase.

RM0090

Secure digital input/output interface (SDIO)
Table 169. Erase commands (continued)

CMD
index
CMD37

Type

Response
format

Argument

Abbreviation

Description

Reserved. This command index cannot be used in order to maintain backward compatibility with older
versions of the MultiMediaCards

CMD38 ac

[31:0] stuff bits

R1

Erases all previously selected write
blocks.

ERASE

Table 170. I/O mode commands
CMD
index

Type

Response
format

Argument

Abbreviation

Description
Used to write and read 8-bit (register) data
fields. The command addresses a card and a
register and provides the data for writing if
the write flag is set. The R4 response
contains data read from the addressed
register. This command accesses
application-dependent registers that are not
defined in the MultiMediaCard standard.

CMD39 ac

[31:16] RCA
[15:15] register
write flag
[14:8] register
address
[7:0] register data

R4

FAST_IO

CMD40 bcr

[31:0] stuff bits

R5

GO_IRQ_STATE Places the system in the interrupt mode.

CMD41 Reserved

Table 171. Lock card
CMD
index

Type

CMD42 adtc

Response
format

Argument

[31:0] stuff bits

R1b

Abbreviation

LOCK_UNLOCK

Description
Sets/resets the password or locks/unlocks
the card. The size of the data block is set
by the SET_BLOCK_LEN command.

CMD43
...
Reserved
CMD54

Table 172. Application-specific commands
CMD
index
CMD55

CMD56

Type

ac

adtc

Argument
[31:16] RCA
[15:0] stuff bits

Response
format
R1

Abbreviation

APP_CMD

Description
Indicates to the card that the next command
bits is an application specific command rather
than a standard command
Used either to transfer a data block to the card
or to get a data block from the card for general
purpose/application-specific commands. The
size of the data block shall be set by the
SET_BLOCK_LEN command.

[31:1] stuff bits
[0]: RD/WR

RM0090 Rev 17

1053/1747
1075

Secure digital input/output interface (SDIO)

RM0090

Table 172. Application-specific commands (continued)
CMD
index

Type

Argument

Response
format

CMD57
...
CMD59

Reserved.

CMD60
...
CMD63

Reserved for manufacturer.

31.5

Abbreviation

Description

Response formats
All responses are sent via the MCCMD command line SDIO_CMD. The response
transmission always starts with the left bit of the bit string corresponding to the response
code word. The code length depends on the response type.
A response always starts with a start bit (always 0), followed by the bit indicating the
direction of transmission (card = 0). A value denoted by x in the tables below indicates a
variable entry. All responses, except for the R3 response type, are protected by a CRC.
Every command code word is terminated by the end bit (always 1).
There are five types of responses. Their formats are defined as follows:

31.5.1

R1 (normal response command)
Code length = 48 bits. The 45:40 bits indicate the index of the command to be responded to,
this value being interpreted as a binary-coded number (between 0 and 63). The status of the
card is coded in 32 bits.
Table 173. R1 response
Bit position

31.5.2

Width (bits

Value

Description

47

1

0

Start bit

46

1

0

Transmission bit

[45:40]

6

X

Command index

[39:8]

32

X

Card status

[7:1]

7

X

CRC7

0

1

1

End bit

R1b
It is identical to R1 with an optional busy signal transmitted on the data line. The card may
become busy after receiving these commands based on its state prior to the command
reception.

31.5.3

R2 (CID, CSD register)
Code length = 136 bits. The contents of the CID register are sent as a response to the
CMD2 and CMD10 commands. The contents of the CSD register are sent as a response to

1054/1747

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)
CMD9. Only the bits [127...1] of the CID and CSD are transferred, the reserved bit [0] of
these registers is replaced by the end bit of the response. The card indicates that an erase
is in progress by holding MCDAT low. The actual erase time may be quite long, and the host
may issue CMD7 to deselect the card.
Table 174. R2 response
Bit position

31.5.4

Width (bits

Value

Description

135

1

0

Start bit

134

1

0

Transmission bit

[133:128]

6

‘111111’

Command index

[127:1]

127

X

Card status

0

1

1

End bit

R3 (OCR register)
Code length: 48 bits. The contents of the OCR register are sent as a response to CMD1.
The level coding is as follows: restricted voltage windows = low, card busy = low.
Table 175. R3 response
Bit position

31.5.5

Width (bits

Value

Description

47

1

0

Start bit

46

1

0

Transmission bit

[45:40]

6

‘111111’

Reserved

[39:8]

32

X

OCR register

[7:1]

7

‘1111111’

Reserved

0

1

1

End bit

R4 (Fast I/O)
Code length: 48 bits. The argument field contains the RCA of the addressed card, the
register address to be read from or written to, and its content.
Table 176. R4 response
Bit position

Width (bits

Value

Description

47

1

0

Start bit

46

1

0

Transmission bit

[45:40]

6

‘100111’

CMD39

[31:16]

16

X

RCA

[15:8]

8

X

register address

[7:0]

8

X

read register contents

[39:8] Argument field

RM0090 Rev 17

1055/1747
1075

Secure digital input/output interface (SDIO)

RM0090

Table 176. R4 response (continued)
Bit position

31.5.6

Width (bits

Value

Description

[7:1]

7

X

CRC7

0

1

1

End bit

R4b
For SD I/O only: an SDIO card receiving the CMD5 will respond with a unique SDIO
response R4. The format is:
Table 177. R4b response
Bit position

Width (bits

Value

Description

47

1

0

Start bit

46

1

0

Transmission bit

[45:40]

6

x

Reserved

39

16

X

Card is ready

[38:36]

3

X

Number of I/O functions

35

1

X

Present memory

[34:32]

3

X

Stuff bits

[31:8]

24

X

I/O ORC

[7:1]

7

X

Reserved

0

1

1

End bit

[39:8] Argument field

Once an SD I/O card has received a CMD5, the I/O portion of that card is enabled to
respond normally to all further commands. This I/O enable of the function within the I/O card
will remain set until a reset, power cycle or CMD52 with write to I/O reset is received by the
card. Note that an SD memory-only card may respond to a CMD5. The proper response for
a memory-only card would be Present memory = 1 and Number of I/O functions = 0. A
memory-only card built to meet the SD Memory Card specification version 1.0 would detect
the CMD5 as an illegal command and not respond. The I/O aware host will send CMD5. If
the card responds with response R4, the host determines the card’s configuration based on
the data contained within the R4 response.

31.5.7

R5 (interrupt request)
Only for MultiMediaCard. Code length: 48 bits. If the response is generated by the host, the
RCA field in the argument will be 0x0.
Table 178. R5 response
Bit position

1056/1747

Width (bits

Value

Description

47

1

0

Start bit

46

1

0

Transmission bit

[45:40]

6

‘101000’

CMD40

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)
Table 178. R5 response (continued)
Bit position

Width (bits

Value

Description

[31:16]

16

X

RCA [31:16] of winning
card or of the host

[15:0]

16

X

Not defined. May be used
for IRQ data

[7:1]

7

X

CRC7

0

1

1

End bit

[39:8] Argument field

31.5.8

R6
Only for SD I/O. The normal response to CMD3 by a memory device. It is shown in
Table 179.
Table 179. R6 response
Bit position

Width (bits)

Value

Description

47

1

0

Start bit

46

1

0

Transmission bit

[45:40]

6

‘101000’

CMD40

[31:16]

16

X

RCA [31:16] of winning card or of the host

[15:0]

16

X

Not defined. May be used for IRQ data

[7:1]

7

X

CRC7

0

1

1

End bit

[39:8] Argument
field

The card [23:8] status bits are changed when CMD3 is sent to an I/O-only card. In this case,
the 16 bits of response are the SD I/O-only values:

31.6

•

Bit [15] COM_CRC_ERROR

•

Bit [14] ILLEGAL_COMMAND

•

Bit [13] ERROR

•

Bits [12:0] Reserved

SDIO I/O card-specific operations
The following features are SD I/O-specific operations:
•

SDIO read wait operation by SDIO_D2 signalling

•

SDIO read wait operation by stopping the clock

•

SDIO suspend/resume operation (write and read suspend)

•

SDIO interrupts

The SDIO supports these operations only if the SDIO_DCTRL[11] bit is set, except for read
suspend that does not need specific hardware implementation.

RM0090 Rev 17

1057/1747
1075

Secure digital input/output interface (SDIO)

31.6.1

RM0090

SDIO I/O read wait operation by SDIO_D2 signaling
It is possible to start the readwait interval before the first block is received: when the data
path is enabled (SDIO_DCTRL[0] bit set), the SDIO-specific operation is enabled
(SDIO_DCTRL[11] bit set), read wait starts (SDI0_DCTRL[10] =0 and SDI_DCTRL[8] =1)
and data direction is from card to SDIO (SDIO_DCTRL[1] = 1), the DPSM directly moves
from Idle to Readwait. In Readwait the DPSM drives SDIO_D2 to 0 after 2 SDIO_CK clock
cycles. In this state, when you set the RWSTOP bit (SDIO_DCTRL[9]), the DPSM remains
in Wait for two more SDIO_CK clock cycles to drive SDIO_D2 to 1 for one clock cycle (in
accordance with SDIO specification). The DPSM then starts waiting again until it receives
data from the card. The DPSM will not start a readwait interval while receiving a block even
if read wait start is set: the readwait interval will start after the CRC is received. The
RWSTOP bit has to be cleared to start a new read wait operation. During the readwait
interval, the SDIO can detect SDIO interrupts on SDIO_D1.

31.6.2

SDIO read wait operation by stopping SDIO_CK
If the SDIO card does not support the previous read wait method, the SDIO can perform a
read wait by stopping SDIO_CK (SDIO_DCTRL is set just like in the method presented in
Section 31.6.1, but SDIO_DCTRL[10] =1): DSPM stops the clock two SDIO_CK cycles after
the end bit of the current received block and starts the clock again after the read wait start bit
is set.
As SDIO_CK is stopped, any command can be issued to the card. During a read/wait
interval, the SDIO can detect SDIO interrupts on SDIO_D1.

31.6.3

SDIO suspend/resume operation
While sending data to the card, the SDIO can suspend the write operation. the
SDIO_CMD[11] bit is set and indicates to the CPSM that the current command is a suspend
command. The CPSM analyzes the response and when the ACK is received from the card
(suspend accepted), it acknowledges the DPSM that goes Idle after receiving the CRC
token of the current block.
The hardware does not save the number of the remaining block to be sent to complete the
suspended operation (resume).
The write operation can be suspended by software, just by disabling the DPSM
(SDIO_DCTRL[0] =0) when the ACK of the suspend command is received from the card.
The DPSM enters then the Idle state.
To suspend a read: the DPSM waits in the Wait_r state as the function to be suspended
sends a complete packet just before stopping the data transaction. The application
continues reading RxFIFO until the FIF0 is empty, and the DPSM goes Idle automatically.

31.6.4

SDIO interrupts
SDIO interrupts are detected on the SDIO_D1 line once the SDIO_DCTRL[11] bit is set.

1058/1747

RM0090 Rev 17

RM0090

31.7

Secure digital input/output interface (SDIO)

CE-ATA specific operations
The following features are CE-ATA specific operations:
•

sending the command completion signal disable to the CE-ATA device

•

receiving the command completion signal from the CE-ATA device

•

signaling the completion of the CE-ATA command to the CPU, using the status bit
and/or interrupt.

The SDIO supports these operations only for the CE-ATA CMD61 command, that is, if
SDIO_CMD[14] is set.

31.7.1

Command completion signal disable
Command completion signal disable is sent 8 bit cycles after the reception of a short
response if the ‘enable CMD completion’ bit, SDIO_CMD[12], is not set and the ‘not interrupt
Enable’ bit, SDIO_CMD[13], is set.
The CPSM enters the Pend state, loading the command shift register with the disable
sequence “00001” and, the command counter with 43. Eight cycles after, a trigger moves
the CPSM to the Send state. When the command counter reaches 48, the CPSM becomes
Idle as no response is awaited.

31.7.2

Command completion signal enable
If the ‘enable CMD completion’ bit SDIO_CMD[12] is set and the ‘not interrupt Enable’ bit
SDIO_CMD[13] is set, the CPSM waits for the command completion signal in the Waitcpl
state.
When ‘0’ is received on the CMD line, the CPSM enters the Idle state. No new command
can be sent for 7 bit cycles. Then, for the last 5 cycles (out of the 7) the CMD line is driven to
‘1’ in push-pull mode.

31.7.3

CE-ATA interrupt
The command completion is signaled to the CPU by the status bit SDIO_STA[23]. This static
bit can be cleared with the clear bit SDIO_ICR[23].
The SDIO_STA[23] status bit can generate an interrupt on each interrupt line, depending on
the mask bit SDIO_MASKx[23].

31.7.4

Aborting CMD61
If the command completion disable signal has not been sent and CMD61 needs to be
aborted, the command state machine must be disabled. It then becomes Idle, and the
CMD12 command can be sent. No command completion disable signal is sent during the
operation.

RM0090 Rev 17

1059/1747
1075

Secure digital input/output interface (SDIO)

31.8

RM0090

HW flow control
The HW flow control functionality is used to avoid FIFO underrun (TX mode) and overrun
(RX mode) errors.
The behavior is to stop SDIO_CK and freeze SDIO state machines. The data transfer is
stalled while the FIFO is unable to transmit or receive data. Only state machines clocked by
SDIOCLK are frozen, the APB2 interface is still alive. The FIFO can thus be filled or emptied
even if flow control is activated.
To enable HW flow control, the SDIO_CLKCR[14] register bit must be set to 1. After reset
Flow Control is disabled.

31.9

SDIO registers
The device communicates to the system via 32-bit-wide control registers accessible via
APB2.
The peripheral registers have to be accessed by words (32 bits).

31.9.1

SDIO power control register (SDIO_POWER)
Address offset: 0x00
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

Reserved

4

3

2

1

0

PWRC
TRL
rw

rw

Bits 31:2 Reserved, must be kept at reset value
Bits 1:0 PWRCTRL: Power supply control bits.
These bits are used to define the current functional state of the card clock:
00: Power-off: the clock to card is stopped.
01: Reserved
10: Reserved power-up
11: Power-on: the card is clocked.

Note:

At least seven HCLK clock periods are needed between two write accesses to this register.
After a data write, data cannot be written to this register for three SDIOCLK clock periods
plus two PCLK2 clock periods.

1060/1747

RM0090 Rev 17

RM0090

31.9.2

Secure digital input/output interface (SDIO)

SDI clock control register (SDIO_CLKCR)
Address offset: 0x04
Reset value: 0x0000 0000

rw

rw

rw

CLKEN

rw

WID
BUS

8

PWRSAV

HWFC_EN

NEGEDGE

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

Reserved

9

BYPASS

The SDIO_CLKCR register controls the SDIO_CK output clock.

rw

rw

rw

7

6

5

4

3

2

1

0

rw

rw

rw

CLKDIV

rw

rw

rw

rw

rw

Bits 31:15 Reserved, must be kept at reset value
Bit 14 HWFC_EN: HW Flow Control enable
0b: HW Flow Control is disabled
1b: HW Flow Control is enabled
When HW Flow Control is enabled, the meaning of the TXFIFOE and RXFIFOF interrupt
signals, please see SDIO Status register definition in Section 31.9.11.
Bit 13 NEGEDGE:SDIO_CK dephasing selection bit
0b: SDIO_CK generated on the rising edge of the master clock SDIOCLK
1b: SDIO_CK generated on the falling edge of the master clock SDIOCLK
Bits 12:11 WIDBUS: Wide bus mode enable bit
00: Default bus mode: SDIO_D0 used
01: 4-wide bus mode: SDIO_D[3:0] used
10: 8-wide bus mode: SDIO_D[7:0] used
Bit 10 BYPASS: Clock divider bypass enable bit
0: Disable bypass: SDIOCLK is divided according to the CLKDIV value before driving the
SDIO_CK output signal.
1: Enable bypass: SDIOCLK directly drives the SDIO_CK output signal.
Bit 9 PWRSAV: Power saving configuration bit
For power saving, the SDIO_CK clock output can be disabled when the bus is idle by setting
PWRSAV:
0: SDIO_CK clock is always enabled
1: SDIO_CK is only enabled when the bus is active
Bit 8 CLKEN: Clock enable bit
0: SDIO_CK is disabled
1: SDIO_CK is enabled
Bits 7:0 CLKDIV: Clock divide factor
This field defines the divide factor between the input clock (SDIOCLK) and the output clock
(SDIO_CK): SDIO_CK frequency = SDIOCLK / [CLKDIV + 2].

RM0090 Rev 17

1061/1747
1075

Secure digital input/output interface (SDIO)
Note:

RM0090

While the SD/SDIO card or MultiMediaCard is in identification mode, the SDIO_CK
frequency must be less than 400 kHz.
The clock frequency can be changed to the maximum card bus frequency when relative
card addresses are assigned to all cards.
After a data write, data cannot be written to this register for three SDIOCLK clock periods
plus two PCLK2 clock periods. SDIO_CK can also be stopped during the read wait interval
for SD I/O cards: in this case the SDIO_CLKCR register does not control SDIO_CK.

31.9.3

SDIO argument register (SDIO_ARG)
Address offset: 0x08
Reset value: 0x0000 0000
The SDIO_ARG register contains a 32-bit command argument, which is sent to a card as
part of a command message.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

CMDARG
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 CMDARG: Command argument
Command argument sent to a card as part of a command message. If a command contains
an argument, it must be loaded into this register before writing a command to the command
register.

31.9.4

SDIO command register (SDIO_CMD)
Address offset: 0x0C
Reset value: 0x0000 0000
The SDIO_CMD register contains the command index and command type bits. The
command index is sent to a card as part of a command message. The command type bits
control the command path state machine (CPSM).

CPSMEN

WAITPEND

WAITINT

rw

rw

rw

rw

rw

rw

Bits 31:15 Reserved, must be kept at reset value
Bit 14 ATACMD: CE-ATA command
If ATACMD is set, the CPSM transfers CMD61.
Bit 13 nIEN: not Interrupt Enable
if this bit is 0, interrupts in the CE-ATA device are enabled.
Bit 12 ENCMDcompl: Enable CMD completion
If this bit is set, the command completion signal is enabled.

1062/1747

RM0090 Rev 17

6

5

4

3

rw

rw

rw

rw

rw

2

1

0

rw

rw

rw

CMDINDEX

SDIOSuspend

rw

7
WAITRESP

ENCMDcompl

8

nIEN

Reserved

9

CE-ATACMD

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

RM0090

Secure digital input/output interface (SDIO)

Bit 11 SDIOSuspend: SD I/O suspend command
If this bit is set, the command to be sent is a suspend command (to be used only with SDIO
card).
Bit 10 CPSMEN: Command path state machine (CPSM) Enable bit
If this bit is set, the CPSM is enabled.
Bit 9 WAITPEND: CPSM Waits for ends of data transfer (CmdPend internal signal).
If this bit is set, the CPSM waits for the end of data transfer before it starts sending a
command.
Bit 8 WAITINT: CPSM waits for interrupt request
If this bit is set, the CPSM disables command timeout and waits for an interrupt request.
Bits 7:6 WAITRESP: Wait for response bits
They are used to configure whether the CPSM is to wait for a response, and if yes, which
kind of response.
00: No response, expect CMDSENT flag
01: Short response, expect CMDREND or CCRCFAIL flag
10: No response, expect CMDSENT flag
11: Long response, expect CMDREND or CCRCFAIL flag
Bits 5:0 CMDINDEX: Command index
The command index is sent to the card as part of a command message.

Note:

After a data write, data cannot be written to this register for three SDIOCLK clock periods
plus two PCLK2 clock periods.
MultiMediaCards can send two kinds of response: short responses, 48 bits long, or long
responses,136 bits long. SD card and SD I/O card can send only short responses, the
argument can vary according to the type of response: the software will distinguish the type
of response according to the sent command. CE-ATA devices send only short responses.

31.9.5

SDIO command response register (SDIO_RESPCMD)
Address offset: 0x10
Reset value: 0x0000 0000
The SDIO_RESPCMD register contains the command index field of the last command
response received. If the command response transmission does not contain the command
index field (long or OCR response), the RESPCMD field is unknown, although it must
contain 111111b (the value of the reserved field from the response).

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

7

6

5

4

3

2

1

0

r

r

RESPCMD
r

r

r

r

Bits 31:6 Reserved, must be kept at reset value
Bits 5:0 RESPCMD: Response command index
Read-only bit field. Contains the command index of the last command response received.

RM0090 Rev 17

1063/1747
1075

Secure digital input/output interface (SDIO)

31.9.6

RM0090

SDIO response 1..4 register (SDIO_RESPx)
Address offset: (0x10 + (4 × x)); x = 1..4
Reset value: 0x0000 0000
The SDIO_RESP1/2/3/4 registers contain the status of a card, which is part of the received
response.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

CARDSTATUSx
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:0 CARDSTATUSx: see Table 180.

The Card Status size is 32 or 127 bits, depending on the response type.
Table 180. Response type and SDIO_RESPx registers
Register

Short response

Long response

SDIO_RESP1

Card Status[31:0]

Card Status [127:96]

SDIO_RESP2

Unused

Card Status [95:64]

SDIO_RESP3

Unused

Card Status [63:32]

SDIO_RESP4

Unused

Card Status [31:1]0b

The most significant bit of the card status is received first. The SDIO_RESP3 register LSB is
always 0b.

31.9.7

SDIO data timer register (SDIO_DTIMER)
Address offset: 0x24
Reset value: 0x0000 0000
The SDIO_DTIMER register contains the data timeout period, in card bus clock periods.
A counter loads the value from the SDIO_DTIMER register, and starts decrementing when
the data path state machine (DPSM) enters the Wait_R or Busy state. If the timer reaches 0
while the DPSM is in either of these states, the timeout status flag is set.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

DATATIME
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 DATATIME: Data timeout period
Data timeout period expressed in card bus clock periods.

Note:

1064/1747

A data transfer must be written to the data timer register and the data length register before
being written to the data control register.

RM0090 Rev 17

RM0090

31.9.8

Secure digital input/output interface (SDIO)

SDIO data length register (SDIO_DLEN)
Address offset: 0x28
Reset value: 0x0000 0000
The SDIO_DLEN register contains the number of data bytes to be transferred. The value is
loaded into the data counter when data transfer starts.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

DATALENGTH
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:25 Reserved, must be kept at reset value
Bits 24:0 DATALENGTH: Data length value
Number of data bytes to be transferred.

Note:

For a block data transfer, the value in the data length register must be a multiple of the block
size (see SDIO_DCTRL). A data transfer must be written to the data timer register and the
data length register before being written to the data control register.
For an SDIO multibyte transfer the value in the data length register must be between 1 and
512.

RM0090 Rev 17

1065/1747
1075

Secure digital input/output interface (SDIO)

31.9.9

RM0090

SDIO data control register (SDIO_DCTRL)
Address offset: 0x2C
Reset value: 0x0000 0000

rw

1

0

DBLOCKSIZE

DTEN

rw

2

DTDIR

rw

3

DTMODE

RWSTOP

RWSTART

rw

4

DMAEN

8

SDIOEN

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

Reserved

9

RWMOD

The SDIO_DCTRL register control the data path state machine (DPSM).
7

rw

rw

rw

rw

rw

6

rw

5

rw

Bits 31:12 Reserved, must be kept at reset value
Bit 11 SDIOEN: SD I/O enable functions
If this bit is set, the DPSM performs an SD I/O-card-specific operation.
Bit 10 RWMOD: Read wait mode
0: Read Wait control stopping SDIO_D2
1: Read Wait control using SDIO_CK
Bit 9 RWSTOP: Read wait stop
0: Read wait in progress if RWSTART bit is set
1: Enable for read wait stop if RWSTART bit is set
Bit 8 RWSTART: Read wait start
If this bit is set, read wait operation starts.
Bits 7:4 DBLOCKSIZE: Data block size
Define the data block length when the block data transfer mode is selected:
0000: (0 decimal) lock length = 20 = 1 byte
0001: (1 decimal) lock length = 21 = 2 bytes
0010: (2 decimal) lock length = 22 = 4 bytes
0011: (3 decimal) lock length = 23 = 8 bytes
0100: (4 decimal) lock length = 24 = 16 bytes
0101: (5 decimal) lock length = 25 = 32 bytes
0110: (6 decimal) lock length = 26 = 64 bytes
0111: (7 decimal) lock length = 27 = 128 bytes
1000: (8 decimal) lock length = 28 = 256 bytes
1001: (9 decimal) lock length = 29 = 512 bytes
1010: (10 decimal) lock length = 210 = 1024 bytes
1011: (11 decimal) lock length = 211 = 2048 bytes
1100: (12 decimal) lock length = 212 = 4096 bytes
1101: (13 decimal) lock length = 213 = 8192 bytes
1110: (14 decimal) lock length = 214 = 16384 bytes
1111: (15 decimal) reserved
Bit 3 DMAEN: DMA enable bit
0: DMA disabled.
1: DMA enabled.

1066/1747

RM0090 Rev 17

rw

RM0090

Secure digital input/output interface (SDIO)

Bit 2 DTMODE: Data transfer mode selection 1: Stream or SDIO multibyte data transfer.
0: Block data transfer
1: Stream or SDIO multibyte data transfer
Bit 1 DTDIR: Data transfer direction selection
0: From controller to card.
1: From card to controller.
Bit 0 DTEN: Data transfer enabled bit
Data transfer starts if 1b is written to the DTEN bit. Depending on the direction bit, DTDIR,
the DPSM moves to the Wait_S, Wait_R state or Readwait if RW Start is set immediately at
the beginning of the transfer. It is not necessary to clear the enable bit after the end of a data
transfer but the SDIO_DCTRL must be updated to enable a new data transfer

Note:

After a data write, data cannot be written to this register for three SDIOCLK clock periods
plus two PCLK2 clock periods.
The meaning of the DTMODE bit changes according to the value of the SDIOEN bit. When
SDIOEN=0 and DTMODE=1, the MultiMediaCard stream mode is enabled, and when
SDIOEN=1 and DTMODE=1, the peripheral enables an SDIO multibyte transfer.

31.9.10

SDIO data counter register (SDIO_DCOUNT)
Address offset: 0x30
Reset value: 0x0000 0000
The SDIO_DCOUNT register loads the value from the data length register (see
SDIO_DLEN) when the DPSM moves from the Idle state to the Wait_R or Wait_S state. As
data is transferred, the counter decrements the value until it reaches 0. The DPSM then
moves to the Idle state and the data status end flag, DATAEND, is set.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

DATACOUNT
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:25 Reserved, must be kept at reset value
Bits 24:0 DATACOUNT: Data count value
When this bit is read, the number of remaining data bytes to be transferred is returned. Write
has no effect.

Note:

This register should be read only when the data transfer is complete.

RM0090 Rev 17

1067/1747
1075

Secure digital input/output interface (SDIO)

31.9.11

RM0090

SDIO status register (SDIO_STA)
Address offset: 0x34
Reset value: 0x0000 0000
The SDIO_STA register is a read-only register. It contains two types of flag:

r

r

3

2

1

0
CCRCFAIL

r

4

DCRCFAIL

r

5

CTIMEOUT

r

6

DTIMEOUT

r

7

TXUNDERR

r

8

RXOVERR

CMDACT

RXFIFOF
r

DBCKEND

TXFIFOE
r

TXACT

RXFIFOE
r

RXACT

TXDAVL
r

TXFIFOHE

RXDAVL
r

TXFIFOF

SDIOIT
r

RXFIFOHF

CEATAEND

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

r

9

CMDREND

Dynamic flags (bits [21:11]): these bits change state depending on the state of the
underlying logic (for example, FIFO full and empty flags are asserted and deasserted
as data while written to the FIFO)

CMDSENT

•

DATAEND

Static flags (bits [23:22,10:0]): these bits remain asserted until they are cleared by
writing to the SDIO Interrupt Clear register (see SDIO_ICR)

STBITERR

•

r

r

r

r

r

r

r

r

r

r

Bits 31:24 Reserved, must be kept at reset value
Bit 23 CEATAEND: CE-ATA command completion signal received for CMD61
Bit 22 SDIOIT: SDIO interrupt received
Bit 21 RXDAVL: Data available in receive FIFO
Bit 20 TXDAVL: Data available in transmit FIFO
Bit 19 RXFIFOE: Receive FIFO empty
Bit 18 TXFIFOE: Transmit FIFO empty
When HW Flow Control is enabled, TXFIFOE signals becomes activated when the FIFO
contains 2 words.
Bit 17 RXFIFOF: Receive FIFO full
When HW Flow Control is enabled, RXFIFOF signals becomes activated 2 words before the
FIFO is full.
Bit 16 TXFIFOF: Transmit FIFO full
Bit 15 RXFIFOHF: Receive FIFO half full: there are at least 8 words in the FIFO
Bit 14 TXFIFOHE: Transmit FIFO half empty: at least 8 words can be written into the FIFO
Bit 13 RXACT: Data receive in progress
Bit 12 TXACT: Data transmit in progress
Bit 11 CMDACT: Command transfer in progress
Bit 10 DBCKEND: Data block sent/received (CRC check passed)
Bit 9 STBITERR: Start bit not detected on all data signals in wide bus mode
Bit 8 DATAEND: Data end (data counter, SDIDCOUNT, is zero)
Bit 7 CMDSENT: Command sent (no response required)
Bit 6 CMDREND: Command response received (CRC check passed)
Bit 5 RXOVERR: Received FIFO overrun error

1068/1747

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)

Bit 4 TXUNDERR: Transmit FIFO underrun error
Bit 3 DTIMEOUT: Data timeout
Bit 2 CTIMEOUT: Command response timeout
The Command TimeOut period has a fixed value of 64 SDIO_CK clock periods.
Bit 1 DCRCFAIL: Data block sent/received (CRC check failed)
Bit 0 CCRCFAIL: Command response received (CRC check failed)

31.9.12

SDIO interrupt clear register (SDIO_ICR)
Address offset: 0x38
Reset value: 0x0000 0000

6

5

4

3

2

1

0

CMDSENTC

CMDRENDC

RXOVERRC

TXUNDERRC

DTIMEOUTC

CTIMEOUTC

DCRCFAILC

CCRCFAILC

rw

7

DATAENDC

SDIOITC

rw

Reserved

8

STBITERRC

CEATAENDC

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

Reserved

9

DBCKENDC

The SDIO_ICR register is a write-only register. Writing a bit with 1b clears the corresponding
bit in the SDIO_STA Status register.

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:24 Reserved, must be kept at reset value
Bit 23 CEATAENDC: CEATAEND flag clear bit
Set by software to clear the CEATAEND flag.
0: CEATAEND not cleared
1: CEATAEND cleared
Bit 22 SDIOITC: SDIOIT flag clear bit
Set by software to clear the SDIOIT flag.
0: SDIOIT not cleared
1: SDIOIT cleared
Bits 21:11 Reserved, must be kept at reset value
Bit 10 DBCKENDC: DBCKEND flag clear bit
Set by software to clear the DBCKEND flag.
0: DBCKEND not cleared
1: DBCKEND cleared
Bit 9 STBITERRC: STBITERR flag clear bit
Set by software to clear the STBITERR flag.
0: STBITERR not cleared
1: STBITERR cleared
Bit 8 DATAENDC: DATAEND flag clear bit
Set by software to clear the DATAEND flag.
0: DATAEND not cleared
1: DATAEND cleared

RM0090 Rev 17

1069/1747
1075

Secure digital input/output interface (SDIO)

Bit 7 CMDSENTC: CMDSENT flag clear bit
Set by software to clear the CMDSENT flag.
0: CMDSENT not cleared
1: CMDSENT cleared
Bit 6 CMDRENDC: CMDREND flag clear bit
Set by software to clear the CMDREND flag.
0: CMDREND not cleared
1: CMDREND cleared
Bit 5 RXOVERRC: RXOVERR flag clear bit
Set by software to clear the RXOVERR flag.
0: RXOVERR not cleared
1: RXOVERR cleared
Bit 4 TXUNDERRC: TXUNDERR flag clear bit
Set by software to clear TXUNDERR flag.
0: TXUNDERR not cleared
1: TXUNDERR cleared
Bit 3 DTIMEOUTC: DTIMEOUT flag clear bit
Set by software to clear the DTIMEOUT flag.
0: DTIMEOUT not cleared
1: DTIMEOUT cleared
Bit 2 CTIMEOUTC: CTIMEOUT flag clear bit
Set by software to clear the CTIMEOUT flag.
0: CTIMEOUT not cleared
1: CTIMEOUT cleared
Bit 1 DCRCFAILC: DCRCFAIL flag clear bit
Set by software to clear the DCRCFAIL flag.
0: DCRCFAIL not cleared
1: DCRCFAIL cleared
Bit 0 CCRCFAILC: CCRCFAIL flag clear bit
Set by software to clear the CCRCFAIL flag.
0: CCRCFAIL not cleared
1: CCRCFAIL cleared

1070/1747

RM0090 Rev 17

RM0090

RM0090

31.9.13

Secure digital input/output interface (SDIO)

SDIO mask register (SDIO_MASK)
Address offset: 0x3C
Reset value: 0x0000 0000

2

1

0

DCRCFAILIE

CCRCFAILIE

rw

3

CTIMEOUTIE

rw

4

DTIMEOUTIE

rw

5

TXUNDERRIE

rw

6

RXOVERRIE

rw

7

CMDRENDIE

rw

8

CMDSENTIE

rw

9

DATAENDIE

rw

CMDACTIE

rw

DBCKENDIE

RXFIFOFIE

rw

TXACTIE

TXFIFOEIE

rw

RXACTIE

RXFIFOEIE

rw

TXFIFOHEIE

TXDAVLIE

rw

TXFIFOFIE

RXDAVLIE

rw

RXFIFOHFIE

SDIOITIE

Reserved

CEATAENDIE

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

STBITERRIE

The interrupt mask register determines which status flags generate an interrupt request by
setting the corresponding bit to 1b.

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:24 Reserved, must be kept at reset value
Bit 23 CEATAENDIE: CE-ATA command completion signal received interrupt enable
Set and cleared by software to enable/disable the interrupt generated when receiving the
CE-ATA command completion signal.
0: CE-ATA command completion signal received interrupt disabled
1: CE-ATA command completion signal received interrupt enabled
Bit 22 SDIOITIE: SDIO mode interrupt received interrupt enable
Set and cleared by software to enable/disable the interrupt generated when receiving the
SDIO mode interrupt.
0: SDIO Mode Interrupt Received interrupt disabled
1: SDIO Mode Interrupt Received interrupt enabled
Bit 21 RXDAVLIE: Data available in Rx FIFO interrupt enable
Set and cleared by software to enable/disable the interrupt generated by the presence of
data available in Rx FIFO.
0: Data available in Rx FIFO interrupt disabled
1: Data available in Rx FIFO interrupt enabled
Bit 20 TXDAVLIE: Data available in Tx FIFO interrupt enable
Set and cleared by software to enable/disable the interrupt generated by the presence of
data available in Tx FIFO.
0: Data available in Tx FIFO interrupt disabled
1: Data available in Tx FIFO interrupt enabled
Bit 19 RXFIFOEIE: Rx FIFO empty interrupt enable
Set and cleared by software to enable/disable interrupt caused by Rx FIFO empty.
0: Rx FIFO empty interrupt disabled
1: Rx FIFO empty interrupt enabled
Bit 18 TXFIFOEIE: Tx FIFO empty interrupt enable
Set and cleared by software to enable/disable interrupt caused by Tx FIFO empty.
0: Tx FIFO empty interrupt disabled
1: Tx FIFO empty interrupt enabled
Bit 17 RXFIFOFIE: Rx FIFO full interrupt enable
Set and cleared by software to enable/disable interrupt caused by Rx FIFO full.
0: Rx FIFO full interrupt disabled
1: Rx FIFO full interrupt enabled

RM0090 Rev 17

1071/1747
1075

Secure digital input/output interface (SDIO)

RM0090

Bit 16 TXFIFOFIE: Tx FIFO full interrupt enable
Set and cleared by software to enable/disable interrupt caused by Tx FIFO full.
0: Tx FIFO full interrupt disabled
1: Tx FIFO full interrupt enabled
Bit 15 RXFIFOHFIE: Rx FIFO half full interrupt enable
Set and cleared by software to enable/disable interrupt caused by Rx FIFO half full.
0: Rx FIFO half full interrupt disabled
1: Rx FIFO half full interrupt enabled
Bit 14 TXFIFOHEIE: Tx FIFO half empty interrupt enable
Set and cleared by software to enable/disable interrupt caused by Tx FIFO half empty.
0: Tx FIFO half empty interrupt disabled
1: Tx FIFO half empty interrupt enabled
Bit 13 RXACTIE: Data receive acting interrupt enable
Set and cleared by software to enable/disable interrupt caused by data being received (data
receive acting).
0: Data receive acting interrupt disabled
1: Data receive acting interrupt enabled
Bit 12 TXACTIE: Data transmit acting interrupt enable
Set and cleared by software to enable/disable interrupt caused by data being transferred
(data transmit acting).
0: Data transmit acting interrupt disabled
1: Data transmit acting interrupt enabled
Bit 11 CMDACTIE: Command acting interrupt enable
Set and cleared by software to enable/disable interrupt caused by a command being
transferred (command acting).
0: Command acting interrupt disabled
1: Command acting interrupt enabled
Bit 10 DBCKENDIE: Data block end interrupt enable
Set and cleared by software to enable/disable interrupt caused by data block end.
0: Data block end interrupt disabled
1: Data block end interrupt enabled
Bit 9 STBITERRIE: Start bit error interrupt enable
Set and cleared by software to enable/disable interrupt caused by start bit error.
0: Start bit error interrupt disabled
1: Start bit error interrupt enabled
Bit 8 DATAENDIE: Data end interrupt enable
Set and cleared by software to enable/disable interrupt caused by data end.
0: Data end interrupt disabled
1: Data end interrupt enabled
Bit 7 CMDSENTIE: Command sent interrupt enable
Set and cleared by software to enable/disable interrupt caused by sending command.
0: Command sent interrupt disabled
1: Command sent interrupt enabled

1072/1747

RM0090 Rev 17

RM0090

Secure digital input/output interface (SDIO)

Bit 6 CMDRENDIE: Command response received interrupt enable
Set and cleared by software to enable/disable interrupt caused by receiving command
response.
0: Command response received interrupt disabled
1: command Response Received interrupt enabled
Bit 5 RXOVERRIE: Rx FIFO overrun error interrupt enable
Set and cleared by software to enable/disable interrupt caused by Rx FIFO overrun error.
0: Rx FIFO overrun error interrupt disabled
1: Rx FIFO overrun error interrupt enabled
Bit 4 TXUNDERRIE: Tx FIFO underrun error interrupt enable
Set and cleared by software to enable/disable interrupt caused by Tx FIFO underrun error.
0: Tx FIFO underrun error interrupt disabled
1: Tx FIFO underrun error interrupt enabled
Bit 3 DTIMEOUTIE: Data timeout interrupt enable
Set and cleared by software to enable/disable interrupt caused by data timeout.
0: Data timeout interrupt disabled
1: Data timeout interrupt enabled
Bit 2 CTIMEOUTIE: Command timeout interrupt enable
Set and cleared by software to enable/disable interrupt caused by command timeout.
0: Command timeout interrupt disabled
1: Command timeout interrupt enabled
Bit 1 DCRCFAILIE: Data CRC fail interrupt enable
Set and cleared by software to enable/disable interrupt caused by data CRC failure.
0: Data CRC fail interrupt disabled
1: Data CRC fail interrupt enabled
Bit 0 CCRCFAILIE: Command CRC fail interrupt enable
Set and cleared by software to enable/disable interrupt caused by command CRC failure.
0: Command CRC fail interrupt disabled
1: Command CRC fail interrupt enabled

31.9.14

SDIO FIFO counter register (SDIO_FIFOCNT)
Address offset: 0x48
Reset value: 0x0000 0000
The SDIO_FIFOCNT register contains the remaining number of words to be written to or
read from the FIFO. The FIFO counter loads the value from the data length register (see
SDIO_DLEN) when the data transfer enable bit, DTEN, is set in the data control register
(SDIO_DCTRL register) and the DPSM is at the Idle state. If the data length is not wordaligned (multiple of 4), the remaining 1 to 3 bytes are regarded as a word.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

FIFOCOUNT
r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:24 Reserved, must be kept at reset value
Bits 23:0

FIFOCOUNT: Remaining number of words to be written to or read from the FIFO.

RM0090 Rev 17

1073/1747
1075

Secure digital input/output interface (SDIO)

31.9.15

RM0090

SDIO data FIFO register (SDIO_FIFO)
Address offset: 0x80
Reset value: 0x0000 0000
The receive and transmit FIFOs can be read or written as 32-bit wide registers. The FIFOs
contain 32 entries on 32 sequential addresses. This allows the CPU to use its load and store
multiple operands to read from/write to the FIFO.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

FIF0Data
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

bits 31:0 FIFOData: Receive and transmit FIFO data
The FIFO data occupies 32 entries of 32-bit words, from address:
SDIO base + 0x080 to SDIO base + 0xFC.

31.9.16

SDIO register map
The following table summarizes the SDIO registers.
Table 181. SDIO register map
Register

0x00

SDIO_POWER

0x04

SDIO_CLKCR

0x08

SDIO_ARG

0x0C

SDIO_CMD

0x10

SDIO_RESPCMD

0x14

SDIO_RESP1

CARDSTATUS1

CLKEN
WAITINT

CLKDIV

BYPASS

PWRSAV

CPSMEN

WAITPEND

CMDINDEX

WAITRESP

ENCMDcompl

SDIOSuspend

nIEN

Reserved

CE-ATACMD

CMDARG

Reserved

RESPCMD

0x18

SDIO_RESP2

CARDSTATUS2

0x1C

SDIO_RESP3

CARDSTATUS3

0x20

SDIO_RESP4

CARDSTATUS4

0x24

SDIO_DTIMER

0x28

SDIO_DLEN

0x2C

SDIO_DCTRL

0x30

SDIO_DCOUNT

1074/1747

WIDBUS

HWFC_EN

Reserved

NEGEDGE

Reserved

PWRCTRL

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Offset

DATATIME

Reserved

RM0090 Rev 17

DTEN

DTDIR

DTMODE

DMAEN

DBLOCKSIZE

RWSTART

RWMOD

DATACOUNT

RWSTOP

SDIOEN

DATALENGTH
Reserved

Reserved

SDIO_STA

0x38
SDIO_ICR
Reserved

CEATAENDC
SDIOITC

0x3C
SDIO_MASK
Reserved

CEATAENDIE
SDIOITIE

0x48
SDIO_FIFOCNT
Reserved

0x80
SDIO_FIFO

RM0090 Rev 17

CMDREND
RXOVERR

CMDRENDC
RXOVERRC

CMDRENDIE
RXOVERRIE

DTIMEOUT
CTIMEOUT
DCRCFAIL
CCRCFAIL

DTIMEOUTC
CTIMEOUTC
DCRCFAILC
CCRCFAILC

DTIMEOUTIE
CTIMEOUTIE
DCRCFAILIE
CCRCFAILIE

TXUNDERRIE TXUNDERRC TXUNDERR

DATAEND
CMDSENT

DATAENDC
CMDSENTC

DATAENDIE
CMDSENTIE

DBCKEND
STBITERR

DBCKENDC

CMDACTIE
STBITERRC

TXACT
CMDACT

TXACTIE
DBCKENDIE

RXACT

RXACTIE

STBITERRIE

TXFIFOHE

TXFIFOHEIE

TXFIFOF
RXFIFOHF

RXFIFOHFIE

TXFIFOFIE

TXFIFOE
RXFIFOF

TXFIFOEIE
RXFIFOFIE

TXDAVL
RXFIFOE

TXDAVLIE

RXDAVL

SDIOIT

CEATAEND

RXFIFOEIE

Reserved

0x34

RXDAVLIE

Register

Reserved

Offset
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

RM0090
Secure digital input/output interface (SDIO)

Table 181. SDIO register map (continued)

FIF0Data
FIFOCOUNT

1075/1747

1075

Controller area network (bxCAN)

32

RM0090

Controller area network (bxCAN)
This section applies to the whole STM32F4xx family, unless otherwise specified.

32.1

bxCAN introduction
The Basic Extended CAN peripheral, named bxCAN, interfaces the CAN network. It
supports the CAN protocols version 2.0A and B. It has been designed to manage a high
number of incoming messages efficiently with a minimum CPU load. It also meets the
priority requirements for transmit messages.
For safety-critical applications, the CAN controller provides all hardware functions for
supporting the CAN Time Triggered Communication option.

32.2

bxCAN main features
•

Supports CAN protocol version 2.0 A, B Active

•

Bit rates up to 1 Mbit/s

•

Supports the Time Triggered Communication option

Transmission
•

Three transmit mailboxes

•

Configurable transmit priority

•

Time Stamp on SOF transmission

Reception
•

Two receive FIFOs with three stages

•

Scalable filter banks:
–

28 filter banks shared between CAN1 and CAN2

•

Identifier list feature

•

Configurable FIFO overrun

•

Time Stamp on SOF reception

Time-triggered communication option
•

Disable automatic retransmission mode

•

16-bit free running timer

•

Time Stamp sent in last two data bytes

Management

1076/1747

•

Maskable interrupts

•

Software-efficient mailbox mapping at a unique address space

RM0090 Rev 17

RM0090

Controller area network (bxCAN)

Dual CAN

32.3

•

CAN1: Master bxCAN for managing the communication between a Slave bxCAN and
the 512-byte SRAM memory

•

CAN2: Slave bxCAN, with no direct access to the SRAM memory.

•

The two bxCAN cells share the 512-byte SRAM memory (see Figure 335: Dual CAN
block diagram)

bxCAN general description
In today’s CAN applications, the number of nodes in a network is increasing and often
several networks are linked together via gateways. Typically the number of messages in the
system (and thus to be handled by each node) has significantly increased. In addition to the
application messages, Network Management and Diagnostic messages have been
introduced.
•

An enhanced filtering mechanism is required to handle each type of message.

Furthermore, application tasks require more CPU time, therefore real-time constraints
caused by message reception have to be reduced.
•

A receive FIFO scheme allows the CPU to be dedicated to application tasks for a long
time period without losing messages.

The standard HLP (Higher Layer Protocol) based on standard CAN drivers requires an
efficient interface to the CAN controller.

MCU
Application

CAN
Controller
CAN
Rx

CAN node n

CAN node 2

CAN node 1

Figure 334. CAN network topology

CAN
Tx

CAN
Transceiver
CAN
High

CAN
Low

CAN Bus

32.3.1

CAN 2.0B active core
The bxCAN module handles the transmission and the reception of CAN messages fully
autonomously. Standard identifiers (11-bit) and extended identifiers (29-bit) are fully
supported by hardware.

RM0090 Rev 17

1077/1747
1120

Controller area network (bxCAN)

32.3.2

RM0090

Control, status and configuration registers
The application uses these registers to:

32.3.3

•

Configure CAN parameters, e.g. baud rate

•

Request transmissions

•

Handle receptions

•

Manage interrupts

•

Get diagnostic information

Tx mailboxes
Three transmit mailboxes are provided to the software for setting up messages. The
transmission Scheduler decides which mailbox has to be transmitted first.

32.3.4

Acceptance filters
The bxCAN provides 28 scalable/configurable identifier filter banks for selecting the
incoming messages the software needs and discarding the others.

Receive FIFO
Two receive FIFOs are used by hardware to store the incoming messages. Three complete
messages can be stored in each FIFO. The FIFOs are managed completely by hardware.

1078/1747

RM0090 Rev 17

RM0090

Controller area network (bxCAN)
Figure 335. Dual CAN block diagram
&$1 0DVWHU ZLWKE\WHV65$0
0DVWHU
7[0DLOER[HV

0DLOER[

0DVWHU
5HFHLYH),)2


0DLOER[

0DVWHU&RQWURO

0DVWHU
5HFHLYH),)2





0DLOER[



0DVWHU6WDWXV
7[6WDWXV

&RQWURO6WDWXV&RQILJXUDWLRQ

5[),)26WDWXV

7UDQVPLVVLRQ
6FKHGXOHU

5[),)26WDWXV
$FFHSWDQFH)LOWHUV

,QWHUUXSW(QDEOH
&$1%$FWLYH&RUH

(UURU6WDWXV

0HPRU\
$FFHVV
&RQWUROOHU

%LW7LPLQJ

)LOWHU





0DVWHU)LOWHUV
WR

)LOWHU0DVWHU









 

6ODYH)LOWHUV
WR

)LOWHU0RGH
7UDQVPLVVLRQ
6FKHGXOHU

)LOWHU6FDOH

6ODYH
5HFHLYH),)2

6ODYH
7[0DLOER[HV

)LOWHU),)2$VVLJQ



)LOWHU$FWLYDWLRQ
0DLOER[

6ODYH
5HFHLYH),)2




0DLOER[




0DLOER[



&RQWURO6WDWXV&RQILJXUDWLRQ

&$1 6ODYH
0DVWHU&RQWURO
0DVWHU6WDWXV
7[6WDWXV
5[),)26WDWXV
5[),)26WDWXV

&$1%$FWLYH&RUH

,QWHUUXSW(QDEOH
(UURU6WDWXV
1RWH &$1VWDUWILOWHUEDQNQXPEHUQLVFRQIL JXUDEOHE\ZULWLQJWR
WKH&$16%>@ELWVLQWKH&$1B)05UHJLVWHU

%LW7LPLQJ

DLE

32.4

bxCAN operating modes
bxCAN has three main operating modes: initialization, normal and Sleep. After a
hardware reset, bxCAN is in Sleep mode to reduce power consumption and an internal pullup is active on CANTX. The software requests bxCAN to enter initialization or Sleep mode
by setting the INRQ or SLEEP bits in the CAN_MCR register. Once the mode has been
entered, bxCAN confirms it by setting the INAK or SLAK bits in the CAN_MSR register and
the internal pull-up is disabled. When neither INAK nor SLAK are set, bxCAN is in normal

RM0090 Rev 17

1079/1747
1120

Controller area network (bxCAN)

RM0090

mode. Before entering normal mode bxCAN always has to synchronize on the CAN bus.
To synchronize, bxCAN waits until the CAN bus is idle, this means 11 consecutive recessive
bits have been monitored on CANRX.

32.4.1

Initialization mode
The software initialization can be done while the hardware is in Initialization mode. To enter
this mode the software sets the INRQ bit in the CAN_MCR register and waits until the
hardware has confirmed the request by setting the INAK bit in the CAN_MSR register.
To leave Initialization mode, the software clears the INQR bit. bxCAN has left Initialization
mode once the INAK bit has been cleared by hardware.
While in Initialization Mode, all message transfers to and from the CAN bus are stopped and
the status of the CAN bus output CANTX is recessive (high).
Entering Initialization Mode does not change any of the configuration registers.
To initialize the CAN Controller, software has to set up the Bit Timing (CAN_BTR) and CAN
options (CAN_MCR) registers.
To initialize the registers associated with the CAN filter banks (mode, scale, FIFO
assignment, activation and filter values), software has to set the FINIT bit (CAN_FMR). Filter
initialization also can be done outside the initialization mode.

Note:

When FINIT=1, CAN reception is deactivated.
The filter values also can be modified by deactivating the associated filter activation bits (in
the CAN_FA1R register).
If a filter bank is not used, it is recommended to leave it non active (leave the corresponding
FACT bit cleared).

32.4.2

Normal mode
Once the initialization is complete, the software must request the hardware to enter Normal
mode to be able to synchronize on the CAN bus and start reception and transmission.
The request to enter Normal mode is issued by clearing the INRQ bit in the CAN_MCR
register. The bxCAN enters Normal mode and is ready to take part in bus activities when it
has synchronized with the data transfer on the CAN bus. This is done by waiting for the
occurrence of a sequence of 11 consecutive recessive bits (Bus Idle state). The switch to
Normal mode is confirmed by the hardware by clearing the INAK bit in the CAN_MSR
register.
The initialization of the filter values is independent from Initialization Mode but must be done
while the filter is not active (corresponding FACTx bit cleared). The filter scale and mode
configuration must be configured before entering Normal Mode.

32.4.3

Sleep mode (low power)
To reduce power consumption, bxCAN has a low-power mode called Sleep mode. This
mode is entered on software request by setting the SLEEP bit in the CAN_MCR register. In
this mode, the bxCAN clock is stopped, however software can still access the bxCAN
mailboxes.
If software requests entry to initialization mode by setting the INRQ bit while bxCAN is in
Sleep mode, it must also clear the SLEEP bit.

1080/1747

RM0090 Rev 17

RM0090

Controller area network (bxCAN)
bxCAN can be woken up (exit Sleep mode) either by software clearing the SLEEP bit or on
detection of CAN bus activity.
On CAN bus activity detection, hardware automatically performs the wakeup sequence by
clearing the SLEEP bit if the AWUM bit in the CAN_MCR register is set. If the AWUM bit is
cleared, software has to clear the SLEEP bit when a wakeup interrupt occurs, in order to exit
from Sleep mode.

Note:

If the wakeup interrupt is enabled (WKUIE bit set in CAN_IER register) a wakeup interrupt
will be generated on detection of CAN bus activity, even if the bxCAN automatically
performs the wakeup sequence.
After the SLEEP bit has been cleared, Sleep mode is exited once bxCAN has synchronized
with the CAN bus, refer to Figure 336: bxCAN operating modes. The Sleep mode is exited
once the SLAK bit has been cleared by hardware.
Figure 336. bxCAN operating modes
2ESET

3LEEP

%
3,

.ORMAL

%0

3

#
9.

%
3,

1
.2
)

%0

!

3,!+ 
).!+  

3,
%%
0

3,

#+

%%

0

).21  !#+

)

.2

1

)
.2
1

!

!

#+

#+

)NITIALIZATION
3,!+ 
).!+  

3,!+ 
).!+  

).21  39.#  3,%%0
AI

1. ACK = The wait state during which hardware confirms a request by setting the INAK or SLAK bits in the
CAN_MSR register
2. SYNC = The state during which bxCAN waits until the CAN bus is idle, meaning 11 consecutive recessive
bits have been monitored on CANRX

32.5

Test mode
Test mode can be selected by the SILM and LBKM bits in the CAN_BTR register. These bits
must be configured while bxCAN is in Initialization mode. Once test mode has been
selected, the INRQ bit in the CAN_MCR register must be reset to enter Normal mode.

32.5.1

Silent mode
The bxCAN can be put in Silent mode by setting the SILM bit in the CAN_BTR register.
In Silent mode, the bxCAN is able to receive valid data frames and valid remote frames, but
it sends only recessive bits on the CAN bus and it cannot start a transmission. If the bxCAN
has to send a dominant bit (ACK bit, overload flag, active error flag), the bit is rerouted
internally so that the CAN Core monitors this dominant bit, although the CAN bus may

RM0090 Rev 17

1081/1747
1120

Controller area network (bxCAN)

RM0090

remain in recessive state. Silent mode can be used to analyze the traffic on a CAN bus
without affecting it by the transmission of dominant bits (Acknowledge Bits, Error Frames).
Figure 337. bxCAN in silent mode
bxCAN
Tx

Rx

=1

CANTX CANRX

32.5.2

Loop back mode
The bxCAN can be set in Loop Back Mode by setting the LBKM bit in the CAN_BTR
register. In Loop Back Mode, the bxCAN treats its own transmitted messages as received
messages and stores them (if they pass acceptance filtering) in a Receive mailbox.
Figure 338. bxCAN in loop back mode
bxCAN
Tx

Rx

CANTX CANRX

This mode is provided for self-test functions. To be independent of external events, the CAN
Core ignores acknowledge errors (no dominant bit sampled in the acknowledge slot of a
data / remote frame) in Loop Back Mode. In this mode, the bxCAN performs an internal
feedback from its Tx output to its Rx input. The actual value of the CANRX input pin is
disregarded by the bxCAN. The transmitted messages can be monitored on the CANTX pin.

32.5.3

Loop back combined with silent mode
It is also possible to combine Loop Back mode and Silent mode by setting the LBKM and
SILM bits in the CAN_BTR register. This mode can be used for a “Hot Selftest”, meaning the
bxCAN can be tested like in Loop Back mode but without affecting a running CAN system
connected to the CANTX and CANRX pins. In this mode, the CANRX pin is disconnected
from the bxCAN and the CANTX pin is held recessive.

1082/1747

RM0090 Rev 17

RM0090

Controller area network (bxCAN)
Figure 339. bxCAN in combined mode
bxCAN
Tx

Rx

=1

CANTX CANRX

32.6

Debug mode
When the microcontroller enters the debug mode (Cortex®-M4 with FPU core halted), the
bxCAN continues to work normally or stops, depending on:
•

the DBG_CAN1_STOP bit for CAN1 or the DBG_CAN2_STOP bit for CAN2 in the
DBG module. For more details, refer to Section 38.16.2: Debug support for timers,
watchdog, bxCAN and I2C.

•

the DBF bit in CAN_MCR. For more details, refer to Section 32.9.2: CAN control and
status registers.

32.7

bxCAN functional description

32.7.1

Transmission handling
In order to transmit a message, the application must select one empty transmit mailbox, set
up the identifier, the data length code (DLC) and the data before requesting the transmission
by setting the corresponding TXRQ bit in the CAN_TIxR register. Once the mailbox has left
empty state, the software no longer has write access to the mailbox registers. Immediately
after the TXRQ bit has been set, the mailbox enters pending state and waits to become the
highest priority mailbox, see Transmit Priority. As soon as the mailbox has the highest
priority it will be scheduled for transmission. The transmission of the message of the
scheduled mailbox will start (enter transmit state) when the CAN bus becomes idle. Once
the mailbox has been successfully transmitted, it will become empty again. The hardware
indicates a successful transmission by setting the RQCP and TXOK bits in the CAN_TSR
register.
If the transmission fails, the cause is indicated by the ALST bit in the CAN_TSR register in
case of an Arbitration Lost, and/or the TERR bit, in case of transmission error detection.

Transmit priority
By identifier
When more than one transmit mailbox is pending, the transmission order is given by the
identifier of the message stored in the mailbox. The message with the lowest identifier value
has the highest priority according to the arbitration of the CAN protocol. If the identifier
values are equal, the lower mailbox number will be scheduled first.
By transmit request order

RM0090 Rev 17

1083/1747
1120

Controller area network (bxCAN)

RM0090

The transmit mailboxes can be configured as a transmit FIFO by setting the TXFP bit in the
CAN_MCR register. In this mode the priority order is given by the transmit request order.
This mode is very useful for segmented transmission.

Abort
A transmission request can be aborted by the user setting the ABRQ bit in the CAN_TSR
register. In pending or scheduled state, the mailbox is aborted immediately. An abort
request while the mailbox is in transmit state can have two results. If the mailbox is
transmitted successfully the mailbox becomes empty with the TXOK bit set in the
CAN_TSR register. If the transmission fails, the mailbox becomes scheduled, the
transmission is aborted and becomes empty with TXOK cleared. In all cases the mailbox
will become empty again at least at the end of the current transmission.

Nonautomatic retransmission mode
This mode has been implemented in order to fulfil the requirement of the Time Triggered
Communication option of the CAN standard. To configure the hardware in this mode the
NART bit in the CAN_MCR register must be set.
In this mode, each transmission is started only once. If the first attempt fails, due to an
arbitration loss or an error, the hardware will not automatically restart the message
transmission.
At the end of the first transmission attempt, the hardware considers the request as
completed and sets the RQCP bit in the CAN_TSR register. The result of the transmission is
indicated in the CAN_TSR register by the TXOK, ALST and TERR bits.
Figure 340. Transmit mailbox states
EMPTY
RQCP=X
TXOK=X
TME = 1

TXRQ=1

PENDING
ABRQ=1

RQCP=0
TXOK=0
TME = 0

Mailbox does not
have highest priority

EMPTY

ABRQ=1

RQCP=1
TXOK=0
TME = 1

CAN Bus = IDLE

Transmit failed * NART

TRANSMIT
RQCP=0
TXOK=0
TME = 0

EMPTY
RQCP=1
TXOK=1
TME = 1

1084/1747

Mailbox has
highest priority

Transmit succeeded

RM0090 Rev 17

SCHEDULED
RQCP=0
TXOK=0
TME = 0

Transmit failed * NART

RM0090

32.7.2

Controller area network (bxCAN)

Time triggered communication mode
In this mode, the internal counter of the CAN hardware is activated and used to generate the
Time Stamp value stored in the CAN_RDTxR/CAN_TDTxR registers, respectively (for Rx
and Tx mailboxes). The internal counter is incremented each CAN bit time (refer to
Section 32.7.7: Bit timing). The internal counter is captured on the sample point of the Start
Of Frame bit in both reception and transmission.

32.7.3

Reception handling
For the reception of CAN messages, three mailboxes organized as a FIFO are provided. In
order to save CPU load, simplify the software and guarantee data consistency, the FIFO is
managed completely by hardware. The application accesses the messages stored in the
FIFO through the FIFO output mailbox.

Valid message
A received message is considered as valid when it has been received correctly according to
the CAN protocol (no error until the last but one bit of the EOF field) and It passed through
the identifier filtering successfully, see Section 32.7.4: Identifier filtering.
Figure 341. Receive FIFO states
EMPTY
FMP=0x00
FOVR=0

Valid Message
Received

Release
Mailbox

PENDING_1
FMP=0x01
FOVR=0

Release
Mailbox
RFOM=1

Valid Message
Received

PENDING_2
FMP=0x10
FOVR=0

Release
Mailbox
RFOM=1

Valid Message
Received

PENDING_3
FMP=0x11
FOVR=0

Valid Message
Received

Release
Mailbox
RFOM=1

OVERRUN
FMP=0x11
FOVR=1

Valid Message
Received

RM0090 Rev 17

1085/1747
1120

Controller area network (bxCAN)

RM0090

FIFO management
Starting from the empty state, the first valid message received is stored in the FIFO which
becomes pending_1. The hardware signals the event setting the FMP[1:0] bits in the
CAN_RFR register to the value 01b. The message is available in the FIFO output mailbox.
The software reads out the mailbox content and releases it by setting the RFOM bit in the
CAN_RFR register. The FIFO becomes empty again. If a new valid message has been
received in the meantime, the FIFO stays in pending_1 state and the new message is
available in the output mailbox.
If the application does not release the mailbox, the next valid message will be stored in the
FIFO which enters pending_2 state (FMP[1:0] = 10b). The storage process is repeated for
the next valid message putting the FIFO into pending_3 state (FMP[1:0] = 11b). At this
point, the software must release the output mailbox by setting the RFOM bit, so that a
mailbox is free to store the next valid message. Otherwise the next valid message received
will cause a loss of message.
Refer also to Section 32.7.5: Message storage

Overrun
Once the FIFO is in pending_3 state (i.e. the three mailboxes are full) the next valid
message reception will lead to an overrun and a message will be lost. The hardware
signals the overrun condition by setting the FOVR bit in the CAN_RFR register. Which
message is lost depends on the configuration of the FIFO:
•

If the FIFO lock function is disabled (RFLM bit in the CAN_MCR register cleared) the
last message stored in the FIFO will be overwritten by the new incoming message. In
this case the latest messages will be always available to the application.

•

If the FIFO lock function is enabled (RFLM bit in the CAN_MCR register set) the most
recent message will be discarded and the software will have the three oldest messages
in the FIFO available.

Reception related interrupts
Once a message has been stored in the FIFO, the FMP[1:0] bits are updated and an
interrupt request is generated if the FMPIE bit in the CAN_IER register is set.
When the FIFO becomes full (i.e. a third message is stored) the FULL bit in the CAN_RFR
register is set and an interrupt is generated if the FFIE bit in the CAN_IER register is set.
On overrun condition, the FOVR bit is set and an interrupt is generated if the FOVIE bit in
the CAN_IER register is set.

32.7.4

Identifier filtering
In the CAN protocol the identifier of a message is not associated with the address of a node
but related to the content of the message. Consequently a transmitter broadcasts its
message to all receivers. On message reception a receiver node decides - depending on
the identifier value - whether the software needs the message or not. If the message is
needed, it is copied into the SRAM. If not, the message must be discarded without
intervention by the software.
To fulfill this requirement, the bxCAN Controller provides 28 configurable and scalable filter
banks (27-0) to the application. This hardware filtering saves CPU resources which would
be otherwise needed to perform filtering by software. Each filter bank x consists of two 32-bit
registers, CAN_FxR0 and CAN_FxR1.

1086/1747

RM0090 Rev 17

RM0090

Controller area network (bxCAN)

Scalable width
To optimize and adapt the filters to the application needs, each filter bank can be scaled
independently. Depending on the filter scale a filter bank provides:
•

One 32-bit filter for the STDID[10:0], EXTID[17:0], IDE and RTR bits.

•

Two 16-bit filters for the STDID[10:0], RTR, IDE and EXTID[17:15] bits.

Refer to Figure 342.
Furthermore, the filters can be configured in mask mode or in identifier list mode.

Mask mode
In mask mode the identifier registers are associated with mask registers specifying which
bits of the identifier are handled as “must match” or as “don’t care”.

Identifier list mode
In identifier list mode, the mask registers are used as identifier registers. Thus instead of
defining an identifier and a mask, two identifiers are specified, doubling the number of single
identifiers. All bits of the incoming identifier must match the bits specified in the filter
registers.

Filter bank scale and mode configuration
The filter banks are configured by means of the corresponding CAN_FMR register. To
configure a filter bank it must be deactivated by clearing the FACT bit in the CAN_FAR
register. The filter scale is configured by means of the corresponding FSCx bit in the
CAN_FS1R register, refer to Figure 342. The identifier list or identifier mask mode for the
corresponding Mask/Identifier registers is configured by means of the FBMx bits in the
CAN_FMR register.
To filter a group of identifiers, configure the Mask/Identifier registers in mask mode.
To select single identifiers, configure the Mask/Identifier registers in identifier list mode.
Filters not used by the application should be left deactivated.
Each filter within a filter bank is numbered (called the Filter Number) from 0 to a maximum
dependent on the mode and the scale of each of the filter banks.
Concerning the filter configuration, refer to Figure 342.

RM0090 Rev 17

1087/1747
1120

Controller area network (bxCAN)

RM0090

FBMx = 0
FBMx = 1

FSCx = 1

Figure 342. Filter bank scale configuration - register organization
Filter
Num.

One 32-Bit Filter - Identifier Mask
ID
Mask
Mapping

CAN_FxR1[31:24]
CAN_FxR2[31:24]
STID[10:3]

CAN_FxR1[23:16]
CAN_FxR2[23:16]
STID[2:0]

CAN_FxR1[15:8]
CAN_FxR2[15:8]

EXID[17:13]

EXID[12:5]

CAN_FxR1[7:0]
CAN_FxR2[7:0]
EXID[4:0]

IDE RTR

n
0

Two 32-Bit Filters - Identifier List
ID
ID
Mapping

CAN_FxR1[31:24]
CAN_FxR2[31:24]
STID[10:3]

CAN_FxR1[23:16]
CAN_FxR2[23:16]
STID[2:0]

CAN_FxR1[15:8]
CAN_FxR2[15:8]

EXID[17:13]

EXID[12:5]

n
n+1

CAN_FxR1[7:0]
CAN_FxR2[7:0]
EXID[4:0]

IDE RTR

0

FSCx = 0

FBMx = 0

Two 16-Bit Filters - Identifier Mask
CAN_FxR1[15:8]
CAN_FxR1[31:24]

CAN_FxR1[7:0]
CAN_FxR1[23:16]

n

ID
Mask
Mapping

CAN_FxR2[15:8]
CAN_FxR2[31:24]

CAN_FxR2[7:0]
CAN_FxR2[23:16]

n+1

STID[10:3]

STID[2:0] RTR IDE EXID[17:15]

FBMx = 1

Four 16-Bit Filters - Identifier List
ID
ID

CAN_FxR1[15:8]
CAN_FxR1[31:24]

CAN_FxR1[7:0]
CAN_FxR1[23:16]

n
n+1

ID
ID
Mapping

CAN_FxR2[15:8]
CAN_FxR2[31:24]

CAN_FxR2[7:0]
CAN_FxR2[23:16]

n+2
n+3

STID[10:3]

STID[2:0] RTR IDE EXID[17:15]

Filter Bank Mode2

Filter Bank Scale
Config. Bits1

ID
Mask

x = filter bank number
ID=Identifier
1
2

These bits are located in the CAN_FS1R register
These bits are located in the CAN_FM1R register

Filter match index
Once a message has been received in the FIFO it is available to the application. Typically,
application data is copied into SRAM locations. To copy the data to the right location the
application has to identify the data by means of the identifier. To avoid this, and to ease the
access to the SRAM locations, the CAN controller provides a Filter Match Index.
This index is stored in the mailbox together with the message according to the filter priority
rules. Thus each received message has its associated filter match index.
The Filter Match index can be used in two ways:
•

Compare the Filter Match index with a list of expected values.

•

Use the Filter Match Index as an index on an array to access the data destination
location.

For nonmasked filters, the software no longer has to compare the identifier.
If the filter is masked the software reduces the comparison to the masked bits only.
The index value of the filter number does not take into account the activation state of the
filter banks. In addition, two independent numbering schemes are used, one for each FIFO.
Refer to Figure 343 for an example.

1088/1747

RM0090 Rev 17

RM0090

Controller area network (bxCAN)
Figure 343. Example of filter numbering
Filter
Bank

FIFO0

0

ID List (32-bit)

1

ID Mask (32-bit)

3

ID List (16-bit)

5

Deactivated
ID List (32-bit)

6

ID Mask (16-bit)

9

ID List (32-bit)

13

ID Mask (32-bit)

Filter
Num.
0

Filter
Bank

FIFO1

2

ID Mask (16-bit)

2

4

ID List (32-bit)

3
4
5
6

7

Deactivated
ID Mask (16-bit)

8

ID Mask (16-bit)

10

Deactivated
ID List (16-bit)

11

ID List (32-bit)

12

ID Mask (32-bit)

1

7
8
9
10
11
12

13

Filter
Num.
0
1
2
3

4
5
6
7
8
9
10
11
12
13

14

ID = Identifier

Filter priority rules
Depending on the filter combination it may occur that an identifier passes successfully
through several filters. In this case the filter match value stored in the receive mailbox is
chosen according to the following priority rules:
•

A 32-bit filter takes priority over a 16-bit filter.

•

For filters of equal scale, priority is given to the Identifier List mode over the Identifier
Mask mode

•

For filters of equal scale and mode, priority is given by the filter number (the lower the
number, the higher the priority).

RM0090 Rev 17

1089/1747
1120

Controller area network (bxCAN)

RM0090
Figure 344. Filtering mechanism - example

Example of 3 filter banks in 32-bit Unidentified List mode and
the remaining in 32-bit Identifier Mask mode
Message Received
Identifier

Ctrl

Data

Filter bank
Num

Receive FIFO

3

1

4

Identifier & Mask

2

Identifier List

0

Identifier
Identifier
Identifier

0
1
4

Identifier

5

Identifier
Mask

2

Identifier #4 Match

Message
Stored

FMI

Identifier
3
Mask
No Match
Found

Filter number stored
in the Filter Match
Index field within the
CAN_RDTxR register

Message Discarded

The example above shows the filtering principle of the bxCAN. On reception of a message,
the identifier is compared first with the filters configured in identifier list mode. If there is a
match, the message is stored in the associated FIFO and the index of the matching filter is
stored in the Filter Match Index. As shown in the example, the identifier matches with
Identifier #2 thus the message content and FMI 2 is stored in the FIFO.
If there is no match, the incoming identifier is then compared with the filters configured in
mask mode.
If the identifier does not match any of the identifiers configured in the filters, the message is
discarded by hardware without disturbing the software.

32.7.5

Message storage
The interface between the software and the hardware for the CAN messages is
implemented by means of mailboxes. A mailbox contains all information related to a
message; identifier, data, control, status and time stamp information.

Transmit mailbox
The software sets up the message to be transmitted in an empty transmit mailbox. The
status of the transmission is indicated by hardware in the CAN_TSR register.

1090/1747

RM0090 Rev 17

RM0090

Controller area network (bxCAN)
Table 182. Transmit mailbox mapping
Offset to transmit mailbox base address

Register name

0

CAN_TIxR

4

CAN_TDTxR

8

CAN_TDLxR

12

CAN_TDHxR

Receive mailbox
When a message has been received, it is available to the software in the FIFO output
mailbox. Once the software has handled the message (e.g. read it) the software must
release the FIFO output mailbox by means of the RFOM bit in the CAN_RFR register to
make the next incoming message available. The filter match index is stored in the MFMI
field of the CAN_RDTxR register. The 16-bit time stamp value is stored in the TIME[15:0]
field of CAN_RDTxR.
Table 183. Receive mailbox mapping
Offset to receive mailbox base
address (bytes)

Register name

0

CAN_RIxR

4

CAN_RDTxR

8

CAN_RDLxR

12

CAN_RDHxR

Figure 345. CAN error state diagram
7HEN 4%# OR 2%#  

%22/2 !#4)6%

%22/2 0! 33)6%

7HEN 4%# AND 2%#  

7HEN   RECESSIVE BITS OCCUR

7HEN 4%#  

"53 /&&
AI

RM0090 Rev 17

1091/1747
1120

Controller area network (bxCAN)

32.7.6

RM0090

Error management
The error management as described in the CAN protocol is handled entirely by hardware
using a Transmit Error Counter (TEC value, in CAN_ESR register) and a Receive Error
Counter (REC value, in the CAN_ESR register), which get incremented or decremented
according to the error condition. For detailed information about TEC and REC management,
refer to the CAN standard.
Both of them may be read by software to determine the stability of the network.
Furthermore, the CAN hardware provides detailed information on the current error status in
CAN_ESR register. By means of the CAN_IER register (ERRIE bit, etc.), the software can
configure the interrupt generation on error detection in a very flexible way.

Bus-Off recovery
The Bus-Off state is reached when TEC is greater than 255, this state is indicated by BOFF
bit in CAN_ESR register. In Bus-Off state, the bxCAN is no longer able to transmit and
receive messages.
Depending on the ABOM bit in the CAN_MCR register bxCAN will recover from Bus-Off
(become error active again) either automatically or on software request. But in both cases
the bxCAN has to wait at least for the recovery sequence specified in the CAN standard
(128 occurrences of 11 consecutive recessive bits monitored on CANRX).
If ABOM is set, the bxCAN will start the recovering sequence automatically after it has
entered Bus-Off state.
If ABOM is cleared, the software must initiate the recovering sequence by requesting
bxCAN to enter and to leave initialization mode.
Note:

In initialization mode, bxCAN does not monitor the CANRX signal, therefore it cannot
complete the recovery sequence. To recover, bxCAN must be in normal mode.

32.7.7

Bit timing
The bit timing logic monitors the serial bus-line and performs sampling and adjustment of
the sample point by synchronizing on the start-bit edge and resynchronizing on the following
edges.
Its operation may be explained simply by splitting nominal bit time into three segments as
follows:
•

Synchronization segment (SYNC_SEG): a bit change is expected to occur within this
time segment. It has a fixed length of one time quantum (1 x tq).

•

Bit segment 1 (BS1): defines the location of the sample point. It includes the
PROP_SEG and PHASE_SEG1 of the CAN standard. Its duration is programmable
between 1 and 16 time quanta but may be automatically lengthened to compensate for
positive phase drifts due to differences in the frequency of the various nodes of the
network.

•

Bit segment 2 (BS2): defines the location of the transmit point. It represents the
PHASE_SEG2 of the CAN standard. Its duration is programmable between 1 and 8
time quanta but may also be automatically shortened to compensate for negative
phase drifts.

The resynchronization Jump Width (SJW) defines an upper bound to the amount of
lengthening or shortening of the bit segments. It is programmable between 1 and 4 time
quanta.

1092/1747

RM0090 Rev 17

RM0090

Controller area network (bxCAN)
A valid edge is defined as the first transition in a bit time from dominant to recessive bus
level provided the controller itself does not send a recessive bit.
If a valid edge is detected in BS1 instead of SYNC_SEG, BS1 is extended by up to SJW so
that the sample point is delayed.
Conversely, if a valid edge is detected in BS2 instead of SYNC_SEG, BS2 is shortened by
up to SJW so that the transmit point is moved earlier.
As a safeguard against programming errors, the configuration of the Bit Timing Register
(CAN_BTR) is only possible while the device is in Standby mode.

Note:

For a detailed description of the CAN bit timing and resynchronization mechanism, refer to
the ISO 11898 standard.
Figure 346. Bit timing
NOMINAL BIT TIME
SYNC_SEG

BIT SEGMENT 1 (BS1)

1 x tq

BIT SEGMENT 2 (BS2)

tBS1

1
BaudRate = ---------------------------------------------NominalBitTime

tBS2

SAMPLE POINT

TRANSMIT POINT

NominalBitTime = 1 × t q + t BS1 + t BS2
with:
tBS1 = tq x (TS1[3:0] + 1),
tBS2 = tq x (TS2[2:0] + 1),
tq = (BRP[9:0] + 1) x tPCLK
where tq refers to the Time quantum
tPCLK = time period of the APB clock,
BRP[9:0], TS1[3:0] and TS2[2:0] are defined in the CAN_BTR Register.

RM0090 Rev 17

1093/1747
1120

Controller area network (bxCAN)

RM0090
Figure 347. CAN frames
Inter-Frame Space
or Overload Frame

Data Frame (Standard identifier)

Inter-Frame Space
Arbitration Field

Ctrl Field

32

6

ID

44 + 8 * N
Data Field

Ack Field
2

CRC Field

8*N

16

EOF

CRC
ACK

SOF

RTR
IDE
r0

DLC

Inter-Frame Space

Arbitration Field Ctrl Field

32

Data Field

6

32

ID

CRC

6

ID

16

Flag Echo Error Delimiter
6

Ack Field
2

7
EOF

ACK

SOF

RTR
IDE
r0

Inter-Frame Space
or Overload Frame

Error Frame

Inter-Frame Space
or Overload Frame

CRC

DLC

EOF
ACK

Remote Frame
44
Ctrl Field
CRC Field

32

CRC Field Ack Field
2
16
7

RTR
r1
r0

SRR
IDE

SOF

Arbitration Field

Error
Flag
6

8*N

DLC

Inter-Frame Space

Data Frame or
Remote Frame

Inter-Frame Space
or Overload Frame

Data Frame (Extended Identifier)
64 + 8 * N

Arbitration Field

7

Notes:
0 <= N <= 8
SOF = Start Of Frame

8

ID = Identifier
RTR = Remote Transmission Request
Any Frame

Inter-Frame Space
Suspend
Intermission Transmission Bus Idle
3
8

Data Frame or
Remote Frame

IDE = Identifier Extension Bit
r0 = Reserved Bit
DLC = Data Length Code
CRC = Cyclic Redundancy Code
Error flag: 6 dominant bits if node is error

End Of Frame or
Error Delimiter or
Overload Delimiter

active else 6 recessive bits.
Overload Frame

Overload Overload
Flag
Echo
6
6

Inter-Frame Space
or Error Frame

Overload
Delimiter
8

Suspend transmission: applies to error
passive nodes only.
EOF = End of Frame
ACK = Acknowledge bit
Ctrl = Control

ai15154

32.8

bxCAN interrupts
Four interrupt vectors are dedicated to bxCAN. Each interrupt source can be independently
enabled or disabled by means of the CAN Interrupt Enable Register (CAN_IER).

1094/1747

RM0090 Rev 17

RM0090

Controller area network (bxCAN)
Figure 348. Event flags and interrupt generation
CAN_IER

CAN_TSR

RQCP0
RQCP1
RQCP2

+

TRANSMIT
INTERRUPT

TMEIE

&

FMPIE0

&

FMP0

FIFO 0
INTERRUPT

CAN_RF0R

FFIE0

&

FOVIE0

&

FMPIE1

&

FULL0

FOVR0

FMP1

+

FIFO 1
INTERRUPT

CAN_RF1R

FFIE1

&

FOVIE1

&

FULL1

FOVR1

+

ERRIE
EWGIE

&

EPVIE

&

EWGF

CAN_ESR

EPVF
BOFIE
BOFF
LECIE
1≤LEC≤6

&

•

•

•

CAN_MSR

STATUS CHANGE
ERROR
INTERRUPT

+

WKUIE

&

SLKIE

&

SLAKI

•

ERRI

&

WKUI

CAN_MSR

&

+

The transmit interrupt can be generated by the following events:
–

Transmit mailbox 0 becomes empty, RQCP0 bit in the CAN_TSR register set.

–

Transmit mailbox 1 becomes empty, RQCP1 bit in the CAN_TSR register set.

–

Transmit mailbox 2 becomes empty, RQCP2 bit in the CAN_TSR register set.

The FIFO 0 interrupt can be generated by the following events:
–

Reception of a new message, FMP0 bits in the CAN_RF0R register are not ‘00’.

–

FIFO0 full condition, FULL0 bit in the CAN_RF0R register set.

–

FIFO0 overrun condition, FOVR0 bit in the CAN_RF0R register set.

The FIFO 1 interrupt can be generated by the following events:
–

Reception of a new message, FMP1 bits in the CAN_RF1R register are not ‘00’.

–

FIFO1 full condition, FULL1 bit in the CAN_RF1R register set.

–

FIFO1 overrun condition, FOVR1 bit in the CAN_RF1R register set.

The error and status change interrupt can be generated by the following events:
–

Error condition, for more details on error conditions refer to the CAN Error Status
register (CAN_ESR).
RM0090 Rev 17

1095/1747
1120

Controller area network (bxCAN)

32.9

RM0090

–

Wakeup condition, SOF monitored on the CAN Rx signal.

–

Entry into Sleep mode.

CAN registers
The peripheral registers have to be accessed by words (32 bits).

32.9.1

Register access protection
Erroneous access to certain configuration registers can cause the hardware to temporarily
disturb the whole CAN network. Therefore the CAN_BTR register can be modified by
software only while the CAN hardware is in initialization mode.
Although the transmission of incorrect data will not cause problems at the CAN network
level, it can severely disturb the application. A transmit mailbox can be only modified by
software while it is in empty state, refer to Figure 340.
The filter values can be modified either deactivating the associated filter banks or by setting
the FINIT bit. Moreover, the modification of the filter configuration (scale, mode and FIFO
assignment) in CAN_FMxR, CAN_FSxR and CAN_FFAR registers can only be done when
the filter initialization mode is set (FINIT=1) in the CAN_FMR register.

32.9.2

CAN control and status registers
Refer to Section 2.2 on page 45 for a list of abbreviations used in register descriptions.

CAN master control register (CAN_MCR)
Address offset: 0x00
Reset value: 0x0001 0002
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16
DBF

Reserved
15

14

13

12

11

RESET

10

9

rw

8

7

6

5

4

3

2

1

0

TTCM

ABOM

AWUM

NART

RFLM

TXFP

SLEEP

INRQ

rw

rw

rw

rw

rw

rw

rw

rw

Reserved
rs

Bits 31:17 Reserved, must be kept at reset value.
Bit 16 DBF: Debug freeze
0: CAN working during debug
1: CAN reception/transmission frozen during debug. Reception FIFOs can still be
accessed/controlled normally.
Bit 15 RESET: bxCAN software master reset
0: Normal operation.
1: Force a master reset of the bxCAN -> Sleep mode activated after reset (FMP bits and
CAN_MCR register are initialized to the reset values). This bit is automatically reset to 0.
Bits 14:8 Reserved, must be kept at reset value.

1096/1747

RM0090 Rev 17

RM0090

Controller area network (bxCAN)

Bit 7 TTCM: Time triggered communication mode
0: Time Triggered Communication mode disabled.
1: Time Triggered Communication mode enabled
Note: For more information on Time Triggered Communication mode refer to Section 32.7.2:
Time triggered communication mode.
Bit 6 ABOM: Automatic bus-off management
This bit controls the behavior of the CAN hardware on leaving the Bus-Off state.
0: The Bus-Off state is left on software request, once 128 occurrences of 11 recessive bits
have been monitored and the software has first set and cleared the INRQ bit of the
CAN_MCR register.
1: The Bus-Off state is left automatically by hardware once 128 occurrences of 11 recessive
bits have been monitored.
For detailed information on the Bus-Off state refer to Section 32.7.6: Error management.
Bit 5 AWUM: Automatic wakeup mode
This bit controls the behavior of the CAN hardware on message reception during Sleep
mode.
0: The Sleep mode is left on software request by clearing the SLEEP bit of the CAN_MCR
register.
1: The Sleep mode is left automatically by hardware on CAN message detection.
The SLEEP bit of the CAN_MCR register and the SLAK bit of the CAN_MSR register are
cleared by hardware.
Bit 4 NART: No automatic retransmission
0: The CAN hardware will automatically retransmit the message until it has been
successfully transmitted according to the CAN standard.
1: A message will be transmitted only once, independently of the transmission result
(successful, error or arbitration lost).
Bit 3 RFLM: Receive FIFO locked mode
0: Receive FIFO not locked on overrun. Once a receive FIFO is full the next incoming
message will overwrite the previous one.
1: Receive FIFO locked against overrun. Once a receive FIFO is full the next incoming
message will be discarded.
Bit 2 TXFP: Transmit FIFO priority
This bit controls the transmission order when several mailboxes are pending at the same
time.
0: Priority driven by the identifier of the message
1: Priority driven by the request order (chronologically)
Bit 1 SLEEP: Sleep mode request
This bit is set by software to request the CAN hardware to enter the Sleep mode. Sleep
mode will be entered as soon as the current CAN activity (transmission or reception of a
CAN frame) has been completed.
This bit is cleared by software to exit Sleep mode.
This bit is cleared by hardware when the AWUM bit is set and a SOF bit is detected on the
CAN Rx signal.
This bit is set after reset - CAN starts in Sleep mode.

RM0090 Rev 17

1097/1747
1120

Controller area network (bxCAN)

RM0090

Bit 0 INRQ: Initialization request
The software clears this bit to switch the hardware into normal mode. Once 11 consecutive
recessive bits have been monitored on the Rx signal the CAN hardware is synchronized and
ready for transmission and reception. Hardware signals this event by clearing the INAK bit in
the CAN_MSR register.
Software sets this bit to request the CAN hardware to enter initialization mode. Once
software has set the INRQ bit, the CAN hardware waits until the current CAN activity
(transmission or reception) is completed before entering the initialization mode. Hardware
signals this event by setting the INAK bit in the CAN_MSR register.

CAN master status register (CAN_MSR)
Address offset: 0x04
Reset value: 0x0000 0C02
31

30

29

28

27

26

25

24

23

22

21

6

5

20

19

18

17

16

Reserved
15

14

13

Reserved.

12

11

10

9

8

RX

SAMP

RXM

TXM

r

r

r

r

7

Reserved

4

3

2

1

0

SLAKI

WKUI

ERRI

SLAK

INAK

rc_w1

rc_w1

rc_w1

r

r

Bits 31:12 Reserved, must be kept at reset value.
Bit 11 RX: CAN Rx signal
Monitors the actual value of the CAN_RX Pin.
Bit 10 SAMP: Last sample point
The value of RX on the last sample point (current received bit value).
Bit 9 RXM: Receive mode
The CAN hardware is currently receiver.
Bit 8 TXM: Transmit mode
The CAN hardware is currently transmitter.
Bits 7:5 Reserved, must be kept at reset value.
Bit 4 SLAKI: Sleep acknowledge interrupt
When SLKIE=1, this bit is set by hardware to signal that the bxCAN has entered Sleep
Mode. When set, this bit generates a status change interrupt if the SLKIE bit in the
CAN_IER register is set.
This bit is cleared by software or by hardware, when SLAK is cleared.
Note: When SLKIE=0, no polling on SLAKI is possible. In this case the SLAK bit can be
polled.
Bit 3 WKUI: Wakeup interrupt
This bit is set by hardware to signal that a SOF bit has been detected while the CAN
hardware was in Sleep mode. Setting this bit generates a status change interrupt if the
WKUIE bit in the CAN_IER register is set.
This bit is cleared by software.

1098/1747

RM0090 Rev 17

RM0090

Controller area network (bxCAN)

Bit 2 ERRI: Error interrupt
This bit is set by hardware when a bit of the CAN_ESR has been set on error detection and
the corresponding interrupt in the CAN_IER is enabled. Setting this bit generates a status
change interrupt if the ERRIE bit in the CAN_IER register is set.
This bit is cleared by software.
Bit 1 SLAK: Sleep acknowledge
This bit is set by hardware and indicates to the software that the CAN hardware is now in
Sleep mode. This bit acknowledges the Sleep mode request from the software (set SLEEP
bit in CAN_MCR register).
This bit is cleared by hardware when the CAN hardware has left Sleep mode (to be
synchronized on the CAN bus). To be synchronized the hardware has to monitor a
sequence of 11 consecutive recessive bits on the CAN RX signal.
Note: The process of leaving Sleep mode is triggered when the SLEEP bit in the CAN_MCR
register is cleared. Refer to the AWUM bit of the CAN_MCR register description for
detailed information for clearing SLEEP bit
Bit 0 INAK: Initialization acknowledge
This bit is set by hardware and indicates to the software that the CAN hardware is now in
initialization mode. This bit acknowledges the initialization request from the software (set
INRQ bit in CAN_MCR register).
This bit is cleared by hardware when the CAN hardware has left the initialization mode (to
be synchronized on the CAN bus). To be synchronized the hardware has to monitor a
sequence of 11 consecutive recessive bits on the CAN RX signal.

CAN transmit status register (CAN_TSR)
Address offset: 0x08
Reset value: 0x1C00 0000
31

30

29

28

27

26

LOW2

LOW1

LOW0

TME2

TME1

TME0

r

r

r

r

r

r

r

r

rs

15

14

13

12

11

10

9

8

7

ABRQ1
rs

Reserved
Res.

25

24

CODE[1:0]

23

TERR1 ALST1 TXOK1 RQCP1 ABRQ0
rc_w1

rc_w1

rc_w1

rc_w1

22

ABRQ2

rs

21

20

Reserved
6

5
Reserved

19

18

17

16

TERR2 ALST2 TXOK2 RQCP2

4

rc_w1

rc_w1

rc_w1

rc_w1

3

2

1

0

TERR0 ALST0 TXOK0 RQCP0
rc_w1

rc_w1

rc_w1

rc_w1

Bit 31 LOW2: Lowest priority flag for mailbox 2
This bit is set by hardware when more than one mailbox are pending for transmission and
mailbox 2 has the lowest priority.
Bit 30 LOW1: Lowest priority flag for mailbox 1
This bit is set by hardware when more than one mailbox are pending for transmission and
mailbox 1 has the lowest priority.
Bit 29 LOW0: Lowest priority flag for mailbox 0
This bit is set by hardware when more than one mailbox are pending for transmission and
mailbox 0 has the lowest priority.
Note: The LOW[2:0] bits are set to zero when only one mailbox is pending.
Bit 28 TME2: Transmit mailbox 2 empty
This bit is set by hardware when no transmit request is pending for mailbox 2.
Bit 27 TME1: Transmit mailbox 1 empty
This bit is set by hardware when no transmit request is pending for mailbox 1.

RM0090 Rev 17

1099/1747
1120

Controller area network (bxCAN)

RM0090

Bit 26 TME0: Transmit mailbox 0 empty
This bit is set by hardware when no transmit request is pending for mailbox 0.
Bits 25:24 CODE[1:0]: Mailbox code
In case at least one transmit mailbox is free, the code value is equal to the number of the
next transmit mailbox free.
In case all transmit mailboxes are pending, the code value is equal to the number of the
transmit mailbox with the lowest priority.
Bit 23 ABRQ2: Abort request for mailbox 2
Set by software to abort the transmission request for the corresponding mailbox.
Cleared by hardware when the mailbox becomes empty.
Setting this bit has no effect when the mailbox is not pending for transmission.
Bits 22:20 Reserved, must be kept at reset value.
Bit 19 TERR2: Transmission error of mailbox 2
This bit is set when the previous TX failed due to an error.
Bit 18 ALST2: Arbitration lost for mailbox 2
This bit is set when the previous TX failed due to an arbitration lost.
Bit 17 TXOK2: Transmission OK of mailbox 2
The hardware updates this bit after each transmission attempt.
0: The previous transmission failed
1: The previous transmission was successful
This bit is set by hardware when the transmission request on mailbox 2 has been completed
successfully. Refer to Figure 340.
Bit 16 RQCP2: Request completed mailbox2
Set by hardware when the last request (transmit or abort) has been performed.
Cleared by software writing a “1” or by hardware on transmission request (TXRQ2 set in
CAN_TMID2R register).
Clearing this bit clears all the status bits (TXOK2, ALST2 and TERR2) for Mailbox 2.
Bit 15 ABRQ1: Abort request for mailbox 1
Set by software to abort the transmission request for the corresponding mailbox.
Cleared by hardware when the mailbox becomes empty.
Setting this bit has no effect when the mailbox is not pending for transmission.
Bits 14:12 Reserved, must be kept at reset value.
Bit 11 TERR1: Transmission error of mailbox1
This bit is set when the previous TX failed due to an error.
Bit 10 ALST1: Arbitration lost for mailbox1
This bit is set when the previous TX failed due to an arbitration lost.
Bit 9 TXOK1: Transmission OK of mailbox1
The hardware updates this bit after each transmission attempt.
0: The previous transmission failed
1: The previous transmission was successful
This bit is set by hardware when the transmission request on mailbox 1 has been completed
successfully. Refer to Figure 340
Bit 8 RQCP1: Request completed mailbox1
Set by hardware when the last request (transmit or abort) has been performed.
Cleared by software writing a “1” or by hardware on transmission request (TXRQ1 set in
CAN_TI1R register).
Clearing this bit clears all the status bits (TXOK1, ALST1 and TERR1) for Mailbox 1.

1100/1747

RM0090 Rev 17

RM0090

Controller area network (bxCAN)

Bit 7 ABRQ0: Abort request for mailbox0
Set by software to abort the transmission request for the corresponding mailbox.
Cleared by hardware when the mailbox becomes empty.
Setting this bit has no effect when the mailbox is not pending for transmission.
Bits 6:4 Reserved, must be kept at reset value.
Bit 3 TERR0: Transmission error of mailbox0
This bit is set when the previous TX failed due to an error.
Bit 2 ALST0: Arbitration lost for mailbox0
This bit is set when the previous TX failed due to an arbitration lost.
Bit 1 TXOK0: Transmission OK of mailbox0
The hardware updates this bit after each transmission attempt.
0: The previous transmission failed
1: The previous transmission was successful
This bit is set by hardware when the transmission request on mailbox 1 has been completed
successfully. Refer to Figure 340
Bit 0 RQCP0: Request completed mailbox0
Set by hardware when the last request (transmit or abort) has been performed.
Cleared by software writing a “1” or by hardware on transmission request (TXRQ0 set in
CAN_TI0R register).
Clearing this bit clears all the status bits (TXOK0, ALST0 and TERR0) for Mailbox 0.

CAN receive FIFO 0 register (CAN_RF0R)
Address offset: 0x0C
Reset value: 0x0000 0000
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

Reserved
15

14

13

12

11

10

9

8

7

RFOM0 FOVR0

Reserved

rs

rc_w1

FULL0
rc_w1

Res.

FMP0[1:0]
r

r

Bits 31:6 Reserved, must be kept at reset value.
Bit 5 RFOM0: Release FIFO 0 output mailbox
Set by software to release the output mailbox of the FIFO. The output mailbox can only be
released when at least one message is pending in the FIFO. Setting this bit when the FIFO
is empty has no effect. If at least two messages are pending in the FIFO, the software has to
release the output mailbox to access the next message.
Cleared by hardware when the output mailbox has been released.
Bit 4 FOVR0: FIFO 0 overrun
This bit is set by hardware when a new message has been received and passed the filter
while the FIFO was full.
This bit is cleared by software.
Bit 3 FULL0: FIFO 0 full
Set by hardware when three messages are stored in the FIFO.
This bit is cleared by software.
Bit 2 Reserved, must be kept at reset value.

RM0090 Rev 17

1101/1747
1120

Controller area network (bxCAN)

RM0090

Bits 1:0 FMP0[1:0]: FIFO 0 message pending
These bits indicate how many messages are pending in the receive FIFO.
FMP is increased each time the hardware stores a new message in to the FIFO. FMP is
decreased each time the software releases the output mailbox by setting the RFOM0 bit.

CAN receive FIFO 1 register (CAN_RF1R)
Address offset: 0x10
Reset value: 0x0000 0000
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

Reserved
15

14

13

12

11

10

9

8

7

RFOM1 FOVR1

Reserved

rs

rc_w1

FULL1
rc_w1

FMP1[1:0]

Res.

r

r

Bits 31:6 Reserved, must be kept at reset value.
Bit 5 RFOM1: Release FIFO 1 output mailbox
Set by software to release the output mailbox of the FIFO. The output mailbox can only be
released when at least one message is pending in the FIFO. Setting this bit when the FIFO
is empty has no effect. If at least two messages are pending in the FIFO, the software has to
release the output mailbox to access the next message.
Cleared by hardware when the output mailbox has been released.
Bit 4 FOVR1: FIFO 1 overrun
This bit is set by hardware when a new message has been received and passed the filter
while the FIFO was full.
This bit is cleared by software.
Bit 3 FULL1: FIFO 1 full
Set by hardware when three messages are stored in the FIFO.
This bit is cleared by software.
Bit 2 Reserved, must be kept at reset value.
Bits 1:0 FMP1[1:0]: FIFO 1 message pending
These bits indicate how many messages are pending in the receive FIFO1.
FMP1 is increased each time the hardware stores a new message in to the FIFO1. FMP is
decreased each time the software releases the output mailbox by setting the RFOM1 bit.

CAN interrupt enable register (CAN_IER)
Address offset: 0x14
Reset value: 0x0000 0000
31

30

29

28

27

26

25

24

23

22

21

20

19

18

Reserved
15

14

ERRIE
rw

1102/1747

13
Reserved

12

11

10

9

8

LEC
IE

BOF
IE

EPV
IE

EWG
IE

rw

rw

rw

rw

17

16

SLKIE

WKUIE

rw

rw

7

6

5

4

3

2

1

0

Res.

FOV
IE1

FF
IE1

FMP
IE1

FOV
IE0

FF
IE0

FMP
IE0

TME
IE

rw

rw

rw

rw

rw

rw

rw

RM0090 Rev 17

RM0090

Controller area network (bxCAN)

Bits 31:18 Reserved, must be kept at reset value.
Bit 17 SLKIE: Sleep interrupt enable
0: No interrupt when SLAKI bit is set.
1: Interrupt generated when SLAKI bit is set.
Bit 16 WKUIE: Wakeup interrupt enable
0: No interrupt when WKUI is set.
1: Interrupt generated when WKUI bit is set.
Bit 15 ERRIE: Error interrupt enable
0: No interrupt will be generated when an error condition is pending in the CAN_ESR.
1: An interrupt will be generation when an error condition is pending in the CAN_ESR.
Bits 14:12 Reserved, must be kept at reset value.
Bit 11 LECIE: Last error code interrupt enable
0: ERRI bit will not be set when the error code in LEC[2:0] is set by hardware on error
detection.
1: ERRI bit will be set when the error code in LEC[2:0] is set by hardware on error detection.
Bit 10 BOFIE: Bus-off interrupt enable
0: ERRI bit will not be set when BOFF is set.
1: ERRI bit will be set when BOFF is set.
Bit 9 EPVIE: Error passive interrupt enable
0: ERRI bit will not be set when EPVF is set.
1: ERRI bit will be set when EPVF is set.
Bit 8 EWGIE: Error warning interrupt enable
0: ERRI bit will not be set when EWGF is set.
1: ERRI bit will be set when EWGF is set.
Bit 7 Reserved, must be kept at reset value.
Bit 6 FOVIE1: FIFO overrun interrupt enable
0: No interrupt when FOVR is set.
1: Interrupt generation when FOVR is set.
Bit 5 FFIE1: FIFO full interrupt enable
0: No interrupt when FULL bit is set.
1: Interrupt generated when FULL bit is set.
Bit 4 FMPIE1: FIFO message pending interrupt enable
0: No interrupt generated when state of FMP[1:0] bits are not 00b.
1: Interrupt generated when state of FMP[1:0] bits are not 00b.
Bit 3 FOVIE0: FIFO overrun interrupt enable
0: No interrupt when FOVR bit is set.
1: Interrupt generated when FOVR bit is set.

RM0090 Rev 17

1103/1747
1120

Controller area network (bxCAN)

RM0090

Bit 2 FFIE0: FIFO full interrupt enable
0: No interrupt when FULL bit is set.
1: Interrupt generated when FULL bit is set.
Bit 1 FMPIE0: FIFO message pending interrupt enable
0: No interrupt generated when state of FMP[1:0] bits are not 00b.
1: Interrupt generated when state of FMP[1:0] bits are not 00b.
Bit 0 TMEIE: Transmit mailbox empty interrupt enable
0: No interrupt when RQCPx bit is set.
1: Interrupt generated when RQCPx bit is set.
Note: Refer to Section 32.8: bxCAN interrupts.

CAN error status register (CAN_ESR)
Address offset: 0x18
Reset value: 0x0000 0000
31

30

29

28

27

26

25

24

23

22

21

REC[7:0]

20

19

18

17

16

r

r

r

TEC[7:0]

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

r

r

r

5

4

3

LEC[2:0]

Reserved

rw

rw

rw

Res.

2

1

0

BOFF

EPVF

EWGF

r

r

r

Bits 31:24 REC[7:0]: Receive error counter
The implementing part of the fault confinement mechanism of the CAN protocol. In case of
an error during reception, this counter is incremented by 1 or by 8 depending on the error
condition as defined by the CAN standard. After every successful reception the counter is
decremented by 1 or reset to 120 if its value was higher than 128. When the counter value
exceeds 127, the CAN controller enters the error passive state.
Bits 23:16 TEC[7:0]: Least significant byte of the 9-bit transmit error counter
The implementing part of the fault confinement mechanism of the CAN protocol.
Bits 15:7 Reserved, must be kept at reset value.
Bits 6:4 LEC[2:0]: Last error code
This field is set by hardware and holds a code which indicates the error condition of the last
error detected on the CAN bus. If a message has been transferred (reception or
transmission) without error, this field will be cleared to ‘0’.
The LEC[2:0] bits can be set to value 0b111 by software. They are updated by hardware to
indicate the current communication status.
000: No Error
001: Stuff Error
010: Form Error
011: Acknowledgment Error
100: Bit recessive Error
101: Bit dominant Error
110: CRC Error
111: Set by software
Bit 3 Reserved, must be kept at reset value.

1104/1747

RM0090 Rev 17

RM0090

Controller area network (bxCAN)

Bit 2 BOFF: Bus-off flag
This bit is set by hardware when it enters the bus-off state. The bus-off state is entered on
TEC overflow, greater than 255, refer to Section 32.7.6: Error management.
Bit 1 EPVF: Error passive flag
This bit is set by hardware when the Error Passive limit has been reached (Receive Error
Counter or Transmit Error Counter>127).
Bit 0 EWGF: Error warning flag
This bit is set by hardware when the warning limit has been reached
(Receive Error Counter or Transmit Error Counter≥96).

CAN bit timing register (CAN_BTR)
Address offset: 0x1C
Reset value: 0x0123 0000
This register can only be accessed by the software when the CAN hardware is in
initialization mode.
31

30

SILM

LBKM

rw

rw

15

14

29

28

27

26

12

Reserved

11

24

SJW[1:0]

Reserved
13

25

10

rw

rw

9

8

23

22

Res.

7

21

20

19

TS2[2:0]

18

17

16

TS1[3:0]

rw

rw

rw

rw

rw

rw

rw

6

5

4

3

2

1

0

rw

rw

rw

rw

BRP[9:0]
rw

rw

rw

rw

rw

rw

Bit 31 SILM: Silent mode (debug)
0: Normal operation
1: Silent Mode
Bit 30 LBKM: Loop back mode (debug)
0: Loop Back Mode disabled
1: Loop Back Mode enabled
Bits 29:26 Reserved, must be kept at reset value.
Bits 25:24 SJW[1:0]: Resynchronization jump width
These bits define the maximum number of time quanta the CAN hardware is allowed to
lengthen or shorten a bit to perform the resynchronization.
tRJW = tq x (SJW[1:0] + 1)
Bit 23 Reserved, must be kept at reset value.
Bits 22:20 TS2[2:0]: Time segment 2
These bits define the number of time quanta in Time Segment 2.
tBS2 = tq x (TS2[2:0] + 1)

RM0090 Rev 17

1105/1747
1120

Controller area network (bxCAN)

RM0090

Bits 19:16 TS1[3:0]: Time segment 1
These bits define the number of time quanta in Time Segment 1
tBS1 = tq x (TS1[3:0] + 1)
For more information on bit timing refer to Section 32.7.7: Bit timing.
Bits 15:10 Reserved, must be kept at reset value.
Bits 9:0 BRP[9:0]: Baud rate prescaler
These bits define the length of a time quanta.
tq = (BRP[9:0]+1) x tPCLK

32.9.3

CAN mailbox registers
This chapter describes the registers of the transmit and receive mailboxes. Refer to
Section 32.7.5: Message storage for detailed register mapping.
Transmit and receive mailboxes have the same registers except:
•

The FMI field in the CAN_RDTxR register.

•

A receive mailbox is always write protected.

•

A transmit mailbox is write-enabled only while empty, corresponding TME bit in the
CAN_TSR register set.

There are 3 TX Mailboxes and 2 RX Mailboxes. Each RX Mailbox allows access to a 3 level
depth FIFO, the access being offered only to the oldest received message in the FIFO.
Each mailbox consist of 4 registers.
Figure 349. RX and TX mailboxes

CAN_RI0R

CAN_RI1R

CAN_TI0R

CAN_TI1R

CAN_TI2R

CAN_RDT0R

CAN_RDT1R

CAN_TDT0R

CAN_TDT1R

CAN_TDT2R

CAN_RL0R

CAN_RL1R

CAN_TDL0R

CAN_TDL1R

CAN_TDL2R

CAN_RH0R

CAN_RH1R

CAN_TDH0R

CAN_TDH1R

CAN_TDH2R

FIFO0

1106/1747

FIFO1

RM0090 Rev 17

Three Tx Mailboxes

RM0090

Controller area network (bxCAN)

CAN TX mailbox identifier register (CAN_TIxR) (x=0..2)
Address offsets: 0x180, 0x190, 0x1A0
Reset value: 0xXXXX XXXX (except bit 0, TXRQ = 0)
All TX registers are write protected when the mailbox is pending transmission (TMEx reset).
This register also implements the TX request control (bit 0) - reset value 0.
31

30

29

28

27

26

25

24

23

22

21

20

19

STID[10:0]/EXID[28:18]
rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

rw

rw

rw

rw

rw

rw

rw

9

8

7

6

5

4

3

EXID[12:0]
rw

rw

rw

rw

rw

rw

rw

18

17

16

rw

rw

EXID[17:13]

rw

rw

rw

rw

rw

rw

rw
2

1

0

IDE

RTR

TXRQ

rw

rw

rw

Bits 31:21 STID[10:0]/EXID[28:18]: Standard identifier or extended identifier
The standard identifier or the MSBs of the extended identifier (depending on the IDE bit
value).
Bits 20:3 EXID[17:0]: Extended identifier
The LSBs of the extended identifier.
Bit 2 IDE: Identifier extension
This bit defines the identifier type of message in the mailbox.
0: Standard identifier.
1: Extended identifier.
Bit 1 RTR: Remote transmission request
0: Data frame
1: Remote frame
Bit 0 TXRQ: Transmit mailbox request
Set by software to request the transmission for the corresponding mailbox.
Cleared by hardware when the mailbox becomes empty.

RM0090 Rev 17

1107/1747
1120

Controller area network (bxCAN)

RM0090

CAN mailbox data length control and time stamp register (CAN_TDTxR)
(x=0..2)
All bits of this register are write protected when the mailbox is not in empty state.
Address offsets: 0x184, 0x194, 0x1A4
Reset value: 0xXXXX XXXX
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

TIME[15:0]
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

Reserved

TGT
rw

Reserved

DLC[3:0]
rw

rw

Bits 31:16 TIME[15:0]: Message time stamp
This field contains the 16-bit timer value captured at the SOF transmission.
Bits 15:9 Reserved, must be kept at reset value.
Bit 8 TGT: Transmit global time
This bit is active only when the hardware is in the Time Trigger Communication mode,
TTCM bit of the CAN_MCR register is set.
0: Time stamp TIME[15:0] is not sent.
1: Time stamp TIME[15:0] value is sent in the last two data bytes of the 8-byte message:
TIME[7:0] in data byte 7 and TIME[15:8] in data byte 6, replacing the data written in
CAN_TDHxR[31:16] register (DATA6[7:0] and DATA7[7:0]). DLC must be programmed as 8
in order these two bytes to be sent over the CAN bus.
Bits 7:4 Reserved, must be kept at reset value.
Bits 3:0 DLC[3:0]: Data length code
This field defines the number of data bytes a data frame contains or a remote frame request.
A message can contain from 0 to 8 data bytes, depending on the value in the DLC field.

1108/1747

RM0090 Rev 17

RM0090

Controller area network (bxCAN)

CAN mailbox data low register (CAN_TDLxR) (x=0..2)
All bits of this register are write protected when the mailbox is not in empty state.
Address offsets: 0x188, 0x198, 0x1A8
Reset value: 0xXXXX XXXX
31

30

29

28

rw

rw

rw

rw

15

14

13

12

27

26

25

24

23

22

21

20

rw

rw

rw

rw

rw

rw

rw

rw

11

10

9

8

7

6

5

4

DATA3[7:0]

rw

rw

rw

rw

18

17

16

rw

rw

rw

rw

3

2

1

0

rw

rw

rw

18

17

16

DATA2[7:0]

DATA1[7:0]
rw

19

DATA0[7:0]
rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:24 DATA3[7:0]: Data byte 3
Data byte 3 of the message.
Bits 23:16 DATA2[7:0]: Data byte 2
Data byte 2 of the message.
Bits 15:8 DATA1[7:0]: Data byte 1

Data byte 1 of the message.
Bits 7:0 DATA0[7:0]: Data byte 0
Data byte 0 of the message.
A message can contain from 0 to 8 data bytes and starts with byte 0.

CAN mailbox data high register (CAN_TDHxR) (x=0..2)
All bits of this register are write protected when the mailbox is not in empty state.
Address offsets: 0x18C, 0x19C, 0x1AC
Reset value: 0xXXXX XXXX
31

30

29

28

rw

rw

rw

rw

15

14

13

12

27

26

25

24

23

22

21

20

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

11

10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

DATA7[7:0]

DATA6[7:0]

DATA5[7:0]
rw

rw

rw

rw

rw

19

DATA4[7:0]
rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:24 DATA7[7:0]: Data byte 7
Data byte 7 of the message.
Note: If TGT of this message and TTCM are active, DATA7 and DATA6 will be replaced by
the TIME stamp value.
Bits 23:16 DATA6[7:0]: Data byte 6
Data byte 6 of the message.
Bits 15:8 DATA5[7:0]: Data byte 5

Data byte 5 of the message.
Bits 7:0 DATA4[7:0]: Data byte 4
Data byte 4 of the message.

RM0090 Rev 17

1109/1747
1120

Controller area network (bxCAN)

RM0090

CAN receive FIFO mailbox identifier register (CAN_RIxR) (x=0..1)
Address offsets: 0x1B0, 0x1C0
Reset value: 0xXXXX XXXX
All RX registers are write protected.
31

30

29

28

27

26

r

r

r

r

r

r

15

14

13

12

11

10

25

24

23

22

21

20

19

r

r

r

r

r

r

r

r

r

r

9

8

7

6

5

4

3

2

1

0

IDE

RTR

r

r

STID[10:0]/EXID[28:18]

r

r

r

r

r

r

17

16

EXID[17:13]

EXID[12:0]
r

18

r

r

r

r

r

r

Res.

Bits 31:21 STID[10:0]/EXID[28:18]: Standard identifier or extended identifier
The standard identifier or the MSBs of the extended identifier (depending on the IDE bit
value).
Bits 20:3 EXID[17:0]: Extended identifier
The LSBs of the extended identifier.
Bit 2 IDE: Identifier extension
This bit defines the identifier type of message in the mailbox.
0: Standard identifier.
1: Extended identifier.
Bit 1 RTR: Remote transmission request
0: Data frame
1: Remote frame
Bit 0 Reserved, must be kept at reset value.

1110/1747

RM0090 Rev 17

RM0090

Controller area network (bxCAN)

CAN receive FIFO mailbox data length control and time stamp register
(CAN_RDTxR) (x=0..1)
Address offsets: 0x1B4, 0x1C4
Reset value: 0xXXXX XXXX
All RX registers are write protected.
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

TIME[15:0]
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

FMI[7:0]
r

Reserved

DLC[3:0]
r

r

Bits 31:16 TIME[15:0]: Message time stamp
This field contains the 16-bit timer value captured at the SOF detection.
Bits 15:8 FMI[7:0]: Filter match index
This register contains the index of the filter the message stored in the mailbox passed
through. For more details on identifier filtering refer to Section 32.7.4: Identifier filtering Filter Match Index paragraph.
Bits 7:4 Reserved, must be kept at reset value.
Bits 3:0 DLC[3:0]: Data length code
This field defines the number of data bytes a data frame contains (0 to 8). It is 0 in the case
of a remote frame request.

RM0090 Rev 17

1111/1747
1120

Controller area network (bxCAN)

RM0090

CAN receive FIFO mailbox data low register (CAN_RDLxR) (x=0..1)
All bits of this register are write protected when the mailbox is not in empty state.
Address offsets: 0x1B8, 0x1C8
Reset value: 0xXXXX XXXX
All RX registers are write protected.
31

30

29

28

27

26

25

24

23

22

21

DATA3[7:0]
r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

DATA1[7:0]
r

r

r

r

19

18

17

16

DATA2[7:0]

r

r

20

r

r

r

r

r

4

3

2

1

0

r

r

r

17

16

DATA0[7:0]
r

r

r

r

r

r

r

r

Bits 31:24 DATA3[7:0]: Data Byte 3
Data byte 3 of the message.
Bits 23:16 DATA2[7:0]: Data Byte 2
Data byte 2 of the message.
Bits 15:8 DATA1[7:0]: Data Byte 1

Data byte 1 of the message.
Bits 7:0 DATA0[7:0]: Data Byte 0
Data byte 0 of the message.
A message can contain from 0 to 8 data bytes and starts with byte 0.

CAN receive FIFO mailbox data high register (CAN_RDHxR) (x=0..1)
Address offsets: 0x1BC, 0x1CC
Reset value: 0xXXXX XXXX
All RX registers are write protected.
31

30

29

28

27

26

25

24

23

22

21

DATA7[7:0]

20

19

18

DATA6[7:0]

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

r

r

DATA5[7:0]
r

r

DATA4[7:0]

Bits 31:24 DATA7[7:0]: Data Byte 7
Data byte 3 of the message.
Bits 23:16 DATA6[7:0]: Data Byte 6
Data byte 2 of the message.
Bits 15:8 DATA5[7:0]: Data Byte 5

Data byte 1 of the message.
Bits 7:0 DATA4[7:0]: Data Byte 4
Data byte 0 of the message.

1112/1747

RM0090 Rev 17

r

r

RM0090

Controller area network (bxCAN)

32.9.4

CAN filter registers
CAN filter master register (CAN_FMR)
Address offset: 0x200
Reset value: 0x2A1C 0E01
All bits of this register are set and cleared by software.

31

30

29

28

27

26

25

24

15

14

13

12

11

10

9

8

23

22

21

20

19

18

17

6

5

4

3

2

1

16

Reserved

Reserved

7

CAN2SB[5:0]
rw

rw

rw

rw

rw

rw

0
FINIT

Reserved

rw

Bits 31:14 Reserved, must be kept at reset value.
Bits 13:8 CAN2SB[5:0]: CAN2 start bank
These bits are set and cleared by software. They define the start bank for the CAN2
interface (Slave) in the range 0 to 27.
Note: When CAN2SB[5:0] = 28d, all the filters to CAN1 can be used.
When CAN2SB[5:0] is set to 0, no filters are assigned to CAN1.
Bits 7:1 Reserved, must be kept at reset value.
Bit 0 FINIT: Filter init mode
Initialization mode for filter banks
0: Active filters mode.
1: Initialization mode for the filters.

RM0090 Rev 17

1113/1747
1120

Controller area network (bxCAN)

RM0090

CAN filter mode register (CAN_FM1R)
Address offset: 0x204
Reset value: 0x0000 0000
This register can be written only when the filter initialization mode is set (FINIT=1) in the
CAN_FMR register.
31

30

29

28

14

13

26

12

rw

rw

Note:

24

23

22

21

20

19

18

17

16

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

11

10

9

8

7

6

5

4

3

2

1

0

FBM9

FBM8

FBM7

FBM6

FBM5

FBM4

FBM3

FBM2

FBM1

FBM0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

FBM15 FBM14 FBM13 FBM12 FBM11 FBM10
rw

25

FBM27 FBM26 FBM25 FBM24 FBM23 FBM22 FBM21 FBM20 FBM19 FBM18 FBM17 FBM16

Reserved
15

27

rw

rw

rw

rw

Refer to Figure 342.
Bits 31:28

Reserved, must be kept at reset value.

Bits 27:0 FBMx: Filter mode
Mode of the registers of Filter x.
0: Two 32-bit registers of filter bank x are in Identifier Mask mode.
1: Two 32-bit registers of filter bank x are in Identifier List mode.

CAN filter scale register (CAN_FS1R)
Address offset: 0x20C
Reset value: 0x0000 0000
This register can be written only when the filter initialization mode is set (FINIT=1) in the
CAN_FMR register.
31

30

29

28

Reserved

27

26

25

24

23

22

21

20

19

18

17

16

FSC27

FSC26

FSC25

FSC24

FSC23

FSC22

FSC21

FSC20

FSC19

FSC18

FSC17

FSC16
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

FSC15

FSC14

FSC13

FSC12

FSC11

FSC10

FSC9

FSC8

FSC7

FSC6

FSC5

FSC4

FSC3

FSC2

FSC1

FSC0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:28 Reserved, must be kept at reset value.
Bits 27:0 FSCx: Filter scale configuration
These bits define the scale configuration of Filters 13-0.
0: Dual 16-bit scale configuration
1: Single 32-bit scale configuration

Note:

1114/1747

Refer to Figure 342.

RM0090 Rev 17

RM0090

Controller area network (bxCAN)

CAN filter FIFO assignment register (CAN_FFA1R)
Address offset: 0x214
Reset value: 0x0000 0000
This register can be written only when the filter initialization mode is set (FINIT=1) in the
CAN_FMR register.
31

30

29

28

Reserved

27

26

25

24

23

22

21

20

19

18

17

16

FFA27

FFA26

FFA25

FFA24

FFA23

FFA22

FFA21

FFA20

FFA19

FFA18

FFA17

FFA16

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

FFA15

FFA14

FFA13

FFA12

FFA11

FFA10

FFA9

FFA8

FFA7

FFA6

FFA5

FFA4

FFA3

FFA2

FFA1

FFA0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

17

16

Bits 31:28 Reserved, must be kept at reset value.
Bits 27:0 FFAx: Filter FIFO assignment for filter x
The message passing through this filter will be stored in the specified FIFO.
0: Filter assigned to FIFO 0
1: Filter assigned to FIFO 1

CAN filter activation register (CAN_FA1R)
Address offset: 0x21C
Reset value: 0x0000 0000
31

30

29

28

Reserved
15

14

13

27

26

25

12

rw

rw

23

22

21

20

19

18

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

11

10

9

8

7

6

5

4

3

2

1

0

FACT8

FACT7

FACT6

FACT5

FACT4

FACT3

FACT2

FACT1

FACT0

rw

rw

rw

rw

rw

rw

rw

rw

rw

FACT15 FACT14 FACT13 FACT12 FACT11 FACT10 FACT9
rw

24

FACT27 FACT26 FACT25 FACT24 FACT23 FACT22 FACT21 FACT20 FACT19 FACT18 FACT17 FACT16

rw

rw

rw

rw

rw

Bits 31:28 Reserved, must be kept at reset value.
Bits 27:0 FACTx: Filter active
The software sets this bit to activate Filter x. To modify the Filter x registers (CAN_FxR[0:7]),
the FACTx bit must be cleared or the FINIT bit of the CAN_FMR register must be set.
0: Filter x is not active
1: Filter x is active

RM0090 Rev 17

1115/1747
1120

Controller area network (bxCAN)

RM0090

Filter bank i register x (CAN_FiRx) (i=0..27, x=1, 2)
Address offsets: 0x240..0x31C
Reset value: 0xXXXX XXXX
There are 28 filter banks, i=0 .. 27. Each filter bank i is composed of two 32-bit registers,
CAN_FiR[2:1].
This register can only be modified when the FACTx bit of the CAN_FAxR register is cleared
or when the FINIT bit of the CAN_FMR register is set.
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

FB31

FB30

FB29

FB28

FB27

FB26

FB25

FB24

FB23

FB22

FB21

FB20

FB19

FB18

FB17

FB16

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

FB15

FB14

FB13

FB12

FB11

FB10

FB9

FB8

FB7

FB6

FB5

FB4

FB3

FB2

FB1

FB0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

In all configurations:
Bits 31:0 FB[31:0]: Filter bits
Identifier
Each bit of the register specifies the level of the corresponding bit of the expected identifier.
0: Dominant bit is expected
1: Recessive bit is expected
Mask
Each bit of the register specifies whether the bit of the associated identifier register must
match with the corresponding bit of the expected identifier or not.
0: Don’t care, the bit is not used for the comparison
1: Must match, the bit of the incoming identifier must have the same level has specified in
the corresponding identifier register of the filter.

Note:

Depending on the scale and mode configuration of the filter the function of each register can
differ. For the filter mapping, functions description and mask registers association, refer to
Section 32.7.4: Identifier filtering on page 1086.
A Mask/Identifier register in mask mode has the same bit mapping as in identifier list
mode.
For the register mapping/addresses of the filter banks refer to Table 184.

1116/1747

RM0090 Rev 17

RM0090

32.9.5

Controller area network (bxCAN)

bxCAN register map
Refer to Section 2.3: Memory map for the register boundary addresses. The registers from
offset 0x200 to 31C are present only in CAN1.

0

Reserved

0

0

0

0

0

0

TS2[2:0]
0

1

0

0x0200x17F
CAN_TI0R

0

1

INRQ
INAK
RQCP0

TXFP

SLEEP

ERRI

FMP0[1:0]

ALST0

FMP1[1:0]

0

0

0

0

0

LEC[2:0]

EPVIE

EWGIE

Reserved

LECIE

0

TS1[3:0]
0

BOFIE

0

0

0

0

0

0

0

0

0

x

x

0

BRP[9:0]

Reserved
1

STID[10:0]/EXID[28:18]
x

x

x

x

x

x

CAN_TDT0R
Reset value

0x188

0

0

0

0

0

0

0

0

0

x

x

x

x

x

x

x

Reserved

Reset value

0x184

0

TXOK0

FULL1

0

Reserved

x

x

x

x

x

x

x

x

x

x

x

TIME[15:0]
x

x

CAN_TDL0R
Reset value

x

EXID[17:0]

x

x

x

x

x

x

x

x

DATA3[7:0]
x

x

x

x

x

x

x

x

x

x

x

x

Reserved
x

x

x

x

x

x

x

x

x

x

x

DATA1[7:0]
x

RM0090 Rev 17

x

DLC[3:0]

Reserved

x

DATA2[7:0]
x

x

TGT

0x180

0

0

0
TMEIE

0

0

0

EWGF

0

0

0

0

TXRQ

0

0

0

Reserved

FOVR1

0

Reserved

RFOM1

0

0

FFIE0

Reset value

0

0

0

FMPIE0

CAN_BTR

0

0

0

EPVF

0

Reserved

0x01C

0

0

Res.

TEC[7:0]

SJW[1:0]

Reset value

REC[7:0]

SILM

CAN_ESR

LBKM

0x018

ERRIE

Reset value

WKUIE

Reserved

SLKIE

CAN_IER

0

0

Reserved

Reset value

0x014

0

FOVIE1

CAN_RF1R

0

BOFF

Reserved

Reset value

0x010

0

0

IDE

0

RTR

0

SLAK

NART

RFLM

0

WKUI

AWUM

0

0

Res..

TERR0

0

0

FULL0

0

1

FOVIE0

0

0

Reserved

0

0

FOVR0

CAN_RF0R

0

0

RFOM0

0

0

SLAKI

0

1

FFIE1

0

TTCM
0

Res.

0

FMPIE1

0

Res.

ABOM

TXM

1

TERR1

1

ABRQ1

1

RQCP2

1

ABRQ2

CODE[1:0]

TME[2:0]
0

ALST2

0

TXOK2

0

Res.

TERR2

Reset value

0x00C

LOW[2:0]

CAN_TSR

0

1

Reset value

0x008

0

ABRQ0

Reserved

0

RXM

CAN_MSR

0

TXOK1

0x004

0

RQCP1

0
RX

1

Reset value

Reserved

SAMP

Reserved

ALST1

CAN_MCR

DBF

0x000

Register

RESET

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 184. bxCAN register map and reset values

x

x

x

x

x

x

x

x

x

x

x

DATA0[7:0]
x

x

x

x

x

x

x

x

1117/1747
1120

Controller area network (bxCAN)

RM0090

0x1B4

0x1C0

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

DATA7[7:0]
x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

DATA7[7:0]
x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

DATA7[7:0]
x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

RM0090 Rev 17

x

x

x

x

x

x

x

0

x

x

x

x

x

x

x

x

x

x

0

x

x

x

x

x

x

x

x

x

x

x

x

DATA4[7:0]
x

x

x

x

x

x

x

x

x

x

x

x

x

x

DLC[3:0]

x

x

x

x

x

x

x

x

x

x

x

x

x

DATA0[7:0]
x

x

x

x

x

x

x

x

DATA4[7:0]
x

x

x

x

x

x

x

x

x

x

x

x

x

x

EXID[17:0]
x

x

DATA0[7:0]

DATA5[7:0]
x

x

DLC[3:0]

Reserved

DATA1[7:0]
x

x

DATA4[7:0]

Reserved
x

x

DATA0[7:0]

FMI[7:0]
x

x

DLC[3:0]

Reserved

DATA5[7:0]

STID[10:0]/EXID[28:18]
x

x

DATA1[7:0]

DATA6[7:0]
x

x

x

DATA2[7:0]
x

x

EXID[17:0]

DATA3[7:0]
x

x

x

TIME[15:0]
x

x

Reserved

STID[10:0]/EXID[28:18]
x

x

DATA5[7:0]

DATA6[7:0]
x

x

DATA1[7:0]

DATA2[7:0]
x

x

EXID[17:0]

DATA3[7:0]
x

x

x

TIME[15:0]
x

x

x

STID[10:0]/EXID[28:18]
x

x

Reserved

DATA6[7:0]
x

x

EXID[17:0]

DATA2[7:0]
x

x

TXRQ

x

TXRQ

x

Reserved

x

IDE

x

RTR

x

IDE

x

RTR

x

CAN_RI1R
Reset value

1118/1747

x

CAN_RDH0R
Reset value

x

DATA3[7:0]

CAN_RDL0R
Reset value

0x1BC

x

CAN_RDT0R
Reset value

0x1B8

x

CAN_RI0R
Reset value

x

Reserved

0x1B0

x

CAN_TDH2R
Reset value

x

TIME[15:0]

CAN_TDL2R
Reset value

0x1AC

x

CAN_TDT2R
Reset value

0x1A8

x

CAN_TI2R
Reset value

0x1A4

x

CAN_TDH1R
Reset value

0x1A0

x

STID[10:0]/EXID[28:18]

CAN_TDL1R
Reset value

0x19C

x

CAN_TDT1R
Reset value

0x198

x

DATA4[7:0]

IDE

CAN_TI1R
Reset value

0x194

x

DATA5[7:0]

RTR

0x190

x

DATA6[7:0]

IDE

Reset value

DATA7[7:0]

RTR

CAN_TDH0R

TGT

0x18C

Register

TGT

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 184. bxCAN register map and reset values (continued)

RM0090

Controller area network (bxCAN)

0x1C4

Register
CAN_RDT1R
Reset value

0x1C8

x

x

CAN_RDL1R
Reset value

0x1CC

TIME[15:0]
x

x

x

x

x

x

x

DATA3[7:0]
x

x

CAN_RDH1R
Reset value

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

CAN_FMR

CAN_FM1R

x

x

0

0

0

0

0

0

0

0

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

DATA0[7:0]
x

x

x

x

x

x

x

x

DATA4[7:0]
x

x

x

x

x

x

x

x

0

Reserved

1

1

1

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Reserved
CAN_FS1R

FSC[27:0]

Reserved
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Reserved
CAN_FFA1R

FFA[27:0]

Reserved
0

0

0

0

0

0

0

0

0x218

0

0

0

0

0

0

0

Reserved
CAN_FA1R

FACT[27:0]

Reserved

Reset value

0

0

0

0

0

0

0

0

0

0

0

0

0x220

Reserved

0x2240x23F

Reserved
CAN_F0R1
Reset value

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

FB[31:0]
x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

CAN_F1R1
Reset value

0

FB[31:0]

CAN_F0R2
Reset value

0x248

x

FBM[27:0]

Reserved

Reset value

0x244

x

CAN2SB[5:0]

0x210

0x240

x

0

Reset value

0x21C

x

Reserved

0x208

0x214

x

Reserved

Reset value

0x20C

x

DATA5[7:0]

Reset value
0x204

x

DLC[3:0]

Reserved

DATA1[7:0]

DATA6[7:0]

0x1D00x1FF

0x200

x

DATA2[7:0]

DATA7[7:0]
x

FMI[7:0]

FINIT

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 184. bxCAN register map and reset values (continued)

x

x

FB[31:0]
x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

RM0090 Rev 17

x

x

1119/1747
1120

Controller area network (bxCAN)

RM0090

Offset
0x24C

Register

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 184. bxCAN register map and reset values (continued)

CAN_F1R2

FB[31:0]

Reset value
.
.
.
.
0x318

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

CAN_F27R1

FB[31:0]
x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

CAN_F27R2
Reset value

1120/1747

x

.
.
.
.

Reset value
0x31C

x

.
.
.
.

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

FB[31:0]
x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

RM0090 Rev 17

x

x

RM0090

33

Ethernet (ETH): media access control (MAC) with DMA controller

Ethernet (ETH): media access control (MAC) with
DMA controller
This section applies only to STM32F42xx/F43xx and STM32F407x/F417x devices.

33.1

Ethernet introduction
Portions Copyright (c) 2004, 2005 Synopsys, Inc. All rights reserved. Used with permission.
The Ethernet peripheral enables the STM32F4xx to transmit and receive data over Ethernet
in compliance with the IEEE 802.3-2002 standard.
The Ethernet provides a configurable, flexible peripheral to meet the needs of various
applications and customers. It supports two industry standard interfaces to the external
physical layer (PHY): the default media independent interface (MII) defined in the IEEE
802.3 specifications and the reduced media independent interface (RMII). It can be used in
number of applications such as switches, network interface cards, etc.
The Ethernet is compliant with the following standards:

33.2

•

IEEE 802.3-2002 for Ethernet MAC

•

IEEE 1588-2008 standard for precision networked clock synchronization

•

AMBA 2.0 for AHB Master/Slave ports

•

RMII specification from RMII consortium

Ethernet main features
The Ethernet (ETH) peripheral includes the following features, listed by category:

RM0090 Rev 17

1121/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

33.2.1

1122/1747

RM0090

MAC core features
•

Supports 10/100 Mbit/s data transfer rates with external PHY interfaces

•

IEEE 802.3-compliant MII interface to communicate with an external Fast Ethernet
PHY

•

Supports both full-duplex and half-duplex operations
–

Supports CSMA/CD Protocol for half-duplex operation

–

Supports IEEE 802.3x flow control for full-duplex operation

–

Optional forwarding of received pause control frames to the user application in fullduplex operation

–

Back-pressure support for half-duplex operation

–

Automatic transmission of zero-quanta pause frame on deassertion of flow control
input in full-duplex operation

•

Preamble and start-of-frame data (SFD) insertion in Transmit, and deletion in Receive
paths

•

Automatic CRC and pad generation controllable on a per-frame basis

•

Options for automatic pad/CRC stripping on receive frames

•

Programmable frame length to support Standard frames with sizes up to 16 KB

•

Programmable interframe gap (40-96 bit times in steps of 8)

•

Supports a variety of flexible address filtering modes:
–

Up to four 48-bit perfect (DA) address filters with masks for each byte

–

Up to three 48-bit SA address comparison check with masks for each byte

–

64-bit Hash filter (optional) for multicast and unicast (DA) addresses

–

Option to pass all multicast addressed frames

–

Promiscuous mode support to pass all frames without any filtering for network
monitoring

–

Passes all incoming packets (as per filter) with a status report

•

Separate 32-bit status returned for transmission and reception packets

•

Supports IEEE 802.1Q VLAN tag detection for reception frames

•

Separate transmission, reception, and control interfaces to the Application

•

Supports mandatory network statistics with RMON/MIB counters (RFC2819/RFC2665)

•

MDIO interface for PHY device configuration and management

•

Detection of LAN wakeup frames and AMD Magic Packet™ frames

•

Receive feature for checksum off-load for received IPv4 and TCP packets
encapsulated by the Ethernet frame

•

Enhanced receive feature for checking IPv4 header checksum and TCP, UDP, or ICMP
checksum encapsulated in IPv4 or IPv6 datagrams

•

Support Ethernet frame time stamping as described in IEEE 1588-2008. Sixty-four-bit
time stamps are given in each frame’s transmit or receive status

•

Two sets of FIFOs: a 2-KB Transmit FIFO with programmable threshold capability, and
a 2-KB Receive FIFO with a configurable threshold (default of 64 bytes)

•

Receive Status vectors inserted into the Receive FIFO after the EOF transfer enables
multiple-frame storage in the Receive FIFO without requiring another FIFO to store
those frames’ Receive Status

•

Option to filter all error frames on reception and not forward them to the application in
RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
Store-and-Forward mode

33.2.2

33.2.3

•

Option to forward under-sized good frames

•

Supports statistics by generating pulses for frames dropped or corrupted (due to
overflow) in the Receive FIFO

•

Supports Store and Forward mechanism for transmission to the MAC core

•

Automatic generation of PAUSE frame control or back pressure signal to the MAC core
based on Receive FIFO-fill (threshold configurable) level

•

Handles automatic retransmission of Collision frames for transmission

•

Discards frames on late collision, excessive collisions, excessive deferral and underrun
conditions

•

Software control to flush Tx FIFO

•

Calculates and inserts IPv4 header checksum and TCP, UDP, or ICMP checksum in
frames transmitted in Store-and-Forward mode

•

Supports internal loopback on the MII for debugging

DMA features
•

Supports all AHB burst types in the AHB Slave Interface

•

Software can select the type of AHB burst (fixed or indefinite burst) in the AHB Master
interface.

•

Option to select address-aligned bursts from AHB master port

•

Optimization for packet-oriented DMA transfers with frame delimiters

•

Byte-aligned addressing for data buffer support

•

Dual-buffer (ring) or linked-list (chained) descriptor chaining

•

Descriptor architecture, allowing large blocks of data transfer with minimum CPU
intervention;

•

each descriptor can transfer up to 8 KB of data

•

Comprehensive status reporting for normal operation and transfers with errors

•

Individual programmable burst size for Transmit and Receive DMA Engines for optimal
host bus utilization

•

Programmable interrupt options for different operational conditions

•

Per-frame Transmit/Receive complete interrupt control

•

Round-robin or fixed-priority arbitration between Receive and Transmit engines

•

Start/Stop modes

•

Current Tx/Rx Buffer pointer as status registers

•

Current Tx/Rx Descriptor pointer as status registers

PTP features
•

Received and transmitted frames time stamping

•

Coarse and fine correction methods

•

Trigger interrupt when system time becomes greater than target time

•

Pulse per second output (product alternate function output)

RM0090 Rev 17

1123/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

33.3

RM0090

Ethernet pins
Table 185 shows the MAC signals and the corresponding MII/RMII signal mapping. All MAC
signals are mapped onto AF11, some signals are mapped onto different I/O pins, and
should be configured in Alternate function mode (for more details, refer to Section 8.3.2: I/O
pin multiplexer and mapping).
Table 185. Alternate function mapping
AF11
Port
ETH

1124/1747

PA0-WKUP

ETH_MII_CRS

PA1

ETH_MII _RX_CLK / ETH_RMII _REF_CLK

PA2

ETH _MDIO

PA3

ETH _MII_COL

PA7

ETH_MII _RX_DV / ETH_RMII _CRS_DV

PB0

ETH _MII_RXD2

PB1

ETH _MII_RXD3

PB5

ETH _PPS_OUT

PB8

ETH _MII_TXD3

PB10

ETH_ MII_RX_ER

PB11

ETH _MII_TX_EN / ETH _RMII_TX_EN

PB12

ETH _MII_TXD0 / ETH _RMII_TXD0

PB13

ETH _MII_TXD1 / ETH _RMII_TXD1

PC1

ETH _MDC

PC2

ETH _MII_TXD2

PC3

ETH _MII_TX_CLK

PC4

ETH_MII_RXD0 / ETH_RMII_RXD0

PC5

ETH _MII_RXD1/ ETH _RMII_RXD1

PE2

ETH_MII_TXD3

PG8

ETH_PPS_OUT

PG11

ETH _MII_TX_EN / ETH _RMII_TX_EN

PG13

ETH _MII_TXD0 / ETH _RMII_TXD0

PG14

ETH _MII_TXD1 / ETH _RMII_TXD1

PH2

ETH _MII_CRS

PH3

ETH _MII_COL

PH6

ETH _MII_RXD2

PH7

ETH _MII_RXD3

PI10

ETH _MII_RX_ER

RM0090 Rev 17

RM0090

33.4

Ethernet (ETH): media access control (MAC) with DMA controller

Ethernet functional description: SMI, MII and RMII
The Ethernet peripheral consists of a MAC 802.3 (media access control) with a dedicated
DMA controller. It supports both default media-independent interface (MII) and reduced
media-independent interface (RMII) through one selection bit (refer to SYSCFG_PMC
register).
The DMA controller interfaces with the Core and memories through the AHB Master and
Slave interfaces. The AHB Master Interface controls data transfers while the AHB Slave
interface accesses Control and Status Registers (CSR) space.
The Transmit FIFO (Tx FIFO) buffers data read from system memory by the DMA before
transmission by the MAC Core. Similarly, the Receive FIFO (Rx FIFO) stores the Ethernet
frames received from the line until they are transferred to system memory by the DMA.
The Ethernet peripheral also includes an SMI to communicate with external PHY. A set of
configuration registers permit the user to select the desired mode and features for the MAC
and the DMA controller.

Note:

The AHB clock frequency must be at least 25 MHz when the Ethernet is used.

"US MATRIX

!(" 3LAVE INTERFACE

Figure 350. ETH block diagram

$-!
CONTROL 
STATUS
REGISTERS

/PERATION
MODE
REGISTER

-EDIA ACCESS
CONTROL
-!# 

)NTERFACE
3ELECT

-!#
CONTROL
REGISTERS

 +BYTE
28 &)&/
%THERNET
$-!

2-))

 +BYTE
48 &)&/

#HECKSUM
040
OFFLOAD
)%%%
0-4

--#

-))

%XTERNAL 0(9

-$#
-$)/

AIC

1. For AHB connections refer to Figure 1: System architecture for STM32F405xx/07xx and
STM32F415xx/17xx devices and Figure 2: System architecture for STM32F42xxx and STM32F43xxx
devices.

33.4.1

Station management interface: SMI
The station management interface (SMI) allows the application to access any PHY registers
through a 2-wire clock and data lines. The interface supports accessing up to 32 PHYs.
The application can select one of the 32 PHYs and one of the 32 registers within any PHY
and send control data or receive status information. Only one register in one PHY can be
addressed at any given time.
Both the MDC clock line and the MDIO data line are implemented as alternate function I/O
in the microcontroller:
•

MDC: a periodic clock that provides the timing reference for the data transfer at the
maximum frequency of 2.5 MHz. The minimum high and low times for MDC must be

RM0090 Rev 17

1125/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

160 ns each, and the minimum period for MDC must be 400 ns. In idle state the SMI
management interface drives the MDC clock signal low.
•

MDIO: data input/output bitstream to transfer status information to/from the PHY device
synchronously with the MDC clock signal

6700&8

 -!#

Figure 351. SMI interface signals


0'&

([WHUQDO
3+<

0',2

DLE

SMI frame format
The frame structure related to a read or write operation is shown in Table 13, the order of bit
transmission must be from left to right.
Table 186. Management frame format
Management frame fields
Preamble
(32 bits)

Start

Operation

PADDR

RADDR

TA

Data (16 bits)

Idle

Read

1... 1

01

10

ppppp

rrrrr

Z0

ddddddddddddddd

Z

Write

1... 1

01

01

ppppp

rrrrr

10

ddddddddddddddd

Z

The management frame consists of eight fields:

1126/1747

•

Preamble: each transaction (read or write) can be initiated with the preamble field that
corresponds to 32 contiguous logic one bits on the MDIO line with 32 corresponding
cycles on MDC. This field is used to establish synchronization with the PHY device.

•

Start: the start of frame is defined by a <01> pattern to verify transitions on the line
from the default logic one state to zero and back to one.

•

Operation: defines the type of transaction (read or write) in progress.

•

PADDR: the PHY address is 5 bits, allowing 32 unique PHY addresses. The MSB bit of
the address is the first transmitted and received.

•

RADDR: the register address is 5 bits, allowing 32 individual registers to be addressed
within the selected PHY device. The MSB bit of the address is the first transmitted and
received.

•

TA: the turn-around field defines a 2-bit pattern between the RADDR and DATA fields
to avoid contention during a read transaction. For a read transaction the MAC controller
drives high-impedance on the MDIO line for the 2 bits of TA. The PHY device must
drive a high-impedance state on the first bit of TA, a zero bit on the second one.

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
For a write transaction, the MAC controller drives a <10> pattern during the TA field.
The PHY device must drive a high-impedance state for the 2 bits of TA.
•

Data: the data field is 16-bit. The first bit transmitted and received must be bit 15 of the
ETH_MIID register.

•

Idle: the MDIO line is driven in high-impedance state. All three-state drivers must be
disabled and the PHY’s pull-up resistor keeps the line at logic one.

SMI write operation
When the application sets the MII Write and Busy bits (in Ethernet MAC MII address register
(ETH_MACMIIAR)), the SMI initiates a write operation into the PHY registers by transferring
the PHY address, the register address in PHY, and the write data (in Ethernet MAC MII data
register (ETH_MACMIIDR). The application should not change the MII Address register
contents or the MII Data register while the transaction is ongoing. Write operations to the MII
Address register or the MII Data Register during this period are ignored (the Busy bit is
high), and the transaction is completed without any error. After the Write operation has
completed, the SMI indicates this by resetting the Busy bit.
Figure 352 shows the frame format for the write operation.
Figure 352. MDIO timing and frame structure - Write cycle

-$#

-$)/

 gS

 

 

3TART
/0
0REAMBLE OF
CODE
FRAME

! ! ! ! ! 2 2

0(9 ADDRESS

2 2 2

2EGISTER ADDRESS 4URN
AROUND

$ $

$ $

DATA

$ATA TO 0(9
AI

SMI read operation
When the user sets the MII Busy bit in the Ethernet MAC MII address register
(ETH_MACMIIAR) with the MII Write bit at 0, the SMI initiates a read operation in the PHY
registers by transferring the PHY address and the register address in PHY. The application
should not change the MII Address register contents or the MII Data register while the
transaction is ongoing. Write operations to the MII Address register or MII Data Register
during this period are ignored (the Busy bit is high) and the transaction is completed without
any error. After the read operation has completed, the SMI resets the Busy bit and then
updates the MII Data register with the data read from the PHY.
Figure 353 shows the frame format for the read operation.

RM0090 Rev 17

1127/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Figure 353. MDIO timing and frame structure - Read cycle

MDC

MDIO

32 1's

0 1 1

0

Start
OP
Preamble of
code
frame

A4 A3 A2 A1 A0 R4 R3

PHY address

R2 R1 R0

Register address Turn
around

Data to PHY

D15 D14

D1 D0

data

Data from PHY

ai15627

SMI clock selection
The MAC initiates the Management Write/Read operation. The SMI clock is a divided clock
whose source is the application clock (AHB clock). The divide factor depends on the clock
range setting in the MII Address register.
Table 187 shows how to set the clock ranges.
Table 187. Clock range

33.4.2

Selection

HCLK clock

MDC clock

000

60-100 MHz

AHB clock / 42

001

100-150 MHz

AHB clock / 62

010

20-35 MHz

AHB clock / 16

011

35-60 MHz

AHB clock / 26

100

150-180 MHz

AHB clock / 102

101, 110, 111

Reserved

-

Media-independent interface: MII
The media-independent interface (MII) defines the interconnection between the MAC
sublayer and the PHY for data transfer at 10 Mbit/s and 100 Mbit/s.

1128/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
Figure 354. Media independent interface signals
48?#,+
48$;=

6700&8

 -!#

48?%.
28?#,+
28$;=
28?%2
28?$6

%XTERNAL
0(9

#23
#/,
-$#
-$)/
AIC

•

MII_TX_CLK: continuous clock that provides the timing reference for the TX data
transfer. The nominal frequency is: 2.5 MHz at 10 Mbit/s speed; 25 MHz at 100 Mbit/s
speed.

•

MII_RX_CLK: continuous clock that provides the timing reference for the RX data
transfer. The nominal frequency is: 2.5 MHz at 10 Mbit/s speed; 25 MHz at 100 Mbit/s
speed.

•

MII_TX_EN: transmission enable indicates that the MAC is presenting nibbles on the
MII for transmission. It must be asserted synchronously (MII_TX_CLK) with the first
nibble of the preamble and must remain asserted while all nibbles to be transmitted are
presented to the MII.

•

MII_TXD[3:0]: transmit data is a bundle of 4 data signals driven synchronously by the
MAC sublayer and qualified (valid data) on the assertion of the MII_TX_EN signal.
MII_TXD[0] is the least significant bit, MII_TXD[3] is the most significant bit. While
MII_TX_EN is deasserted the transmit data must have no effect upon the PHY.

•

MII_CRS: carrier sense is asserted by the PHY when either the transmit or receive
medium is non idle. It shall be deasserted by the PHY when both the transmit and
receive media are idle. The PHY must ensure that the MII_CS signal remains asserted
throughout the duration of a collision condition. This signal is not required to transition
synchronously with respect to the TX and RX clocks. In full duplex mode the state of
this signal is don’t care for the MAC sublayer.

•

MII_COL: collision detection must be asserted by the PHY upon detection of a collision
on the medium and must remain asserted while the collision condition persists. This
signal is not required to transition synchronously with respect to the TX and RX clocks.
In full duplex mode the state of this signal is don’t care for the MAC sublayer.

•

MII_RXD[3:0]: reception data is a bundle of 4 data signals driven synchronously by the
PHY and qualified (valid data) on the assertion of the MII_RX_DV signal. MII_RXD[0] is
the least significant bit, MII_RXD[3] is the most significant bit. While MII_RX_EN is
deasserted and MII_RX_ER is asserted, a specific MII_RXD[3:0] value is used to
transfer specific information from the PHY (see Table 189).

•

MII_RX_DV: receive data valid indicates that the PHY is presenting recovered and
decoded nibbles on the MII for reception. It must be asserted synchronously
(MII_RX_CLK) with the first recovered nibble of the frame and must remain asserted
through the final recovered nibble. It must be deasserted prior to the first clock cycle

RM0090 Rev 17

1129/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

that follows the final nibble. In order to receive the frame correctly, the MII_RX_DV
signal must encompass the frame, starting no later than the SFD field.
•

MII_RX_ER: receive error must be asserted for one or more clock periods
(MII_RX_CLK) to indicate to the MAC sublayer that an error was detected somewhere
in the frame. This error condition must be qualified by MII_RX_DV assertion as
described in Table 189.
Table 188. TX interface signal encoding
MII_TX_EN

MII_TXD[3:0]

Description

0

0000 through 1111

Normal inter-frame

1

0000 through 1111

Normal data transmission

Table 189. RX interface signal encoding
MII_RX_DV

MII_RX_ERR

MII_RXD[3:0]

Description

0

0

0000 through 1111

Normal inter-frame

0

1

0000

Normal inter-frame

0

1

0001 through 1101

0

1

1110

False carrier indication

0

1

1111

Reserved

1

0

0000 through 1111

Normal data reception

1

1

0000 through 1111

Data reception with errors

Reserved

MII clock sources
To generate both TX_CLK and RX_CLK clock signals, the external PHY must be clocked
with an external 25 MHz as shown in Figure 355. Instead of using an external 25 MHz
quartz to provide this clock, the STM32F4xxmicrocontroller can output this signal on its
MCO pin. In this case, the PLL multiplier has to be configured so as to get the desired
frequency on the MCO pin, from the 25 MHz external quartz.

1130/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

0+]

670
0&8

0$&

Figure 355. MII clock sources

7;B&/.
5;B&/.



+6(

([WHUQDO
3+<

0&2

0+]

)RU0ELWV

DLE

Reduced media-independent interface: RMII
The reduced media-independent interface (RMII) specification reduces the pin count
between the microcontroller Ethernet peripheral and the external Ethernet in 10/100 Mbit/s.
According to the IEEE 802.3u standard, an MII contains 16 pins for data and control. The
RMII specification is dedicated to reduce the pin count to 7 pins (a 62.5% decrease in pin
count).
The RMII is instantiated between the MAC and the PHY. This helps translation of the MAC’s
MII into the RMII. The RMII block has the following characteristics:
•

It supports 10-Mbit/s and 100-Mbit/s operating rates

•

The clock reference must be doubled to 50 MHz

•

The same clock reference must be sourced externally to both MAC and external
Ethernet PHY

•

It provides independent 2-bit wide (dibit) transmit and receive data paths
Figure 356. Reduced media-independent interface signals

7;'>@
7;B(1

670
0&8

0$&

33.4.3

5;'>@
&56B
'9

([WHUQDO
3+<

0'&
0',2
5()
B &/.
&ORFNVRXUFH
DLE

RM0090 Rev 17

1131/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

RMII clock sources
Either clock the PHY from an external 50 MHz clock or use a PHY with an embedded PLL to
generate the 50 MHz frequency.

670
0&8
0+]


0$&

Figure 357. RMII clock sources

3//

([WHUQDO
3+<

5()B&/.

0+]

)RU0ELWV
069

33.4.4

MII/RMII selection
The mode, MII or RMII, is selected using the configuration bit 23, MII_RMII_SEL, in the
SYSCFG_PMC register. The application has to set the MII/RMII mode while the Ethernet
controller is under reset or before enabling the clocks.

MII/RMII internal clock scheme
The clock scheme required to support both the MII and RMII, as well as 10 and 100 Mbit/s
operations is described in Figure 358.
Figure 358. Clock scheme
MAC
MII_TX_CLK as AF
(25 MHz or 2.5 MHz)

MII_RX_CLK as AF
(25 MHz or 2.5 MHz)
RMII_REF_CK as AF
(50 MHz)

GPIO and AF
controller

25 MHz or 2.5 MHz

50 MHz Sync. divider
/2 for 100 Mb/s
/20 for 10 Mb/s
GPIO and AF
controller

25 MHz or 2.5 MHz

0
1

25 MHz
or 2.5 MHz

MACTXCLK

TX

MACRXCLK

RX

0 MII
1 RMII(1)
0
1

25 MHz
or 2.5 MHz

AHB

RMII
HCLK
HCLK
must be greater
than 25 MHz
ai15650

1. The MII/RMII selection is controlled through bit 23, MII_RMII_SEL, in the SYSCFG_PMC register.

To save a pin, the two input clock signals, RMII_REF_CK and MII_RX_CLK, are multiplexed
on the same GPIO pin.

1132/1747

RM0090 Rev 17

RM0090

33.5

Ethernet (ETH): media access control (MAC) with DMA controller

Ethernet functional description: MAC 802.3
The IEEE 802.3 International Standard for local area networks (LANs) employs the
CSMA/CD (carrier sense multiple access with collision detection) as the access method.
The Ethernet peripheral consists of a MAC 802.3 (media access control) controller with
media independent interface (MII) and a dedicated DMA controller.
The MAC block implements the LAN CSMA/CD sublayer for the following families of
systems: 10 Mbit/s and 100 Mbit/s of data rates for baseband and broadband systems. Halfand full-duplex operation modes are supported. The collision detection access method is
applied only to the half-duplex operation mode. The MAC control frame sublayer is
supported.
The MAC sublayer performs the following functions associated with a data link control
procedure:
•

•

Data encapsulation (transmit and receive)
–

Framing (frame boundary delimitation, frame synchronization)

–

Addressing (handling of source and destination addresses)

–

Error detection

Media access management
–

Medium allocation (collision avoidance)

–

Contention resolution (collision handling)

Basically there are two operating modes of the MAC sublayer:

33.5.1

•

Half-duplex mode: the stations contend for the use of the physical medium, using the
CSMA/CD algorithms.

•

Full duplex mode: simultaneous transmission and reception without contention
resolution (CSMA/CD algorithm are unnecessary) when all the following conditions are
met:
–

physical medium capability to support simultaneous transmission and reception

–

exactly 2 stations connected to the LAN

–

both stations configured for full-duplex operation

MAC 802.3 frame format
The MAC block implements the MAC sublayer and the optional MAC control sublayer
(10/100 Mbit/s) as specified by the IEEE 802.3-2002 standard.
Two frame formats are specified for data communication systems using the CSMA/CD
MAC:
•

Basic MAC frame format

•

Tagged MAC frame format (extension of the basic MAC frame format)

RM0090 Rev 17

1133/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Figure 360 and Figure 361 describe the frame structure (untagged and tagged) that
includes the following fields:
•

Preamble: 7-byte field used for synchronization purposes (PLS circuitry)
Hexadecimal value: 55-55-55-55-55-55-55
Bit pattern: 01010101 01010101 01010101 01010101 01010101 01010101 01010101
(right-to-left bit transmission)

•

Start frame delimiter (SFD): 1-byte field used to indicate the start of a frame.
Hexadecimal value: D5
Bit pattern: 11010101 (right-to-left bit transmission)

•

Destination and Source Address fields: 6-byte fields to indicate the destination and
source station addresses as follows (see Figure 359):
–

Each address is 48 bits in length

–

The first LSB bit (I/G) in the destination address field is used to indicate an
individual (I/G = 0) or a group address (I/G = 1). A group address could identify
none, one or more, or all the stations connected to the LAN. In the source address
the first bit is reserved and reset to 0.

–

The second bit (U/L) distinguishes between locally (U/L = 1) or globally (U/L = 0)
administered addresses. For broadcast addresses this bit is also 1.

–

Each byte of each address field must be transmitted least significant bit first.

The address designation is based on the following types:
•

Individual address: this is the physical address associated with a particular station on
the network.

•

Group address. A multidestination address associated with one or more stations on a
given network. There are two kinds of multicast address:
–

Multicast-group address: an address associated with a group of logically related
stations.

–

Broadcast address: a distinguished, predefined multicast address (all 1’s in the
destination address field) that always denotes all the stations on a given LAN.
Figure 359. Address field format

MSB

46-bit address

U/L I/G LSB

I/G = 0 Individual address
I/G = 1 Group address
U/L = 0 Globally administered address
U/L = 1 Locally administered address

Bit transmission order (right to left)
ai15628

•

QTag Prefix: 4-byte field inserted between the Source address field and the MAC Client
Length/Type field. This field is an extension of the basic frame (untagged) to obtain the
tagged MAC frame. The untagged MAC frames do not include this field. The
extensions for tagging are as follows:
–

1134/1747

2-byte constant Length/Type field value consistent with the Type interpretation
(greater than 0x0600) equal to the value of the 802.1Q Tag Protocol Type (0x8100

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
hexadecimal). This constant field is used to distinguish tagged and untagged MAC
frames.
–

•

2-byte field containing the Tag control information field subdivided as follows: a 3bit user priority, a canonical format indicator (CFI) bit and a 12-bit VLAN Identifier.
The length of the tagged MAC frame is extended by 4 bytes by the QTag Prefix.

MAC client length/type: 2-byte field with different meaning (mutually exclusive),
depending on its value:
–

If the value is less than or equal to maxValidFrame (0d1500) then this field
indicates the number of MAC client data bytes contained in the subsequent data
field of the 802.3 frame (length interpretation).

–

If the value is greater than or equal to MinTypeValue (0d1536 decimal, 0x0600)
then this field indicates the nature of the MAC client protocol (Type interpretation)
related to the Ethernet frame.

Regardless of the interpretation of the length/type field, if the length of the data field is
less than the minimum required for proper operation of the protocol, a PAD field is
added after the data field but prior to the FCS (frame check sequence) field. The
length/type field is transmitted and received with the higher-order byte first.
For length/type field values in the range between maxValidLength and minTypeValue
(boundaries excluded), the behavior of the MAC sublayer is not specified: they may or
may not be passed by the MAC sublayer.
•

Data and PAD fields: n-byte data field. Full data transparency is provided, it means that
any arbitrary sequence of byte values may appear in the data field. The size of the
PAD, if any, is determined by the size of the data field. Max and min length of the data
and PAD field are:
–

Maximum length = 1500 bytes

–

Minimum length for untagged MAC frames = 46 bytes

–

Minimum length for tagged MAC frames = 42 bytes

When the data field length is less than the minimum required, the PAD field is added to
match the minimum length (42 bytes for tagged frames, 46 bytes for untagged frames).
•

Frame check sequence: 4-byte field that contains the cyclic redundancy check (CRC)
value. The CRC computation is based on the following fields: source address,
destination address, QTag prefix, length/type, LLC data and PAD (that is, all fields
except the preamble, SFD). The generating polynomial is the following:
G( x) = x

32

+x

26

+x

23

+x

22

+x

16

+x

12

+x

11

+x

10

8

7

5

4

2

+x +x +x +x +x +x+1

The CRC value of a frame is computed as follows:
•

The first 2 bits of the frame are complemented

•

The n-bits of the frame are the coefficients of a polynomial M(x) of degree (n – 1). The
first bit of the destination address corresponds to the xn – 1 term and the last bit of the
data field corresponds to the x0 term

•

M(x) is multiplied by x32 and divided by G(x), producing a remainder R(x) of degree
≤ 31

•

The coefficients of R(x) are considered as a 32-bit sequence

•

The bit sequence is complemented and the result is the CRC

•

The 32-bits of the CRC value are placed in the frame check sequence. The x32 term is
the first transmitted, the x0 term is the last one

RM0090 Rev 17

1135/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Figure 360. MAC frame format
7 bytes

Preamble

1 byte

SFD

6 bytes

Destination address

6 bytes

Source address

2 bytes

MAC client length/type

Bytes within
frame transmitted
top to bottom

MAC client data
46-1500 bytes

PAD

4 bytes

Frame check sequence

MSB

LSB
Bit transmission order (right to left)
ai15629

Figure 361. Tagged MAC frame format
7 bytes
1 byte
6 bytes
6 bytes
QTag Prefix
4 bytes
2 bytes
42-1500 bytes

Preamble
SFD

bytes within
frame transmitted
top to bottom

Destination address
MSB

Source address
Length/type = 802.1QTagType
Tag control information

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

MAC client length/type
MAC client data
User priority CFI

Pad
4 bytes

LSB

1

VLAN identifier (VID, 12 bits)

Frame check sequence

MSB

LSB
Bit transmission order (r ight to left)
ai15630

Each byte of the MAC frame, except the FCS field, is transmitted low-order bit first.
An invalid MAC frame is defined by one of the following conditions:

1136/1747

•

The frame length is inconsistent with the expected value as specified by the length/type
field. If the length/type field contains a type value, then the frame length is assumed to
be consistent with this field (no invalid frame)

•

The frame length is not an integer number of bytes (extra bits)

•

The CRC value computed on the incoming frame does not match the included FCS

RM0090 Rev 17

RM0090

33.5.2

Ethernet (ETH): media access control (MAC) with DMA controller

MAC frame transmission
The DMA controls all transactions for the transmit path. Ethernet frames read from the
system memory are pushed into the FIFO by the DMA. The frames are then popped out and
transferred to the MAC core. When the end-of-frame is transferred, the status of the
transmission is taken from the MAC core and transferred back to the DMA. The Transmit
FIFO has a depth of 2 Kbyte. FIFO-fill level is indicated to the DMA so that it can initiate a
data fetch in required bursts from the system memory, using the AHB interface. The data
from the AHB Master interface is pushed into the FIFO.
When the SOF is detected, the MAC accepts the data and begins transmitting to the MII.
The time required to transmit the frame data to the MII after the application initiates
transmission is variable, depending on delay factors like IFG delay, time to transmit
preamble/SFD, and any back-off delays for Half-duplex mode. After the EOF is transferred
to the MAC core, the core completes normal transmission and then gives the status of
transmission back to the DMA. If a normal collision (in Half-duplex mode) occurs during
transmission, the MAC core makes the transmit status valid, then accepts and drops all
further data until the next SOF is received. The same frame should be retransmitted from
SOF on observing a Retry request (in the Status) from the MAC. The MAC issues an
underflow status if the data are not provided continuously during the transmission. During
the normal transfer of a frame, if the MAC receives an SOF without getting an EOF for the
previous frame, then the SOF is ignored and the new frame is considered as the
continuation of the previous frame.
There are two modes of operation for popping data towards the MAC core:
•

In Threshold mode, as soon as the number of bytes in the FIFO crosses the configured
threshold level (or when the end-of-frame is written before the threshold is crossed),
the data is ready to be popped out and forwarded to the MAC core. The threshold level
is configured using the TTC bits of ETH_DMABMR.

•

In Store-and-forward mode, only after a complete frame is stored in the FIFO, the frame
is popped towards the MAC core. If the Tx FIFO size is smaller than the Ethernet frame
to be transmitted, then the frame is popped towards the MAC core when the Tx FIFO
becomes almost full.

The application can flush the Transmit FIFO of all contents by setting the FTF
(ETH_DMAOMR register [20]) bit. This bit is self-clearing and initializes the FIFO pointers to
the default state. If the FTF bit is set during a frame transfer to the MAC core, then transfer
is stopped as the FIFO is considered to be empty. Hence an underflow event occurs at the
MAC transmitter and the corresponding Status word is forwarded to the DMA.

Automatic CRC and pad generation
When the number of bytes received from the application falls below 60 (DA+SA+LT+Data),
zeros are appended to the transmitting frame to make the data length exactly 46 bytes to
meet the minimum data field requirement of IEEE 802.3. The MAC can be programmed not
to append any padding. The cyclic redundancy check (CRC) for the frame check sequence
(FCS) field is calculated and appended to the data being transmitted. When the MAC is
programmed to not append the CRC value to the end of Ethernet frames, the computed
CRC is not transmitted. An exception to this rule is that when the MAC is programmed to
append pads for frames (DA+SA+LT+Data) less than 60 bytes, CRC will be appended at the
end of the padded frames.

RM0090 Rev 17

1137/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

The CRC generator calculates the 32-bit CRC for the FCS field of the Ethernet frame. The
encoding is defined by the following polynomial.
G( x) = x

32

+x

26

+x

23

+x

22

+x

16

+x

12

+x

11

+x

10

8

7

5

4

2

+x +x +x +x +x +x+1

Transmit protocol
The MAC controls the operation of Ethernet frame transmission. It performs the following
functions to meet the IEEE 802.3/802.3z specifications. It:
•

generates the preamble and SFD

•

generates the jam pattern in Half-duplex mode

•

controls the Jabber timeout

•

controls the flow for Half-duplex mode (back pressure)

•

generates the transmit frame status

•

contains time stamp snapshot logic in accordance with IEEE 1588

When a new frame transmission is requested, the MAC sends out the preamble and SFD,
followed by the data. The preamble is defined as 7 bytes of 0b10101010 pattern, and the
SFD is defined as 1 byte of 0b10101011 pattern. The collision window is defined as 1 slot
time (512 bit times for 10/100 Mbit/s Ethernet). The jam pattern generation is applicable only
to Half-duplex mode, not to Full-duplex mode.
In MII mode, if a collision occurs at any time from the beginning of the frame to the end of
the CRC field, the MAC sends a 32-bit jam pattern of 0x5555 5555 on the MII to inform all
other stations that a collision has occurred. If the collision is seen during the preamble
transmission phase, the MAC completes the transmission of the preamble and SFD and
then sends the jam pattern.
A jabber timer is maintained to cut off the transmission of Ethernet frames if more than 2048
(default) bytes have to be transferred. The MAC uses the deferral mechanism for flow
control (back pressure) in Half-duplex mode. When the application requests to stop
receiving frames, the MAC sends a JAM pattern of 32 bytes whenever it senses the
reception of a frame, provided that transmit flow control is enabled. This results in a collision
and the remote station backs off. The application requests flow control by setting the BPA bit
(bit 0) in the ETH_MACFCR register. If the application requests a frame to be transmitted,
then it is scheduled and transmitted even when back pressure is activated. Note that if back
pressure is kept activated for a long time (and more than 16 consecutive collision events
occur) then the remote stations abort their transmissions due to excessive collisions. If IEEE
1588 time stamping is enabled for the transmit frame, this block takes a snapshot of the
system time when the SFD is put onto the transmit MII bus.

Transmit scheduler
The MAC is responsible for scheduling the frame transmission on the MII. It maintains the
interframe gap between two transmitted frames and follows the truncated binary exponential
backoff algorithm for Half-duplex mode. The MAC enables transmission after satisfying the
IFG and backoff delays. It maintains an idle period of the configured interframe gap (IFG bits
in the ETH_MACCR register) between any two transmitted frames. If frames to be
transmitted arrive sooner than the configured IFG time, the MII waits for the enable signal
from the MAC before starting the transmission on it. The MAC starts its IFG counter as soon
as the carrier signal of the MII goes inactive. At the end of the programmed IFG value, the
MAC enables transmission in Full-duplex mode. In Half-duplex mode and when IFG is

1138/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
configured for 96 bit times, the MAC follows the rule of deference specified in Section
4.2.3.2.1 of the IEEE 802.3 specification. The MAC resets its IFG counter if a carrier is
detected during the first two-thirds (64-bit times for all IFG values) of the IFG interval. If the
carrier is detected during the final one third of the IFG interval, the MAC continues the IFG
count and enables the transmitter after the IFG interval. The MAC implements the truncated
binary exponential backoff algorithm when it operates in Half-duplex mode.

Transmit flow control
When the Transmit Flow Control Enable bit (TFE bit in ETH_MACFCR) is set, the MAC
generates Pause frames and transmits them as necessary, in Full-duplex mode. The Pause
frame is appended with the calculated CRC, and is sent. Pause frame generation can be
initiated in two ways.
A pause frame is sent either when the application sets the FCB bit in the ETH_MACFCR
register or when the receive FIFO is full (packet buffer).
•

If the application has requested flow control by setting the FCB bit in ETH_MACFCR,
the MAC generates and transmits a single Pause frame. The value of the pause time in
the generated frame contains the programmed pause time value in ETH_MACFCR. To
extend the pause or end the pause prior to the time specified in the previously
transmitted Pause frame, the application must request another Pause frame
transmission after programming the Pause Time value (PT in ETH_MACFCR register)
with the appropriate value.

•

If the application has requested flow control when the receive FIFO is full, the MAC
generates and transmits a Pause frame. The value of the pause time in the generated
frame is the programmed pause time value in ETH_MACFCR. If the receive FIFO
remains full at a configurable number of slot-times (PLT bits in ETH_MACFCR) before
this Pause time runs out, a second Pause frame is transmitted. The process is
repeated as long as the receive FIFO remains full. If this condition is no more satisfied
prior to the sampling time, the MAC transmits a Pause frame with zero pause time to
indicate to the remote end that the receive buffer is ready to receive new data frames.

Single-packet transmit operation
The general sequence of events for a transmit operation is as follows:
1.

If the system has data to be transferred, the DMA controller fetches them from the
memory through the AHB Master interface and starts forwarding them to the FIFO. It
continues to receive the data until the end of frame is transferred.

2.

When the threshold level is crossed or a full packet of data is received into the FIFO,
the frame data are popped and driven to the MAC core. The DMA continues to transfer
data from the FIFO until a complete packet has been transferred to the MAC. Upon
completion of the frame, the DMA controller is notified by the status coming from the
MAC.

Transmit operation—Two packets in the buffer
1.

Because the DMA must update the descriptor status before releasing it to the Host,
there can be at the most two frames inside a transmit FIFO. The second frame is
fetched by the DMA and put into the FIFO only if the OSF (operate on second frame)
bit is set. If this bit is not set, the next frame is fetched from the memory only after the
MAC has completely processed the frame and the DMA has released the descriptors.

2.

If the OSF bit is set, the DMA starts fetching the second frame immediately after
completing the transfer of the first frame to the FIFO. It does not wait for the status to
be updated. In the meantime, the second frame is received into the FIFO while the first

RM0090 Rev 17

1139/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

frame is being transmitted. As soon as the first frame has been transferred and the
status is received from the MAC, it is pushed to the DMA. If the DMA has already
completed sending the second packet to the FIFO, the second transmission must wait
for the status of the first packet before proceeding to the next frame.

Retransmission during collision
While a frame is being transferred to the MAC, a collision event may occur on the MAC line
interface in Half-duplex mode. The MAC would then indicate a retry attempt by giving the
status even before the end of frame is received. Then the retransmission is enabled and the
frame is popped out again from the FIFO. After more than 96 bytes have been popped
towards the MAC core, the FIFO controller frees up that space and makes it available to the
DMA to push in more data. This means that the retransmission is not possible after this
threshold is crossed or when the MAC core indicates a late collision event.

Transmit FIFO flush operation
The MAC provides a control to the software to flush the Transmit FIFO through the use of Bit
20 in the Operation mode register. The Flush operation is immediate and the Tx FIFO and
the corresponding pointers are cleared to the initial state even if the Tx FIFO is in the middle
of transferring a frame to the MAC Core. This results in an underflow event in the MAC
transmitter, and the frame transmission is aborted. The status of such a frame is marked
with both underflow and frame flush events (TDES0 bits 13 and 1). No data are coming to
the FIFO from the application (DMA) during the Flush operation. Transfer transmit status
words are transferred to the application for the number of frames that is flushed (including
partial frames). Frames that are completely flushed have the Frame flush status bit (TDES0
13) set. The Flush operation is completed when the application (DMA) has accepted all of
the Status words for the frames that were flushed. The Transmit FIFO Flush control register
bit is then cleared. At this point, new frames from the application (DMA) are accepted. All
data presented for transmission after a Flush operation are discarded unless they start with
an SOF marker.

Transmit status word
At the end of the Ethernet frame transfer to the MAC core and after the core has completed
the transmission of the frame, the transmit status is given to the application. The detailed
description of the Transmit Status is the same as for bits [23:0] in TDES0. If IEEE 1588 time
stamping is enabled, a specific frames’ 64-bit time stamp is returned, along with the transmit
status.

Transmit checksum offload
Communication protocols such as TCP and UDP implement checksum fields, which helps
determine the integrity of data transmitted over a network. Because the most widespread
use of Ethernet is to encapsulate TCP and UDP over IP datagrams, the Ethernet controller
has a transmit checksum offload feature that supports checksum calculation and insertion in
the transmit path, and error detection in the receive path. This section explains the operation
of the checksum offload feature for transmitted frames.
Note:

1140/1747

The checksum for TCP, UDP or ICMP is calculated over a complete frame, then inserted
into its corresponding header field. Due to this requirement, this function is enabled only
when the Transmit FIFO is configured for Store-and-forward mode (that is, when the TSF bit

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
is set in the ETH_ETH_DMAOMR register). If the core is configured for Threshold (cutthrough) mode, the Transmit checksum offload is bypassed.
You must make sure the Transmit FIFO is deep enough to store a complete frame before
that frame is transferred to the MAC Core transmitter. If the FIFO depth is less than the input
Ethernet frame size, the payload (TCP/UDP/ICMP) checksum insertion function is bypassed
and only the frame’s IPv4 Header checksum is modified, even in Store-and-forward mode.
The transmit checksum offload supports two types of checksum calculation and insertion.
This checksum can be controlled for each frame by setting the CIC bits (Bits 28:27 in
TDES1, described in TDES1: Transmit descriptor Word1 on page 1174).
See IETF specifications RFC 791, RFC 793, RFC 768, RFC 792, RFC 2460 and RFC 4443
for IPv4, TCP, UDP, ICMP, IPv6 and ICMPv6 packet header specifications, respectively.
•

IP header checksum
In IPv4 datagrams, the integrity of the header fields is indicated by the 16-bit header
checksum field (the eleventh and twelfth bytes of the IPv4 datagram). The checksum
offload detects an IPv4 datagram when the Ethernet frame’s Type field has the value
0x0800 and the IP datagram’s Version field has the value 0x4. The input frame’s
checksum field is ignored during calculation and replaced by the calculated value. IPv6
headers do not have a checksum field; thus, the checksum offload does not modify
IPv6 header fields. The result of this IP header checksum calculation is indicated by the
IP Header Error status bit in the Transmit status (Bit 16). This status bit is set whenever
the values of the Ethernet Type field and the IP header’s Version field are not
consistent, or when the Ethernet frame does not have enough data, as indicated by the
IP header Length field. In other words, this bit is set when an IP header error is
asserted under the following circumstances:
a)

For IPv4 datagrams:

–

The received Ethernet type is 0x0800, but the IP header’s Version field does not
equal 0x4

–

The IPv4 Header Length field indicates a value less than 0x5 (20 bytes)

–

The total frame length is less than the value given in the IPv4 Header Length field

b)

For IPv6 datagrams:

–

The Ethernet type is 0x86DD but the IP header Version field does not equal 0x6

–

The frame ends before the IPv6 header (40 bytes) or extension header (as given
in the corresponding Header Length field in an extension header) has been
completely received. Even when the checksum offload detects such an IP header

RM0090 Rev 17

1141/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

error, it inserts an IPv4 header checksum if the Ethernet Type field indicates an
IPv4 payload.
•

TCP/UDP/ICMP checksum
The TCP/UDP/ICMP checksum processes the IPv4 or IPv6 header (including
extension headers) and determines whether the encapsulated payload is TCP, UDP or
ICMP.
Note that:
a)

For non-TCP, -UDP, or -ICMP/ICMPv6 payloads, this checksum is bypassed and
nothing further is modified in the frame.

b)

Fragmented IP frames (IPv4 or IPv6), IP frames with security features (such as an
authentication header or encapsulated security payload), and IPv6 frames with
routing headers are bypassed and not processed by the checksum.

The checksum is calculated for the TCP, UDP, or ICMP payload and inserted into its
corresponding field in the header. It can work in the following two modes:
–

In the first mode, the TCP, UDP, or ICMPv6 pseudo-header is not included in the
checksum calculation and is assumed to be present in the input frame’s checksum
field. The checksum field is included in the checksum calculation, and then
replaced by the final calculated checksum.

–

In the second mode, the checksum field is ignored, the TCP, UDP, or ICMPv6
pseudo-header data are included into the checksum calculation, and the
checksum field is overwritten with the final calculated value.

Note that: for ICMP-over-IPv4 packets, the checksum field in the ICMP packet must
always be 0x0000 in both modes, because pseudo-headers are not defined for such
packets. If it does not equal 0x0000, an incorrect checksum may be inserted into the
packet.
The result of this operation is indicated by the payload checksum error status bit in the
Transmit Status vector (bit 12). The payload checksum error status bit is set when
either of the following is detected:
–

the frame has been forwarded to the MAC transmitter in Store-and-forward mode
without the end of frame being written to the FIFO

–

the packet ends before the number of bytes indicated by the payload length field in
the IP header is received.

When the packet is longer than the indicated payload length, the bytes are ignored as
stuff bytes, and no error is reported. When the first type of error is detected, the TCP,
UDP or ICMP header is not modified. For the second error type, still, the calculated
checksum is inserted into the corresponding header field.

MII/RMII transmit bit order
Each nibble from the MII is transmitted on the RMII a dibit at a time with the order of dibit
transmission shown in Figure 362. Lower order bits (D1 and D0) are transmitted first
followed by higher order bits (D2 and D3).

1142/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

LSB
D0

LSB

RMII_TXD[1:0]

Figure 362. Transmission bit order

MSB
D1

Bibit stream

D0

D1
MII_TXD[3:0]
D2

D3

MSB

Nibble stream

ai15632

MII/RMII transmit timing diagrams
Figure 363. Transmission with no collision
MII_TX_CLK

MII_TX_EN

MII_TXD[3:0]

PR

EA

MB

LE

MII_CS

MII_COL

Low
ai15631

RM0090 Rev 17

1143/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Figure 364. Transmission with collision
MII_TX_CLK

MII_TX_EN

MII_TXD[3:0]

PR

EAM

BLE

SFD

DA

DA

JAM

JAM

JAM

JAM

MII_CS

MII_COL

ai15651

Figure 365 shows a frame transmission in MII and RMII.
Figure 365. Frame transmission in MMI and RMII modes
MII_RX_CLK

MII_TX_EN

MII_TXD[3:0]

RMII_REF_CLK

RMII_TX_EN

RMII_TXD[1:0]

ai15652

33.5.3

MAC frame reception
The MAC received frames are pushes into the Rx FIFO. The status (fill level) of this FIFO is
indicated to the DMA once it crosses the configured receive threshold (RTC in the
ETH_DMAOMR register) so that the DMA can initiate pre-configured burst transfers
towards the AHB interface.
In the default Cut-through mode, when 64 bytes (configured with the RTC bits in the
ETH_DMAOMR register) or a full packet of data are received into the FIFO, the data are
popped out and the DMA is notified of its availability. Once the DMA has initiated the
transfer to the AHB interface, the data transfer continues from the FIFO until a complete

1144/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
packet has been transferred. Upon completion of the EOF frame transfer, the status word is
popped out and sent to the DMA controller.
In Rx FIFO Store-and-forward mode (configured by the RSF bit in the ETH_DMAOMR
register), a frame is read only after being written completely into the Receive FIFO. In this
mode, all error frames are dropped (if the core is configured to do so) such that only valid
frames are read and forwarded to the application. In Cut-through mode, some error frames
are not dropped, because the error status is received at the end of the frame, by which time
the start of that frame has already been read of the FIFO.
A receive operation is initiated when the MAC detects an SFD on the MII. The core strips the
preamble and SFD before proceeding to process the frame. The header fields are checked
for the filtering and the FCS field used to verify the CRC for the frame. The frame is dropped
in the core if it fails the address filter.

Receive protocol
The received frame preamble and SFD are stripped. Once the SFD has been detected, the
MAC starts sending the Ethernet frame data to the receive FIFO, beginning with the first
byte following the SFD (destination address). If IEEE 1588 time stamping is enabled, a
snapshot of the system time is taken when any frame's SFD is detected on the MII. Unless
the MAC filters out and drops the frame, this time stamp is passed on to the application.
If the received frame length/type field is less than 0x600 and if the MAC is programmed for
the auto CRC/pad stripping option, the MAC sends the data of the frame to RxFIFO up to
the count specified in the length/type field, then starts dropping bytes (including the FCS
field). If the Length/Type field is greater than or equal to 0x600, the MAC sends all received
Ethernet frame data to Rx FIFO, regardless of the value on the programmed auto-CRC strip
option. The MAC watchdog timer is enabled by default, that is, frames above 2048 bytes
(DA + SA + LT + Data + pad + FCS) are cut off. This feature can be disabled by
programming the watchdog disable (WD) bit in the MAC configuration register. However,
even if the watchdog timer is disabled, frames greater than 16 KB in size are cut off and a
watchdog timeout status is given.

Receive CRC: automatic CRC and pad stripping
The MAC checks for any CRC error in the receiving frame. It calculates the 32-bit CRC for
the received frame that includes the Destination address field through the FCS field. The
encoding is defined by the following polynomial.
G( x) = x

32

+x

26

+x

23

+x

22

+x

16

+x

12

+x

11

+x

10

8

7

5

4

2

+x +x +x +x +x +x+1

Regardless of the auto-pad/CRC strip, the MAC receives the entire frame to compute the
CRC check for the received frame.

Receive checksum offload
Both IPv4 and IPv6 frames in the received Ethernet frames are detected and processed for
data integrity. You can enable the receive checksum offload by setting the IPCO bit in the
ETH_MACCR register. The MAC receiver identifies IPv4 or IPv6 frames by checking for
value 0x0800 or 0x86DD, respectively, in the received Ethernet frame Type field. This
identification applies to VLAN-tagged frames as well. The receive checksum offload
calculates IPv4 header checksums and checks that they match the received IPv4 header
checksums. The IP Header Error bit is set for any mismatch between the indicated payload

RM0090 Rev 17

1145/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

type (Ethernet Type field) and the IP header version, or when the received frame does not
have enough bytes, as indicated by the IPv4 header’s Length field (or when fewer than 20
bytes are available in an IPv4 or IPv6 header). The receive checksum offload also identifies
a TCP, UDP or ICMP payload in the received IP datagrams (IPv4 or IPv6) and calculates the
checksum of such payloads properly, as defined in the TCP, UDP or ICMP specifications. It
includes the TCP/UDP/ICMPv6 pseudo-header bytes for checksum calculation and checks
whether the received checksum field matches the calculated value. The result of this
operation is given as a Payload Checksum Error bit in the receive status word. This status
bit is also set if the length of the TCP, UDP or ICMP payload does not match the expected
payload length given in the IP header. As mentioned in TCP/UDP/ICMP checksum on
page 1142, the receive checksum offload bypasses the payload of fragmented IP
datagrams, IP datagrams with security features, IPv6 routing headers, and payloads other
than TCP, UDP or ICMP. This information (whether the checksum is bypassed or not) is
given in the receive status, as described in the RDES0: Receive descriptor Word0 section.
In this configuration, the core does not append any payload checksum bytes to the received
Ethernet frames.
As mentioned in RDES0: Receive descriptor Word0 on page 1181, the meaning of certain
register bits changes as shown in Table 190.
Table 190. Frame statuses
Bit 18:
Bit 27: Header Bit 28: Payload
Ethernet frame checksum error checksum error

Frame status

0

0

0

The frame is an IEEE 802.3 frame (Length
field value is less than 0x0600).

1

0

0

IPv4/IPv6 Type frame in which no checksum
error is detected.

1

0

1

IPv4/IPv6 Type frame in which a payload
checksum error (as described for PCE) is
detected

1

1

0

IPv4/IPv6 Type frame in which IP header
checksum error (as described for IPCO HCE)
is detected.

1

1

1

IPv4/IPv6 Type frame in which both PCE and
IPCO HCE are detected.

0

0

1

IPv4/IPv6 Type frame in which there is no IP
HCE and the payload check is bypassed due
to unsupported payload.

0

1

1

Type frame which is neither IPv4 or IPv6
(checksum offload bypasses the checksum
check completely)

0

1

0

Reserved

Receive frame controller
If the RA bit is reset in the MAC CSR frame filter register, the MAC performs frame filtering
based on the destination/source address (the application still needs to perform another level
of filtering if it decides not to receive any bad frames like runt, CRC error frames, etc.). On
detecting a filter-fail, the frame is dropped and not transferred to the application. When the
filtering parameters are changed dynamically, and in case of (DA-SA) filter-fail, the rest of
1146/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
the frame is dropped and the Rx Status Word is immediately updated (with zero frame
length, CRC error and Runt Error bits set), indicating the filter fail. In Ethernet power down
mode, all received frames are dropped, and are not forwarded to the application.

Receive flow control
The MAC detects the receiving Pause frame and pauses the frame transmission for the
delay specified within the received Pause frame (only in Full-duplex mode). The Pause
frame detection function can be enabled or disabled with the RFCE bit in ETH_MACFCR.
Once receive flow control has been enabled, the received frame destination address begins
to be monitored for any match with the multicast address of the control frame
(0x0180 C200 0001). If a match is detected (the destination address of the received frame
matches the reserved control frame destination address), the MAC then decides whether or
not to transfer the received control frame to the application, based on the level of the PCF
bit in ETH_MACFFR.
The MAC also decodes the type, opcode, and Pause Timer fields of the receiving control
frame. If the byte count of the status indicates 64 bytes, and if there is no CRC error, the
MAC transmitter pauses the transmission of any data frame for the duration of the decoded
Pause time value, multiplied by the slot time (64 byte times for both 10/100 Mbit/s modes).
Meanwhile, if another Pause frame is detected with a zero Pause time value, the MAC
resets the Pause time and manages this new pause request.
If the received control frame matches neither the type field (0x8808), the opcode (0x00001),
nor the byte length (64 bytes), or if there is a CRC error, the MAC does not generate a
Pause.
In the case of a pause frame with a multicast destination address, the MAC filters the frame
based on the address match.
For a pause frame with a unicast destination address, the MAC filtering depends on whether
the DA matched the contents of the MAC address 0 register and whether the UPDF bit in
ETH_MACFCR is set (detecting a pause frame even with a unicast destination address).
The PCF register bits (bits [7:6] in ETH_MACFFR) control filtering for control frames in
addition to address filtering.

Receive operation multiframe handling
Since the status is available immediately following the data, the FIFO is capable of storing
any number of frames into it, as long as it is not full.

Error handling
If the Rx FIFO is full before it receives the EOF data from the MAC, an overflow is declared
and the whole frame is dropped, and the overflow counter in the (ETH_DMAMFBOCR
register) is incremented. The status indicates a partial frame due to overflow. The Rx FIFO
can filter error and undersized frames, if enabled (using the FEF and FUGF bits in
ETH_DMAOMR).
If the Receive FIFO is configured to operate in Store-and-forward mode, all error frames can
be filtered and dropped.
In Cut-through mode, if a frame's status and length are available when that frame's SOF is
read from the Rx FIFO, then the complete erroneous frame can be dropped. The DMA can
flush the error frame being read from the FIFO, by enabling the receive frame flash bit. The
data transfer to the application (DMA) is then stopped and the rest of the frame is internally
read and dropped. The next frame transfer can then be started, if available.

RM0090 Rev 17

1147/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Receive status word
At the end of the Ethernet frame reception, the MAC outputs the receive status to the
application (DMA). The detailed description of the receive status is the same as for
bits[31:0] in RDES0, given in RDES0: Receive descriptor Word0 on page 1181.

Frame length interface
In case of switch applications, data transmission and reception between the application and
MAC happen as complete frame transfers. The application layer should be aware of the
length of the frames received from the ingress port in order to transfer the frame to the
egress port. The MAC core provides the frame length of each received frame inside the
status at the end of each frame reception.
Note:

A frame length value of 0 is given for partial frames written into the Rx FIFO due to overflow.

MII/RMII receive bit order
Each nibble is transmitted to the MII from the dibit received from the RMII in the nibble
transmission order shown in Figure 366. The lower-order bits (D0 and D1) are received first,
followed by the higher-order bits (D2 and D3).

LSB
D0

LSB

RMII_RXD[1:0]

Figure 366. Receive bit order

MSB
D1

Di-bit stream

D0

D1
MII_RXD[3:0]
D2

MSB

D3
Nibble stream

1148/1747

RM0090 Rev 17

ai15633

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
Figure 367. Reception with no error
MII_RX_CLK

MII_RX_DV

MII_RXD[3:0]

PREAMBLE

SFD

FCS

MII_RX_ERR
ai15634

Figure 368. Reception with errors
MII_RX_CLK

MII_RX_DV

MII_RXD[3:0]

PREAMBLE

SFD

DA

DA

XX

XX

XX

MII_RX_ERR
ai15635

Figure 369. Reception with false carrier indication
MII_RX_CLK

MII_RX_DV

MII_RXD[3:0]

XX

XX

XX

XX

0E

XX

XX

XX

XX

MII_RX_ERR
ai15636

33.5.4

MAC interrupts
Interrupts can be generated from the MAC core as a result of various events.
The ETH_MACSR register describes the events that can cause an interrupt from the MAC
core. You can prevent each event from asserting the interrupt by setting the corresponding
mask bits in the Interrupt Mask register.

RM0090 Rev 17

1149/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

The interrupt register bits only indicate the block from which the event is reported. You have
to read the corresponding status registers and other registers to clear the interrupt. For
example, bit 3 of the Interrupt register, set high, indicates that the Magic packet or Wake-onLAN frame is received in Power-down mode. You must read the ETH_MACPMTCSR
Register to clear this interrupt event.
Figure 370. MAC core interrupt masking scheme
TSTS

TSTI

AND
TSTIM

OR

Interrupt

PMTS
PMTI

AND

PMTIM

ai15637

33.5.5

MAC filtering
Address filtering
Address filtering checks the destination and source addresses on all received frames and
the address filtering status is reported accordingly. Address checking is based on different
parameters (Frame filter register) chosen by the application. The filtered frame can also be
identified: multicast or broadcast frame.
Address filtering uses the station's physical (MAC) address and the Multicast Hash table for
address checking purposes.

Unicast destination address filter
The MAC supports up to 4 MAC addresses for unicast perfect filtering. If perfect filtering is
selected (HU bit in the Frame filter register is reset), the MAC compares all 48 bits of the
received unicast address with the programmed MAC address for any match. Default
MacAddr0 is always enabled, other addresses MacAddr1–MacAddr3 are selected with an
individual enable bit. Each byte of these other addresses (MacAddr1–MacAddr3) can be
masked during comparison with the corresponding received DA byte by setting the
corresponding Mask Byte Control bit in the register. This helps group address filtering for the
DA. In Hash filtering mode (when HU bit is set), the MAC performs imperfect filtering for
unicast addresses using a 64-bit Hash table. For hash filtering, the MAC uses the 6 upper
CRC (see note 1 below) bits of the received destination address to index the content of the
Hash table. A value of 000000 selects bit 0 in the selected register, and a value of 111111
selects bit 63 in the Hash Table register. If the corresponding bit (indicated by the 6-bit CRC)
is set to 1, the unicast frame is said to have passed the Hash filter; otherwise, the frame has
failed the Hash filter.
Note:

This CRC is a 32-bit value coded by the following polynomial (for more details refer to
Section 33.5.3: MAC frame reception):
G(x) = x

1150/1747

32

+x

26

+x

23

+x

22

+x

16

+x

12

+x

11

+x

10

RM0090 Rev 17

8

7

5

4

2

+x +x +x +x +x +x+1

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Multicast destination address filter
The MAC can be programmed to pass all multicast frames by setting the PAM bit in the
Frame filter register. If the PAM bit is reset, the MAC performs the filtering for multicast
addresses based on the HM bit in the Frame filter register. In Perfect filtering mode, the
multicast address is compared with the programmed MAC destination address registers (1–
3). Group address filtering is also supported. In Hash filtering mode, the MAC performs
imperfect filtering using a 64-bit Hash table. For hash filtering, the MAC uses the 6 upper
CRC (see note 1 below) bits of the received multicast address to index the content of the
Hash table. A value of 000000 selects bit 0 in the selected register and a value of 111111
selects bit 63 in the Hash Table register. If the corresponding bit is set to 1, then the
multicast frame is said to have passed the Hash filter; otherwise, the frame has failed the
Hash filter.
Note:

This CRC is a 32-bit value coded by the following polynomial (for more details refer to
Section 33.5.3: MAC frame reception):
G(x) = x

32

+x

26

+x

23

+x

22

+x

16

+x

12

+x

11

+x

10

8

7

5

4

2

+x +x +x +x +x +x+1

Hash or perfect address filter
The DA filter can be configured to pass a frame when its DA matches either the Hash filter
or the Perfect filter by setting the HPF bit in the Frame filter register and setting the
corresponding HU or HM bits. This configuration applies to both unicast and multicast
frames. If the HPF bit is reset, only one of the filters (Hash or Perfect) is applied to the
received frame.

Broadcast address filter
The MAC does not filter any broadcast frames in the default mode. However, if the MAC is
programmed to reject all broadcast frames by setting the BFD bit in the Frame filter register,
any broadcast frames are dropped.

Unicast source address filter
The MAC can also perform perfect filtering based on the source address field of the
received frames. By default, the MAC compares the SA field with the values programmed in
the SA registers. The MAC address registers [1:3] can be configured to contain SA instead
of DA for comparison, by setting bit 30 in the corresponding register. Group filtering with SA
is also supported. The frames that fail the SA filter are dropped by the MAC if the SAF bit in
the Frame filter register is set. Otherwise, the result of the SA filter is given as a status bit in
the Receive Status word (see RDES0: Receive descriptor Word0).
When the SAF bit is set, the result of the SA and DA filters is AND’ed to decide whether the
frame needs to be forwarded. This means that either of the filter fail result will drop the
frame. Both filters have to pass the frame for the frame to be forwarded to the application.

Inverse filtering operation
For both destination and source address filtering, there is an option to invert the filter-match
result at the final output. These are controlled by the DAIF and SAIF bits in the Frame filter
register, respectively. The DAIF bit is applicable for both Unicast and Multicast DA frames.
The result of the unicast/multicast destination address filter is inverted in this mode.
Similarly, when the SAIF bit is set, the result of the unicast SA filter is inverted. Table 191
and Table 192 summarize destination and source address filtering based on the type of
RM0090 Rev 17

1151/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

frame received.
Table 191. Destination address filtering
Frame
type

HPF

HU

DAIF

HM

PAM DB

DA filter operation

1

X

X

X

X

X

X

Pass

Broadcast 0

X

X

X

X

X

0

Pass

0

X

X

X

X

X

1

Fail

1

X

X

X

X

X

X

Pass all frames

0

X

0

0

X

X

X

Pass on perfect/group filter match

0

X

0

1

X

X

X

Fail on perfect/Group filter match

0

0

1

0

X

X

X

Pass on hash filter match

0

0

1

1

X

X

X

Fail on hash filter match

0

1

1

0

X

X

X

Pass on hash or perfect/Group filter
match

0

1

1

1

X

X

X

Fail on hash or perfect/Group filter match

1

X

X

X

X

X

X

Pass all frames

X

X

X

X

X

1

X

Pass all frames

0

X

X

0

0

0

X

Pass on Perfect/Group filter match and
drop PAUSE control frames if PCF = 0x

0

0

X

0

1

0

X

Pass on hash filter match and drop
PAUSE control frames if PCF = 0x

0

1

X

0

1

0

X

Pass on hash or perfect/Group filter
match and drop PAUSE control frames if
PCF = 0x

0

X

X

1

0

0

X

Fail on perfect/Group filter match and
drop PAUSE control frames if PCF = 0x

0

0

X

1

1

0

X

Fail on hash filter match and drop PAUSE
control frames if PCF = 0x

0

1

X

1

1

0

X

Fail on hash or perfect/Group filter match
and drop PAUSE control frames if PCF =
0x

Unicast

Multicast

1152/1747

PM

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
Table 192. Source address filtering
Frame type

Unicast

33.5.6

PM

SAIF

SAF

SA filter operation

1

X

X

Pass all frames

0

0

0

Pass status on perfect/Group filter match but do not drop
frames that fail

0

1

0

Fail status on perfect/group filter match but do not drop frame

0

0

1

Pass on perfect/group filter match and drop frames that fail

0

1

1

Fail on perfect/group filter match and drop frames that fail

MAC loopback mode
The MAC supports loopback of transmitted frames onto its receiver. By default, the MAC
loopback function is disabled, but this feature can be enabled by programming the
Loopback bit in the MAC ETH_MACCR register.

33.5.7

MAC management counters: MMC
The MAC management counters (MMC) maintain a set of registers for gathering statistics
on the received and transmitted frames. These include a control register for controlling the
behavior of the registers, two 32-bit registers containing generated interrupts (receive and
transmit), and two 32-bit registers containing masks for the Interrupt register (receive and
transmit). These registers are accessible from the application. Each register is 32 bits wide.
Section 33.8: Ethernet register descriptions describes the various counters and lists the
addresses of each of the statistics counters. This address is used for read/write accesses to
the desired transmit/receive counter.
The Receive MMC counters are updated for frames that pass address filtering. Dropped
frames statistics are not updated unless the dropped frames are runt frames of less than 6
bytes (DA bytes are not received fully).

Good transmitted and received frames
Transmitted frames are considered “good” if transmitted successfully. In other words, a
transmitted frame is good if the frame transmission is not aborted due to any of the following
errors:
+ Jabber Timeout
+ No Carrier/Loss of Carrier
+ Late Collision
+ Frame Underflow
+ Excessive Deferral
+ Excessive Collision

RM0090 Rev 17

1153/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Received frames are considered “good” if none of the following errors exists:
+ CRC error
+ Runt Frame (shorter than 64 bytes)
+ Alignment error (in 10/ 100 Mbit/s only)
+ Length error (non-Type frames only)
+ Out of Range (non-Type frames only, longer than maximum size)
+ MII_RXER Input error
The maximum frame size depends on the frame type, as follows:
+ Untagged frame maxsize = 1518
+ VLAN Frame maxsize = 1522

33.5.8

Power management: PMT
This section describes the power management (PMT) mechanisms supported by the MAC.
PMT supports the reception of network (remote) wakeup frames and Magic Packet frames.
PMT generates interrupts for wakeup frames and Magic Packets received by the MAC. The
PMT block is enabled with remote wakeup frame enable and Magic Packet enable. These
enable bits (WFE and MPE) are in the ETH_MACPMTCSR register and are programmed by
the application. When the power down mode is enabled in the PMT, then all received frames
are dropped by the MAC and they are not forwarded to the application. The MAC comes out
of the power down mode only when either a Magic Packet or a Remote wakeup frame is
received and the corresponding detection is enabled.

Remote wakeup frame filter register
There are eight wakeup frame filter registers. To write on each of them, load the wakeup
frame filter register value by value. The wanted values of the wakeup frame filter are loaded
by sequentially loading eight times the wakeup frame filter register. The read operation is
identical to the write operation. To read the eight values, you have to read eight times the
wakeup frame filter register to reach the last register. Each read/write points the wakeup
frame filter register to the next filter register.

1154/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
Figure 371. Wakeup frame filter register
Wakeup frame filter reg0

Filter 0 Byte Mask

Wakeup frame filter reg1

Filter 1 Byte Mask

Wakeup frame filter reg2

Filter 2 Byte Mask

Wakeup frame filter reg3

Filter 3 Byte Mask

Wakeup frame filter reg4
Wakeup frame filter reg5

RSVD

Filter 3
Command

Filter 3 Offset

RSVD

Filter 2
Command

Filter 2 Offset

RSVD

Filter 1
Command

Filter 1 Offset

RSVD

Filter 0
Command

Filter 0 Offset

Wakeup frame filter reg6

Filter 1 CRC - 16

Filter 0 CRC - 16

Wakeup frame filter reg7

Filter 3 CRC - 16

Filter 2 CRC - 16
ai15647

•

Filter i Byte Mask
This register defines which bytes of the frame are examined by filter i (0, 1, 2, and 3) in
order to determine whether or not the frame is a wakeup frame. The MSB (thirty-first
bit) must be zero. Bit j [30:0] is the Byte Mask. If bit j (byte number) of the Byte Mask is
set, then Filter i Offset + j of the incoming frame is processed by the CRC block;
otherwise Filter i Offset + j is ignored.

•

Filter i Command
This 4-bit command controls the filter i operation. Bit 3 specifies the address type,
defining the pattern’s destination address type. When the bit is set, the pattern applies
to only multicast frames. When the bit is reset, the pattern applies only to unicast
frames. Bit 2 and bit 1 are reserved. Bit 0 is the enable bit for filter i; if bit 0 is not set,
filter i is disabled.

•

Filter i Offset
This register defines the offset (within the frame) from which the frames are examined
by filter i. This 8-bit pattern offset is the offset for the filter i first byte to be examined.
The minimum allowed is 12, which refers to the 13th byte of the frame (offset value 0
refers to the first byte of the frame).

•

Filter i CRC-16
This register contains the CRC_16 value calculated from the pattern, as well as the
byte mask programmed to the wakeup filter register block.

Remote wakeup frame detection
When the MAC is in sleep mode and the remote wakeup bit is enabled in the
ETH_MACPMTCSR register, normal operation is resumed after receiving a remote wakeup
frame. The application writes all eight wakeup filter registers, by performing a sequential
write to the wakeup frame filter register address. The application enables remote wakeup by
writing a 1 to bit 2 in the ETH_MACPMTCSR register. PMT supports four programmable
filters that provide different receive frame patterns. If the incoming frame passes the
address filtering of Filter Command, and if Filter CRC-16 matches the incoming examined
pattern, then the wakeup frame is received. Filter_offset (minimum value 12, which refers to
the 13th byte of the frame) determines the offset from which the frame is to be examined.
Filter Byte Mask determines which bytes of the frame must be examined. The thirty-first bit
of Byte Mask must be set to zero. The wakeup frame is checked only for length error, FCS
error, dribble bit error, MII error, collision, and to ensure that it is not a runt frame. Even if the
RM0090 Rev 17

1155/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

wakeup frame is more than 512 bytes long, if the frame has a valid CRC value, it is
considered valid. Wakeup frame detection is updated in the ETH_MACPMTCSR register for
every remote wakeup frame received. If enabled, a PMT interrupt is generated to indicate
the reception of a remote wakeup frame.

Magic packet detection
The Magic Packet frame is based on a method that uses Advanced Micro Device’s Magic
Packet technology to power up the sleeping device on the network. The MAC receives a
specific packet of information, called a Magic Packet, addressed to the node on the network.
Only Magic Packets that are addressed to the device or a broadcast address are checked to
determine whether they meet the wakeup requirements. Magic Packets that pass address
filtering (unicast or broadcast) are checked to determine whether they meet the remote
Wake-on-LAN data format of 6 bytes of all ones followed by a MAC address appearing 16
times. The application enables Magic Packet wakeup by writing a 1 to bit 1 in the
ETH_MACPMTCSR register. The PMT block constantly monitors each frame addressed to
the node for a specific Magic Packet pattern. Each received frame is checked for a
0xFFFF FFFF FFFF pattern following the destination and source address field. The PMT
block then checks the frame for 16 repetitions of the MAC address without any breaks or
interruptions. In case of a break in the 16 repetitions of the address, the 0xFFFF FFFF FFFF
pattern is scanned for again in the incoming frame. The 16 repetitions can be anywhere in
the frame, but must be preceded by the synchronization stream (0xFFFF FFFF FFFF). The
device also accepts a multicast frame, as long as the 16 duplications of the MAC address
are detected. If the MAC address of a node is 0x0011 2233 4455, then the MAC scans for
the data sequence:
Destination address source address ……………….. FFFF FFFF FFFF
0011 2233 4455 0011 2233 4455 0011 2233 4455 0011 2233 4455
0011 2233 4455 0011 2233 4455 0011 2233 4455 0011 2233 4455
0011 2233 4455 0011 2233 4455 0011 2233 4455 0011 2233 4455
0011 2233 4455 0011 2233 4455 0011 2233 4455 0011 2233 4455
…CRC
Magic Packet detection is updated in the ETH_MACPMTCSR register for received Magic
Packet. If enabled, a PMT interrupt is generated to indicate the reception of a Magic Packet.

System consideration during power-down
The Ethernet PMT block is able to detect frames while the system is in the Stop mode,
provided that the EXTI line 19 is enabled.
The MAC receiver state machine should remain enabled during the power-down mode. This
means that the RE bit has to remain set in the ETH_MACCR register because it is involved
in magic packet/ wake-on-LAN frame detection. The transmit state machine should however
be turned off during the power-down mode by clearing the TE bit in the ETH_MACCR
register. Moreover, the Ethernet DMA should be disabled during the power-down mode,
because it is not necessary to copy the magic packet/wake-on-LAN frame into the SRAM.
To disable the Ethernet DMA, clear the ST bit and the SR bit (for the transmit DMA and the
receive DMA, respectively) in the ETH_DMAOMR register.
The recommended power-down and wakeup sequences are as follows:

1156/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
1.

Disable the transmit DMA and wait for any previous frame transmissions to complete.
These transmissions can be detected when the transmit interrupt ETH_DMASR
register[0] is received.

2.

Disable the MAC transmitter and MAC receiver by clearing the RE and TE bits in the
ETH_MACCR configuration register.

3.

Wait for the receive DMA to have emptied all the frames in the Rx FIFO.

4.

Disable the receive DMA.

5.

Configure and enable the EXTI line 19 to generate either an event or an interrupt.

6.

If you configure the EXTI line 19 to generate an interrupt, you also have to correctly
configure the ETH_WKUP_IRQ Handler function, which should clear the pending bit of
the EXTI line 19.

7.

Enable Magic packet/Wake-on-LAN frame detection by setting the MFE/ WFE bit in the
ETH_MACPMTCSR register.

8.

Enable the MAC power-down mode, by setting the PD bit in the ETH_MACPMTCSR
register.

9.

Enable the MAC Receiver by setting the RE bit in the ETH_MACCR register.

10. Enter the system’s Stop mode (for more details refer to Section 5.3.4: Stop mode):
11. On receiving a valid wakeup frame, the Ethernet peripheral exits the power-down
mode.
12. Read the ETH_MACPMTCSR to clear the power management event flag, enable the
MAC transmitter state machine, and the receive and transmit DMA.
13. Configure the system clock: enable the HSE and set the clocks.

33.5.9

Precision time protocol (IEEE1588 PTP)
The IEEE 1588 standard defines a protocol that allows precise clock synchronization in
measurement and control systems implemented with technologies such as network
communication, local computing and distributed objects. The protocol applies to systems
that communicate by local area networks supporting multicast messaging, including (but not
limited to) Ethernet. This protocol is used to synchronize heterogeneous systems that
include clocks of varying inherent precision, resolution and stability. The protocol supports
system-wide synchronization accuracy in the submicrosecond range with minimum network
and local clock computing resources. The message-based protocol, known as the precision
time protocol (PTP), is transported over UDP/IP. The system or network is classified into
Master and Slave nodes for distributing the timing/clock information. The protocol’s
technique for synchronizing a slave node to a master node by exchanging PTP messages is
described in Figure 372.

RM0090 Rev 17

1157/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Figure 372. Networked time synchronization
Master clock time

Slave clock time
Sync message

t1

t2m

Follow_up message
containing value of t1

Data at
slave clock
t2 t2

t1, t2
t3m

Delay_Req message

t3 t1, t2, t3

t4
Delay_Resp message
containing value of t4

time

t1, t2, t3, t4
ai15669

1.

The master broadcasts PTP Sync messages to all its nodes. The Sync message
contains the master’s reference time information. The time at which this message
leaves the master’s system is t1. For Ethernet ports, this time has to be captured at the
MII.

2.

A slave receives the Sync message and also captures the exact time, t2, using its
timing reference.

3.

The master then sends the slave a Follow_up message, which contains the t1
information for later use.

4.

The slave sends the master a Delay_Req message, noting the exact time, t3, at which
this frame leaves the MII.

5.

The master receives this message and captures the exact time, t4, at which it enters its
system.

6.

The master sends the t4 information to the slave in the Delay_Resp message.

7.

The slave uses the four values of t1, t2, t3, and t4 to synchronize its local timing
reference to the master’s timing reference.

Most of the protocol implementation occurs in the software, above the UDP layer. As
described above, however, hardware support is required to capture the exact time when
specific PTP packets enter or leave the Ethernet port at the MII. This timing information has
to be captured and returned to the software for a proper, high-accuracy implementation of
PTP.

Reference timing source
To get a snapshot of the time, the core requires a reference time in 64-bit format (split into
two 32-bit channels, with the upper 32 bits providing time in seconds, and the lower 32 bits
indicating time in nanoseconds) as defined in the IEEE 1588 specification.
The PTP reference clock input is used to internally generate the reference time (also called
the System Time) and to capture time stamps. The frequency of this reference clock must

1158/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
be greater than or equal to the resolution of time stamp counter. The synchronization
accuracy target between the master node and the slaves is around 100 ns.
The generation, update and modification of the System Time are described in System Time
correction methods.
The accuracy depends on the PTP reference clock input period, the characteristics of the
oscillator (drift) and the frequency of the synchronization procedure.
Due to the synchronization from the Tx and Rx clock input domain to the PTP reference
clock domain, the uncertainty on the time stamp latched value is 1 reference clock period. If
we add the uncertainty due to resolution, we will add half the period for time stamping.

Transmission of frames with the PTP feature
When a frame’s SFD is output on the MII, a time stamp is captured. Frames for which time
stamp capture is required are controllable on a per-frame basis. In other words, each
transmitted frame can be marked to indicate whether a time stamp must be captured or not
for that frame. The transmitted frames are not processed to identify PTP frames. Frame
control is exercised through the control bits in the transmit descriptor. Captured time stamps
are returned to the application in the same way as the status is provided for frames. The
time stamp is sent back along with the Transmit status of the frame, inside the
corresponding transmit descriptor, thus connecting the time stamp automatically to the
specific PTP frame. The 64-bit time stamp information is written back to the TDES2 and
TDES3 fields, with TDES2 holding the time stamp’s 32 least significant bits.

Reception of frames with the PTP feature
When the IEEE 1588 time stamping feature is enabled, the Ethernet MAC captures the time
stamp of all frames received on the MII. The MAC provides the time stamp as soon as the
frame reception is complete. Captured time stamps are returned to the application in the
same way as the frame status is provided. The time stamp is sent back along with the
Receive status of the frame, inside the corresponding receive descriptor. The 64-bit time
stamp information is written back to the RDES2 and RDES3 fields, with RDES2 holding the
time stamp’s 32 least significant bits.

System Time correction methods
The 64-bit PTP time is updated using the PTP input reference clock, HCLK. This PTP time
is used as a source to take snapshots (time stamps) of the Ethernet frames being
transmitted or received at the MII. The System Time counter can be initialized or corrected
using either the Coarse or the Fine correction method.
In the Coarse correction method, the initial value or the offset value is written to the Time
stamp update register (refer to Section 33.8.3: IEEE 1588 time stamp registers on
page 1215). For initialization, the System Time counter is written with the value in the Time
stamp update registers, whereas for system time correction, the offset value (Time stamp
update register) is added to or subtracted from the system time.
In the Fine correction method, the slave clock (reference clock) frequency drift with respect
to the master clock (as defined in IEEE 1588) is corrected over a period of time, unlike in the
Coarse correction method where it is corrected in a single clock cycle. The longer correction
time helps maintain linear time and does not introduce drastic changes (or a large jitter) in
the reference time between PTP Sync message intervals. In this method, an accumulator
sums up the contents of the Addend register as shown in Figure 373. The arithmetic carry
that the accumulator generates is used as a pulse to increment the system time counter.

RM0090 Rev 17

1159/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

The accumulator and the addend are 32-bit registers. Here, the accumulator acts as a highprecision frequency multiplier or divider. Figure 373 shows this algorithm.
Figure 373. System time update using the Fine correction method
Addend register

Addend update

+
Accumulator register

Constant value

Increment Subsecond
register

+

Subsecond register

Increment Second register
Second register

ai15670

The system time update logic requires a 50 MHz clock frequency to achieve 20 ns accuracy.
The frequency division is the ratio of the reference clock frequency to the required clock
frequency. Hence, if the reference clock (HCLK) is, let us say, 66 MHz, the ratio is calculated
as 66 MHz/50 MHz = 1.32. Hence, the default addend value to be set in the register is
232/1.32, which is equal to 0xC1F0 7C1F.
If the reference clock drifts lower, to 65 MHz for example, the ratio is 65/50 or 1.3 and the
value to set in the addend register is 232/1.30 equal to 0xC4EC 4EC4. If the clock drifts
higher, to 67 MHz for example, the addend register must be set to 0xBF0 B7672. When the
clock drift is zero, the default addend value of 0xC1F0 7C1F (232/1.32) should be
programmed.
In Figure 373, the constant value used to increment the subsecond register is 0d43. This
makes an accuracy of 20 ns in the system time (in other words, it is incremented by 20 ns
steps).
The software has to calculate the drift in frequency based on the Sync messages, and to
update the Addend register accordingly. Initially, the slave clock is set with
FreqCompensationValue0 in the Addend register. This value is as follows:
FreqCompensationValue0 = 232 / FreqDivisionRatio
If MasterToSlaveDelay is initially assumed to be the same for consecutive Sync messages,
the algorithm described below must be applied. After a few Sync cycles, frequency lock
occurs. The slave clock can then determine a precise MasterToSlaveDelay value and resynchronize with the master using the new value.

1160/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
The algorithm is as follows:
•

At time MasterSyncTime (n) the master sends the slave clock a Sync message. The
slave receives this message when its local clock is SlaveClockTime (n) and computes
MasterClockTime (n) as:
MasterClockTime (n) = MasterSyncTime (n) + MasterToSlaveDelay (n)

•

The master clock count for current Sync cycle, MasterClockCount (n) is given by:
MasterClockCount (n) = MasterClockTime (n) – MasterClockTime (n – 1) (assuming
that MasterToSlaveDelay is the same for Sync cycles n and n – 1)

•

The slave clock count for current Sync cycle, SlaveClockCount (n) is given by:
SlaveClockCount (n) = SlaveClockTime (n) – SlaveClockTime (n – 1)

•

The difference between master and slave clock counts for current Sync cycle,
ClockDiffCount (n) is given by:
ClockDiffCount (n) = MasterClockCount (n) – SlaveClockCount (n)

•

The frequency-scaling factor for slave clock, FreqScaleFactor (n) is given by:
FreqScaleFactor (n) = (MasterClockCount (n) + ClockDiffCount (n)) /
SlaveClockCount (n)

•

The frequency compensation value for Addend register, FreqCompensationValue (n) is
given by:
FreqCompensationValue (n) = FreqScaleFactor (n) × FreqCompensationValue (n – 1)

In theory, this algorithm achieves lock in one Sync cycle; however, it may take several
cycles, due to changing network propagation delays and operating conditions.
This algorithm is self-correcting: if for any reason the slave clock is initially set to a value
from the master that is incorrect, the algorithm corrects it at the cost of more Sync cycles.

Programming steps for system time generation initialization
The time stamping feature can be enabled by setting bit 0 in the Time stamp control register
(ETH__PTPTSCR). However, it is essential to initialize the time stamp counter after this bit
is set to start time stamp operation. The proper sequence is the following:
1.

Mask the Time stamp trigger interrupt by setting bit 9 in the MACIMR register.

2.

Program Time stamp register bit 0 to enable time stamping.

3.

Program the Subsecond increment register based on the PTP clock frequency.

4.

If you are using the Fine correction method, program the Time stamp addend register
and set Time stamp control register bit 5 (addend register update).

5.

Poll the Time stamp control register until bit 5 is cleared.

6.

To select the Fine correction method (if required), program Time stamp control register
bit 1.

7.

Program the Time stamp high update and Time stamp low update registers with the
appropriate time value.

8.

Set Time stamp control register bit 2 (Time stamp init).

9.

The Time stamp counter starts operation as soon as it is initialized with the value
written in the Time stamp update register.

10. Enable the MAC receiver and transmitter for proper time stamping.
Note:

If time stamp operation is disabled by clearing bit 0 in the ETH_PTPTSCR register, the
above steps must be repeated to restart the time stamp operation.

RM0090 Rev 17

1161/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Programming steps for system time update in the Coarse correction method
To synchronize or update the system time in one process (coarse correction method),
perform the following steps:
1.

Write the offset (positive or negative) in the Time stamp update high and low registers.

2.

Set bit 3 (TSSTU) in the Time stamp control register.

3.

The value in the Time stamp update registers is added to or subtracted from the system
time when the TSSTU bit is cleared.

Programming steps for system time update in the Fine correction method
To synchronize or update the system time to reduce system-time jitter (fine correction
method), perform the following steps:
1.

With the help of the algorithm explained in Section : System Time correction methods,
calculate the rate by which you want to speed up or slow down the system time
increments.

2.

Update the time stamp.

3.

Wait the time you want the new value of the Addend register to be active. You can do
this by activating the Time stamp trigger interrupt after the system time reaches the
target value.

4.

Program the required target time in the Target time high and low registers. Unmask the
Time stamp interrupt by clearing bit 9 in the ETH_MACIMR register.

5.

Set Time stamp control register bit 4 (TSARU).

6.

When this trigger causes an interrupt, read the ETH_MACSR register.

7.

Reprogram the Time stamp addend register with the old value and set ETH_TPTSCR
bit 5 again.

PTP trigger internal connection with TIM2
The MAC provides a trigger interrupt when the system time becomes greater than the target
time. Using an interrupt introduces a known latency plus an uncertainty in the command
execution time.
In order to avoid this uncertainty, a PTP trigger output signal is set high when the system
time is greater than the target time. It is internally connected to the TIM2 input trigger. With
this signal, the input capture feature, the output compare feature and the waveforms of the
timer can be used, triggered by the synchronized PTP system time. No uncertainty is
introduced since the clock of the timer (PCLK1: TIM2 APB1 clock) and PTP reference clock
(HCLK) are synchronous.
This PTP trigger signal is connected to the TIM2 ITR1 input selectable by software. The
connection is enabled through bits 11 and 10 in the TIM2 option register (TIM2_OR).
Figure 374 shows the connection.
Figure 374. PTP trigger output to TIM2 ITR1 connection
PTP trigger
Ethernet MAC

ITR1
TIM2
ai15671

1162/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

PTP pulse-per-second output signal
This PTP pulse output is used to check the synchronization between all nodes in the
network. To be able to test the difference between the local slave clock and the master
reference clock, both clocks were given a pulse-per-second (PPS) output signal that may be
connected to an oscilloscope if necessary. The deviation between the two signals can
therefore be measured. The pulse width of the PPS output is 125 ms.
The PPS output is enabled through a GPIO alternate function. (GPIO_AFR register).
The default frequency of the PPS output is 1 Hz. PPSFREQ[3:0] (in ETH_PTPPPSCR) can
be used to set the frequency of the PPS output to 2PPSFREQ Hz.
When set to 1 Hz, the PPS pulse width is 125 ms with binary rollover (TSSSR=0, bit 9 in
ETH_PTPTSCR) and 100 ms with digital rollover (TSSSR=1). When set to 2 Hz and higher,
the duty cycle of the PPS output is 50% with binary rollover.
With digital rollover (TSSSR=1), it is recommended not to use the PPS output with a
frequency other than 1 Hz as it would have irregular waveforms (though its average
frequency would always be correct during any one-second window).
Figure 375. PPS output
PPS output
Ethernet MAC
ai15672

33.6

Ethernet functional description: DMA controller operation
The DMA has independent transmit and receive engines, and a CSR space. The transmit
engine transfers data from system memory into the Tx FIFO while the receive engine
transfers data from the Rx FIFO into system memory. The controller utilizes descriptors to
efficiently move data from source to destination with minimum CPU intervention. The DMA
is designed for packet-oriented data transfers such as frames in Ethernet. The controller can
be programmed to interrupt the CPU in cases such as frame transmit and receive transfer
completion, and other normal/error conditions. The DMA and the STM32F4xx communicate
through two data structures:
•

Control and status registers (CSR)

•

Descriptor lists and data buffers.

Control and status registers are described in detail in Section 33.8: Ethernet register
descriptions. Descriptors are described in detail in Normal Tx DMA descriptors.
The DMA transfers the received data frames to the receive buffer in the STM32F4xx
memory, and transmits data frames from the transmit buffer in the STM32F4xx memory.
Descriptors that reside in the STM32F4xx memory act as pointers to these buffers. There
are two descriptor lists: one for reception, and one for transmission. The base address of
each list is written into DMA Registers 3 and 4, respectively. A descriptor list is forwardlinked (either implicitly or explicitly). The last descriptor may point back to the first entry to
create a ring structure. Explicit chaining of descriptors is accomplished by configuring the
second address chained in both the receive and transmit descriptors (RDES1[14] and
TDES0[20]). The descriptor lists reside in the Host’s physical memory space. Each
descriptor can point to a maximum of two buffers. This enables the use of two physically

RM0090 Rev 17

1163/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

addressed buffers, instead of two contiguous buffers in memory. A data buffer resides in the
Host’s physical memory space, and consists of an entire frame or part of a frame, but cannot
exceed a single frame. Buffers contain only data. The buffer status is maintained in the
descriptor. Data chaining refers to frames that span multiple data buffers. However, a single
descriptor cannot span multiple frames. The DMA skips to the next frame buffer when the
end of frame is detected. Data chaining can be enabled or disabled. The descriptor ring and
chain structure is shown in Figure 376.
Figure 376. Descriptor ring and chain structure
Ring structure

Chain structure
Buffer 1

Descriptor 0

Buffer 2

Descriptor 0

Buffer 1

Buffer 1
Descriptor 1

Buffer 2

Descriptor 1

Buffer 1

Buffer 1
Descriptor 2

Buffer 2
Descriptor 2

Buffer 1

Buffer 1
Descriptor n

Buffer 2
Next descriptor
ai15638

33.6.1

Initialization of a transfer using DMA
Initialization for the MAC is as follows:

33.6.2

1.

Write to ETH_DMABMR to set STM32F4xx bus access parameters.

2.

Write to the ETH_DMAIER register to mask unnecessary interrupt causes.

3.

The software driver creates the transmit and receive descriptor lists. Then it writes to
both the ETH_DMARDLAR and ETH_DMATDLAR registers, providing the DMA with
the start address of each list.

4.

Write to MAC Registers 1, 2, and 3 to choose the desired filtering options.

5.

Write to the MAC ETH_MACCR register to configure and enable the transmit and
receive operating modes. The PS and DM bits are set based on the auto-negotiation
result (read from the PHY).

6.

Write to the ETH_DMAOMR register to set bits 13 and 1 and start transmission and
reception.

7.

The transmit and receive engines enter the running state and attempt to acquire
descriptors from the respective descriptor lists. The receive and transmit engines then
begin processing receive and transmit operations. The transmit and receive processes
are independent of each other and can be started or stopped separately.

Host bus burst access
The DMA attempts to execute fixed-length burst transfers on the AHB master interface if
configured to do so (FB bit in ETH_DMABMR). The maximum burst length is indicated and
limited by the PBL field (ETH_DMABMR [13:8]). The receive and transmit descriptors are

1164/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
always accessed in the maximum possible burst size (limited by PBL) for the 16 bytes to be
read.
The Transmit DMA initiates a data transfer only when there is sufficient space in the
Transmit FIFO to accommodate the configured burst or the number of bytes until the end of
frame (when it is less than the configured burst length). The DMA indicates the start address
and the number of transfers required to the AHB Master Interface. When the AHB Interface
is configured for fixed-length burst, then it transfers data using the best combination of
INCR4, INCR8, INCR16 and SINGLE transactions. Otherwise (no fixed-length burst), it
transfers data using INCR (undefined length) and SINGLE transactions.
The Receive DMA initiates a data transfer only when sufficient data for the configured burst
is available in Receive FIFO or when the end of frame (when it is less than the configured
burst length) is detected in the Receive FIFO. The DMA indicates the start address and the
number of transfers required to the AHB master interface. When the AHB interface is
configured for fixed-length burst, then it transfers data using the best combination of INCR4,
INCR8, INCR16 and SINGLE transactions. If the end of frame is reached before the fixedburst ends on the AHB interface, then dummy transfers are performed in order to complete
the fixed-length burst. Otherwise (FB bit in ETH_DMABMR is reset), it transfers data using
INCR (undefined length) and SINGLE transactions.
When the AHB interface is configured for address-aligned beats, both DMA engines ensure
that the first burst transfer the AHB initiates is less than or equal to the size of the configured
PBL. Thus, all subsequent beats start at an address that is aligned to the configured PBL.
The DMA can only align the address for beats up to size 16 (for PBL > 16), because the
AHB interface does not support more than INCR16.

33.6.3

Host data buffer alignment
The transmit and receive data buffers do not have any restrictions on start address
alignment. In our system with 32-bit memory, the start address for the buffers can be aligned
to any of the four bytes. However, the DMA always initiates transfers with address aligned to
the bus width with dummy data for the byte lanes not required. This typically happens during
the transfer of the beginning or end of an Ethernet frame.
•

Example of buffer read:
If the Transmit buffer address is 0x0000 0FF2, and 15 bytes need to be transferred,
then the DMA will read five full words from address 0x0000 0FF0, but when transferring
data to the Transmit FIFO, the extra bytes (the first two bytes) will be dropped or
ignored. Similarly, the last 3 bytes of the last transfer will also be ignored. The DMA
always ensures it transfers a full 32-bit data items to the Transmit FIFO, unless it is the
end of frame.

•

Example of buffer write:
If the Receive buffer address is 0x0000 0FF2, and 16 bytes of a received frame need to
be transferred, then the DMA will write five full 32-bit data items from address
0x0000 0FF0. But the first 2 bytes of the first transfer and the last 2 bytes of the third
transfer will have dummy data.

33.6.4

Buffer size calculations
The DMA does not update the size fields in the transmit and receive descriptors. The DMA
updates only the status fields (xDES0) of the descriptors. The driver has to calculate the
sizes. The transmit DMA transfers the exact number of bytes (indicated by buffer size field in
TDES1) towards the MAC core. If a descriptor is marked as first (FS bit in TDES0 is set),

RM0090 Rev 17

1165/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

then the DMA marks the first transfer from the buffer as the start of frame. If a descriptor is
marked as last (LS bit in TDES0), then the DMA marks the last transfer from that data buffer
as the end of frame. The receive DMA transfers data to a buffer until the buffer is full or the
end of frame is received. If a descriptor is not marked as last (LS bit in RDES0), then the
buffer(s) that correspond to the descriptor are full and the amount of valid data in a buffer is
accurately indicated by the buffer size field minus the data buffer pointer offset when the
descriptor’s FS bit is set. The offset is zero when the data buffer pointer is aligned to the
databus width. If a descriptor is marked as last, then the buffer may not be full (as indicated
by the buffer size in RDES1). To compute the amount of valid data in this final buffer, the
driver must read the frame length (FL bits in RDES0[29:16]) and subtract the sum of the
buffer sizes of the preceding buffers in this frame. The receive DMA always transfers the
start of next frame with a new descriptor.
Note:

Even when the start address of a receive buffer is not aligned to the system databus width
the system should allocate a receive buffer of a size aligned to the system bus width. For
example, if the system allocates a 1024 byte (1 KB) receive buffer starting from address
0x1000, the software can program the buffer start address in the receive descriptor to have
a 0x1002 offset. The receive DMA writes the frame to this buffer with dummy data in the first
two locations (0x1000 and 0x1001). The actual frame is written from location 0x1002. Thus,
the actual useful space in this buffer is 1022 bytes, even though the buffer size is
programmed as 1024 bytes, due to the start address offset.

33.6.5

DMA arbiter
The arbiter inside the DMA takes care of the arbitration between transmit and receive
channel accesses to the AHB master interface. Two types of arbitrations are possible:
round-robin, and fixed-priority. When round-robin arbitration is selected (DA bit in
ETH_DMABMR is reset), the arbiter allocates the databus in the ratio set by the PM bits in
ETH_DMABMR, when both transmit and receive DMAs request access simultaneously.
When the DA bit is set, the receive DMA always gets priority over the transmit DMA for data
access.

33.6.6

Error response to DMA
For any data transfer initiated by a DMA channel, if the slave replies with an error response,
that DMA stops all operations and updates the error bits and the fatal bus error bit in the
Status register (ETH_DMASR register). That DMA controller can resume operation only
after soft- or hard-resetting the peripheral and re-initializing the DMA.

33.6.7

Tx DMA configuration
TxDMA operation: default (non-OSF) mode
The transmit DMA engine in default mode proceeds as follows:

1166/1747

1.

The user sets up the transmit descriptor (TDES0-TDES3) and sets the OWN bit
(TDES0[31]) after setting up the corresponding data buffer(s) with Ethernet frame data.

2.

Once the ST bit (ETH_DMAOMR register[13]) is set, the DMA enters the Run state.

3.

While in the Run state, the DMA polls the transmit descriptor list for frames requiring
transmission. After polling starts, it continues in either sequential descriptor ring order
or chained order. If the DMA detects a descriptor flagged as owned by the CPU, or if an
error condition occurs, transmission is suspended and both the Transmit Buffer

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
Unavailable (ETH_DMASR register[2]) and Normal Interrupt Summary (ETH_DMASR
register[16]) bits are set. The transmit engine proceeds to Step 9.
4.

If the acquired descriptor is flagged as owned by DMA (TDES0[31] is set), the DMA
decodes the transmit data buffer address from the acquired descriptor.

5.

The DMA fetches the transmit data from the STM32F4xx memory and transfers the
data.

6.

If an Ethernet frame is stored over data buffers in multiple descriptors, the DMA closes
the intermediate descriptor and fetches the next descriptor. Steps 3, 4, and 5 are
repeated until the end of Ethernet frame data is transferred.

7.

When frame transmission is complete, if IEEE 1588 time stamping was enabled for the
frame (as indicated in the transmit status) the time stamp value is written to the transmit
descriptor (TDES2 and TDES3) that contains the end-of-frame buffer. The status
information is then written to this transmit descriptor (TDES0). Because the OWN bit is
cleared during this step, the CPU now owns this descriptor. If time stamping was not
enabled for this frame, the DMA does not alter the contents of TDES2 and TDES3.

8.

Transmit Interrupt (ETH_DMASR register [0]) is set after completing the transmission
of a frame that has Interrupt on Completion (TDES1[31]) set in its last descriptor. The
DMA engine then returns to Step 3.

9.

In the Suspend state, the DMA tries to re-acquire the descriptor (and thereby returns to
Step 3) when it receives a transmit poll demand, and the Underflow Interrupt Status bit
is cleared.

Figure 377 shows the TxDMA transmission flow in default mode.

RM0090 Rev 17

1167/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Figure 377. TxDMA operation in Default mode
Start TxDMA

Start

Stop TxDMA

(Re-)fetch next
descriptor

(AHB)
error?

Poll demand

Yes

No

TxDMA suspended

No

Own
bit set?
Yes

Transfer data from
buffer(s)

(AHB)
error?

Yes

No

No

Frame xfer
complete?
Yes

Close intermediate
descriptor

Wait for Tx status

Time stamp
present?

Yes

Write time stamp to
TDES2 and TDES3

No

Write status word
to TDES0

No

(AHB)
error?

No

(AHB)
error?

Yes

Yes

ai15639

TxDMA operation: OSF mode
While in the Run state, the transmit process can simultaneously acquire two frames without
closing the Status descriptor of the first (if the OSF bit is set in ETH_DMAOMR register[2]).
As the transmit process finishes transferring the first frame, it immediately polls the transmit
descriptor list for the second frame. If the second frame is valid, the transmit process
transfers this frame before writing the first frame’s status information. In OSF mode, the
Run-state transmit DMA operates according to the following sequence:

1168/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
1.

The DMA operates as described in steps 1–6 of the TxDMA (default mode).

2.

Without closing the previous frame’s last descriptor, the DMA fetches the next
descriptor.

3.

If the DMA owns the acquired descriptor, the DMA decodes the transmit buffer address
in this descriptor. If the DMA does not own the descriptor, the DMA goes into Suspend
mode and skips to Step 7.

4.

The DMA fetches the Transmit frame from the STM32F4xx memory and transfers the
frame until the end of frame data are transferred, closing the intermediate descriptors if
this frame is split across multiple descriptors.

5.

The DMA waits for the transmission status and time stamp of the previous frame. When
the status is available, the DMA writes the time stamp to TDES2 and TDES3, if such
time stamp was captured (as indicated by a status bit). The DMA then writes the status,
with a cleared OWN bit, to the corresponding TDES0, thus closing the descriptor. If
time stamping was not enabled for the previous frame, the DMA does not alter the
contents of TDES2 and TDES3.

6.

If enabled, the Transmit interrupt is set, the DMA fetches the next descriptor, then
proceeds to Step 3 (when Status is normal). If the previous transmission status shows
an underflow error, the DMA goes into Suspend mode (Step 7).

7.

In Suspend mode, if a pending status and time stamp are received by the DMA, it
writes the time stamp (if enabled for the current frame) to TDES2 and TDES3, then
writes the status to the corresponding TDES0. It then sets relevant interrupts and
returns to Suspend mode.

8.

The DMA can exit Suspend mode and enter the Run state (go to Step 1 or Step 2
depending on pending status) only after receiving a Transmit Poll demand
(ETH_DMATPDR register).

Figure 378 shows the basic flowchart in OSF mode.

RM0090 Rev 17

1169/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Figure 378. TxDMA operation in OSF mode

Start TxDMA

Start

Stop TxDMA

(Re-)fetch next
descriptor

(AHB)
error?

Poll
demand

Yes

No

TxDMA suspended

Own
bit set?

No

Yes
Previous frame
status available

Transfer data from
buffer(s)

(AHB)
error?

Time stamp
present?

Yes

No

Yes
No

Write time stamp to
TDES2 & TDES3
for previous frame

Frame xfer
complete?

No

Yes

No

Yes

Wait for previous
frame’s Tx status

Close intermediate
descriptor

(AHB)
error?

Yes

Time stamp
present?

No

Yes

Write time stamp to
TDES2 & TDES3
for previous frame

No

Write status word to
prev. frame’s TDES0

Write status word to
prev. frame’s TDES0

No

Second
frame?

(AHB)
error?

No

No

(AHB)
error?

Yes

(AHB)
error?
Yes

Yes

ai15640

Transmit frame processing
The transmit DMA expects that the data buffers contain complete Ethernet frames,
excluding preamble, pad bytes, and FCS fields. The DA, SA, and Type/Len fields contain
valid data. If the transmit descriptor indicates that the MAC core must disable CRC or pad
insertion, the buffer must have complete Ethernet frames (excluding preamble), including
the CRC bytes. Frames can be data-chained and span over several buffers. Frames have to
be delimited by the first descriptor (TDES0[28]) and the last descriptor (TDES0[29]). As the
transmission starts, TDES0[28] has to be set in the first descriptor. When this occurs, the
frame data are transferred from the memory buffer to the Transmit FIFO. Concurrently, if the
last descriptor (TDES0[29]) of the current frame is cleared, the transmit process attempts to
acquire the next descriptor. The transmit process expects TDES0[28] to be cleared in this
descriptor. If TDES0[29] is cleared, it indicates an intermediary buffer. If TDES0[29] is set, it

1170/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
indicates the last buffer of the frame. After the last buffer of the frame has been transmitted,
the DMA writes back the final status information to the transmit descriptor 0 (TDES0) word
of the descriptor that has the last segment set in transmit descriptor 0 (TDES0[29]). At this
time, if Interrupt on Completion (TDES0[30]) is set, Transmit Interrupt (in ETH_DMASR
register [0]) is set, the next descriptor is fetched, and the process repeats. Actual frame
transmission begins after the Transmit FIFO has reached either a programmable transmit
threshold (ETH_DMAOMR register[16:14]), or a full frame is contained in the FIFO. There is
also an option for the Store and forward mode (ETH_DMAOMR register[21]). Descriptors
are released (OWN bit TDES0[31] is cleared) when the DMA finishes transferring the frame.

Transmit polling suspended
Transmit polling can be suspended by either of the following conditions:
•

The DMA detects a descriptor owned by the CPU (TDES0[31]=0) and the Transmit
buffer unavailable flag is set (ETH_DMASR register[2]). To resume, the driver must
give descriptor ownership to the DMA and then issue a Poll Demand command.

•

A frame transmission is aborted when a transmit error due to underflow is detected.
The appropriate Transmit Descriptor 0 (TDES0) bit is set. If the second condition
occurs, both the Abnormal Interrupt Summary (in ETH_DMASR register [15]) and
Transmit Underflow bits (in ETH_DMASR register[5]) are set, and the information is
written to Transmit Descriptor 0, causing the suspension. If the DMA goes into
Suspend state due to the first condition, then both the Normal Interrupt Summary
(ETH_DMASR register [16]) and Transmit Buffer Unavailable (ETH_DMASR
register[2]) bits are set. In both cases, the position in the transmit list is retained. The
retained position is that of the descriptor following the last descriptor closed by the
DMA. The driver must explicitly issue a Transmit Poll Demand command after rectifying
the suspension cause.

Normal Tx DMA descriptors
The normal transmit descriptor structure consists of four 32-bit words as shown in
Figure 379. The bit descriptions of TDES0, TDES1, TDES2 and TDES3 are given below.
Note that enhanced descriptors must be used if time stamping is activated (ETH_PTPTSCR
bit 0, TSE=1) or if IPv4 checksum offload is activated (ETH_MACCR bit 10, IPCO=1).
Figure 379. Normal transmit descriptor
31

TDES 0
TDES 1
TDES 2

TDES 3

O
W
N

0
Ctrl
[30:26]

Reserved
[31:29]

T
T Res.
S 24
E

Ctrl
[23:20]

T
Reserved T
[19:18]
S
S

Buffer 2 byte count
[28:16]

Reserved
[15:13]

Status [16:0]
Buffer 1 byte count
[12:0]

Buffer 1 address [31:0] / Time stamp low [31:0]

Buffer 2 address [31:0] or Next descriptor address [31:0] / Time stamp high [31:0]
ai15642b

RM0090 Rev 17

1171/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller
•

RM0090

TDES0: Transmit descriptor Word0
The application software has to program the control bits [30:26]+[23:20] plus the OWN
bit [31] during descriptor initialization. When the DMA updates the descriptor (or writes
it back), it resets all the control bits plus the OWN bit, and reports only the status bits.
31

30

29

28

27

26

25

OWN

IC

LS

FS

DC

DP

TTSE

rw

rw

rw

rw

rw

rw

rw

15

14

13

12

11

10

9

ES

JT

FF

IPE

LCA

NC

rw

rw

rw

rw

rw

rw

24
Res

23

22
CIC

21

20

TER

TCH

19

18
Res.

rw

rw

rw

rw

8

7

6

5

4

3

LCO

EC

VF

rw

rw

rw

rw

rw

rw

rw

CC

17

16

TTSS

IHE

rw

rw

2

1

0

ED

UF

DB

rw

rw

rw

Bit 31 OWN: Own bit
When set, this bit indicates that the descriptor is owned by the DMA. When this bit is reset, it
indicates that the descriptor is owned by the CPU. The DMA clears this bit either when it
completes the frame transmission or when the buffers allocated in the descriptor are read
completely. The ownership bit of the frame’s first descriptor must be set after all subsequent
descriptors belonging to the same frame have been set.
Bit 30 IC: Interrupt on completion
When set, this bit sets the Transmit Interrupt (Register 5[0]) after the present frame has
been transmitted.
Bit 29 LS: Last segment
When set, this bit indicates that the buffer contains the last segment of the frame.
Bit 28 FS: First segment
When set, this bit indicates that the buffer contains the first segment of a frame.
Bit 27 DC: Disable CRC
When this bit is set, the MAC does not append a cyclic redundancy check (CRC) to the end
of the transmitted frame. This is valid only when the first segment (TDES0[28]) is set.
Bit 26 DP: Disable pad
When set, the MAC does not automatically add padding to a frame shorter than 64 bytes.
When this bit is reset, the DMA automatically adds padding and CRC to a frame shorter than
64 bytes, and the CRC field is added despite the state of the DC (TDES0[27]) bit. This is
valid only when the first segment (TDES0[28]) is set.
Bit 25 TTSE: Transmit time stamp enable
When TTSE is set and when TSE is set (ETH_PTPTSCR bit 0), IEEE1588 hardware time
stamping is activated for the transmit frame described by the descriptor. This field is only valid
when the First segment control bit (TDES0[28]) is set.
Bit 24 Reserved, must be kept at reset value.
Bits 23:22 CIC: Checksum insertion control
These bits control the checksum calculation and insertion. Bit encoding is as shown below:
00: Checksum Insertion disabled
01: Only IP header checksum calculation and insertion are enabled
10: IP header checksum and payload checksum calculation and insertion are enabled, but
pseudo-header checksum is not calculated in hardware
11: IP Header checksum and payload checksum calculation and insertion are enabled, and
pseudo-header checksum is calculated in hardware.

1172/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Bit 21 TER: Transmit end of ring
When set, this bit indicates that the descriptor list reached its final descriptor. The DMA
returns to the base address of the list, creating a descriptor ring.
Bit 20 TCH: Second address chained
When set, this bit indicates that the second address in the descriptor is the next descriptor
address rather than the second buffer address. When TDES0[20] is set, TBS2
(TDES1[28:16]) is a “don’t care” value. TDES0[21] takes precedence over TDES0[20].
Bits 19:18 Reserved, must be kept at reset value.
Bit 17 TTSS: Transmit time stamp status
This field is used as a status bit to indicate that a time stamp was captured for the described
transmit frame. When this bit is set, TDES2 and TDES3 have a time stamp value captured for the
transmit frame. This field is only valid when the descriptor’s Last segment control bit (TDES0[29])
is set.
Note that when enhanced descriptors are enabled (EDFE=1 in ETH_DMABMR), TTSS=1
indicates that TDES6 and TDES7 have the time stamp value.
Bit 16 IHE: IP header error
When set, this bit indicates that the MAC transmitter detected an error in the IP datagram
header. The transmitter checks the header length in the IPv4 packet against the number of
header bytes received from the application and indicates an error status if there is a
mismatch. For IPv6 frames, a header error is reported if the main header length is not 40
bytes. Furthermore, the Ethernet length/type field value for an IPv4 or IPv6 frame must
match the IP header version received with the packet. For IPv4 frames, an error status is
also indicated if the Header Length field has a value less than 0x5.
Bit 15 ES: Error summary
Indicates the logical OR of the following bits:
TDES0[14]: Jabber timeout
TDES0[13]: Frame flush
TDES0[11]: Loss of carrier
TDES0[10]: No carrier
TDES0[9]: Late collision
TDES0[8]: Excessive collision
TDES0[2]:Excessive deferral
TDES0[1]: Underflow error
TDES0[16]: IP header error
TDES0[12]: IP payload error
Bit 14 JT: Jabber timeout
When set, this bit indicates the MAC transmitter has experienced a jabber timeout. This bit is
only set when the MAC configuration register’s JD bit is not set.
Bit 13 FF: Frame flushed
When set, this bit indicates that the DMA/MTL flushed the frame due to a software Flush
command given by the CPU.
Bit 12 IPE: IP payload error
When set, this bit indicates that MAC transmitter detected an error in the TCP, UDP, or ICMP
IP datagram payload. The transmitter checks the payload length received in the IPv4 or IPv6
header against the actual number of TCP, UDP or ICMP packet bytes received from the
application and issues an error status in case of a mismatch.

RM0090 Rev 17

1173/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Bit 11 LCA: Loss of carrier
When set, this bit indicates that a loss of carrier occurred during frame transmission (that is,
the MII_CRS signal was inactive for one or more transmit clock periods during frame
transmission). This is valid only for the frames transmitted without collision when the MAC
operates in Half-duplex mode.
Bit 10 NC: No carrier
When set, this bit indicates that the Carrier Sense signal form the PHY was not asserted
during transmission.
Bit 9 LCO: Late collision
When set, this bit indicates that frame transmission was aborted due to a collision occurring
after the collision window (64 byte times, including preamble, in MII mode). This bit is not
valid if the Underflow Error bit is set.
Bit 8 EC: Excessive collision
When set, this bit indicates that the transmission was aborted after 16 successive collisions
while attempting to transmit the current frame. If the RD (Disable retry) bit in the MAC
Configuration register is set, this bit is set after the first collision, and the transmission of the
frame is aborted.
Bit 7 VF: VLAN frame
When set, this bit indicates that the transmitted frame was a VLAN-type frame.
Bits 6:3 CC: Collision count
This 4-bit counter value indicates the number of collisions occurring before the frame was
transmitted. The count is not valid when the Excessive collisions bit (TDES0[8]) is set.
Bit 2 ED: Excessive deferral
When set, this bit indicates that the transmission has ended because of excessive deferral
of over 24 288 bit times if the Deferral check (DC) bit in the MAC Control register is set high.
Bit 1 UF: Underflow error
When set, this bit indicates that the MAC aborted the frame because data arrived late from
the RAM memory. Underflow error indicates that the DMA encountered an empty transmit
buffer while transmitting the frame. The transmission process enters the Suspended state
and sets both Transmit underflow (Register 5[5]) and Transmit interrupt (Register 5[0]).
Bit 0 DB: Deferred bit
When set, this bit indicates that the MAC defers before transmission because of the
presence of the carrier. This bit is valid only in Half-duplex mode.

•

TDES1: Transmit descriptor Word1

31 30 29 28 27 26 25
Reserved

24

23 22 21 20 19 18 17 16 15 14 13 12 11 10
TBS2

rw

rw

rw

rw

rw

rw

rw rw rw rw

rw

rw

rw

Reserved

31:29 Reserved, must be kept at reset value.

1174/1747

RM0090 Rev 17

9

8

7

6

5

4

3

2

1

0

rw rw rw rw rw rw

rw

rw

TBS1
rw

rw

rw

rw

rw

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

28:16 TBS2: Transmit buffer 2 size
These bits indicate the second data buffer size in bytes. This field is not valid if TDES0[20] is
set.
15:13 Reserved, must be kept at reset value.
12:0 TBS1: Transmit buffer 1 size
These bits indicate the first data buffer byte size, in bytes. If this field is 0, the DMA ignores
this buffer and uses Buffer 2 or the next descriptor, depending on the value of TCH
(TDES0[20]).

•

TDES2: Transmit descriptor Word2
TDES2 contains the address pointer to the first buffer of the descriptor or it contains
time stamp data.

31 30 29 28 27 26 25

24

23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

TBAP1/TBAP/TTSL
rw

Bits 31:0 TBAP1: Transmit buffer 1 address pointer / Transmit frame time stamp low
These bits have two different functions: they indicate to the DMA the location of data in
memory, and after all data are transferred, the DMA can then use these bits to pass back time
stamp data.
TBAP: When the software makes this descriptor available to the DMA (at the moment that the
OWN bit is set to 1 in TDES0), these bits indicate the physical address of Buffer 1. There is no
limitation on the buffer address alignment. See Host data buffer alignment on page 1165 for further
details on buffer address alignment.
TTSL: Before it clears the OWN bit in TDES0, the DMA updates this field with the 32 least
significant bits of the time stamp captured for the corresponding transmit frame (overwriting
the value for TBAP1). This field has the time stamp only if time stamping is activated for this
frame (see TTSE, TDES0 bit 25) and if the Last segment control bit (LS) in the descriptor is
set.

•

TDES3: Transmit descriptor Word3
TDES3 contains the address pointer either to the second buffer of the descriptor or the
next descriptor, or it contains time stamp data.

31 30 29 28 27 26 25

24

23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

TBAP2/TBAP2/TTSH
rw

RM0090 Rev 17

1175/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Bits 31:0 TBAP2: Transmit buffer 2 address pointer (Next descriptor address) / Transmit frame time
stamp high
These bits have two different functions: they indicate to the DMA the location of data in
memory, and after all data are transferred, the DMA can then use these bits to pass back
time stamp data.
TBAP2: When the software makes this descriptor available to the DMA (at the moment when
the OWN bit is set to 1 in TDES0), these bits indicate the physical address of Buffer 2 when a
descriptor ring structure is used. If the Second address chained (TDES1 [20]) bit is set, this
address contains the pointer to the physical memory where the next descriptor is present. The
buffer address pointer must be aligned to the bus width only when TDES1 [20] is set. (LSBs are
ignored internally.)
TTSH: Before it clears the OWN bit in TDES0, the DMA updates this field with the 32 most
significant bits of the time stamp captured for the corresponding transmit frame (overwriting
the value for TBAP2). This field has the time stamp only if time stamping is activated for this
frame (see TDES0 bit 25, TTSE) and if the Last segment control bit (LS) in the descriptor is
set.

Enhanced Tx DMA descriptors
Enhanced descriptors (enabled with EDFE=1, ETHDMABMR bit 7), must be used if time
stamping is activated (TSE=1, ETH_PTPTSCR bit 0) or if IPv4 checksum offload is
activated (IPCO=1, ETH_MACCR bit 10).
Enhanced descriptors comprise eight 32-bit words, twice the size of normal descriptors.
TDES0, TDES1, TDES2 and TDES3 have the same definitions as for normal transmit
descriptors (refer to Normal Tx DMA descriptors). TDES6 and TDES7 hold the time stamp.
TDES4, TDES5, TDES6 and TDES7 are defined below.
When the Enhanced descriptor mode is selected, the software needs to allocate 32-bytes (8
words) of memory for every descriptor. When time stamping or IPv4 checksum offload are
not being used, the enhanced descriptor format may be disabled and the software can use
normal descriptors with the default size of 16 bytes.

1176/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
Figure 380. Enhanced transmit descriptor


4$%3 
4$%3 
4$%3 

4$%3 

/
7
.


#TRL
;=

2ESERVED
;=

4
4 2ES
3 
%

#TRL
;=

4
2ESERVED 4
;=
3
3

"UFFER  BYTE COUNT
;=

3TATUS ;=

2ESERVED
;=

"UFFER  BYTE COUNT
;=

"UFFER  ADDRESS ;=

"UFFER  ADDRESS ;= OR .EXT DESCRIPTOR ADDRESS ;=

4$%3 

2ESERVED

4$%3 

2ESERVED

4$%3 

4IME STAMP LOW ;=

4$%3 

4IME STAMP HIGH ;=

AIB

•

TDES4: Transmit descriptor Word4
Reserved

•

TDES5: Transmit descriptor Word5
Reserved

•

TDES6: Transmit descriptor Word6

31 30 29 28 27 26 25

24

23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

TTSL
rw

Bits 31:0 TTSL: Transmit frame time stamp low
This field is updated by DMA with the 32 least significant bits of the time stamp captured
for the corresponding transmit frame. This field has the time stamp only if the Last segment
control bit (LS) in the descriptor is set.

RM0090 Rev 17

1177/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller
•

RM0090

TDES7: Transmit descriptor Word7

31 30 29 28 27 26 25

24

23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

TTSH
rw

Bits 31:0 TTSH: Transmit frame time stamp high
This field is updated by DMA with the 32 most significant bits of the time stamp captured for
the corresponding transmit frame. This field has the time stamp only if the Last segment control
bit (LS) in the descriptor is set.

33.6.8

Rx DMA configuration
The Receive DMA engine’s reception sequence is illustrated in Figure 381 and described
below:
1.

The CPU sets up Receive descriptors (RDES0-RDES3) and sets the OWN bit
(RDES0[31]).

2.

Once the SR (ETH_DMAOMR register[1]) bit is set, the DMA enters the Run state.
While in the Run state, the DMA polls the receive descriptor list, attempting to acquire
free descriptors. If the fetched descriptor is not free (is owned by the CPU), the DMA
enters the Suspend state and jumps to Step 9.

3.

The DMA decodes the receive data buffer address from the acquired descriptors.

4.

Incoming frames are processed and placed in the acquired descriptor’s data buffers.

5.

When the buffer is full or the frame transfer is complete, the Receive engine fetches the
next descriptor.

6.

If the current frame transfer is complete, the DMA proceeds to step 7. If the DMA does
not own the next fetched descriptor and the frame transfer is not complete (EOF is not
yet transferred), the DMA sets the Descriptor error bit in RDES0 (unless flushing is
disabled). The DMA closes the current descriptor (clears the OWN bit) and marks it as
intermediate by clearing the Last segment (LS) bit in the RDES1 value (marks it as last
descriptor if flushing is not disabled), then proceeds to step 8. If the DMA owns the next
descriptor but the current frame transfer is not complete, the DMA closes the current
descriptor as intermediate and returns to step 4.

7.

If IEEE 1588 time stamping is enabled, the DMA writes the time stamp (if available) to
the current descriptor’s RDES2 and RDES3. It then takes the received frame’s status
and writes the status word to the current descriptor’s RDES0, with the OWN bit cleared
and the Last segment bit set.

8.

The Receive engine checks the latest descriptor’s OWN bit. If the CPU owns the
descriptor (OWN bit is at 0) the Receive buffer unavailable bit (in ETH_DMASR
register[7]) is set and the DMA Receive engine enters the Suspended state (step 9). If
the DMA owns the descriptor, the engine returns to step 4 and awaits the next frame.

9.

Before the Receive engine enters the Suspend state, partial frames are flushed from
the Receive FIFO (you can control flushing using bit 24 in the ETH_DMAOMR
register).

10. The Receive DMA exits the Suspend state when a Receive Poll demand is given or the
start of next frame is available from the Receive FIFO. The engine proceeds to step 2
and re-fetches the next descriptor.

1178/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
The DMA does not acknowledge accepting the status until it has completed the time stamp
write-back and is ready to perform status write-back to the descriptor. If software has
enabled time stamping through CSR, when a valid time stamp value is not available for the
frame (for example, because the receive FIFO was full before the time stamp could be
written to it), the DMA writes all ones to RDES2 and RDES3. Otherwise (that is, if time
stamping is not enabled), RDES2 and RDES3 remain unchanged.
Figure 381. Receive DMA operation
Start RxDMA

Stop RxDMA

(Re-)Fetch next
descriptor

Poll demand /
new frame available

(AHB)
error?

RxDMA suspended

Yes

No

Yes
Frame transfer
complete?

Yes

Start

No

Own bit set?

No

Yes

Flush disabled ?

Frame data
available ?

No

Yes

Flush the
remaining frame

Write data to buffer(s)

No

Wait for frame data

(AHB)
error?

Yes

No
Fetch next descriptor

(AHB)
error?

Yes

No
Flush
disabled ?
No

Set descriptor error

No

Yes

Own bit set
for next desc?

Frame transfer
complete?

No

Yes

Yes

Close RDES0 as
intermediate descriptor

Time stamp
present?

Yes

Write time stamp to
RDES2 & RDES3

No
Close RDES0 as last
descriptor

No

(AHB)
error?

No

(AHB)
error?

Yes

Yes

ai15643

RM0090 Rev 17

1179/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Receive descriptor acquisition
The receive engine always attempts to acquire an extra descriptor in anticipation of an
incoming frame. Descriptor acquisition is attempted if any of the following conditions is/are
satisfied:
•

The receive Start/Stop bit (ETH_DMAOMR register[1]) has been set immediately after
the DMA has been placed in the Run state.

•

The data buffer of the current descriptor is full before the end of the frame currently
being transferred

•

The controller has completed frame reception, but the current receive descriptor has
not yet been closed.

•

The receive process has been suspended because of a CPU-owned buffer
(RDES0[31] = 0) and a new frame is received.

•

A Receive poll demand has been issued.

Receive frame processing
The MAC transfers the received frames to the STM32F4xx memory only when the frame
passes the address filter and the frame size is greater than or equal to the configurable
threshold bytes set for the Receive FIFO, or when the complete frame is written to the FIFO
in Store-and-forward mode. If the frame fails the address filtering, it is dropped in the MAC
block itself (unless Receive All ETH_MACFFR [31] bit is set). Frames that are shorter than
64 bytes, because of collision or premature termination, can be purged from the Receive
FIFO. After 64 (configurable threshold) bytes have been received, the DMA block begins
transferring the frame data to the receive buffer pointed to by the current descriptor. The
DMA sets the first descriptor (RDES0[9]) after the DMA AHB Interface becomes ready to
receive a data transfer (if DMA is not fetching transmit data from the memory), to delimit the
frame. The descriptors are released when the OWN (RDES0[31]) bit is reset to 0, either as
the data buffer fills up or as the last segment of the frame is transferred to the receive buffer.
If the frame is contained in a single descriptor, both the last descriptor (RDES0[8]) and first
descriptor (RDES0[9]) bits are set. The DMA fetches the next descriptor, sets the last
descriptor (RDES0[8]) bit, and releases the RDES0 status bits in the previous frame
descriptor. Then the DMA sets the receive interrupt bit (ETH_DMASR register [6]). The
same process repeats unless the DMA encounters a descriptor flagged as being owned by
the CPU. If this occurs, the receive process sets the receive buffer unavailable bit
(ETH_DMASR register[7]) and then enters the Suspend state. The position in the receive
list is retained.

Receive process suspended
If a new receive frame arrives while the receive process is in Suspend state, the DMA refetches the current descriptor in the STM32F4xx memory. If the descriptor is now owned by
the DMA, the receive process re-enters the Run state and starts frame reception. If the
descriptor is still owned by the host, by default, the DMA discards the current frame at the
top of the Rx FIFO and increments the missed frame counter. If more than one frame is
stored in the Rx FIFO, the process repeats. The discarding or flushing of the frame at the
top of the Rx FIFO can be avoided by setting the DMA Operation mode register bit 24
(DFRF). In such conditions, the receive process sets the receive buffer unavailable status
bit and returns to the Suspend state.

1180/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Normal Rx DMA descriptors
The normal receive descriptor structure consists of four 32-bit words (16 bytes). These are
shown in Figure 382. The bit descriptions of RDES0, RDES1, RDES2 and RDES3 are given
below.
Note that enhanced descriptors must be used if time stamping is activated (TSE=1,
ETH_PTPTSCR bit 0) or if IPv4 checksum offload is activated (IPCO=1, ETH_MACCR bit
10).
Figure 382. Normal Rx DMA descriptor structure
31

0

O
RDES 0 W
N
RDES 1 CT Reserved
RL [30:29]

Status [30:0]
Buffer 2 byte count
[28:16]

CTRL Res.
[15:14]

Buffer 1 byte count
[12:0]

Buffer 1 address [31:0]

RDES 2

Buffer 2 address [31:0] or Next descriptor address [31:0]

RDES 3

ai15644

•

RDES0: Receive descriptor Word0

1

0
PCE/ESA

LCO

2

CE

IPHCE/TSV

3

DE

LS

4

RE

FS

FL

5

FT

6

RWT

7

LE

8

OE

9

SAF

23 22 21 20 19 18 17 16 15 14 13 12 11 10

ES

24

DE

AFM

OWN

31 30 29 28 27 26 25

VLAN

RDES0 contains the received frame status, the frame length and the descriptor
ownership information.

rw

Bit 31 OWN: Own bit
When set, this bit indicates that the descriptor is owned by the DMA of the MAC Subsystem.
When this bit is reset, it indicates that the descriptor is owned by the Host. The DMA clears this bit
either when it completes the frame reception or when the buffers that are associated with this
descriptor are full.
Bit 30 AFM: Destination address filter fail
When set, this bit indicates a frame that failed the DA filter in the MAC Core.
Bits 29:16 FL: Frame length
These bits indicate the byte length of the received frame that was transferred to host memory
(including CRC). This field is valid only when last descriptor (RDES0[8]) is set and descriptor error
(RDES0[14]) is reset.
This field is valid when last descriptor (RDES0[8]) is set. When the last descriptor and error
summary bits are not set, this field indicates the accumulated number of bytes that have been
transferred for the current frame.

RM0090 Rev 17

1181/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Bit 15 ES: Error summary
Indicates the logical OR of the following bits:
RDES0[1]: CRC error
RDES0[3]: Receive error
RDES0[4]: Watchdog timeout
RDES0[6]: Late collision
RDES0[7]: Giant frame (This is not applicable when RDES0[7] indicates an IPV4 header
checksum error.)
RDES0[11]: Overflow error
RDES0[14]: Descriptor error.
This field is valid only when the last descriptor (RDES0[8]) is set.
Bit 14 DE: Descriptor error
When set, this bit indicates a frame truncation caused by a frame that does not fit within the current
descriptor buffers, and that the DMA does not own the next descriptor. The frame is truncated.
This field is valid only when the last descriptor (RDES0[8]) is set.
Bit 13 SAF: Source address filter fail
When set, this bit indicates that the SA field of frame failed the SA filter in the MAC Core.
Bit 12 LE: Length error
When set, this bit indicates that the actual length of the received frame does not match the value in
the Length/ Type field. This bit is valid only when the Frame type (RDES0[5]) bit is reset.
Bit 11 OE: Overflow error
When set, this bit indicates that the received frame was damaged due to buffer overflow.
Bit 10 VLAN: VLAN tag
When set, this bit indicates that the frame pointed to by this descriptor is a VLAN frame tagged
by the MAC core.
Bit 9 FS: First descriptor
When set, this bit indicates that this descriptor contains the first buffer of the frame. If the size of the
first buffer is 0, the second buffer contains the beginning of the frame. If the size of the second buffer
is also 0, the next descriptor contains the beginning of the frame.
Bit 8 LS: Last descriptor
When set, this bit indicates that the buffers pointed to by this descriptor are the last buffers of
the frame.
Bit 7 IPHCE/TSV: IPv header checksum error / time stamp valid
If IPHCE is set, it indicates an error in the IPv4 or IPv6 header. This error can be due to inconsistent
Ethernet Type field and IP header Version field values, a header checksum mismatch in IPv4, or an
Ethernet frame lacking the expected number of IP header bytes. This bit can take on special
meaning as specified in Table 193.
If enhanced descriptor format is enabled (EDFE=1, bit 7 of ETH_DMABMR), this bit takes on
the TSV function (otherwise it is IPHCE). When TSV is set, it indicates that a snapshot of the
timestamp is written in descriptor words 6 (RDES6) and 7 (RDES7). TSV is valid only when
the Last descriptor bit (RDES0[8]) is set.
Bit 6 LCO: Late collision
When set, this bit indicates that a late collision has occurred while receiving the frame in Halfduplex mode.

1182/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Bit 5 FT: Frame type
When set, this bit indicates that the Receive frame is an Ethernet-type frame (the LT field is greater
than or equal to 0x0600). When this bit is reset, it indicates that the received frame is an
IEEE802.3 frame. This bit is not valid for Runt frames less than 14 bytes. When the normal
descriptor format is used (ETH_DMABMR EDFE=0), FT can take on special meaning as
specified in Table 193.
Bit 4 RWT: Receive watchdog timeout
When set, this bit indicates that the Receive watchdog timer has expired while receiving the
current frame and the current frame is truncated after the watchdog timeout.
Bit 3 RE: Receive error
When set, this bit indicates that the RX_ERR signal is asserted while RX_DV is asserted
during frame reception.
Bit 2 DE: Dribble bit error
When set, this bit indicates that the received frame has a non-integer multiple of bytes (odd
nibbles). This bit is valid only in MII mode.
Bit 1 CE: CRC error
When set, this bit indicates that a cyclic redundancy check (CRC) error occurred on the
received frame. This field is valid only when the last descriptor (RDES0[8]) is set.
Bit 0 PCE/ESA: Payload checksum error / extended status available
When set, it indicates that the TCP, UDP or ICMP checksum the core calculated does not
match the received encapsulated TCP, UDP or ICMP segment’s Checksum field. This bit is
also set when the received number of payload bytes does not match the value indicated in
the Length field of the encapsulated IPv4 or IPv6 datagram in the received Ethernet frame.
This bit can take on special meaning as specified in Table 193.
If the enhanced descriptor format is enabled (EDFE=1, bit 7 in ETH_DMABMR), this bit takes
on the ESA function (otherwise it is PCE). When ESA is set, it indicates that the extended
status is available in descriptor word 4 (RDES4). ESA is valid only when the last descriptor bit
(RDES0[8]) is set.

Bits 5, 7, and 0 reflect the conditions discussed in Table 193.

RM0090 Rev 17

1183/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Table 193. Receive descriptor 0 - encoding for bits 7, 5 and 0 (normal descriptor
format only, EDFE=0)
Bit 5:
frame
type

•

Bit 7: IPC Bit 0: payload
checksum
checksum
error
error

0

0

0

IEEE 802.3 Type frame (Length field value is less than
0x0600.)

1

0

0

IPv4/IPv6 Type frame, no checksum error detected

1

0

1

IPv4/IPv6 Type frame with a payload checksum error (as
described for PCE) detected

1

1

0

IPv4/IPv6 Type frame with an IP header checksum error (as
described for IPC CE) detected

1

1

1

IPv4/IPv6 Type frame with both IP header and payload checksum
errors detected

0

0

1

IPv4/IPv6 Type frame with no IP header checksum error and the
payload check bypassed, due to an unsupported payload

0

1

1

A Type frame that is neither IPv4 or IPv6 (the checksum offload
engine bypasses checksum completely.)

0

1

0

Reserved

RDES1: Receive descriptor Word1

rw

RBS2
rw

rw

rw

rw

rw

rw

rw rw rw rw

rw

rw

rw

rw

rw

Reserved

rw

23 22 21 20 19 18 17 16 15 14 13 12 11 10
RER

rw

24

RCH

DIC

31 30 29 28 27 26 25
RBS2

Frame status

9

8

7

6

5

4

3

2

1

0

rw rw rw rw rw rw

rw

rw

RBS
rw

rw

rw

rw

rw

Bit 31 DIC: Disable interrupt on completion
When set, this bit prevents setting the Status register’s RS bit (CSR5[6]) for the received frame
ending in the buffer indicated by this descriptor. This, in turn, disables the assertion of the interrupt to
Host due to RS for that frame.
Bits 30:29 Reserved, must be kept at reset value.
Bits 28:16 RBS2: Receive buffer 2 size
These bits indicate the second data buffer size, in bytes. The buffer size must be a multiple of 4,
8, or 16, depending on the bus widths (32, 64 or 128, respectively), even if the value of RDES3
(buffer2 address pointer) is not aligned to bus width. If the buffer size is not an appropriate multiple
of 4, 8 or 16, the resulting behavior is undefined. This field is not valid if RDES1 [14] is set.
Bit 15 RER: Receive end of ring
When set, this bit indicates that the descriptor list reached its final descriptor. The DMA returns to
the base address of the list, creating a descriptor ring.

1184/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Bit 14 RCH: Second address chained
When set, this bit indicates that the second address in the descriptor is the next descriptor address
rather than the second buffer address. When this bit is set, RBS2 (RDES1[28:16]) is a “don’t
care” value. RDES1[15] takes precedence over RDES1[14].
Bit 13 Reserved, must be kept at reset value.
Bits 12:0 RBS1: Receive buffer 1 size
Indicates the first data buffer size in bytes. The buffer size must be a multiple of 4, 8 or 16,
depending upon the bus widths (32, 64 or 128), even if the value of RDES2 (buffer1 address
pointer) is not aligned. When the buffer size is not a multiple of 4, 8 or 16, the resulting behavior is
undefined. If this field is 0, the DMA ignores this buffer and uses Buffer 2 or next descriptor
depending on the value of RCH (bit 14).

•

RDES2: Receive descriptor Word2
RDES2 contains the address pointer to the first data buffer in the descriptor, or it
contains time stamp data.

31 30 29 28 27 26 25

24

23 22 21 20 19 18 17 16 15 14 13 12 11 10

rw

rw

rw

9

8

7

6

5

4

3

2

1

0

rw

rw

rw rw rw rw rw rw

rw

rw

RBP1 / RTSL
rw

rw

rw

rw

rw

rw

rw rw rw rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 RBAP1 / RTSL: Receive buffer 1 address pointer / Receive frame time stamp low
These bits take on two different functions: the application uses them to indicate to the DMA
where to store the data in memory, and then after transferring all the data the DMA may use
these bits to pass back time stamp data.
RBAP1: When the software makes this descriptor available to the DMA (at the moment that
the OWN bit is set to 1 in RDES0), these bits indicate the physical address of Buffer 1. There are
no limitations on the buffer address alignment except for the following condition: the DMA uses the
configured value for its address generation when the RDES2 value is used to store the start of
frame. Note that the DMA performs a write operation with the RDES2[3/2/1:0] bits as 0 during the
transfer of the start of frame but the frame data is shifted as per the actual Buffer address pointer.
The DMA ignores RDES2[3/2/1:0] (corresponding to bus width of 128/64/32) if the address pointer
is to a buffer where the middle or last part of the frame is stored.
RTSL: Before it clears the OWN bit in RDES0, the DMA updates this field with the 32 least
significant bits of the time stamp captured for the corresponding receive frame (overwriting
the value for RBAP1). This field has the time stamp only if time stamping is activated for this
frame and if the Last segment control bit (LS) in the descriptor is set.

RM0090 Rev 17

1185/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller
•

RM0090

RDES3: Receive descriptor Word3
RDES3 contains the address pointer either to the second data buffer in the descriptor
or to the next descriptor, or it contains time stamp data.

31 30 29 28 27 26 25

24

23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw rw rw rw rw rw

rw

rw

RBP2 / RTSH
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw rw rw rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 RBAP2 / RTSH: Receive buffer 2 address pointer (next descriptor address) / Receive frame
time stamp high
These bits take on two different functions: the application uses them to indicate to the DMA
the location of where to store the data in memory, and then after transferring all the data the
DMA may use these bits to pass back time stamp data.
RBAP1: When the software makes this descriptor available to the DMA (at the moment that
the OWN bit is set to 1 in RDES0), these bits indicate the physical address of buffer 2 when a
descriptor ring structure is used. If the second address chained (RDES1 [24]) bit is set, this address
contains the pointer to the physical memory where the next descriptor is present. If RDES1 [24] is
set, the buffer (next descriptor) address pointer must be bus width-aligned (RDES3[3, 2, or
1:0] = 0, corresponding to a bus width of 128, 64 or 32. LSBs are ignored internally.)
However, when RDES1 [24] is reset, there are no limitations on the RDES3 value, except for the
following condition: the DMA uses the configured value for its buffer address generation when the
RDES3 value is used to store the start of frame. The DMA ignores RDES3[3, 2, or 1:0]
(corresponding to a bus width of 128, 64 or 32) if the address pointer is to a buffer where the
middle or last part of the frame is stored.
RTSH: Before it clears the OWN bit in RDES0, the DMA updates this field with the 32 most
significant bits of the time stamp captured for the corresponding receive frame (overwriting
the value for RBAP2). This field has the time stamp only if time stamping is activated and if
the Last segment control bit (LS) in the descriptor is set.

Enhanced Rx DMA descriptors format with IEEE1588 time stamp
Enhanced descriptors (enabled with EDFE=1, ETHDMABMR bit 7), must be used if time
stamping is activated (TSE=1, ETH_PTPTSCR bit 0) or if IPv4 checksum offload is
activated (IPCO=1, ETH_MACCR bit 10).
Enhanced descriptors comprise eight 32-bit words, twice the size of normal descriptors.
RDES0, RDES1, RDES2 and RDES3 have the same definitions as for normal receive
descriptors (refer to Normal Rx DMA descriptors). RDES4 contains extended status while
RDES6 and RDES7 hold the time stamp. RDES4, RDES5, RDES6 and RDES7 are defined
below.
When the Enhanced descriptor mode is selected, the software needs to allocate 32 bytes (8
words) of memory for every descriptor. When time stamping or IPv4 checksum offload are
not being used, the enhanced descriptor format may be disabled and the software can use
normal descriptors with the default size of 16 bytes.

1186/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
Figure 383. Enhanced receive descriptor field format with IEEE1588 time stamp
enabled
31

0

O
RDES 0 W
N
RDES 1 CT Reserved
RL [30:29]

Status [30:0]
Buffer 2 byte count
[28:16]

RDES 2

CTRL Res.
[15:14]

Buffer 1 byte count
[12:0]

Buffer 1 address [31:0]

RDES 3

Buffer 2 address [31:0] or Next descriptor address [31:0]

RDES 4

Extended Status [31:0]

RDES 5

Reserved

RDES 6

Time stamp low [31:0]

Time stamp high [31:0]

RDES 7

ai17104

•

RDES4: Receive descriptor Word4

rw

rw

PMT
rw

rw

rw

rw

7

6

5

4

3

IPPE

PV

8

IPHE

Reserved

9

IPCB

23 22 21 20 19 18 17 16 15 14 13 12 11 10

IPV4PR

24

PFT

31 30 29 28 27 26 25

IPV6PR

The extended status, shown below, is valid only when there is status related to IPv4
checksum or time stamp available as indicated by bit 0 in RDES0.
2

1

0

IPPT

rw rw rw rw rw rw

rw

rw

Bits 31:14 Reserved, must be kept at reset value.
Bit 13 PV: PTP version
When set, indicates that the received PTP message uses the IEEE 1588 version 2 format.
When cleared, it uses version 1 format. This is valid only if the message type is non-zero.
Bit 12 PFT: PTP frame type
When set, this bit indicates that the PTP message is sent directly over Ethernet. When this bit
is cleared and the message type is non-zero, it indicates that the PTP message is sent over
UDP-IPv4 or UDP-IPv6. The information on IPv4 or IPv6 can be obtained from bits 6 and 7.

RM0090 Rev 17

1187/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Bits 11:8 PMT: PTP message type
These bits are encoded to give the type of the message received.
– 0000: No PTP message received
– 0001: SYNC (all clock types)
– 0010: Follow_Up (all clock types)
– 0011: Delay_Req (all clock types)
– 0100: Delay_Resp (all clock types)
– 0101: Pdelay_Req (in peer-to-peer transparent clock) or Announce (in ordinary or boundary
clock)
– 0110: Pdelay_Resp (in peer-to-peer transparent clock) or Management (in ordinary or
boundary clock)
– 0111: Pdelay_Resp_Follow_Up (in peer-to-peer transparent clock) or Signaling (for ordinary
or boundary clock)
– 1xxx - Reserved
Bit 7 IPV6PR: IPv6 packet received
When set, this bit indicates that the received packet is an IPv6 packet.
Bit 6 IPV4PR: IPv4 packet received
When set, this bit indicates that the received packet is an IPv4 packet.
Bit 5 IPCB: IP checksum bypassed
When set, this bit indicates that the checksum offload engine is bypassed.
Bit 4 IPPE: IP payload error
When set, this bit indicates that the 16-bit IP payload checksum (that is, the TCP, UDP, or
ICMP checksum) that the core calculated does not match the corresponding checksum field
in the received segment. It is also set when the TCP, UDP, or ICMP segment length does not
match the payload length value in the IP Header field.
Bit 3 IPHE: IP header error
When set, this bit indicates either that the 16-bit IPv4 header checksum calculated by the
core does not match the received checksum bytes, or that the IP datagram version is not
consistent with the Ethernet Type value.
Bits 2:0 IPPT: IP payload type

if IPv4 checksum offload is activated (IPCO=1, ETH_MACCR bit 10), these bits
–
–
–
–
–

•

indicate the type of payload encapsulated in the IP datagram. These bits are ‘00’ if there is an IP
header error or fragmented IP.
000: Unknown or did not process IP payload
001: UDP
010: TCP
011: ICMP
1xx: Reserved

RDES5: Receive descriptor Word5
Reserved.

•

RDES6: Receive descriptor Word6
The table below describes the fields that have different meaning for RDES6 when the
receive descriptor is closed and time stamping is enabled.

1188/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

31 30 29 28 27 26 25

24

23 22 21 20 19 18 17 16 15 14 13 12 11 10

rw

rw

rw

9

8

7

6

5

4

3

2

1

0

rw

rw

rw rw rw rw rw rw

rw

rw

RTSL
rw

rw

rw

rw

rw

rw

rw rw rw rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

.

Bits 31:0 RTSL: Receive frame time stamp low
The DMA updates this field with the 32 least significant bits of the time stamp captured for the
corresponding receive frame. The DMA updates this field only for the last descriptor of the receive
frame indicated by last descriptor status bit (RDES0[8]). When this field and the RTSH field in
RDES7 show all ones, the time stamp must be treated as corrupt.

•

RDES7: Receive descriptor Word7
The table below describes the fields that have a different meaning for RDES7 when the
receive descriptor is closed and time stamping is enabled.

31 30 29 28 27 26 25

24

23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw rw rw rw rw rw

rw

rw

RTSH
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw rw rw rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

.

Bits 31:0 RTSH: Receive frame time stamp high
The DMA updates this field with the 32 most significant bits of the time stamp captured for the
corresponding receive frame. The DMA updates this field only for the last descriptor of the receive
frame indicated by last descriptor status bit (RDES0[8]).
When this field and RDES7’s RTSL field show all ones, the time stamp must be treated as
corrupt.

33.6.9

DMA interrupts
Interrupts can be generated as a result of various events. The ETH_DMASR register
contains all the bits that might cause an interrupt. The ETH_DMAIER register contains an
enable bit for each of the events that can cause an interrupt.
There are two groups of interrupts, Normal and Abnormal, as described in the ETH_DMASR
register. Interrupts are cleared by writing a 1 to the corresponding bit position. When all the
enabled interrupts within a group are cleared, the corresponding summary bit is cleared. If
the MAC core is the cause for assertion of the interrupt, then any of the TSTS or PMTS bits
in the ETH_DMASR register is set high.
Interrupts are not queued and if the interrupt event occurs before the driver has responded
to it, no additional interrupts are generated. For example, the Receive Interrupt bit
(ETH_DMASR register [6]) indicates that one or more frames were transferred to the
STM32F4xx buffer. The driver must scan all descriptors, from the last recorded position to
the first one owned by the DMA.
An interrupt is generated only once for simultaneous, multiple events. The driver must scan
the ETH_DMASR register for the cause of the interrupt. The interrupt is not generated again
unless a new interrupting event occurs, after the driver has cleared the appropriate bit in the
ETH_DMASR register. For example, the controller generates a Receive interrupt
(ETH_DMASR register[6]) and the driver begins reading the ETH_DMASR register. Next,
receive buffer unavailable (ETH_DMASR register[7]) occurs. The driver clears the Receive
interrupt. Even then, a new interrupt is generated, due to the active or pending Receive
buffer unavailable interrupt.

RM0090 Rev 17

1189/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Figure 384. Interrupt scheme
TS
TBUS
TBUIE

MMCI

AND

TIE

PMTI
TSTI

AND
RS
RIE

OR

AND

AND

ERS
ERIE

NIS
NISE

AND
OR

Interrupt

FBES
FBEIE

TPSS
TPSSIE

AND
TJTS

AND
ROS
ROIE

TUS
TUIE

AISE

AND

AND

AND
RBU

RWTS
RWTIE

TJTIE

AIS
AND

RBUIE

OR
AND

RPSS
RPSSIE

AND

AND

ETS
ETIE

AND
AI15646

33.7

Ethernet interrupts
The Ethernet controller has two interrupt vectors: one dedicated to normal Ethernet
operations and the other, used only for the Ethernet wakeup event (with wakeup frame or
Magic Packet detection) when it is mapped on EXTI lIne19.
The first Ethernet vector is reserved for interrupts generated by the MAC and the DMA as
listed in the MAC interrupts and DMA interrupts sections.
The second vector is reserved for interrupts generated by the PMT on wakeup events. The
mapping of a wakeup event on EXTI line19 causes the STM32F4xx to exit the low-power
mode, and generates an interrupt.
When an Ethernet wakeup event mapped on EXTI Line19 occurs and the MAC PMT
interrupt is enabled and the EXTI Line19 interrupt, with detection on rising edge, is also
enabled, both interrupts are generated.
A watchdog timer (see ETH_DMARSWTR register) is given for flexible control of the RS bit
(ETH_DMASR register). When this watchdog timer is programmed with a non-zero value, it
gets activated as soon as the RxDMA completes a transfer of a received frame to system
memory without asserting the Receive Status because it is not enabled in the corresponding
Receive descriptor (RDES1[31]). When this timer runs out as per the programmed value,
the RS bit is set and the interrupt is asserted if the corresponding RIE is enabled in the
ETH_DMAIER register. This timer is disabled before it runs out, when a frame is transferred
to memory and the RS is set because it is enabled for that descriptor.

1190/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Note:

Reading the PMT control and status register automatically clears the Wakeup Frame
Received and Magic Packet Received PMT interrupt flags. However, since the registers for
these flags are in the CLK_RX domain, there may be a significant delay before this update
is visible by the firmware. The delay is especially long when the RX clock is slow (in 10 Mbit
mode) and when the AHB bus is high-frequency.
Since interrupt requests from the PMT to the CPU are based on the same registers in the
CLK_RX domain, the CPU may spuriously call the interrupt routine a second time even after
reading PMT_CSR. Thus, it may be necessary that the firmware polls the Wakeup Frame
Received and Magic Packet Received bits and exits the interrupt service routine only when
they are found to be at ‘0’.

33.8

Ethernet register descriptions
The peripheral registers can be accessed by bytes (8-bit), half-words (16-bit) or words (32bits).

33.8.1

MAC register description
Ethernet MAC configuration register (ETH_MACCR)
Address offset: 0x0000
Reset value: 0x0000 8000

rw

6

5

rw

rw

rw

rw

1

0
Reserved

rw

2
RE

rw

3
TE

rw

4
DC

rw

7

BL

rw

8

APCS

DM

IPCO

rw

9
RD

LM

rw

FES

rw

ROD

CSD

rw

IFG

Reserved

JD

rw

Reserved

WD

rw

Reserved

Reserved

CSTF

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

Reserved

The MAC configuration register is the operation mode register of the MAC. It establishes
receive and transmit operating modes.

Bits 31:26 Reserved, must be kept at reset value.
CSTF: CRC stripping for Type frames
Bit 25 When set, the last 4 bytes (FCS) of all frames of Ether type (type field greater than
0x0600) will be stripped and dropped before forwarding the frame to the application.
Bit 24Reserved, must be kept at reset value.
Bit 23 WD: Watchdog disable
When this bit is set, the MAC disables the watchdog timer on the receiver, and can receive
frames of up to 16 384 bytes.
When this bit is reset, the MAC allows no more than 2 048 bytes of the frame being received
and cuts off any bytes received after that.
Bit 22 JD: Jabber disable
When this bit is set, the MAC disables the jabber timer on the transmitter, and can transfer
frames of up to 16 384 bytes.
When this bit is reset, the MAC cuts off the transmitter if the application sends out more than
2 048 bytes of data during transmission.
Bits 21:20 Reserved, must be kept at reset value.

RM0090 Rev 17

1191/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Bits 19:17 IFG: Interframe gap
These bits control the minimum interframe gap between frames during transmission.
000: 96 bit times
001: 88 bit times
010: 80 bit times
….
111: 40 bit times
Note: In Half-duplex mode, the minimum IFG can be configured for 64 bit times (IFG = 100)
only. Lower values are not considered.
Bit 16 CSD: Carrier sense disable
When set high, this bit makes the MAC transmitter ignore the MII CRS signal during frame
transmission in Half-duplex mode. No error is generated due to Loss of Carrier or No Carrier
during such transmission.
When this bit is low, the MAC transmitter generates such errors due to Carrier Sense and
even aborts the transmissions.
Bit 15 Reserved, must be kept at reset value.
Bit 14 FES: Fast Ethernet speed
Indicates the speed in Fast Ethernet (MII) mode:
0: 10 Mbit/s
1: 100 Mbit/s
Bit 13 ROD: Receive own disable
When this bit is set, the MAC disables the reception of frames in Half-duplex mode.
When this bit is reset, the MAC receives all packets that are given by the PHY while
transmitting.
This bit is not applicable if the MAC is operating in Full-duplex mode.
Bit 12 LM: Loopback mode
When this bit is set, the MAC operates in loopback mode at the MII. The MII receive clock
input (RX_CLK) is required for the loopback to work properly, as the transmit clock is not
looped-back internally.
Bit 11 DM: Duplex mode
When this bit is set, the MAC operates in a Full-duplex mode where it can transmit and
receive simultaneously.
Bit 10 IPCO: IPv4 checksum offload
When set, this bit enables IPv4 checksum checking for received frame payloads'
TCP/UDP/ICMP headers. When this bit is reset, the checksum offload function in the
receiver is disabled and the corresponding PCE and IP HCE status bits (see Table 190 on
page 1146) are always cleared.
Bit 9 RD: Retry disable
When this bit is set, the MAC attempts only 1 transmission. When a collision occurs on the
MII, the MAC ignores the current frame transmission and reports a Frame Abort with
excessive collision error in the transmit frame status.
When this bit is reset, the MAC attempts retries based on the settings of BL.
Note: This bit is applicable only in the Half-duplex mode.
Bit 8 Reserved, must be kept at reset value.

1192/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Bit 7 APCS: Automatic pad/CRC stripping
When this bit is set, the MAC strips the Pad/FCS field on incoming frames only if the
length’s field value is less than or equal to 1 500 bytes. All received frames with length field
greater than or equal to 1 501 bytes are passed on to the application without stripping the
Pad/FCS field.
When this bit is reset, the MAC passes all incoming frames unmodified.
Bits 6:5 BL: Back-off limit
The Back-off limit determines the random integer number (r) of slot time delays (4 096 bit
times for 1000 Mbit/s and 512 bit times for 10/100 Mbit/s) the MAC waits before
rescheduling a transmission attempt during retries after a collision.
Note: This bit is applicable only to Half-duplex mode.
00: k = min (n, 10)
01: k = min (n, 8)
10: k = min (n, 4)
11: k = min (n, 1),
where n = retransmission attempt. The random integer r takes the value in the range 0 ≤ r <
2k
Bit 4 DC: Deferral check
When this bit is set, the deferral check function is enabled in the MAC. The MAC issues a
Frame Abort status, along with the excessive deferral error bit set in the transmit frame
status when the transmit state machine is deferred for more than 24 288 bit times in 10/100Mbit/s mode. Deferral begins when the transmitter is ready to transmit, but is prevented
because of an active CRS (carrier sense) signal on the MII. Defer time is not cumulative. If
the transmitter defers for 10 000 bit times, then transmits, collides, backs off, and then has
to defer again after completion of back-off, the deferral timer resets to 0 and restarts.
When this bit is reset, the deferral check function is disabled and the MAC defers until the
CRS signal goes inactive. This bit is applicable only in Half-duplex mode.
Bit 3 TE: Transmitter enable
When this bit is set, the transmit state machine of the MAC is enabled for transmission on
the MII. When this bit is reset, the MAC transmit state machine is disabled after the
completion of the transmission of the current frame, and does not transmit any further
frames.
Bit 2 RE: Receiver enable
When this bit is set, the receiver state machine of the MAC is enabled for receiving frames
from the MII. When this bit is reset, the MAC receive state machine is disabled after the
completion of the reception of the current frame, and will not receive any further frames from
the MII.
Bits 1:0 Reserved, must be kept at reset value.

RM0090 Rev 17

1193/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Ethernet MAC frame filter register (ETH_MACFFR)
Address offset: 0x0004
Reset value: 0x0000 0000

rw

3

2

1

0

HU

rw

4

PM

rw

5

HM

rw

Reserved

rw

6

PAM

SAF

SAIF

rw

RA

7

DAIF

8

BFD

9

PCF

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
HPF

The MAC frame filter register contains the filter controls for receiving frames. Some of the
controls from this register go to the address check block of the MAC, which performs the
first level of address filtering. The second level of filtering is performed on the incoming
frame, based on other controls such as pass bad frames and pass control frames.

rw

rw

rw

rw

rw

rw

Bit 31 RA: Receive all
When this bit is set, the MAC receiver passes all received frames on to the application,
irrespective of whether they have passed the address filter. The result of the SA/DA filtering
is updated (pass or fail) in the corresponding bits in the receive status word. When this bit is
reset, the MAC receiver passes on to the application only those frames that have passed the
SA/DA address filter.
Bits 30:11 Reserved, must be kept at reset value.
Bit 10 HPF: Hash or perfect filter
When this bit is set and if the HM or HU bit is set, the address filter passes frames that
match either the perfect filtering or the hash filtering.
When this bit is cleared and if the HU or HM bit is set, only frames that match the Hash filter
are passed.
Bit 9 SAF: Source address filter
The MAC core compares the SA field of the received frames with the values programmed in
the enabled SA registers. If the comparison matches, then the SAMatch bit in the RxStatus
word is set high. When this bit is set high and the SA filter fails, the MAC drops the frame.
When this bit is reset, the MAC core forwards the received frame to the application. It also
forwards the updated SA Match bit in RxStatus depending on the SA address comparison.
Bit 8 SAIF: Source address inverse filtering
When this bit is set, the address check block operates in inverse filtering mode for the SA
address comparison. The frames whose SA matches the SA registers are marked as failing
the SA address filter.
When this bit is reset, frames whose SA does not match the SA registers are marked as
failing the SA address filter.

1194/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Bits 7:6 PCF: Pass control frames
These bits control the forwarding of all control frames (including unicast and multicast
PAUSE frames). Note that the processing of PAUSE control frames depends only on RFCE
in Flow Control Register[2].
00: MAC prevents all control frames from reaching the application
01: MAC forwards all control frames to application except Pause control frames
10: MAC forwards all control frames to application even if they fail the address filter
11: MAC forwards control frames that pass the address filter.
These bits control the forwarding of all control frames (including unicast and multicast
PAUSE frames). Note that the processing of PAUSE control frames depends only on RFCE
in Flow Control Register[2].
00 or 01: MAC prevents all control frames from reaching the application
10: MAC forwards all control frames to application even if they fail the address filter
11: MAC forwards control frames that pass the address filter.
Bit 5 BFD: Broadcast frames disable
When this bit is set, the address filters filter all incoming broadcast frames.
When this bit is reset, the address filters pass all received broadcast frames.
Bit 4 PAM: Pass all multicast
When set, this bit indicates that all received frames with a multicast destination address (first
bit in the destination address field is '1') are passed.
When reset, filtering of multicast frame depends on the HM bit.
Bit 3 DAIF: Destination address inverse filtering
When this bit is set, the address check block operates in inverse filtering mode for the DA
address comparison for both unicast and multicast frames.
When reset, normal filtering of frames is performed.
Bit 2 HM: Hash multicast
When set, MAC performs destination address filtering of received multicast frames
according to the hash table.
When reset, the MAC performs a perfect destination address filtering for multicast frames,
that is, it compares the DA field with the values programmed in DA registers.
Bit 1 HU: Hash unicast
When set, MAC performs destination address filtering of unicast frames according to the
hash table.
When reset, the MAC performs a perfect destination address filtering for unicast frames, that
is, it compares the DA field with the values programmed in DA registers.
Bit 0 PM: Promiscuous mode
When this bit is set, the address filters pass all incoming frames regardless of their
destination or source address. The SA/DA filter fails status bits in the receive status word
are always cleared when PM is set.

Ethernet MAC hash table high register (ETH_MACHTHR)
Address offset: 0x0008
Reset value: 0x0000 0000
The 64-bit Hash table is used for group address filtering. For hash filtering, the contents of
the destination address in the incoming frame are passed through the CRC logic, and the
upper 6 bits in the CRC register are used to index the contents of the Hash table. This CRC
is a 32-bit value coded by the following polynomial (for more details refer to Section 33.5.3:

RM0090 Rev 17

1195/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

MAC frame reception):
G( x) = x

32

+x

26

+x

23

+x

22

+x

16

+x

12

+x

11

+x

10

8

7

5

4

2

+x +x +x +x +x +x+1

The most significant bit determines the register to be used (hash table high/hash table low),
and the other 5 bits determine which bit within the register. A hash value of 0b0 0000 selects
bit 0 in the selected register, and a value of 0b1 1111 selects bit 31 in the selected register.
For example, if the DA of the incoming frame is received as 0x1F52 419C B6AF (0x1F is the
first byte received on the MII interface), then the internally calculated 6-bit Hash value is
0x2C and the HTH register bit[12] is checked for filtering. If the DA of the incoming frame is
received as 0xA00A 9800 0045, then the calculated 6-bit Hash value is 0x07 and the HTL
register bit[7] is checked for filtering.
If the corresponding bit value in the register is 1, the frame is accepted. Otherwise, it is
rejected. If the PAM (pass all multicast) bit is set in the ETH_MACFFR register, then all
multicast frames are accepted regardless of the multicast hash values.
The Hash table high register contains the higher 32 bits of the multicast Hash table.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

HTH
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 HTH: Hash table high
This field contains the upper 32 bits of Hash table.

Ethernet MAC hash table low register (ETH_MACHTLR)
Address offset: 0x000C
Reset value: 0x0000 0000
The Hash table low register contains the lower 32 bits of the multi-cast Hash table.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

HTL
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 HTL: Hash table low
This field contains the lower 32 bits of the Hash table.

Ethernet MAC MII address register (ETH_MACMIIAR)
Address offset: 0x0010
Reset value: 0x0000 0000
The MII address register controls the management cycles to the external PHY through the
management interface.
9

PA
Reserved

1196/1747

rw

rw

rw

RM0090 Rev 17

8

7

6

MR
rw

rw

rw

rw

rw

rw

rw

5
Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

4

3

2

CR
rw rw rw

1

0

MW MB
rw

rc_
w1

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:11 PA: PHY address
This field tells which of the 32 possible PHY devices are being accessed.
Bits 10:6 MR: MII register
These bits select the desired MII register in the selected PHY device.
Bit 5 Reserved, must be kept at reset value.
Bits 4:2 CR: Clock range
The CR clock range selection determines the HCLK frequency and is used to decide the
frequency of the MDC clock:
Selection HCLK MDC Clock
000 60-100 MHz HCLK/42
001 100-150 MHzHCLK/62
010 20-35 MHz HCLK/16
011 35-60 MHz HCLK/26
100 150-168 MHz HCLK/102
101, 110, 111 Reserved Bit 1 MW: MII write
When set, this bit tells the PHY that this will be a Write operation using the MII Data register. If
this bit is not set, this will be a Read operation, placing the data in the MII Data register.
Bit 0 MB: MII busy
This bit should read a logic 0 before writing to ETH_MACMIIAR and ETH_MACMIIDR. This
bit must also be reset to 0 during a Write to ETH_MACMIIAR. During a PHY register access,
this bit is set to 0b1 by the application to indicate that a read or write access is in progress.
ETH_MACMIIDR (MII Data) should be kept valid until this bit is cleared by the MAC during a
PHY Write operation. The ETH_MACMIIDR is invalid until this bit is cleared by the MAC
during a PHY Read operation. The ETH_MACMIIAR (MII Address) should not be written to
until this bit is cleared.

Ethernet MAC MII data register (ETH_MACMIIDR)
Address offset: 0x0014
Reset value: 0x0000 0000
The MAC MII Data register stores write data to be written to the PHY register located at the
address specified in ETH_MACMIIAR. ETH_MACMIIDR also stores read data from the PHY
register located at the address specified by ETH_MACMIIAR.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

rw

rw

7

6

5

4

3

2

1

0

rw

rw rw rw rw

rw

rw

MD
rw

rw

rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 MD: MII data
This contains the 16-bit data value read from the PHY after a Management Read operation,
or the 16-bit data value to be written to the PHY before a Management Write operation.

RM0090 Rev 17

1197/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Ethernet MAC flow control register (ETH_MACFCR)
Address offset: 0x0018
Reset value: 0x0000 0000

rw

6

3

2

1

0

TFCE

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

7

RFCE

Reserved

8

ZQPD

PT

9

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

UPFD

The Flow control register controls the generation and reception of the control (Pause
Command) frames by the MAC. A write to a register with the Busy bit set to '1' causes the
MAC to generate a pause control frame. The fields of the control frame are selected as
specified in the 802.3x specification, and the Pause Time value from this register is used in
the Pause Time field of the control frame. The Busy bit remains set until the control frame is
transferred onto the cable. The Host must make sure that the Busy bit is cleared before
writing to the register.
5

4

FCB/
BPA

rw rw rw rw

rw

rc_w1
/rw

PLT

Bits 31:16 PT: Pause time
This field holds the value to be used in the Pause Time field in the transmit control frame. If
the Pause Time bits is configured to be double-synchronized to the MII clock domain, then
consecutive write operations to this register should be performed only after at least 4 clock
cycles in the destination clock domain.
Bits 15:8 Reserved, must be kept at reset value.
Bit 7 ZQPD: Zero-quanta pause disable
When set, this bit disables the automatic generation of Zero-quanta pause control frames on
the deassertion of the flow-control signal from the FIFO layer.
When this bit is reset, normal operation with automatic Zero-quanta pause control frame
generation is enabled.
Bit 6 Reserved, must be kept at reset value.
Bits 5:4 PLT: Pause low threshold
This field configures the threshold of the Pause timer at which the Pause frame is
automatically retransmitted. The threshold values should always be less than the Pause
Time configured in bits[31:16]. For example, if PT = 100H (256 slot-times), and PLT = 01,
then a second PAUSE frame is automatically transmitted if initiated at 228 (256 – 28) slottimes after the first PAUSE frame is transmitted.
Selection Threshold
00 Pause time minus 4 slot times
01 Pause time minus 28 slot times
10 Pause time minus 144 slot times
11 Pause time minus 256 slot times
Slot time is defined as time taken to transmit 512 bits (64 bytes) on the MII interface.
Bit 3 UPFD: Unicast pause frame detect
When this bit is set, the MAC detects the Pause frames with the station’s unicast address
specified in the ETH_MACA0HR and ETH_MACA0LR registers, in addition to detecting
Pause frames with the unique multicast address.
When this bit is reset, the MAC detects only a Pause frame with the unique multicast
address specified in the 802.3x standard.

1198/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Bit 2 RFCE: Receive flow control enable
When this bit is set, the MAC decodes the received Pause frame and disables its transmitter
for a specified (Pause Time) time.
When this bit is reset, the decode function of the Pause frame is disabled.
Bit 1 TFCE: Transmit flow control enable
In Full-duplex mode, when this bit is set, the MAC enables the flow control operation to
transmit Pause frames. When this bit is reset, the flow control operation in the MAC is
disabled, and the MAC does not transmit any Pause frames.
In Half-duplex mode, when this bit is set, the MAC enables the back-pressure operation.
When this bit is reset, the back pressure feature is disabled.
Bit 0 FCB/BPA: Flow control busy/back pressure activate
This bit initiates a Pause Control frame in Full-duplex mode and activates the back pressure
function in Half-duplex mode if TFCE bit is set.
In Full-duplex mode, this bit should be read as 0 before writing to the Flow control register.
To initiate a Pause control frame, the Application must set this bit to 1. During a transfer of
the Control frame, this bit continues to be set to signify that a frame transmission is in
progress. After completion of the Pause control frame transmission, the MAC resets this bit
to 0. The Flow control register should not be written to until this bit is cleared.
In Half-duplex mode, when this bit is set (and TFCE is set), back pressure is asserted by the
MAC core. During back pressure, when the MAC receives a new frame, the transmitter
starts sending a JAM pattern resulting in a collision. When the MAC is configured to Fullduplex mode, the BPA is automatically disabled.

Ethernet MAC VLAN tag register (ETH_MACVLANTR)
Address offset: 0x001C
Reset value: 0x0000 0000
The VLAN tag register contains the IEEE 802.1Q VLAN Tag to identify the VLAN frames.
The MAC compares the 13th and 14th bytes of the receiving frame (Length/Type) with
0x8100, and the following 2 bytes are compared with the VLAN tag; if a match occurs, the
received VLAN bit in the receive frame status is set. The legal length of the frame is
increased from 1518 bytes to 1522 bytes.

Reserved

9

VLANTC

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

rw

8

7

6

5

4

3

2

rw rw

rw

rw rw

1

0

VLANTI
rw rw

rw rw

RM0090 Rev 17

rw

rw rw

rw rw

rw rw

1199/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Bits 31:17 Reserved, must be kept at reset value.
Bit 16 VLANTC: 12-bit VLAN tag comparison
When this bit is set, a 12-bit VLAN identifier, rather than the complete 16-bit VLAN tag, is
used for comparison and filtering. Bits[11:0] of the VLAN tag are compared with the
corresponding field in the received VLAN-tagged frame.
When this bit is reset, all 16 bits of the received VLAN frame’s fifteenth and sixteenth bytes
are used for comparison.
Bits 15:0 VLANTI: VLAN tag identifier (for receive frames)
This contains the 802.1Q VLAN tag to identify VLAN frames, and is compared to the fifteenth
and sixteenth bytes of the frames being received for VLAN frames. Bits[15:13] are the user
priority, Bit[12] is the canonical format indicator (CFI) and bits[11:0] are the VLAN tag’s VLAN
identifier (VID) field. When the VLANTC bit is set, only the VID (bits[11:0]) is used for
comparison.
If VLANTI (VLANTI[11:0] if VLANTC is set) is all zeros, the MAC does not check the fifteenth
and sixteenth bytes for VLAN tag comparison, and declares all frames with a Type field
value of 0x8100 as VLAN frames.

Ethernet MAC remote wakeup frame filter register (ETH_MACRWUFFR)
Address offset: 0x0028
Reset value: 0x0000 0000
This is the address through which the remote wakeup frame filter registers are written/read
by the application. The Wakeup frame filter register is actually a pointer to eight (not
transparent) such wakeup frame filter registers. Eight sequential write operations to this
address with the offset (0x0028) will write all wakeup frame filter registers. Eight sequential
read operations from this address with the offset (0x0028) will read all wakeup frame filter
registers. This register contains the higher 16 bits of the 7th MAC address. Refer to Remote
wakeup frame filter register section for additional information.
Figure 385. Ethernet MAC remote wakeup frame filter register (ETH_MACRWUFFR)
Wakeup frame filter reg0

Filter 0 Byte Mask

Wakeup frame filter reg1

Filter 1 Byte Mask

Wakeup frame filter reg2

Filter 2 Byte Mask

Wakeup frame filter reg3

Filter 3 Byte Mask

Wakeup frame filter reg4
Wakeup frame filter reg5

RSVD

Filter 3
Command

Filter 3 Offset

RSVD

Filter 2
Command

Filter 2 Offset

RSVD

Filter 1
Command

Filter 1 Offset

RSVD

Filter 0
Command

Filter 0 Offset

Wakeup frame filter reg6

Filter 1 CRC - 16

Filter 0 CRC - 16

Wakeup frame filter reg7

Filter 3 CRC - 16

Filter 2 CRC - 16

ai15648

1200/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Ethernet MAC PMT control and status register (ETH_MACPMTCSR)
Address offset: 0x002C
Reset value: 0x0000 0000

3

2

1

0
PD

rc_r rc_r

4

MPE

5

WFE

rw

6

Reserved

Res.

7

MPR

rs

8
Reserved

Reserved

GU

9

WFFRPR

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

WFR

The ETH_MACPMTCSR programs the request wakeup events and monitors the wakeup
events.

rw

rw

rs

Bit 31 WFFRPR: Wakeup frame filter register pointer reset
When set, it resets the Remote wakeup frame filter register pointer to 0b000. It is
automatically cleared after 1 clock cycle.
Bits 30:10 Reserved, must be kept at reset value.
Bit 9 GU: Global unicast
When set, it enables any unicast packet filtered by the MAC (DAF) address recognition to be
a wakeup frame.
Bits 8:7 Reserved, must be kept at reset value.
Bit 6 WFR: Wakeup frame received
When set, this bit indicates the power management event was generated due to reception of
a wakeup frame. This bit is cleared by a read into this register.
Bit 5 MPR: Magic packet received
When set, this bit indicates the power management event was generated by the reception of
a Magic Packet. This bit is cleared by a read into this register.
Bits 4:3 Reserved, must be kept at reset value.
Bit 2 WFE: Wakeup frame enable
When set, this bit enables the generation of a power management event due to wakeup
frame reception.
Bit 1 MPE: Magic Packet enable
When set, this bit enables the generation of a power management event due to Magic
Packet reception.
Bit 0 PD: Power down
When this bit is set, all received frames will be dropped. This bit is cleared automatically
when a magic packet or wakeup frame is received, and Power-down mode is disabled.
Frames received after this bit is cleared are forwarded to the application. This bit must only
be set when either the Magic Packet Enable or Wakeup Frame Enable bit is set high.

RM0090 Rev 17

1201/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Ethernet MAC debug register (ETH_MACDBGR)
Address offset: 0x0034
Reset value: 0x0000 0000
This debug register gives the status of all the main modules of the transmit and receive data
paths and the FIFOs. An all-zero status indicates that the MAC core is in Idle state (and
FIFOs are empty) and no activity is going on in the data paths.

ro

ro

ro

ro

ro

ro

ro

ro

ro

ro

3

ro

2

1

ro

0
MMRPEA

4

MSFRWCS

5

RFWRA

6

Reserved

7
Reserved

Reserved

ro

8

RFRCS

9
RFFL

MMTEA

MTFCS

MTP

TFRS

ro

TFWA

TFNE

ro

Reserved

TFF

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

ro

ro

Bits 31:26 Reserved, must be kept at reset value.
Bit 25 TFF: Tx FIFO full
When high, it indicates that the Tx FIFO is full and hence no more frames will be accepted
for transmission.
Bit 24 TFNE: Tx FIFO not empty
When high, it indicates that the TxFIFO is not empty and has some data left for
transmission.
Bit 23 Reserved, must be kept at reset value.
Bit 22 TFWA: Tx FIFO write active
When high, it indicates that the TxFIFO write controller is active and transferring data to the
TxFIFO.
Bits 21:20 TFRS: Tx FIFO read status
This indicates the state of the TxFIFO read controller:
00: Idle state
01: Read state (transferring data to the MAC transmitter)
10: Waiting for TxStatus from MAC transmitter
11: Writing the received TxStatus or flushing the TxFIFO
Bit 19 MTP: MAC transmitter in pause
When high, it indicates that the MAC transmitter is in Pause condition (in full-duplex mode
only) and hence will not schedule any frame for transmission
Bits 18:17 MTFCS: MAC transmit frame controller status
This indicates the state of the MAC transmit frame controller:
00: Idle
01: Waiting for Status of previous frame or IFG/backoff period to be over
10: Generating and transmitting a Pause control frame (in full duplex mode)
11: Transferring input frame for transmission
Bit 16 MMTEA: MAC MII transmit engine active
When high, it indicates that the MAC MII transmit engine is actively transmitting data and
that it is not in the Idle state.
Bits 15:10 Reserved, must be kept at reset value.

1202/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Bits 9:8 RFFL: Rx FIFO fill level
This gives the status of the Rx FIFO fill-level:
00: RxFIFO empty
01: RxFIFO fill-level below flow-control de-activate threshold
10: RxFIFO fill-level above flow-control activate threshold
11: RxFIFO full
Bit 7 Reserved, must be kept at reset value.
Bits 6:5 RFRCS: Rx FIFO read controller status
It gives the state of the Rx FIFO read controller:
00: IDLE state
01: Reading frame data
10: Reading frame status (or time-stamp)
11: Flushing the frame data and status
Bit 4 RFWRA: Rx FIFO write controller active
When high, it indicates that the Rx FIFO write controller is active and transferring a received
frame to the FIFO.
Bit 3 Reserved, must be kept at reset value.
Bits 2:1 MSFRWCS: MAC small FIFO read / write controllers status
When high, these bits indicate the respective active state of the small FIFO read and write
controllers of the MAC receive frame controller module.
Bit 0 MMRPEA: MAC MII receive protocol engine active
When high, it indicates that the MAC MII receive protocol engine is actively receiving data
and is not in the Idle state.

RM0090 Rev 17

1203/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Ethernet MAC interrupt status register (ETH_MACSR)
Address offset: 0x0038
Reset value: 0x0000 0000
The ETH_MACSR register contents identify the events in the MAC that can generate an
interrupt.
15

14

13

12

Reserved

11

10

9
TSTS
rc_r

8

7

Reserved

6

5

MMCTS MMCRS
r

r

4

3

MMCS

PMTS

r

r

2

1

0

Reserved

Bits 15:10 Reserved, must be kept at reset value.
Bit 9 TSTS: Time stamp trigger status
This bit is set high when the system time value equals or exceeds the value specified in the
Target time high and low registers. This bit is cleared by reading the ETH_PTPTSSR
register.
Bits 8:7 Reserved, must be kept at reset value.
Bit 6 MMCTS: MMC transmit status
This bit is set high whenever an interrupt is generated in the ETH_MMCTIR Register. This bit
is cleared when all the bits in this interrupt register (ETH_MMCTIR) are cleared.
Bit 5 MMCRS: MMC receive status
This bit is set high whenever an interrupt is generated in the ETH_MMCRIR register. This bit
is cleared when all the bits in this interrupt register (ETH_MMCRIR) are cleared.
Bit 4 MMCS: MMC status
This bit is set high whenever any of bits 6:5 is set high. It is cleared only when both bits are
low.
Bit 3 PMTS: PMT status
This bit is set whenever a Magic packet or Wake-on-LAN frame is received in Power-down
mode (see bits 5 and 6 in the ETH_MACPMTCSR register Ethernet MAC PMT control and
status register (ETH_MACPMTCSR) on page 1201). This bit is cleared when both bits[6:5],
of this last register, are cleared due to a read operation to the ETH_MACPMTCSR register.
Bits 2:0 Reserved, must be kept at reset value.

1204/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Ethernet MAC interrupt mask register (ETH_MACIMR)
Address offset: 0x003C
Reset value: 0x0000 0000
The ETH_MACIMR register bits make it possible to mask the interrupt signal due to the
corresponding event in the ETH_MACSR register.
15

14

13

12

11

10

9

8

7

6

TSTIM

Reserved

5

4

3

Reserved

rw

2

1

PMTIM

0

Reserved

rw

Bits 15:10 Reserved, must be kept at reset value.
Bit 9 TSTIM: Time stamp trigger interrupt mask
When set, this bit disables the time stamp interrupt generation.
Bits 8:4 Reserved, must be kept at reset value.
Bit 3 PMTIM: PMT interrupt mask
When set, this bit disables the assertion of the interrupt signal due to the setting of the PMT
Status bit in ETH_MACSR.
Bits 2:0 Reserved, must be kept at reset value.

Ethernet MAC address 0 high register (ETH_MACA0HR)
Address offset: 0x0040
Reset value: 0x8000 FFFF
The MAC address 0 high register holds the upper 16 bits of the 6-byte first MAC address of
the station. Note that the first DA byte that is received on the MII interface corresponds to
the LS Byte (bits [7:0]) of the MAC address low register. For example, if 0x1122 3344 5566
is received (0x11 is the first byte) on the MII as the destination address, then the MAC
address 0 register [47:0] is compared with 0x6655 4433 2211.

MO

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

MACA0H

Reserved

1

8

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bit 31 MO: Always 1.
Bits 30:16 Reserved, must be kept at reset value.
Bits 15:0 MACA0H: MAC address0 high [47:32]
This field contains the upper 16 bits (47:32) of the 6-byte MAC address0. This is used by the
MAC for filtering for received frames and for inserting the MAC address in the transmit flow
control (Pause) frames.

RM0090 Rev 17

1205/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Ethernet MAC address 0 low register (ETH_MACA0LR)
Address offset: 0x0044
Reset value: 0xFFFF FFFF
The MAC address 0 low register holds the lower 32 bits of the 6-byte first MAC address of
the station.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

MACA0L
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 MACA0L: MAC address0 low [31:0]
This field contains the lower 32 bits of the 6-byte MAC address0. This is used by the MAC
for filtering for received frames and for inserting the MAC address in the transmit flow control
(Pause) frames.

Ethernet MAC address 1 high register (ETH_MACA1HR)
Address offset: 0x0048
Reset value: 0x0000 FFFF
The MAC address 1 high register holds the upper 16 bits of the 6-byte second MAC address
of the station.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
AE SA
rw

rw

MBC
rw

rw

rw

rw

9

Reserved
rw

rw

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

MACA1H
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bit 31 AE: Address enable
When this bit is set, the address filters use the MAC address1 for perfect filtering. When this
bit is cleared, the address filters ignore the address for filtering.
Bit 30 SA: Source address
When this bit is set, the MAC address1[47:0] is used for comparison with the SA fields of the
received frame.
When this bit is cleared, the MAC address1[47:0] is used for comparison with the DA fields
of the received frame.

1206/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Bits 29:24 MBC: Mask byte control
These bits are mask control bits for comparison of each of the MAC address1 bytes. When
they are set high, the MAC core does not compare the corresponding byte of received
DA/SA with the contents of the MAC address1 registers. Each bit controls the masking of the
bytes as follows:
– Bit 29: ETH_MACA1HR [15:8]
– Bit 28: ETH_MACA1HR [7:0]
– Bit 27: ETH_MACA1LR [31:24]
…
– Bit 24: ETH_MACA1LR [7:0]
Bits 23:16 Reserved, must be kept at reset value.
Bits 15:0 MACA1H: MAC address1 high [47:32]
This field contains the upper 16 bits (47:32) of the 6-byte second MAC address.

Ethernet MAC address1 low register (ETH_MACA1LR)
Address offset: 0x004C
Reset value: 0xFFFF FFFF
The MAC address 1 low register holds the lower 32 bits of the 6-byte second MAC address
of the station.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

MACA1L
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 MACA1L: MAC address1 low [31:0]
This field contains the lower 32 bits of the 6-byte MAC address1. The content of this field is
undefined until loaded by the application after the initialization process.

Ethernet MAC address 2 high register (ETH_MACA2HR)
Address offset: 0x0050
Reset value: 0x0000 FFFF
The MAC address 2 high register holds the upper 16 bits of the 6-byte second MAC address
of the station.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
AE SA
rw

rw

MBC
rw

rw

rw

rw

9

Reserved
rw

rw

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

MACA2H
rw

rw

rw

RM0090 Rev 17

rw

rw

rw

rw

rw

rw

1207/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Bit 31 AE: Address enable
When this bit is set, the address filters use the MAC address2 for perfect filtering. When
reset, the address filters ignore the address for filtering.
Bit 30 SA: Source address
When this bit is set, the MAC address 2 [47:0] is used for comparison with the SA fields of
the received frame.
When this bit is reset, the MAC address 2 [47:0] is used for comparison with the DA fields of
the received frame.
Bits 29:24 MBC: Mask byte control
These bits are mask control bits for comparison of each of the MAC address2 bytes. When
set high, the MAC core does not compare the corresponding byte of received DA/SA with the
contents of the MAC address 2 registers. Each bit controls the masking of the bytes as
follows:
– Bit 29: ETH_MACA2HR [15:8]
– Bit 28: ETH_MACA2HR [7:0]
– Bit 27: ETH_MACA2LR [31:24]
…
– Bit 24: ETH_MACA2LR [7:0]
Bits 23:16 Reserved, must be kept at reset value.
Bits 15:0 MACA2H: MAC address2 high [47:32]
This field contains the upper 16 bits (47:32) of the 6-byte MAC address2.

Ethernet MAC address 2 low register (ETH_MACA2LR)
Address offset: 0x0054
Reset value: 0xFFFF FFFF
The MAC address 2 low register holds the lower 32 bits of the 6-byte second MAC address
of the station.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

MACA2L
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 MACA2L: MAC address2 low [31:0]
This field contains the lower 32 bits of the 6-byte second MAC address2. The content of this
field is undefined until loaded by the application after the initialization process.

1208/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Ethernet MAC address 3 high register (ETH_MACA3HR)
Address offset: 0x0058
Reset value: 0x0000 FFFF
The MAC address 3 high register holds the upper 16 bits of the 6-byte second MAC address
of the station.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
AE SA
rw

rw

MBC
rw

rw

rw

rw

9

Reserved
rw

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

MACA3H

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bit 31 AE: Address enable
When this bit is set, the address filters use the MAC address3 for perfect filtering. When this
bit is cleared, the address filters ignore the address for filtering.
Bit 30 SA: Source address
When this bit is set, the MAC address 3 [47:0] is used for comparison with the SA fields of the
received frame.
When this bit is cleared, the MAC address 3[47:0] is used for comparison with the DA fields
of the received frame.
Bits 29:24 MBC: Mask byte control
These bits are mask control bits for comparison of each of the MAC address3 bytes. When
these bits are set high, the MAC core does not compare the corresponding byte of received
DA/SA with the contents of the MAC address 3 registers. Each bit controls the masking of the
bytes as follows:
– Bit 29: ETH_MACA3HR [15:8]
– Bit 28: ETH_MACA3HR [7:0]
– Bit 27: ETH_MACA3LR [31:24]
…
– Bit 24: ETH_MACA3LR [7:0]
Bits 23:16 Reserved, must be kept at reset value.
Bits 15:0 MACA3H: MAC address3 high [47:32]
This field contains the upper 16 bits (47:32) of the 6-byte MAC address3.

Ethernet MAC address 3 low register (ETH_MACA3LR)
Address offset: 0x005C
Reset value: 0xFFFF FFFF
The MAC address 3 low register holds the lower 32 bits of the 6-byte second MAC address
of the station.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

MACA3L
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 MACA3L: MAC address3 low [31:0]
This field contains the lower 32 bits of the 6-byte second MAC address3. The content of this
field is undefined until loaded by the application after the initialization process.

RM0090 Rev 17

1209/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

33.8.2

RM0090

MMC register description
Ethernet MMC control register (ETH_MMCCR)
Address offset: 0x0100
Reset value: 0x0000 0000

5

4

3

2

1

0
CR

6

CSR

Reserved

7

ROR

8

MCF

9

MCP

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

MCFHP

The Ethernet MMC Control register establishes the operating mode of the management
counters.

rw

rw

rw

rw

rw

rw

Bits 31:6 Reserved, must be kept at reset value.
MCFHP: MMC counter Full-Half preset
When MCFHP is low and bit4 is set, all MMC counters get preset to almost-half value. All
Bit 5 frame-counters get preset to 0x7FFF_FFF0 (half - 16)
When MCFHP is high and bit4 is set, all MMC counters get preset to almost-full value. All
frame-counters get preset to 0xFFFF_FFF0 (full - 16)
MCP: MMC counter preset
When set, all counters will be initialized or preset to almost full or almost half as per
Bit 4 Bit5 above. This bit will be cleared automatically after 1 clock cycle. This bit along
with bit5 is useful for debugging and testing the assertion of interrupts due to MMC
counter becoming half-full or full.
Bit 3 MCF: MMC counter freeze
When set, this bit freezes all the MMC counters to their current value. (None of the MMC
counters are updated due to any transmitted or received frame until this bit is cleared to 0. If
any MMC counter is read with the Reset on Read bit set, then that counter is also cleared in
this mode.)
Bit 2 ROR: Reset on read
When this bit is set, the MMC counters is reset to zero after read (self-clearing after reset).
The counters are cleared when the least significant byte lane (bits [7:0]) is read.
Bit 1 CSR: Counter stop rollover
When this bit is set, the counter does not roll over to zero after it reaches the maximum
value.
Bit 0 CR: Counter reset
When it is set, all counters are reset. This bit is cleared automatically after 1 clock cycle.

Ethernet MMC receive interrupt register (ETH_MMCRIR)
Address offset: 0x0104
Reset value: 0x0000 0000
The Ethernet MMC receive interrupt register maintains the interrupts generated when
receive statistic counters reach half their maximum values. (MSB of the counter is set.) It is
a 32-bit wide register. An interrupt bit is cleared when the respective MMC counter that

1210/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

16 15 14 13 12 11 10

Reserved

9

8

7

Reserved

rc_r

6

5
RFCES

17
RGUFS

31 30 29 28 27 26 25 24 23 22 21 20 19 18

RFAES

caused the interrupt is read. The least significant byte lane (bits [7:0]) of the respective
counter must be read in order to clear the interrupt bit.
4

3

2

1

0

Reserved

rc_r rc_r

Bits 31:18 Reserved, must be kept at reset value.
Bit 17 RGUFS: Received Good Unicast Frames Status
This bit is set when the received, good unicast frames, counter reaches half the maximum
value.
Bits 16:7 Reserved, must be kept at reset value.
Bit 6 RFAES: Received frames alignment error status
This bit is set when the received frames, with alignment error, counter reaches half the
maximum value.
Bit 5 RFCES: Received frames CRC error status
This bit is set when the received frames, with CRC error, counter reaches half the maximum
value.
Bits 4:0 Reserved, must be kept at reset value.

Ethernet MMC transmit interrupt register (ETH_MMCTIR)
Address offset: 0x0108
Reset value: 0x0000 0000

20 19 18 17 16

Reserved

Reserved

rc_r

15

14
TGFSCS

21

TGFMSCS

31 30 29 28 27 26 25 24 23 22

TGFS

The Ethernet MMC transmit Interrupt register maintains the interrupts generated when
transmit statistic counters reach half their maximum values. (MSB of the counter is set.) It is
a 32-bit wide register. An interrupt bit is cleared when the respective MMC counter that
caused the interrupt is read. The least significant byte lane (bits [7:0]) of the respective
counter must be read in order to clear the interrupt bit.
13 12 11 10

9

8

7

6

5

4

3

2

1

0

Reserved

rc_r rc_r

Bits 31:22 Reserved, must be kept at reset value.
Bit 21 TGFS: Transmitted good frames status
This bit is set when the transmitted, good frames, counter reaches half the maximum value.
Bits 20:16 Reserved, must be kept at reset value.

RM0090 Rev 17

1211/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Bit 15 TGFMSCS: Transmitted good frames more single collision status
This bit is set when the transmitted, good frames after more than a single collision, counter
reaches half the maximum value.
Bit 14 TGFSCS: Transmitted good frames single collision status
This bit is set when the transmitted, good frames after a single collision, counter reaches half
the maximum value.
Bits 13:0 Reserved, must be kept at reset value.

Ethernet MMC receive interrupt mask register (ETH_MMCRIMR)
Address offset: 0x010C
Reset value: 0x0000 0000

Reserved

16 15 14 13 12 11 10

rw

Reserved

9

8

7

6

5
RFCEM

17

RFAEM

31 30 29 28 27 26 25 24 23 22 21 20 19 18

RGUFM

The Ethernet MMC receive interrupt mask register maintains the masks for interrupts
generated when the receive statistic counters reach half their maximum value. (MSB of the
counter is set.) It is a 32-bit wide register.

rw

rw

4

3

2

1

0

Reserved

Bits 31:18 Reserved, must be kept at reset value.
Bit 17 RGUFM: Received good unicast frames mask
Setting this bit masks the interrupt when the received, good unicast frames, counter
reaches half the maximum value.
Bits 16:7 Reserved, must be kept at reset value.
Bit 6 RFAEM: Received frames alignment error mask
Setting this bit masks the interrupt when the received frames, with alignment error, counter
reaches half the maximum value.
Bit 5 RFCEM: Received frame CRC error mask
Setting this bit masks the interrupt when the received frames, with CRC error, counter
reaches half the maximum value.
Bits 4:0 Reserved, must be kept at reset value.

1212/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Ethernet MMC transmit interrupt mask register (ETH_MMCTIMR)
Address offset: 0x0110
Reset value: 0x0000 0000

20 19 18 17 16

Reserved

Reserved

rw

15

14
TGFSCM

21

TGFMSCM

31 30 29 28 27 26 25 24 23 22

TGFM

The Ethernet MMC transmit interrupt mask register maintains the masks for interrupts
generated when the transmit statistic counters reach half their maximum value. (MSB of the
counter is set). It is a 32-bit wide register.

rw

rw

13 12 11 10

9

8

7

6

5

4

3

2

1

0

Reserved

Bits 31:22 Reserved, must be kept at reset value.
Bit 21 TGFM: Transmitted good frames mask
Setting this bit masks the interrupt when the transmitted, good frames, counter reaches half
the maximum value.
Bits 20:16 Reserved, must be kept at reset value.
Bit 15 TGFMSCM: Transmitted good frames more single collision mask
Setting this bit masks the interrupt when the transmitted good frames after more than a
single collision counter reaches half the maximum value.
Bit 14 TGFSCM: Transmitted good frames single collision mask
Setting this bit masks the interrupt when the transmitted good frames after a single collision
counter reaches half the maximum value.
Bits 13:0 Reserved, must be kept at reset value.

Ethernet MMC transmitted good frames after a single collision counter
register (ETH_MMCTGFSCCR)
Address offset: 0x014C
Reset value: 0x0000 0000
This register contains the number of successfully transmitted frames after a single collision
in Half-duplex mode.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

TGFSCC
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:0 TGFSCC: Transmitted good frames single collision counter
Transmitted good frames after a single collision counter.

RM0090 Rev 17

1213/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Ethernet MMC transmitted good frames after more than a single collision
counter register (ETH_MMCTGFMSCCR)
Address offset: 0x0150
Reset value: 0x0000 0000
This register contains the number of successfully transmitted frames after more than a
single collision in Half-duplex mode.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

TGFMSCC
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:0 TGFMSCC: Transmitted good frames more single collision counter
Transmitted good frames after more than a single collision counter

Ethernet MMC transmitted good frames counter register (ETH_MMCTGFCR)
Address offset: 0x0168
Reset value: 0x0000 0000
This register contains the number of good frames transmitted.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

TGFC
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:0 TGFC: Transmitted good frames counter

Ethernet MMC received frames with CRC error counter register
(ETH_MMCRFCECR)
Address offset: 0x0194
Reset value: 0x0000 0000
This register contains the number of frames received with CRC error.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

RFCEC
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:0 RFCEC: Received frames CRC error counter
Received frames with CRC error counter

1214/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Ethernet MMC received frames with alignment error counter register
(ETH_MMCRFAECR)
Address offset: 0x0198
Reset value: 0x0000 0000
This register contains the number of frames received with alignment (dribble) error.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

RFAEC
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:0 RFAEC: Received frames alignment error counter
Received frames with alignment error counter

MMC received good unicast frames counter register (ETH_MMCRGUFCR)
Address offset: 0x01C4
Reset value: 0x0000 0000
This register contains the number of good unicast frames received.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

RGUFC
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:0 RGUFC: Received good unicast frames counter

33.8.3

IEEE 1588 time stamp registers
This section describes the registers required to support precision network clock
synchronization functions under the IEEE 1588 standard.

Ethernet PTP time stamp control register (ETH_PTPTSCR)
Address offset: 0x0700
Reset value: 0x0000 00002000

rw

rw

RM0090 Rev 17

2

1

0

TSE

rw

3

TSFCU

rw

4

TSSTI

rw

5

TSITE

TSSSR

TSSARFE

rw

6

TSSTU

TSPTPPSV2E

rw

7

Reserved

TSSPTPOEFE

rw

TSSIPV6FE

rw

8

TSSEME

TSSMRME

rw

9

TSSIPV4FE

TSCNT

Reserved

TSPFFMAE

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

TTSARU

This register controls the time stamp generation and update logic.

rw

rw

rw

rw

rw

rw

1215/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Bits 31:19 Reserved, must be kept at reset value.
Bit 18 TSPFFMAE: Time stamp PTP frame filtering MAC address enable
When set, this bit uses the MAC address (except for MAC address 0) to filter the PTP frames
when PTP is sent directly over Ethernet.
Bits 17:16 TSCNT: Time stamp clock node type
The following are the available types of clock node:
00: Ordinary clock
01: Boundary clock
10: End-to-end transparent clock
11: Peer-to-peer transparent clock
Bit 15 TSSMRME: Time stamp snapshot for message relevant to master enable
When this bit is set, the snapshot is taken for messages relevant to the master node only.
When this bit is cleared the snapshot is taken for messages relevant to the slave node only.
This is valid only for the ordinary clock and boundary clock nodes.
Bit 14 TSSEME: Time stamp snapshot for event message enable
When this bit is set, the time stamp snapshot is taken for event messages only (SYNC,
Delay_Req, Pdelay_Req or Pdelay_Resp). When this bit is cleared the snapshot is taken for
all other messages except for Announce, Management and Signaling.
Bit 13 TSSIPV4FE: Time stamp snapshot for IPv4 frames enable
When this bit is set, the time stamp snapshot is taken for IPv4 frames.
Bit 12 TSSIPV6FE: Time stamp snapshot for IPv6 frames enable
When this bit is set, the time stamp snapshot is taken for IPv6 frames.
Bit 11 TSSPTPOEFE: Time stamp snapshot for PTP over ethernet frames enable
When this bit is set, the time stamp snapshot is taken for frames which have PTP messages
in Ethernet frames (PTP over Ethernet) also. By default snapshots are taken for UDPIPEthernet PTP packets.
Bit 10 TSPTPPSV2E: Time stamp PTP packet snooping for version2 format enable
When this bit is set, the PTP packets are snooped using the version 2 format. When the bit is
cleared, the PTP packets are snooped using the version 1 format.
Note: IEEE 1588 Version 1 and Version 2 formats as indicated in IEEE standard 1588-2008
(Revision of IEEE STD. 1588-2002).
Bit 9 TSSSR: Time stamp subsecond rollover: digital or binary rollover control
When this bit is set, the Time stamp low register rolls over when the subsecond counter
reaches the value 0x3B9A C9FF (999 999 999 in decimal), and increments the Time Stamp
(high) seconds.
When this bit is cleared, the rollover value of the subsecond register reaches 0x7FFF FFFF.
The subsecond increment has to be programmed correctly depending on the PTP’s
reference clock frequency and this bit value.
Bit 8 TSSARFE: Time stamp snapshot for all received frames enable
When this bit is set, the time stamp snapshot is enabled for all frames received by the core.
Bits 7:6 Reserved, must be kept at reset value.
Bit 5 TSARU: Time stamp addend register update
When this bit is set, the Time stamp addend register’s contents are updated to the PTP block
for fine correction. This bit is cleared when the update is complete. This register bit must be
read as zero before you can set it.

1216/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Bit 4 TSITE: Time stamp interrupt trigger enable
When this bit is set, a time stamp interrupt is generated when the system time becomes
greater than the value written in the Target time register. When the Time stamp trigger
interrupt is generated, this bit is cleared.
Bit 3 TSSTU: Time stamp system time update
When this bit is set, the system time is updated (added to or subtracted from) with the value
specified in the Time stamp high update and Time stamp low update registers. Both the
TSSTU and TSSTI bits must be read as zero before you can set this bit. Once the update is
completed in hardware, this bit is cleared.
Bit 2 TSSTI: Time stamp system time initialize
When this bit is set, the system time is initialized (overwritten) with the value specified in the
Time stamp high update and Time stamp low update registers. This bit must be read as zero
before you can set it. When initialization is complete, this bit is cleared.
Bit 1 TSFCU: Time stamp fine or coarse update
When set, this bit indicates that the system time stamp is to be updated using the Fine
Update method. When cleared, it indicates the system time stamp is to be updated using the
Coarse method.
Bit 0 TSE: Time stamp enable
When this bit is set, time stamping is enabled for transmit and receive frames. When this bit
is cleared, the time stamp function is suspended and time stamps are not added for transmit
and receive frames. Because the maintained system time is suspended, you must always
initialize the time stamp feature (system time) after setting this bit high.

The table below indicates the messages for which a snapshot is taken depending on the
clock, enable master and enable snapshot for event message register settings.
Table 194. Time stamp snapshot dependency on registers bits
TSCNT
(bits 17:16)

TSSMRME
(bit 15)(1)

TSSEME
(bit 14)

00 or 01

X(2)

0

SYNC, Follow_Up, Delay_Req, Delay_Resp

00 or 01

1

1

Delay_Req

00 or 01

0

1

SYNC

10

N/A

0

SYNC, Follow_Up, Delay_Req, Delay_Resp

10

N/A

1

SYNC, Follow_Up

11

N/A

0

SYNC, Follow_Up, Delay_Req, Delay_Resp,
Pdelay_Req, Pdelay_Resp

11

N/A

1

SYNC, Pdelay_Req, Pdelay_Resp

Messages for which snapshots are taken

1. N/A = not applicable.
2. X = don’t care.

RM0090 Rev 17

1217/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Ethernet PTP subsecond increment register (ETH_PTPSSIR)
Address offset: 0x0704
Reset value: 0x0000 0000
This register contains the 8-bit value by which the subsecond register is incremented. In
Coarse update mode (TSFCU bit in ETH_PTPTSCR), the value in this register is added to
the system time every clock cycle of HCLK. In Fine update mode, the value in this register is
added to the system time whenever the accumulator gets an overflow.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

STSSI

Reserved
rw

rw

rw

rw

rw

Bits 31:8 Reserved, must be kept at reset value.
Bits 7:0 STSSI: System time subsecond increment
The value programmed in this register is added to the contents of the subsecond value of the
system time in every update.
For example, to achieve 20 ns accuracy, the value is: 20 / 0.467 = ~ 43 (or 0x2A).

Ethernet PTP time stamp high register (ETH_PTPTSHR)
Address offset: 0x0708
Reset value: 0x0000 0000
This register contains the most significant (higher) 32 time bits. This read-only register
contains the seconds system time value. The Time stamp high register, along with Time
stamp low register, indicates the current value of the system time maintained by the MAC.
Though it is updated on a continuous basis.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

STS
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:0 STS: System time second
The value in this field indicates the current value in seconds of the System Time maintained
by the core.

1218/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Ethernet PTP time stamp low register (ETH_PTPTSLR)
Address offset: 0x070C
Reset value: 0x0000 0000
This register contains the least significant (lower) 32 time bits. This read-only register
contains the subsecond system time value.
30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

STPNS

31

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

STSS

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bit 31 STPNS: System time positive or negative sign
This bit indicates a positive or negative time value. When set, the bit indicates that time
representation is negative. When cleared, it indicates that time representation is positive.
Because the system time should always be positive, this bit is normally zero.
Bits 30:0 STSS: System time subseconds
The value in this field has the subsecond time representation, with 0.46 ns accuracy.

Ethernet PTP time stamp high update register (ETH_PTPTSHUR)
Address offset: 0x0710
Reset value: 0x0000 0000
This register contains the most significant (higher) 32 bits of the time to be written to, added
to, or subtracted from the System Time value. The Time stamp high update register, along
with the Time stamp update low register, initializes or updates the system time maintained
by the MAC. You have to write both of these registers before setting the TSSTI or TSSTU
bits in the Time stamp control register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

TSUS
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 TSUS: Time stamp update second
The value in this field indicates the time, in seconds, to be initialized or added to the system
time.

RM0090 Rev 17

1219/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Ethernet PTP time stamp low update register (ETH_PTPTSLUR)
Address offset: 0x0714
Reset value: 0x0000 0000
This register contains the least significant (lower) 32 bits of the time to be written to, added
to, or subtracted from the System Time value.
30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

TSUPNS

31

TSUSS

rw

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 31 TSUPNS: Time stamp update positive or negative sign
This bit indicates positive or negative time value. When set, the bit indicates that time
representation is negative. When cleared, it indicates that time representation is positive.
When TSSTI is set (system time initialization) this bit should be zero. If this bit is set when
TSSTU is set, the value in the Time stamp update registers is subtracted from the system
time. Otherwise it is added to the system time.
Bits 30:0 TSUSS: Time stamp update subseconds
The value in this field indicates the subsecond time to be initialized or added to the system
time. This value has an accuracy of 0.46 ns (in other words, a value of 0x0000_0001 is
0.46 ns).

Ethernet PTP time stamp addend register (ETH_PTPTSAR)
Address offset: 0x0718
Reset value: 0x0000 0000
This register is used by the software to readjust the clock frequency linearly to match the
master clock frequency. This register value is used only when the system time is configured
for Fine update mode (TSFCU bit in ETH_PTPTSCR). This register content is added to a
32-bit accumulator in every clock cycle and the system time is updated whenever the
accumulator overflows.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

TSA
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 TSA: Time stamp addend
This register indicates the 32-bit time value to be added to the Accumulator register to
achieve time synchronization.

1220/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Ethernet PTP target time high register (ETH_PTPTTHR)
Address offset: 0x071C
Reset value: 0x0000 0000
This register contains the higher 32 bits of time to be compared with the system time for
interrupt event generation. The Target time high register, along with Target time low register,
is used to schedule an interrupt event (TSARU bit in ETH_PTPTSCR) when the system time
exceeds the value programmed in these registers.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

TTSH
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 TTSH: Target time stamp high
This register stores the time in seconds. When the time stamp value matches or exceeds
both Target time stamp registers, the MAC, if enabled, generates an interrupt.

Ethernet PTP target time low register (ETH_PTPTTLR)
Address offset: 0x0720
Reset value: 0x0000 0000
This register contains the lower 32 bits of time to be compared with the system time for
interrupt event generation.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

TTSL
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 TTSL: Target time stamp low
This register stores the time in (signed) nanoseconds. When the value of the time stamp
matches or exceeds both Target time stamp registers, the MAC, if enabled, generates an
interrupt.

Ethernet PTP time stamp status register (ETH_PTPTSSR)
Address offset: 0x0728
Reset value: 0x0000 0000

Reserved

RM0090 Rev 17

9

8

7

6

5

4

3

2

1

0
TSSO

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

TSTTR

This register contains the time stamp status register.

ro

ro

1221/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Bits 31:2 Reserved, must be kept at reset value.
Bit 1 TSTTR: Time stamp target time reached
When set, this bit indicates that the value of the system time is greater than or equal to the
value specified in the Target time high and low registers. This bit is cleared when the
ETH_PTPTSSR register is read.
Bit 0 TSSO: Time stamp second overflow
When set, this bit indicates that the second value of the time stamp has overflowed beyond
0xFFFF FFFF.

Ethernet PTP PPS control register (ETH_PTPPPSCR)
Address offset: 0x072C
Reset value: 0x0000 0000
This register controls the frequency of the PPS output.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

7

6

5

4

3

2

1

0

PPSFREQ
ro

ro

Bits 31:4 Reserved, must be kept at reset value.
Bits 3:0 PPSFREQ: PPS frequency selection
The PPS output frequency is set to 2PPSFREQ Hz.
0000: 1 Hz with a pulse width of 125 ms for binary rollover and, of 100 ms for digital rollover
0001: 2 Hz with 50% duty cycle for binary rollover (digital rollover not recommended)
0010: 4 Hz with 50% duty cycle for binary rollover (digital rollover not recommended)
0011: 8 Hz with 50% duty cycle for binary rollover (digital rollover not recommended)
0100: 16 Hz with 50% duty cycle for binary rollover (digital rollover not recommended)
...
1111: 32768 Hz with 50% duty cycle for binary rollover (digital rollover not recommended)
Note: If digital rollover is used (TSSSR=1, bit 9 in ETH_PTPTSCR), it is recommended not to
use the PPS output with a frequency other than 1 Hz. Otherwise, with digital rollover,
the PPS output has irregular waveforms at higher frequencies (though its average
frequency will always be correct during any one-second window).

1222/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

33.8.4

DMA register description
This section defines the bits for each DMA register. Non-32 bit accesses are allowed as long
as the address is word-aligned.

Ethernet DMA bus mode register (ETH_DMABMR)
Address offset: 0x1000
Reset value: 0x0002 0101

rw

rw

rw

rw

rw

rw

rw

rw

rw

PM
rw

rw

8

PBL
rw

rw

rw

rw

rw

rw

7

6

5

4

3

2

rw

rw

DSL
rw

rw

rw

1

0
SR

USP

rw

RDP

9

EDFE

FPM

rw

FB

MB

Reserved

AAB

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

DA

The bus mode register establishes the bus operating modes for the DMA.

rw

rs

Bits 31:27 Reserved, must be kept at reset value.
Bit 26 MB: Mixed burst
When this bit is set high and the FB bit is low, the AHB master interface starts all bursts of a
length greater than 16 with INCR (undefined burst). When this bit is cleared, it reverts to
fixed burst transfers (INCRx and SINGLE) for burst lengths of 16 and below.
Bit 25 AAB: Address-aligned beats
When this bit is set high and the FB bit equals 1, the AHB interface generates all bursts
aligned to the start address LS bits. If the FB bit equals 0, the first burst (accessing the data
buffer’s start address) is not aligned, but subsequent bursts are aligned to the address.
Bit 24 FPM: 4xPBL mode
When set high, this bit multiplies the PBL value programmed (bits [22:17] and bits [13:8])
four times. Thus the DMA transfers data in a maximum of 4, 8, 16, 32, 64 and 128 beats
depending on the PBL value.
Bit 23 USP: Use separate PBL
When set high, it configures the RxDMA to use the value configured in bits [22:17] as PBL
while the PBL value in bits [13:8] is applicable to TxDMA operations only. When this bit is
cleared, the PBL value in bits [13:8] is applicable for both DMA engines.
Bits 22:17 RDP: Rx DMA PBL
These bits indicate the maximum number of beats to be transferred in one RxDMA
transaction. This is the maximum value that is used in a single block read/write operation.
The RxDMA always attempts to burst as specified in RDP each time it starts a burst transfer
on the host bus. RDP can be programmed with permissible values of 1, 2, 4, 8, 16, and 32.
Any other value results in undefined behavior.
These bits are valid and applicable only when USP is set high.
Bit 16 FB: Fixed burst
This bit controls whether the AHB Master interface performs fixed burst transfers or not.
When set, the AHB uses only SINGLE, INCR4, INCR8 or INCR16 during start of normal
burst transfers. When reset, the AHB uses SINGLE and INCR burst transfer operations.

RM0090 Rev 17

1223/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Bits 15:14 PM: Rx Tx priority ratio
RxDMA requests are given priority over TxDMA requests in the following ratio:
00: 1:1
01: 2:1
10: 3:1
11: 4:1
This is valid only when the DA bit is cleared.
Bits 13:8 PBL: Programmable burst length
These bits indicate the maximum number of beats to be transferred in one DMA transaction.
This is the maximum value that is used in a single block read/write operation. The DMA
always attempts to burst as specified in PBL each time it starts a burst transfer on the host
bus. PBL can be programmed with permissible values of 1, 2, 4, 8, 16, and 32. Any other
value results in undefined behavior. When USP is set, this PBL value is applicable for
TxDMA transactions only.
The PBL values have the following limitations:
– The maximum number of beats (PBL) possible is limited by the size of the Tx FIFO and Rx
FIFO.
– The FIFO has a constraint that the maximum beat supported is half the depth of the FIFO.
– If the PBL is common for both transmit and receive DMA, the minimum Rx FIFO and Tx
FIFO depths must be considered.
– Do not program out-of-range PBL values, because the system may not behave properly.
Bit 7 EDFE: Enhanced descriptor format enable
When this bit is set, the enhanced descriptor format is enabled and the descriptor size is
increased to 32 bytes (8 words). This is required when time stamping is activated (TSE=1,
ETH_PTPTSCR bit 0) or if IPv4 checksum offload is activated (IPCO=1, ETH_MACCR bit
10).
Bits 6:2 DSL: Descriptor skip length
This bit specifies the number of words to skip between two unchained descriptors. The
address skipping starts from the end of current descriptor to the start of next descriptor.
When DSL value equals zero, the descriptor table is taken as contiguous by the DMA, in
Ring mode.
Bit 1 DA: DMA Arbitration
0: Round-robin with Rx:Tx priority given in bits [15:14]
1: Rx has priority over Tx
Bit 0 SR: Software reset
When this bit is set, the MAC DMA controller resets all MAC Subsystem internal registers
and logic. It is cleared automatically after the reset operation has completed in all of the core
clock domains. Read a 0 value in this bit before re-programming any register of the core.

Ethernet DMA transmit poll demand register (ETH_DMATPDR)
Address offset: 0x1004
Reset value: 0x0000 0000
This register is used by the application to instruct the DMA to poll the transmit descriptor list.
The transmit poll demand register enables the Transmit DMA to check whether or not the
current descriptor is owned by DMA. The Transmit Poll Demand command is given to wake
up the TxDMA if it is in Suspend mode. The TxDMA can go into Suspend mode due to an
underflow error in a transmitted frame or due to the unavailability of descriptors owned by

1224/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller
transmit DMA. You can issue this command anytime and the TxDMA resets it once it starts
re-fetching the current descriptor from host memory.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

TPD
rw_wt

Bits 31:0 TPD: Transmit poll demand
When these bits are written with any value, the DMA reads the current descriptor pointed to
by the ETH_DMACHTDR register. If that descriptor is not available (owned by Host),
transmission returns to the Suspend state and ETH_DMASR register bit 2 is asserted. If the
descriptor is available, transmission resumes.

EHERNET DMA receive poll demand register (ETH_DMARPDR)
Address offset: 0x1008
Reset value: 0x0000 0000
This register is used by the application to instruct the DMA to poll the receive descriptor list.
The Receive poll demand register enables the receive DMA to check for new descriptors.
This command is given to wake up the RxDMA from Suspend state. The RxDMA can go into
Suspend state only due to the unavailability of descriptors owned by it.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

RPD
rw_wt

Bits 31:0 RPD: Receive poll demand
When these bits are written with any value, the DMA reads the current descriptor pointed to
by the ETH_DMACHRDR register. If that descriptor is not available (owned by Host),
reception returns to the Suspended state and ETH_DMASR register bit 7 is not asserted. If
the descriptor is available, the Receive DMA returns to active state.

Ethernet DMA receive descriptor list address register (ETH_DMARDLAR)
Address offset: 0x100C
Reset value: 0x0000 0000
The Receive descriptor list address register points to the start of the receive descriptor list.
The descriptor lists reside in the STM32F4xx's physical memory space and must be wordaligned. The DMA internally converts it to bus-width aligned address by making the
corresponding LS bits low. Writing to the ETH_DMARDLAR register is permitted only when
reception is stopped. When stopped, the ETH_DMARDLAR register must be written to
before the receive Start command is given.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

SRL
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

RM0090 Rev 17

rw

rw

rw

1225/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Bits 31:0 SRL: Start of receive list
This field contains the base address of the first descriptor in the receive descriptor list. The
LSB bits [1/2/3:0] for 32/64/128-bit bus width) are internally ignored and taken as all-zero by
the DMA. Hence these LSB bits are read only.

Ethernet DMA transmit descriptor list address register (ETH_DMATDLAR)
Address offset: 0x1010
Reset value: 0x0000 0000
The Transmit descriptor list address register points to the start of the transmit descriptor list.
The descriptor lists reside in the STM32F4xx's physical memory space and must be wordaligned. The DMA internally converts it to bus-width-aligned address by taking the
corresponding LSB to low. Writing to the ETH_DMATDLAR register is permitted only when
transmission has stopped. Once transmission has stopped, the ETH_DMATDLAR register
can be written before the transmission Start command is given.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

STL
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 STL: Start of transmit list
This field contains the base address of the first descriptor in the transmit descriptor list. The
LSB bits [1/2/3:0] for 32/64/128-bit bus width) are internally ignored and taken as all-zero
by the DMA. Hence these LSB bits are read-only.

Ethernet DMA status register (ETH_DMASR)
Address offset: 0x1014
Reset value: 0x0000 0000

1226/1747

r

r

r

r

r

r

r

rc- rc- rc- rcw1 w1 w1 w1

RM0090 Rev 17

4

3

2

1

0

TBUS

TPSS

TS

5

ROS

RBUS

6

TJTS

7

RS

8

TUS

9

RPSS

ETS

ERS

FBES

AIS

NIS
r

Reserved

r

RPS

r

TPS

r

EBS

MMCS

r

Reserved

TSTS

PMTS

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

RWTS

The Status register contains all the status bits that the DMA reports to the application. The
ETH_DMASR register is usually read by the software driver during an interrupt service
routine or polling. Most of the fields in this register cause the host to be interrupted. The
ETH_DMASR register bits are not cleared when read. Writing 1 to (unreserved) bits in
ETH_DMASR register[16:0] clears them and writing 0 has no effect. Each field (bits [16:0])
can be masked by masking the appropriate bit in the ETH_DMAIER register.

rc- rc- rc- rc- rc- rc- rc- rc- rc- rc- rcw1 w1 w1 w1 w1 w1 w1 w1 w1 w1 w1

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Bits 31:30 Reserved, must be kept at reset value.
Bit 29 TSTS: Time stamp trigger status
This bit indicates an interrupt event in the MAC core's Time stamp generator block. The
software must read the MAC core’s status register, clearing its source (bit 9), to reset this bit
to 0. When this bit is high an interrupt is generated if enabled.
Bit 28 PMTS: PMT status
This bit indicates an event in the MAC core’s PMT. The software must read the
corresponding registers in the MAC core to get the exact cause of interrupt and clear its
source to reset this bit to 0. The interrupt is generated when this bit is high if enabled.
Bit 27 MMCS: MMC status
This bit reflects an event in the MMC of the MAC core. The software must read the
corresponding registers in the MAC core to get the exact cause of interrupt and clear the
source of interrupt to make this bit as 0. The interrupt is generated when this bit is high if
enabled.
Bit 26 Reserved, must be kept at reset value.
Bits 25:23 EBS: Error bits status
These bits indicate the type of error that caused a bus error (error response on the AHB
interface). Valid only with the fatal bus error bit (ETH_DMASR register [13]) set. This field
does not generate an interrupt.
Bit 231 Error during data transfer by TxDMA
0 Error during data transfer by RxDMA
Bit 24 1 Error during read transfer
0 Error during write transfer
Bit 25 1 Error during descriptor access
0 Error during data buffer access
Bits 22:20 TPS: Transmit process state
These bits indicate the Transmit DMA FSM state. This field does not generate an interrupt.
000: Stopped; Reset or Stop Transmit Command issued
001: Running; Fetching transmit transfer descriptor
010: Running; Waiting for status
011: Running; Reading Data from host memory buffer and queuing it to transmit buffer (Tx
FIFO)
100, 101: Reserved for future use
110: Suspended; Transmit descriptor unavailable or transmit buffer underflow
111: Running; Closing transmit descriptor
Bits 19:17 RPS: Receive process state
These bits indicate the Receive DMA FSM state. This field does not generate an interrupt.
000: Stopped: Reset or Stop Receive Command issued
001: Running: Fetching receive transfer descriptor
010: Reserved for future use
011: Running: Waiting for receive packet
100: Suspended: Receive descriptor unavailable
101: Running: Closing receive descriptor
110: Reserved for future use
111: Running: Transferring the receive packet data from receive buffer to host memory

RM0090 Rev 17

1227/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Bit 16 NIS: Normal interrupt summary
The normal interrupt summary bit value is the logical OR of the following when the
corresponding interrupt bits are enabled in the ETH_DMAIER register:
– ETH_DMASR [0]: Transmit interrupt
– ETH_DMASR [2]: Transmit buffer unavailable
– ETH_DMASR [6]: Receive interrupt
– ETH_DMASR [14]: Early receive interrupt
Only unmasked bits affect the normal interrupt summary bit.
This is a sticky bit and it must be cleared (by writing a 1 to this bit) each time a corresponding
bit that causes NIS to be set is cleared.
Bit 15 AIS: Abnormal interrupt summary
The abnormal interrupt summary bit value is the logical OR of the following when the
corresponding interrupt bits are enabled in the ETH_DMAIER register:
– ETH_DMASR [1]:Transmit process stopped
– ETH_DMASR [3]:Transmit jabber timeout
– ETH_DMASR [4]: Receive FIFO overflow
– ETH_DMASR [5]: Transmit underflow
– ETH_DMASR [7]: Receive buffer unavailable
– ETH_DMASR [8]: Receive process stopped
– ETH_DMASR [9]: Receive watchdog timeout
– ETH_DMASR [10]: Early transmit interrupt
– ETH_DMASR [13]: Fatal bus error
Only unmasked bits affect the abnormal interrupt summary bit.
This is a sticky bit and it must be cleared each time a corresponding bit that causes AIS to be
set is cleared.
Bit 14 ERS: Early receive status
This bit indicates that the DMA had filled the first data buffer of the packet. Receive Interrupt
ETH_DMASR [6] automatically clears this bit.
Bit 13 FBES: Fatal bus error status
This bit indicates that a bus error occurred, as detailed in [25:23]. When this bit is set, the
corresponding DMA engine disables all its bus accesses.
Bits 12:11 Reserved, must be kept at reset value.
Bit 10 ETS: Early transmit status
This bit indicates that the frame to be transmitted was fully transferred to the Transmit FIFO.
Bit 9

RWTS: Receive watchdog timeout status
This bit is asserted when a frame with a length greater than 2 048 bytes is received.

Bit 8 RPSS: Receive process stopped status
This bit is asserted when the receive process enters the Stopped state.
Bit 7 RBUS: Receive buffer unavailable status
This bit indicates that the next descriptor in the receive list is owned by the host and cannot
be acquired by the DMA. Receive process is suspended. To resume processing receive
descriptors, the host should change the ownership of the descriptor and issue a Receive Poll
Demand command. If no Receive Poll Demand is issued, receive process resumes when the
next recognized incoming frame is received. ETH_DMASR [7] is set only when the previous
receive descriptor was owned by the DMA.

1228/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Bit 6 RS: Receive status
This bit indicates the completion of the frame reception. Specific frame status information
has been posted in the descriptor. Reception remains in the Running state.
Bit 5 TUS: Transmit underflow status
This bit indicates that the transmit buffer had an underflow during frame transmission.
Transmission is suspended and an underflow error TDES0[1] is set.
Bit 4 ROS: Receive overflow status
This bit indicates that the receive buffer had an overflow during frame reception. If the partial
frame is transferred to the application, the overflow status is set in RDES0[11].
Bit 3 TJTS: Transmit jabber timeout status
This bit indicates that the transmit jabber timer expired, meaning that the transmitter had
been excessively active. The transmission process is aborted and placed in the Stopped
state. This causes the transmit jabber timeout TDES0[14] flag to be asserted.
Bit 2 TBUS: Transmit buffer unavailable status
This bit indicates that the next descriptor in the transmit list is owned by the host and cannot
be acquired by the DMA. Transmission is suspended. Bits [22:20] explain the transmit
process state transitions. To resume processing transmit descriptors, the host should
change the ownership of the bit of the descriptor and then issue a Transmit Poll Demand
command.
Bit 1 TPSS: Transmit process stopped status
This bit is set when the transmission is stopped.
Bit 0 TS: Transmit status
This bit indicates that frame transmission is finished and TDES1[31] is set in the first
descriptor.

Ethernet DMA operation mode register (ETH_DMAOMR)
Address offset: 0x1018
Reset value: 0x0000 0000

RM0090 Rev 17

rw

rw

4

3
RTC

5

rw

rw

2

1
SR

rw

6

rw

rw

0
Reserved

rw

7

OSF

rw

Reserved

8

FUGF

rw

9

Reserved

rs

Reserved

ST

rw

TTC

rw

FTF

rw

TSF

DFRF

rw

Reserved

RSF

Reserved

DTCEFD

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

FEF

The operation mode register establishes the Transmit and Receive operating modes and
commands. The ETH_DMAOMR register should be the last CSR to be written as part of
DMA initialization.

1229/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Bits 31:27 Reserved, must be kept at reset value.
Bit 26 DTCEFD: Dropping of TCP/IP checksum error frames disable
When this bit is set, the core does not drop frames that only have errors detected by the
receive checksum offload engine. Such frames do not have any errors (including FCS error)
in the Ethernet frame received by the MAC but have errors in the encapsulated payload only.
When this bit is cleared, all error frames are dropped if the FEF bit is reset.
Bit 25 RSF: Receive store and forward
When this bit is set, a frame is read from the Rx FIFO after the complete frame has been
written to it, ignoring RTC bits. When this bit is cleared, the Rx FIFO operates in Cut-through
mode, subject to the threshold specified by the RTC bits.
Bit 24 DFRF: Disable flushing of received frames
When this bit is set, the RxDMA does not flush any frames due to the unavailability of
receive descriptors/buffers as it does normally when this bit is cleared (see Receive process
suspended on page 1180).
Bits 23:22 Reserved, must be kept at reset value.
Bit 21 TSF: Transmit store and forward
When this bit is set, transmission starts when a full frame resides in the Transmit FIFO.
When this bit is set, the TTC values specified by the ETH_DMAOMR register bits [16:14] are
ignored.
When this bit is cleared, the TTC values specified by the ETH_DMAOMR register bits
[16:14] are taken into account.
This bit should be changed only when transmission is stopped.
Bit 20 FTF: Flush transmit FIFO
When this bit is set, the transmit FIFO controller logic is reset to its default values and thus
all data in the Tx FIFO are lost/flushed. This bit is cleared internally when the flushing
operation is complete. The Operation mode register should not be written to until this bit is
cleared.
Bits 19:17 Reserved, must be kept at reset value.
Bits 16:14 TTC: Transmit threshold control
These three bits control the threshold level of the Transmit FIFO. Transmission starts when
the frame size within the Transmit FIFO is larger than the threshold. In addition, full frames
with a length less than the threshold are also transmitted. These bits are used only when the
TSF bit (Bit 21) is cleared.
000: 64
001: 128
010: 192
011: 256
100: 40
101: 32
110: 24
111: 16

1230/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Bit 13 ST: Start/stop transmission
When this bit is set, transmission is placed in the Running state, and the DMA checks the
transmit list at the current position for a frame to be transmitted. Descriptor acquisition is
attempted either from the current position in the list, which is the transmit list base address
set by the ETH_DMATDLAR register, or from the position retained when transmission was
stopped previously. If the current descriptor is not owned by the DMA, transmission enters
the Suspended state and the transmit buffer unavailable bit (ETH_DMASR [2]) is set. The
Start Transmission command is effective only when transmission is stopped. If the command
is issued before setting the DMA ETH_DMATDLAR register, the DMA behavior is
unpredictable.
When this bit is cleared, the transmission process is placed in the Stopped state after
completing the transmission of the current frame. The next descriptor position in the transmit
list is saved, and becomes the current position when transmission is restarted. The Stop
Transmission command is effective only when the transmission of the current frame is
complete or when the transmission is in the Suspended state.
Bits 12:8 Reserved, must be kept at reset value.
Bit 7 FEF: Forward error frames
When this bit is set, all frames except runt error frames are forwarded to the DMA.
When this bit is cleared, the Rx FIFO drops frames with error status (CRC error, collision
error, giant frame, watchdog timeout, overflow). However, if the frame’s start byte (write)
pointer is already transferred to the read controller side (in Threshold mode), then the
frames are not dropped. The Rx FIFO drops the error frames if that frame's start byte is not
transferred (output) on the ARI bus.
Bit 6 FUGF: Forward undersized good frames
When this bit is set, the Rx FIFO forwards undersized frames (frames with no error and
length less than 64 bytes) including pad-bytes and CRC).
When this bit is cleared, the Rx FIFO drops all frames of less than 64 bytes, unless such a
frame has already been transferred due to lower value of receive threshold (e.g., RTC = 01).
Bit 5 Reserved, must be kept at reset value.
Bits 4:3 RTC: Receive threshold control
These two bits control the threshold level of the Receive FIFO. Transfer (request) to DMA
starts when the frame size within the Receive FIFO is larger than the threshold. In addition,
full frames with a length less than the threshold are transferred automatically.
Note: Note that value of 11 is not applicable if the configured Receive FIFO size is 128 bytes.
Note: These bits are valid only when the RSF bit is zero, and are ignored when the RSF bit is
set to 1.
00: 64
01: 32
10: 96
11: 128

RM0090 Rev 17

1231/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Bit 2 OSF: Operate on second frame
When this bit is set, this bit instructs the DMA to process a second frame of Transmit data
even before status for first frame is obtained.
Bit 1 SR: Start/stop receive
When this bit is set, the receive process is placed in the Running state. The DMA attempts to
acquire the descriptor from the receive list and processes incoming frames. Descriptor
acquisition is attempted from the current position in the list, which is the address set by the
DMA ETH_DMARDLAR register or the position retained when the receive process was
previously stopped. If no descriptor is owned by the DMA, reception is suspended and the
receive buffer unavailable bit (ETH_DMASR [7]) is set. The Start Receive command is
effective only when reception has stopped. If the command was issued before setting the
DMA ETH_DMARDLAR register, the DMA behavior is unpredictable.
When this bit is cleared, RxDMA operation is stopped after the transfer of the current frame.
The next descriptor position in the receive list is saved and becomes the current position
when the receive process is restarted. The Stop Receive command is effective only when
the Receive process is in either the Running (waiting for receive packet) or the Suspended
state.
Bit 0 Reserved, must be kept at reset value.

1232/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Ethernet DMA interrupt enable register (ETH_DMAIER)
Address offset: 0x101C
Reset value: 0x0000 0000

5

4

3

2

1

0

RIE

TUIE

ROIE

TJTIE

TBUIE

TPSIE

TIE

rw

6

RBUIE

rw

7

RPSIE

ERIE

FBEIE

rw

8

RWTIE

AISE

rw

Reserved

NISE

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

Reserved

9

ETIE

The Interrupt enable register enables the interrupts reported by ETH_DMASR. Setting a bit
to 1 enables a corresponding interrupt. After a hardware or software reset, all interrupts are
disabled.

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:17 Reserved, must be kept at reset value.
Bit 16 NISE: Normal interrupt summary enable
When this bit is set, a normal interrupt is enabled. When this bit is cleared, a normal
interrupt is disabled. This bit enables the following bits:
– ETH_DMASR [0]: Transmit Interrupt
– ETH_DMASR [2]: Transmit buffer unavailable
– ETH_DMASR [6]: Receive interrupt
– ETH_DMASR [14]: Early receive interrupt
Bit 15 AISE: Abnormal interrupt summary enable
When this bit is set, an abnormal interrupt is enabled. When this bit is cleared, an abnormal
interrupt is disabled. This bit enables the following bits:
– ETH_DMASR [1]: Transmit process stopped
– ETH_DMASR [3]: Transmit jabber timeout
– ETH_DMASR [4]: Receive overflow
– ETH_DMASR [5]: Transmit underflow
– ETH_DMASR [7]: Receive buffer unavailable
– ETH_DMASR [8]: Receive process stopped
– ETH_DMASR [9]: Receive watchdog timeout
– ETH_DMASR [10]: Early transmit interrupt
– ETH_DMASR [13]: Fatal bus error
Bit 14 ERIE: Early receive interrupt enable
When this bit is set with the normal interrupt summary enable bit (ETH_DMAIER
register[16]), the early receive interrupt is enabled.
When this bit is cleared, the early receive interrupt is disabled.
Bit 13 FBEIE: Fatal bus error interrupt enable
When this bit is set with the abnormal interrupt summary enable bit (ETH_DMAIER
register[15]), the fatal bus error interrupt is enabled.
When this bit is cleared, the fatal bus error enable interrupt is disabled.
Bits 12:11 Reserved, must be kept at reset value.
Bit 10 ETIE: Early transmit interrupt enable
When this bit is set with the abnormal interrupt summary enable bit (ETH_DMAIER register
[15]), the early transmit interrupt is enabled.
When this bit is cleared, the early transmit interrupt is disabled.

RM0090 Rev 17

1233/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Bit 9 RWTIE: receive watchdog timeout interrupt enable
When this bit is set with the abnormal interrupt summary enable bit (ETH_DMAIER
register[15]), the receive watchdog timeout interrupt is enabled.
When this bit is cleared, the receive watchdog timeout interrupt is disabled.
Bit 8 RPSIE: Receive process stopped interrupt enable
When this bit is set with the abnormal interrupt summary enable bit (ETH_DMAIER
register[15]), the receive stopped interrupt is enabled. When this bit is cleared, the receive
stopped interrupt is disabled.
Bit 7 RBUIE: Receive buffer unavailable interrupt enable
When this bit is set with the abnormal interrupt summary enable bit (ETH_DMAIER
register[15]), the receive buffer unavailable interrupt is enabled.
When this bit is cleared, the receive buffer unavailable interrupt is disabled.
Bit 6 RIE: Receive interrupt enable
When this bit is set with the normal interrupt summary enable bit (ETH_DMAIER
register[16]), the receive interrupt is enabled.
When this bit is cleared, the receive interrupt is disabled.
Bit 5 TUIE: Underflow interrupt enable
When this bit is set with the abnormal interrupt summary enable bit (ETH_DMAIER
register[15]), the transmit underflow interrupt is enabled.
When this bit is cleared, the underflow interrupt is disabled.
Bit 4

ROIE: Overflow interrupt enable
When this bit is set with the abnormal interrupt summary enable bit (ETH_DMAIER
register[15]), the receive overflow interrupt is enabled.
When this bit is cleared, the overflow interrupt is disabled.

Bit 3 TJTIE: Transmit jabber timeout interrupt enable
When this bit is set with the abnormal interrupt summary enable bit (ETH_DMAIER
register[15]), the transmit jabber timeout interrupt is enabled.
When this bit is cleared, the transmit jabber timeout interrupt is disabled.
Bit 2 TBUIE: Transmit buffer unavailable interrupt enable
When this bit is set with the normal interrupt summary enable bit (ETH_DMAIER
register[16]), the transmit buffer unavailable interrupt is enabled.
When this bit is cleared, the transmit buffer unavailable interrupt is disabled.
Bit 1 TPSIE: Transmit process stopped interrupt enable
When this bit is set with the abnormal interrupt summary enable bit (ETH_DMAIER
register[15]), the transmission stopped interrupt is enabled.
When this bit is cleared, the transmission stopped interrupt is disabled.
Bit 0 TIE: Transmit interrupt enable
When this bit is set with the normal interrupt summary enable bit (ETH_DMAIER
register[16]), the transmit interrupt is enabled.
When this bit is cleared, the transmit interrupt is disabled.

The Ethernet interrupt is generated only when the TSTS or PMTS bits of the DMA Status
register is asserted with their corresponding interrupt are unmasked, or when the NIS/AIS
Status bit is asserted and the corresponding Interrupt Enable bits (NISE/AISE) are enabled.

1234/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Ethernet DMA missed frame and buffer overflow counter register
(ETH_DMAMFBOCR)
Address offset: 0x1020
Reset value: 0x0000 0000
The DMA maintains two counters to track the number of missed frames during reception.
This register reports the current value of the counter. The counter is used for diagnostic
purposes. Bits [15:0] indicate missed frames due to the STM32F4xx buffer being
unavailable (no receive descriptor was available). Bits [27:17] indicate missed frames due to
Rx FIFO overflow conditions and runt frames (good frames of less than 64 bytes).
9

OMFC

MFA

Reserved

OFOC

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

8

7

6

5

4

3

2

1

0

MFC

rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r

Bits 31:29 Reserved, must be kept at reset value.
Bit 28 OFOC: Overflow bit for FIFO overflow counter
Bits 27:17 MFA: Missed frames by the application
Indicates the number of frames missed by the application
Bit 16 OMFC: Overflow bit for missed frame counter
Bits 15:0 MFC: Missed frames by the controller
Indicates the number of frames missed by the Controller due to the host receive buffer being
unavailable. This counter is incremented each time the DMA discards an incoming frame.

Ethernet DMA receive status watchdog timer register (ETH_DMARSWTR)
Address offset: 0x1024
Reset value: 0x0000 0000
This register, when written with a non-zero value, enables the watchdog timer for the receive
status (RS, ETH_DMASR[6]).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

rw

rw

rw

4

3

2

1

0

rw

rw

rw

RSWTC
Reserved
rw

rw

Bits 31:8 Reserved, must be kept at reset value.
Bits 7:0 RSWTC: Receive status (RS) watchdog timer count
Indicates the number of HCLK clock cycles multiplied by 256 for which the watchdog timer
is set. The watchdog timer gets triggered with the programmed value after the RxDMA
completes the transfer of a frame for which the RS status bit is not set due to the setting of
RDES1[31] in the corresponding descriptor. When the watchdog timer runs out, the RS bit
is set and the timer is stopped. The watchdog timer is reset when the RS bit is set high due
to automatic setting of RS as per RDES1[31] of any received frame.

RM0090 Rev 17

1235/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Ethernet DMA current host transmit descriptor register (ETH_DMACHTDR)
Address offset: 0x1048
Reset value: 0x0000 0000
The Current host transmit descriptor register points to the start address of the current
transmit descriptor read by the DMA.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

HTDAP
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:0 HTDAP: Host transmit descriptor address pointer
Cleared . Pointer updated by DMA during operation.

Ethernet DMA current host receive descriptor register (ETH_DMACHRDR)
Address offset: 0x104C
Reset value: 0x0000 0000
The Current host receive descriptor register points to the start address of the current receive
descriptor read by the DMA.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

HRDAP
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:0 HRDAP: Host receive descriptor address pointer
Cleared On Reset. Pointer updated by DMA during operation.

Ethernet DMA current host transmit buffer address register
(ETH_DMACHTBAR)
Address offset: 0x1050
Reset value: 0x0000 0000
The Current host transmit buffer address register points to the current transmit buffer
address being read by the DMA.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

HTBAP
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:0 HTBAP: Host transmit buffer address pointer
Cleared On Reset. Pointer updated by DMA during operation.

1236/1747

RM0090 Rev 17

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

Ethernet DMA current host receive buffer address register
(ETH_DMACHRBAR)
Address offset: 0x1054
Reset value: 0x0000 0000
The current host receive buffer address register points to the current receive buffer address
being read by the DMA.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

HRBAP
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:0 HRBAP: Host receive buffer address pointer
Cleared On Reset. Pointer updated by DMA during operation.

33.8.5

Ethernet register maps
Table 195 gives the ETH register map and reset values.

0

2

1

3

RE

PM

Reserved

4

TE

5

DC

6
BL

9
RD

HU

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

HTL[31:0]
0

0

0

0

0

0

0

0

0

0

0

0

0

PA

Reserved
0

ETH_MACFCR

0

0

0

0

PT
0

0

0

0

0

0

0

0

0

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

PLT

MD

Reserved

Reset value

0

MR

M M
W B

0

0

0

0

0

Reserved
0

0

0

0

0

0

0

0

FCB/BPA

0

TFCE

0

RFCE

0

CR

0

Reserved

0

Reset value
0x28

0
HM

0

ETH_MACMIIDR

ETH_
MACVLANTR

0

UPFD

0x1C

0

DAIF

0

ETH_MACMIIAR

Reset value

0

0

VLANTC

0x18

0

PAM

0

Reset value
0x14

0

BFD

11

10
IPCO

0

7

12

DM

0

8

13

LM

0

APCS

14

ROD

0

Reserved

15

FES

16

Reserved

17

CSD

18

19

20

21
Reserved

23

JD

22

24

WD

0

HTH[31:0]

ETH_MACHTLR
Reset value

0

Reserved

ETH_MACHTHR
Reset value

0

PCF

0

0

ZQPD

0x10

Reset value

0

SAF

0x0C

ETH_MACFFR

0

SAIF

0x08

0

IFG

HPF

0x04

0

0
RA

Reset value

25

26

27

28

29

Reserved

eserved

ETH_MACCR

CSTF

0x00

Register

30

Offset

31

Table 195. Ethernet register map and reset values

0

0

0

0

0

0

0

0

0

0

0

0

VLANTI

0

0

0

0

0

0

0

0

0

0

0

ETH_
MACRWUFFR

Frame filter reg0\Frame filter reg1\Frame filter reg2\Frame filter reg3\Frame filter reg4\...\Frame filter reg7

Reset value

0

RM0090 Rev 17

1237/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

Reserved

ETH_MACIMR

Reserved

0

0

0

0

0

0

0

AE

SA
0

0

0

0

0

0

1

1

1

1

1

1

1

1

1

MBC[6:0]
0

0

0

0

1

1

1

0

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

Reserved
0

0

1

1

ETH_MACA2HR
Reset value

0

0

1

1

1

1

1

1

1

1

1

1

MBC
0

0

0

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

Reserved

0

0

0

0

0

1

1

1

1

1

1

1

1

1

1

MBC
0

0

0

1

1

1

1

2

1

0

MPE

PD
MMRPEA

3

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

MACA3H

Reserved

0

0

0

1

ETH_MACA3LR

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

0

0

0

0

0

0

MACA3L
1

1

1

1

1

1

1

1

1

1

1

ETH_MMCCR

1238/1747

WFE

1

CR

1

Reset value

1

1

1

1

1

1

Reserved

0

RM0090 Rev 17

0

0

0

0

Reserved

Reserved

Reserved

RFCEM

Reserved

RGUFM

0

TGFSCS

Reserved

TGFMSCS

Reserved

TGFS

0

RFCES

Reserved

RFAES

RGUFS

Reserved

RFAEM

Reset value

MSFRWCS

1

CSR

1

ETH_MACA3HR

ETH_MMCRIMR

4

1

MACA2L

Reset value

0x10C

Reserved
0

MACA2H

ETH_MACA2LR

ETH_MMCTIR

Reserve
d

MACA1L

Reset value

0x108

Reserve
d

MACA1H

ETH_MACA1LR

ETH_MMCRIR

Reserved

5

Reserved

Reset value

0x104

RFWRA

6

MPR

RFRCS

7

8
Reserved

WFR

0

0

ROR

0x100

1

0

Reset value

0

0

MCF

0x5C

0

0

MCP

0x58

1

Reset value

Reset value

0

0

MACA0L

ETH_MACA1HR

Reset value

0

0

MCFHP

0x54

0

SA

0x50

1

Reset value

0

0

MACA0H

ETH_MACA0LR

AE

0x4C

Reset value

Reserved

SA

0x48

ETH_MACA0HR

AE

0x44

0

0
MO

Reset value
0x40

0

TSTIM

0

0

PMTS

0

PMTIM

0

ETH_MACSR

Reserved

RFFL

MMTEA
0

MMCS

0

0

MMCRS

0

0

MMCTS

0

MTFCS

MTP

TFRS
0

Reserved

0

Reset value

0x3C

9
GU

11

10

12

13

14

15

16

17

18

19

20

21

23

22
TFWA

0

Reserved

TSTS

0x38

24

25
0

Reset value

Reserved

ETH_
MACDBGR

TFNEGU

0
Reserved

0x34

26

0

Reserved

TFF

Reset value

27

ETH_
MACPMTCSR

28

31

0x2C

29

Register

30

Offset

WFFRPR

Table 195. Ethernet register map and reset values (continued)

0

0

Reserved

RM0090

Ethernet (ETH): media access control (MAC) with DMA controller

0x710

0x714

0x718
0x71C
0x720

0

5

6

7

8

9

11

10

12

14

13

15

TGFSCM

16

TGFMSCM

17

18

19

20

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

TGFC
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

RFCEC
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

RFAEC
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Reserved

0

0

0

0

0

0

TSE

0

TSFCU

0

TSSTI

0

TSSTU

0

Reserved

RGUFC

0

0

0

0

STSSI

Reserved

Reset value
ETH_PTPTSHR

0

0

0

0

0

0

0

0

STS[31:0]

Reset value

0

ETH_PTPTSLR
Reset value

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

STSS
0

0

0

0

0

0

0

0

0

0

0

0

0

0

ETH_PTPTSHU
R

0

0

TSUS

Reset value

0

ETH_PTPTSLU
R
Reset value

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

TSUSS
0

0

0

0

0

0

0

0

0

0

0

0

0

0

ETH_PTPTSAR

0

0

TSA
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

ETH_PTPTTHR

0

0

TTSH
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

ETH_PTPTTLR
Reset value

2

0

ETH_PTPSSIR

Reset value

3

0

ETH_PTPTSCR

Reset value

4

0

STPNS

0x70C

0

TSITE

0

TGFMSCC

TSUPNS

0x708

0

TTSARU

0

Reset value
0x704

0

TSSSR

0x700

0

TSSARFE

Reset value

0

TSPTPPSV2E

Reset value

0

TSSPTPOEFE

0x1C4

ETH_MMCRGU
FCR

Reset value

0

TSSIPV6FE

0x198

ETH_MMCRFAE
CR

Reset value

0

TSSIPV4FE

0x194

ETH_MMCRFC
ECR

Reset value

0

TSSEME

0x168

ETH_MMCTGF
CR

Reset value

0

Reserved

TGFSCC

TSCNT

0x150

ETH_MMCTGF
MSCCR

0

TSPFFMAE

0x14C

Reserved

TSSMRME

Reserved

Reset value
ETH_MMCTGFS
CCR

21

22

23

24

25

26

27

28

29

ETH_MMCTIMR

TGFM

0x110

Register

30

Offset

31

Table 195. Ethernet register map and reset values (continued)

0

0

TTSL
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

RM0090 Rev 17

0

0

1239/1747
1240

Ethernet (ETH): media access control (MAC) with DMA controller

RM0090

1

0
TSSO

0

0

ETH_PTPPPSC
R

PPS
FREQ

Reserved

Reserved

Reset value

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Reset value

0

0

0

0

0

0

0

0

ETH_
DMARSWTR

0

0

0

0

0

0x104C

0x1050

0x1054

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

TPSS

TS

SR

TBUS
OSF

TJTS

TBUIE

TPSIE

TIE

RTC

0

0

0

0

0

0

0

0

0

MFC
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

HTBAP
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

HRBAP
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Refer to Section 2.3: Memory map for the register boundary addresses.

1240/1747

0

HRDAP

ETH_
DMACHRBAR
Reset value

0

HTDAP

ETH_
DMACHTBAR
Reset value

0

RSWTC

ETH_
DMACHRDR
Reset value

0

0

Reserved

ETH_
DMACHTDR
Reset value

0

0

0

Reset value
0x1048

0

0

0

OMFC

MFA

OFOC

Reserve
d

Reserved

0

TJTIE

0

ROS

0

0

ROIE

0

TUS

0

0
Reserved

0

RS

0

0
FUGF

0

0

TUIE

Reserved

0

0

RIE

Reserve
d

RBUS

0

FEF

0

RBUIE

0

RPSS

TPS
Reserved

0

0

RPSIE

0

0

RWTIE

0

0

0

Reserved

0

0

ETIE

ETH_DMAIER

0

0

Reserved

0

0

0

FBES

0

0

0

FBEIE

0

0

0

ERS

0

0

ERIE

0

0

AIS

0

0

TTC

0

0

AISE

0

Reserved

0

NIS

0

Reset value

0x1024

0

NISE

0

RPS

0

FTF

0

TSF

0

EBS

0

DFRF

0

ETH_DMAOMR

ETH_DMAMFB
OCR

0

STL

Reset value

0x1020

0

SRL

RSF

Reset value

0

Reserved

ETH_DMASR

0

DTCEFD

Reset value

0

RPD

MMCS

0x1010

Reset value

0

SR

0

EDFE

FB

0

DA

0

DSL

RWTS

0

PBL

0

TPD

PMTS

Reset value

ETH_DMATDLA
R

0x101C

0

PM

ST

Reset value

0x100C

0x1018

0

ETH_DMARPDR
ETH_DMARDLA
R

0x1014

0

TSTS

0x1008

0

ETH_DMATPDR

Reserved

0x1004

0

RDP

0

ETS

Reset value

USP

Reserved

FPM

ETH_DMABMR

AAB

0
MB

0x1000

2

3

4

5

6

7

8

9

11

10

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

Reset value

ETH_PTPTSSR

Reserved

0x72C

TSTTR

0x728

Register

30

Offset

31

Table 195. Ethernet register map and reset values (continued)

RM0090 Rev 17

RM0090

34

USB on-the-go full-speed (OTG_FS)

USB on-the-go full-speed (OTG_FS)
This section applies to the whole STM32F4xx family, unless otherwise specified.

34.1

OTG_FS introduction
Portions Copyright (c) 2004, 2005 Synopsys, Inc. All rights reserved. Used with permission.
This section presents the architecture and the programming model of the OTG_FS
controller.
The following acronyms are used throughout this section:
FS

Full-speed

LS

Low-speed

MAC

Media access controller

OTG

On-the-go

PFC

Packet FIFO controller

PHY

Physical layer

USB

Universal serial bus

UTMI

USB 2.0 transceiver macrocell interface (UTMI)

References are made to the following documents:
•

USB On-The-Go Supplement, Revision 1.3

•

Universal Serial Bus Revision 2.0 Specification

The OTG_FS is a dual-role device (DRD) controller that supports both device and host
functions and is fully compliant with the On-The-Go Supplement to the USB 2.0
Specification. It can also be configured as a host-only or device-only controller, fully
compliant with the USB 2.0 Specification. In host mode, the OTG_FS supports full-speed
(FS, 12 Mbits/s) and low-speed (LS, 1.5 Mbits/s) transfers whereas in device mode, it only
supports full-speed (FS, 12 Mbits/s) transfers. The OTG_FS supports both HNP and SRP.
The only external device required is a charge pump for VBUS in host mode.

RM0090 Rev 17

1241/1747
1379

USB on-the-go full-speed (OTG_FS)

34.2

RM0090

OTG_FS main features
The main features can be divided into three categories: general, host-mode and devicemode features.

34.2.1

General features
The OTG_FS interface general features are the following:
•

It is USB-IF certified to the Universal Serial Bus Specification Rev 2.0

•

It includes full support (PHY) for the optional On-The-Go (OTG) protocol detailed in the
On-The-Go Supplement Rev 1.3 specification

•

•

Integrated support for A-B Device Identification (ID line)

–

Integrated support for host Negotiation Protocol (HNP) and Session Request
Protocol (SRP)

–

It allows host to turn VBUS off to conserve battery power in OTG applications

–

It supports OTG monitoring of VBUS levels with internal comparators

–

It supports dynamic host-peripheral switch of role

It is software-configurable to operate as:
–

SRP capable USB FS Peripheral (B-device)

–

SRP capable USB FS/LS host (A-device)

–

USB On-The-Go Full-Speed Dual Role device

It supports FS SOF and LS Keep-alives with
–

SOF pulse PAD connectivity (OTG_FS_SOF)

–

SOF pulse internal connection to timer2 (TIM2)

–

Configurable framing period

–

Configurable end of frame interrupt

•

It includes power saving features such as system stop during USB Suspend, switch-off
of clock domains internal to the digital core, PHY and DFIFO power management

•

It features a dedicated RAM of 1.25 Kbytes with advanced FIFO control:

•

1242/1747

–

–

Configurable partitioning of RAM space into different FIFOs for flexible and
efficient use of RAM

–

Each FIFO can hold multiple packets

–

Dynamic memory allocation

–

Configurable FIFO sizes that are not powers of 2 to allow the use of contiguous
memory locations

It guarantees max USB bandwidth for up to one frame (1ms) without system
intervention

RM0090 Rev 17

RM0090

34.2.2

USB on-the-go full-speed (OTG_FS)

Host-mode features
The OTG_FS interface main features and requirements in host-mode are the following:
•

External charge pump for VBUS voltage generation.

•

Up to 8 host channels (pipes): each channel is dynamically reconfigurable to allocate
any type of USB transfer.

•

Built-in hardware scheduler holding:

•

34.2.3

–

Up to 8 interrupt plus isochronous transfer requests in the periodic hardware
queue

–

Up to 8 control plus bulk transfer requests in the non-periodic hardware queue

Management of a shared RX FIFO, a periodic TX FIFO and a nonperiodic TX FIFO for
efficient usage of the USB data RAM.

Peripheral-mode features
The OTG_FS interface main features in peripheral-mode are the following:
•

1 bidirectional control endpoint0

•

3 IN endpoints (EPs) configurable to support Bulk, Interrupt or Isochronous transfers

•

3 OUT endpoints configurable to support Bulk, Interrupt or Isochronous transfers

•

Management of a shared Rx FIFO and a Tx-OUT FIFO for efficient usage of the USB
data RAM

•

Management of up to 4 dedicated Tx-IN FIFOs (one for each active IN EP) to put less
load on the application

•

Support for the soft disconnect feature.

RM0090 Rev 17

1243/1747
1379

USB on-the-go full-speed (OTG_FS)

34.3

RM0090

OTG_FS functional description
Figure 386. OTG full-speed block diagram

&RUWH[ŠFRUH

27*B)6B'3
3RZHU
DQG
FORFN
FRQWUROOHU

86%VXVSHQG

870,)6

6\VWHPFORFNGRPDLQ

27*
)6
3+<

27*B)6B'0
27*B)6B,'
27*B)6B9%86

86%FORFN
GRPDLQ

5$0EXV

86%FORFNDW0+]

86%
27*)6
FRUH

8QLYHUVDOVHULDOEXV

27*B)6B62)
.E\WH
86%GDWD
),)2V
069

34.3.1

OTG pins
Table 196. OTG_FS input/output pins
Signal name

34.3.2

Signal type

Description

OTG_FS_DP

Digital input/output USB OTG D+ line

OTG_FS_DM

Digital input/output USB OTG D- line

OTG_FS_ID

Digital input

USB OTG ID

OTG_FS_VBUS

Analog input

USB OTG VBUS

OTG_FS_SOF

Digital output

USB OTG Start Of Frame (visibility)

OTG full-speed core
The USB OTG FS receives the 48 MHz ±0.25% clock from the reset and clock controller
(RCC), via an external quartz. The USB clock is used for driving the 48 MHz domain at fullspeed (12 Mbit/s) and must be enabled prior to configuring the OTG FS core.
The CPU reads and writes from/to the OTG FS core registers through the AHB peripheral
bus. It is informed of USB events through the single USB OTG interrupt line described in
Section 34.15: OTG_FS interrupts.

1244/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
The CPU submits data over the USB by writing 32-bit words to dedicated OTG_FS locations
(push registers). The data are then automatically stored into Tx-data FIFOs configured
within the USB data RAM. There is one Tx-FIFO push register for each in-endpoint
(peripheral mode) or out-channel (host mode).
The CPU receives the data from the USB by reading 32-bit words from dedicated OTG_FS
addresses (pop registers). The data are then automatically retrieved from a shared Rx-FIFO
configured within the 1.25 KB USB data RAM. There is one Rx-FIFO pop register for each
out-endpoint or in-channel.
The USB protocol layer is driven by the serial interface engine (SIE) and serialized over the
USB by the full-/low-speed transceiver module within the on-chip physical layer (PHY).

34.3.3

Full-speed OTG PHY
The embedded full-speed OTG PHY is controlled by the OTG FS core and conveys USB
control & data signals through the full-speed subset of the UTMI+ Bus (UTMIFS). It provides
the physical support to USB connectivity.
The full-speed OTG PHY includes the following components:

Caution:

•

FS/LS transceiver module used by both host and device. It directly drives transmission
and reception on the single-ended USB lines.

•

integrated ID pull-up resistor used to sample the ID line for A/B device identification.

•

DP/DM integrated pull-up and pull-down resistors controlled by the OTG_FS core
depending on the current role of the device. As a peripheral, it enables the DP pull-up
resistor to signal full-speed peripheral connections as soon as VBUS is sensed to be at
a valid level (B-session valid). In host mode, pull-down resistors are enabled on both
DP/DM. Pull-up and pull-down resistors are dynamically switched when the device’s
role is changed via the host negotiation protocol (HNP).

•

Pull-up/pull-down resistor ECN circuit. The DP pull-up consists of 2 resistors controlled
separately from the OTG_FS as per the resistor Engineering Change Notice applied to
USB Rev2.0. The dynamic trimming of the DP pull-up strength allows for better noise
rejection and Tx/Rx signal quality.

•

VBUS sensing comparators with hysteresis used to detect VBUS Valid, A-B Session
Valid and session-end voltage thresholds. They are used to drive the session request
protocol (SRP), detect valid startup and end-of-session conditions, and constantly
monitor the VBUS supply during USB operations.

•

VBUS pulsing method circuit used to charge/discharge VBUS through resistors during
the SRP (weak drive).

To guarantee a correct operation for the USB OTG FS peripheral, the AHB frequency should
be higher than 14.2 MHz.

RM0090 Rev 17

1245/1747
1379

USB on-the-go full-speed (OTG_FS)

34.4

RM0090

OTG dual role device (DRD)
Figure 387. OTG A-B device connection
6$$
 6 TO 6$$
VOLTAGE REGULATOR

6$$

'0)/
'0)/ )21

%.

/VERCURRENT

34-03342
#URRENT LIMITED
POWER DISTRIBUTION

SWITCH

 6 0WR

0!
0!
/3#?).
/3#?/54

0!
0!

6"53
$$0
)$
633

53"MICRO !" CONNECTOR

34- -#5

-36

1. External voltage regulator only needed when building a VBUS powered device
2. STMPS2141STR needed only if the application has to support a VBUS powered device. A basic power
switch can be used if 5 V are available on the application board.

34.4.1

ID line detection
The host or peripheral (the default) role is assumed depending on the ID input pin
(OTG_FS_ID). The ID line status is determined on plugging in the USB, depending on which
side of the USB cable is connected to the micro-AB receptacle.

34.4.2

•

If the B-side of the USB cable is connected with a floating ID wire, the integrated pullup resistor detects a high ID level and the default Peripheral role is confirmed. In this
configuration the OTG_FS complies with the standard FSM described by section 6.8.2:
On-The-Go B-device of the On-The-Go Specification Rev1.3 supplement to the
USB2.0.

•

If the A-side of the USB cable is connected with a grounded ID, the OTG_FS issues an
ID line status change interrupt (CIDSCHG bit in OTG_FS_GINTSTS) for host software
initialization, and automatically switches to the host role. In this configuration the
OTG_FS complies with the standard FSM described by section 6.8.1: On-The-Go Adevice of the On-The-Go Specification Rev1.3 supplement to the USB2.0.

HNP dual role device
The HNP capable bit in the Global USB configuration register (HNPCAP bit in OTG_FS_
GUSBCFG) enables the OTG_FS core to dynamically change its role from A-host to Aperipheral and vice-versa, or from B-Peripheral to B-host and vice-versa according to the
host negotiation protocol (HNP). The current device status can be read by the combined
values of the Connector ID Status bit in the Global OTG control and status register (CIDSTS
bit in OTG_FS_GOTGCTL) and the current mode of operation bit in the global interrupt and
status register (CMOD bit in OTG_FS_GINTSTS).
The HNP program model is described in detail in Section 34.17: OTG_FS programming
model.

1246/1747

RM0090 Rev 17

RM0090

34.4.3

USB on-the-go full-speed (OTG_FS)

SRP dual role device
The SRP capable bit in the global USB configuration register (SRPCAP bit in
OTG_FS_GUSBCFG) enables the OTG_FS core to switch off the generation of VBUS for
the A-device to save power. Note that the A-device is always in charge of driving VBUS
regardless of the host or peripheral role of the OTG_FS.
the SRP A/B-device program model is described in detail in Section 34.17: OTG_FS
programming model.

34.5

USB peripheral
This section gives the functional description of the OTG_FS in the USB peripheral mode.
The OTG_FS works as an USB peripheral in the following circumstances:
•

OTG B-Peripheral
–

•

OTG A-Peripheral
–

•

If the ID line is present, functional and connected to the B-side of the USB cable,
and the HNP-capable bit in the Global USB Configuration register (HNPCAP bit in
OTG_FS_GUSBCFG) is cleared (see On-The-Go Rev1.3 par. 6.8.3).

Peripheral only (see Figure 388: USB peripheral-only connection)
–

Note:

OTG A-device state after the HNP switches the OTG_FS to its peripheral role

B-device
–

•

OTG B-device default state if B-side of USB cable is plugged in

The force device mode bit in the Global USB configuration register (FDMOD in
OTG_FS_GUSBCFG) is set to 1, forcing the OTG_FS core to work as a USB
peripheral-only (see On-The-Go Rev1.3 par. 6.8.3). In this case, the ID line is
ignored even if present on the USB connector.

To build a bus-powered device implementation in case of the B-device or peripheral-only
configuration, an external regulator has to be added that generates the VDD chip-supply
from VBUS.
The VBUS pin can be freed by disabling the VBUS sensing option. This is done by setting the
NOVBUSSENS bit in the OTG_FS_GCCFG register. In this case the VBUS is considered
internally to be always at VBUS valid level (5 V).

RM0090 Rev 17

1247/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Figure 388. USB peripheral-only connection
6$$

6 TO 6$$
6OLATGE REGULATOR

6"53

0!
/3#?).

0!
0!

$$0
633

/3#?/54

53" 3TD " CONNECTOR

34- -#5

-36

1. Use a regulator to build a bus-powered device.

34.5.1

SRP-capable peripheral
The SRP capable bit in the Global USB configuration register (SRPCAP bit in
OTG_FS_GUSBCFG) enables the OTG_FS to support the session request protocol (SRP).
In this way, it allows the remote A-device to save power by switching off VBUS while the USB
session is suspended.
The SRP peripheral mode program model is described in detail in the B-device session
request protocol section.

34.5.2

Peripheral states
Powered state
The VBUS input detects the B-Session valid voltage by which the USB peripheral is allowed
to enter the powered state (see USB2.0 par9.1). The OTG_FS then automatically connects
the DP pull-up resistor to signal full-speed device connection to the host and generates the
session request interrupt (SRQINT bit in OTG_FS_GINTSTS) to notify the powered state.
The VBUS input also ensures that valid VBUS levels are supplied by the host during USB
operations. If a drop in VBUS below B-session valid happens to be detected (for instance
because of a power disturbance or if the host port has been switched off), the OTG_FS
automatically disconnects and the session end detected (SEDET bit in
OTG_FS_GOTGINT) interrupt is generated to notify that the OTG_FS has exited the
powered state.
In the powered state, the OTG_FS expects to receive some reset signaling from the host.
No other USB operation is possible. When a reset signaling is received the reset detected
interrupt (USBRST in OTG_FS_GINTSTS) is generated. When the reset signaling is
complete, the enumeration done interrupt (ENUMDNE bit in OTG_FS_GINTSTS) is
generated and the OTG_FS enters the Default state.

1248/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

Soft disconnect
The powered state can be exited by software with the soft disconnect feature. The DP pullup resistor is removed by setting the soft disconnect bit in the device control register (SDIS
bit in OTG_FS_DCTL), causing a device disconnect detection interrupt on the host side
even though the USB cable was not really removed from the host port.

Default state
In the Default state the OTG_FS expects to receive a SET_ADDRESS command from the
host. No other USB operation is possible. When a valid SET_ADDRESS command is
decoded on the USB, the application writes the corresponding number into the device
address field in the device configuration register (DAD bit in OTG_FS_DCFG). The
OTG_FS then enters the address state and is ready to answer host transactions at the
configured USB address.

Suspended state
The OTG_FS peripheral constantly monitors the USB activity. After counting 3 ms of USB
idleness, the early suspend interrupt (ESUSP bit in OTG_FS_GINTSTS) is issued, and
confirmed 3 ms later, if appropriate, by the suspend interrupt (USBSUSP bit in
OTG_FS_GINTSTS). The device suspend bit is then automatically set in the device status
register (SUSPSTS bit in OTG_FS_DSTS) and the OTG_FS enters the suspended state.
The suspended state may optionally be exited by the device itself. In this case the
application sets the remote wakeup signaling bit in the device control register (RWUSIG bit
in OTG_FS_DCTL) and clears it after 1 to 15 ms.
When a resume signaling is detected from the host, the resume interrupt (WKUPINT bit in
OTG_FS_GINTSTS) is generated and the device suspend bit is automatically cleared.

34.5.3

Peripheral endpoints
The OTG_FS core instantiates the following USB endpoints:
•

•

Control endpoint 0:
–

Bidirectional and handles control messages only

–

Separate set of registers to handle in and out transactions

–

Proper control (OTG_FS_DIEPCTL0/OTG_FS_DOEPCTL0), transfer
configuration (OTG_FS_DIEPTSIZ0/OTG_FS_DIEPTSIZ0), and status-interrupt
(OTG_FS_DIEPINTx/)OTG_FS_DOEPINT0) registers. The available set of bits
inside the control and transfer size registers slightly differs from that of other
endpoints

3 IN endpoints
–

Each of them can be configured to support the isochronous, bulk or interrupt
transfer type

–

Each of them has proper control (OTG_FS_DIEPCTLx), transfer configuration
(OTG_FS_DIEPTSIZx), and status-interrupt (OTG_FS_DIEPINTx) registers

–

The Device IN endpoints common interrupt mask register (OTG_FS_DIEPMSK) is
available to enable/disable a single kind of endpoint interrupt source on all of the
IN endpoints (EP0 included)

–

Support for incomplete isochronous IN transfer interrupt (IISOIXFR bit in
OTG_FS_GINTSTS), asserted when there is at least one isochronous IN endpoint

RM0090 Rev 17

1249/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

on which the transfer is not completed in the current frame. This interrupt is
asserted along with the end of periodic frame interrupt
(OTG_FS_GINTSTS/EOPF).
•

3 OUT endpoints
–

Each of them can be configured to support the isochronous, bulk or interrupt
transfer type

–

Each of them has a proper control (OTG_FS_DOEPCTLx), transfer configuration
(OTG_FS_DOEPTSIZx) and status-interrupt (OTG_FS_DOEPINTx) register

–

Device Out endpoints common interrupt mask register (OTG_FS_DOEPMSK) is
available to enable/disable a single kind of endpoint interrupt source on all of the
OUT endpoints (EP0 included)

–

Support for incomplete isochronous OUT transfer interrupt (INCOMPISOOUT bit
in OTG_FS_GINTSTS), asserted when there is at least one isochronous OUT
endpoint on which the transfer is not completed in the current frame. This interrupt
is asserted along with the end of periodic frame interrupt
(OTG_FS_GINTSTS/EOPF).

Endpoint control
•

The following endpoint controls are available to the application through the device
endpoint-x IN/OUT control register (DIEPCTLx/DOEPCTLx):
–

Endpoint enable/disable

–

Endpoint activate in current configuration

–

Program USB transfer type (isochronous, bulk, interrupt)

–

Program supported packet size

–

Program Tx-FIFO number associated with the IN endpoint

–

Program the expected or transmitted data0/data1 PID (bulk/interrupt only)

–

Program the even/odd frame during which the transaction is received or
transmitted (isochronous only)

–

Optionally program the NAK bit to always negative-acknowledge the host
regardless of the FIFO status

–

Optionally program the STALL bit to always stall host tokens to that endpoint

–

Optionally program the SNOOP mode for OUT endpoint not to check the CRC
field of received data

Endpoint transfer
The device endpoint-x transfer size registers (DIEPTSIZx/DOEPTSIZx) allow the application
to program the transfer size parameters and read the transfer status. Programming must be
done before setting the endpoint enable bit in the endpoint control register. Once the
endpoint is enabled, these fields are read-only as the OTG FS core updates them with the
current transfer status.
The following transfer parameters can be programmed:

1250/1747

•

Transfer size in bytes

•

Number of packets that constitute the overall transfer size

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

Endpoint status/interrupt
The device endpoint-x interrupt registers (DIEPINTx/DOPEPINTx) indicate the status of an
endpoint with respect to USB- and AHB-related events. The application must read these
registers when the OUT endpoint interrupt bit or the IN endpoint interrupt bit in the core
interrupt register (OEPINT bit in OTG_FS_GINTSTS or IEPINT bit in OTG_FS_GINTSTS,
respectively) is set. Before the application can read these registers, it must first read the
device all endpoints interrupt (OTG_FS_DAINT) register to get the exact endpoint number
for the device endpoint-x interrupt register. The application must clear the appropriate bit in
this register to clear the corresponding bits in the DAINT and GINTSTS registers
The peripheral core provides the following status checks and interrupt generation:

34.6

•

Transfer completed interrupt, indicating that data transfer was completed on both the
application (AHB) and USB sides

•

Setup stage has been done (control-out only)

•

Associated transmit FIFO is half or completely empty (in endpoints)

•

NAK acknowledge has been transmitted to the host (isochronous-in only)

•

IN token received when Tx-FIFO was empty (bulk-in/interrupt-in only)

•

Out token received when endpoint was not yet enabled

•

Babble error condition has been detected

•

Endpoint disable by application is effective

•

Endpoint NAK by application is effective (isochronous-in only)

•

More than 3 back-to-back setup packets were received (control-out only)

•

Timeout condition detected (control-in only)

•

Isochronous out packet has been dropped, without generating an interrupt

USB host
This section gives the functional description of the OTG_FS in the USB host mode. The
OTG_FS works as a USB host in the following circumstances:
•

OTG A-host
–

•

OTG B-host

•

A-device

–
–

•

OTG B-device after HNP switching to the host role
If the ID line is present, functional and connected to the A-side of the USB cable,
and the HNP-capable bit is cleared in the Global USB Configuration register
(HNPCAP bit in OTG_FS_GUSBCFG). Integrated pull-down resistors are
automatically set on the DP/DM lines.

Host only (see figure Figure 389: USB host-only connection).
–

Note:

OTG A-device default state when the A-side of the USB cable is plugged in

The force host mode bit in the global USB configuration register (FHMOD bit in
OTG_FS_GUSBCFG) forces the OTG_FS core to work as a USB host-only. In this
case, the ID line is ignored even if present on the USB connector. Integrated pulldown resistors are automatically set on the DP/DM lines.

On-chip 5 V VBUS generation is not supported. For this reason, a charge pump or, if 5 V are
available on the application board, a basic power switch must be added externally to drive

RM0090 Rev 17

1251/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

the 5 V VBUS line. The external charge pump can be driven by any GPIO output. This is
required for the OTG A-host, A-device and host-only configurations.
The VBUS input ensures that valid VBUS levels are supplied by the charge pump during USB
operations while the charge pump overcurrent output can be input to any GPIO pin
configured to generate port interrupts. The overcurrent ISR must promptly disable the VBUS
generation.
The VBUS pin can be freed by disabling the VBUS sensing option. This is done by setting the
NOVBUSSENS bit in the OTG_FS_GCCFG register. In this case the VBUS is considered
internally to be always at VBUS valid level (5 V).
Figure 389. USB host-only connection

9''
9

*3,2,54

(1

67036675
93ZU
&XUUHQWOLPLWHG
SRZHUGLVWULEXWLRQ
2YHUFXUUHQW

VZLWFK
9%86
'0

26&B,1

'3
966

26&B287

86%6WG$FRQQHFWRU

*3,2

06Y9

1. STMPS2141STR needed only if the application has to support a VBUS powered device. A basic power
switch can be used if 5 V are available on the application board.
2. VDD range is between 2 V and 3.6 V.

34.6.1

SRP-capable host
SRP support is available through the SRP capable bit in the global USB configuration
register (SRPCAP bit in OTG_FS_GUSBCFG). With the SRP feature enabled, the host can
save power by switching off the VBUS power while the USB session is suspended.
The SRP host mode program model is described in detail in the A-device session request
protocol section.

34.6.2

USB host states
Host port power
On-chip 5 V VBUS generation is not supported. For this reason, a charge pump or, if 5 V are
available on the application board, a basic power switch, must be added externally to drive
the 5 V VBUS line. The external charge pump can be driven by any GPIO output. When the
application decides to power on VBUS using the chosen GPIO, it must also set the port
power bit in the host port control and status register (PPWR bit in OTG_FS_HPRT).

1252/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

VBUS valid
When HNP or SRP is enabled the VBUS sensing pin (PA9) pin should be connected to
VBUS. The VBUS input ensures that valid VBUS levels are supplied by the charge pump
during USB operations. Any unforeseen VBUS voltage drop below the VBUS valid threshold
(4.25 V) leads to an OTG interrupt triggered by the session end detected bit (SEDET bit in
OTG_FS_GOTGINT). The application is then required to remove the VBUS power and clear
the port power bit.
When HNP and SRP are both disabled, the VBUS sensing pin (PA9) should not be
connected to VBUS. This pin can be can be used as GPIO.
The charge pump overcurrent flag can also be used to prevent electrical damage. Connect
the overcurrent flag output from the charge pump to any GPIO input and configure it to
generate a port interrupt on the active level. The overcurrent ISR must promptly disable the
VBUS generation and clear the port power bit.

Host detection of a peripheral connection
If SRP or HNP are enabled, even if USB peripherals or B-devices can be attached at any
time, the OTG_FS will not detect any bus connection until VBUS is no longer sensed at a
valid level (5 V). When VBUS is at a valid level and a remote B-device is attached, the
OTG_FS core issues a host port interrupt triggered by the device connected bit in the host
port control and status register (PCDET bit in OTG_FS_HPRT).
When HNP and SRP are both disabled, USB peripherals or B-device are detected as soon
as they are connected. The OTG_FS core issues a host port interrupt triggered by the
device connected bit in the host port control and status (PCDET bit in OTG_FS_HPRT).

Host detection of peripheral a disconnection
The peripheral disconnection event triggers the disconnect detected interrupt (DISCINT bit
in OTG_FS_GINTSTS).

Host enumeration
After detecting a peripheral connection the host must start the enumeration process by
sending USB reset and configuration commands to the new peripheral.
Before starting to drive a USB reset, the application waits for the OTG interrupt triggered by
the debounce done bit (DBCDNE bit in OTG_FS_GOTGINT), which indicates that the bus is
stable again after the electrical debounce caused by the attachment of a pull-up resistor on
DP (FS) or DM (LS).
The application drives a USB reset signaling (single-ended zero) over the USB by keeping
the port reset bit set in the host port control and status register (PRST bit in
OTG_FS_HPRT) for a minimum of 10 ms and a maximum of 20 ms. The application takes
care of the timing count and then of clearing the port reset bit.
Once the USB reset sequence has completed, the host port interrupt is triggered by the port
enable/disable change bit (PENCHNG bit in OTG_FS_HPRT). This informs the application
that the speed of the enumerated peripheral can be read from the port speed field in the
host port control and status register (PSPD bit in OTG_FS_HPRT) and that the host is
starting to drive SOFs (FS) or Keep alives (LS). The host is now ready to complete the
peripheral enumeration by sending peripheral configuration commands.

RM0090 Rev 17

1253/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Host suspend
The application decides to suspend the USB activity by setting the port suspend bit in the
host port control and status register (PSUSP bit in OTG_FS_HPRT). The OTG_FS core
stops sending SOFs and enters the suspended state.
The suspended state can be optionally exited on the remote device’s initiative (remote
wakeup). In this case the remote wakeup interrupt (WKUPINT bit in OTG_FS_GINTSTS) is
generated upon detection of a remote wakeup signaling, the port resume bit in the host port
control and status register (PRES bit in OTG_FS_HPRT) self-sets, and resume signaling is
automatically driven over the USB. The application must time the resume window and then
clear the port resume bit to exit the suspended state and restart the SOF.
If the suspended state is exited on the host initiative, the application must set the port
resume bit to start resume signaling on the host port, time the resume window and finally
clear the port resume bit.

34.6.3

Host channels
The OTG_FS core instantiates 8 host channels. Each host channel supports an USB host
transfer (USB pipe). The host is not able to support more than 8 transfer requests at the
same time. If more than 8 transfer requests are pending from the application, the host
controller driver (HCD) must re-allocate channels when they become available from
previous duty, that is, after receiving the transfer completed and channel halted interrupts.
Each host channel can be configured to support in/out and any type of periodic/nonperiodic
transaction. Each host channel makes us of proper control (HCCHARx), transfer
configuration (HCTSIZx) and status/interrupt (HCINTx) registers with associated mask
(HCINTMSKx) registers.

Host channel control
•

The following host channel controls are available to the application through the host
channel-x characteristics register (HCCHARx):
–

Channel enable/disable

–

Program the FS/LS speed of target USB peripheral

–

Program the address of target USB peripheral

–

Program the endpoint number of target USB peripheral

–

Program the transfer IN/OUT direction

–

Program the USB transfer type (control, bulk, interrupt, isochronous)

–

Program the maximum packet size (MPS)

–

Program the periodic transfer to be executed during odd/even frames

Host channel transfer
The host channel transfer size registers (HCTSIZx) allow the application to program the
transfer size parameters, and read the transfer status. Programming must be done before
setting the channel enable bit in the host channel characteristics register. Once the endpoint

1254/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
is enabled the packet count field is read-only as the OTG FS core updates it according to
the current transfer status.
•

The following transfer parameters can be programmed:
–

transfer size in bytes

–

number of packets making up the overall transfer size

–

initial data PID

Host channel status/interrupt
The host channel-x interrupt register (HCINTx) indicates the status of an endpoint with
respect to USB- and AHB-related events. The application must read these register when the
host channels interrupt bit in the core interrupt register (HCINT bit in OTG_FS_GINTSTS) is
set. Before the application can read these registers, it must first read the host all channels
interrupt (HCAINT) register to get the exact channel number for the host channel-x interrupt
register. The application must clear the appropriate bit in this register to clear the
corresponding bits in the HAINT and GINTSTS registers. The mask bits for each interrupt
source of each channel are also available in the OTG_FS_HCINTMSK-x register.
•

34.6.4

The host core provides the following status checks and interrupt generation:
–

Transfer completed interrupt, indicating that the data transfer is complete on both
the application (AHB) and USB sides

–

Channel has stopped due to transfer completed, USB transaction error or disable
command from the application

–

Associated transmit FIFO is half or completely empty (IN endpoints)

–

ACK response received

–

NAK response received

–

STALL response received

–

USB transaction error due to CRC failure, timeout, bit stuff error, false EOP

–

Babble error

–

fraMe overrun

–

dAta toggle error

Host scheduler
The host core features a built-in hardware scheduler which is able to autonomously re-order
and manage the USB transaction requests posted by the application. At the beginning of
each frame the host executes the periodic (isochronous and interrupt) transactions first,
followed by the nonperiodic (control and bulk) transactions to achieve the higher level of
priority granted to the isochronous and interrupt transfer types by the USB specification.
The host processes the USB transactions through request queues (one for periodic and one
for nonperiodic). Each request queue can hold up to 8 entries. Each entry represents a
pending transaction request from the application, and holds the IN or OUT channel number
along with other information to perform a transaction on the USB. The order in which the
requests are written to the queue determines the sequence of the transactions on the USB
interface.
At the beginning of each frame, the host processes the periodic request queue first, followed
by the nonperiodic request queue. The host issues an incomplete periodic transfer interrupt
(IPXFR bit in OTG_FS_GINTSTS) if an isochronous or interrupt transaction scheduled for
the current frame is still pending at the end of the current frame. The OTG HS core is fully

RM0090 Rev 17

1255/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

responsible for the management of the periodic and nonperiodic request queues.The
periodic transmit FIFO and queue status register (HPTXSTS) and nonperiodic transmit
FIFO and queue status register (HNPTXSTS) are read-only registers which can be used by
the application to read the status of each request queue. They contain:
•

The number of free entries currently available in the periodic (nonperiodic) request
queue (8 max)

•

Free space currently available in the periodic (nonperiodic) Tx-FIFO (out-transactions)

•

IN/OUT token, host channel number and other status information.

As request queues can hold a maximum of 8 entries each, the application can push to
schedule host transactions in advance with respect to the moment they physically reach the
SB for a maximum of 8 pending periodic transactions plus 8 pending nonperiodic
transactions.
To post a transaction request to the host scheduler (queue) the application must check that
there is at least 1 entry available in the periodic (nonperiodic) request queue by reading the
PTXQSAV bits in the OTG_FS_HNPTXSTS register or NPTQXSAV bits in the
OTG_FS_HNPTXSTS register.

34.7

SOF trigger
Figure 390. SOF connectivity

34- -#5

3/& PULSE OUTPUT TO
EXTERNAL AUDIO CONTROL

0!
)42

4)-

3/&
PULSE

3/&GEN

6"53

0!

$

0!

$

0!

)$
633

53" MICRO !" CONNECTOR

0!

-36

The OTG FS core provides means to monitor, track and configure SOF framing in the host
and peripheral, as well as an SOF pulse output connectivity feature.
Such utilities are especially useful for adaptive audio clock generation techniques, where
the audio peripheral needs to synchronize to the isochronous stream provided by the PC, or
the host needs to trim its framing rate according to the requirements of the audio peripheral.

34.7.1

Host SOFs
In host mode the number of PHY clocks occurring between the generation of two
consecutive SOF (FS) or Keep-alive (LS) tokens is programmable in the host frame interval
register (HFIR), thus providing application control over the SOF framing period. An interrupt

1256/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
is generated at any start of frame (SOF bit in OTH_FS_GINTSTS). The current frame
number and the time remaining until the next SOF are tracked in the host frame number
register (HFNUM).
An SOF pulse signal, generated at any SOF starting token and with a width of 20 HCLK
cycles, can be made available externally on the OTG_FS_SOF pin using the SOFOUTEN
bit in the global control and configuration register. The SOF pulse is also internally
connected to the input trigger of timer 2 (TIM2), so that the input capture feature, the output
compare feature and the timer can be triggered by the SOF pulse. The TIM2 connection is
enabled through the ITR1_RMP bits of TIM2_OR register.

34.7.2

Peripheral SOFs
In device mode, the start of frame interrupt is generated each time an SOF token is received
on the USB (SOF bit in OTH_FS_GINTSTS). The corresponding frame number can be read
from the device status register (FNSOF bit in OTG_FS_DSTS). An SOF pulse signal with a
width of 20 HCLK cycles is also generated and can be made available externally on the
OTG_FS_SOF pin by using the SOF output enable bit in the global control and configuration
register (SOFOUTEN bit in OTG_FS_GCCFG). The SOF pulse signal is also internally
connected to the TIM2 input trigger, so that the input capture feature, the output compare
feature and the timer can be triggered by the SOF pulse. The TIM2 connection is enabled
through the ITR1_RMP bits of the TIM2 option register (TIM2_OR).
The end of periodic frame interrupt (GINTSTS/EOPF) is used to notify the application when
80%, 85%, 90% or 95% of the time frame interval elapsed depending on the periodic frame
interval field in the device configuration register (PFIVL bit in OTG_FS_DCFG). This feature
can be used to determine if all of the isochronous traffic for that frame is complete.

34.8

OTG low-power modes
Table 197 below defines the STM32 low power modes and their compatibility with the OTG.
Table 197. Compatibility of STM32 low power modes with the OTG

Mode

Description

USB compatibility

Run

MCU fully active

Required when USB not in
suspend state.

Sleep

USB suspend exit causes the device to exit Sleep mode. Peripheral
registers content is kept.

Available while USB is in
suspend state.

Stop

USB suspend exit causes the device to exit Stop mode. Peripheral
registers content is kept(1).

Available while USB is in
suspend state.

Standby

Powered-down. The peripheral must be reinitialized after exiting
Standby mode.

Not compatible with USB
applications.

1. Within Stop mode there are different possible settings. Some restrictions may also exist, please refer to Section 5: Power
controller (PWR) to understand which (if any) restrictions apply when using OTG.

RM0090 Rev 17

1257/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

The power consumption of the OTG PHY is controlled by three bits in the general core
configuration register:
•

PHY power down (GCCFG/PWRDWN)
It switches on/off the full-speed transceiver module of the PHY. It must be preliminarily
set to allow any USB operation.

•

A-VBUS sensing enable (GCCFG/VBUSASEN)
It switches on/off the VBUS comparators associated with A-device operations. It must
be set when in A-device (USB host) mode and during HNP.

•

B-VBUS sensing enable (GCCFG/VBUSASEN)
It switches on/off the VBUS comparators associated with B-device operations. It must
be set when in B-device (USB peripheral) mode and during HNP.

Power reduction techniques are available while in the USB suspended state, when the USB
session is not yet valid or the device is disconnected.
•

Stop PHY clock (STPPCLK bit in OTG_FS_PCGCCTL)
When setting the stop PHY clock bit in the clock gating control register, most of the
48 MHz clock domain internal to the OTG full-speed core is switched off by clock
gating. The dynamic power consumption due to the USB clock switching activity is cut
even if the 48 MHz clock input is kept running by the application
Most of the transceiver is also disabled, and only the part in charge of detecting the
asynchronous resume or remote wakeup event is kept alive.

•

Gate HCLK (GATEHCLK bit in OTG_FS_PCGCCTL)
When setting the Gate HCLK bit in the clock gating control register, most of the system
clock domain internal to the OTG_FS core is switched off by clock gating. Only the
register read and write interface is kept alive. The dynamic power consumption due to
the USB clock switching activity is cut even if the system clock is kept running by the
application for other purposes.

•

USB system stop
When the OTG_FS is in the USB suspended state, the application may decide to
drastically reduce the overall power consumption by a complete shut down of all the
clock sources in the system. USB System Stop is activated by first setting the Stop
PHY clock bit and then configuring the system deep sleep mode in the power control
system module (PWR).
The OTG_FS core automatically reactivates both system and USB clocks by
asynchronous detection of remote wakeup (as an host) or resume (as a device)
signaling on the USB.

To save dynamic power, the USB data FIFO is clocked only when accessed by the OTG_FS
core.

34.9

Dynamic update of the OTG_FS_HFIR register
The USB core embeds a dynamic trimming capability of SOF framing period in host mode
allowing to synchronize an external device with the SOF frames.
When the OTG_FS_HFIR register is changed within a current SOF frame, the SOF period
correction is applied in the next frame as described in Figure 391.

1258/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
Figure 391. Updating OTG_FS_HFIR dynamically
/LD /4'?&3?()&2 VALUE
  PERIODS

/4'?&3?()&2 VALUE
  PERIODS ()&2 WRITE LATENCY

.EW /4'?&3?()&2 VALUE
  PERIODS

3/&
RELOAD
,ATENCY

/4'?&3?(&)2
WRITE

x

x






x






x











&RAME
TIMER








/4'?&3?(&)2
VALUE

AI

34.10

USB data FIFOs
The USB system features 1.25 Kbyte of dedicated RAM with a sophisticated FIFO control
mechanism. The packet FIFO controller module in the OTG_FS core organizes RAM space
into Tx-FIFOs into which the application pushes the data to be temporarily stored before the
USB transmission, and into a single Rx FIFO where the data received from the USB are
temporarily stored before retrieval (popped) by the application. The number of instructed
FIFOs and how these are organized inside the RAM depends on the device’s role. In
peripheral mode an additional Tx-FIFO is instructed for each active IN endpoint. Any FIFO
size is software configured to better meet the application requirements.

RM0090 Rev 17

1259/1747
1379

USB on-the-go full-speed (OTG_FS)

34.11

RM0090

Peripheral FIFO architecture
Figure 392. Device-mode FIFO address mapping and AHB FIFO access mapping
Single data
FIFO
IN endpoint Tx FIFO #n
DFIFO push access
from AHB

Dedicated Tx
FIFO #n control
(optional)

Tx FIFO #n
packet

..
.

..
.

MAC pop

DIEPTXF2[31:16]
DIEPTXFx[15:0]

..
.
DIEPTXF2[15:0]

IN endpoint Tx FIFO #1
DFIFO push access
from AHB

Dedicated Tx
FIFO #1 control
(optional)

Tx FIFO #1 packet
DIEPTXF1[31:16]
DIEPTXF1[15:0]

MAC pop
IN endpoint Tx FIFO #0
DFIFO push access
from AHB

Dedicated Tx
FIFO #0 control
(optional)

Tx FIFO #0 packet

MAC pop

Any OUT endpoint DFIFO pop
access from AHB

GNPTXFSIZ[31:16]

GNPTXFSIZ[15:0]

Rx FIFO control

Rx packets

GRXFSIZ[31:16]
(Rx start
A1 = 0 address
fixed to 0)

MAC push

ai15611

34.11.1

Peripheral Rx FIFO
The OTG peripheral uses a single receive FIFO that receives the data directed to all OUT
endpoints. Received packets are stacked back-to-back until free space is available in the
Rx-FIFO. The status of the received packet (which contains the OUT endpoint destination
number, the byte count, the data PID and the validity of the received data) is also stored by
the core on top of the data payload. When no more space is available, host transactions are
NACKed and an interrupt is received on the addressed endpoint. The size of the receive
FIFO is configured in the receive FIFO Size register (GRXFSIZ).
The single receive FIFO architecture makes it more efficient for the USB peripheral to fill in
the receive RAM buffer:
•

All OUT endpoints share the same RAM buffer (shared FIFO)

•

The OTG FS core can fill in the receive FIFO up to the limit for any host sequence of
OUT tokens

The application keeps receiving the Rx-FIFO non-empty interrupt (RXFLVL bit in
OTG_FS_GINTSTS) as long as there is at least one packet available for download. It reads
the packet information from the receive status read and pop register (GRXSTSP) and finally
pops data off the receive FIFO by reading from the endpoint-related pop address.

1260/1747

RM0090 Rev 17

RM0090

34.11.2

USB on-the-go full-speed (OTG_FS)

Peripheral Tx FIFOs
The core has a dedicated FIFO for each IN endpoint. The application configures FIFO sizes
by writing the non periodic transmit FIFO size register (OTG_FS_TX0FSIZ) for IN endpoint0
and the device IN endpoint transmit FIFOx registers (DIEPTXFx) for IN endpoint-x.

34.12

Host FIFO architecture
Figure 393. Host-mode FIFO address mapping and AHB FIFO access mapping
Single data
FIFO

Any periodic channel
DFIFO push access
from AHB

Periodic Tx
FIFO control
(optional)

Periodic Tx packets

MAC pop

HPTXFSIZ[15:0]

Periodic Tx packets

Any non-periodic
channel DFIFO push
access from AHB

NPTXFSIZ[31:16]

Non-periodic
Tx FIFO control
NPTXFSIZ[15:0]

MAC pop
Rx packets
Any channel DFIFO pop
access from AHB

HPTXFSIZ[31:16]

RXFSIZ[31:16]

Rx FIFO control
Rx start address
fixed to 0
A1 = 0

MAC push

ai15610

34.12.1

Host Rx FIFO
The host uses one receiver FIFO for all periodic and nonperiodic transactions. The FIFO is
used as a receive buffer to hold the received data (payload of the received packet) from the
USB until it is transferred to the system memory. Packets received from any remote IN
endpoint are stacked back-to-back until free space is available. The status of each received
packet with the host channel destination, byte count, data PID and validity of the received
data are also stored into the FIFO. The size of the receive FIFO is configured in the receive
FIFO size register (GRXFSIZ).
The single receive FIFO architecture makes it highly efficient for the USB host to fill in the
receive data buffer:
•

All IN configured host channels share the same RAM buffer (shared FIFO)

•

The OTG FS core can fill in the receive FIFO up to the limit for any sequence of IN
tokens driven by the host software

The application receives the Rx FIFO not-empty interrupt as long as there is at least one
packet available for download. It reads the packet information from the receive status read
and pop register and finally pops the data off the receive FIFO.

RM0090 Rev 17

1261/1747
1379

USB on-the-go full-speed (OTG_FS)

34.12.2

RM0090

Host Tx FIFOs
The host uses one transmit FIFO for all non-periodic (control and bulk) OUT transactions
and one transmit FIFO for all periodic (isochronous and interrupt) OUT transactions. FIFOs
are used as transmit buffers to hold the data (payload of the transmit packet) to be
transmitted over the USB. The size of the periodic (nonperiodic) Tx FIFO is configured in the
host periodic (nonperiodic) transmit FIFO size (HPTXFSIZ/HNPTXFSIZ) register.
The two Tx FIFO implementation derives from the higher priority granted to the periodic type
of traffic over the USB frame. At the beginning of each frame, the built-in host scheduler
processes the periodic request queue first, followed by the nonperiodic request queue.
The two transmit FIFO architecture provides the USB host with separate optimization for
periodic and nonperiodic transmit data buffer management:
•

All host channels configured to support periodic (nonperiodic) transactions in the OUT
direction share the same RAM buffer (shared FIFOs)

•

The OTG FS core can fill in the periodic (nonperiodic) transmit FIFO up to the limit for
any sequence of OUT tokens driven by the host software

The OTG_FS core issues the periodic Tx FIFO empty interrupt (PTXFE bit in
OTG_FS_GINTSTS) as long as the periodic Tx-FIFO is half or completely empty,
depending on the value of the periodic Tx-FIFO empty level bit in the AHB configuration
register (PTXFELVL bit in OTG_FS_GAHBCFG). The application can push the transmission
data in advance as long as free space is available in both the periodic Tx FIFO and the
periodic request queue. The host periodic transmit FIFO and queue status register
(HPTXSTS) can be read to know how much space is available in both.
OTG_FS core issues the non periodic Tx FIFO empty interrupt (NPTXFE bit in
OTG_FS_GINTSTS) as long as the nonperiodic Tx FIFO is half or completely empty
depending on the non periodic Tx FIFO empty level bit in the AHB configuration register
(TXFELVL bit in OTG_FS_GAHBCFG). The application can push the transmission data as
long as free space is available in both the nonperiodic Tx FIFO and nonperiodic request
queue. The host nonperiodic transmit FIFO and queue status register (HNPTXSTS) can be
read to know how much space is available in both.

34.13

FIFO RAM allocation

34.13.1

Device mode
Receive FIFO RAM allocation: the application should allocate RAM for SETUP Packets:
10 locations must be reserved in the receive FIFO to receive SETUP packets on control
endpoint. The core does not use these locations, which are reserved for SETUP packets, to
write any other data. One location is to be allocated for Global OUT NAK. Status information
is written to the FIFO along with each received packet. Therefore, a minimum space of
(Largest Packet Size / 4) + 1 must be allocated to receive packets. If multiple isochronous
endpoints are enabled, then at least two (Largest Packet Size / 4) + 1 spaces must be
allocated to receive back-to-back packets. Typically, two (Largest Packet Size / 4) + 1
spaces are recommended so that when the previous packet is being transferred to the CPU,
the USB can receive the subsequent packet.
Along with the last packet for each endpoint, transfer complete status information is also
pushed to the FIFO. Typically, one location for each OUT endpoint is recommended.

1262/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
Transmit FIFO RAM allocation: the minimum RAM space required for each IN Endpoint
Transmit FIFO is the maximum packet size for that particular IN endpoint.

Note:

More space allocated in the transmit IN Endpoint FIFO results in better performance on the
USB.

34.13.2

Host mode
Receive FIFO RAM allocation
Status information is written to the FIFO along with each received packet. Therefore, a
minimum space of (Largest Packet Size / 4) + 1 must be allocated to receive packets. If
multiple isochronous channels are enabled, then at least two (Largest Packet Size / 4) + 1
spaces must be allocated to receive back-to-back packets. Typically, two (Largest Packet
Size / 4) + 1 spaces are recommended so that when the previous packet is being
transferred to the CPU, the USB can receive the subsequent packet.
Along with the last packet in the host channel, transfer complete status information is also
pushed to the FIFO. So one location must be allocated for this.

Transmit FIFO RAM allocation
The minimum amount of RAM required for the host Non-periodic Transmit FIFO is the
largest maximum packet size among all supported non-periodic OUT channels.
Typically, two Largest Packet Sizes worth of space is recommended, so that when the
current packet is under transfer to the USB, the CPU can get the next packet.
The minimum amount of RAM required for host periodic Transmit FIFO is the largest
maximum packet size out of all the supported periodic OUT channels. If there is at least one
Isochronous OUT endpoint, then the space must be at least two times the maximum packet
size of that channel.
Note:

More space allocated in the Transmit Non-periodic FIFO results in better performance on
the USB.

34.14

USB system performance
Best USB and system performance is achieved owing to the large RAM buffers, the highly
configurable FIFO sizes, the quick 32-bit FIFO access through AHB push/pop registers and,
especially, the advanced FIFO control mechanism. Indeed, this mechanism allows the
OTG_FS to fill in the available RAM space at best regardless of the current USB sequence.
With these features:
•

•

The application gains good margins to calibrate its intervention in order to optimize the
CPU bandwidth usage:
–

It can accumulate large amounts of transmission data in advance compared to
when they are effectively sent over the USB

–

It benefits of a large time margin to download data from the single receive FIFO

The USB Core is able to maintain its full operating rate, that is to provide maximum fullspeed bandwidth with a great margin of autonomy versus application intervention:
–

It has a large reserve of transmission data at its disposal to autonomously manage
the sending of data over the USB

RM0090 Rev 17

1263/1747
1379

USB on-the-go full-speed (OTG_FS)
–

RM0090

It has a lot of empty space available in the receive buffer to autonomously fill it in
with the data coming from the USB

As the OTG_FS core is able to fill in the 1.25 Kbyte RAM buffer very efficiently, and as
1.25 Kbyte of transmit/receive data is more than enough to cover a full speed frame, the
USB system is able to withstand the maximum full-speed data rate for up to one USB frame
(1 ms) without any CPU intervention.

34.15

OTG_FS interrupts
When the OTG_FS controller is operating in one mode, either device or host, the application
must not access registers from the other mode. If an illegal access occurs, a mode
mismatch interrupt is generated and reflected in the Core interrupt register (MMIS bit in the
OTG_FS_GINTSTS register). When the core switches from one mode to the other, the
registers in the new mode of operation must be reprogrammed as they would be after a
power-on reset.
Figure 394 shows the interrupt hierarchy.

1264/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
Figure 394. Interrupt hierarchy
:DNHXSLQWHUUXSW
27*B)6B:.83



$1'

*OREDOLQWHUUXSW
27*B)6

25

$1'

27

2(

*,
17

3,1
,(3 3
,1
7

+&
,1
7
+3
57
,1
7

*OREDOLQWHUUXSWPDVN ELW
27*B$+%&)*
$+%FRQILJXUDWLRQUHJLVWHU

27*B*,17676
&RUHUHJLVWHULQWHUUXSW


 



 







27*B*,1706.
&RUHLQWHUUXSWPDVNUHJLVWHU

27*B*27*,17
27*LQWHUUXSWUHJLVWHU

(3 
287HQGSRLQWV

(3 
,1HQGSRLQWV

27*B'$,1706.
'HYLFHDOOHQGSRLQWVLQWHUUXSWPDVN
UHJLVWHU
27*B'$,17
'HYLFHDOOHQGSRLQWVLQWHUUXSWUHJLVWHU

27*B',(306.
27*B'2(306.
'HYLFH,1287HQGSRLQWVFRPPRQ
LQWHUUXSWPDVNUHJLVWHU

[ 



27*B',(3,17[
27*B'2(3,17[
'HYLFH,1287HQGSRLQWLQWHUUXSW
UHJLVWHUV

[ +&

27*B+357
+RVWSRUWFRQWURODQGVWDWXVUHJLVWHU

27*B+$,1706.
+RVWDOOFKDQQHOVLQWHUUXSWPDVNUHJLVWHU
27*B+$,17
+RVWDOOFKDQQHOVLQWHUUXSWUHJLVWHU
[ 



27*B+&7,1706.[
+RVWFKDQQHOVLQWHUUXSWPDVNUHJLVWHUV

[ +&

27*B+&7,17[
+RVWFKDQQHOVLQWHUUXSWUHJLVWHUV

06Y9

1. OTG_FS_WKUP become active (high state) when resume condition occurs during L1 SLEEP or L2 SUSPEND states.

RM0090 Rev 17

1265/1747
1379

USB on-the-go full-speed (OTG_FS)

34.16

RM0090

OTG_FS control and status registers
By reading from and writing to the control and status registers (CSRs) through the AHB
slave interface, the application controls the OTG_FS controller. These registers are 32 bits
wide, and the addresses are 32-bit block aligned. The OTG_FS registers must be accessed
by words (32 bits).
CSRs are classified as follows:
•

Core global registers

•

Host-mode registers

•

Host global registers

•

Host port CSRs

•

Host channel-specific registers

•

Device-mode registers

•

Device global registers

•

Device endpoint-specific registers

•

Power and clock-gating registers

•

Data FIFO (DFIFO) access registers

Only the Core global, Power and clock-gating, Data FIFO access, and host port control and
status registers can be accessed in both host and device modes. When the OTG_FS
controller is operating in one mode, either device or host, the application must not access
registers from the other mode. If an illegal access occurs, a mode mismatch interrupt is
generated and reflected in the Core interrupt register (MMIS bit in the OTG_FS_GINTSTS
register). When the core switches from one mode to the other, the registers in the new mode
of operation must be reprogrammed as they would be after a power-on reset.

1266/1747

RM0090 Rev 17

RM0090

34.16.1

USB on-the-go full-speed (OTG_FS)

CSR memory map
The host and device mode registers occupy different addresses. All registers are
implemented in the AHB clock domain.
Figure 395. CSR memory map
0000h
Core global CSRs (1 Kbyte)
0400h
Host mode CSRs (1 Kbyte)
0800h
Device mode CSRs (1.5 Kbyte)
0E00h
Power and clock gating CSRs (0.5 Kbyte)
1000h
Device EP 0/Host channel 0 FIFO (4 Kbyte)
2000h
Device EP1/Host channel 1 FIFO (4 Kbyte)
DFIFO
push/pop
to this region

3000h

Device EP (x – 1)(1)/Host channel (x – 1)(1) FIFO (4 Kbyte)
Device EP x(1)/Host channel x(1) FIFO (4 Kbyte)

Reserved

2 0000h
Direct access to data FIFO RAM
for debugging (128 Kbyte)

DFIFO
debug read/
write to this
region

3 FFFFh
ai15615b

1. x = 3 in device mode and x = 7 in host mode.

Global CSR map
These registers are available in both host and device modes.
Table 198. Core global control and status registers (CSRs)
Acronym

Address
offset

Register name

OTG_FS_GOTGCTL

0x000

OTG_FS control and status register (OTG_FS_GOTGCTL) on page 1272

OTG_FS_GOTGINT

0x004

OTG_FS interrupt register (OTG_FS_GOTGINT) on page 1273

OTG_FS_GAHBCFG

0x008

OTG_FS AHB configuration register (OTG_FS_GAHBCFG) on page 1275

OTG_FS_GUSBCFG

0x00C

OTG_FS USB configuration register (OTG_FS_GUSBCFG) on page 1276

OTG_FS_GRSTCTL

0x010

OTG_FS reset register (OTG_FS_GRSTCTL) on page 1278

RM0090 Rev 17

1267/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Table 198. Core global control and status registers (CSRs) (continued)
Address
offset

Acronym

Register name

OTG_FS_GINTSTS

0x014

OTG_FS core interrupt register (OTG_FS_GINTSTS) on page 1280

OTG_FS_GINTMSK

0x018

OTG_FS interrupt mask register (OTG_FS_GINTMSK) on page 1284

OTG_FS_GRXSTSR

0x01C

OTG_FS_GRXSTSP

0x020

OTG_FS Receive status debug read/OTG status read and pop registers
(OTG_FS_GRXSTSR/OTG_FS_GRXSTSP) on page 1287

OTG_FS_GRXFSIZ

0x024

OTG_FS Receive FIFO size register (OTG_FS_GRXFSIZ) on page 1288

OTG_FS_HNPTXFSIZ/
OTG_FS_DIEPTXF0(1)

0x028

OTG_FS Host non-periodic transmit FIFO size register
(OTG_FS_HNPTXFSIZ)/Endpoint 0 Transmit FIFO size
(OTG_FS_DIEPTXF0)

OTG_FS_HNPTXSTS

0x02C

OTG_FS non-periodic transmit FIFO/queue status register
(OTG_FS_HNPTXSTS) on page 1289

OTG_FS_GCCFG

0x038

OTG_FS general core configuration register (OTG_FS_GCCFG) on
page 1290

OTG_FS_CID

0x03C

OTG_FS core ID register (OTG_FS_CID) on page 1291

OTG_FS_HPTXFSIZ

0x100

OTG_FS Host periodic transmit FIFO size register (OTG_FS_HPTXFSIZ) on
page 1292

OTG_FS_DIEPTXFx

0x104
0x108
0x10C

OTG_FS device IN endpoint transmit FIFO size register
(OTG_FS_DIEPTXFx) (x = 1..3, where x is the FIFO_number) on page 1292

1. The general rule is to use OTG_FS_HNPTXFSIZ for host mode and OTG_FS_DIEPTXF0 for device mode.

Host-mode CSR map
These registers must be programmed every time the core changes to host mode.
Table 199. Host-mode control and status registers (CSRs)
Acronym

Offset
address

Register name

OTG_FS_HCFG

0x400

OTG_FS Host configuration register (OTG_FS_HCFG) on page 1293

OTG_FS_HFIR

0x404

OTG_FS Host frame interval register (OTG_FS_HFIR) on page 1293

OTG_FS_HFNUM

0x408

OTG_FS Host frame number/frame time remaining register
(OTG_FS_HFNUM) on page 1294

OTG_FS_HPTXSTS

0x410

OTG_FS_Host periodic transmit FIFO/queue status register
(OTG_FS_HPTXSTS) on page 1294

OTG_FS_HAINT

0x414

OTG_FS Host all channels interrupt register (OTG_FS_HAINT) on
page 1295

OTG_FS_HAINTMSK

0x418

OTG_FS Host all channels interrupt mask register (OTG_FS_HAINTMSK)
on page 1296

1268/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
Table 199. Host-mode control and status registers (CSRs) (continued)
Offset
address

Acronym

Register name

OTG_FS_HPRT

0x440

OTG_FS Host port control and status register (OTG_FS_HPRT) on
page 1296

OTG_FS_HCCHARx

0x500
0x520
...
0x5E0

OTG_FS Host channel-x characteristics register (OTG_FS_HCCHARx)
(x = 0..7, where x = Channel_number) on page 1299

OTG_FS_HCINTx

0x508

OTG_FS Host channel-x interrupt register (OTG_FS_HCINTx) (x = 0..7,
where x = Channel_number) on page 1300

OTG_FS_HCINTMSKx

0x50C

OTG_FS Host channel-x interrupt mask register (OTG_FS_HCINTMSKx)
(x = 0..7, where x = Channel_number) on page 1301

OTG_FS_HCTSIZx

0x510

OTG_FS Host channel-x transfer size register (OTG_FS_HCTSIZx)
(x = 0..7, where x = Channel_number) on page 1302

Device-mode CSR map
These registers must be programmed every time the core changes to device mode.
Table 200. Device-mode control and status registers
Acronym

Offset
address

Register name

OTG_FS_DCFG

0x800

OTG_FS device configuration register (OTG_FS_DCFG) on page 1303

OTG_FS_DCTL

0x804

OTG_FS device control register (OTG_FS_DCTL) on page 1304

OTG_FS_DSTS

0x808

OTG_FS device status register (OTG_FS_DSTS) on page 1305

OTG_FS_DIEPMSK

0x810

OTG_FS device IN endpoint common interrupt mask register
(OTG_FS_DIEPMSK) on page 1306

OTG_FS_DOEPMSK

0x814

OTG_FS device OUT endpoint common interrupt mask register
(OTG_FS_DOEPMSK) on page 1307

OTG_FS_DAINT

0x818

OTG_FS device all endpoints interrupt register (OTG_FS_DAINT) on
page 1308

OTG_FS_DAINTMSK

0x81C

OTG_FS all endpoints interrupt mask register (OTG_FS_DAINTMSK) on
page 1309

OTG_FS_DVBUSDIS

0x828

OTG_FS device VBUS discharge time register (OTG_FS_DVBUSDIS) on
page 1309

OTG_FS_DVBUSPULSE

0x82C

OTG_FS device VBUS pulsing time register (OTG_FS_DVBUSPULSE)
on page 1309

OTG_FS_DIEPEMPMSK

0x834

OTG_FS device IN endpoint FIFO empty interrupt mask register:
(OTG_FS_DIEPEMPMSK) on page 1310

OTG_FS_DIEPCTL0

0x900

OTG_FS device control IN endpoint 0 control register
(OTG_FS_DIEPCTL0) on page 1310

RM0090 Rev 17

1269/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Table 200. Device-mode control and status registers (continued)
Acronym

Offset
address

Register name

OTG_FS_DIEPCTLx

0x920
0x940
0x960

OTG device endpoint x control register (OTG_FS_DIEPCTLx) (x = 1..3,
where x = Endpoint_number) on page 1312

OTG_FS_DIEPINTx

0x908

OTG_FS device endpoint-x interrupt register (OTG_FS_DIEPINTx)
(x = 0..3, where x = Endpoint_number) on page 1319

OTG_FS_DIEPTSIZ0

0x910

OTG_FS device IN endpoint 0 transfer size register
(OTG_FS_DIEPTSIZ0) on page 1321

OTG_FS_DTXFSTSx

0x918

OTG_FS device IN endpoint transmit FIFO status register
(OTG_FS_DTXFSTSx) (x = 0..3, where x = Endpoint_number) on
page 1325

OTG_FS_DIEPTSIZx

0x930
0x950
0x970

OTG_FS device OUT endpoint-x transfer size register
(OTG_FS_DOEPTSIZx) (x = 1..3, where x = Endpoint_number) on
page 1325

OTG_FS_DOEPCTL0

0xB00

OTG_FS device control OUT endpoint 0 control register
(OTG_FS_DOEPCTL0) on page 1315

OTG_FS_DOEPCTLx

0xB20
0xB40
0xB60

OTG device endpoint x control register (OTG_FS_DIEPCTLx) (x = 1..3,
where x = Endpoint_number) on page 1312

OTG_FS_DOEPINTx

0xB08

OTG_FS device endpoint-x interrupt register (OTG_FS_DOEPINTx)
(x = 0..3, where x = Endpoint_number) on page 1320

OTG_FS_DOEPTSIZ0

0xB10

OTG_FS device OUT endpoint 0 transfer size register
(OTG_FS_DOEPTSIZ0) on page 1323

OTG_FS_DOEPTSIZx

0xB30
0xB50
0xB70

OTG_FS device OUT endpoint-x transfer size register
(OTG_FS_DOEPTSIZx) (x = 1..3, where x = Endpoint_number) on
page 1325

Data FIFO (DFIFO) access register map
These registers, available in both host and device modes, are used to read or write the FIFO
space for a specific endpoint or a channel, in a given direction. If a host channel is of type
IN, the FIFO can only be read on the channel. Similarly, if a host channel is of type OUT, the
FIFO can only be written on the channel.
Table 201. Data FIFO (DFIFO) access register map
FIFO access register section

1270/1747

Address range

Access

Device IN Endpoint 0/Host OUT Channel 0: DFIFO Write Access
Device OUT Endpoint 0/Host IN Channel 0: DFIFO Read Access

0x1000–0x1FFC

w
r

Device IN Endpoint 1/Host OUT Channel 1: DFIFO Write Access
Device OUT Endpoint 1/Host IN Channel 1: DFIFO Read Access

0x2000–0x2FFC

w
r

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
Table 201. Data FIFO (DFIFO) access register map (continued)
FIFO access register section

Address range

...

...

Device IN Endpoint x(1)/Host OUT Channel x(1): DFIFO Write Access
0xX000–0xXFFC
Device OUT Endpoint x(1)/Host IN Channel x(1): DFIFO Read Access

Access
...
w
r

1. Where x is 3 in device mode and 7 in host mode.

Power and clock gating CSR map
There is a single register for power and clock gating. It is available in both host and device
modes.
Table 202. Power and clock gating control and status registers
Register name

Acronym

Power and clock gating control register

Offset address: 0xE00–0xFFF

OTG_FS_PCGCCTL

0xE00-0xE04

-

0xE05–0xFFF

Reserved

RM0090 Rev 17

1271/1747
1379

USB on-the-go full-speed (OTG_FS)

34.16.2

RM0090

OTG_FS global registers
These registers are available in both host and device modes, and do not need to be
reprogrammed when switching between these modes.
Bit values in the register descriptions are expressed in binary unless otherwise specified.

OTG_FS control and status register (OTG_FS_GOTGCTL)
Address offset: 0x000
Reset value: 0x0001 0000

rw

rw

rw

r

6

5

4

Reserved

3

2

1

0

SRQ

r

7

SRQSCS

CIDSTS

r

HNPRQ

DBCT

r

HNGSCS

ASVLD

r

Reserved

8

DHNPEN

BSVLD

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

Reserved

9

HSHNPEN

The OTG_FS_GOTGCTL register controls the behavior and reflects the status of the OTG
function of the core.

rw

r

Bits 31:20 Reserved, must be kept at reset value.
Bit 19 BSVLD: B-session valid
Indicates the device mode transceiver status.
0: B-session is not valid.
1: B-session is valid.
In OTG mode, you can use this bit to determine if the device is connected or disconnected.
Note: Only accessible in device mode.
Bit 18 ASVLD: A-session valid
Indicates the host mode transceiver status.
0: A-session is not valid
1: A-session is valid
Note: Only accessible in host mode.
Bit 17 DBCT: Long/short debounce time
Indicates the debounce time of a detected connection.
0: Long debounce time, used for physical connections (100 ms + 2.5 µs)
1: Short debounce time, used for soft connections (2.5 µs)
Note: Only accessible in host mode.
Bit 16 CIDSTS: Connector ID status
Indicates the connector ID status on a connect event.
0: The OTG_FS controller is in A-device mode
1: The OTG_FS controller is in B-device mode
Note: Accessible in both device and host modes.
Bits 15:12 Reserved, must be kept at reset value.

1272/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

Bit 11 DHNPEN: Device HNP enabled
The application sets this bit when it successfully receives a SetFeature.SetHNPEnable
command from the connected USB host.
0: HNP is not enabled in the application
1: HNP is enabled in the application
Note: Only accessible in device mode.
Bit 10 HSHNPEN: host set HNP enable
The application sets this bit when it has successfully enabled HNP (using the
SetFeature.SetHNPEnable command) on the connected device.
0: Host Set HNP is not enabled
1: Host Set HNP is enabled
Note: Only accessible in host mode.
Bit 9 HNPRQ: HNP request
The application sets this bit to initiate an HNP request to the connected USB host. The
application can clear this bit by writing a 0 when the host negotiation success status change
bit in the OTG_FS_GOTGINT register (HNSSCHG bit in OTG_FS_GOTGINT) is set. The
core clears this bit when the HNSSCHG bit is cleared.
0: No HNP request
1: HNP request
Note: Only accessible in device mode.
Bit 8 HNGSCS: Host negotiation success
The core sets this bit when host negotiation is successful. The core clears this bit when the
HNP Request (HNPRQ) bit in this register is set.
0: Host negotiation failure
1: Host negotiation success
Note: Only accessible in device mode.
Bits 7:2 Reserved, must be kept at reset value.
Bit 1 SRQ: Session request
The application sets this bit to initiate a session request on the USB. The application can
clear this bit by writing a 0 when the host negotiation success status change bit in the
OTG_FS_GOTGINT register (HNSSCHG bit in OTG_FS_GOTGINT) is set. The core clears
this bit when the HNSSCHG bit is cleared.
If you use the USB 1.1 full-speed serial transceiver interface to initiate the session request,
the application must wait until VBUS discharges to 0.2 V, after the B-Session Valid bit in this
register (BSVLD bit in OTG_FS_GOTGCTL) is cleared.
0: No session request
1: Session request
Note: Only accessible in device mode.
Bit 0 SRQSCS: Session request success
The core sets this bit when a session request initiation is successful.
0: Session request failure
1: Session request success
Note: Only accessible in device mode.

OTG_FS interrupt register (OTG_FS_GOTGINT)
Address offset: 0x04
Reset value: 0x0000 0000

RM0090 Rev 17

1273/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Reserved

rc_ rc_ rc_
w1 w1 w1

9

8

7

6

5

4

Reserved

rc_ rc_
w1 w1

3

2
SEDET

16 15 14 13 12 11 10

SRSSCHG

17

HNSSCHG

18

HNGDET

Reserved

19

ADTOCHG

31 30 29 28 27 26 25 24 23 22 21 20

DBCDNE

The application reads this register whenever there is an OTG interrupt and clears the bits in
this register to clear the OTG interrupt.
1

0

Res.

rc_
w1

Bits 31:20 Reserved, must be kept at reset value.
Bit 19 DBCDNE: Debounce done
The core sets this bit when the debounce is completed after the device connect. The
application can start driving USB reset after seeing this interrupt. This bit is only valid when
the HNP Capable or SRP Capable bit is set in the OTG_FS_GUSBCFG register (HNPCAP
bit or SRPCAP bit in OTG_FS_GUSBCFG, respectively).
Note: Only accessible in host mode.
Bit 18 ADTOCHG: A-device timeout change
The core sets this bit to indicate that the A-device has timed out while waiting for the B-device
to connect.
Note: Accessible in both device and host modes.
Bit 17 HNGDET: Host negotiation detected
The core sets this bit when it detects a host negotiation request on the USB.
Note: Accessible in both device and host modes.
Bits 16:10 Reserved, must be kept at reset value.
Bit 9 HNSSCHG: Host negotiation success status change
The core sets this bit on the success or failure of a USB host negotiation request. The
application must read the host negotiation success bit of the OTG_FS_GOTGCTL register
(HNGSCS in OTG_FS_GOTGCTL) to check for success or failure.
Note: Accessible in both device and host modes.
Bits 7:3 Reserved, must be kept at reset value.
Bit 8 SRSSCHG: Session request success status change
The core sets this bit on the success or failure of a session request. The application must
read the session request success bit in the OTG_FS_GOTGCTL register (SRQSCS bit in
OTG_FS_GOTGCTL) to check for success or failure.
Note: Accessible in both device and host modes.
Bit 2 SEDET: Session end detected
The core sets this bit to indicate that the level of the voltage on VBUS is no longer valid for a
B-Peripheral session when VBUS < 0.8 V.
Bits 1:0 Reserved, must be kept at reset value.

1274/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

OTG_FS AHB configuration register (OTG_FS_GAHBCFG)
Address offset: 0x008
Reset value: 0x0000 0000

8

7

rw

rw

6

5

4

3

2

1

0
GINTMSK

Reserved

9

TXFELVL

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

PTXFELVL

This register can be used to configure the core after power-on or a change in mode. This
register mainly contains AHB system-related configuration parameters. Do not change this
register after the initial programming. The application must program this register before
starting any transactions on either the AHB or the USB.

Reserved

rw

Bits 31:9 Reserved, must be kept at reset value.
Bit 8 PTXFELVL: Periodic TxFIFO empty level
Indicates when the periodic TxFIFO empty interrupt bit in the OTG_FS_GINTSTS register
(PTXFE bit in OTG_FS_GINTSTS) is triggered.
0: PTXFE (in OTG_FS_GINTSTS) interrupt indicates that the Periodic TxFIFO is half empty
1: PTXFE (in OTG_FS_GINTSTS) interrupt indicates that the Periodic TxFIFO is completely
empty
Note: Only accessible in host mode.
Bit 7 TXFELVL: TxFIFO empty level
In device mode, this bit indicates when IN endpoint Transmit FIFO empty interrupt (TXFE in
OTG_FS_DIEPINTx.) is triggered.
0: the TXFE (in OTG_FS_DIEPINTx) interrupt indicates that the IN Endpoint TxFIFO is half
empty
1: the TXFE (in OTG_FS_DIEPINTx) interrupt indicates that the IN Endpoint TxFIFO is
completely empty
In host mode, this bit indicates when the nonperiodic Tx FIFO empty interrupt (NPTXFE bit in
OTG_FS_GINTSTS) is triggered:
0: the NPTXFE (in OTG_FS_GINTSTS) interrupt indicates that the nonperiodic Tx FIFO is
half empty
1: the NPTXFE (in OTG_FS_GINTSTS) interrupt indicates that the nonperiodic Tx FIFO is
completely empty
Bits 6:1 Reserved, must be kept at reset value.
Bit 0 GINTMSK: Global interrupt mask
The application uses this bit to mask or unmask the interrupt line assertion to itself.
Irrespective of this bit’s setting, the interrupt status registers are updated by the core.
0: Mask the interrupt assertion to the application.
1: Unmask the interrupt assertion to the application.
Note: Accessible in both device and host modes.

RM0090 Rev 17

1275/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

OTG_FS USB configuration register (OTG_FS_GUSBCFG)
Address offset: 0x00C
Reset value: 0x0000 1440

12

FHMOD

rw

rw

rw

11

10

TRDT
rw

rw

rw

7

Res.

6
PHYSEL

29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13

FDMOD

8
SRPCAP

30

Reserved

9
HNPCAP

31
CTXPKT

This register can be used to configure the core after power-on or a changing to host mode
or device mode. It contains USB and USB-PHY related configuration parameters. The
application must program this register before starting any transactions on either the AHB or
the USB. Do not make changes to this register after the initial programming.
5

4

3

Reserved

r

2

1

0

TOCAL
rw

Bit 31 CTXPKT: Corrupt Tx packet
This bit is for debug purposes only. Never set this bit to 1.
Note: Accessible in both device and host modes.
Bit 30 FDMOD: Force device mode
Writing a 1 to this bit forces the core to device mode irrespective of the OTG_FS_ID input
pin.
0: Normal mode
1: Force device mode
After setting the force bit, the application must wait at least 25 ms before the change takes
effect.
Note: Accessible in both device and host modes.
Bit 29 FHMOD: Force host mode
Writing a 1 to this bit forces the core to host mode irrespective of the OTG_FS_ID input pin.
0: Normal mode
1: Force host mode
After setting the force bit, the application must wait at least 25 ms before the change takes
effect.
Note: Accessible in both device and host modes.
Bits 28:14 Reserved, must be kept at reset value.
Bits 13:10 TRDT: USB turnaround time
These bits allow setting the turnaround time in PHY clocks. They must be configured
according to Table 203: TRDT values, depending on the application AHB frequency. Higher
TRDT values allow stretching the USB response time to IN tokens in order to compensate for
longer AHB read access latency to the Data FIFO.
Note: Only accessible in device mode.
Bit 9 HNPCAP: HNP-capable
The application uses this bit to control the OTG_FS controller’s HNP capabilities.
0: HNP capability is not enabled.
1: HNP capability is enabled.
Note: Accessible in both device and host modes.

1276/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

Bit 8 SRPCAP: SRP-capable
The application uses this bit to control the OTG_FS controller’s SRP capabilities. If the core
operates as a non-SRP-capable
B-device, it cannot request the connected A-device (host) to activate VBUS and start a
session.
0: SRP capability is not enabled.
1: SRP capability is enabled.
Note: Accessible in both device and host modes.
Bit 7 Reserved, must be kept at reset value.
Bit 6 PHYSEL: Full speed serial transceiver select
This bit is always 1 with read-only access.
Bits 5:3 Reserved, must be kept at reset value.
Bits 2:0 TOCAL: FS timeout calibration
The number of PHY clocks that the application programs in this field is added to the fullspeed interpacket timeout duration in the core to account for any additional delays
introduced by the PHY. This can be required, because the delay introduced by the PHY in
generating the line state condition can vary from one PHY to another.
The USB standard timeout value for full-speed operation is 16 to 18 (inclusive) bit times. The
application must program this field based on the speed of enumeration. The number of bit
times added per PHY clock is 0.25 bit times.

Table 203. TRDT values
AHB frequency range (MHz)
TRDT minimum value
Min.

Max

14.2

15

0xF

15

16

0xE

16

17.2

0xD

17.2

18.5

0xC

18.5

20

0xB

20

21.8

0xA

21.8

24

0x9

24

27.5

0x8

27.5

32

0x7

32

-

0x6

RM0090 Rev 17

1277/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

OTG_FS reset register (OTG_FS_GRSTCTL)
Address offset: 0x010
Reset value: 0x8000 0000

TXFNUM
rw

rs

rs

3

2

1

0
CSRST

6

HSRST

7

FCRST

4
RXFFLSH

r

8

Reserved

9

Reserved

5
TXFFLSH

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
AHBIDL

The application uses this register to reset various hardware features inside the core.

rs

rs

rs

Bit 31 AHBIDL: AHB master idle
Indicates that the AHB master state machine is in the Idle condition.
Note: Accessible in both device and host modes.
Bits 30:11 Reserved, must be kept at reset value.
Bits 10:6 TXFNUM: TxFIFO number
This is the FIFO number that must be flushed using the TxFIFO Flush bit. This field must not
be changed until the core clears the TxFIFO Flush bit.
00000:
–
Non-periodic TxFIFO flush in host mode
–
Tx FIFO 0 flush in device mode
00001:
–
Periodic TxFIFO flush in host mode
–
TXFIFO 1 flush in device mode
00010: TXFIFO 2 flush in device mode
...
00101: TXFIFO 15 flush in device mode
10000: Flush all the transmit FIFOs in device or host mode.
Note: Accessible in both device and host modes.
Bit 5 TXFFLSH: TxFIFO flush
This bit selectively flushes a single or all transmit FIFOs, but cannot do so if the core is in the
midst of a transaction.
The application must write this bit only after checking that the core is neither writing to the
TxFIFO nor reading from the TxFIFO. Verify using these registers:
Read—NAK Effective Interrupt ensures the core is not reading from the FIFO
Write—AHBIDL bit in OTG_FS_GRSTCTL ensures the core is not writing anything to the
FIFO.
Note: Accessible in both device and host modes.
Bit 4 RXFFLSH: RxFIFO flush
The application can flush the entire RxFIFO using this bit, but must first ensure that the core
is not in the middle of a transaction.
The application must only write to this bit after checking that the core is neither reading from
the RxFIFO nor writing to the RxFIFO.
The application must wait until the bit is cleared before performing any other operations. This
bit requires 8 clocks (slowest of PHY or AHB clock) to clear.
Note: Accessible in both device and host modes.
Bit 3 Reserved, must be kept at reset value.

1278/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

Bit 2 FCRST: Host frame counter reset
The application writes this bit to reset the frame number counter inside the core. When the
frame counter is reset, the subsequent SOF sent out by the core has a frame number of 0.
Note: Only accessible in host mode.
Bit 1 HSRST: HCLK soft reset
The application uses this bit to flush the control logic in the AHB Clock domain. Only AHB
Clock Domain pipelines are reset.
FIFOs are not flushed with this bit.
All state machines in the AHB clock domain are reset to the Idle state after terminating the
transactions on the AHB, following the protocol.
CSR control bits used by the AHB clock domain state machines are cleared.
To clear this interrupt, status mask bits that control the interrupt status and are generated by
the AHB clock domain state machine are cleared.
Because interrupt status bits are not cleared, the application can get the status of any core
events that occurred after it set this bit.
This is a self-clearing bit that the core clears after all necessary logic is reset in the core. This
can take several clocks, depending on the core’s current state.
Note: Accessible in both device and host modes.
Bit 0 CSRST: Core soft reset
Resets the HCLK and PCLK domains as follows:
Clears the interrupts and all the CSR register bits except for the following bits:
– RSTPDMODL bit in OTG_FS_PCGCCTL
– GAYEHCLK bit in OTG_FS_PCGCCTL
– PWRCLMP bit in OTG_FS_PCGCCTL
– STPPCLK bit in OTG_FS_PCGCCTL
– FSLSPCS bit in OTG_FS_HCFG
– DSPD bit in OTG_FS_DCFG
All module state machines (except for the AHB slave unit) are reset to the Idle state, and all
the transmit FIFOs and the receive FIFO are flushed.
Any transactions on the AHB Master are terminated as soon as possible, after completing the
last data phase of an AHB transfer. Any transactions on the USB are terminated immediately.
The application can write to this bit any time it wants to reset the core. This is a self-clearing
bit and the core clears this bit after all the necessary logic is reset in the core, which can take
several clocks, depending on the current state of the core. Once this bit has been cleared,
the software must wait at least 3 PHY clocks before accessing the PHY domain
(synchronization delay). The software must also check that bit 31 in this register is set to 1
(AHB Master is Idle) before starting any operation.
Typically, the software reset is used during software development and also when you
dynamically change the PHY selection bits in the above listed USB configuration registers.
When you change the PHY, the corresponding clock for the PHY is selected and used in the
PHY domain. Once a new clock is selected, the PHY domain has to be reset for proper
operation.
Note: Accessible in both device and host modes.

RM0090 Rev 17

1279/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

OTG_FS core interrupt register (OTG_FS_GINTSTS)
Address offset: 0x014
Reset value: 0x0400 0020
This register interrupts the application for system-level events in the current mode (device
mode or host mode).
Some of the bits in this register are valid only in host mode, while others are valid in device
mode only. This register also indicates the current mode. To clear the interrupt status bits of
the rc_w1 type, the application must write 1 into the bit.
The FIFO status interrupts are read-only; once software reads from or writes to the FIFO
while servicing these interrupts, FIFO interrupt conditions are cleared automatically.

4

r

r

r

1

0

MMIS

5

CMOD

SOF

6

r

rc_w1

2

RXFLVL

7

OTGINT

3

r

rc_w1

Reserved

ESUSP

8

NPTXFE

rc_w1

USBSUSP

USBRST

ISOODRP

ENUMDNE

r

EOPF

r

9

GINAKEFF

rc_w1

Reserved

Res.

IEPINT

Reserved

r

OEPINT

HPRTINT

r

IISOIXFR

HCINT

r

IPXFR/INCOMPISOOUT

PTXFE

Reserved

DISCINT

rc_w1

CIDSCHG

SRQINT

WKUINT

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

GONAKEFF

The application must clear the OTG_FS_GINTSTS register at initialization before
unmasking the interrupt bit to avoid any interrupts generated prior to initialization.

r

Bit 31 WKUPINT: Resume/remote wakeup detected interrupt
In device mode, this interrupt is asserted when a resume is detected on the USB. In host
mode, this interrupt is asserted when a remote wakeup is detected on the USB.
Note: Accessible in both device and host modes.
Bit 30 SRQINT: Session request/new session detected interrupt
In host mode, this interrupt is asserted when a session request is detected from the device.
In device mode, this interrupt is asserted when VBUS is in the valid range for a B-peripheral
device. Accessible in both device and host modes.
Bit 29 DISCINT: Disconnect detected interrupt
Asserted when a device disconnect is detected.
Note: Only accessible in host mode.
Bit 28 CIDSCHG: Connector ID status change
The core sets this bit when there is a change in connector ID status.
Note: Accessible in both device and host modes.
Bit 27 Reserved, must be kept at reset value.
Bit 26 PTXFE: Periodic TxFIFO empty
Asserted when the periodic transmit FIFO is either half or completely empty and there is
space for at least one entry to be written in the periodic request queue. The half or
completely empty status is determined by the periodic TxFIFO empty level bit in the
OTG_FS_GAHBCFG register (PTXFELVL bit in OTG_FS_GAHBCFG).
Note: Only accessible in host mode.

1280/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

Bit 25 HCINT: Host channels interrupt
The core sets this bit to indicate that an interrupt is pending on one of the channels of the
core (in host mode). The application must read the OTG_FS_HAINT register to determine
the exact number of the channel on which the interrupt occurred, and then read the
corresponding OTG_FS_HCINTx register to determine the exact cause of the interrupt. The
application must clear the appropriate status bit in the OTG_FS_HCINTx register to clear
this bit.
Note: Only accessible in host mode.
Bit 24 HPRTINT: Host port interrupt
The core sets this bit to indicate a change in port status of one of the OTG_FS controller
ports in host mode. The application must read the OTG_FS_HPRT register to determine the
exact event that caused this interrupt. The application must clear the appropriate status bit in
the OTG_FS_HPRT register to clear this bit.
Note: Only accessible in host mode.
Bits 23:22 Reserved, must be kept at reset value.
Bit 21 IPXFR: Incomplete periodic transfer
In host mode, the core sets this interrupt bit when there are incomplete periodic transactions
still pending, which are scheduled for the current frame.
INCOMPISOOUT: Incomplete isochronous OUT transfer
In device mode, the core sets this interrupt to indicate that there is at least one isochronous
OUT endpoint on which the transfer is not completed in the current frame. This interrupt is
asserted along with the End of periodic frame interrupt (EOPF) bit in this register.
Bit 20 IISOIXFR: Incomplete isochronous IN transfer
The core sets this interrupt to indicate that there is at least one isochronous IN endpoint on
which the transfer is not completed in the current frame. This interrupt is asserted along with
the End of periodic frame interrupt (EOPF) bit in this register.
Note: Only accessible in device mode.
Bit 19 OEPINT: OUT endpoint interrupt
The core sets this bit to indicate that an interrupt is pending on one of the OUT endpoints of
the core (in device mode). The application must read the OTG_FS_DAINT register to
determine the exact number of the OUT endpoint on which the interrupt occurred, and then
read the corresponding OTG_FS_DOEPINTx register to determine the exact cause of the
interrupt. The application must clear the appropriate status bit in the corresponding
OTG_FS_DOEPINTx register to clear this bit.
Note: Only accessible in device mode.
Bit 18 IEPINT: IN endpoint interrupt
The core sets this bit to indicate that an interrupt is pending on one of the IN endpoints of the
core (in device mode). The application must read the OTG_FS_DAINT register to determine
the exact number of the IN endpoint on which the interrupt occurred, and then read the
corresponding OTG_FS_DIEPINTx register to determine the exact cause of the interrupt.
The application must clear the appropriate status bit in the corresponding
OTG_FS_DIEPINTx register to clear this bit.
Note: Only accessible in device mode.
Bits 17:16 Reserved, must be kept at reset value.
Bit 15 EOPF: End of periodic frame interrupt
Indicates that the period specified in the periodic frame interval field of the OTG_FS_DCFG
register (PFIVL bit in OTG_FS_DCFG) has been reached in the current frame.
Note: Only accessible in device mode.

RM0090 Rev 17

1281/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Bit 14 ISOODRP: Isochronous OUT packet dropped interrupt
The core sets this bit when it fails to write an isochronous OUT packet into the RxFIFO
because the RxFIFO does not have enough space to accommodate a maximum size packet
for the isochronous OUT endpoint.
Note: Only accessible in device mode.
Bit 13 ENUMDNE: Enumeration done
The core sets this bit to indicate that speed enumeration is complete. The application must
read the OTG_FS_DSTS register to obtain the enumerated speed.
Note: Only accessible in device mode.
Bit 12 USBRST: USB reset
The core sets this bit to indicate that a reset is detected on the USB.
Note: Only accessible in device mode.
Bit 11 USBSUSP: USB suspend
The core sets this bit to indicate that a suspend was detected on the USB. The core enters
the Suspended state when there is no activity on the data lines for a period of 3 ms.
Note: Only accessible in device mode.
Bit 10 ESUSP: Early suspend
The core sets this bit to indicate that an Idle state has been detected on the USB for 3 ms.
Note: Only accessible in device mode.
Bits 9:8 Reserved, must be kept at reset value.
Bit 7 GONAKEFF: Global OUT NAK effective
Indicates that the Set global OUT NAK bit in the OTG_FS_DCTL register (SGONAK bit in
OTG_FS_DCTL), set by the application, has taken effect in the core. This bit can be cleared
by writing the Clear global OUT NAK bit in the OTG_FS_DCTL register (CGONAK bit in
OTG_FS_DCTL).
Note: Only accessible in device mode.
Bit 6 GINAKEFF: Global IN non-periodic NAK effective
Indicates that the Set global non-periodic IN NAK bit in the OTG_FS_DCTL register
(SGINAK bit in OTG_FS_DCTL), set by the application, has taken effect in the core. That is,
the core has sampled the Global IN NAK bit set by the application. This bit can be cleared by
clearing the Clear global non-periodic IN NAK bit in the OTG_FS_DCTL register (CGINAK
bit in OTG_FS_DCTL).
This interrupt does not necessarily mean that a NAK handshake is sent out on the USB. The
STALL bit takes precedence over the NAK bit.
Note: Only accessible in device mode.
Bit 5 NPTXFE: Non-periodic TxFIFO empty
This interrupt is asserted when the non-periodic TxFIFO is either half or completely empty,
and there is space for at least one entry to be written to the non-periodic transmit request
queue. The half or completely empty status is determined by the non-periodic TxFIFO empty
level bit in the OTG_FS_GAHBCFG register (TXFELVL bit in OTG_FS_GAHBCFG).
Note: Accessible in host mode only.
Bit 4 RXFLVL: RxFIFO non-empty
Indicates that there is at least one packet pending to be read from the RxFIFO.
Note: Accessible in both host and device modes.

1282/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

Bit 3 SOF: Start of frame
In host mode, the core sets this bit to indicate that an SOF (FS), or Keep-Alive (LS) is
transmitted on the USB. The application must write a 1 to this bit to clear the interrupt.
In device mode, in the core sets this bit to indicate that an SOF token has been received on
the USB. The application can read the Device Status register to get the current frame
number. This interrupt is seen only when the core is operating in FS.
Note: Accessible in both host and device modes.
Bit 2 OTGINT: OTG interrupt
The core sets this bit to indicate an OTG protocol event. The application must read the OTG
Interrupt Status (OTG_FS_GOTGINT) register to determine the exact event that caused this
interrupt. The application must clear the appropriate status bit in the OTG_FS_GOTGINT
register to clear this bit.
Note: Accessible in both host and device modes.
Bit 1 MMIS: Mode mismatch interrupt
The core sets this bit when the application is trying to access:
–
A host mode register, when the core is operating in device mode
–
A device mode register, when the core is operating in host mode
The register access is completed on the AHB with an OKAY response, but is ignored by the
core internally and does not affect the operation of the core.
Note: Accessible in both host and device modes.
Bit 0 CMOD: Current mode of operation
Indicates the current mode.
0: Device mode
1: Host mode
Note: Accessible in both host and device modes.

RM0090 Rev 17

1283/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

OTG_FS interrupt mask register (OTG_FS_GINTMSK)
Address offset: 0x018
Reset value: 0x0000 0000

Bit 31 WUIM: Resume/remote wakeup detected interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both host and device modes.
Bit 30 SRQIM: Session request/new session detected interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both host and device modes.
Bit 29 DISCINT: Disconnect detected interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 28 CIDSCHGM: Connector ID status change mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both host and device modes.
Bit 27 Reserved, must be kept at reset value.
Bit 26 PTXFEM: Periodic TxFIFO empty mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in host mode.
Bit 25 HCIM: Host channels interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in host mode.
Bit 24 PRTIM: Host port interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in host mode.

1284/1747

RM0090 Rev 17

3

2

1

MMISM

rw

rw

rw

rw

rw

rw

rw

0

Reserved

4

OTGINT

rw

5

SOFM

rw

6

RXFLVLM

rw

7

NPTXFEM

rw

8

GINAKEFFM

ESUSPM

rw

9

Reserved

USBSUSPM

rw

USBRST

rw

ISOODRPM

rw

ENUMDNEM

rw

EOPFM

IEPINT

rw

Reserved

OEPINT

rw

IISOIXFRM

rw

IPXFRM/IISOOXFRM

rw

Reserved

rw

PRTIM

rw

HCIM

rw

PTXFEM

DISCINT

CIDSCHGM

rw

Reserved

WUIM

SRQIM

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

GONAKEFFM

This register works with the Core interrupt register to interrupt the application. When an
interrupt bit is masked, the interrupt associated with that bit is not generated. However, the
Core Interrupt (OTG_FS_GINTSTS) register bit corresponding to that interrupt is still set.

RM0090

USB on-the-go full-speed (OTG_FS)

Bits 23:22 Reserved, must be kept at reset value.
Bit 21 IPXFRM: Incomplete periodic transfer mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in host mode.
IISOOXFRM: Incomplete isochronous OUT transfer mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 20 IISOIXFRM: Incomplete isochronous IN transfer mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 19 OEPINT: OUT endpoints interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 18 IEPINT: IN endpoints interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bits 17:16 Reserved, must be kept at reset value.
Bit 15 EOPFM: End of periodic frame interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 14 ISOODRPM: Isochronous OUT packet dropped interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 13 ENUMDNEM: Enumeration done mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 12 USBRST: USB reset mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 11 USBSUSPM: USB suspend mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.

RM0090 Rev 17

1285/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Bit 10 ESUSPM: Early suspend mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bits 9:8 Reserved, must be kept at reset value.
Bit 7 GONAKEFFM: Global OUT NAK effective mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 6 GINAKEFFM: Global non-periodic IN NAK effective mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 5 NPTXFEM: Non-periodic TxFIFO empty mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in Host mode.
Bit 4 RXFLVLM: Receive FIFO non-empty mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both device and host modes.
Bit 3 SOFM: Start of frame mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both device and host modes.
Bit 2 OTGINT: OTG interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both device and host modes.
Bit 1 MMISM: Mode mismatch interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both device and host modes.
Bit 0 Reserved, must be kept at reset value.

1286/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

OTG_FS Receive status debug read/OTG status read and pop registers
(OTG_FS_GRXSTSR/OTG_FS_GRXSTSP)
Address offset for Read: 0x01C
Address offset for Pop: 0x020
Reset value: 0x0000 0000
A read to the Receive status debug read register returns the contents of the top of the
Receive FIFO. A read to the Receive status read and pop register additionally pops the top
data entry out of the RxFIFO.
The receive status contents must be interpreted differently in host and device modes. The
core ignores the receive status pop/read when the receive FIFO is empty and returns a
value of 0x0000 0000. The application must only pop the Receive Status FIFO when the
Receive FIFO non-empty bit of the Core interrupt register (RXFLVL bit in
OTG_FS_GINTSTS) is asserted.
Host mode
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

7

6

5

4

3

2

1

PKTSTS

DPID

BCNT

CHNUM

r

r

r

r

0

Bits 31:21 Reserved, must be kept at reset value.
Bits 20:17 PKTSTS: Packet status
Indicates the status of the received packet
0010: IN data packet received
0011: IN transfer completed (triggers an interrupt)
0101: Data toggle error (triggers an interrupt)
0111: Channel halted (triggers an interrupt)
Others: Reserved
Bits 16:15 DPID: Data PID
Indicates the Data PID of the received packet
00: DATA0
10: DATA1
01: DATA2
11: MDATA
Bits 14:4 BCNT: Byte count
Indicates the byte count of the received IN data packet.
Bits 3:0 CHNUM: Channel number
Indicates the channel number to which the current received packet belongs.

RM0090 Rev 17

1287/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Device mode
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

7

6

5

4

3

2

1

FRMNUM

PKTSTS

DPID

BCNT

EPNUM

r

r

r

r

r

0

Bits 31:25 Reserved, must be kept at reset value.
Bits 24:21 FRMNUM: Frame number
This is the least significant 4 bits of the frame number in which the packet is received on the
USB. This field is supported only when isochronous OUT endpoints are supported.
Bits 20:17 PKTSTS: Packet status
Indicates the status of the received packet
0001: Global OUT NAK (triggers an interrupt)
0010: OUT data packet received
0011: OUT transfer completed (triggers an interrupt)
0100: SETUP transaction completed (triggers an interrupt)
0110: SETUP data packet received
Others: Reserved
Bits 16:15 DPID: Data PID
Indicates the Data PID of the received OUT data packet
00: DATA0
10: DATA1
01: DATA2
11: MDATA
Bits 14:4 BCNT: Byte count
Indicates the byte count of the received data packet.
Bits 3:0 EPNUM: Endpoint number
Indicates the endpoint number to which the current received packet belongs.

OTG_FS Receive FIFO size register (OTG_FS_GRXFSIZ)
Address offset: 0x024
Reset value: 0x0000 0200
The application can program the RAM size that must be allocated to the RxFIFO.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

RXFD

Reserved

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 RXFD: RxFIFO depth
This value is in terms of 32-bit words.
Minimum value is 16
Maximum value is 256
The power-on reset value of this register is specified as the largest Rx data FIFO depth.

1288/1747

RM0090 Rev 17

0

RM0090

USB on-the-go full-speed (OTG_FS)

OTG_FS Host non-periodic transmit FIFO size register
(OTG_FS_HNPTXFSIZ)/Endpoint 0 Transmit FIFO size (OTG_FS_DIEPTXF0)
Address offset: 0x028
Reset value: 0x0000 0200
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

NPTXFD/TX0FD

NPTXFSA/TX0FSA

rw

rw

5

4

3

2

1

0

Host mode
Bits 31:16 NPTXFD: Non-periodic TxFIFO depth
This value is in terms of 32-bit words.
Minimum value is 16
Maximum value is 256
Bits 15:0 NPTXFSA: Non-periodic transmit RAM start address
This field contains the memory start address for non-periodic transmit FIFO RAM.

Device mode
Bits 31:16 TX0FD: Endpoint 0 TxFIFO depth
This value is in terms of 32-bit words.
Minimum value is 16
Maximum value is 256
Bits 15:0 TX0FSA: Endpoint 0 transmit RAM start address
This field contains the memory start address for the endpoint 0 transmit FIFO RAM.

OTG_FS non-periodic transmit FIFO/queue status register
(OTG_FS_HNPTXSTS)
Address offset: 0x02C
Reset value: 0x0008 0200
Note:

In Device mode, this register is not valid.
This read-only register contains the free space information for the non-periodic TxFIFO and
the non-periodic transmit request queue.

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

NPTXQTOP

NPTQXSAV

NPTXFSAV

r

r

r

RM0090 Rev 17

6

5

4

3

2

1

0

1289/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Bit 31 Reserved, must be kept at reset value.
Bits 30:24 NPTXQTOP: Top of the non-periodic transmit request queue
Entry in the non-periodic Tx request queue that is currently being processed by the MAC.
Bits 30:27: Channel/endpoint number
Bits 26:25:
–
00: IN/OUT token
–
01: Zero-length transmit packet (device IN/host OUT)
–
11: Channel halt command
Bit 24: Terminate (last entry for selected channel/endpoint)
Bits 23:16 NPTQXSAV: Non-periodic transmit request queue space available
Indicates the amount of free space available in the non-periodic transmit request queue.
This queue holds both IN and OUT requests in host mode. Device mode has only IN
requests.
00: Non-periodic transmit request queue is full
01: 1 location available
10: 2 locations available
bxn: n locations available (0 ≤ n ≤ 8)
Others: Reserved
Bits 15:0 NPTXFSAV: Non-periodic TxFIFO space available
Indicates the amount of free space available in the non-periodic TxFIFO.
Values are in terms of 32-bit words.
00: Non-periodic TxFIFO is full
01: 1 word available
10: 2 words available
0xn: n words available (where 0 ≤ n ≤ 256)
Others: Reserved

OTG_FS general core configuration register (OTG_FS_GCCFG)
Address offset: 0x038
Reset value: 0x0000 XXXX

1290/1747

VBUSASEN

rw

rw

rw

.PWRDWN

VBUSBSEN

rw

Reserved

SOFOUTEN

Reserved

NOVBUSSENS

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

rw

RM0090 Rev 17

9

8

7

Reserved

6

5

4

3

2

1

0

RM0090

USB on-the-go full-speed (OTG_FS)

Bits 31:22 Reserved, must be kept at reset value.
Bit 21 NOVBUSSENS: VBUS sensing disable option
When this bit is set, VBUS is considered internally to be always at VBUS valid level (5 V). This
option removes the need for a dedicated VBUS pad, and leave this pad free to be used for
other purposes such as a shared functionality. VBUS connection can be remapped on
another general purpose input pad and monitored by software.
This option is only suitable for host-only or device-only applications.
0: VBUS sensing available by hardware
1: VBUS sensing not available by hardware.
Bit 20 SOFOUTEN: SOF output enable
0: SOF pulse not available on PAD (OTG_FS_SOF)
1: SOF pulse available on PAD (OTG_FS_SOF)
Bit 19 VBUSBSEN: Enable the VBUS sensing “B” device
0: VBUS sensing “B” disabled
1: VBUS sensing “B” enabled
Bit 18 VBUSASEN: Enable the VBUS sensing “A” device
0: VBUS sensing “A” disabled
1: VBUS sensing “A” enabled
Bit 17 Reserved, must be kept at reset value.
Bit 16 PWRDWN: Power down
Used to activate the transceiver in transmission/reception
0: Power down active
1: Power down deactivated (“Transceiver active”)
Bits 15:0 Reserved, must be kept at reset value.

OTG_FS core ID register (OTG_FS_CID)
Address offset: 0x03C
Reset value:0x0000 1200
This is a register containing the Product ID as reset value.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

PRODUCT_ID
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 PRODUCT_ID: Product ID field
Application-programmable ID field.

RM0090 Rev 17

1291/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

OTG_FS Host periodic transmit FIFO size register (OTG_FS_HPTXFSIZ)
Address offset: 0x100
Reset value: 0x0200 0400
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

PTXFSIZ
r

r

r

r

r

r

r

r

r

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

PTXSA
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:16 PTXFD: Host periodic TxFIFO depth
This value is in terms of 32-bit words.
Minimum value is 16
Bits 15:0 PTXSA: Host periodic TxFIFO start address
The power-on reset value of this register is the sum of the largest Rx data FIFO depth and
largest non-periodic Tx data FIFO depth.

OTG_FS device IN endpoint transmit FIFO size register (OTG_FS_DIEPTXFx)
(x = 1..3, where x is the FIFO_number)
Address offset: 0x104 + 0x04 * (x -1)
Reset value: 0x0200 0200
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

INEPTXFD
r

r

r

r

r

r

r

r

r

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

INEPTXSA
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:16 INEPTXFD: IN endpoint TxFIFO depth
This value is in terms of 32-bit words.
Minimum value is 16
The power-on reset value of this register is specified as the largest IN endpoint FIFO
number depth.
Bits 15:0 INEPTXSA: IN endpoint FIFOx transmit RAM start address
This field contains the memory start address for IN endpoint transmit FIFOx. The address
must be aligned with a 32-bit memory location.

1292/1747

RM0090 Rev 17

RM0090

34.16.3

USB on-the-go full-speed (OTG_FS)

Host-mode registers
Bit values in the register descriptions are expressed in binary unless otherwise specified.
Host-mode registers affect the operation of the core in the host mode. Host mode registers
must not be accessed in device mode, as the results are undefined. Host mode registers
can be categorized as follows:

OTG_FS Host configuration register (OTG_FS_HCFG)
Address offset: 0x400
Reset value: 0x0000 0000
This register configures the core after power-on. Do not make changes to this register after
initializing the host.
8

7

6

5

4

3

2

1

Reserved

r

0
FSLSPCS

9

FSLSS

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

rw

rw

Bits 31:3 Reserved, must be kept at reset value.
Bit 2 FSLSS: FS- and LS-only support
The application uses this bit to control the core’s enumeration speed. Using this bit, the
application can make the core enumerate as an FS host, even if the connected device
supports HS traffic. Do not make changes to this field after initial programming.
1: FS/LS-only, even if the connected device can support HS (read-only)
Bits 1:0 FSLSPCS: FS/LS PHY clock select
When the core is in FS host mode
01: PHY clock is running at 48 MHz
Others: Reserved
When the core is in LS host mode
00: Reserved
01: Select 48 MHz PHY clock frequency
10: Select 6 MHz PHY clock frequency
11: Reserved
Note: The FSLSPCS must be set on a connection event according to the speed of the
connected device (after changing this bit, a software reset must be performed).

OTG_FS Host frame interval register (OTG_FS_HFIR)
Address offset: 0x404
Reset value: 0x0000 EA60
This register stores the frame interval information for the current speed to which the
OTG_FS controller has enumerated.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

rw

rw

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

FRIVL
rw

rw

rw

RM0090 Rev 17

rw

rw

rw

rw

1293/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 FRIVL: Frame interval
The value that the application programs to this field specifies the interval between two
consecutive SOFs (FS) or Keep-Alive tokens (LS). This field contains the number of PHY
clocks that constitute the required frame interval. The application can write a value to this
register only after the Port enable bit of the host port control and status register (PENA bit in
OTG_FS_HPRT) has been set. If no value is programmed, the core calculates the value
based on the PHY clock specified in the FS/LS PHY Clock Select field of the host
configuration register (FSLSPCS in OTG_FS_HCFG). Do not change the value of this field
after the initial configuration.
– Frame interval = 1 ms × (FRIVL - 1)

OTG_FS Host frame number/frame time remaining register (OTG_FS_HFNUM)
Address offset: 0x408
Reset value: 0x0000 3FFF
This register indicates the current frame number. It also indicates the time remaining (in
terms of the number of PHY clocks) in the current frame.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

FTREM
r

r

r

r

r

r

r

r

r

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

FRNUM
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:16 FTREM: Frame time remaining
Indicates the amount of time remaining in the current frame, in terms of PHY clocks. This
field decrements on each PHY clock. When it reaches zero, this field is reloaded with the
value in the Frame interval register and a new SOF is transmitted on the USB.
Bits 15:0 FRNUM: Frame number
This field increments when a new SOF is transmitted on the USB, and is cleared to 0 when
it reaches 0x3FFF.

OTG_FS_Host periodic transmit FIFO/queue status register
(OTG_FS_HPTXSTS)
Address offset: 0x410
Reset value: 0x0008 0100
This read-only register contains the free space information for the periodic TxFIFO and the
periodic transmit request queue.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
PTXQTOP
r

r

r

1294/1747

r

r

9

PTXQSAV
r

r

r

r

r

r

r

r

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

PTXFSAVL
r

r

r

r

r

RM0090 Rev 17

r

r

r

r

r

r

r

RM0090

USB on-the-go full-speed (OTG_FS)

Bits 31:24 PTXQTOP: Top of the periodic transmit request queue
This indicates the entry in the periodic Tx request queue that is currently being processed by
the MAC.
This register is used for debugging.
Bit 31: Odd/Even frame
–
0: send in even frame
–
1: send in odd frame
Bits 30:27: Channel/endpoint number
Bits 26:25: Type
–
00: IN/OUT
–
01: Zero-length packet
–
11: Disable channel command
Bit 24: Terminate (last entry for the selected channel/endpoint)
Bits 23:16 PTXQSAV: Periodic transmit request queue space available
Indicates the number of free locations available to be written in the periodic transmit request
queue. This queue holds both IN and OUT requests.
00: Periodic transmit request queue is full
01: 1 location available
10: 2 locations available
bxn: n locations available (0 ≤ n ≤ 8)
Others: Reserved
Bits 15:0 PTXFSAVL: Periodic transmit data FIFO space available
Indicates the number of free locations available to be written to in the periodic TxFIFO.
Values are in terms of 32-bit words
0000: Periodic TxFIFO is full
0001: 1 word available
0010: 2 words available
bxn: n words available (where 0 ≤ n ≤ PTXFD)
Others: Reserved

OTG_FS Host all channels interrupt register (OTG_FS_HAINT)
Address offset: 0x414
Reset value: 0x0000 000
When a significant event occurs on a channel, the host all channels interrupt register
interrupts the application using the host channels interrupt bit of the Core interrupt register
(HCINT bit in OTG_FS_GINTSTS). This is shown in Figure 394. There is one interrupt bit
per channel, up to a maximum of 16 bits. Bits in this register are set and cleared when the
application sets and clears bits in the corresponding host channel-x interrupt register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

HAINT
r

r

r

r

r

r

r

r

r

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 HAINT: Channel interrupts
One bit per channel: Bit 0 for Channel 0, bit 15 for Channel 15

RM0090 Rev 17

1295/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

OTG_FS Host all channels interrupt mask register (OTG_FS_HAINTMSK)
Address offset: 0x418
Reset value: 0x0000 0000
The host all channel interrupt mask register works with the host all channel interrupt register
to interrupt the application when an event occurs on a channel. There is one interrupt mask
bit per channel, up to a maximum of 16 bits.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

HAINTM

Reserved

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 HAINTM: Channel interrupt mask
0: Masked interrupt
1: Unmasked interrupt
One bit per channel: Bit 0 for channel 0, bit 15 for channel 15

OTG_FS Host port control and status register (OTG_FS_HPRT)
Address offset: 0x440
Reset value: 0x0000 0000
This register is available only in host mode. Currently, the OTG host supports only one port.

rw

rw

rw

rw

r

r

Bits 31:19 Reserved, must be kept at reset value.
Bits 18:17 PSPD: Port speed
Indicates the speed of the device attached to this port.
01: Full speed
10: Low speed
11: Reserved

1296/1747

RM0090 Rev 17

rs

rw

rc_
w1

r

2

1

0
PCSTS

rw

3

PENA

6

PCDET

7

PENCHNG

POCA

rw

4

POCCHNG

r

5

PRES

r

8
PRST

PTCTL

Reserved

9

Reserved

PPWR

PSPD

PLSTS

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

PSUSP

A single register holds USB port-related information such as USB reset, enable, suspend,
resume, connect status, and test mode for each port. It is shown in Figure 394. The rc_w1
bits in this register can trigger an interrupt to the application through the host port interrupt
bit of the core interrupt register (HPRTINT bit in OTG_FS_GINTSTS). On a Port Interrupt,
the application must read this register and clear the bit that caused the interrupt. For the
rc_w1 bits, the application must write a 1 to the bit to clear the interrupt.

rc_ rc_ rc_
w1 w0 w1

r

RM0090

USB on-the-go full-speed (OTG_FS)

Bits 16:13 PTCTL: Port test control
The application writes a nonzero value to this field to put the port into a Test mode, and the
corresponding pattern is signaled on the port.
0000: Test mode disabled
0001: Test_J mode
0010: Test_K mode
0011: Test_SE0_NAK mode
0100: Test_Packet mode
0101: Test_Force_Enable
Others: Reserved
Bit 12 PPWR: Port power
The application uses this field to control power to this port, and the core clears this bit on an
overcurrent condition.
0: Power off
1: Power on
Bits 11:10 PLSTS: Port line status
Indicates the current logic level USB data lines
Bit 10: Logic level of OTG_FS_DP
Bit 11: Logic level of OTG_FS_DM
Bit 9 Reserved, must be kept at reset value.
Bit 8 PRST: Port reset
When the application sets this bit, a reset sequence is started on this port. The application
must time the reset period and clear this bit after the reset sequence is complete.
0: Port not in reset
1: Port in reset
The application must leave this bit set for a minimum duration of at least 10 ms to start a
reset on the port. The application can leave it set for another 10 ms in addition to the
required minimum duration, before clearing the bit, even though there is no maximum limit
set by the USB standard.
Bit 7 PSUSP: Port suspend
The application sets this bit to put this port in Suspend mode. The core only stops sending
SOFs when this is set. To stop the PHY clock, the application must set the Port clock stop
bit, which asserts the suspend input pin of the PHY.
The read value of this bit reflects the current suspend status of the port. This bit is cleared
by the core after a remote wakeup signal is detected or the application sets the Port reset bit
or Port resume bit in this register or the Resume/remote wakeup detected interrupt bit or
Disconnect detected interrupt bit in the Core interrupt register (WKUINT or DISCINT in
OTG_FS_GINTSTS, respectively).
0: Port not in Suspend mode
1: Port in Suspend mode
Bit 6 PRES: Port resume
The application sets this bit to drive resume signaling on the port. The core continues to
drive the resume signal until the application clears this bit.
If the core detects a USB remote wakeup sequence, as indicated by the Port
resume/remote wakeup detected interrupt bit of the Core interrupt register (WKUINT bit in
OTG_FS_GINTSTS), the core starts driving resume signaling without application
intervention and clears this bit when it detects a disconnect condition. The read value of this
bit indicates whether the core is currently driving resume signaling.
0: No resume driven
1: Resume driven

RM0090 Rev 17

1297/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Bit 5 POCCHNG: Port overcurrent change
The core sets this bit when the status of the Port overcurrent active bit (bit 4) in this register
changes.
Bit 4 POCA: Port overcurrent active
Indicates the overcurrent condition of the port.
0: No overcurrent condition
1: Overcurrent condition
Bit 3 PENCHNG: Port enable/disable change
The core sets this bit when the status of the Port enable bit 2 in this register changes.
Bit 2 PENA: Port enable
A port is enabled only by the core after a reset sequence, and is disabled by an overcurrent
condition, a disconnect condition, or by the application clearing this bit. The application
cannot set this bit by a register write. It can only clear it to disable the port. This bit does not
trigger any interrupt to the application.
0: Port disabled
1: Port enabled
Bit 1 PCDET: Port connect detected
The core sets this bit when a device connection is detected to trigger an interrupt to the
application using the host port interrupt bit in the Core interrupt register (HPRTINT bit in
OTG_FS_GINTSTS). The application must write a 1 to this bit to clear the interrupt.
Bit 0 PCSTS: Port connect status
0: No device is attached to the port
1: A device is attached to the port

1298/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

OTG_FS Host channel-x characteristics register (OTG_FS_HCCHARx)
(x = 0..7, where x = Channel_number)
Address offset: 0x500 + 0x20 * x
Reset value: 0x0000 0000

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

EPDIR

rw

Reserved

ODDFRM

rs

MCNT

LSDEV

CHDIS

rs

DAD

EPTYP

CHENA

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

rw

9

8

7

6

EPNUM
rw

rw

rw

5

4

3

2

1

0

rw

rw

rw

rw

rw

MPSIZ
rw

rw

rw

rw

rw

rw

rw

Bit 31 CHENA: Channel enable
This field is set by the application and cleared by the OTG host.
0: Channel disabled
1: Channel enabled
Bit 30 CHDIS: Channel disable
The application sets this bit to stop transmitting/receiving data on a channel, even before
the transfer for that channel is complete. The application must wait for the Channel disabled
interrupt before treating the channel as disabled.
Bit 29 ODDFRM: Odd frame
This field is set (reset) by the application to indicate that the OTG host must perform a
transfer in an odd frame. This field is applicable for only periodic (isochronous and interrupt)
transactions.
0: Even frame
1: Odd frame
Bits 28:22 DAD: Device address
This field selects the specific device serving as the data source or sink.
Bits 21:20 MCNT: Multicount
This field indicates to the host the number of transactions that must be executed per frame
for this periodic endpoint. For non-periodic transfers, this field is not used
00: Reserved. This field yields undefined results
01: 1 transaction
10: 2 transactions per frame to be issued for this endpoint
11: 3 transactions per frame to be issued for this endpoint
Note: This field must be set to at least 01.
Bits 19:18 EPTYP: Endpoint type
Indicates the transfer type selected.
00: Control
01: Isochronous
10: Bulk
11: Interrupt
Bit 17 LSDEV: Low-speed device
This field is set by the application to indicate that this channel is communicating to a lowspeed device.
Bit 16 Reserved, must be kept at reset value.

RM0090 Rev 17

1299/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Bit 15 EPDIR: Endpoint direction
Indicates whether the transaction is IN or OUT.
0: OUT
1: IN
Bits 14:11 EPNUM: Endpoint number
Indicates the endpoint number on the device serving as the data source or sink.
Bits 10:0 MPSIZ: Maximum packet size
Indicates the maximum packet size of the associated endpoint.

OTG_FS Host channel-x interrupt register (OTG_FS_HCINTx) (x = 0..7, where
x = Channel_number)
Address offset: 0x508 + 0x20 * x
Reset value: 0x0000 0000

Bit 10 DTERR: Data toggle error
Bit 9 FRMOR: Frame overrun
Bit 8 BBERR: Babble error
Bit 7 TXERR: Transaction error
Indicates one of the following errors occurred on the USB.
CRC check failure
Timeout
Bit stuff error
False EOP
Bit 6 Reserved, must be kept at reset value.
Bit 5 ACK: ACK response received/transmitted interrupt
Bit 4 NAK: NAK response received interrupt
Bit 3 STALL: STALL response received interrupt

1300/1747

RM0090 Rev 17

3

rc_ rc_ rc_
w1 w1 w1

2

1

0

CHH

TXERR

Bits 31:11 Reserved, must be kept at reset value.

4

XFRC

BBERR

rc_ rc_ rc_ rc_
w1 w1 w1 w1

5

Reserved

FRMOR

Reserved

6

NAK

7

STALL

8

ACK

9

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
DTERR

This register indicates the status of a channel with respect to USB- and AHB-related events.
It is shown in Figure 394. The application must read this register when the host channels
interrupt bit in the Core interrupt register (HCINT bit in OTG_FS_GINTSTS) is set. Before
the application can read this register, it must first read the host all channels interrupt
(OTG_FS_HAINT) register to get the exact channel number for the host channel-x interrupt
register. The application must clear the appropriate bit in this register to clear the
corresponding bits in the OTG_FS_HAINT and OTG_FS_GINTSTS registers.

rc_ rc_
w1 w1

RM0090

USB on-the-go full-speed (OTG_FS)

Bit 2 Reserved, must be kept at reset value.
Bit 1 CHH: Channel halted
Indicates the transfer completed abnormally either because of any USB transaction error or
in response to disable request by the application.
Bit 0 XFRC: Transfer completed
Transfer completed normally without any errors.

OTG_FS Host channel-x interrupt mask register (OTG_FS_HCINTMSKx)
(x = 0..7, where x = Channel_number)
Address offset: 0x50C + 0x20 * x
Reset value: 0x0000 0000

rw

3

rw

rw

rw

2

1

0

CHHM

TXERRM

rw

4

XFRCM

BBERRM

rw

5

Reserved

FRMORM

rw

Reserved

6

NAKM

7

STALLM

8

ACKM

9

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
DTERRM

This register reflects the mask for each channel status described in the previous section.

rw

rw

Bits 31:11 Reserved, must be kept at reset value.
Bit 10 DTERRM: Data toggle error mask
0: Masked interrupt
1: Unmasked interrupt
Bit 9 FRMORM: Frame overrun mask
0: Masked interrupt
1: Unmasked interrupt
Bit 8 BBERRM: Babble error mask
0: Masked interrupt
1: Unmasked interrupt
Bit 7 TXERRM: Transaction error mask
0: Masked interrupt
1: Unmasked interrupt
Bit 6 Reserved, must be kept at reset value.
Bit 5 ACKM: ACK response received/transmitted interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 4 NAKM: NAK response received interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 3 STALLM: STALL response received interrupt mask
0: Masked interrupt
1: Unmasked interrupt

RM0090 Rev 17

1301/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Bit 2 Reserved, must be kept at reset value.
Bit 1 CHHM: Channel halted mask
0: Masked interrupt
1: Unmasked interrupt
Bit 0 XFRCM: Transfer completed mask
0: Masked interrupt
1: Unmasked interrupt

OTG_FS Host channel-x transfer size register (OTG_FS_HCTSIZx) (x = 0..7,
where x = Channel_number)
Address offset: 0x510 + 0x20 * x
Reset value: 0x0000 0000

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

rw

DPID
rw

PKTCNT

rw

rw

rw

rw

rw

rw

rw

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

XFRSIZ
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bit 31 Reserved, must be kept at reset value.
Bits 30:29 DPID: Data PID
The application programs this field with the type of PID to use for the initial transaction. The
host maintains this field for the rest of the transfer.
00: DATA0
01: DATA2
10: DATA1
11: MDATA (non-control)/SETUP (control)
Bits 28:19 PKTCNT: Packet count
This field is programmed by the application with the expected number of packets to be
transmitted (OUT) or received (IN).
The host decrements this count on every successful transmission or reception of an OUT/IN
packet. Once this count reaches zero, the application is interrupted to indicate normal
completion.
Bits 18:0 XFRSIZ: Transfer size
For an OUT, this field is the number of data bytes the host sends during the transfer.
For an IN, this field is the buffer size that the application has reserved for the transfer. The
application is expected to program this field as an integer multiple of the maximum packet
size for IN transactions (periodic and non-periodic).

1302/1747

RM0090 Rev 17

RM0090

34.16.4

USB on-the-go full-speed (OTG_FS)

Device-mode registers
OTG_FS device configuration register (OTG_FS_DCFG)
Address offset: 0x800
Reset value: 0x0220 0000

rw

rw

rw

6

5

4

rw

rw

rw

DAD

Reserved

7

rw

rw

3

2

rw

1

0
DSPD

8

NZLSOHSK

9

PFIVL

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

Reserved

This register configures the core in device mode after power-on or after certain control
commands or enumeration. Do not make changes to this register after initial programming.

rw

rw

Bits 31:13 Reserved, must be kept at reset value.
Bits 12:11 PFIVL: Periodic frame interval
Indicates the time within a frame at which the application must be notified using the end of
periodic frame interrupt. This can be used to determine if all the isochronous traffic for that
frame is complete.
00: 80% of the frame interval
01: 85% of the frame interval
10: 90% of the frame interval
11: 95% of the frame interval
Bits 10:4 DAD: Device address
The application must program this field after every SetAddress control command.
Bit 3 Reserved, must be kept at reset value.
Bit 2 NZLSOHSK: Non-zero-length status OUT handshake
The application can use this field to select the handshake the core sends on receiving a
nonzero-length data packet during the OUT transaction of a control transfer’s Status stage.
1: Send a STALL handshake on a nonzero-length status OUT transaction and do not send
the received OUT packet to the application.
0: Send the received OUT packet to the application (zero-length or nonzero-length) and
send a handshake based on the NAK and STALL bits for the endpoint in the Device
endpoint control register.
Bits 1:0 DSPD: Device speed
Indicates the speed at which the application requires the core to enumerate, or the
maximum speed the application can support. However, the actual bus speed is determined
only after the chirp sequence is completed, and is based on the speed of the USB host to
which the core is connected.
00: Reserved
01: Reserved
10: Reserved
11: Full speed (USB 1.1 transceiver clock is 48 MHz)

RM0090 Rev 17

1303/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

OTG_FS device control register (OTG_FS_DCTL)
Address offset: 0x804

w

w

rw

rw

rw

3

2

1

0

SDIS

w

4

RWUSIG

w

5

GINSTS

SGINAK

rw

6

TCTL

CGINAK

7

SGONAK

8

CGONAK

Reserved

9

POPRGDNE

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

GONSTS

Reset value: 0x0000 0000

r

r

rw

rw

Bits 31:12 Reserved, must be kept at reset value.
Bit 11 POPRGDNE: Power-on programming done
The application uses this bit to indicate that register programming is completed after a
wakeup from power down mode.
Bit 10 CGONAK: Clear global OUT NAK
Writing 1 to this field clears the Global OUT NAK.
Bit 9 SGONAK: Set global OUT NAK
Writing 1 to this field sets the Global OUT NAK.
The application uses this bit to send a NAK handshake on all OUT endpoints.
The application must set the this bit only after making sure that the Global OUT NAK
effective bit in the Core interrupt register (GONAKEFF bit in OTG_FS_GINTSTS) is cleared.
Bit 8 CGINAK: Clear global IN NAK
Writing 1 to this field clears the Global IN NAK.
Bit 7 SGINAK: Set global IN NAK
Writing 1 to this field sets the Global non-periodic IN NAK. The application uses this bit to
send a NAK handshake on all non-periodic IN endpoints.
The application must set this bit only after making sure that the Global IN NAK effective bit
in the Core interrupt register (GINAKEFF bit in OTG_FS_GINTSTS) is cleared.
Bits 6:4 TCTL: Test control
000: Test mode disabled
001: Test_J mode
010: Test_K mode
011: Test_SE0_NAK mode
100: Test_Packet mode
101: Test_Force_Enable
Others: Reserved
Bit 3 GONSTS: Global OUT NAK status
0: A handshake is sent based on the FIFO Status and the NAK and STALL bit settings.
1: No data is written to the RxFIFO, irrespective of space availability. Sends a NAK
handshake on all packets, except on SETUP transactions. All isochronous OUT packets are
dropped.

1304/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

Bit 2 GINSTS: Global IN NAK status
0: A handshake is sent out based on the data availability in the transmit FIFO.
1: A NAK handshake is sent out on all non-periodic IN endpoints, irrespective of the data
availability in the transmit FIFO.
Bit 1 SDIS: Soft disconnect
The application uses this bit to signal the USB OTG core to perform a soft disconnect. As
long as this bit is set, the host does not see that the device is connected, and the device
does not receive signals on the USB. The core stays in the disconnected state until the
application clears this bit.
0: Normal operation. When this bit is cleared after a soft disconnect, the core generates a
device connect event to the USB host. When the device is reconnected, the USB host
restarts device enumeration.
1: The core generates a device disconnect event to the USB host.
Bit 0 RWUSIG: Remote wakeup signaling
When the application sets this bit, the core initiates remote signaling to wake up the USB
host. The application must set this bit to instruct the core to exit the Suspend state. As
specified in the USB 2.0 specification, the application must clear this bit 1 ms to 15 ms after
setting it.

Table 204 contains the minimum duration (according to device state) for which the Soft
disconnect (SDIS) bit must be set for the USB host to detect a device disconnect. To
accommodate clock jitter, it is recommended that the application add some extra delay to
the specified minimum duration.
Table 204. Minimum duration for soft disconnect
Operating speed

Device state

Minimum duration

Full speed

Suspended

1 ms + 2.5 µs

Full speed

Idle

2.5 µs

Full speed

Not Idle or Suspended (Performing transactions)

2.5 µs

OTG_FS device status register (OTG_FS_DSTS)
Address offset: 0x808
Reset value: 0x0000 0010
This register indicates the status of the core with respect to USB-related events. It must be
read on interrupts from the device all interrupts (OTG_FS_DAINT) register.

FNSOF

Reserved
r

r

r

r

r

r

r

r

7

6

5

Reserved
r

r

r

r

r

r

4

3

2

r

1

r

0
SUSPSTS

8

ENUMSPD

9

EERR

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

r

r

Bits 31:22 Reserved, must be kept at reset value.
Bits 21:8 FNSOF: Frame number of the received SOF
Bits 7:4 Reserved, must be kept at reset value.

RM0090 Rev 17

1305/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Bit 3 EERR: Erratic error
The core sets this bit to report any erratic errors.
Due to erratic errors, the OTG_FS controller goes into Suspended state and an interrupt is
generated to the application with Early suspend bit of the OTG_FS_GINTSTS register
(ESUSP bit in OTG_FS_GINTSTS). If the early suspend is asserted due to an erratic error,
the application can only perform a soft disconnect recover.
Bits 2:1 ENUMSPD: Enumerated speed
Indicates the speed at which the OTG_FS controller has come up after speed detection
through a chirp sequence.
01: Reserved
10: Reserved
11: Full speed (PHY clock is running at 48 MHz)
Others: reserved
Bit 0 SUSPSTS: Suspend status
In device mode, this bit is set as long as a Suspend condition is detected on the USB. The
core enters the Suspended state when there is no activity on the USB data lines for a period
of 3 ms. The core comes out of the suspend:
– When there is an activity on the USB data lines
– When the application writes to the Remote wakeup signaling bit in the OTG_FS_DCTL
register (RWUSIG bit in OTG_FS_DCTL).

OTG_FS device IN endpoint common interrupt mask register
(OTG_FS_DIEPMSK)
Address offset: 0x810
Reset value: 0x0000 0000

Bits 31:14 Reserved, must be kept at reset value.
Bit 13 NAKM: NAK interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bits 12:7 Reserved, must be kept at reset value.
Bit 6 INEPNEM: IN endpoint NAK effective mask
0: Masked interrupt
1: Unmasked interrupt

1306/1747

RM0090 Rev 17

6

5

4

3

2

1

0

Reserved

EPDM

XFRCM

7

TOM

rw

Reserved

8

ITTXFEMSK

Reserved

9

INEPNEM

NAKM

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

INEPNMM

This register works with each of the OTG_FS_DIEPINTx registers for all endpoints to
generate an interrupt per IN endpoint. The IN endpoint interrupt for a specific status in the
OTG_FS_DIEPINTx register can be masked by writing to the corresponding bit in this
register. Status bits are masked by default.

rw

rw

rw

rw

rw

rw

RM0090

USB on-the-go full-speed (OTG_FS)

Bit 5 INEPNMM: IN token received with EP mismatch mask
0: Masked interrupt
1: Unmasked interrupt
Bit 4 ITTXFEMSK: IN token received when TxFIFO empty mask
0: Masked interrupt
1: Unmasked interrupt
Bit 3 TOM: Timeout condition mask (Non-isochronous endpoints)
0: Masked interrupt
1: Unmasked interrupt
Bit 2 Reserved, must be kept at reset value.
Bit 1 EPDM: Endpoint disabled interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 0 XFRCM: Transfer completed interrupt mask
0: Masked interrupt
1: Unmasked interrupt

OTG_FS device OUT endpoint common interrupt mask register
(OTG_FS_DOEPMSK)
Address offset: 0x814
Reset value: 0x0000 0000

3

rw

rw

rw

2

1

0

EPDM

4

XFRCM

5

Reserved

6

STUPM

rw

7

OTEPDM

rw

Reserved

8

Reserved

rw

9

OUTPKTERRM

BERRM

Reserved

NAK

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

STSPHSRXM

This register works with each of the OTG_FS_DOEPINTx registers for all endpoints to
generate an interrupt per OUT endpoint. The OUT endpoint interrupt for a specific status in
the OTG_FS_DOEPINTx register can be masked by writing into the corresponding bit in this
register. Status bits are masked by default.

rw

rw

Bits 31:14 Reserved, must be kept at reset value.
Bit 13 NAKMSK: NAK interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 12 BERRM: Babble error interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bits 11:9 Reserved, must be kept at reset value.
Bit 8 OUTPKTERRM: Out packet error mask
0: Masked interrupt
1: Unmasked interrupt

RM0090 Rev 17

1307/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Bits 7:6 Reserved, must be kept at reset value.
Bit 5 STSPHSRXM: Status phase received for control write mask
0: Masked interrupt
1: Unmasked interrupt
Bit 4 OTEPDM: OUT token received when endpoint disabled mask
Applies to control OUT endpoints only.
0: Masked interrupt
1: Unmasked interrupt
Bit 3 STUPM: SETUP phase done mask
Applies to control endpoints only.
0: Masked interrupt
1: Unmasked interrupt
Bit 2 Reserved, must be kept at reset value.
Bit 1 EPDM: Endpoint disabled interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 0 XFRCM: Transfer completed interrupt mask
0: Masked interrupt
1: Unmasked interrupt

OTG_FS device all endpoints interrupt register (OTG_FS_DAINT)
Address offset: 0x818
Reset value: 0x0000 0000
When a significant event occurs on an endpoint, a OTG_FS_DAINT register interrupts the
application using the Device OUT endpoints interrupt bit or Device IN endpoints interrupt bit
of the OTG_FS_GINTSTS register (OEPINT or IEPINT in OTG_FS_GINTSTS,
respectively). There is one interrupt bit per endpoint, up to a maximum of 16 bits for OUT
endpoints and 16 bits for IN endpoints. For a bidirectional endpoint, the corresponding IN
and OUT interrupt bits are used. Bits in this register are set and cleared when the
application sets and clears bits in the corresponding Device Endpoint-x interrupt register
(OTG_FS_DIEPINTx/OTG_FS_DOEPINTx).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

OEPINT
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:16 OEPINT: OUT endpoint interrupt bits
One bit per OUT endpoint:
Bit 16 for OUT endpoint 0, bit 19 for OUT endpoint 3.
Bits 15:0 IEPINT: IN endpoint interrupt bits
One bit per IN endpoint:
Bit 0 for IN endpoint 0, bit 3 for endpoint 3.

1308/1747

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

IEPINT

RM0090 Rev 17

r

r

r

r

RM0090

USB on-the-go full-speed (OTG_FS)

OTG_FS all endpoints interrupt mask register (OTG_FS_DAINTMSK)
Address offset: 0x81C
Reset value: 0x0000 0000
The OTG_FS_DAINTMSK register works with the Device endpoint interrupt register to
interrupt the application when an event occurs on a device endpoint. However, the
OTG_FS_DAINT register bit corresponding to that interrupt is still set.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

OEPM
rw

rw

rw

rw

rw

rw

rw

rw

rw

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

IEPM
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:16 OEPM: OUT EP interrupt mask bits
One per OUT endpoint:
Bit 16 for OUT EP 0, bit 19 for OUT EP 3
0: Masked interrupt
1: Unmasked interrupt
Bits 15:0 IEPM: IN EP interrupt mask bits
One bit per IN endpoint:
Bit 0 for IN EP 0, bit 3 for IN EP 3
0: Masked interrupt
1: Unmasked interrupt

OTG_FS device VBUS discharge time register (OTG_FS_DVBUSDIS)
Address offset: 0x0828
Reset value: 0x0000 17D7
This register specifies the VBUS discharge time after VBUS pulsing during SRP.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

VBUSDT
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 VBUSDT: Device VBUS discharge time
Specifies the VBUS discharge time after VBUS pulsing during SRP. This value equals:
VBUS discharge time in PHY clocks / 1 024
Depending on your VBUS load, this value may need adjusting.

OTG_FS device VBUS pulsing time register (OTG_FS_DVBUSPULSE)
Address offset: 0x082C
Reset value: 0x0000 05B8
This register specifies the VBUS pulsing time during SRP.

RM0090 Rev 17

1309/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

rw

rw

rw

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

DVBUSP

Reserved

rw

rw

rw

rw

Bits 31:12 Reserved, must be kept at reset value.
Bits 11:0 DVBUSP: Device VBUS pulsing time
Specifies the VBUS pulsing time during SRP. This value equals:
VBUS pulsing time in PHY clocks / 1 024

OTG_FS device IN endpoint FIFO empty interrupt mask register:
(OTG_FS_DIEPEMPMSK)
Address offset: 0x834
Reset value: 0x0000 0000
This register is used to control the IN endpoint FIFO empty interrupt generation
(TXFE_OTG_FS_DIEPINTx).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

INEPTXFEM

Reserved

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 INEPTXFEM: IN EP Tx FIFO empty interrupt mask bits
These bits act as mask bits for OTG_FS_DIEPINTx.
TXFE interrupt one bit per IN endpoint:
Bit 0 for IN endpoint 0, bit 3 for IN endpoint 3
0: Masked interrupt
1: Unmasked interrupt

OTG_FS device control IN endpoint 0 control register (OTG_FS_DIEPCTL0)
Address offset: 0x900
Reset value: 0x0000 0000
This section describes the OTG_FS_DIEPCTL0 register. Nonzero control endpoints use
registers for endpoints 1–3.

1310/1747

rw

rw

rw

rs

r

r

r

USBAEP

Reserved

rw

EPTYP

NAKSTS

w

STALL

w

TXFNUM

Reserved

CNAK

r

SNAK

EPDIS

r

Reserved

EPENA

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

r

RM0090 Rev 17

9

8

7

Reserved

6

5

4

3

2

1

0

MPSIZ
rw

rw

RM0090

USB on-the-go full-speed (OTG_FS)

Bit 31 EPENA: Endpoint enable
The application sets this bit to start transmitting data on the endpoint 0.
The core clears this bit before setting any of the following interrupts on this endpoint:
– Endpoint disabled
– Transfer completed
Bit 30 EPDIS: Endpoint disable
The application sets this bit to stop transmitting data on an endpoint, even before the
transfer for that endpoint is complete. The application must wait for the Endpoint disabled
interrupt before treating the endpoint as disabled. The core clears this bit before setting the
Endpoint disabled interrupt. The application must set this bit only if Endpoint enable is
already set for this endpoint.
Bits 29:28 Reserved, must be kept at reset value.
Bit 27 SNAK: Set NAK
A write to this bit sets the NAK bit for the endpoint.
Using this bit, the application can control the transmission of NAK handshakes on an
endpoint. The core can also set this bit for an endpoint after a SETUP packet is received on
that endpoint.
Bit 26 CNAK: Clear NAK
A write to this bit clears the NAK bit for the endpoint.
Bits 25:22 TXFNUM: TxFIFO number
This value is set to the FIFO number that is assigned to IN endpoint 0.
Bit 21 STALL: STALL handshake
The application can only set this bit, and the core clears it when a SETUP token is received
for this endpoint. If a NAK bit, a Global IN NAK or Global OUT NAK is set along with this bit,
the STALL bit takes priority.
Bit 20 Reserved, must be kept at reset value.
Bits 19:18 EPTYP: Endpoint type
Hardcoded to ‘00’ for control.
Bit 17 NAKSTS: NAK status
Indicates the following:
0: The core is transmitting non-NAK handshakes based on the FIFO status
1: The core is transmitting NAK handshakes on this endpoint.
When this bit is set, either by the application or core, the core stops transmitting data, even
if there are data available in the TxFIFO. Irrespective of this bit’s setting, the core always
responds to SETUP data packets with an ACK handshake.
Bit 16 Reserved, must be kept at reset value.

RM0090 Rev 17

1311/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Bit 15 USBAEP: USB active endpoint
This bit is always set to 1, indicating that control endpoint 0 is always active in all
configurations and interfaces.
Bits 14:2 Reserved, must be kept at reset value.
Bits 1:0 MPSIZ: Maximum packet size
The application must program this field with the maximum packet size for the current logical
endpoint.
00: 64 bytes
01: 32 bytes
10: 16 bytes
11: 8 bytes

OTG device endpoint x control register (OTG_FS_DIEPCTLx) (x = 1..3, where
x = Endpoint_number)
Address offset: 0x900 + 0x20 * x
Reset value: 0x0000 0000
The application uses this register to control the behavior of each logical endpoint other than
endpoint 0.

rw

rw

rw

rw

rw

rw

USBAEP

rw

EONUM/DPID

w

NAKSTS

w

EPTYP

w

STALL

CNAK

w

TXFNUM

Reserved

SNAK

rs

SODDFRM

EPDIS

rs

SD0PID/SEVNFRM

EPENA

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

r

r

rw

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

MPSIZ

Reserved

rw

rw

rw

rw

rw

rw

Bit 31 EPENA: Endpoint enable
The application sets this bit to start transmitting data on an endpoint.
The core clears this bit before setting any of the following interrupts on this endpoint:
–
SETUP phase done
–
Endpoint disabled
–
Transfer completed
Bit 30 EPDIS: Endpoint disable
The application sets this bit to stop transmitting/receiving data on an endpoint, even before
the transfer for that endpoint is complete. The application must wait for the Endpoint
disabled interrupt before treating the endpoint as disabled. The core clears this bit before
setting the Endpoint disabled interrupt. The application must set this bit only if Endpoint
enable is already set for this endpoint.
Bit 29 SODDFRM: Set odd frame
Applies to isochronous IN and OUT endpoints only.
Writing to this field sets the Even/Odd frame (EONUM) field to odd frame.

1312/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

Bit 28 SD0PID: Set DATA0 PID
Applies to interrupt/bulk IN endpoints only.
Writing to this field sets the endpoint data PID (DPID) field in this register to DATA0.
SEVNFRM: Set even frame
Applies to isochronous IN endpoints only.
Writing to this field sets the Even/Odd frame (EONUM) field to even frame.
Bit 27 SNAK: Set NAK
A write to this bit sets the NAK bit for the endpoint.
Using this bit, the application can control the transmission of NAK handshakes on an
endpoint. The core can also set this bit for OUT endpoints on a Transfer completed
interrupt, or after a SETUP is received on the endpoint.
Bit 26 CNAK: Clear NAK
A write to this bit clears the NAK bit for the endpoint.
Bits 25:22 TXFNUM: TxFIFO number
These bits specify the FIFO number associated with this endpoint. Each active IN endpoint
must be programmed to a separate FIFO number.
This field is valid only for IN endpoints.
Bit 21 STALL: STALL handshake
Applies to non-control, non-isochronous IN endpoints only (access type is rw).
The application sets this bit to stall all tokens from the USB host to this endpoint. If a NAK
bit, Global IN NAK, or Global OUT NAK is set along with this bit, the STALL bit takes priority.
Only the application can clear this bit, never the core.
Bit 20 Reserved, must be kept at reset value.
Bits 19:18 EPTYP: Endpoint type
This is the transfer type supported by this logical endpoint.
00: Control
01: Isochronous
10: Bulk
11: Interrupt
Bit 17 NAKSTS: NAK status
It indicates the following:
0: The core is transmitting non-NAK handshakes based on the FIFO status.
1: The core is transmitting NAK handshakes on this endpoint.
When either the application or the core sets this bit:
For non-isochronous IN endpoints: The core stops transmitting any data on an IN endpoint,
even if there are data available in the TxFIFO.
For isochronous IN endpoints: The core sends out a zero-length data packet, even if there
are data available in the TxFIFO.
Irrespective of this bit’s setting, the core always responds to SETUP data packets with an
ACK handshake.

RM0090 Rev 17

1313/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Bit 16 EONUM: Even/odd frame
Applies to isochronous IN endpoints only.
Indicates the frame number in which the core transmits/receives isochronous data for this
endpoint. The application must program the even/odd frame number in which it intends to
transmit/receive isochronous data for this endpoint using the SEVNFRM and SODDFRM
fields in this register.
0: Even frame
1: Odd frame
DPID: Endpoint data PID
Applies to interrupt/bulk IN endpoints only.
Contains the PID of the packet to be received or transmitted on this endpoint. The
application must program the PID of the first packet to be received or transmitted on this
endpoint, after the endpoint is activated. The application uses the SD0PID register field to
program either DATA0 or DATA1 PID.
0: DATA0
1: DATA1
Bit 15 USBAEP: USB active endpoint
Indicates whether this endpoint is active in the current configuration and interface. The core
clears this bit for all endpoints (other than EP 0) after detecting a USB reset. After receiving
the SetConfiguration and SetInterface commands, the application must program endpoint
registers accordingly and set this bit.
Bits 14:11 Reserved, must be kept at reset value.
Bits 10:0 MPSIZ: Maximum packet size
The application must program this field with the maximum packet size for the current logical
endpoint. This value is in bytes.

1314/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

OTG_FS device control OUT endpoint 0 control register
(OTG_FS_DOEPCTL0)
Address offset: 0xB00
Reset value: 0x0000 8000
This section describes the OTG_FS_DOEPCTL0 register. Nonzero control endpoints use
registers for endpoints 1–3.

r

r

r

USBAEP

rw

Reserved

rs

EPTYP

NAKSTS

w

SNPM

w

Reserved

STALL

CNAK

r

SNAK

EPDIS

w

Reserved

EPENA

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

MPSIZ

Reserved

r

r

r

Bit 31 EPENA: Endpoint enable
The application sets this bit to start transmitting data on endpoint 0.
The core clears this bit before setting any of the following interrupts on this endpoint:

–

SETUP phase done

–

Endpoint disabled

–

Transfer completed

Bit 30 EPDIS: Endpoint disable
The application cannot disable control OUT endpoint 0.
Bits 29:28 Reserved, must be kept at reset value.
Bit 27 SNAK: Set NAK
A write to this bit sets the NAK bit for the endpoint.
Using this bit, the application can control the transmission of NAK handshakes on an
endpoint. The core can also set this bit on a Transfer completed interrupt, or after a SETUP
is received on the endpoint.
Bit 26 CNAK: Clear NAK
A write to this bit clears the NAK bit for the endpoint.
Bits 25:22 Reserved, must be kept at reset value.
Bit 21 STALL: STALL handshake
The application can only set this bit, and the core clears it, when a SETUP token is received
for this endpoint. If a NAK bit or Global OUT NAK is set along with this bit, the STALL bit
takes priority. Irrespective of this bit’s setting, the core always responds to SETUP data
packets with an ACK handshake.
Bit 20 SNPM: Snoop mode
This bit configures the endpoint to Snoop mode. In Snoop mode, the core does not check
the correctness of OUT packets before transferring them to application memory.
Bits 19:18 EPTYP: Endpoint type
Hardcoded to 2’b00 for control.

RM0090 Rev 17

1315/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Bit 17 NAKSTS: NAK status
Indicates the following:
0: The core is transmitting non-NAK handshakes based on the FIFO status.
1: The core is transmitting NAK handshakes on this endpoint.
When either the application or the core sets this bit, the core stops receiving data, even if
there is space in the RxFIFO to accommodate the incoming packet. Irrespective of this bit’s
setting, the core always responds to SETUP data packets with an ACK handshake.
Bit 16 Reserved, must be kept at reset value.
Bit 15 USBAEP: USB active endpoint
This bit is always set to 1, indicating that a control endpoint 0 is always active in all
configurations and interfaces.
Bits 14:2 Reserved, must be kept at reset value.
Bits 1:0 MPSIZ: Maximum packet size
The maximum packet size for control OUT endpoint 0 is the same as what is programmed in
control IN endpoint 0.
00: 64 bytes
01: 32 bytes
10: 16 bytes
11: 8 bytes

OTG_FS device endpoint-x control register (OTG_FS_DOEPCTLx) (x = 1..3,
where x = Endpoint_number)
Address offset for OUT endpoints: 0xB00 + 0x20 * x
Reset value: 0x0000 0000
The application uses this register to control the behavior of each logical endpoint other than
endpoint 0.

w

rw

rw

rw

rw

USBAEP

CNAK

w

EONUM/DPID

SNAK

w

NAKSTS

SD0PID/SEVNFRM

w

EPTYP

SODDFRM/SD1PID

rs

SNPM

EPDIS

rs

Reserved

STALL

EPENA

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

r

r

rw

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

MPSIZ

Reserved

rw

rw

rw

rw

rw

rw

Bit 31 EPENA: Endpoint enable
Applies to IN and OUT endpoints.
The application sets this bit to start transmitting data on an endpoint.
The core clears this bit before setting any of the following interrupts on this endpoint:
–
SETUP phase done
–
Endpoint disabled
–
Transfer completed

1316/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

Bit 30 EPDIS: Endpoint disable
The application sets this bit to stop transmitting/receiving data on an endpoint, even before
the transfer for that endpoint is complete. The application must wait for the Endpoint
disabled interrupt before treating the endpoint as disabled. The core clears this bit before
setting the Endpoint disabled interrupt. The application must set this bit only if Endpoint
enable is already set for this endpoint.
Bit 29 SD1PID: Set DATA1 PID
Applies to interrupt/bulk IN and OUT endpoints only. Writing to this field sets the endpoint
data PID (DPID) field in this register to DATA1.
SODDFRM: Set odd frame
Applies to isochronous IN and OUT endpoints only. Writing to this field sets the Even/Odd
frame (EONUM) field to odd frame.
Bit 28 SD0PID: Set DATA0 PID
Applies to interrupt/bulk OUT endpoints only.
Writing to this field sets the endpoint data PID (DPID) field in this register to DATA0.
SEVNFRM: Set even frame
Applies to isochronous OUT endpoints only.
Writing to this field sets the Even/Odd frame (EONUM) field to even frame.
Bit 27 SNAK: Set NAK
A write to this bit sets the NAK bit for the endpoint.
Using this bit, the application can control the transmission of NAK handshakes on an
endpoint. The core can also set this bit for OUT endpoints on a Transfer Completed
interrupt, or after a SETUP is received on the endpoint.
Bit 26 CNAK: Clear NAK
A write to this bit clears the NAK bit for the endpoint.
Bits 25:22 Reserved, must be kept at reset value.
Bit 21 STALL: STALL handshake
Applies to non-control, non-isochronous OUT endpoints only (access type is rw).
The application sets this bit to stall all tokens from the USB host to this endpoint. If a NAK
bit, Global IN NAK, or Global OUT NAK is set along with this bit, the STALL bit takes
priority. Only the application can clear this bit, never the core.
Bit 20 SNPM: Snoop mode
This bit configures the endpoint to Snoop mode. In Snoop mode, the core does not check
the correctness of OUT packets before transferring them to application memory.
Bits 19:18 EPTYP: Endpoint type
This is the transfer type supported by this logical endpoint.
00: Control
01: Isochronous
10: Bulk
11: Interrupt

RM0090 Rev 17

1317/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Bit 17 NAKSTS: NAK status
Indicates the following:
0: The core is transmitting non-NAK handshakes based on the FIFO status.
1: The core is transmitting NAK handshakes on this endpoint.
When either the application or the core sets this bit:
The core stops receiving any data on an OUT endpoint, even if there is space in the
RxFIFO to accommodate the incoming packet.
Irrespective of this bit’s setting, the core always responds to SETUP data packets with an
ACK handshake.
Bit 16 EONUM: Even/odd frame
Applies to isochronous IN and OUT endpoints only.
Indicates the frame number in which the core transmits/receives isochronous data for this
endpoint. The application must program the even/odd frame number in which it intends to
transmit/receive isochronous data for this endpoint using the SEVNFRM and SODDFRM
fields in this register.
0: Even frame
1: Odd frame
DPID: Endpoint data PID
Applies to interrupt/bulk OUT endpoints only.
Contains the PID of the packet to be received or transmitted on this endpoint. The
application must program the PID of the first packet to be received or transmitted on this
endpoint, after the endpoint is activated. The application uses the SD0PID register field to
program either DATA0 or DATA1 PID.
0: DATA0
1: DATA1
Bit 15 USBAEP: USB active endpoint
Indicates whether this endpoint is active in the current configuration and interface. The core
clears this bit for all endpoints (other than EP 0) after detecting a USB reset. After receiving
the SetConfiguration and SetInterface commands, the application must program endpoint
registers accordingly and set this bit.
Bits 14:11 Reserved, must be kept at reset value.
Bits 10:0 MPSIZ: Maximum packet size
The application must program this field with the maximum packet size for the current logical
endpoint. This value is in bytes.

1318/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

OTG_FS device endpoint-x interrupt register (OTG_FS_DIEPINTx) (x = 0..3,
where x = Endpoint_number)
Address offset: 0x908 + 0x20 * x
Reset value: 0x0000 0080

r

4

3

2

Reserved

rc_ rc_ rc_ rc_
w1 w1 w1 w1

1

0
XFRC

5

EPDISD

6

TOC

Reserved

7
TXFE

8

ITTXFE

rc_
w1

9

INEPNE

rc_
w1

PKTDRPSTS

Reserved

Reserved

NAK

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

INEPNM

This register indicates the status of an endpoint with respect to USB- and AHB-related
events. It is shown in Figure 394. The application must read this register when the IN
endpoints interrupt bit of the Core interrupt register (IEPINT in OTG_FS_GINTSTS) is set.
Before the application can read this register, it must first read the device all endpoints
interrupt (OTG_FS_DAINT) register to get the exact endpoint number for the Device
endpoint-x interrupt register. The application must clear the appropriate bit in this register to
clear the corresponding bits in the OTG_FS_DAINT and OTG_FS_GINTSTS registers.

rc_ rc_
w1 w1

Bits 31:14 Reserved, must be kept at reset value.
Bit 13 NAK: NAK input
The core generates this interrupt when a NAK is transmitted or received by the device. In
case of isochronous IN endpoints the interrupt gets generated when a zero length packet is
transmitted due to unavailability of data in the Tx FIFO.
Bit 12 Reserved, must be kept at reset value.
Bit 11 PKTDRPSTS: Packet dropped status
This bit indicates to the application that an ISOC OUT packet has been dropped. This bit
does not have an associated mask bit and does not generate an interrupt.
Bits 10:8 Reserved, must be kept at reset value.
Bit 7 TXFE: Transmit FIFO empty
This interrupt is asserted when the TxFIFO for this endpoint is either half or completely
empty. The half or completely empty status is determined by the TxFIFO Empty Level bit in
the OTG_FS_GAHBCFG register (TXFELVL bit in OTG_FS_GAHBCFG).
Bit 6 INEPNE: IN endpoint NAK effective
This bit can be cleared when the application clears the IN endpoint NAK by writing to the
CNAK bit in OTG_FS_DIEPCTLx.
This interrupt indicates that the core has sampled the NAK bit set (either by the application
or by the core). The interrupt indicates that the IN endpoint NAK bit set by the application
has taken effect in the core.
This interrupt does not guarantee that a NAK handshake is sent on the USB. A STALL bit
takes priority over a NAK bit.
Bit 5 INEPNM: IN token received with EP mismatch.
Indicates that the data in the top of the non-periodic TxFIFO belongs to an endpoint other
than the one for which the IN token was received. This interrupt is asserted on the endpoint
for which the IN token was received.

RM0090 Rev 17

1319/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Bit 4 ITTXFE: IN token received when TxFIFO is empty
Applies to non-periodic IN endpoints only.
Indicates that an IN token was received when the associated TxFIFO (periodic/non-periodic)
was empty. This interrupt is asserted on the endpoint for which the IN token was received.
Bit 3 TOC: Timeout condition
Applies only to Control IN endpoints.
Indicates that the core has detected a timeout condition on the USB for the last IN token on
this endpoint.
Bit 2 Reserved, must be kept at reset value.
Bit 1 EPDISD: Endpoint disabled interrupt
This bit indicates that the endpoint is disabled per the application’s request.
Bit 0 XFRC: Transfer completed interrupt
This field indicates that the programmed transfer is complete on the AHB as well as on the
USB, for this endpoint.

OTG_FS device endpoint-x interrupt register (OTG_FS_DOEPINTx) (x = 0..3,
where x = Endpoint_number)
Address offset: 0xB08 + 0x20 * x
Reset value: 0x0000 0080

0
XFRC

STUP

1
EPDISD

OTEPDIS

2

rc_w1

3

rc_w1

4

Reserved

5

rc_w1

6

rc_w1

7

STSPHSRX

8

Reserved

9

rc_w1 OUTPKTERR

BERR
rc_w1

Reserved

Reserved

NAK
rc_w1

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

rc_w1

This register indicates the status of an endpoint with respect to USB- and AHB-related
events. It is shown in Figure 394. The application must read this register when the OUT
Endpoints Interrupt bit of the OTG_FS_GINTSTS register (OEPINT bit in
OTG_FS_GINTSTS) is set. Before the application can read this register, it must first read
the OTG_FS_DAINT register to get the exact endpoint number for the OTG_FS_DOEPINTx
register. The application must clear the appropriate bit in this register to clear the
corresponding bits in the OTG_FS_DAINT and OTG_FS_GINTSTS registers.

Bits 31:14 Reserved, must be kept at reset value.
Bit 13 NAK: NAK input
The core generates this interrupt when a NAK is transmitted or received by the device. In
case of isochronous IN endpoints the interrupt gets generated when a zero length packet is
transmitted due to unavailability of data in the Tx FIFO.
Bit 12 BERR: Babble error interrupt
The core generates this interrupt when babble is received for the endpoint.
Bits 11:9 Reserved, must be kept at reset value.

1320/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

Bit 8 OUTPKTERR: OUT packet error
This interrupt is asserted when the core detects an overflow or a CRC error for an OUT
packet. This interrupt is valid only when thresholding is enabled.
Bits 7:6 Reserved, must be kept at reset value.
Bit 5 STSPHSRX: Status phase received for control write
This interrupt is generated only after the core has transferred all the data that the host has
sent during the data phase of a control write transfer, to the system memory buffer. The
interrupt indicates to the application that the host has switched from data phase to the status
phase of a control write transfer. The application can use this interrupt to ACK or STALL the
status phase, after it has decoded the data phase.
Bit 4 OTEPDIS: OUT token received when endpoint disabled
Applies only to control OUT endpoints.
Indicates that an OUT token was received when the endpoint was not yet enabled. This
interrupt is asserted on the endpoint for which the OUT token was received.
Bit 3 STUP: SETUP phase done
Applies to control OUT endpoint only.
Indicates that the SETUP phase for the control endpoint is complete and no more back-toback SETUP packets were received for the current control transfer. On this interrupt, the
application can decode the received SETUP data packet.
Bit 2 Reserved, must be kept at reset value.
Bit 1 EPDISD: Endpoint disabled interrupt
This bit indicates that the endpoint is disabled per the application’s request.
Bit 0 XFRC: Transfer completed interrupt
This field indicates that the programmed transfer is complete on the AHB as well as on the
USB, for this endpoint.

OTG_FS device IN endpoint 0 transfer size register (OTG_FS_DIEPTSIZ0)
Address offset: 0x910
Reset value: 0x0000 0000
The application must modify this register before enabling endpoint 0. Once endpoint 0 is
enabled using the endpoint enable bit in the device control endpoint 0 control registers
(EPENA in OTG_FS_DIEPCTL0), the core modifies this register. The application can only
read this register once the core has cleared the Endpoint enable bit.
Nonzero endpoints use the registers for endpoints 1–3.
31 30 29 28 27 26 25 24 23 22 21
Reserved

20

19

PKTCNT
rw

rw

18 17 16 15 14 13 12 11 10
Reserved

RM0090 Rev 17

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

XFRSIZ
rw

rw

rw

rw

1321/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Bits 31:21 Reserved, must be kept at reset value.
Bits 20:19 PKTCNT: Packet count
Indicates the total number of USB packets that constitute the Transfer Size amount of data
for endpoint 0.
This field is decremented every time a packet (maximum size or short packet) is read from
the TxFIFO.
Bits 18:7 Reserved, must be kept at reset value.
Bits 6:0 XFRSIZ: Transfer size
Indicates the transfer size in bytes for endpoint 0. The core interrupts the application only
after it has exhausted the transfer size amount of data. The transfer size can be set to the
maximum packet size of the endpoint, to be interrupted at the end of each packet.
The core decrements this field every time a packet from the external memory is written to
the TxFIFO.

1322/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

OTG_FS device OUT endpoint 0 transfer size register (OTG_FS_DOEPTSIZ0)
Address offset: 0xB10
Reset value: 0x0000 0000
The application must modify this register before enabling endpoint 0. Once endpoint 0 is
enabled using the Endpoint enable bit in the OTG_FS_DOEPCTL0 registers (EPENA bit in
OTG_FS_DOEPCTL0), the core modifies this register. The application can only read this
register once the core has cleared the Endpoint enable bit.
Nonzero endpoints use the registers for endpoints 1–3.
30

29

28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

STUPCNT
rw

rw

Reserved

PKTCNT

Reserved

31

9

8

7

6

5

4

2

1

0

rw

rw

rw

XFRSIZ

Reserved

rw

3

rw

rw

rw

rw

Bit 31 Reserved, must be kept at reset value.
Bits 30:29 STUPCNT: SETUP packet count
This field specifies the number of back-to-back SETUP data packets the endpoint can
receive.
01: 1 packet
10: 2 packets
11: 3 packets
Bits 28:20 Reserved, must be kept at reset value.
Bit 19 PKTCNT: Packet count
This field is decremented to zero after a packet is written into the RxFIFO.
Bits 18:7 Reserved, must be kept at reset value.
Bits 6:0 XFRSIZ: Transfer size
Indicates the transfer size in bytes for endpoint 0. The core interrupts the application only
after it has exhausted the transfer size amount of data. The transfer size can be set to the
maximum packet size of the endpoint, to be interrupted at the end of each packet.
The core decrements this field every time a packet is read from the RxFIFO and written to
the external memory.

RM0090 Rev 17

1323/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

OTG_FS device endpoint-x transfer size register (OTG_FS_DIEPTSIZx)
(x = 1..3, where x = Endpoint_number)
Address offset: 0x910 + 0x20 * x
Reset value: 0x0000 0000
The application must modify this register before enabling the endpoint. Once the endpoint is
enabled using the Endpoint enable bit in the OTG_FS_DIEPCTLx registers (EPENA bit in
OTG_FS_DIEPCTLx), the core modifies this register. The application can only read this
register once the core has cleared the Endpoint enable bit.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
PKTCNT

Reserved

rw

rw

rw

rw

rw

rw

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

XFRSIZ
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:29 Reserved, must be kept at reset value.
Bit 28:19 PKTCNT: Packet count
Indicates the total number of USB packets that constitute the Transfer Size amount of data
for this endpoint.
This field is decremented every time a packet (maximum size or short packet) is read from
the TxFIFO.
Bits 18:0 XFRSIZ: Transfer size
This field contains the transfer size in bytes for the current endpoint. The core only interrupts
the application after it has exhausted the transfer size amount of data. The transfer size can
be set to the maximum packet size of the endpoint, to be interrupted at the end of each
packet.
The core decrements this field every time a packet from the external memory is written to
the TxFIFO.

1324/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)

OTG_FS device IN endpoint transmit FIFO status register
(OTG_FS_DTXFSTSx) (x = 0..3, where x = Endpoint_number)
Address offset for IN endpoints: 0x918 + 0x20 * x
This read-only register contains the free space information for the Device IN endpoint
TxFIFO.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

INEPTFSAV

Reserved

r

r

r

r

r

r

r

r

r

31:16 Reserved, must be kept at reset value.
15:0 INEPTFSAV: IN endpoint TxFIFO space available
Indicates the amount of free space available in the Endpoint TxFIFO.
Values are in terms of 32-bit words:
0x0: Endpoint TxFIFO is full
0x1: 1 word available
0x2: 2 words available
0xn: n words available
Others: Reserved

OTG_FS device OUT endpoint-x transfer size register (OTG_FS_DOEPTSIZx)
(x = 1..3, where x = Endpoint_number)
Address offset: 0xB10 + 0x20 * x
Reset value: 0x0000 0000
The application must modify this register before enabling the endpoint. Once the endpoint is
enabled using Endpoint Enable bit of the OTG_FS_DOEPCTLx registers (EPENA bit in
OTG_FS_DOEPCTLx), the core modifies this register. The application can only read this
register once the core has cleared the Endpoint enable bit.

Reserved

31

30

29

RXDPID/S
TUPCNT
r/rw

28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
PKTCNT

9

8

7

6

5

4

3

2

1

0

XFRSIZ

r/rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 31 Reserved, must be kept at reset value.
Bits 30:29 RXDPID: Received data PID
Applies to isochronous OUT endpoints only.
This is the data PID received in the last packet for this endpoint.
00: DATA0
01: DATA2
10: DATA1
11: MDATA

RM0090 Rev 17

1325/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

STUPCNT: SETUP packet count
Applies to control OUT Endpoints only.
This field specifies the number of back-to-back SETUP data packets the endpoint can
receive.
01: 1 packet
10: 2 packets
11: 3 packets
Bit 28:19 PKTCNT: Packet count
Indicates the total number of USB packets that constitute the Transfer Size amount of data
for this endpoint.
This field is decremented every time a packet (maximum size or short packet) is written to
the RxFIFO.
Bits 18:0 XFRSIZ: Transfer size
This field contains the transfer size in bytes for the current endpoint. The core only interrupts
the application after it has exhausted the transfer size amount of data. The transfer size can
be set to the maximum packet size of the endpoint, to be interrupted at the end of each
packet.
The core decrements this field every time a packet is read from the RxFIFO and written to
the external memory.

34.16.5

OTG_FS power and clock gating control register
(OTG_FS_PCGCCTL)
Address offset: 0xE00
Reset value: 0x0000 0000

28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

Reserved

9

8

7

6

5

4

rw

3

2

1

0
STPPCLK

29

Reserved

30

PHYSUSP

31

GATEHCLK

This register is available in host and device modes.

rw rw

Bit 31:5 Reserved, must be kept at reset value.
Bit 4 PHYSUSP: PHY Suspended
Indicates that the PHY has been suspended. This bit is updated once the PHY is suspended
after the application has set the STPPCLK bit (bit 0).
Bits 3:2 Reserved, must be kept at reset value.
Bit 1 GATEHCLK: Gate HCLK
The application sets this bit to gate HCLK to modules other than the AHB Slave and Master
and wakeup logic when the USB is suspended or the session is not valid. The application
clears this bit when the USB is resumed or a new session starts.
Bit 0 STPPCLK: Stop PHY clock
The application sets this bit to stop the PHY clock when the USB is suspended, the session
is not valid, or the device is disconnected. The application clears this bit when the USB is
resumed or a new session starts.

1326/1747

RM0090 Rev 17

RM0090

34.16.6

USB on-the-go full-speed (OTG_FS)

OTG_FS register map
The table below gives the USB OTG register map and reset values.

OTG_FS_GRXS
TSR (host
mode)
0x01C

PKTSTS

Reserved

Reset value
OTG_FS_GRXS
TSR (Device
mode)
Reset value

0

Reserved

FRMNUM
0

0

0

0

0

0

0

0

ESUSP
0

USBRST

USBSUSPM

ESUSPM

0

0

0

0

0

0

0

0

0

0

0

0

DPID
0

RM0090 Rev 17

0

0

0

0

0

0

0

0

SEDET

SRQ

SRQSCS
GINTMSK

FCRST

CSRST

0

1

0

0

0

0

0

0

0

0

0

0

0

0

CHNUM
0

0

0

0

0

0

BCNT
0

HSRST

0

BCNT
0

0

CMOD

USBRST

USBSUSP
0

0

Reserved

ENUMDNE

0

DPID

PKTSTS
0

0

ENUMDNEM

0

EOPF

IEPINT

0

ISOODRP

OEPINT

0

0

EOPFM

IEPINT

IISOIXFRM

0

0

ISOODRPM

OEPINT

0

Reserved

IISOIXFR

0

Reserved

IPXFR/INCOMPISOOUT

0

IPXFRM/IISOOXFRM

0

Reserved

HCINT

HPRTINT
PRTIM

0

Reserved

PTXFE

HCIM

Reserved

PTXFEM

Reserved

0

0

0

MMIS

0

0

OTGINT

0

0

MMISM

0

0

OTGINT

0

0

0

SOF

DISCINT

CIDSCHGM

Reset value

0

0

0

SOFM

OTG_FS_GINT
MSK

1

TXFFLSH

0

0

0

RXFLVL

0

0

TXFNUM

Reserved

0

NPTXFE

0

0

RXFLVLM

0

0

NPTXFEM

DISCINT

CIDSCHG

Reset value

1

GINAKEFF

OTG_FS_GINTS
TS

0

TOCAL

GONAKEFF

1

1

SRPCAP

AHBIDL

Reset value

0

HNPCAP

FHMOD

OTG_FS_GRST
CTL

0

Reserved

FDMOD

0

TRDT

Reserved

Reserved

CTXPKT

0

SRQINT

0x018

0

WKUINT

0x014

Reset value

WUIM

0x010

OTG_FS_GUSB
CFG

SRQIM

0x00C

0
Reserved

0

Reserved

Reserved

0

Reset value

Res.

0

RXFFLSH

Reserved

0

GINAKEFFM

OTG_FS_GAHB
CFG

0

Reserved

0

GONAKEFFM

0x008

0

Reserved

TXFELVL

0

0

PHYSEL

0

0

Reserved

0

Reset value

HNPRQ

Reserved

HNGSCS

OTG_FS_GOTG
INT

0

SRSSCHG

0x004

0

HNSSCHG

1

Reserved

PTXFELVL

0

DHNPEN

DBCT

CIDSTS

0

Reset value

Reserved

HSHNPEN

ASVLD

0

HNGDET

Reserved

BSVLD

OTG_FS_GOTG
CTL

DBCDNE

0x000

Register

ADTOCHG

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 205. OTG_FS register map and reset values

0

0

EPNUM
0

0

0

0

0

0

0

0

0

1327/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

OTG_FS_GRXS
TSR (host
mode)
0x020

Reset value

0

OTG_FS_GRXS
TSPR (Device
mode)

FRMNUM

Reserved

Reset value

0x024

PKTSTS

Reserved

0

OTG_FS_GRXF
SIZ

0

0

0

0

DPID
0

PKTSTS

0

0

0

0

0

0

0

0x100

0

0

0

0

0

NPTXQTOP
0

0

0

0

0

0x104

0

0

0

0x10C

0x400

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0x408

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

1

1

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

1

0

0

0

INEPTXSA
0

0

0

0

0

0

0

0

0

0

0

1

INEPTXFD
0

0

INEPTXSA

INEPTXFD
0

0

PTXSA

INEPTXFD

OTG_FS_DIEPT
XF3

0

0

0

0

INEPTXSA
0

0

0

OTG_FS_HCFG

0

0

0

0

0

0

0

0

1

0

0

0

0

Reserved

OTG_FS_HFIR

OTG_FS_HFNU
M

1328/1747

1

1

1

0

1

0

1

FTREM
0

0

0

0

0

0

0

0

0

0

0

0

FRIVL

Reserved

Reset value

Reset value

0

EPNUM

Reset value
0x404

0

NPTXFSAV

PTXFSIZ

OTG_FS_DIEPT
XF2

Reset value

0

PRODUCT_ID

OTG_FS_DIEPT
XF1

Reset value

0

BCNT

NPTQXSAV

Reserved

OTG_FS_HPTX
FSIZ

Reset value

0x108

0

OTG_FS_CID

Reset value

0

.PWRDWN

0

VBUSASEN

0

Reserved

0

VBUSBSEN

0

OTG_FS_
GCCFG

Reset value

0

NPTXFSA/TX0FSA

SOFOUTEN

0

Reset value
0x03C

0

NPTXFD/TX0FD

Reset value

0x038

0

CHNUM

RXFD

NOVBUSSENS

OTG_FS_HNPT
XSTS

0

0

Res.

0x02C

0

Reserved

OTG_FS_HNPT
XFSIZ/
OTG_FS_DIEPT
XF0
Reset value

0

DPID

Reset value

0x028

BCNT

FSLSPCS

Register

FSLSS

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 205. OTG_FS register map and reset values (continued)

0

0

1

1

0

0

0

0

0

1

1

1

1

1

1

1

FRNUM
0

0

0

0

0

0

RM0090 Rev 17

0

0

0

1

1

1

1

1

1

1

RM0090

USB on-the-go full-speed (OTG_FS)

Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
HAINT

Reserved

0

0

0

0

0

0

0

0

DAD
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

MCNT
0

0

0

DAD
0

0

MCNT

DAD
0

0

MCNT
0

0

0

MCNT
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Reserved

Reset value

RM0090 Rev 17

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

PRES

0

0

POCCHNG

0

0

PRST

0

0

PSUSP

0

0

0

0

0

EPNUM
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

MPSIZ
0

0

0

0

0

EPNUM
0

0

MPSIZ

EPNUM
0

0

MPSIZ

EPNUM
0

0

MPSIZ

EPNUM
0

0

MPSIZ

EPNUM
0

0

MPSIZ

EPNUM
0

0

MPSIZ

EPNUM
0

PCSTS

ODDFRM

Reset value

0

DAD
0

0

PENA

OTG_FS_HCCH
AR7

0

EPDIR

0

0

EPDIR

0

0

0

EPDIR

0

0

0

EPDIR

ODDFRM

Reset value

0

MCNT

0

EPDIR

OTG_FS_HCCH
AR6

0

0

EPDIR

0

DAD

0

EPDIR

0

0

EPDIR

0

0

LSDEV

ODDFRM

Reset value

0

0

Reserved

OTG_FS_HCCH
AR5

0

LSDEV

0

0

0

Reserved

0

0

MCNT

0

LSDEV

ODDFRM

0

0

0

Reserved

CHDIS

Reset value

DAD

0

LSDEV

OTG_FS_HCCH
AR4

0

Reserved

0

0

LSDEV

ODDFRM

0

0

0

Reserved

CHDIS

0

0

0

LSDEV

CHENA

Reset value

0

0

0

Reserved

OTG_FS_HCCH
AR3

0

MCNT

0

LSDEV

0

0

0

0

PTCTL

Reserved

ODDFRM

0

DAD

0

0

LSDEV

CHDIS

0

0

EPTYP

CHENA

Reset value

0

EPTYP

OTG_FS_HCCH
AR2

0

EPTYP

0

0

EPTYP

ODDFRM

0

0

EPTYP

CHDIS

0

0

EPTYP

ODDFRM

CHENA

Reset value

0

MCNT

EPTYP

CHDIS

OTG_FS_HCCH
AR1

DAD

EPTYP

CHENA

0

OTG_FS_HCINT
0

0

0

0

0

MPSIZ
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0
XFRC

0x508

0

CHENA

0x5E0

0

CHDIS

0x5C0

Reset value

CHENA

0x5A0

OTG_FS_HCCH
AR0

CHDIS

0x580

0

CHENA

0x560

0

TXERR

PSPD

Reserved

Reserved

OTG_FS_HPRT

CHDIS

0x540

0

Reserved

0

CHENA

0x520

0

HAINTM

Reset value

0x500

0

Reserved

Reset value

0x440

0

PCDET

OTG_FS_HAINT
MSK

0

CHH

Reset value

0

0

0

Reserved

0

POCA

0

PENCHNG

0

NAK

0

STALL

0

ACK

0

BBERR

0x418

0

PTXFSAVL

FRMOR

0x414

0

OTG_FS_HAINT

PTXQSAV

PLSTS

Reset value

PTXQTOP

Reserved

OTG_FS_HPTX
STS

DTERR

0x410

Register

PPWR

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 205. OTG_FS register map and reset values (continued)

0

0

1329/1747
1379

0x54C

0x56C

0x58C
OTG_FS_HCINT
MSK2

OTG_FS_HCINT
MSK3

OTG_FS_HCINT
MSK4

1330/1747
Reserved

Reset value

Reserved

Reset value

Reserved

Reset value

RM0090 Rev 17

0

BBERR
TXERR

DTERR
FRMOR
BBERR
TXERR

Reserved
CHH
XFRC

Reserved
CHH
XFRC

Reserved
CHH
XFRC

Reserved
CHH
XFRC

XFRC

XFRC

XFRC

0
0

XFRCM

0
0

XFRCM

0
0

XFRCM

0
0

XFRCM

0

0

XFRCM

0

CHH

0

Reserved

0

CHH

0

Reserved

0

0

CHH

0

0

Reserved

0

0

CHHM

0

0

Reserved

0

0

CHHM

0

0

Reserved

0

0

CHHM

0

0

Reserved

0

0

CHHM

0

0

Reserved

0

0

CHHM

0

0

Reserved

NAK

0

STALL

0

STALL

0

STALL

0

STALL

0

STALL

0

STALL

0

STALL

0

0
STALLM

0

0

STALLM

0

0

STALLM

0

0

STALLM

0

0

STALLM

0

ACK

0

NAK

0

ACK

0

NAK

0

ACK

0

NAK

0

ACK

0

NAK

0

ACK

0

NAK

0

ACK

0
NAK

0

0

ACK

0

0

NAKM

0

0

ACKM

0

NAKM

0

0

ACKM

0

0

NAKM

0

0

ACKM

0

0

NAKM

0

0

ACKM

0

0

NAKM

0

0

ACKM

TXERR
0

Reserved

BBERR
0

Reserved

TXERR

0

0

Reserved

BBERR

0

0

Reserved

TXERR

0

Reserved

BBERR

0

0

Reserved

TXERR

0

0

Reserved

BBERR

DTERR
FRMOR

0

0

Reserved

TXERR

DTERR
FRMOR

0

Reserved

BBERR

DTERR
FRMOR

0

Reserved

TXERRM

DTERR
FRMOR

0

Reserved

BBERRM

DTERR

0

Reserved

TXERRM

Reset value
BBERRM

Reserved

TXERRM

Reset value

BBERRM

OTG_FS_HCINT
MSK1
Reserved

TXERRM

0x52C
OTG_FS_HCINT
MSK0
Reserved

BBERRM

Reset value
FRMOR

Reset value

TXERRM

0x50C
OTG_FS_HCINT
7
Reserved

DTERR

Reset value

BBERRM

0x5E8
OTG_FS_HCINT
6
Reserved

FRMOR

0x5C8
OTG_FS_HCINT
5
Reserved

DTERRM

Reset value

FRMORM

0x5A8
OTG_FS_HCINT
4
Reserved

DTERRM

Reset value

FRMORM

Reset value

DTERRM

0x588
OTG_FS_HCINT
3
Reserved

FRMORM

Reset value

DTERRM

0x568
OTG_FS_HCINT
2
Reserved

FRMORM

0x548
OTG_FS_HCINT
1

DTERRM

0x528

Register

FRMORM

Offset
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

USB on-the-go full-speed (OTG_FS)
RM0090

Table 205. OTG_FS register map and reset values (continued)

0

0

RM0090

USB on-the-go full-speed (OTG_FS)

0

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

OTG_FS_DCTL

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

PKTCNT
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DTERRM

BBERRM

TXERRM

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

XFRSIZ
0

0

0

0

0

0

0

0

0

0

0

0

PKTCNT
0

0

XFRSIZ

PKTCNT
0

0

XFRSIZ

0

0

0

XFRSIZ
0

0

0

0

0

0

Reserved

Reset value

0x804

0

XFRSIZ
0

0

CHHM

0

XFRCM

0

XFRCM

0

XFRCM

0

0

0

0

0

0

0

0

Reserved

Reset value

RM0090 Rev 17

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DSPD

0

DPID
0

0

0

0

0
RWUSIG

0

0

PKTCNT

DPID
0

0

XFRSIZ
0

0

Reserved

0

0

CHHM

0

Reserved

0

CHHM

0

Reserved

0

DPID
0

0

NZLSOHSK

0

0

PKTCNT

DPID
0

0

XFRSIZ
0

0

0

SDIS

0

0

0

0

GINSTS

OTG_FS_DCFG

0

NAKM

0x800

0

STALLM

Reset value

0

DPID
0

0

PKTCNT
0

STALLM

OTG_FS_HCTSI
Z7

0

STALLM

0x5F0

0

0

0

0

0

Reserved

Reset value

DPID

0

0

0

GONSTS

OTG_FS_HCTSI
Z6

0

ACKM

0x5D0

0

NAKM

Reset value

0

0

0

XFRSIZ

PKTCNT
0

ACKM

OTG_FS_HCTSI
Z5

0

NAKM

0x5B0

0

0

ACKM

Reset value

DPID

0

0

TCTL

OTG_FS_HCTSI
Z4

0

0

DAD

0x590

0

0

0

SGINAK

Reset value

0

0

0

CGINAK

OTG_FS_HCTSI
Z3

0

0

0

SGONAK

0x570

0

0

CGONAK

Reset value

0

PKTCNT

0

0

PFIVL

OTG_FS_HCTSI
Z2

DPID

0

POPRGDNE

0x550

Reserved

Reset value

Reserved

OTG_FS_HCTSI
Z1

Reserved

0x530

Reserved

Reset value

Reserved

OTG_FS_HCTSI
Z0

Reserved

0x510

Reserved

Reset value

Reserved

Reserved

0

Reserved

OTG_FS_HCINT
MSK7

0

Reserved

0x5EC

0
TXERRM

Reset value

0
BBERRM

Reserved

0

TXERRM

OTG_FS_HCINT
MSK6

0

BBERRM

0x5CC

FRMORM

Reset value

DTERRM

Reserved

FRMORM

OTG_FS_HCINT
MSK5

DTERRM

0x5AC

Register

FRMORM

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 205. OTG_FS register map and reset values (continued)

0

0

0

0

1331/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

0x828

0

0

0

0

0

0

0

0

OTG_FS_DAINT
MSK

0

0

0

0

0

0

0

0

OTG_FS_DVBU
SDIS

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0x918

0x938

0

0

0

0

0

0

0

1

0

1

1

TG_FS_DTXFST
S0

0

0

0

0

0

0

0

0

0

0

0

0

0

USBAEP

NAKSTS

EPTY
P

Reserved

0

STALL

0

TXFNUM

Reserved

SNAK

CNAK

EPDIS
0

Reserved

EPENA

0

0

0

0

0

0

0

0

0

USBAEP

0

NAKSTS

SNAK

CNAK

0

EONUM/DPID

SD0PID/SEVNFRM
0

TXFNUM

EPTYP

SODDFRM/SD1PID
0

STALL

EPDIS
0

0

0

0

0

0

0

0

0

1

0

SUSPSTS

ENUMSPD

XFRCM

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

1

1

0

1

1

1

0

1

1

0

1

1

1

0

0

0

0

0

0

0

0

0

0

0

0

0

MPSI
Z
0

0

INEPTFSAV

Reserved

EPENA
0

0

1

0

Reset value

0

Reserved

Reserved

OTG_FS_DIEPC
TL1

EERR
0

INEPTXFEM

Reserved

Reset value

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

MPSIZ

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

INEPTFSAV
0

RM0090 Rev 17

1

Reserved

Reserved

Reset value

1332/1747

0

0

OTG_FS_DIEPC
TL0

TG_FS_DTXFST
S1

0

DVBUSP

Reset value

0x920

0

Reserved

Reset value

0x900

0

VBUSDT
0

OTG_FS_DIEPE
MPMSK

0

IEPM

Reset value

0x834

0

IEPINT

0

Reserved

OTG_FS_DVBU
SPULSE

0

0

0

OEPM
0

0

EPDM

Reserved

OEPINT
0

Reset value

0x82C

0

Reserved

Reserved

OTG_FS_DAINT

Reset value

0

XFRCM

0x81C

0

OUTPKTERRM

OTG_FS_DOEP
MSK

Reset value

0

0

Reset value
0x818

0

EPDM

Reserved

Reset value

0x814

0

TOM

0

Reserved

0

ITTXFEMSK

0

STUPM

0

OTEPDM

0

INEPNEM

0

INEPNMM

OTG_FS_DIEPM
SK

0

Reserved

0x810

0

NAKM

Reset value

Reserved

Reserved

FNSOF

Reserved

STSPHSRXM

OTG_FS_DSTS

BERRM

0x808

Register

NAKMSK

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 205. OTG_FS register map and reset values (continued)

0

0

0

0

0

1

0

0

0

RM0090

USB on-the-go full-speed (OTG_FS)

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

MPSIZ

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

MPSI
Z

Reserved

0

0

MPSIZ

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

MPSIZ

Reserved

0

0

0

0

0

0

0

MPSIZ

Reserved

0

0

0

XFRC

0

0

EPDISD

0

RM0090 Rev 17

0

0

0

0

0

0

0

0

0

0

TOC

USBAEP

0

Reset value

0

Reserved

NAKSTS

Reserved

NAKSTS

EONUM/DPID

0

Reserved

0

ITTXFE

0

0

PKTDRPSTS

0

0

USBAEP

0

0

USBAEP

0

0

NAKSTS

0

EPTYP

SNPM
SNPM

STALL
STALL
0

0

0

1

EONUM/DPID

0

0

0

NAKSTS

0

0

0

EONUM/DPID

0

0

0

EPTYP

0

Reserved

0

EPTY
P

EPTYP

0

0

SNPM

0

Reserved

0

STALL

0

Reserved

0

SNPM

0

Reserved

STALL

SNAK

CNAK

SNAK

CNAK

SNAK

0

CNAK

EPDIS
0

0

SNAK

0

0

CNAK

Reset value

Reserved

OTG_FS_DOEP
CTL3

0

SODDFRM

0

SD0PID/SEVNFRM

Reset value

0

SODDFRM

OTG_FS_DOEP
CTL2

0

SD0PID/SEVNFRM

0

0

SODDFRM

EPDIS

EPENA

Reset value

0

SD0PID/SEVNFRM

EPENA

OTG_FS_DOEP
CTL1

EPDIS

0

OTG_FS_DIEPI
NT0

0

Reserved

NAK

0x908

0

EPENA

0xB60

Reset value

EPDIS

0xB40

OTG_FS_DOEP
CTL0

0

INEPTFSAV
0

EPENA

0xB20

0

Reserved

Reset value

0xB00

0

INEPNM

0

USBAEP

0

0

TXFE

TG_FS_DTXFST
S3

0

NAKSTS

0

EONUM/DPID

0

EPTYP

0

STALL

0

Reserved

SNAK

CNAK

0

SODDFRM

EPDIS

0

SD0PID/SEVNFRM

EPENA

Reset value

TXFNUM

0

Reserved

0x978

0

OTG_FS_DIEPC
TL3

0

INEPTFSAV

Reserved

Reset value

0x960

0

INEPNE

0

USBAEP

0

MPSIZ

Reserved

Reserved

TG_FS_DTXFST
S2

0

NAKSTS

0

EONUM/DPID

0

EPTYP

0

STALL

0

Reserved

SNAK

CNAK

0

SODDFRM

0

SD0PID/SEVNFRM

Reset value

TXFNUM

USBAEP

0x958

OTG_FS_DIEPC
TL2

EPDIS

0x940

Register

EPENA

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 205. OTG_FS register map and reset values (continued)

1

0

0

0

0

1333/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

0x910

OTG_FS_DIEPT
SIZ0

PKTC
NT

Reserved

Reset value

0x930

OTG_FS_DIEPT
SIZ1

0

0x950

OTG_FS_DIEPT
SIZ2

0

0x970

OTG_FS_DIEPT
SIZ3
Reset value

1334/1747

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

EPDISD

XFRC
XFRC
XFRC

EPDISD
EPDISD

XFRC

0

0
XFRC

EPDISD

0

EPDISD

Reserved
Reserved
Reserved

0

0

0
XFRC

0

0

EPDISD

0

0

0

0
XFRC

0

Reserved

TOC

0

Reserved

0

Reserved

ITTXFE

0

STUP

STUP

TOC

TOC

INEPNM

ITTXFE
ITTXFE
OTEPDIS
0
OTEPDIS

0
STSPHSRX

Reserved

STSPHSRX

INEPNM

INEPNM

TXFE

INEPNE

TXFE

Reserved

INEPNE

TXFE

INEPNE

Reserved

Reserved

PKTDRPSTS
PKTDRPSTS

Reserved
OUTPKTERR

0

0

0

0

XFRSIZ

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

XFRSIZ
0

0

0

0

0

0

0

0

0

0

0

0

PKTCNT

Reserved

0

0

XFRSIZ
0

0

0

0

PKTCNT

Reserved

Reset value

0

0

0

PKTCNT

Reserved

Reset value

0

1

0

EPDISD

Reset value

0

Reserved

Reserved

0

STUP

OTG_FS_DOEPI
NT3

0

0

STUP

0xB68

0

0

OTEPDIS

Reset value

1

STSPHSRX

Reserved

0

OTEPDIS

OTG_FS_DOEPI
NT2

0

0

STSPHSRX

0xB48

0

0

Reserved

Reset value

0

Reserved

Reserved

1

0
OUTPKTERR

OTG_FS_DOEPI
NT1

0

OUTPKTERR

0xB28

0

OUTPKTERR

Reset value

NAK

Reserved

BERR

OTG_FS_DOEPI
NT0

0

NAK

0xB08

0

BERR

Reset value

PKTDRPSTS

Reserved

Reserved

NAK

OTG_FS_DIEPI
NT3

0

NAK

0x968

0

Reserved

Reset value

Reserved

Reserved

Reserved

NAK

OTG_FS_DIEPI
NT2

0

BERR

0x948

0

Reserved

Reset value

Reserved

NAK

Reserved

Reserved

OTG_FS_DIEPI
NT1

NAK

0x928

Register

BERR

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 205. OTG_FS register map and reset values (continued)

0

0

0

XFRSIZ
0

0

0

0

0

0

RM0090 Rev 17

0

0

0

0

0

0

0

0

0

RM0090

USB on-the-go full-speed (OTG_FS)

0xB70

OTG_FS_DOEP
TSIZ3
Reset value

0xE00

OTG_FS_PCGC
CTL

0

0

0

0

0

0

0

0

0

0

Reserved

PKTCNT

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

XFRSIZ
0

0

0

0

0

0

0

0

0

0

0

0

PKTCNT
0

0

XFRSIZ

PKTCNT
0

0

STPPCLK

Reset value

0

PKTCNT

0

GATEHCLK

OTG_FS_DOEP
TSIZ2

0

0

0

0

XFRSIZ
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Reserved

Reserved

0xB50

0

XFRSIZ

Reserved

PHYSUSP

Reset value

0

Reserved

RXDPID/
STUPCNT

OTG_FS_DOEP
TSIZ1

Reserved

0xB30

STUP
CNT

RXDPID/
STUPCNT

Reset value

Reserved

OTG_FS_DOEP
TSIZ0

RXDPID/
STUPCNT

0xB10

Register

Reserved

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 205. OTG_FS register map and reset values (continued)

Reset value

Refer to Section 2.3: Memory map for the register boundary addresses.

RM0090 Rev 17

1335/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

34.17

OTG_FS programming model

34.17.1

Core initialization
The application must perform the core initialization sequence. If the cable is connected
during power-up, the current mode of operation bit in the OTG_FS_GINTSTS (CMOD bit in
OTG_FS_GINTSTS) reflects the mode. The OTG_FS controller enters host mode when an
“A” plug is connected or device mode when a “B” plug is connected.
This section explains the initialization of the OTG_FS controller after power-on. The
application must follow the initialization sequence irrespective of host or device mode
operation. All core global registers are initialized according to the core’s configuration:
1.

2.

3.

Program the following fields in the OTG_FS_GAHBCFG register:
–

Global interrupt mask bit GINTMSK = 1

–

RxFIFO non-empty (RXFLVL bit in OTG_FS_GINTSTS)

–

Periodic TxFIFO empty level

Program the following fields in the OTG_FS_GUSBCFG register:
–

HNP capable bit

–

SRP capable bit

–

FS timeout calibration field

–

USB turnaround time field

The software must unmask the following bits in the OTG_FS_GINTMSK register:
OTG interrupt mask
Mode mismatch interrupt mask

4.

1336/1747

The software can read the CMOD bit in OTG_FS_GINTSTS to determine whether the
OTG_FS controller is operating in host or device mode.

RM0090 Rev 17

RM0090

34.17.2

USB on-the-go full-speed (OTG_FS)

Host initialization
To initialize the core as host, the application must perform the following steps:
1.

Program the HPRTINT in the OTG_FS_GINTMSK register to unmask

2.

Program the OTG_FS_HCFG register to select full-speed host

3.

Program the PPWR bit in OTG_FS_HPRT to 1. This drives VBUS on the USB.

4.

Wait for the PCDET interrupt in OTG_FS_HPRT0. This indicates that a device is
connecting to the port.

5.

Program the PRST bit in OTG_FS_HPRT to 1. This starts the reset process.

6.

Wait at least 10 ms for the reset process to complete.

7.

Program the PRST bit in OTG_FS_HPRT to 0.

8.

Wait for the PENCHNG interrupt in OTG_FS_HPRT.

9.

Read the PSPD bit in OTG_FS_HPRT to get the enumerated speed.

10. Program the HFIR register with a value corresponding to the selected PHY clock 1
11. Program the FSLSPCS field in the OTG_FS_HCFG register following the speed of the
device detected in step 9. If FSLSPCS has been changed a port reset must be
performed.
12. Program the OTG_FS_GRXFSIZ register to select the size of the receive FIFO.
13. Program the OTG_FS_HNPTXFSIZ register to select the size and the start address of
the Non-periodic transmit FIFO for non-periodic transactions.
14. Program the OTG_FS_HPTXFSIZ register to select the size and start address of the
periodic transmit FIFO for periodic transactions.
To communicate with devices, the system software must initialize and enable at least one
channel.

34.17.3

Device initialization
The application must perform the following steps to initialize the core as a device on powerup or after a mode change from host to device.
1.

2.

Program the following fields in the OTG_FS_DCFG register:
–

Device speed

–

Non-zero-length status OUT handshake

Program the OTG_FS_GINTMSK register to unmask the following interrupts:
–

USB reset

–

Enumeration done

–

Early suspend

–

USB suspend

–

SOF

3.

Program the VBUSBSEN bit in the OTG_FS_GCCFG register to enable VBUS sensing
in “B” device mode and supply the 5 volts across the pull-up resistor on the DP line.

4.

Wait for the USBRST interrupt in OTG_FS_GINTSTS. It indicates that a reset has been
detected on the USB that lasts for about 10 ms on receiving this interrupt.

Wait for the ENUMDNE interrupt in OTG_FS_GINTSTS. This interrupt indicates the end of
reset on the USB. On receiving this interrupt, the application must read the OTG_FS_DSTS
register to determine the enumeration speed and perform the steps listed in Endpoint

RM0090 Rev 17

1337/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

initialization on enumeration completion on page 1354.
At this point, the device is ready to accept SOF packets and perform control transfers on
control endpoint 0.

34.17.4

Host programming model
Channel initialization
The application must initialize one or more channels before it can communicate with
connected devices. To initialize and enable a channel, the application must perform the
following steps:
1.

Program the OTG_FS_GINTMSK register to unmask the following:

2.

Channel interrupt
–

Non-periodic transmit FIFO empty for OUT transactions (applicable when
operating in pipelined transaction-level with the packet count field programmed
with more than one).

–

Non-periodic transmit FIFO half-empty for OUT transactions (applicable when
operating in pipelined transaction-level with the packet count field programmed
with more than one).

3.

Program the OTG_FS_HAINTMSK register to unmask the selected channels’
interrupts.

4.

Program the OTG_FS_HCINTMSK register to unmask the transaction-related
interrupts of interest given in the host channel interrupt register.

5.

Program the selected channel’s OTG_FS_HCTSIZx register with the total transfer size,
in bytes, and the expected number of packets, including short packets. The application
must program the PID field with the initial data PID (to be used on the first OUT
transaction or to be expected from the first IN transaction).

6.

Program the OTG_FS_HCCHARx register of the selected channel with the device’s
endpoint characteristics, such as type, speed, direction, and so forth. (The channel can
be enabled by setting the channel enable bit to 1 only when the application is ready to
transmit or receive any packet).

Halting a channel
The application can disable any channel by programming the OTG_FS_HCCHARx register
with the CHDIS and CHENA bits set to 1. This enables the OTG_FS host to flush the posted
requests (if any) and generates a channel halted interrupt. The application must wait for the
CHH interrupt in OTG_FS_HCINTx before reallocating the channel for other transactions.
The OTG_FS host does not interrupt the transaction that has already been started on the
USB.
Before disabling a channel, the application must ensure that there is at least one free space
available in the non-periodic request queue (when disabling a non-periodic channel) or the
periodic request queue (when disabling a periodic channel). The application can simply
flush the posted requests when the Request queue is full (before disabling the channel), by
programming the OTG_FS_HCCHARx register with the CHDIS bit set to 1, and the CHENA
bit cleared to 0.
The application is expected to disable a channel on any of the following conditions:

1338/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
1.

When an STALL, TXERR, BBERR or DTERR interrupt in OTG_FS_HCINTx is received
for an IN or OUT channel. The application must be able to receive other interrupts
(DTERR, Nak, Data, TXERR) for the same channel before receiving the halt.

2.

When a DISCINT (Disconnect Device) interrupt in OTG_FS_GINTSTS is received.
(The application is expected to disable all enabled channels).

3.

When the application aborts a transfer before normal completion.

Operational model
The application must initialize a channel before communicating to the connected device.
This section explains the sequence of operation to be performed for different types of USB
transactions.
•

Writing the transmit FIFO
The OTG_FS host automatically writes an entry (OUT request) to the periodic/nonperiodic request queue, along with the last word write of a packet. The application must
ensure that at least one free space is available in the periodic/non-periodic request
queue before starting to write to the transmit FIFO. The application must always write
to the transmit FIFO in words. If the packet size is non-word aligned, the application
must use padding. The OTG_FS host determines the actual packet size based on the
programmed maximum packet size and transfer size.
Figure 396. Transmit FIFO write task
Start

Read GNPTXSTS/HPTXFSIZ
registers for available FIFO
and queue spaces

Wait for NPTXFE/PTXFE
interrupt in
OTG_FS_GINTSTS

No

1 MPS or
LPS FIFO space
available?

Yes

Yes

Write 1 packet
data to
transmit FIFO

More packets
to send?
No
MPS: Maximum packet size
LPS: Last packet size

Done
ai15673b

•

Reading the receive FIFO

RM0090 Rev 17

1339/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

The application must ignore all packet statuses other than IN data packet (bx0010).
Figure 397. Receive FIFO read task
Start

No

RXFLVL
interrupt ?

Yes

Unmask RXFLVL
interrupt

Read the received
packet from the
Receive FIFO

Mask RXFLVL
interrupt

Unmask RXFLVL
interrupt

Read
OTG_FS_GRXSTSP

PKTSTS
0b0010?

No
No

Yes
Yes

BCNT > 0?

ai15674

•

Bulk and control OUT/SETUP transactions
A typical bulk or control OUT/SETUP pipelined transaction-level operation is shown in
Figure 398. See channel 1 (ch_1). Two bulk OUT packets are transmitted. A control
SETUP transaction operates in the same way but has only one packet. The
assumptions are:

•

–

The application is attempting to send two maximum-packet-size packets (transfer
size = 1, 024 bytes).

–

The non-periodic transmit FIFO can hold two packets (128 bytes for FS).

–

The non-periodic request queue depth = 4.

Normal bulk and control OUT/SETUP operations
The sequence of operations in (channel 1) is as follows:

1340/1747

a)

Initialize channel 1

b)

Write the first packet for channel 1

c)

Along with the last word write, the core writes an entry to the non-periodic request
queue

d)

As soon as the non-periodic queue becomes non-empty, the core attempts to
send an OUT token in the current frame

e)

Write the second (last) packet for channel 1

f)

The core generates the XFRC interrupt as soon as the last transaction is
completed successfully

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
g)

In response to the XFRC interrupt, de-allocate the channel for other transfers

h)

Handling non-ACK responses

Figure 398. Normal bulk/control OUT/SETUP and bulk/control IN transactions
Application
1
init _reg(ch_2)

set _ch_en
(ch _2)

1

Host

USB

Device

init_reg(ch _1)

write_tx_fifo
(ch_1)

1
MPS

2
2

set _ch_en
(ch _2)

AHB

4

3

Non-Periodic Request
Queue
Assume that this queue
can hold 4 entries.

ch_1
write_tx_fifo
(ch_1)

1
MPS

5

ch_2
ch_1
ch_2

OU T

D AT A0
MPS

3
AC K

set _ch_en
(ch _2)

IN

4

D AT A0

5
RXFLVL interrupt

1
MPS

read_rx_sts
read_rx_fifo

ch_1
ch_2

set _ch_en
(ch _2)

ch_2

ACK
O UT

D AT A1
MPS

ch_2

7

ACK

XFRC interrupt

6
IN

De-allocate
(ch_1)

D AT A1
RXFLVL interrupt

1
MPS

read_rx_stsre
ad_rx_fifo

RXFLVL interrupt

read_rx_sts

Disable
(ch _2)

7

6

8

ACK

ch_2

XFRC interrupt

9
RXFLVL interrupt

read_rx_sts

De-allocate
(ch _2)

11

CHH interrupt
r

10
12

13
ai15675

The channel-specific interrupt service routine for bulk and control OUT/SETUP
transactions is shown in the following code samples.
•

Interrupt service routine for bulk/control OUT/SETUP and bulk/control IN
transactions
a)

Bulk/Control OUT/SETUP

Unmask (NAK/TXERR/STALL/XFRC)
if (XFRC)
{

RM0090 Rev 17

1341/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Reset Error Count
Mask ACK
De-allocate Channel
}
else if (STALL)
{
Transfer Done = 1
Unmask CHH
Disable Channel
}
else if (NAK or TXERR )
{
Rewind Buffer Pointers
Unmask CHH
Disable Channel
if (TXERR)
{
Increment Error Count
Unmask ACK
}
else
{
Reset Error Count
}
}
else if (CHH)
{
Mask CHH
if (Transfer Done or (Error_count == 3))
{
De-allocate Channel
}
else
{
Re-initialize Channel
}
}
else if (ACK)
{
Reset Error Count
Mask ACK
}
The application is expected to write the data packets into the transmit FIFO as and
when the space is available in the transmit FIFO and the Request queue. The
application can make use of the NPTXFE interrupt in OTG_FS_GINTSTS to find the
transmit FIFO space.
b)

Bulk/Control IN

Unmask (TXERR/XFRC/BBERR/STALL/DTERR)
if (XFRC)
{
Reset Error Count

1342/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
Unmask CHH
Disable Channel
Reset Error Count
Mask ACK
}
else if (TXERR or BBERR or STALL)
{
Unmask CHH
Disable Channel
if (TXERR)
{
Increment Error Count
Unmask ACK
}
}
else if (CHH)
{
Mask CHH
if (Transfer Done or (Error_count == 3))
{
De-allocate Channel
}
else
{
Re-initialize Channel
}
}
else if (ACK)
{
Reset Error Count
Mask ACK
}
else if (DTERR)
{
Reset Error Count
}
The application is expected to write the requests as and when the Request queue space is
available and until the XFRC interrupt is received.
•

Bulk and control IN transactions
A typical bulk or control IN pipelined transaction-level operation is shown in Figure 399.
See channel 2 (ch_2). The assumptions are:
–

The application is attempting to receive two maximum-packet-size packets
(transfer size = 1 024 bytes).

–

The receive FIFO can contain at least one maximum-packet-size packet and two
status words per packet (72 bytes for FS).

–

The non-periodic request queue depth = 4.

RM0090 Rev 17

1343/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Figure 399. Bulk/control IN transactions
Application
1
init _reg(ch_2)

set _ch_en
(ch _2)

1

set _ch_en
(ch _2)

write_tx_fifo
(ch_1)

2
2

AHB

Host

USB

Device

init_reg(ch _1)

1
MPS

4

3

Non-Periodic Request
Queue
Assume that this queue
can hold 4 entries.

ch_1
write_tx_fifo
(ch_1)

5

1
MPS

ch_2
ch_1
ch_2

OU T

D AT A0
MPS

3
AC K

set _ch_en
(ch _2)

IN

4

D AT A0

5
RXFLVL interrupt

1
MPS

read_rx_sts
read_rx_fifo

ch_1
ch_2

set _ch_en
(ch _2)

ch_2

ACK
O UT

D AT A1
MPS

ch_2

7

ACK

XFRC interrupt

6
IN

De-allocate
(ch_1)

D AT A1
RXFLVL interrupt

1
MPS

read_rx_stsre
ad_rx_fifo

RXFLVL interrupt

read_rx_sts

Disable
(ch _2)

7

6

8

ACK

ch_2

XFRC interrupt

9
RXFLVL interrupt

read_rx_sts

De-allocate
(ch _2)

11

CHH interrupt
r

10
12

13
ai15675

The sequence of operations is as follows:

1344/1747

a)

Initialize channel 2.

b)

Set the CHENA bit in HCCHAR2 to write an IN request to the non-periodic request
queue.

c)

The core attempts to send an IN token after completing the current OUT
transaction.

d)

The core generates an RXFLVL interrupt as soon as the received packet is written
to the receive FIFO.

e)

In response to the RXFLVL interrupt, mask the RXFLVL interrupt and read the
received packet status to determine the number of bytes received, then read the
receive FIFO accordingly. Following this, unmask the RXFLVL interrupt.

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
f)

The core generates the RXFLVL interrupt for the transfer completion status entry
in the receive FIFO.

g)

The application must read and ignore the receive packet status when the receive
packet status is not an IN data packet (PKTSTS in GRXSTSR ≠ 0b0010).

h)

The core generates the XFRC interrupt as soon as the receive packet status is
read.

i)

In response to the XFRC interrupt, disable the channel and stop writing the
OTG_FS_HCCHAR2 register for further requests. The core writes a channel
disable request to the non-periodic request queue as soon as the
OTG_FS_HCCHAR2 register is written.

j)

The core generates the RXFLVL interrupt as soon as the halt status is written to
the receive FIFO.

k)

Read and ignore the receive packet status.

l)

The core generates a CHH interrupt as soon as the halt status is popped from the
receive FIFO.

m) In response to the CHH interrupt, de-allocate the channel for other transfers.
n)
•

Handling non-ACK responses

Control transactions
Setup, Data, and Status stages of a control transfer must be performed as three
separate transfers. Setup-, Data- or Status-stage OUT transactions are performed
similarly to the bulk OUT transactions explained previously. Data- or Status-stage IN
transactions are performed similarly to the bulk IN transactions explained previously.
For all three stages, the application is expected to set the EPTYP field in
OTG_FS_HCCHAR1 to Control. During the Setup stage, the application is expected to
set the PID field in OTG_FS_HCTSIZ1 to SETUP.

•

Interrupt OUT transactions
A typical interrupt OUT operation is shown in Figure 400. The assumptions are:
–

The application is attempting to send one packet in every frame (up to 1 maximum
packet size), starting with the odd frame (transfer size = 1 024 bytes)

–

The periodic transmit FIFO can hold one packet (1 KB)

–

Periodic request queue depth = 4

The sequence of operations is as follows:
a)

Initialize and enable channel 1. The application must set the ODDFRM bit in
OTG_FS_HCCHAR1.

b)

Write the first packet for channel 1.

c)

Along with the last word write of each packet, the OTG_FS host writes an entry to
the periodic request queue.

d)

The OTG_FS host attempts to send an OUT token in the next (odd) frame.

e)

The OTG_FS host generates an XFRC interrupt as soon as the last packet is
transmitted successfully.

f)

In response to the XFRC interrupt, reinitialize the channel for the next transfer.

RM0090 Rev 17

1345/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Figure 400. Normal interrupt OUT/IN transactions
Application
1
init_reg(ch _2)

set_ch_en
(ch_2)

1

AHB

Host

USB

Device

init _reg(ch_1)

write_tx_fifo
(ch_1)

2

Periodic Request Queue
Assume that this queue
can hold 4 entries.

3

1
MPS

4

ch_1

2

ch_2

3
OU T

DATA0
M PS

Odd
(micro)
frame

5
6

ACK

XFRC interrupt

4

init _reg(ch_1)
write_tx_fifo
(ch_1)

IN

5

1
MPS

DATA0

RXFLVL interrupt
read_rx_sts
read_rx_fifo

ACK

1
MPS

6

RXFLVL interrupt
read_rx_sts

init_reg(ch _2)

7

XFRC interrupt

8

ch_1
ch_2

9

set_ch_en
(ch_2)

OU T

XFRC interrupt

init _reg(ch_1)

write_tx_fifo
(ch_1)

1
MPS

DATA1
MPS

Even
(micro)
frame

ACK
IN

DATA1

ai15676

•

Interrupt service routine for interrupt OUT/IN transactions
a)

Interrupt OUT

Unmask (NAK/TXERR/STALL/XFRC/FRMOR)
if (XFRC)
{
Reset Error Count
Mask ACK
De-allocate Channel
}
else
if (STALL or FRMOR)
{
Mask ACK
Unmask CHH
1346/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
Disable Channel
if (STALL)
{
Transfer Done = 1
}
}
else
if (NAK or TXERR)
{
Rewind Buffer Pointers
Reset Error Count
Mask ACK
Unmask CHH
Disable Channel
}
else
if (CHH)
{
Mask CHH
if (Transfer Done or (Error_count == 3))
{
De-allocate Channel
}
else
{
Re-initialize Channel (in next b_interval - 1 Frame)
}
}
else
if (ACK)
{
Reset Error Count
Mask ACK
}
The application uses the NPTXFE interrupt in OTG_FS_GINTSTS to find the transmit
FIFO space.
b)

Interrupt IN

Unmask (NAK/TXERR/XFRC/BBERR/STALL/FRMOR/DTERR)
if (XFRC)
{
Reset Error Count
Mask ACK
if (OTG_FS_HCTSIZx.PKTCNT == 0)
{
De-allocate Channel
}
else
{
Transfer Done = 1
Unmask CHH
Disable Channel

RM0090 Rev 17

1347/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

}
}
else
if (STALL or FRMOR or NAK or DTERR or BBERR)
{
Mask ACK
Unmask CHH
Disable Channel
if (STALL or BBERR)
{
Reset Error Count
Transfer Done = 1
}
else
if (!FRMOR)
{
Reset Error Count
}
}
else
if (TXERR)
{
Increment Error Count
Unmask ACK
Unmask CHH
Disable Channel
}
else
if (CHH)
{
Mask CHH
if (Transfer Done or (Error_count == 3))
{
De-allocate Channel
}
else
Re-initialize Channel (in next b_interval - 1 /Frame)
}
}
else
if (ACK)
{
Reset Error Count
Mask ACK

1348/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
}
•

Interrupt IN transactions
The assumptions are:

•

–

The application is attempting to receive one packet (up to 1 maximum packet size)
in every frame, starting with odd (transfer size = 1 024 bytes).

–

The receive FIFO can hold at least one maximum-packet-size packet and two
status words per packet (1 031 bytes).

–

Periodic request queue depth = 4.

Normal interrupt IN operation
The sequence of operations is as follows:
a)

Initialize channel 2. The application must set the ODDFRM bit in
OTG_FS_HCCHAR2.

b)

Set the CHENA bit in OTG_FS_HCCHAR2 to write an IN request to the periodic
request queue.

c)

The OTG_FS host writes an IN request to the periodic request queue for each
OTG_FS_HCCHAR2 register write with the CHENA bit set.

d)

The OTG_FS host attempts to send an IN token in the next (odd) frame.

e)

As soon as the IN packet is received and written to the receive FIFO, the OTG_FS
host generates an RXFLVL interrupt.

f)

In response to the RXFLVL interrupt, read the received packet status to determine
the number of bytes received, then read the receive FIFO accordingly. The
application must mask the RXFLVL interrupt before reading the receive FIFO, and
unmask after reading the entire packet.

g)

The core generates the RXFLVL interrupt for the transfer completion status entry
in the receive FIFO. The application must read and ignore the receive packet
status when the receive packet status is not an IN data packet (PKTSTS in
GRXSTSR ≠ 0b0010).

h)

The core generates an XFRC interrupt as soon as the receive packet status is
read.

i)

In response to the XFRC interrupt, read the PKTCNT field in OTG_FS_HCTSIZ2.
If the PKTCNT bit in OTG_FS_HCTSIZ2 is not equal to 0, disable the channel
before re-initializing the channel for the next transfer, if any). If PKTCNT bit in

RM0090 Rev 17

1349/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

OTG_FS_HCTSIZ2 = 0, reinitialize the channel for the next transfer. This time, the
application must reset the ODDFRM bit in OTG_FS_HCCHAR2.
•

Isochronous OUT transactions
A typical isochronous OUT operation is shown in Figure 401. The assumptions are:
–

The application is attempting to send one packet every frame (up to 1 maximum
packet size), starting with an odd frame. (transfer size = 1 024 bytes).

–

The periodic transmit FIFO can hold one packet (1 KB).

–

Periodic request queue depth = 4.

The sequence of operations is as follows:

1350/1747

a)

Initialize and enable channel 1. The application must set the ODDFRM bit in
OTG_FS_HCCHAR1.

b)

Write the first packet for channel 1.

c)

Along with the last word write of each packet, the OTG_FS host writes an entry to
the periodic request queue.

d)

The OTG_FS host attempts to send the OUT token in the next frame (odd).

e)

The OTG_FS host generates the XFRC interrupt as soon as the last packet is
transmitted successfully.

f)

In response to the XFRC interrupt, reinitialize the channel for the next transfer.

g)

Handling non-ACK responses

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
Figure 401. Normal isochronous OUT/IN transactions
AHB

Application
1
init_reg(ch _2)

set_ch_en
(ch_2)

1

Host

USB

Device

init _reg(ch_1)

write_tx_fifo
(ch_1)

2

Periodic Request Queue
Assume that this queue
can hold 4 entries.

3

1
MPS

4

ch_1

2

ch_2

3
OU T

DATA0
M PS

Odd
(micro)
frame

5
6

ACK

XFRC interrupt

4

init _reg(ch_1)
write_tx_fifo
(ch_1)

IN

5

1
MPS

DATA0

RXFLVL interrupt
read_rx_sts
read_rx_fifo

ACK

1
MPS

6

RXFLVL interrupt
read_rx_sts

init_reg(ch _2)

7

XFRC interrupt

8

ch_1
ch_2

9

set_ch_en
(ch_2)

OU T

XFRC interrupt

init _reg(ch_1)

1
MPS

write_tx_fifo
(ch_1)

DATA1
MPS

Even
(micro)
frame

ACK
IN

DATA1

ai15676

•

Interrupt service routine for isochronous OUT/IN transactions
Code sample: Isochronous OUT

Unmask (FRMOR/XFRC)
if (XFRC)
{
De-allocate Channel
}
else
if (FRMOR)
{
Unmask CHH
Disable Channel
}

RM0090 Rev 17

1351/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

else
if (CHH)
{
Mask CHH
De-allocate Channel
}
Code sample: Isochronous IN
Unmask (TXERR/XFRC/FRMOR/BBERR)
if (XFRC or FRMOR)
{
if (XFRC and (OTG_FS_HCTSIZx.PKTCNT == 0))
{
Reset Error Count
De-allocate Channel
}
else
{
Unmask CHH
Disable Channel
}
}
else
if (TXERR or BBERR)
{
Increment Error Count
Unmask CHH
Disable Channel
}
else
if (CHH)
{
Mask CHH
if (Transfer Done or (Error_count == 3))
{
De-allocate Channel
}
else
{
Re-initialize Channel
}
}

1352/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
•

Isochronous IN transactions
The assumptions are:
–

The application is attempting to receive one packet (up to 1 maximum packet size)
in every frame starting with the next odd frame (transfer size = 1 024 bytes).

–

The receive FIFO can hold at least one maximum-packet-size packet and two
status word per packet (1 031 bytes).

–

Periodic request queue depth = 4.

The sequence of operations is as follows:

•

a)

Initialize channel 2. The application must set the ODDFRM bit in
OTG_FS_HCCHAR2.

b)

Set the CHENA bit in OTG_FS_HCCHAR2 to write an IN request to the periodic
request queue.

c)

The OTG_FS host writes an IN request to the periodic request queue for each
OTG_FS_HCCHAR2 register write with the CHENA bit set.

d)

The OTG_FS host attempts to send an IN token in the next odd frame.

e)

As soon as the IN packet is received and written to the receive FIFO, the OTG_FS
host generates an RXFLVL interrupt.

f)

In response to the RXFLVL interrupt, read the received packet status to determine
the number of bytes received, then read the receive FIFO accordingly. The
application must mask the RXFLVL interrupt before reading the receive FIFO, and
unmask it after reading the entire packet.

g)

The core generates an RXFLVL interrupt for the transfer completion status entry in
the receive FIFO. This time, the application must read and ignore the receive
packet status when the receive packet status is not an IN data packet (PKTSTS bit
in OTG_FS_GRXSTSR ≠ 0b0010).

h)

The core generates an XFRC interrupt as soon as the receive packet status is
read.

i)

In response to the XFRC interrupt, read the PKTCNT field in OTG_FS_HCTSIZ2.
If PKTCNT ≠ 0 in OTG_FS_HCTSIZ2, disable the channel before re-initializing the
channel for the next transfer, if any. If PKTCNT = 0 in OTG_FS_HCTSIZ2,
reinitialize the channel for the next transfer. This time, the application must reset
the ODDFRM bit in OTG_FS_HCCHAR2.

Selecting the queue depth
Choose the periodic and non-periodic request queue depths carefully to match the
number of periodic/non-periodic endpoints accessed.
The non-periodic request queue depth affects the performance of non-periodic
transfers. The deeper the queue (along with sufficient FIFO size), the more often the
core is able to pipeline non-periodic transfers. If the queue size is small, the core is
able to put in new requests only when the queue space is freed up.
The core’s periodic request queue depth is critical to perform periodic transfers as
scheduled. Select the periodic queue depth, based on the number of periodic transfers
scheduled in a microframe. If the periodic request queue depth is smaller than the
periodic transfers scheduled in a microframe, a frame overrun condition occurs.

•

Handling babble conditions
OTG_FS controller handles two cases of babble: packet babble and port babble.
Packet babble occurs if the device sends more data than the maximum packet size for

RM0090 Rev 17

1353/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

the channel. Port babble occurs if the core continues to receive data from the device at
EOF2 (the end of frame 2, which is very close to SOF).
When OTG_FS controller detects a packet babble, it stops writing data into the Rx
buffer and waits for the end of packet (EOP). When it detects an EOP, it flushes already
written data in the Rx buffer and generates a Babble interrupt to the application.
When OTG_FS controller detects a port babble, it flushes the RxFIFO and disables the
port. The core then generates a Port disabled interrupt (HPRTINT in
OTG_FS_GINTSTS, PENCHNG in OTG_FS_HPRT). On receiving this interrupt, the
application must determine that this is not due to an overcurrent condition (another
cause of the Port Disabled interrupt) by checking POCA in OTG_FS_HPRT, then
perform a soft reset. The core does not send any more tokens after it has detected a
port babble condition.

34.17.5

Device programming model
Endpoint initialization on USB reset
1.

Set the NAK bit for all OUT endpoints
–

2.

3.

4.

SNAK = 1 in OTG_FS_DOEPCTLx (for all OUT endpoints)

Unmask the following interrupt bits
–

INEP0 = 1 in OTG_FS_DAINTMSK (control 0 IN endpoint)

–

OUTEP0 = 1 in OTG_FS_DAINTMSK (control 0 OUT endpoint)

–

STUP = 1 in DOEPMSK

–

XFRC = 1 in DOEPMSK

–

XFRC = 1 in DIEPMSK

–

TOC = 1 in DIEPMSK

Set up the Data FIFO RAM for each of the FIFOs
–

Program the OTG_FS_GRXFSIZ register, to be able to receive control OUT data
and setup data. If thresholding is not enabled, at a minimum, this must be equal to
1 max packet size of control endpoint 0 + 2 words (for the status of the control
OUT data packet) + 10 words (for setup packets).

–

Program the OTG_FS_TX0FSIZ register (depending on the FIFO number chosen)
to be able to transmit control IN data. At a minimum, this must be equal to 1 max
packet size of control endpoint 0.

Program the following fields in the endpoint-specific registers for control OUT endpoint
0 to receive a SETUP packet
–

STUPCNT = 3 in OTG_FS_DOEPTSIZ0 (to receive up to 3 back-to-back SETUP
packets)

At this point, all initialization required to receive SETUP packets is done.

Endpoint initialization on enumeration completion

1354/1747

1.

On the Enumeration Done interrupt (ENUMDNE in OTG_FS_GINTSTS), read the
OTG_FS_DSTS register to determine the enumeration speed.

2.

Program the MPSIZ field in OTG_FS_DIEPCTL0 to set the maximum packet size. This
step configures control endpoint 0. The maximum packet size for a control endpoint
depends on the enumeration speed.

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
At this point, the device is ready to receive SOF packets and is configured to perform control
transfers on control endpoint 0.

Endpoint initialization on SetAddress command
This section describes what the application must do when it receives a SetAddress
command in a SETUP packet.
1.

Program the OTG_FS_DCFG register with the device address received in the
SetAddress command

2.

Program the core to send out a status IN packet

Endpoint initialization on SetConfiguration/SetInterface command
This section describes what the application must do when it receives a SetConfiguration or
SetInterface command in a SETUP packet.
1.

When a SetConfiguration command is received, the application must program the
endpoint registers to configure them with the characteristics of the valid endpoints in
the new configuration.

2.

When a SetInterface command is received, the application must program the endpoint
registers of the endpoints affected by this command.

3.

Some endpoints that were active in the prior configuration or alternate setting are not
valid in the new configuration or alternate setting. These invalid endpoints must be
deactivated.

4.

Unmask the interrupt for each active endpoint and mask the interrupts for all inactive
endpoints in the OTG_FS_DAINTMSK register.

5.

Set up the Data FIFO RAM for each FIFO.

6.

After all required endpoints are configured; the application must program the core to
send a status IN packet.

At this point, the device core is configured to receive and transmit any type of data packet.

Endpoint activation
This section describes the steps required to activate a device endpoint or to configure an
existing device endpoint to a new type.
1.

2.

Program the characteristics of the required endpoint into the following fields of the
OTG_FS_DIEPCTLx register (for IN or bidirectional endpoints) or the
OTG_FS_DOEPCTLx register (for OUT or bidirectional endpoints).
–

Maximum packet size

–

USB active endpoint = 1

–

Endpoint start data toggle (for interrupt and bulk endpoints)

–

Endpoint type

–

TxFIFO number

Once the endpoint is activated, the core starts decoding the tokens addressed to that
endpoint and sends out a valid handshake for each valid token received for the
endpoint.

RM0090 Rev 17

1355/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Endpoint deactivation
This section describes the steps required to deactivate an existing endpoint.
1.

In the endpoint to be deactivated, clear the USB active endpoint bit in the
OTG_FS_DIEPCTLx register (for IN or bidirectional endpoints) or the
OTG_FS_DOEPCTLx register (for OUT or bidirectional endpoints).

2.

Once the endpoint is deactivated, the core ignores tokens addressed to that endpoint,
which results in a timeout on the USB.

Note:

The application must meet the following conditions to set up the device core to handle
traffic:
NPTXFEM and RXFLVLM in the OTG_FS_GINTMSK register must be cleared.

34.17.6

Operational model
SETUP and OUT data transfers
This section describes the internal data flow and application-level operations during data
OUT transfers and SETUP transactions.
•

Packet read

This section describes how to read packets (OUT data and SETUP packets) from the
receive FIFO.

1356/1747

1.

On catching an RXFLVL interrupt (OTG_FS_GINTSTS register), the application must
read the Receive status pop register (OTG_FS_GRXSTSP).

2.

The application can mask the RXFLVL interrupt (in OTG_FS_GINTSTS) by writing to
RXFLVL = 0 (in OTG_FS_GINTMSK), until it has read the packet from the receive
FIFO.

3.

If the received packet’s byte count is not 0, the byte count amount of data is popped
from the receive Data FIFO and stored in memory. If the received packet byte count is
0, no data is popped from the receive data FIFO.

4.

The receive FIFO’s packet status readout indicates one of the following:
a)

Global OUT NAK pattern:
PKTSTS = Global OUT NAK, BCNT = 0x000, EPNUM = Don’t Care (0x0),
DPID = Don’t Care (0b00).
These data indicate that the global OUT NAK bit has taken effect.

b)

SETUP packet pattern:
PKTSTS = SETUP, BCNT = 0x008, EPNUM = Control EP Num, DPID = D0.
These data indicate that a SETUP packet for the specified endpoint is now
available for reading from the receive FIFO.

c)

Setup stage done pattern:
PKTSTS = Setup Stage Done, BCNT = 0x0, EPNUM = Control EP Num,
DPID = Don’t Care (0b00).
These data indicate that the Setup stage for the specified endpoint has completed
and the Data stage has started. After this entry is popped from the receive FIFO,
the core asserts a Setup interrupt on the specified control OUT endpoint.

d)

Data OUT packet pattern:
PKTSTS = DataOUT, BCNT = size of the received data OUT packet (0 ≤ BCNT
≤ 1 024), EPNUM = EPNUM on which the packet was received, DPID = Actual
Data PID.

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
e)

Data transfer completed pattern:
PKTSTS = Data OUT Transfer Done, BCNT = 0x0, EPNUM = OUT EP Num
on which the data transfer is complete, DPID = Don’t Care (0b00).
These data indicate that an OUT data transfer for the specified OUT endpoint has
completed. After this entry is popped from the receive FIFO, the core asserts a
Transfer Completed interrupt on the specified OUT endpoint.

5.

After the data payload is popped from the receive FIFO, the RXFLVL interrupt
(OTG_FS_GINTSTS) must be unmasked.

6.

Steps 1–5 are repeated every time the application detects assertion of the interrupt line
due to RXFLVL in OTG_FS_GINTSTS. Reading an empty receive FIFO can result in
undefined core behavior.

Figure 402 provides a flowchart of the above procedure.
Figure 402. Receive FIFO packet read

wait until RXFLVL in OTG_FS_GINTSTSG

rd_data = rd_reg (OTG_FS_GRXSTSP);

Y

rd_data.BCNT = 0

rcv_out_pkt ()

N

packet
store in
memory

mem[0: word_cnt – 1] =
rd_rxfifo(rd_data.EPNUM,
word_cnt)

word_cnt =
BCNT[11:2]
C
+
(BCNT[1] | BCNT[1])

ai15677b

•

SETUP transactions

This section describes how the core handles SETUP packets and the application’s
sequence for handling SETUP transactions.
•

Application requirements

1.

To receive a SETUP packet, the STUPCNT field (OTG_FS_DOEPTSIZx) in a control
OUT endpoint must be programmed to a non-zero value. When the application
programs the STUPCNT field to a non-zero value, the core receives SETUP packets
and writes them to the receive FIFO, irrespective of the NAK status and EPENA bit
setting in OTG_FS_DOEPCTLx. The STUPCNT field is decremented every time the
control endpoint receives a SETUP packet. If the STUPCNT field is not programmed to
a proper value before receiving a SETUP packet, the core still receives the SETUP
packet and decrements the STUPCNT field, but the application may not be able to

RM0090 Rev 17

1357/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

determine the correct number of SETUP packets received in the Setup stage of a
control transfer.
–
2.

The application must always allocate some extra space in the Receive data FIFO, to be
able to receive up to three SETUP packets on a control endpoint.
–

The space to be reserved is 10 words. Three words are required for the first
SETUP packet, 1 word is required for the Setup stage done word and 6 words are
required to store two extra SETUP packets among all control endpoints.

–

3 words per SETUP packet are required to store 8 bytes of SETUP data and 4
bytes of SETUP status (Setup packet pattern). The core reserves this space in the
receive data FIFO to write SETUP data only, and never uses this space for data
packets.

3.

The application must read the 2 words of the SETUP packet from the receive FIFO.

4.

The application must read and discard the Setup stage done word from the receive
FIFO.

•

Internal data flow

1.

When a SETUP packet is received, the core writes the received data to the receive
FIFO, without checking for available space in the receive FIFO and irrespective of the
endpoint’s NAK and STALL bit settings.
–

2.

3.

The core internally sets the IN NAK and OUT NAK bits for the control IN/OUT
endpoints on which the SETUP packet was received.

For every SETUP packet received on the USB, 3 words of data are written to the
receive FIFO, and the STUPCNT field is decremented by 1.
–

The first word contains control information used internally by the core

–

The second word contains the first 4 bytes of the SETUP command

–

The third word contains the last 4 bytes of the SETUP command

When the Setup stage changes to a Data IN/OUT stage, the core writes an entry
(Setup stage done word) to the receive FIFO, indicating the completion of the Setup
stage.

4.

On the AHB side, SETUP packets are emptied by the application.

5.

When the application pops the Setup stage done word from the receive FIFO, the core
interrupts the application with an STUP interrupt (OTG_FS_DOEPINTx), indicating it
can process the received SETUP packet.
–

•
1.

The core clears the endpoint enable bit for control OUT endpoints.

Application programming sequence
Program the OTG_FS_DOEPTSIZx register.
–

STUPCNT = 3

2.

Wait for the RXFLVL interrupt (OTG_FS_GINTSTS) and empty the data packets from
the receive FIFO.

3.

Assertion of the STUP interrupt (OTG_FS_DOEPINTx) marks a successful completion
of the SETUP Data Transfer.
–

1358/1747

STUPCNT = 3 in OTG_FS_DOEPTSIZx

On this interrupt, the application must read the OTG_FS_DOEPTSIZx register to
determine the number of SETUP packets received and process the last received
SETUP packet.

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
Figure 403. Processing a SETUP packet
Wait for STUP in OTG_FS_DOEPINTx

rem_supcnt =
rd_reg(DOEPTSIZx)

setup_cmd[31:0] = mem[4 – 2 * rem_supcnt]
setup_cmd[63:32] = mem[5 – 2 * rem_supcnt]

Find setup cmd type

Read

ctrl-rd/wr/2 stage

Write

2-stage
setup_np_in_pkt
Data IN phase

setup_np_in_pkt
Status IN phase

rcv_out_pkt
Data OUT phase
ai15678

•

Handling more than three back-to-back SETUP packets

Per the USB 2.0 specification, normally, during a SETUP packet error, a host does not send
more than three back-to-back SETUP packets to the same endpoint. However, the USB 2.0
specification does not limit the number of back-to-back SETUP packets a host can send to
the same endpoint. When this condition occurs, the OTG_FS controller generates an
interrupt (B2BSTUP in OTG_FS_DOEPINTx).
•

Setting the global OUT NAK

Internal data flow
1.

When the application sets the Global OUT NAK (SGONAK bit in OTG_FS_DCTL), the
core stops writing data, except SETUP packets, to the receive FIFO. Irrespective of the
space availability in the receive FIFO, non-isochronous OUT tokens receive a NAK
handshake response, and the core ignores isochronous OUT data packets

2.

The core writes the Global OUT NAK pattern to the receive FIFO. The application must
reserve enough receive FIFO space to write this data pattern.

3.

When the application pops the Global OUT NAK pattern word from the receive FIFO,
the core sets the GONAKEFF interrupt (OTG_FS_GINTSTS).

4.

Once the application detects this interrupt, it can assume that the core is in Global OUT
NAK mode. The application can clear this interrupt by clearing the SGONAK bit in
OTG_FS_DCTL.

RM0090 Rev 17

1359/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Application programming sequence
1.

To stop receiving any kind of data in the receive FIFO, the application must set the
Global OUT NAK bit by programming the following field:
–

SGONAK = 1 in OTG_FS_DCTL

2.

Wait for the assertion of the GONAKEFF interrupt in OTG_FS_GINTSTS. When
asserted, this interrupt indicates that the core has stopped receiving any type of data
except SETUP packets.

3.

The application can receive valid OUT packets after it has set SGONAK in
OTG_FS_DCTL and before the core asserts the GONAKEFF interrupt
(OTG_FS_GINTSTS).

4.

The application can temporarily mask this interrupt by writing to the GINAKEFFM bit in
the OTG_FS_GINTMSK register.
–

5.

Whenever the application is ready to exit the Global OUT NAK mode, it must clear the
SGONAK bit in OTG_FS_DCTL. This also clears the GONAKEFF interrupt
(OTG_FS_GINTSTS).
–

6.

OTG_FS_DCTL = 1 in CGONAK

If the application has masked this interrupt earlier, it must be unmasked as follows:
–

•

GINAKEFFM = 0 in the OTG_FS_GINTMSK register

GINAKEFFM = 1 in GINTMSK

Disabling an OUT endpoint

The application must use this sequence to disable an OUT endpoint that it has enabled.
Application programming sequence
1.

Before disabling any OUT endpoint, the application must enable Global OUT NAK
mode in the core.
–

2.
3.

4.

5.

Wait for the GONAKEFF interrupt (OTG_FS_GINTSTS)
Disable the required OUT endpoint by programming the following fields:
–

EPDIS = 1 in OTG_FS_DOEPCTLx

–

SNAK = 1 in OTG_FS_DOEPCTLx

Wait for the EPDISD interrupt (OTG_FS_DOEPINTx), which indicates that the OUT
endpoint is completely disabled. When the EPDISD interrupt is asserted, the core also
clears the following bits:
–

EPDIS = 0 in OTG_FS_DOEPCTLx

–

EPENA = 0 in OTG_FS_DOEPCTLx

The application must clear the Global OUT NAK bit to start receiving data from other
non-disabled OUT endpoints.
–

•

SGONAK = 1 in OTG_FS_DCTL

SGONAK = 0 in OTG_FS_DCTL

Generic non-isochronous OUT data transfers

This section describes a regular non-isochronous OUT data transfer (control, bulk, or
interrupt).
Application requirements
1.

1360/1747

Before setting up an OUT transfer, the application must allocate a buffer in the memory
to accommodate all data to be received as part of the OUT transfer.

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
2.

3.

For OUT transfers, the transfer size field in the endpoint’s transfer size register must be
a multiple of the maximum packet size of the endpoint, adjusted to the word boundary.
–

transfer size[EPNUM] = n × (MPSIZ[EPNUM] + 4 – (MPSIZ[EPNUM] mod 4))

–

packet count[EPNUM] = n

–

n>0

On any OUT endpoint interrupt, the application must read the endpoint’s transfer size
register to calculate the size of the payload in the memory. The received payload size
can be less than the programmed transfer size.
–

Payload size in memory = application programmed initial transfer size – core
updated final transfer size

–

Number of USB packets in which this payload was received = application
programmed initial packet count – core updated final packet count

Internal data flow
1.

The application must set the transfer size and packet count fields in the endpointspecific registers, clear the NAK bit, and enable the endpoint to receive the data.

2.

Once the NAK bit is cleared, the core starts receiving data and writes it to the receive
FIFO, as long as there is space in the receive FIFO. For every data packet received on
the USB, the data packet and its status are written to the receive FIFO. Every packet
(maximum packet size or short packet) written to the receive FIFO decrements the
packet count field for that endpoint by 1.
–

OUT data packets received with bad data CRC are flushed from the receive FIFO
automatically.

–

After sending an ACK for the packet on the USB, the core discards nonisochronous OUT data packets that the host, which cannot detect the ACK, resends. The application does not detect multiple back-to-back data OUT packets
on the same endpoint with the same data PID. In this case the packet count is not
decremented.

–

If there is no space in the receive FIFO, isochronous or non-isochronous data
packets are ignored and not written to the receive FIFO. Additionally, nonisochronous OUT tokens receive a NAK handshake reply.

–

In all the above three cases, the packet count is not decremented because no data
are written to the receive FIFO.

3.

When the packet count becomes 0 or when a short packet is received on the endpoint,
the NAK bit for that endpoint is set. Once the NAK bit is set, the isochronous or nonisochronous data packets are ignored and not written to the receive FIFO, and nonisochronous OUT tokens receive a NAK handshake reply.

4.

After the data are written to the receive FIFO, the application reads the data from the
receive FIFO and writes it to external memory, one packet at a time per endpoint.

5.

At the end of every packet write on the AHB to external memory, the transfer size for
the endpoint is decremented by the size of the written packet.

6.

The OUT data transfer completed pattern for an OUT endpoint is written to the receive
FIFO on one of the following conditions:

7.

–

The transfer size is 0 and the packet count is 0

–

The last OUT data packet written to the receive FIFO is a short packet
(0 ≤ packet size < maximum packet size)

When either the application pops this entry (OUT data transfer completed), a transfer
completed interrupt is generated for the endpoint and the endpoint enable is cleared.

RM0090 Rev 17

1361/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

Application programming sequence
1.

Program the OTG_FS_DOEPTSIZx register for the transfer size and the corresponding
packet count.

2.

Program the OTG_FS_DOEPCTLx register with the endpoint characteristics, and set
the EPENA and CNAK bits.

3.

–

EPENA = 1 in OTG_FS_DOEPCTLx

–

CNAK = 1 in OTG_FS_DOEPCTLx

Wait for the RXFLVL interrupt (in OTG_FS_GINTSTS) and empty the data packets
from the receive FIFO.
–

This step can be repeated many times, depending on the transfer size.

4.

Asserting the XFRC interrupt (OTG_FS_DOEPINTx) marks a successful completion of
the non-isochronous OUT data transfer.

5.

Read the OTG_FS_DOEPTSIZx register to determine the size of the received data
payload.

•

Generic isochronous OUT data transfer

This section describes a regular isochronous OUT data transfer.
Application requirements
1.

All the application requirements for non-isochronous OUT data transfers also apply to
isochronous OUT data transfers.

2.

For isochronous OUT data transfers, the transfer size and packet count fields must
always be set to the number of maximum-packet-size packets that can be received in a
single frame and no more. Isochronous OUT data transfers cannot span more than 1
frame.

3.

The application must read all isochronous OUT data packets from the receive FIFO
(data and status) before the end of the periodic frame (EOPF interrupt in
OTG_FS_GINTSTS).

4.

To receive data in the following frame, an isochronous OUT endpoint must be enabled
after the EOPF (OTG_FS_GINTSTS) and before the SOF (OTG_FS_GINTSTS).

Internal data flow
1.

The internal data flow for isochronous OUT endpoints is the same as that for nonisochronous OUT endpoints, but for a few differences.

2.

When an isochronous OUT endpoint is enabled by setting the Endpoint Enable and
clearing the NAK bits, the Even/Odd frame bit must also be set appropriately. The core
receives data on an isochronous OUT endpoint in a particular frame only if the
following condition is met:
–

3.

1362/1747

EONUM (in OTG_FS_DOEPCTLx) = SOFFN[0] (in OTG_FS_DSTS)

When the application completely reads an isochronous OUT data packet (data and
status) from the receive FIFO, the core updates the RXDPID field in
OTG_FS_DOEPTSIZx with the data PID of the last isochronous OUT data packet read
from the receive FIFO.

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
Application programming sequence
1.

Program the OTG_FS_DOEPTSIZx register for the transfer size and the corresponding
packet count

2.

Program the OTG_FS_DOEPCTLx register with the endpoint characteristics and set
the Endpoint Enable, ClearNAK, and Even/Odd frame bits.

3.

–

EPENA = 1

–

CNAK = 1

–

EONUM = (0: Even/1: Odd)

Wait for the RXFLVL interrupt (in OTG_FS_GINTSTS) and empty the data packets
from the receive FIFO
–

This step can be repeated many times, depending on the transfer size.

4.

The assertion of the XFRC interrupt (in OTG_FS_DOEPINTx) marks the completion of
the isochronous OUT data transfer. This interrupt does not necessarily mean that the
data in memory are good.

5.

This interrupt cannot always be detected for isochronous OUT transfers. Instead, the
application can detect the IISOOXFRM interrupt in OTG_FS_GINTSTS.

6.

Read the OTG_FS_DOEPTSIZx register to determine the size of the received transfer
and to determine the validity of the data received in the frame. The application must
treat the data received in memory as valid only if one of the following conditions is met:
–

RXDPID = D0 (in OTG_FS_DOEPTSIZx) and the number of USB packets in
which this payload was received = 1

–

RXDPID = D1 (in OTG_FS_DOEPTSIZx) and the number of USB packets in
which this payload was received = 2

–

RXDPID = D2 (in OTG_FS_DOEPTSIZx) and the number of USB packets in
which this payload was received = 3
The number of USB packets in which this payload was received =
Application programmed initial packet count – Core updated final packet count

The application can discard invalid data packets.
•

Incomplete isochronous OUT data transfers

This section describes the application programming sequence when isochronous OUT data
packets are dropped inside the core.
Internal data flow
1.

2.

For isochronous OUT endpoints, the XFRC interrupt (in OTG_FS_DOEPINTx) may not
always be asserted. If the core drops isochronous OUT data packets, the application
could fail to detect the XFRC interrupt (OTG_FS_DOEPINTx) under the following
circumstances:
–

When the receive FIFO cannot accommodate the complete ISO OUT data packet,
the core drops the received ISO OUT data

–

When the isochronous OUT data packet is received with CRC errors

–

When the isochronous OUT token received by the core is corrupted

–

When the application is very slow in reading the data from the receive FIFO

When the core detects an end of periodic frame before transfer completion to all
isochronous OUT endpoints, it asserts the incomplete Isochronous OUT data interrupt
(IISOOXFRM in OTG_FS_GINTSTS), indicating that an XFRC interrupt (in
OTG_FS_DOEPINTx) is not asserted on at least one of the isochronous OUT

RM0090 Rev 17

1363/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

endpoints. At this point, the endpoint with the incomplete transfer remains enabled, but
no active transfers remain in progress on this endpoint on the USB.
Application programming sequence
1.

Asserting the IISOOXFRM interrupt (OTG_FS_GINTSTS) indicates that in the current
frame, at least one isochronous OUT endpoint has an incomplete transfer.

2.

If this occurs because isochronous OUT data is not completely emptied from the
endpoint, the application must ensure that the application empties all isochronous OUT
data (data and status) from the receive FIFO before proceeding.
–

3.

When all data are emptied from the receive FIFO, the application can detect the
XFRC interrupt (OTG_FS_DOEPINTx). In this case, the application must reenable the endpoint to receive isochronous OUT data in the next frame.

When it receives an IISOOXFRM interrupt (in OTG_FS_GINTSTS), the application
must read the control registers of all isochronous OUT endpoints
(OTG_FS_DOEPCTLx) to determine which endpoints had an incomplete transfer in the
current microframe. An endpoint transfer is incomplete if both the following conditions
are met:
–

EONUM bit (in OTG_FS_DOEPCTLx) = SOFFN[0] (in OTG_FS_DSTS)

–

EPENA = 1 (in OTG_FS_DOEPCTLx)

4.

The previous step must be performed before the SOF interrupt (in OTG_FS_GINTSTS)
is detected, to ensure that the current frame number is not changed.

5.

For isochronous OUT endpoints with incomplete transfers, the application must discard
the data in the memory and disable the endpoint by setting the EPDIS bit in
OTG_FS_DOEPCTLx.

6.

Wait for the EPDIS interrupt (in OTG_FS_DOEPINTx) and enable the endpoint to
receive new data in the next frame.
–

•

Because the core can take some time to disable the endpoint, the application may
not be able to receive the data in the next frame after receiving bad isochronous
data.

Stalling a non-isochronous OUT endpoint

This section describes how the application can stall a non-isochronous endpoint.
1.

Put the core in the Global OUT NAK mode.

2.

Disable the required endpoint
–

When disabling the endpoint, instead of setting the SNAK bit in
OTG_FS_DOEPCTL, set STALL = 1 (in OTG_FS_DOEPCTL).
The STALL bit always takes precedence over the NAK bit.

3.

When the application is ready to end the STALL handshake for the endpoint, the
STALL bit (in OTG_FS_DOEPCTLx) must be cleared.

4.

If the application is setting or clearing a STALL for an endpoint due to a
SetFeature.Endpoint Halt or ClearFeature.Endpoint Halt command, the STALL bit must
be set or cleared before the application sets up the Status stage transfer on the control
endpoint.

Examples
This section describes and depicts some fundamental transfer types and scenarios.
•

1364/1747

Bulk OUT transaction

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
Figure 404 depicts the reception of a single Bulk OUT Data packet from the USB to the AHB
and describes the events involved in the process.
Figure 404. Bulk OUT transaction
Host

USB

Application

Device

init_ out_ ep
1

2

XFRSIZ = 64 bytes
PKTCNT = 1

wr_reg (DOEPTSIZx)
EPENA= 1
CNAK = 1

O UT

64 bytes

wr_reg(D OEPCTLx)

3
4

AC K

5

6

xact _1
RXFLVL iintr
T L x.N A
K
=
1
PKTCN
T0

D OE P C

XFRSIZ
=0
r
OU T
NA K

7

idle until intr

rcv_out _pkt()

XF
int r RC

8

On new xfer
or RxFIFO
not empty

idle until intr

ai15679b

After a SetConfiguration/SetInterface command, the application initializes all OUT endpoints
by setting CNAK = 1 and EPENA = 1 (in OTG_FS_DOEPCTLx), and setting a suitable
XFRSIZ and PKTCNT in the OTG_FS_DOEPTSIZx register.
1.

host attempts to send data (OUT token) to an endpoint.

2.

When the core receives the OUT token on the USB, it stores the packet in the RxFIFO
because space is available there.

3.

After writing the complete packet in the RxFIFO, the core then asserts the RXFLVL
interrupt (in OTG_FS_GINTSTS).

4.

On receiving the PKTCNT number of USB packets, the core internally sets the NAK bit
for this endpoint to prevent it from receiving any more packets.

5.

The application processes the interrupt and reads the data from the RxFIFO.

6.

When the application has read all the data (equivalent to XFRSIZ), the core generates
an XFRC interrupt (in OTG_FS_DOEPINTx).

7.

The application processes the interrupt and uses the setting of the XFRC interrupt bit
(in OTG_FS_DOEPINTx) to determine that the intended transfer is complete.

IN data transfers
•

Packet write

This section describes how the application writes data packets to the endpoint FIFO when
dedicated transmit FIFOs are enabled.
1.

The application can either choose the polling or the interrupt mode.

RM0090 Rev 17

1365/1747
1379

USB on-the-go full-speed (OTG_FS)

2.

RM0090

–

In polling mode, the application monitors the status of the endpoint transmit data
FIFO by reading the OTG_FS_DTXFSTSx register, to determine if there is enough
space in the data FIFO.

–

In interrupt mode, the application waits for the TXFE interrupt (in
OTG_FS_DIEPINTx) and then reads the OTG_FS_DTXFSTSx register, to
determine if there is enough space in the data FIFO.

–

To write a single non-zero length data packet, there must be space to write the
entire packet in the data FIFO.

–

To write zero length packet, the application must not look at the FIFO space.

Using one of the above mentioned methods, when the application determines that
there is enough space to write a transmit packet, the application must first write into the
endpoint control register, before writing the data into the data FIFO. Typically, the
application, must do a read modify write on the OTG_FS_DIEPCTLx register to avoid
modifying the contents of the register, except for setting the Endpoint Enable bit.

The application can write multiple packets for the same endpoint into the transmit FIFO, if
space is available. For periodic IN endpoints, the application must write packets only for one
microframe. It can write packets for the next periodic transaction only after getting transfer
complete for the previous transaction.
•

Setting IN endpoint NAK

Internal data flow
1.

2.

When the application sets the IN NAK for a particular endpoint, the core stops
transmitting data on the endpoint, irrespective of data availability in the endpoint’s
transmit FIFO.
Non-isochronous IN tokens receive a NAK handshake reply
–

Isochronous IN tokens receive a zero-data-length packet reply

3.

The core asserts the INEPNE (IN endpoint NAK effective) interrupt in
OTG_FS_DIEPINTx in response to the SNAK bit in OTG_FS_DIEPCTLx.

4.

Once this interrupt is seen by the application, the application can assume that the
endpoint is in IN NAK mode. This interrupt can be cleared by the application by setting
the CNAK bit in OTG_FS_DIEPCTLx.

Application programming sequence
1.

To stop transmitting any data on a particular IN endpoint, the application must set the
IN NAK bit. To set this bit, the following field must be programmed.
–

2.

Wait for assertion of the INEPNE interrupt in OTG_FS_DIEPINTx. This interrupt
indicates that the core has stopped transmitting data on the endpoint.

3.

The core can transmit valid IN data on the endpoint after the application has set the
NAK bit, but before the assertion of the NAK Effective interrupt.

4.

The application can mask this interrupt temporarily by writing to the INEPNEM bit in
DIEPMSK.
–

5.

6.

INEPNEM = 0 in DIEPMSK

To exit Endpoint NAK mode, the application must clear the NAK status bit (NAKSTS) in
OTG_FS_DIEPCTLx. This also clears the INEPNE interrupt (in OTG_FS_DIEPINTx).
–

1366/1747

SNAK = 1 in OTG_FS_DIEPCTLx

CNAK = 1 in OTG_FS_DIEPCTLx

If the application masked this interrupt earlier, it must be unmasked as follows:

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
–
•

INEPNEM = 1 in DIEPMSK

IN endpoint disable

Use the following sequence to disable a specific IN endpoint that has been previously
enabled.
Application programming sequence
1.
2.

The application must stop writing data on the AHB for the IN endpoint to be disabled.
The application must set the endpoint in NAK mode.
–

SNAK = 1 in OTG_FS_DIEPCTLx

3.

Wait for the INEPNE interrupt in OTG_FS_DIEPINTx.

4.

Set the following bits in the OTG_FS_DIEPCTLx register for the endpoint that must be
disabled.

5.

–

EPDIS = 1 in OTG_FS_DIEPCTLx

–

SNAK = 1 in OTG_FS_DIEPCTLx

Assertion of the EPDISD interrupt in OTG_FS_DIEPINTx indicates that the core has
completely disabled the specified endpoint. Along with the assertion of the interrupt, the
core also clears the following bits:
–

EPENA = 0 in OTG_FS_DIEPCTLx

–

EPDIS = 0 in OTG_FS_DIEPCTLx

6.

The application must read the OTG_FS_DIEPTSIZx register for the periodic IN EP, to
calculate how much data on the endpoint were transmitted on the USB.

7.

The application must flush the data in the Endpoint transmit FIFO, by setting the
following fields in the OTG_FS_GRSTCTL register:
–

TXFNUM (in OTG_FS_GRSTCTL) = Endpoint transmit FIFO number

–

TXFFLSH in (OTG_FS_GRSTCTL) = 1

The application must poll the OTG_FS_GRSTCTL register, until the TXFFLSH bit is cleared
by the core, which indicates the end of flush operation. To transmit new data on this
endpoint, the application can re-enable the endpoint at a later point.
•

Generic non-periodic IN data transfers

Application requirements
1.

Before setting up an IN transfer, the application must ensure that all data to be
transmitted as part of the IN transfer are part of a single buffer.

2.

For IN transfers, the Transfer Size field in the Endpoint Transfer Size register denotes a
payload that constitutes multiple maximum-packet-size packets and a single short
packet. This short packet is transmitted at the end of the transfer.
–

To transmit a few maximum-packet-size packets and a short packet at the end of
the transfer:
Transfer size[EPNUM] = x × MPSIZ[EPNUM] + sp
If (sp > 0), then packet count[EPNUM] = x + 1.
Otherwise, packet count[EPNUM] = x

–

To transmit a single zero-length data packet:
Transfer size[EPNUM] = 0
Packet count[EPNUM] = 1

–

To transmit a few maximum-packet-size packets and a zero-length data packet at
the end of the transfer, the application must split the transfer into two parts. The

RM0090 Rev 17

1367/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

first sends maximum-packet-size data packets and the second sends the zerolength data packet alone.
First transfer: transfer size[EPNUM] = x × MPSIZ[epnum]; packet count = n;
Second transfer: transfer size[EPNUM] = 0; packet count = 1;
3.

Once an endpoint is enabled for data transfers, the core updates the Transfer size
register. At the end of the IN transfer, the application must read the Transfer size
register to determine how much data posted in the transmit FIFO have already been
sent on the USB.

4.

Data fetched into transmit FIFO = Application-programmed initial transfer size – coreupdated final transfer size
–

Data transmitted on USB = (application-programmed initial packet count – Core
updated final packet count) × MPSIZ[EPNUM]

–

Data yet to be transmitted on USB = (Application-programmed initial transfer size
– data transmitted on USB)

Internal data flow
1.

The application must set the transfer size and packet count fields in the endpointspecific registers and enable the endpoint to transmit the data.

2.

The application must also write the required data to the transmit FIFO for the endpoint.

3.

Every time a packet is written into the transmit FIFO by the application, the transfer size
for that endpoint is decremented by the packet size. The data is fetched from the
memory by the application, until the transfer size for the endpoint becomes 0. After
writing the data into the FIFO, the “number of packets in FIFO” count is incremented
(this is a 3-bit count, internally maintained by the core for each IN endpoint transmit
FIFO. The maximum number of packets maintained by the core at any time in an IN
endpoint FIFO is eight). For zero-length packets, a separate flag is set for each FIFO,
without any data in the FIFO.

4.

Once the data are written to the transmit FIFO, the core reads them out upon receiving
an IN token. For every non-isochronous IN data packet transmitted with an ACK
handshake, the packet count for the endpoint is decremented by one, until the packet
count is zero. The packet count is not decremented on a timeout.

5.

For zero length packets (indicated by an internal zero length flag), the core sends out a
zero-length packet for the IN token and decrements the packet count field.

6.

If there are no data in the FIFO for a received IN token and the packet count field for
that endpoint is zero, the core generates an “IN token received when TxFIFO is empty”
(ITTXFE) Interrupt for the endpoint, provided that the endpoint NAK bit is not set. The
core responds with a NAK handshake for non-isochronous endpoints on the USB.

7.

The core internally rewinds the FIFO pointers and no timeout interrupt is generated.

8.

When the transfer size is 0 and the packet count is 0, the transfer complete (XFRC)
interrupt for the endpoint is generated and the endpoint enable is cleared.

Application programming sequence

1368/1747

1.

Program the OTG_FS_DIEPTSIZx register with the transfer size and corresponding
packet count.

2.

Program the OTG_FS_DIEPCTLx register with the endpoint characteristics and set the
CNAK and EPENA (Endpoint Enable) bits.

3.

When transmitting non-zero length data packet, the application must poll the
OTG_FS_DTXFSTSx register (where x is the FIFO number associated with that

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
endpoint) to determine whether there is enough space in the data FIFO. The
application can optionally use TXFE (in OTG_FS_DIEPINTx) before writing the data.
•

Generic periodic IN data transfers

This section describes a typical periodic IN data transfer.
Application requirements
1.

Application requirements 1, 2, 3, and 4 of Generic non-periodic IN data transfers on
page 1367 also apply to periodic IN data transfers, except for a slight modification of
requirement 2.
–

The application can only transmit multiples of maximum-packet-size data packets
or multiples of maximum-packet-size packets, plus a short packet at the end. To
transmit a few maximum-packet-size packets and a short packet at the end of the
transfer, the following conditions must be met:
transfer size[EPNUM] = x × MPSIZ[EPNUM] + sp
(where x is an integer ≥ 0, and 0 ≤ sp < MPSIZ[EPNUM])
If (sp > 0), packet count[EPNUM] = x + 1
Otherwise, packet count[EPNUM] = x;
MCNT[EPNUM] = packet count[EPNUM]

–

The application cannot transmit a zero-length data packet at the end of a transfer.
It can transmit a single zero-length data packet by itself. To transmit a single zerolength data packet:

–

transfer size[EPNUM] = 0
packet count[EPNUM] = 1
MCNT[EPNUM] = packet count[EPNUM]

2.

3.

The application can only schedule data transfers one frame at a time.
–

(MCNT – 1) × MPSIZ ≤ XFERSIZ ≤ MCNT × MPSIZ

–

PKTCNT = MCNT (in OTG_FS_DIEPTSIZx)

–

If XFERSIZ < MCNT × MPSIZ, the last data packet of the transfer is a short
packet.

–

Note that: MCNT is in OTG_FS_DIEPTSIZx, MPSIZ is in OTG_FS_DIEPCTLx,
PKTCNT is in OTG_FS_DIEPTSIZx and XFERSIZ is in OTG_FS_DIEPTSIZx

The complete data to be transmitted in the frame must be written into the transmit FIFO
by the application, before the IN token is received. Even when 1 word of the data to be
transmitted per frame is missing in the transmit FIFO when the IN token is received, the
core behaves as when the FIFO is empty. When the transmit FIFO is empty:
–

A zero data length packet would be transmitted on the USB for isochronous IN
endpoints

–

A NAK handshake would be transmitted on the USB for interrupt IN endpoints

Internal data flow

RM0090 Rev 17

1369/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

1.

The application must set the transfer size and packet count fields in the endpointspecific registers and enable the endpoint to transmit the data.

2.

The application must also write the required data to the associated transmit FIFO for
the endpoint.

3.

Every time the application writes a packet to the transmit FIFO, the transfer size for that
endpoint is decremented by the packet size. The data are fetched from application
memory until the transfer size for the endpoint becomes 0.

4.

When an IN token is received for a periodic endpoint, the core transmits the data in the
FIFO, if available. If the complete data payload (complete packet, in dedicated FIFO
mode) for the frame is not present in the FIFO, then the core generates an IN token
received when TxFIFO empty interrupt for the endpoint.

5.

6.

–

A zero-length data packet is transmitted on the USB for isochronous IN endpoints

–

A NAK handshake is transmitted on the USB for interrupt IN endpoints

The packet count for the endpoint is decremented by 1 under the following conditions:
–

For isochronous endpoints, when a zero- or non-zero-length data packet is
transmitted

–

For interrupt endpoints, when an ACK handshake is transmitted

–

When the transfer size and packet count are both 0, the transfer completed
interrupt for the endpoint is generated and the endpoint enable is cleared.

At the “Periodic frame Interval” (controlled by PFIVL in OTG_FS_DCFG), when the
core finds non-empty any of the isochronous IN endpoint FIFOs scheduled for the
current frame non-empty, the core generates an IISOIXFR interrupt in
OTG_FS_GINTSTS.

Application programming sequence
1.

Program the OTG_FS_DIEPCTLx register with the endpoint characteristics and set the
CNAK and EPENA bits.

2.

Write the data to be transmitted in the next frame to the transmit FIFO.

3.

Asserting the ITTXFE interrupt (in OTG_FS_DIEPINTx) indicates that the application
has not yet written all data to be transmitted to the transmit FIFO.

4.

If the interrupt endpoint is already enabled when this interrupt is detected, ignore the
interrupt. If it is not enabled, enable the endpoint so that the data can be transmitted on
the next IN token attempt.

5.

Asserting the XFRC interrupt (in OTG_FS_DIEPINTx) with no ITTXFE interrupt in
OTG_FS_DIEPINTx indicates the successful completion of an isochronous IN transfer.
A read to the OTG_FS_DIEPTSIZx register must give transfer size = 0 and packet
count = 0, indicating all data were transmitted on the USB.

6.

Asserting the XFRC interrupt (in OTG_FS_DIEPINTx), with or without the ITTXFE
interrupt (in OTG_FS_DIEPINTx), indicates the successful completion of an interrupt
IN transfer. A read to the OTG_FS_DIEPTSIZx register must give transfer size = 0 and
packet count = 0, indicating all data were transmitted on the USB.

7.

Asserting the incomplete isochronous IN transfer (IISOIXFR) interrupt in
OTG_FS_GINTSTS with none of the aforementioned interrupts indicates the core did
not receive at least 1 periodic IN token in the current frame.

•

Incomplete isochronous IN data transfers

This section describes what the application must do on an incomplete isochronous IN data
transfer.

1370/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
Internal data flow
1.

An isochronous IN transfer is treated as incomplete in one of the following conditions:
a)

The core receives a corrupted isochronous IN token on at least one isochronous
IN endpoint. In this case, the application detects an incomplete isochronous IN
transfer interrupt (IISOIXFR in OTG_FS_GINTSTS).

b)

The application is slow to write the complete data payload to the transmit FIFO
and an IN token is received before the complete data payload is written to the
FIFO. In this case, the application detects an IN token received when TxFIFO
empty interrupt in OTG_FS_DIEPINTx. The application can ignore this interrupt,
as it eventually results in an incomplete isochronous IN transfer interrupt
(IISOIXFR in OTG_FS_GINTSTS) at the end of periodic frame.
The core transmits a zero-length data packet on the USB in response to the
received IN token.

2.

The application must stop writing the data payload to the transmit FIFO as soon as
possible.

3.

The application must set the NAK bit and the disable bit for the endpoint.

4.

The core disables the endpoint, clears the disable bit, and asserts the Endpoint Disable
interrupt for the endpoint.

Application programming sequence
1.

The application can ignore the IN token received when TxFIFO empty interrupt in
OTG_FS_DIEPINTx on any isochronous IN endpoint, as it eventually results in an
incomplete isochronous IN transfer interrupt (in OTG_FS_GINTSTS).

2.

Assertion of the incomplete isochronous IN transfer interrupt (in OTG_FS_GINTSTS)
indicates an incomplete isochronous IN transfer on at least one of the isochronous IN
endpoints.

3.

The application must read the Endpoint Control register for all isochronous IN
endpoints to detect endpoints with incomplete IN data transfers.

4.

The application must stop writing data to the Periodic Transmit FIFOs associated with
these endpoints on the AHB.

5.

Program the following fields in the OTG_FS_DIEPCTLx register to disable the
endpoint:

6.

–

SNAK = 1 in OTG_FS_DIEPCTLx

–

EPDIS = 1 in OTG_FS_DIEPCTLx

The assertion of the Endpoint Disabled interrupt in OTG_FS_DIEPINTx indicates that
the core has disabled the endpoint.
–

•

At this point, the application must flush the data in the associated transmit FIFO or
overwrite the existing data in the FIFO by enabling the endpoint for a new transfer
in the next microframe. To flush the data, the application must use the
OTG_FS_GRSTCTL register.

Stalling non-isochronous IN endpoints

This section describes how the application can stall a non-isochronous endpoint.
Application programming sequence

RM0090 Rev 17

1371/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

1.

Disable the IN endpoint to be stalled. Set the STALL bit as well.

2.

EPDIS = 1 in OTG_FS_DIEPCTLx, when the endpoint is already enabled
–

STALL = 1 in OTG_FS_DIEPCTLx

–

The STALL bit always takes precedence over the NAK bit

3.

Assertion of the Endpoint Disabled interrupt (in OTG_FS_DIEPINTx) indicates to the
application that the core has disabled the specified endpoint.

4.

The application must flush the non-periodic or periodic transmit FIFO, depending on
the endpoint type. In case of a non-periodic endpoint, the application must re-enable
the other non-periodic endpoints that do not need to be stalled, to transmit data.

5.

Whenever the application is ready to end the STALL handshake for the endpoint, the
STALL bit must be cleared in OTG_FS_DIEPCTLx.

6.

If the application sets or clears a STALL bit for an endpoint due to a
SetFeature.Endpoint Halt command or ClearFeature.Endpoint Halt command, the
STALL bit must be set or cleared before the application sets up the Status stage
transfer on the control endpoint.

Special case: stalling the control OUT endpoint
The core must stall IN/OUT tokens if, during the data stage of a control transfer, the host
sends more IN/OUT tokens than are specified in the SETUP packet. In this case, the
application must enable the ITTXFE interrupt in OTG_FS_DIEPINTx and the OTEPDIS
interrupt in OTG_FS_DOEPINTx during the data stage of the control transfer, after the core
has transferred the amount of data specified in the SETUP packet. Then, when the
application receives this interrupt, it must set the STALL bit in the corresponding endpoint
control register, and clear this interrupt.

34.17.7

Worst case response time
When the OTG_FS controller acts as a device, there is a worst case response time for any
tokens that follow an isochronous OUT. This worst case response time depends on the AHB
clock frequency.
The core registers are in the AHB domain, and the core does not accept another token
before updating these register values. The worst case is for any token following an
isochronous OUT, because for an isochronous transaction, there is no handshake and the
next token could come sooner. This worst case value is 7 PHY clocks when the AHB clock
is the same as the PHY clock. When the AHB clock is faster, this value is smaller.
If this worst case condition occurs, the core responds to bulk/interrupt tokens with a NAK
and drops isochronous and SETUP tokens. The host interprets this as a timeout condition
for SETUP and retries the SETUP packet. For isochronous transfers, the Incomplete
isochronous IN transfer interrupt (IISOIXFR) and Incomplete isochronous OUT transfer
interrupt (IISOOXFR) inform the application that isochronous IN/OUT packets were
dropped.

Choosing the value of TRDT in OTG_FS_GUSBCFG
The value in TRDT (OTG_FS_GUSBCFG) is the time it takes for the MAC, in terms of PHY
clocks after it has received an IN token, to get the FIFO status, and thus the first data from
the PFC block. This time involves the synchronization delay between the PHY and AHB
clocks. The worst case delay for this is when the AHB clock is the same as the PHY clock.
In this case, the delay is 5 clocks.

1372/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
Once the MAC receives an IN token, this information (token received) is synchronized to the
AHB clock by the PFC (the PFC runs on the AHB clock). The PFC then reads the data from
the SPRAM and writes them into the dual clock source buffer. The MAC then reads the data
out of the source buffer (4 deep).
If the AHB is running at a higher frequency than the PHY, the application can use a smaller
value for TRDT (in OTG_FS_GUSBCFG).
Figure 405 has the following signals:
•

tkn_rcvd: Token received information from MAC to PFC

•

dynced_tkn_rcvd: Doubled sync tkn_rcvd, from PCLK to HCLK domain

•

spr_read: Read to SPRAM

•

spr_addr: Address to SPRAM

•

spr_rdata: Read data from SPRAM

•

srcbuf_push: Push to the source buffer

•

srcbuf_rdata: Read data from the source buffer. Data seen by MAC

Refer to Table 203: TRDT values for the values of TRDT versus AHB clock frequency.
Figure 405. TRDT max timing case
0ns

1

50ns

2

3

100ns

4

150ns

5

6

200ns

7

8

HCLK

PCLK

tkn_rcvd

dsynced_tkn_rcvd

spr_read

spr_addr

A1

D1

spr_rdata

srcbuf_push

srcbuf_rdata

D1

5 Clocks
ai15680

RM0090 Rev 17

1373/1747
1379

USB on-the-go full-speed (OTG_FS)

34.17.8

RM0090

OTG programming model
The OTG_FS controller is an OTG device supporting HNP and SRP. When the core is
connected to an “A” plug, it is referred to as an A-device. When the core is connected to a
“B” plug it is referred to as a B-device. In host mode, the OTG_FS controller turns off VBUS
to conserve power. SRP is a method by which the B-device signals the A-device to turn on
VBUS power. A device must perform both data-line pulsing and VBUS pulsing, but a host can
detect either data-line pulsing or VBUS pulsing for SRP. HNP is a method by which the Bdevice negotiates and switches to host role. In Negotiated mode after HNP, the B-device
suspends the bus and reverts to the device role.

A-device session request protocol
The application must set the SRP-capable bit in the Core USB configuration register. This
enables the OTG_FS controller to detect SRP as an A-device.
Figure 406. A-device SRP
Suspend
DRV_VBUS

6
1
5

2
VBUS_VALID

VBUS pulsing

A_VALID

3

D+

D-

4
Data line pulsing

7
Connect

Low
ai15681

1. DRV_VBUS = VBUS drive signal to the PHY
VBUS_VALID = VBUS valid signal from PHY
A_VALID = A-peripheral VBUS level signal to PHY
D+ = Data plus line
D- = Data minus line

1374/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
1.

To save power, the application suspends and turns off port power when the bus is idle
by writing the port suspend and port power bits in the host port control and status
register.

2.

PHY indicates port power off by deasserting the VBUS_VALID signal.

3.

The device must detect SE0 for at least 2 ms to start SRP when VBUS power is off.

4.

To initiate SRP, the device turns on its data line pull-up resistor for 5 to 10 ms. The
OTG_FS controller detects data-line pulsing.

5.

The device drives VBUS above the A-device session valid (2.0 V minimum) for VBUS
pulsing.
The OTG_FS controller interrupts the application on detecting SRP. The Session
request detected bit is set in Global interrupt status register (SRQINT set in
OTG_FS_GINTSTS).

6.

The application must service the Session request detected interrupt and turn on the
port power bit by writing the port power bit in the host port control and status register.
The PHY indicates port power-on by asserting the VBUS_VALID signal.

7.

When the USB is powered, the device connects, completing the SRP process.

B-device session request protocol
The application must set the SRP-capable bit in the Core USB configuration register. This
enables the OTG_FS controller to initiate SRP as a B-device. SRP is a means by which the
OTG_FS controller can request a new session from the host.
Figure 407. B-device SRP
Suspend
VBUS_VALID

6
1

2

B_VALID

3
DISCHRG_VBUS
4

SESS_END

5

DP

DM

8
Data line pulsing

Connect

Low
7
VBUS pulsing

CHRG_VBUS

ai15682

1. VBUS_VALID = VBUS valid signal from PHY
B_VALID = B-peripheral valid session to PHY
DISCHRG_VBUS = discharge signal to PHY
SESS_END = session end signal to PHY
CHRG_VBUS = charge VBUS signal to PHY
DP = Data plus line
DM = Data minus line

RM0090 Rev 17

1375/1747
1379

USB on-the-go full-speed (OTG_FS)
1.

RM0090

To save power, the host suspends and turns off port power when the bus is idle.
The OTG_FS controller sets the early suspend bit in the Core interrupt register after 3
ms of bus idleness. Following this, the OTG_FS controller sets the USB suspend bit in
the Core interrupt register.
The OTG_FS controller informs the PHY to discharge VBUS.

2.

The PHY indicates the session’s end to the device. This is the initial condition for SRP.
The OTG_FS controller requires 2 ms of SE0 before initiating SRP.
For a USB 1.1 full-speed serial transceiver, the application must wait until VBUS
discharges to 0.2 V after BSVLD (in OTG_FS_GOTGCTL) is deasserted. This
discharge time can be obtained from the transceiver vendor and varies from one
transceiver to another.

3.

The USB OTG core informs the PHY to speed up VBUS discharge.

4.

The application initiates SRP by writing the session request bit in the OTG Control and
status register. The OTG_FS controller perform data-line pulsing followed by VBUS
pulsing.

5.

The host detects SRP from either the data-line or VBUS pulsing, and turns on VBUS.
The PHY indicates VBUS power-on to the device.

6.

The OTG_FS controller performs VBUS pulsing.
The host starts a new session by turning on VBUS, indicating SRP success. The
OTG_FS controller interrupts the application by setting the session request success
status change bit in the OTG interrupt status register. The application reads the session
request success bit in the OTG control and status register.

7.

When the USB is powered, the OTG_FS controller connects, completing the SRP
process.

A-device host negotiation protocol
HNP switches the USB host role from the A-device to the B-device. The application must set
the HNP-capable bit in the Core USB configuration register to enable the OTG_FS
controller to perform HNP as an A-device.
Figure 408. A-device HNP
1
OTG core

Host

Device

Suspend 2
DP

4
3

Host
6

5
Reset

DM

Traffic

8
7

Connect

Traffic

DPPULLDOWN

DMPULLDOWN
ai15683

1. DPPULLDOWN = signal from core to PHY to enable/disable the pull-down on the DP line inside the PHY.
DMPULLDOWN = signal from core to PHY to enable/disable the pull-down on the DM line inside the PHY.

1376/1747

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
1.

The OTG_FS controller sends the B-device a SetFeature b_hnp_enable descriptor to
enable HNP support. The B-device’s ACK response indicates that the B-device
supports HNP. The application must set host Set HNP Enable bit in the OTG Control
and status register to indicate to the OTG_FS controller that the B-device supports
HNP.

2.

When it has finished using the bus, the application suspends by writing the Port
suspend bit in the host port control and status register.

3.

When the B-device observes a USB suspend, it disconnects, indicating the initial
condition for HNP. The B-device initiates HNP only when it must switch to the host role;
otherwise, the bus continues to be suspended.
The OTG_FS controller sets the host negotiation detected interrupt in the OTG
interrupt status register, indicating the start of HNP.
The OTG_FS controller deasserts the DM pull down and DM pull down in the PHY to
indicate a device role. The PHY enables the OTG_FS_DP pull-up resistor to indicate a
connect for B-device.
The application must read the current mode bit in the OTG Control and status register
to determine device mode operation.

4.

The B-device detects the connection, issues a USB reset, and enumerates the
OTG_FS controller for data traffic.

5.

The B-device continues the host role, initiating traffic, and suspends the bus when
done.
The OTG_FS controller sets the early suspend bit in the Core interrupt register after 3
ms of bus idleness. Following this, the OTG_FS controller sets the USB Suspend bit in
the Core interrupt register.

6.

In Negotiated mode, the OTG_FS controller detects the suspend, disconnects, and
switches back to the host role. The OTG_FS controller asserts the DM pull down and
DM pull down in the PHY to indicate its assumption of the host role.

7.

The OTG_FS controller sets the Connector ID status change interrupt in the OTG
Interrupt Status register. The application must read the connector ID status in the OTG
Control and Status register to determine the OTG_FS controller operation as an Adevice. This indicates the completion of HNP to the application. The application must
read the Current mode bit in the OTG control and status register to determine host
mode operation.

8.

The B-device connects, completing the HNP process.

RM0090 Rev 17

1377/1747
1379

USB on-the-go full-speed (OTG_FS)

RM0090

B-device host negotiation protocol
HNP switches the USB host role from B-device to A-device. The application must set the
HNP-capable bit in the Core USB configuration register to enable the OTG_FS controller to
perform HNP as a B-device.
Figure 409. B-device HNP
1
OTG core

Device

Host

Suspend 2
DP

4
3

Device
6

5
Reset

DM

Traffic

8
7

Connect

Traffic

DPPULLDOWN

DMPULLDOWN
ai15684

1. DPPULLDOWN = signal from core to PHY to enable/disable the pull-down on the DP line inside the PHY.
DMPULLDOWN = signal from core to PHY to enable/disable the pull-down on the DM line inside the PHY.

1.

The A-device sends the SetFeature b_hnp_enable descriptor to enable HNP support.
The OTG_FS controller’s ACK response indicates that it supports HNP. The application
must set the device HNP enable bit in the OTG Control and status register to indicate
HNP support.
The application sets the HNP request bit in the OTG Control and status register to
indicate to the OTG_FS controller to initiate HNP.

2.

When it has finished using the bus, the A-device suspends by writing the Port suspend
bit in the host port control and status register.
The OTG_FS controller sets the Early suspend bit in the Core interrupt register after 3
ms of bus idleness. Following this, the OTG_FS controller sets the USB suspend bit in
the Core interrupt register.
The OTG_FS controller disconnects and the A-device detects SE0 on the bus,
indicating HNP. The OTG_FS controller asserts the DP pull down and DM pull down in
the PHY to indicate its assumption of the host role.
The A-device responds by activating its OTG_FS_DP pull-up resistor within 3 ms of
detecting SE0. The OTG_FS controller detects this as a connect.
The OTG_FS controller sets the host negotiation success status change interrupt in the
OTG Interrupt status register, indicating the HNP status. The application must read the
host negotiation success bit in the OTG Control and status register to determine host
negotiation success. The application must read the current Mode bit in the Core
interrupt register (OTG_FS_GINTSTS) to determine host mode operation.

3.

1378/1747

The application sets the reset bit (PRST in OTG_FS_HPRT) and the OTG_FS
controller issues a USB reset and enumerates the A-device for data traffic.

RM0090 Rev 17

RM0090

USB on-the-go full-speed (OTG_FS)
4.

The OTG_FS controller continues the host role of initiating traffic, and when done,
suspends the bus by writing the Port suspend bit in the host port control and status
register.

5.

In Negotiated mode, when the A-device detects a suspend, it disconnects and switches
back to the host role. The OTG_FS controller deasserts the DP pull down and DM pull
down in the PHY to indicate the assumption of the device role.

6.

The application must read the current mode bit in the Core interrupt
(OTG_FS_GINTSTS) register to determine the host mode operation.

7.

The OTG_FS controller connects, completing the HNP process.

RM0090 Rev 17

1379/1747
1379

USB on-the-go high-speed (OTG_HS)

35

RM0090

USB on-the-go high-speed (OTG_HS)
This section applies to the whole STM32F4xx family, unless otherwise specified.

35.1

OTG_HS introduction
Portions Copyright (c) 2004, 2005 Synopsys, Inc. All rights reserved. Used with permission.
This section presents the architecture and the programming model of the OTG_HS
controller.
The following acronyms are used throughout the section:
FS

full-speed

HS

High-speed

LS

Low-speed

USB

Universal serial bus

OTG

On-the-go

PHY

Physical layer

MAC

Media access controller

PFC

Packet FIFO controller

UTMI

USB Transceiver Macrocell Interface

ULPI

UTMI+ Low Pin Interface

References are made to the following documents:
•

USB On-The-Go Supplement, Revision 1.3

•

Universal Serial Bus Revision 2.0 Specification

The OTG_HS is a dual-role device (DRD) controller that supports both peripheral and host
functions and is fully compliant with the On-The-Go Supplement to the USB 2.0
Specification. It can also be configured as a host-only or peripheral-only controller, fully
compliant with the USB 2.0 Specification. In host mode, the OTG_HS supports high-speed
(HS, 480 Mbits/s), full-speed (FS, 12 Mbits/s) and low-speed (LS, 1.5 Mbits/s) transfers
whereas in peripheral mode, it only supports high-speed (HS, 480Mbits/s) and full-speed
(FS, 12 Mbits/s) transfers. The OTG_HS supports both HNP and SRP. The only external
device required is a charge pump for VBUS in OTG mode.

35.2

OTG_HS main features
The main features can be divided into three categories: general, host-mode and peripheralmode features.

1380/1747

RM0090 Rev 17

RM0090

35.2.1

USB on-the-go high-speed (OTG_HS)

General features
The OTG_HS interface main features are the following:
•

It is USB-IF certified in compliance with the Universal Serial Bus Revision 2.0
Specification

•

It supports 3 PHY interfaces
–

An on-chip full-speed PHY

–

An ULPI interface for external high-speed PHY.

•

It supports the host negotiation protocol (HNP) and the session request protocol (SRP)

•

It allows the host to turn VBUS off to save power in OTG applications, with no need for
external components

•

It allows to monitor VBUS levels using internal comparators

•

It supports dynamic host-peripheral role switching

•

It is software-configurable to operate as:

•

–

An SRP-capable USB HS/FS peripheral (B-device)

–

An SRP-capable USB HS/FS/low-speed host (A-device)

–

An USB OTG FS dual-role device

It supports HS/FS SOFs as well as low-speed (LS) keep-alive tokens with:
–

SOF pulse PAD output capability

–

SOF pulse internal connection to timer 2 (TIM2)

–

Configurable framing period

–

Configurable end-of-frame interrupt

•

It embeds an internal DMA with shareholding support and software selectable AHB
burst type in DMA mode

•

It has power saving features such as system clock stop during USB suspend, switching
off of the digital core internal clock domains, PHY and DFIFO power management

•

It features a dedicated 4-Kbyte data RAM with advanced FIFO management:

•

–

The memory partition can be configured into different FIFOs to allow flexible and
efficient use of RAM

–

Each FIFO can contain multiple packets

–

Memory allocation is performed dynamically

–

The FIFO size can be configured to values that are not powers of 2 to allow the
use of contiguous memory locations

It ensures a maximum USB bandwidth of up to one frame without application
intervention

RM0090 Rev 17

1381/1747
1541

USB on-the-go high-speed (OTG_HS)

35.2.2

RM0090

Host-mode features
The OTG_HS interface features in host mode are the following:

35.2.3

•

It requires an external charge pump to generate VBUS

•

It has up to 12 host channels (pipes), each channel being dynamically reconfigurable to
support any kind of USB transfer

•

It features a built-in hardware scheduler holding:
–

Up to 8 interrupt plus isochronous transfer requests in the periodic hardware
queue

–

Up to 8 control plus bulk transfer requests in the nonperiodic hardware queue

•

It manages a shared RX FIFO, a periodic TX FIFO, and a nonperiodic TX FIFO for
efficient usage of the USB data RAM

•

It features dynamic trimming capability of SOF framing period in host mode.

Peripheral-mode features
The OTG_HS interface main features in peripheral mode are the following:

35.3

•

It has 1 bidirectional control endpoint 0

•

It has 5 IN endpoints (EP) configurable to support bulk, interrupt or isochronous
transfers

•

It has 5 OUT endpoints configurable to support bulk, interrupt or isochronous transfers

•

It manages a shared Rx FIFO and a Tx-OUT FIFO for efficient usage of the USB data
RAM

•

It manages up to 6 dedicated Tx-IN FIFOs (one for each IN-configured EP) to reduce
the application load

•

It features soft disconnect capability

OTG_HS functional description
Figure 410 shows the OTG_HS interface block diagram.

1382/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
Figure 410. USB OTG interface block diagram

27*B+6B'3

27*)63+<
WUDQVFHLYHU

27*B+6B'0

VHULDO

3HULSKHUDO

$+% DSSOLFDWLRQEXV

0HPRU\

27*B+6B,'

27*B+6
86%27*+6FRUH

&38

27*GHWHFWLRQV

$+%
PDVWHU
LQWHUIDFH

27*B+6B9%86

(;7,

,QWHUUXSWDV\QFZDNHXS
,QWHUUXSW(3RXW
,QWHUUXSW(3LQ

$+%
VODYH
LQWHUIDFH

8/3,B&.
8/3,B',5
8/3,B673
8/3,B1;7
8/3,B'

19,&

,QWHUUXSWJOREDO

86% ''
8/3,3+<
H[WHUQDO
FRPSRQHQW

8/3,LQWHUIDFH SLQV

'DWD),)2
5$0LQWHUIDFH

3HULSKHUDO

27*B+6B62)

'DWD),)2
6LQJOHSRUW5$0
635$0
06Y9

1. The USB DMA cannot directly address the internal Flash memory.

35.3.1

OTG pins
Table 206. OTG_HS input/output pins
Signal name

35.3.2

Signal type

Description

OTG_HS_DP

Digital input/output USB OTG D+ line

OTG_HS_DM

Digital input/output USB OTG D- line

OTG_HS_ID

Digital input

USB OTG ID

OTG_HS_VBUS

Analog input

USB OTG VBUS

OTG_HS_SOF

Digital output

USB OTG Start Of Frame (visibility)

OTG_HS_ULPI_CK

Digital input

USB OTG ULPI clock

OTG_HS_ULPI_DIR

Digital input

USB OTG ULPI data bus direction control

OTG_HS_ULPI_STP

Digital output

USB OTG ULPI data stream stop

OTG_HS_ULPI_NXT

Digital input

USB OTG ULPI next data stream request

OTG_HS_ULPI_D[0..7]

Digital input/output USB OTG ULPI 8-bit bi-directional data bus

High-speed OTG PHY
The USB OTG HS core embeds an ULPI interface to connect an external HS phy.
RM0090 Rev 17

1383/1747
1541

USB on-the-go high-speed (OTG_HS)

35.3.3

RM0090

Embedded Full-speed OTG PHY
The full-speed OTG PHY includes the following components:
•

FS/LS transceiver module used by both host and Device. It directly drives transmission
and reception on the single-ended USB lines.

•

Integrated ID pull-up resistor used to sample the ID line for A/B Device identification.

•

DP/DM integrated pull-up and pull-down resistors controlled by the OTG_HS core
depending on the current role of the device. As a peripheral, it enables the DP pull-up
resistor to signal full-speed peripheral connections as soon as VBUS is sensed to be at
a valid level (B-session valid). In host mode, pull-down resistors are enabled on both
DP/DM. Pull-up and pull-down resistors are dynamically switched when the peripheral
role is changed via the host negotiation protocol (HNP).

•

Pull-up/pull-down resistor ECN circuit
The DP pull-up consists of 2 resistors controlled separately from the OTG_HS as per
the resistor Engineering Change Notice applied to USB Rev2.0. The dynamic trimming
of the DP pull-up strength allows to achieve a better noise rejection and Tx/Rx signal
quality.

•

VBUS sensing comparators with hysteresis used to detect VBUS_VALID, A-B Session
Valid and session-end voltage thresholds. They are used to drive the session request
protocol (SRP), detect valid startup and end-of-session conditions, and constantly
monitor the VBUS supply during USB operations.

•

VBUS pulsing method circuit used to charge/discharge VBUS through resistors during
the SRP (weak drive).

Caution:

To guarantee a correct operation for the USB OTG HS peripheral, the AHB frequency
should be higher than 30 MHz.

35.4

OTG dual-role device

35.4.1

ID line detection
The host or peripheral (the default) role depends on the level of the ID input line. It is
determined when the USB cable is plugged in and depends on which side of the USB cable
is connected to the micro-AB receptacle:

35.4.2

•

If the B-side of the USB cable is connected with a floating ID wire, the integrated pullup resistor detects a high ID level and the default peripheral role is confirmed. In this
configuration the OTG_HS conforms to the FSM standard described in
section 6.8.2. On-The-Go B-device of the USB On-The-Go Supplement, Revision 1.3.

•

If the A-side of the USB cable is connected with a grounded ID, the OTG_HS issues an
ID line status change interrupt (CIDSCHG bit in the OTG_HS_GINTSTS register) for
host software initialization, and automatically switches to host role. In this configuration
the OTG_HS conforms to the FSM standard described by section 6.8.1: On-The-Go ADevice of the USB On-The-Go Supplement, Revision 1.3.

HNP dual role device
The HNP capable bit in the Global USB configuration register (HNPCAP bit in the
OTG_HS_ GUSBCFG register) configures the OTG_HS core to dynamically change from
A-host to A-device role and vice-versa, or from B-device to B-host role and vice-versa,
according to the host negotiation protocol (HNP). The current device status is defined by the

1384/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
combination of the Connector ID Status bit in the Global OTG control and status register
(CIDSTS bit in OTG_HS_GOTGCTL) and the current mode of operation bit in the global
interrupt and status register (CMOD bit in OTG_HS_GINTSTS).
The HNP programming model is described in detail in Section 35.13: OTG_HS
programming model.

35.4.3

SRP dual-role device
The SRP capable bit in the global USB configuration register (SRPCAP bit in
OTG_HS_GUSBCFG) configures the OTG_HS core to switch VBUS off for the A-device in
order to save power. The A-device is always in charge of driving VBUS regardless of the
OTG_HS role (host or peripheral). The SRP A/B-device program model is described in
detail in Section 35.13: OTG_HS programming model.

35.5

USB functional description in peripheral mode
The OTG_HS operates as an USB peripheral in the following circumstances:
•

OTG B-device
OTG B-device default state if the B-side of USB cable is plugged in

•

OTG A-device
OTG A-device state after the HNP switches the OTG_HS to peripheral role

•

B-Device
If the ID line is present, functional and connected to the B-side of the USB cable, and
the HNP-capable bit in the Global USB Configuration register (HNPCAP bit in
OTG_HS_GUSBCFG) is cleared (see On-The-Go specification Revision 1.3 section
6.8.3).

•

Peripheral only (see Figure 388: USB peripheral-only connection)
The force peripheral mode bit in the Global USB configuration register (FDMOD in
OTG_HS_GUSBCFG) is set to 1, forcing the OTG_HS core to operate in USB
peripheral-only mode (see On-The-Go specification Revision 1.3 section 6.8.3). In this
case, the ID line is ignored even if it is available on the USB connector.

Note:

To build a bus-powered device architecture in the B-Device or peripheral-only configuration,
an external regulator must be added to generate the VDD supply voltage from VBUS.

35.5.1

SRP-capable peripheral
The SRP capable bit in the Global USB configuration register (SRPCAP bit in
OTG_HS_GUSBCFG) configures the OTG_HS to support the session request protocol
(SRP). As a result, it allows the remote A-device to save power by switching VBUS off when
the USB session is suspended.
The SRP peripheral mode program model is described in detail in Section : B-device
session request protocol.

RM0090 Rev 17

1385/1747
1541

USB on-the-go high-speed (OTG_HS)

35.5.2

RM0090

Peripheral states
Powered state
The VBUS input detects the B-session valid voltage used to put the USB peripheral in the
Powered state (see USB2.0 specification section 9.1). The OTG_HS then automatically
connects the DP pull-up resistor to signal full-speed device connection to the host, and
generates the session request interrupt (SRQINT bit in OTG_HS_GINTSTS) to notify the
Powered state. The VBUS input also ensures that valid VBUS levels are supplied by the host
during USB operations. If VBUS drops below the B-session valid voltage (for example
because power disturbances occurred or the host port has been switched off), the OTG_HS
automatically disconnects and the session end detected (SEDET bit in
OTG_HS_GOTGINT) interrupt is generated to notify that the OTG_HS has exited the
Powered state.
In Powered state, the OTG_HS expects a reset from the host. No other USB operations are
possible. When a reset is received, the reset detected interrupt (USBRST in
OTG_HS_GINTSTS) is generated. When the reset is complete, the enumeration done
interrupt (ENUMDNE bit in OTG_HS_GINTSTS) is generated and the OTG_HS enters the
Default state.

Soft disconnect
The Powered state can be exited by software by using the soft disconnect feature. The DP
pull-up resistor is removed by setting the Soft disconnect bit in the device control register
(SDIS bit in OTG_HS_DCTL), thus generating a device disconnect detection interrupt on
the host side even though the USB cable was not really unplugged from the host port.

Default state
In Default state the OTG_HS expects to receive a SET_ADDRESS command from the host.
No other USB operations are possible. When a valid SET_ADDRESS command is decoded
on the USB, the application writes the corresponding number into the device address field in
the device configuration register (DAD bit in OTG_HS_DCFG). The OTG_HS then enters
the address state and is ready to answer host transactions at the configured USB address.

Suspended state
The OTG_HS peripheral constantly monitors the USB activity. When the USB remains idle
for 3 ms, the early suspend interrupt (ESUSP bit in OTG_HS_GINTSTS) is issued. It is
confirmed 3 ms later, if appropriate, by generating a suspend interrupt (USBSUSP bit in
OTG_HS_GINTSTS). The device suspend bit is then automatically set in the device status
register (SUSPSTS bit in OTG_HS_DSTS) and the OTG_HS enters the Suspended state.
The device can also exit from the Suspended state by itself. In this case the application sets
the remote wakeup signaling bit in the device control register (RWUSIG bit in
OTG_HS_DCTL) and clears it after 1 to 15 ms.
When a resume signaling is detected from the host, the resume interrupt (WKUPINT bit in
OTG_HS_GINTSTS) is generated and the device suspend bit is automatically cleared.

1386/1747

RM0090 Rev 17

RM0090

35.5.3

USB on-the-go high-speed (OTG_HS)

Peripheral endpoints
The OTG_HS core instantiates the following USB endpoints:
•

Control endpoint 0
This endpoint is bidirectional and handles control messages only.
It has a separate set of registers to handle IN and OUT transactions, as well as
dedicated control (OTG_HS_DIEPCTL0/OTG_HS_DOEPCTL0), transfer configuration
(OTG_HS_DIEPTSIZ0/OTG_HS_DIEPTSIZ0), and status-interrupt
(OTG_HS_DIEPINTx/)OTG_HS_DOEPINT0) registers. The bits available inside the
control and transfer size registers slightly differ from other endpoints.

•

•

5 IN endpoints
–

They can be configured to support the isochronous, bulk or interrupt transfer type.

–

They feature dedicated control (OTG_HS_DIEPCTLx), transfer configuration
(OTG_HS_DIEPTSIZx), and status-interrupt (OTG_HS_DIEPINTx) registers.

–

The Device IN endpoints common interrupt mask register (OTG_HS_DIEPMSK)
allows to enable/disable a single endpoint interrupt source on all of the
IN endpoints (EP0 included).

–

They support incomplete isochronous IN transfer interrupt (IISOIXFR bit in
OTG_HS_GINTSTS). This interrupt is asserted when there is at least one
isochronous IN endpoint for which the transfer is not completed in the current
frame. This interrupt is asserted along with the end of periodic frame interrupt
(OTG_HS_GINTSTS/EOPF).

5 OUT endpoints
–

They can be configured to support the isochronous, bulk or interrupt transfer type.

–

They feature dedicated control (OTG_HS_DOEPCTLx), transfer configuration
(OTG_HS_DOEPTSIZx) and status-interrupt (OTG_HS_DOEPINTx) registers.

–

The Device Out endpoints common interrupt mask register
(OTG_HS_DOEPMSK) allows to enable/disable a single endpoint interrupt source
on all OUT endpoints (EP0 included).

–

They support incomplete isochronous OUT transfer interrupt (INCOMPISOOUT
bit in OTG_HS_GINTSTS). This interrupt is asserted when there is at least one
isochronous OUT endpoint on which the transfer is not completed in the current
frame. This interrupt is asserted along with the end of periodic frame interrupt
(OTG_HS_GINTSTS/EOPF).

RM0090 Rev 17

1387/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Endpoint controls
The following endpoint controls are available through the device endpoint-x IN/OUT control
register (DIEPCTLx/DOEPCTLx):
•

Endpoint enable/disable

•

Endpoint activation in current configuration

•

Program the USB transfer type (isochronous, bulk, interrupt)

•

Program the supported packet size

•

Program the Tx-FIFO number associated with the IN endpoint

•

Program the expected or transmitted data0/data1 PID (bulk/interrupt only)

•

Program the even/odd frame during which the transaction is received or transmitted
(isochronous only)

•

Optionally program the NAK bit to always send a negative acknowledge to the host
regardless of the FIFO status

•

Optionally program the STALL bit to always stall host tokens to that endpoint

•

Optionally program the Snoop mode for OUT endpoint where the received data CRC is
not checked

Endpoint transfer
The device endpoint-x transfer size registers (DIEPTSIZx/DOEPTSIZx) allow the application
to program the transfer size parameters and read the transfer status.
The programming operation must be performed before setting the endpoint enable bit in the
endpoint control register.
Once the endpoint is enabled, these fields are read-only as the OTG FS core updates them
with the current transfer status.
The following transfer parameters can be programmed:
•

Transfer size in bytes

•

Number of packets constituting the overall transfer size.

Endpoint status/interrupt
The device endpoint-x interrupt registers (DIEPINTx/DOPEPINTx) indicate the status of an
endpoint with respect to USB- and AHB-related events. The application must read these
registers when the OUT endpoint interrupt bit or the IN endpoint interrupt bit in the core
interrupt register (OEPINT bit in OTG_HS_GINTSTS or IEPINT bit in OTG_HS_GINTSTS,
respectively) is set. Before the application can read these registers, it must first read the
device all endpoints interrupt register (OTG_HS_DAINT) to get the exact endpoint number
for the device endpoint-x interrupt register. The application must clear the appropriate bit in
this register to clear the corresponding bits in the DAINT and GINTSTS registers.

1388/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
The peripheral core provides the following status checks and interrupt generation:
•

Transfer completed interrupt, indicating that data transfer has completed on both the
application (AHB) and USB sides

•

Setup stage done (control-out only)

•

Associated transmit FIFO is half or completely empty (in endpoints)

•

NAK acknowledge transmitted to the host (isochronous-in only)

•

IN token received when Tx-FIFO was empty (bulk-in/interrupt-in only)

•

OUT token received when endpoint was not yet enabled

•

Babble error condition detected

•

Endpoint disable by application is effective

•

Endpoint NAK by application is effective (isochronous-in only)

•

More than 3 back-to-back setup packets received (control-out only)

•

Timeout condition detected (control-in only)

•

Isochronous out packet dropped without generating an interrupt

RM0090 Rev 17

1389/1747
1541

USB on-the-go high-speed (OTG_HS)

35.6

RM0090

USB functional description on host mode
This section gives the functional description of the OTG_HS in the USB host mode. The
OTG_HS works as a USB host in the following circumstances:
•

OTG A-host
OTG A-device default state when the A-side of the USB cable is plugged in

•

OTG B-host
OTG B-device after HNP switching to the host role

•

A-device
If the ID line is present, functional and connected to the A-side of the USB cable, and
the HNP-capable bit is cleared in the Global USB Configuration register (HNPCAP bit
in OTG_HS_GUSBCFG). Integrated pull-down resistors are automatically set on the
DP/DM lines.

•

Host only (Figure 389: USB host-only connection).
The force host mode bit in the global USB configuration register (FHMOD bit in
OTG_HS_GUSBCFG) forces the OTG_HS core to operate in USB host-only mode. In
this case, the ID line is ignored even if it is available on the USB connector. Integrated
pull-down resistors are automatically set on the OTG_HS_FS_DP/OTG_HS_FS_DM
lines.

Note:

On-chip 5 V VBUS generation is not supported. As a result, a charge pump or a basic power
switch (if a 5 V supply is available on the application board) must be added externally to
drive the 5 V VBUS line. The external charge pump can be driven by any GPIO output. This
is required for the OTG A-host, A-device and host-only configurations.
The VBUS input ensures that valid VBUS levels are supplied by the charge pump during USB
operations while the charge pump overcurrent output can be input to any GPIO pin
configured to generate port interrupts. The overcurrent ISR must promptly disable the VBUS
generation.
Figure 411. USB host-only connection
9''
9

*3,2,54

(1

67036675
93ZU
&XUUHQWOLPLWHG
2YHUFXUUHQW SRZHUGLVWULEXWLRQ
VZLWFK 
9%86
'0

26&B,1

'3
966

26&B287

86%6WG$FRQQHFWRU

*3,2

06Y9

35.6.1

SRP-capable host
SRP support is available through the SRP capable bit in the global USB configuration
register (SRPCAP bit in OTG_HS_GUSBCFG). When the SRP feature is enabled, the host
can save power by switching off the VBUS power while the USB session is suspended. The

1390/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
SRP host mode program model is described in detail in Section : A-device session request
protocol.

35.6.2

USB host states
Host port power
On-chip 5 V VBUS generation is not supported. As a result, a charge pump or a basic power
switch (if a 5 V supply voltage is available on the application board) must be added
externally to drive the 5 V VBUS line. The external charge pump can be driven by any GPIO
output. When the application powers on VBUS through the selected GPIO, it must also set
the port power bit in the host port control and status register (PPWR bit in OTG_HS_HPRT).

VBUS valid
When SRP or HNP is enabled the VBUS sensing pin (PB13) pin should be connected to
VBUS. The VBUS input ensures that valid VBUS levels are supplied by the charge pump
during USB operations. Any unforeseen VBUS voltage drop below the VBUS valid threshold
(4.25 V) generates an OTG interrupt triggered by the session end detected bit (SEDET bit in
OTG_HS_GOTGINT). The application must then switch the VBUS power off and clear the
port power bit.
When HNP and SRP are both disabled, the VBUS sensing pin (PB13) should not be
connected to VBUS. This pin can be can be used as GPIO.
The charge pump overcurrent flag can also be used to prevent electrical damage. Connect
the overcurrent flag output from the charge pump to any GPIO input, and configure it to
generate a port interrupt on the active level. The overcurrent ISR must promptly disable the
VBUS generation and clear the port power bit.

Detection of peripheral connection by the host
If SRP or HNP are enabled, even if USB peripherals or B-devices can be attached at any
time, the OTG_HS does not detect a bus connection until the end of the VBUS sensing
(VBUS over 4.75 V).
When VBUS is at a valid level and a remote B-device is attached, the OTG_HS core issues a
host port interrupt triggered by the device connected bit in the host port control and status
register (PCDET bit in OTG_HS_HPRT).
When HNP and SRP are both disabled, USB peripherals or B-device are detected as
soon as they are connected. The OTG_FS core issues a host port interrupt triggered by
the device connected bit in the host port control and status (PCDET bit in OTG_FS_HPRT).

Detection of peripheral disconnection by the host
The peripheral disconnection event triggers the disconnect detected interrupt (DISCINT bit
in OTG_HS_GINTSTS).

Host enumeration
After detecting a peripheral connection, the host must start the enumeration process by
issuing an USB reset and configuration commands to the new peripheral.
Before sending an USB reset, the application waits for the OTG interrupt triggered by the
debounce done bit (DBCDNE bit in OTG_HS_GOTGINT), which indicates that the bus is

RM0090 Rev 17

1391/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

stable again after the electrical debounce caused by the attachment of a pull-up resistor on
OTG_HS_FS_DP (full speed) or OTG_HS_FS_DM (low speed).
The application issues an USB reset (single-ended zero) via the USB by keeping the port
reset bit set in the Host port control and status register (PRST bit in OTG_HS_HPRT) for a
minimum of 10 ms and a maximum of 20 ms. The application monitors the time and then
clears the port reset bit.
Once the USB reset sequence has completed, the host port interrupt is triggered by the port
enable/disable change bit (PENCHNG bit in OTG_HS_HPRT) to inform the application that
the speed of the enumerated peripheral can be read from the port speed field in the host
port control and status register (PSPD bit in OTG_HS_HPRT), and that the host is starting
to drive SOFs (full speed) or keep-alive tokens (low speed). The host is then ready to
complete the peripheral enumeration by sending peripheral configuration commands.

Host suspend
The application can decide to suspend the USB activity by setting the port suspend bit in the
host port control and status register (PSUSP bit in OTG_HS_HPRT). The OTG_HS core
stops sending SOFs and enters the Suspended state.
The Suspended state can be exited on the remote device initiative (remote wakeup). In this
case the remote wakeup interrupt (WKUPINT bit in OTG_HS_GINTSTS) is generated upon
detection of a remote wakeup event, the port resume bit in the host port control and status
register (PRES bit in OTG_HS_HPRT) is set, and a resume signaling is automatically
issued on the USB. The application must monitor the resume window duration, and then
clear the port resume bit to exit the Suspended state and restart the SOF.
If the Suspended state is exited on the host initiative, the application must set the port
resume bit to start resume signaling on the host port, monitor the resume window duration
and then clear the port resume bit.

35.6.3

Host channels
The OTG_HS core instantiates 12 host channels. Each host channel supports an USB host
transfer (USB pipe). The host is not able to support more than 8 transfer requests
simultaneously. If more than 8 transfer requests are pending from the application, the host
controller driver (HCD) must re-allocate channels when they become available, that is, after
receiving the transfer completed and channel halted interrupts.
Each host channel can be configured to support IN/OUT and any type of
periodic/nonperiodic transaction. Each host channel has dedicated control (HCCHARx),
transfer configuration (HCTSIZx) and status/interrupt (HCINTx) registers with associated
mask (HCINTMSKx) registers.

1392/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

Host channel controls
The following host channel controls are available through the host channel-x characteristics
register (HCCHARx):
•

Channel enable/disable

•

Program the HS/FS/LS speed of target USB peripheral

•

Program the address of target USB peripheral

•

Program the endpoint number of target USB peripheral

•

Program the transfer IN/OUT direction

•

Program the USB transfer type (control, bulk, interrupt, isochronous)

•

Program the maximum packet size (MPS)

•

Program the periodic transfer to be executed during odd/even frames

Host channel transfer
The host channel transfer size registers (HCTSIZx) allow the application to program the
transfer size parameters, and read the transfer status.
The programming operation must be performed before setting the channel enable bit in the
host channel characteristics register. Once the endpoint is enabled, the packet count field is
read-only as the OTG HS core updates it according to the current transfer status.
The following transfer parameters can be programmed:
•

Transfer size in bytes

•

Number of packets constituting the overall transfer size

•

Initial data PID

Host channel status/interrupt
The host channel-x interrupt register (HCINTx) indicates the status of an endpoint with
respect to USB- and AHB-related events. The application must read these register when the
host channels interrupt bit in the core interrupt register (HCINT bit in OTG_HS_GINTSTS) is
set. Before the application can read these registers, it must first read the host all channels
interrupt (HCAINT) register to get the exact channel number for the host channel-x interrupt
register. The application must clear the appropriate bit in this register to clear the
corresponding bits in the HAINT and GINTSTS registers. The mask bits for each interrupt
source of each channel are also available in the OTG_HS_HCINTMSK-x register.

RM0090 Rev 17

1393/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

The host core provides the following status checks and interrupt generation:

35.6.4

•

Transfer completed interrupt, indicating that the data transfer is complete on both the
application (AHB) and USB sides

•

Channel stopped due to transfer completed, USB transaction error or disable
command from the application

•

Associated transmit FIFO half or completely empty (IN endpoints)

•

ACK response received

•

NAK response received

•

STALL response received

•

USB transaction error due to CRC failure, timeout, bit stuff error, false EOP

•

Babble error

•

Frame overrun

•

Data toggle error

Host scheduler
The host core features a built-in hardware scheduler which is able to autonomously re-order
and manage the USB the transaction requests posted by the application. At the beginning of
each frame the host executes the periodic (isochronous and interrupt) transactions first,
followed by the nonperiodic (control and bulk) transactions to achieve the higher level of
priority granted to the isochronous and interrupt transfer types by the USB specification.
The host processes the USB transactions through request queues (one for periodic and one
for nonperiodic). Each request queue can hold up to 8 entries. Each entry represents a
pending transaction request from the application, and holds the IN or OUT channel number
along with other information to perform a transaction on the USB. The order in which the
requests are written to the queue determines the sequence of the transactions on the USB
interface.
At the beginning of each frame, the host processes the periodic request queue first, followed
by the nonperiodic request queue. The host issues an incomplete periodic transfer interrupt
(IPXFR bit in OTG_HS_GINTSTS) if an isochronous or interrupt transaction scheduled for
the current frame is still pending at the end of the current frame. The OTG HS core is fully
responsible for the management of the periodic and nonperiodic request queues.The
periodic transmit FIFO and queue status register (HPTXSTS) and nonperiodic transmit
FIFO and queue status register (HNPTXSTS) are read-only registers which can be used by
the application to read the status of each request queue. They contain:
•

The number of free entries currently available in the periodic (nonperiodic) request
queue (8 max)

•

Free space currently available in the periodic (nonperiodic) Tx-FIFO (out-transactions)

•

IN/OUT token, host channel number and other status information.

As request queues can hold a maximum of 8 entries each, the application can push to
schedule host transactions in advance with respect to the moment they physically reach the
USB for a maximum of 8 pending periodic transactions plus 8 pending nonperiodic
transactions.
To post a transaction request to the host scheduler (queue) the application must check that
there is at least 1 entry available in the periodic (nonperiodic) request queue by reading the
PTXQSAV bits in the OTG_HS_HNPTXSTS register or NPTQXSAV bits in the
OTG_HS_HNPTXSTS register.

1394/1747

RM0090 Rev 17

RM0090

35.7

USB on-the-go high-speed (OTG_HS)

SOF trigger
The OTG FS core allows to monitor, track and configure SOF framing in the host and
peripheral. It also features an SOF pulse output connectivity.
These capabilities are particularly useful to implement adaptive audio clock generation
techniques, where the audio peripheral needs to synchronize to the isochronous stream
provided by the PC, or the host needs trimming its framing rate according to the
requirements of the audio peripheral.

35.7.1

Host SOFs
In host mode the number of PHY clocks occurring between the generation of two
consecutive SOF (FS) or keep-alive (LS) tokens is programmable in the host frame interval
register (OTG_HS_HFIR), thus providing application control over the SOF framing period.
An interrupt is generated at any start of frame (SOF bit in OTG_HS_GINTSTS). The current
frame number and the time remaining until the next SOF are tracked in the host frame
number register (OTG_HS_HFNUM).
An SOF pulse signal is generated at any SOF starting token and with a width of 20 HCLK
cycles. It can be made available externally on the SOF pin using the SOFOUTEN bit in the
global control and configuration register. The SOF pulse is also internally connected to the
input trigger of timer 2 (TIM2), so that the input capture feature, the output compare feature
and the timer can be triggered by the SOF pulse. The TIM2 connection is enabled through
ITR1_RMP bits of TIM2_OR register.
Figure 412. SOF trigger output to TIM2 ITR1 connection

SOF output pulse

VBUS
DP
ITR1
TIM2

SOF
pulse

DM
ID

USB Micro-AB connector

OTG_HS_Core

ai16092

35.7.2

Peripheral SOFs
In peripheral mode, the start of frame interrupt is generated each time an SOF token is
received on the USB (SOF bit in OTG_HS_GINTSTS). The corresponding frame number
can be read from the device status register (FNSOF bit in OTG_HS_DSTS). An SOF pulse
signal with a width of 20 HCLK cycles is also generated and can be made available
externally on the SOF pin by using the SOF output enable bit in the global control and
configuration register (SOFOUTEN bit in OTG_HS_GCCFG). The SOF pulse signal is also
internally connected to the TIM2 input trigger, so that the input capture feature, the output
compare feature and the timer can be triggered by the SOF pulse (see Figure 412). The
TIM2 connection is enabled through ITR1_RMP bits of TIM2_OR register.

RM0090 Rev 17

1395/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

The end of periodic frame interrupt (GINTSTS/EOPF) is used to notify the application when
80%, 85%, 90% or 95% of the time frame interval elapsed depending on the periodic frame
interval field in the device configuration register (PFIVL bit in OTG_HS_DCFG).
This feature can be used to determine if all of the isochronous traffic for that frame is
complete.

35.8

OTG_HS low-power modes
Table 207 below defines the STM32 low power modes and their compatibility with the OTG.
Table 207. Compatibility of STM32 low power modes with the OTG

Mode

Description

USB compatibility

Run

MCU fully active

Required when USB not in
suspend state.

Sleep

USB suspend exit causes the device to exit Sleep mode. Peripheral
registers content is kept.

Available while USB is in
suspend state.

Stop

USB suspend exit causes the device to exit Stop mode. Peripheral
registers content is kept(1).

Available while USB is in
suspend state.

Standby

Powered-down. The peripheral must be reinitialized after exiting
Standby mode.

Not compatible with USB
applications.

1. Within Stop mode there are different possible settings. Some restrictions may also exist, please refer to Section 5: Power
controller (PWR) to understand which (if any) restrictions apply when using OTG.

The following bits and procedures reduce power consumption.
The power consumption of the OTG PHY is controlled by three bits in the general core
configuration register:
•

PHY power down (GCCFG/PWRDWN)
This bit switches on/off the PHY full-speed transceiver module. It must be preliminarily
set to allow any USB operation.

•

A-VBUS sensing enable (GCCFG/VBUSASEN)
This bit switches on/off the VBUS comparators associated with A-device operations. It
must be set when in A-device (USB host) mode and during HNP.

•

B-VBUS sensing enable (GCCFG/VBUSASEN)
This bit switches on/off the VBUS comparators associated with B-device operations. It
must be set when in B-device (USB peripheral) mode and during HNP.
Power reduction techniques are available in the USB suspended state, when the USB
session is not yet valid or the device is disconnected.

•

Stop PHY clock (STPPCLK bit in OTG_HS_PCGCCTL)
–

1396/1747

When setting the stop PHY clock bit in the clock gating control register, most of the
clock domain internal to the OTG high-speed core is switched off by clock gating.

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
The dynamic power consumption due to the USB clock switching activity is cut
even if the clock input is kept running by the application
–
•

Most of the transceiver is also disabled, and only the part in charge of detecting
the asynchronous resume or remote wakeup event is kept alive.

Gate HCLK (GATEHCLK bit in OTG_HS_PCGCCTL)
When setting the Gate HCLK bit in the clock gating control register, most of the system
clock domain internal to the OTG_HS core is switched off by clock gating. Only the
register read and write interface is kept alive. The dynamic power consumption due to
the USB clock switching activity is cut even if the system clock is kept running by the
application for other purposes.

•

35.9

USB system stop
–

When the OTG_HS is in USB suspended state, the application can decide to
drastically reduce the overall power consumption by shutting down all the clock
sources in the system. USB System Stop is activated by first setting the Stop PHY
clock bit and then configuring the system deep sleep mode in the powercontrol
system module (PWR).

–

The OTG_HS core automatically reactivates both system and USB clocks by
asynchronous detection of remote wakeup (as an host) or resume (as a Device)
signaling on the USB.

Dynamic update of the OTG_HS_HFIR register
The USB core embeds a dynamic trimming capability of micro-SOF framing period in host
mode allowing to synchronize an external device with the micro-SOF frames.
When the OTG_HS_HFIR register is changed within a current micro-SOF frame, the SOF
period correction is applied in the next frame as described in Figure 413.
Figure 413. Updating OTG_HS_HFIR dynamically
/LD /4'?(3?()&2 VALUE
  PERIODS

/4'?(3?()&2 VALUE
  PERIODS ()&2 WRITE LATENCY

.EW /4'?(3?()&2 VALUE
  PERIODS

3/&
RELOAD
,ATENCY

/4'?(3?(&)2
WRITE

x

x






x






x











&RAME
TIMER








/4'?(3?(&)2
VALUE

x

AIB

RM0090 Rev 17

1397/1747
1541

USB on-the-go high-speed (OTG_HS)

35.10

FIFO RAM allocation

35.10.1

Peripheral mode

RM0090

Receive FIFO RAM
For Receive FIFO RAM, the application should allocate RAM for SETUP packets: 10
locations must be reserved in the receive FIFO to receive SETUP packets on control
endpoints. These locations are reserved for SETUP packets and are not used by the core to
write any other data.
One location must be allocated for Global OUT NAK. Status information are also written to
the FIFO along with each received packet. Therefore, a minimum space of (Largest Packet
Size / 4) + 1 must be allocated to receive packets. If a high-bandwidth endpoint or multiple
isochronous endpoints are enabled, at least two spaces of (Largest Packet Size / 4) + 1
must be allotted to receive back-to-back packets. Typically, two (Largest Packet Size / 4) + 1
spaces are recommended so that when the previous packet is being transferred to AHB, the
USB can receive the subsequent packet.
Along with each endpoints last packet, transfer complete status information are also pushed
to the FIFO. Typically, one location for each OUT endpoint is recommended.

Transmit FIFO RAM
For Transmit FIFO RAM, the minimum RAM space required for each IN Endpoint Transmit
FIFO is the maximum packet size for this IN endpoint.
Note:

More space allocated in the transmit IN Endpoint FIFO results in a better performance on
the USB.

35.10.2

Host mode
Receive FIFO RAM
For Receive FIFO RAM allocation, Status information are written to the FIFO along with
each received packet. Therefore, a minimum space of (Largest Packet Size / 4) + 1 must be
allocated to receive packets. If a high-bandwidth channel or multiple isochronous channels
are enabled, at least two spaces of (Largest Packet Size / 4) + 1 must be allocated to
receive back-to-back packets. Typically, two (Largest Packet Size / 4) + 1 spaces are
recommended so that when the previous packet is being transferred to AHB, the USB can
receive the subsequent packet.
Along with each host channels last packet, transfer complete status information are also
pushed to the FIFO. As a consequence, one location must be allocated to store this data.

Transmit FIFO RAM
For Transmit FIFO RAM allocation, the minimum amount of RAM required for the host
nonperiodic Transmit FIFO is the largest maximum packet size for all supported nonperiodic
OUT channels. Typically, a space corresponding to two Largest Packet Size is
recommended, so that when the current packet is being transferred to the USB, the AHB
can transmit the subsequent packet.
The minimum amount of RAM required for Host periodic Transmit FIFO is the largest
maximum packet size for all supported periodic OUT channels. If there is at least one High

1398/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
Bandwidth Isochronous OUT endpoint, then the space must be at least two times the
maximum packet size for that channel.

Note:

More space allocated in the Transmit nonperiodic FIFO results in better performance on the
USB.
When operating in DMA mode, the DMA address register for each host channel (HCDMAn)
is stored in the SPRAM (FIFO). One location for each channel must be reserved for this.

35.11

OTG_HS interrupts
When the OTG_HS controller is operating in one mode, either peripheral or host, the
application must not access registers from the other mode. If an illegal access occurs, a
mode mismatch interrupt is generated and reflected in the Core interrupt register (MMIS bit
in the OTG_HS_GINTSTS register). When the core switches from one mode to the other,
the registers in the new mode of operation must be reprogrammed as they would be after a
power-on reset.
Figure 414 shows the interrupt hierarchy.

RM0090 Rev 17

1399/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Figure 414. Interrupt hierarchy
:DNHXSLQWHUUXSW
27*B+6B:.83



$1'

*OREDOLQWHUUXSW
27*B+6

25

27*B$+%&)*
$+%FRQILJXUDWLRQUHJLVWHU

$1'

27

2(

*,
17

3,1
,(3 3
,1
7

+&
,1
7
+3
57
,1
7

*OREDOLQWHUUXSWPDVN ELW

27*B*,17676
&RUHUHJLVWHULQWHUUXSW


 



 







27*B*,1706.
&RUHLQWHUUXSWPDVNUHJLVWHU

27*B*27*,17
27*LQWHUUXSWUHJLVWHU

(3B287LQWHUUXSW27*B+6B(3B287
(3B,1LQWHUUXSW27*B+6B(3B,1
(3 
287HQGSRLQWV

(3 
,1HQGSRLQWV


(3287


(3,1

27*B'$,1706.
'HYLFHDOOHQGSRLQWVLQWHUUXSWPDVN
UHJLVWHU

27*B'($&+,1706.
'HYLFHHDFKHQGSRLQWLQWHUUXSWPDVN
UHJLVWHU

27*B'$,17
'HYLFHDOOHQGSRLQWVLQWHUUXSWUHJLVWHU

27*B'($&+,17
'HYLFHHDFKHQGSRLQWLQWHUUXSWUHJLVWHU

27*B',(306.
27*B'2(306.
'HYLFH,1287HQGSRLQWVFRPPRQ
LQWHUUXSWPDVNUHJLVWHU

27*B',(3($&+06.
27*B'2(3($&+06.
'HYLFHHDFK,1287HQGSRLQWLQWHUUXSW
PDVNUHJLVWHUV

[ 



27*B',(3,17[
27*B'2(3,17[
'HYLFH,1287HQGSRLQWLQWHUUXSW
UHJLVWHUV

[ +&

27*B+357
+RVWSRUWFRQWURODQGVWDWXVUHJLVWHU

27*B+$,1706.
+RVWDOOFKDQQHOVLQWHUUXSWPDVNUHJLVWHU
27*B+$,17
+RVWDOOFKDQQHOVLQWHUUXSWUHJLVWHU
[ 



27*B+&7,1706.[
+RVWFKDQQHOVLQWHUUXSWPDVNUHJLVWHUV

[ +&

27*B+&7,17[
+RVWFKDQQHOVLQWHUUXSWUHJLVWHUV

06Y9

1. OTG_HS_WKUP becomes active (high state) when resume condition occurs during L1 SLEEP or L2 SUSPEND states.

1400/1747

RM0090 Rev 17

RM0090

35.12

USB on-the-go high-speed (OTG_HS)

OTG_HS control and status registers
By reading from and writing to the control and status registers (CSRs) through the AHB
slave interface, the application controls the OTG_HS controller. These registers are 32 bits
wide, and the addresses are 32-bit block aligned. The OTG_HS registers must be accessed
by words (32 bits). CSRs are classified as follows:
•

Core global registers

•

Host-mode registers

•

Host global registers

•

Host port CSRs

•

Host channel-specific registers

•

Device-mode registers

•

Device global registers

•

Device endpoint-specific registers

•

Power and clock-gating registers

•

Data FIFO (DFIFO) access registers

Only the Core global, Power and clock-gating, Data FIFO access, and host port control and
status registers can be accessed in both host and peripheral modes. When the OTG_HS
controller is operating in one mode, either peripheral or host, the application must not
access registers from the other mode. If an illegal access occurs, a mode mismatch
interrupt is generated and reflected in the Core interrupt register (MMIS bit in the
OTG_HS_GINTSTS register). When the core switches from one mode to the other, the
registers in the new mode of operation must be reprogrammed as they would be after a
power-on reset.

35.12.1

CSR memory map
The host and peripheral mode registers occupy different addresses. All registers are
implemented in the AHB clock domain.

RM0090 Rev 17

1401/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090
Figure 415. CSR memory map

0000h
Core global CSRs (1 Kbyte)
0400h
Host mode CSRs (1 Kbyte)
0800h
Device mode CSRs (1.5 Kbyte)
0E00h
Power and clock gating CSRs (0.5 Kbyte)
1000h
Device EP 0/Host channel 0 FIFO (4 Kbyte)
2000h
Device EP1/Host channel 1 FIFO (4 Kbyte)
DFIFO
push/pop
to this region

3000h

Device EP (x – 1)(1)/Host channel (x – 1)(1) FIFO (4 Kbyte)
Device EP x(1)/Host channel x(1) FIFO (4 Kbyte)

Reserved

2 0000h
Direct access to data FIFO RAM
for debugging (128 Kbyte)

DFIFO
debug read/
write to this
region

3 FFFFh
ai15615b

1. x = 5 in peripheral mode and x = 11 in host mode.

Global CSR map
These registers are available in both host and peripheral modes.
Table 208. Core global control and status registers (CSRs)
Acronym

Address
offset

Register name

OTG_HS_GOTGCTL

0x000

OTG_HS control and status register (OTG_HS_GOTGCTL) on page 1407

OTG_HS_GOTGINT

0x004

OTG_HS interrupt register (OTG_HS_GOTGINT) on page 1408

OTG_HS_GAHBCFG

0x008

OTG_HS AHB configuration register (OTG_HS_GAHBCFG) on page 1410

OTG_HS_GUSBCFG

0x00C

OTG_HS USB configuration register (OTG_HS_GUSBCFG) on page 1411

OTG_HS_GRSTCTL

0x010

OTG_HS reset register (OTG_HS_GRSTCTL) on page 1414

OTG_HS_GINTSTS

0x014

OTG_HS core interrupt register (OTG_HS_GINTSTS) on page 1417

OTG_HS_GINTMSK

0x018

OTG_HS interrupt mask register (OTG_HS_GINTMSK) on page 1421

1402/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
Table 208. Core global control and status registers (CSRs) (continued)
Address
offset

Acronym

Register name

OTG_HS_GRXSTSR

0x01C

OTG_HS_GRXSTSP

0x020

OTG_HS Receive status debug read/OTG status read and pop registers
(OTG_HS_GRXSTSR/OTG_HS_GRXSTSP) on page 1424

OTG_HS_GRXFSIZ

0x024

OTG_HS Receive FIFO size register (OTG_HS_GRXFSIZ) on page 1425

OTG_HS_GNPTXFSIZ/
OTG_HS_TX0FSIZ

0x028

OTG_HS nonperiodic transmit FIFO size/Endpoint 0 transmit FIFO size
register (OTG_HS_GNPTXFSIZ/OTG_HS_TX0FSIZ) on page 1426

OTG_HS_GNPTXSTS

0x02C

OTG_HS nonperiodic transmit FIFO/queue status register
(OTG_HS_GNPTXSTS) on page 1426

OTG_HS_GCCFG

0x038

OTG_HS general core configuration register (OTG_HS_GCCFG) on
page 1427

OTG_HS_CID

0x03C

OTG_HS core ID register (OTG_HS_CID) on page 1428

OTG_HS_HPTXFSIZ

0x100

OTG_HS Host periodic transmit FIFO size register (OTG_HS_HPTXFSIZ)
on page 1428

OTG_HS_DIEPTXFx

0x104
0x108
...
0x118

OTG_HS device IN endpoint transmit FIFO size register
(OTG_HS_DIEPTXFx) (x = 1..7, where x is the FIFO_number) on page 1429

Host-mode CSR map
These registers must be programmed every time the core changes to host mode.
Table 209. Host-mode control and status registers (CSRs)
Acronym

Offset
address

Register name

OTG_HS_HCFG

0x400

OTG_HS host configuration register (OTG_HS_HCFG) on page 1429

OTG_HS_HFIR

0x404

OTG_HS Host frame interval register (OTG_HS_HFIR) on page 1431

OTG_HS_HFNUM

0x408

OTG_HS host frame number/frame time remaining register
(OTG_HS_HFNUM) on page 1431

OTG_HS_HPTXSTS

0x410

OTG_HS_Host periodic transmit FIFO/queue status register
(OTG_HS_HPTXSTS) on page 1432

OTG_HS_HAINT

0x414

OTG_HS Host all channels interrupt register (OTG_HS_HAINT) on
page 1433

OTG_HS_HAINTMSK

0x418

OTG_HS host all channels interrupt mask register (OTG_HS_HAINTMSK)
on page 1433

OTG_HS_HPRT

0x440

OTG_HS host port control and status register (OTG_HS_HPRT) on
page 1434

RM0090 Rev 17

1403/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Table 209. Host-mode control and status registers (CSRs) (continued)
Acronym

Offset
address

Register name

OTG_HS_HCCHARx

0x500
0x520
...
0x660

OTG_HS host channel-x characteristics register (OTG_HS_HCCHARx)
(x = 0..11, where x = Channel_number) on page 1436

OTG_HS_HCSPLTx

0x504

OTG_HS host channel-x split control register (OTG_HS_HCSPLTx)
(x = 0..11, where x = Channel_number) on page 1438

OTG_HS_HCINTx

0x508

OTG_HS host channel-x interrupt register (OTG_HS_HCINTx) (x = 0..11,
where x = Channel_number) on page 1439

OTG_HS_HCINTMSKx 0x50C

OTG_HS host channel-x interrupt mask register (OTG_HS_HCINTMSKx)
(x = 0..11, where x = Channel_number) on page 1440

OTG_HS_HCTSIZx

0x510

OTG_HS host channel-x transfer size register (OTG_HS_HCTSIZx)
(x = 0..11, where x = Channel_number) on page 1441

OTG_HS_HCDMAx

0x514

OTG_HS host channel-x DMA address register (OTG_HS_HCDMAx)
(x = 0..11, where x = Channel_number) on page 1442

Device-mode CSR map
These registers must be programmed every time the core changes to peripheral mode.
Table 210. Device-mode control and status registers
Acronym

Offset
address

Register name

OTG_HS_DCFG

0x800

OTG_HS device configuration register (OTG_HS_DCFG) on
page 1442

OTG_HS_DCTL

0x804

OTG_HS device control register (OTG_HS_DCTL) on page 1444

OTG_HS_DSTS

0x808

OTG_HS device status register (OTG_HS_DSTS) on page 1446

OTG_HS_DIEPMSK

0x810

OTG_HS device IN endpoint common interrupt mask register
(OTG_HS_DIEPMSK) on page 1447

OTG_HS_DOEPMSK

0x814

OTG_HS device OUT endpoint common interrupt mask register
(OTG_HS_DOEPMSK) on page 1448

OTG_HS_DAINT

0x818

OTG_HS device all endpoints interrupt register (OTG_HS_DAINT)
on page 1449

OTG_HS_DAINTMSK

0x81C

OTG_HS all endpoints interrupt mask register
(OTG_HS_DAINTMSK) on page 1450

OTG_HS_DVBUSDIS

0x828

OTG_HS device VBUS discharge time register
(OTG_HS_DVBUSDIS) on page 1450

OTG_HS_DVBUSPULSE

0x82C

OTG_HS device VBUS pulsing time register
(OTG_HS_DVBUSPULSE) on page 1451

OTG_HS_DTHRCTL

0x830

OTG_HS Device threshold control register (OTG_HS_DTHRCTL)
on page 1452

1404/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
Table 210. Device-mode control and status registers (continued)
Acronym

Offset
address

Register name

OTG_HS_DIEPEMPMSK

0x834

OTG_HS device IN endpoint FIFO empty interrupt mask register:
(OTG_HS_DIEPEMPMSK) on page 1453

OTG_HS_DEACHINT

0x838

OTG_HS device each endpoint interrupt register
(OTG_HS_DEACHINT) on page 1453

OTG_HS_DEACHINTMSK

0x83C

OTG_HS device each endpoint interrupt register mask
(OTG_HS_DEACHINTMSK) on page 1454

OTG_HS_DIEPEACHMSK1

0x844

OTG_HS device each in endpoint-1 interrupt register
(OTG_HS_DIEPEACHMSK1) on page 1454

OTG_HS_DOEPEACHMSK1

0x884

OTG_HS device each OUT endpoint-1 interrupt register
(OTG_HS_DOEPEACHMSK1) on page 1455

OTG_HS_DIEPCTLx

0x900
0x920
...
0x9C0

OTG device endpoint-x control register (OTG_HS_DIEPCTLx) (x =
0..7, where x = Endpoint_number) on page 1456

OTG_HS_DIEPINTx

0x908

OTG_HS device endpoint-x interrupt register
(OTG_HS_DIEPINTx) (x = 0..7, where x = Endpoint_number) on
page 1463

OTG_HS_DIEPTSIZ0

0x910

OTG_HS device IN endpoint 0 transfer size register
(OTG_HS_DIEPTSIZ0) on page 1466

OTG_HS_DIEPDMAx/
OTG_HS_DOEPDMAx

0x914/0xB14

OTG_HS device endpoint-x DMA address register
(OTG_HS_DIEPDMAx / OTG_HS_DOEPDMAx) (x = 1..5, where
x = Endpoint_number) on page 1470

OTG_HS_DTXFSTSx

0x918

OTG_HS device IN endpoint transmit FIFO status register
(OTG_HS_DTXFSTSx) (x = 0..5, where x = Endpoint_number) on
page 1469

OTG_HS_DIEPTSIZx

0x930
0x950
...
0x9D0

OTG_HS device endpoint-x transfer size register
(OTG_HS_DOEPTSIZx) (x = 1..5, where x = Endpoint_number)
on page 1469

OTG_HS_DOEPCTL0

0xB00

OTG_HS device control OUT endpoint 0 control register
(OTG_HS_DOEPCTL0) on page 1459

OTG_HS_DOEPSIZ0

0xB10

OTG_HS device OUT endpoint 0 transfer size register
(OTG_HS_DOEPTSIZ0) on page 1467

OTG_HS_DOEPCTLx

0xB20
0xB40
...
0xBC0

OTG device endpoint-x control register (OTG_HS_DIEPCTLx) (x =
0..7, where x = Endpoint_number) on page 1456

RM0090 Rev 17

1405/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Table 210. Device-mode control and status registers (continued)
Acronym

Offset
address

Register name

OTG_HS_DOEPINTx

0xB08
0xB28
...
0xBC8

OTG_HS device endpoint-x interrupt register
(OTG_HS_DOEPINTx) (x = 0..7, where x = Endpoint_number) on
page 1465

OTG_HS_DOEPTSIZx

0xB30
0xB50
...
0xBD0

OTG_HS device endpoint-x transfer size register
(OTG_HS_DOEPTSIZx) (x = 1..5, where x = Endpoint_number)
on page 1469

Data FIFO (DFIFO) access register map
These registers, available in both host and peripheral modes, are used to read or write the
FIFO space for a specific endpoint or a channel, in a given direction. If a host channel is of
type IN, the FIFO can only be read on the channel. Similarly, if a host channel is of type
OUT, the FIFO can only be written on the channel.
Table 211. Data FIFO (DFIFO) access register map
FIFO access register section

Address range

Access

Device IN Endpoint 0/Host OUT Channel 0: DFIFO Write Access
Device OUT Endpoint 0/Host IN Channel 0: DFIFO Read Access

0x1000–0x1FFC

w
r

Device IN Endpoint 1/Host OUT Channel 1: DFIFO Write Access
Device OUT Endpoint 1/Host IN Channel 1: DFIFO Read Access

0x2000–0x2FFC

w
r

...

...

...

Device IN Endpoint x(1)/Host OUT Channel x(1): DFIFO Write Access
0xX000–0xXFFC
Device OUT Endpoint x(1)/Host IN Channel x(1): DFIFO Read Access

w
r

1. Where x is 5 in peripheral mode and 11 in host mode.

Power and clock gating CSR map
There is a single register for power and clock gating. It is available in both host and
peripheral modes.
Table 212. Power and clock gating control and status registers
Register name

Acronym

Power and clock gating control register

OTG_HS_PCGCCTL

Reserved

35.12.2

-

Offset address: 0xE00–0xFFF
0xE00-0xE04
0xE05–0xFFF

OTG_HS global registers
These registers are available in both host and peripheral modes, and do not need to be
reprogrammed when switching between these modes.

1406/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
Bit values in the register descriptions are expressed in binary unless otherwise specified.

OTG_HS control and status register (OTG_HS_GOTGCTL)
Address offset: 0x000
Reset value: 0x0001 0000

rw

rw

rw

r

6

5

4

Reserved

3

1

0

SRQ

r

7

SRQSCS

CIDSTS

r

HNPRQ

DBCT

r

HNGSCS

ASVLD

r

Reserved

8

DHNPEN

BSVLD

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

Reserved

9

HSHNPEN

The OTG control and status register controls the behavior and reflects the status of the OTG
function of the core.
2

rw

r

Bits 31:20 Reserved, must be kept at reset value.
Bit 19 BSVLD: B-session valid
Indicates the peripheral mode transceiver status.
0: B-session is not valid.
1: B-session is valid.
In OTG mode, you can use this bit to determine if the device is connected or disconnected.
Note: Only accessible in peripheral mode.
Bit 18 ASVLD: A-session valid
Indicates the host mode transceiver status.
0: A-session is not valid
1: A-session is valid
Note: Only accessible in host mode.
Bit 17 DBCT: Long/short debounce time
Indicates the debounce time of a detected connection.
0: Long debounce time, used for physical connections (100 ms + 2.5 µs)
1: Short debounce time, used for soft connections (2.5 µs)
Note: Only accessible in host mode.
Bit 16 CIDSTS: Connector ID status
Indicates the connector ID status on a connect event.
0: The OTG_HS controller is in A-device mode
1: The OTG_HS controller is in B-device mode
Note: Accessible in both peripheral and host modes.
Bits 15:12 Reserved, must be kept at reset value.
Bit 11 DHNPEN: Device HNP enabled
The application sets this bit when it successfully receives a SetFeature.SetHNPEnable
command from the connected USB host.
0: HNP is not enabled in the application
1: HNP is enabled in the application
Note: Only accessible in peripheral mode.

RM0090 Rev 17

1407/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bit 10 HSHNPEN: Host set HNP enable
The application sets this bit when it has successfully enabled HNP (using the
SetFeature.SetHNPEnable command) on the connected device.
0: Host Set HNP is not enabled
1: Host Set HNP is enabled
Note: Only accessible in host mode.
Bit 9 HNPRQ: HNP request
The application sets this bit to initiate an HNP request to the connected USB host. The
application can clear this bit by writing a 0 when the host negotiation success status change
bit in the OTG interrupt register (HNSSCHG bit in OTG_HS_GOTGINT) is set. The core
clears this bit when the HNSSCHG bit is cleared.
0: No HNP request
1: HNP request
Note: Only accessible in peripheral mode.
Bit 8 HNGSCS: Host negotiation success
The core sets this bit when host negotiation is successful. The core clears this bit when the
HNP Request (HNPRQ) bit in this register is set.
0: Host negotiation failure
1: Host negotiation success
Note: Only accessible in peripheral mode.
Bits 7:2 Reserved, must be kept at reset value.
Bit 1 SRQ: Session request
The application sets this bit to initiate a session request on the USB. The application can
clear this bit by writing a 0 when the host negotiation success status change bit in the OTG
Interrupt register (HNSSCHG bit in OTG_HS_GOTGINT) is set. The core clears this bit when
the HNSSCHG bit is cleared.
If you use the USB 1.1 full-speed serial transceiver interface to initiate the session request,
the application must wait until VBUS discharges to 0.2 V, after the B-Session Valid bit in this
register (BSVLD bit in OTG_HS_GOTGCTL) is cleared.
0: No session request
1: Session request
Note: Only accessible in peripheral mode.
Bit 0 SRQSCS: Session request success
The core sets this bit when a session request initiation is successful.
0: Session request failure
1: Session request success
Note: Only accessible in peripheral mode.

OTG_HS interrupt register (OTG_HS_GOTGINT)
Address offset: 0x04
Reset value: 0x0000 0000
The application reads this register whenever there is an OTG interrupt and clears the bits in
this register to clear the OTG interrupt.

1408/1747

RM0090 Rev 17

19

18

17

16 15 14 13 12 11 10

9

8

Reserved

HNGDET

Reserved

HNSSCHG

SRSSCHG

rc_ rc_ rc_
w1 w1 w1

7

6

5

4

Reserved

rc_ rc_
w1 w1

3

2
SEDET

31 30 29 28 27 26 25 24 23 22 21 20

ADTOCHG

USB on-the-go high-speed (OTG_HS)

DBCDNE

RM0090

1

0

Res.

rc_
w1

Bits 31:20 Reserved, must be kept at reset value.
Bit 19 DBCDNE: Debounce done
The core sets this bit when the debounce is completed after the device connect. The
application can start driving USB reset after seeing this interrupt. This bit is only valid when
the HNP Capable or SRP Capable bit is set in the Core USB Configuration register (HNPCAP
bit or SRPCAP bit in OTG_HS_GUSBCFG, respectively).
Note: Only accessible in host mode.
Bit 18 ADTOCHG: A-device timeout change
The core sets this bit to indicate that the A-device has timed out while waiting for the B-device
to connect.
Note: Accessible in both peripheral and host modes.
Bit 17 HNGDET: Host negotiation detected
The core sets this bit when it detects a host negotiation request on the USB.
Note: Accessible in both peripheral and host modes.
Bits 16:10 Reserved, must be kept at reset value.
Bit 9 HNSSCHG: Host negotiation success status change
The core sets this bit on the success or failure of a USB host negotiation request. The
application must read the host negotiation success bit of the OTG Control and Status register
(HNGSCS in OTG_HS_GOTGCTL) to check for success or failure.
Note: Accessible in both peripheral and host modes.
Bits 7:3 Reserved, must be kept at reset value.
Bit 8 SRSSCHG: Session request success status change
The core sets this bit on the success or failure of a session request. The application must
read the session request success bit in the OTG Control and status register (SRQSCS bit in
OTG_HS_GOTGCTL) to check for success or failure.
Note: Accessible in both peripheral and host modes.
Bit 2 SEDET: Session end detected
The core sets this bit to indicate that the level of the voltage on VBUS is no longer valid for a
B-device session when VBUS < 0.8 V.
Bits 1:0 Reserved, must be kept at reset value.

RM0090 Rev 17

1409/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

OTG_HS AHB configuration register (OTG_HS_GAHBCFG)
Address offset: 0x008
Reset value: 0x0000 0000

rw

rw

6

5

4

rw

3

2

1

HBSTLEN

rw

rw

rw

0
GINT

7

DMAEN

8

Reserved

Reserved

9

TXFELVL

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

PTXFELVL

This register can be used to configure the core after power-on or a change in mode. This
register mainly contains AHB system-related configuration parameters. Do not change this
register after the initial programming. The application must program this register before
starting any transactions on either the AHB or the USB.

rw

rw

Bits 31:20 Reserved, must be kept at reset value.
Bit 8 PTXFELVL: Periodic TxFIFO empty level
Indicates when the periodic TxFIFO empty interrupt bit in the Core interrupt register (PTXFE
bit in OTG_HS_GINTSTS) is triggered.
0: PTXFE (in OTG_HS_GINTSTS) interrupt indicates that the Periodic TxFIFO is half empty
1: PTXFE (in OTG_HS_GINTSTS) interrupt indicates that the Periodic TxFIFO is completely
empty
Note: Only accessible in host mode.
Bit 7 TXFELVL: TxFIFO empty level
In peripheral mode, this bit indicates when the IN endpoint Transmit FIFO empty interrupt
(TXFE in OTG_HS_DIEPINTx.) is triggered.
0: TXFE (in OTG_HS_DIEPINTx) interrupt indicates that the IN Endpoint TxFIFO is half
empty
1: TXFE (in OTG_HS_DIEPINTx) interrupt indicates that the IN Endpoint TxFIFO is
completely empty
Note: Only accessible in peripheral mode.
Bit 6 Reserved, must be kept at reset value.
Bits5 DMAEN: DMA enable
0: The core operates in slave mode
1: The core operates in DMA mode
Bits 4:1 HBSTLEN: Burst length/type
0000 Single
0001 INCR
0011 INCR4
0101 INCR8
0111 INCR16
Others: Reserved
Bit 0 GINT: Global interrupt mask
This bit is used to mask or unmask the interrupt line assertion to the application. Irrespective
of this bit setting, the interrupt status registers are updated by the core.
0: Mask the interrupt assertion to the application.
1: Unmask the interrupt assertion to the application
Note: Accessible in both peripheral and host modes.

1410/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

OTG_HS USB configuration register (OTG_HS_GUSBCFG)
Address offset: 0x00C
Reset value: 0x0000 1440

rw

rw

rw

rw

6
PHYSEL

rw

TRDT

7

Reserved

rw

8
SRPCAP

rw

Reserved

rw

PHYLPCS

ULPIFSLS

rw

Reserved

ULPIAR

rw

ULPICSM

rw

ULPIEVBUSI

rw

ULPIEVBUSD

rw

TSDPS

rw

PTCI

FHMOD

rw

PCCI

FDMOD

rw

ULPIIPD

CTXPKT

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

Reserved

9
HNPCAP

This register can be used to configure the core after power-on or a changing to host mode
or peripheral mode. It contains USB and USB-PHY related configuration parameters. The
application must program this register before starting any transactions on either the AHB or
the USB. Do not make changes to this register after the initial programming.

rw

5

4

3

Reserved

2

1

0

TOCAL

rw

Bit 31 CTXPKT: Corrupt Tx packet
This bit is for debug purposes only. Never set this bit to 1.
Note: Accessible in both peripheral and host modes.
Bit 30 FDMOD: Forced peripheral mode
Writing a 1 to this bit forces the core to peripheral mode irrespective of the OTG_HS_ID input
pin.
0: Normal mode
1: Forced peripheral mode
After setting the force bit, the application must wait at least 25 ms before the change takes
effect.
Note: Accessible in both peripheral and host modes.
Bit 29 FHMOD: Forced host mode
Writing a 1 to this bit forces the core to host mode irrespective of the OTG_HS_ID input pin.
0: Normal mode
1: Forced host mode
After setting the force bit, the application must wait at least 25 ms before the change takes
effect.
Note: Accessible in both peripheral and host modes.
Bits 28:26 Reserved, must be kept at reset value.
Bit 25 ULPIIPD: ULPI interface protect disable
This bit controls the circuitry built in the PHY to protect the ULPI interface when the link tristates stp and data. Any pull-up or pull-down resistors employed by this feature can be
disabled. Please refer to the ULPI specification for more details.
0: Enables the interface protection circuit
1: Disables the interface protection circuit
Bit 24 PTCI: Indicator pass through
This bit controls whether the complement output is qualified with the internal VBUS valid
comparator before being used in the VBUS state in the RX CMD. Please refer to the ULPI
specification for more details.
0: Complement Output signal is qualified with the Internal VBUS valid comparator
1: Complement Output signal is not qualified with the Internal VBUS valid comparator

RM0090 Rev 17

1411/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bit 23 PCCI: Indicator complement
This bit controls the PHY to invert the ExternalVbusIndicator input signal, and generate the
complement output. Please refer to the ULPI specification for more details.
0: PHY does not invert the ExternalVbusIndicator signal
1: PHY inverts ExternalVbusIndicator signal
Bit 22 TSDPS: TermSel DLine pulsing selection
This bit selects utmi_termselect to drive the data line pulse during SRP (session request
protocol).
0: Data line pulsing using utmi_txvalid (default)
1: Data line pulsing using utmi_termsel
Bit 21 ULPIEVBUSI: ULPI external VBUS indicator
This bit indicates to the ULPI PHY to use an external VBUS overcurrent indicator.
0: PHY uses an internal VBUS valid comparator
1: PHY uses an external VBUS valid comparator
Bit 20 ULPIEVBUSD: ULPI External VBUS Drive
This bit selects between internal or external supply to drive 5 V on VBUS, in the ULPI PHY.
0: PHY drives VBUS using internal charge pump (default)
1: PHY drives VBUS using external supply.
Bit 19 ULPICSM: ULPI Clock SuspendM
This bit sets the ClockSuspendM bit in the interface control register on the ULPI PHY. This bit
applies only in the serial and carkit modes.
0: PHY powers down the internal clock during suspend
1: PHY does not power down the internal clock
Bit 18 ULPIAR: ULPI Auto-resume
This bit sets the AutoResume bit in the interface control register on the ULPI PHY.
0: PHY does not use AutoResume feature
1: PHY uses AutoResume feature
Bit 17 ULPIFSLS: ULPI FS/LS select
The application uses this bit to select the FS/LS serial interface for the ULPI PHY. This bit is
valid only when the FS serial transceiver is selected on the ULPI PHY.
0: ULPI interface
1: ULPI FS/LS serial interface
Bit 16

Reserved, must be kept at reset value.

Bit 15 PHYLPCS: PHY Low-power clock select
This bit selects either 480 MHz or 48 MHz (low-power) PHY mode. In FS and LS modes, the
PHY can usually operate on a 48 MHz clock to save power.
0: 480 MHz internal PLL clock
1: 48 MHz external clock
In 480 MHz mode, the UTMI interface operates at either 60 or 30 MHz, depending on
whether the 8- or 16-bit data width is selected. In 48 MHz mode, the UTMI interface operates
at 48 MHz in FS and LS modes.
Bit 14 Reserved, must be kept at reset value.
Bits 13:10 TRDT: USB turnaround time
These bits allows to set the turnaround time in PHY clocks. They must be configured
according to Table 213: TRDT values, depending on the application AHB frequency. Higher
TRDT values allow stretching the USB response time to IN tokens in order to compensate for
longer AHB read access latency to the Data FIFO.

1412/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

Bit 9 HNPCAP: HNP-capable
The application uses this bit to control the OTG_HS controller’s HNP capabilities.
0: HNP capability is not enabled
1: HNP capability is enabled
Note: Accessible in both peripheral and host modes.
Bit 8 SRPCAP: SRP-capable
The application uses this bit to control the OTG_HS controller’s SRP capabilities. If the core
operates as a nonSRP-capable B-device, it cannot request the connected A-device (host) to
activate VBUS and start a session.
0: SRP capability is not enabled
1: SRP capability is enabled
Note: Accessible in both peripheral and host modes.
Bit 7 Reserved, must be kept at reset value.
Bit 6 PHYSEL: USB 2.0 high-speed ULPI PHY or USB 1.1 full-speed serial transceiver select
0: USB 2.0 high-speed ULPI PHY
1: USB 1.1 full-speed serial transceiver
Bits 5:3

Reserved, must be kept at reset value.

Bits 2:0 TOCAL: FS timeout calibration
The number of PHY clocks that the application programs in this field is added to the fullspeed interpacket timeout duration in the core to account for any additional delays
introduced by the PHY. This can be required, because the delay introduced by the PHY in
generating the line state condition can vary from one PHY to another.
The USB standard timeout value for full-speed operation is 16 to 18 (inclusive) bit times. The
application must program this field based on the speed of enumeration. The number of bit
times added per PHY clock is 0.25 bit times.

Table 213. TRDT values
AHB frequency range (MHz)
TRDT minimum value
Min.

Max

30

-

RM0090 Rev 17

0x9

1413/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

OTG_HS reset register (OTG_HS_GRSTCTL)
Address offset: 0x010
Reset value: 0x8000 0000

6

4

TXFNUM
rw

rs

rs

3

2

1

0
CSRST

7

HSRST

8

FCRST

5

RXFFLSH

r

9

Reserved

r

Reserved

AHBIDL

DMAREQ

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

TXFFLSH

The application uses this register to reset various hardware features inside the core.

rs

rs

rs

Bit 31 AHBIDL: AHB master idle
Indicates that the AHB master state machine is in the Idle condition.
Note: Accessible in both peripheral and host modes.
Bit 30 DMAREQ: DMA request signal
This bit indicates that the DMA request is in progress. Used for debug.
Bits 29:11

Reserved, must be kept at reset value.

Bits 10:6 TXFNUM: TxFIFO number
This is the FIFO number that must be flushed using the TxFIFO Flush bit. This field must not
be changed until the core clears the TxFIFO Flush bit.
″
00000:
–
Nonperiodic TxFIFO flush in host mode
–
Tx FIFO 0 flush in peripheral mode
″
00001:
–
Periodic TxFIFO flush in host mode
–
TXFIFO 1 flush in peripheral mode
″
00010: TXFIFO 2 flush in peripheral mode
...
″
00101: TXFIFO 15 flush in peripheral mode
″
10000: Flush all the transmit FIFOs in peripheral or host mode.
Note: Accessible in both peripheral and host modes.
Bit 5 TXFFLSH: TxFIFO flush
This bit selectively flushes a single or all transmit FIFOs, but cannot do so if the core is in the
midst of a transaction.
The application must write this bit only after checking that the core is neither writing to the
TxFIFO nor reading from the TxFIFO. Verify using these registers:
–
Read: the NAK effective interrupt ensures the core is not reading from the FIFO
–
Write: the AHBIDL bit in OTG_HS_GRSTCTL ensures that the core is not writing
anything to the FIFO
Note: Accessible in both peripheral and host modes.

1414/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

Bit 4 RXFFLSH: RxFIFO flush
The application can flush the entire RxFIFO using this bit, but must first ensure that the core
is not in the middle of a transaction.
The application must only write to this bit after checking that the core is neither reading from
the RxFIFO nor writing to the RxFIFO.
The application must wait until the bit is cleared before performing any other operation. This
bit requires 8 clocks (slowest of PHY or AHB clock) to be cleared.
Note: Accessible in both peripheral and host modes.
Bit 3

Reserved, must be kept at reset value.

RM0090 Rev 17

1415/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bit 2 FCRST: Host frame counter reset
The application writes this bit to reset the (micro) frame number counter inside the core.
When the (micro) frame counter is reset, the subsequent SOF sent out by the core has a
frame number of 0.
Note: Only accessible in host mode.
Bit 1 HSRST: HCLK soft reset
The application uses this bit to flush the control logic in the AHB Clock domain. Only AHB
Clock Domain pipelines are reset.
FIFOs are not flushed with this bit.
All state machines in the AHB clock domain are reset to the Idle state after terminating the
transactions on the AHB, following the protocol.
CSR control bits used by the AHB clock domain state machines are cleared.
To clear this interrupt, status mask bits that control the interrupt status and are generated by
the AHB clock domain state machine are cleared.
Because interrupt status bits are not cleared, the application can get the status of any core
events that occurred after it set this bit.
This is a self-clearing bit that the core clears after all necessary logic is reset in the core. This
can take several clocks, depending on the core’s current state.
Note: Accessible in both peripheral and host modes.
Bit 0 CSRST: Core soft reset
Resets the HCLK and PCLK domains as follows:
Clears the interrupts and all the CSR register bits except for the following bits:
– RSTPDMODL bit in OTG_HS_PCGCCTL
– GAYEHCLK bit in OTG_HS_PCGCCTL
– PWRCLMP bit in OTG_HS_PCGCCTL
– STPPCLK bit in OTG_HS_PCGCCTL
– FSLSPCS bit in OTG_HS_HCFG
– DSPD bit in OTG_HS_DCFG
All module state machines (except for the AHB slave unit) are reset to the Idle state, and all
the transmit FIFOs and the receive FIFO are flushed.
Any transactions on the AHB Master are terminated as soon as possible, after completing the
last data phase of an AHB transfer. Any transactions on the USB are terminated immediately.
The application can write to this bit any time it wants to reset the core. This is a self-clearing bit
and the core clears this bit after all the necessary logic is reset in the core, which can take
several clocks, depending on the current state of the core. Once this bit has been cleared, the
software must wait at least 3 PHY clocks before accessing the PHY domain (synchronization
delay). The software must also check that bit 31 in this register is set to 1 (AHB Master is Idle)
before starting any operation.
Typically, the software reset is used during software development and also when you
dynamically change the PHY selection bits in the above listed USB configuration registers.
When you change the PHY, the corresponding clock for the PHY is selected and used in the
PHY domain. Once a new clock is selected, the PHY domain has to be reset for proper
operation.
Note: Accessible in both peripheral and host modes.

1416/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

OTG_HS core interrupt register (OTG_HS_GINTSTS)
Address offset: 0x014
Reset value: 0x0400 0020
This register interrupts the application for system-level events in the current mode
(peripheral mode or host mode).
Some of the bits in this register are valid only in host mode, while others are valid in
peripheral mode only. This register also indicates the current mode. To clear the interrupt
status bits of the rc_w1 type, the application must write 1 into the bit.
The FIFO status interrupts are read-only; once software reads from or writes to the FIFO
while servicing these interrupts, FIFO interrupt conditions are cleared automatically.

r

r

r

1

0

MMIS

4

CMOD

SOF

5

r

rc_w1

2

RXFLVL

6

OTGINT

3

r

rc_w1

7

NPTXFE

8

Reserved

ESUSP

9

GINAKEFF

rc_w1

USBSUSP

USBRST

ISOODRP

ENUMDNE

r

EOPF

r

Reserved

IEPINT

rc_w1

OEPINT

IISOIXFR

r

IPXFR/INCOMPISOOUT

HPRTINT

r

DATAFSUSP

HCINT

r

Reserved

PTXFE

Reserved

DISCINT

rc_w1

CIDSCHG

SRQINT

WKUINT

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

GONAKEFF

The application must clear the OTG_HS_GINTSTS register at initialization before
unmasking the interrupt bit to avoid any interrupts generated prior to initialization.

r

Bit 31 WKUPINT: Resume/remote wakeup detected interrupt
In peripheral mode, this interrupt is asserted when a resume is detected on the USB. In host
mode, this interrupt is asserted when a remote wakeup is detected on the USB.
Note: Accessible in both peripheral and host modes.
Bit 30 SRQINT: Session request/new session detected interrupt
In host mode, this interrupt is asserted when a session request is detected from the device.
In peripheral mode, this interrupt is asserted when VBUS is in the valid range for a B-device
device. Accessible in both peripheral and host modes.
Bit 29 DISCINT: Disconnect detected interrupt
Asserted when a device disconnect is detected.
Note: Only accessible in host mode.
Bit 28 CIDSCHG: Connector ID status change
The core sets this bit when there is a change in connector ID status.
Note: Accessible in both peripheral and host modes.
Bit 27 Reserved, must be kept at reset value.
Bit 26 PTXFE: Periodic TxFIFO empty
Asserted when the periodic transmit FIFO is either half or completely empty and there is
space for at least one entry to be written in the periodic request queue. The half or
completely empty status is determined by the periodic TxFIFO empty level bit in the Core
AHB configuration register (PTXFELVL bit in OTG_HS_GAHBCFG).
Note: Only accessible in host mode.

RM0090 Rev 17

1417/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bit 25 HCINT: Host channels interrupt
The core sets this bit to indicate that an interrupt is pending on one of the channels of the
core (in host mode). The application must read the host all channels interrupt
(OTG_HS_HAINT) register to determine the exact number of the channel on which the
interrupt occurred, and then read the corresponding host channel-x interrupt
(OTG_HS_HCINTx) register to determine the exact cause of the interrupt. The application
must clear the appropriate status bit in the OTG_HS_HCINTx register to clear this bit.
Note: Only accessible in host mode.
Bit 24 HPRTINT: Host port interrupt
The core sets this bit to indicate a change in port status of one of the OTG_HS controller
ports in host mode. The application must read the host port control and status
(OTG_HS_HPRT) register to determine the exact event that caused this interrupt. The
application must clear the appropriate status bit in the host port control and status register to
clear this bit.
Note: Only accessible in host mode.
Bits 23 Reserved, must be kept at reset value.
Bit 22 DATAFSUSP: Data fetch suspended
This interrupt is valid only in DMA mode. This interrupt indicates that the core has stopped
fetching data for IN endpoints due to the unavailability of TxFIFO space or request queue
space. This interrupt is used by the application for an endpoint mismatch algorithm. For
example, after detecting an endpoint mismatch, the application:
–
Sets a global nonperiodic IN NAK handshake
–
Disables IN endpoints
–
Flushes the FIFO
–
Determines the token sequence from the IN token sequence learning queue
–
Re-enables the endpoints
–
Clears the global nonperiodic IN NAK handshake If the global nonperiodic IN NAK
is cleared, the core has not yet fetched data for the IN endpoint, and the IN token is
received: the core generates an “IN token received when FIFO empty” interrupt.
The OTG then sends a NAK response to the host. To avoid this scenario, the
application can check the FetSusp interrupt in OTG_FS_GINTSTS, which ensures
that the FIFO is full before clearing a global NAK handshake. Alternatively, the
application can mask the “IN token received when FIFO empty” interrupt when
clearing a global IN NAK handshake.
Bit 21 IPXFR: Incomplete periodic transfer
In host mode, the core sets this interrupt bit when there are incomplete periodic transactions
still pending, which are scheduled for the current frame.
Note: Only accessible in host mode.
INCOMPISOOUT: Incomplete isochronous OUT transfer
In peripheral mode, the core sets this interrupt to indicate that there is at least one
isochronous OUT endpoint on which the transfer is not completed in the current frame. This
interrupt is asserted along with the End of periodic frame interrupt (EOPF) bit in this register.
Note: Only accessible in peripheral mode.
Bit 20 IISOIXFR: Incomplete isochronous IN transfer
The core sets this interrupt to indicate that there is at least one isochronous IN endpoint on
which the transfer is not completed in the current frame. This interrupt is asserted along with
the End of periodic frame interrupt (EOPF) bit in this register.
Note: Only accessible in peripheral mode.

1418/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

Bit 19 OEPINT: OUT endpoint interrupt
The core sets this bit to indicate that an interrupt is pending on one of the OUT endpoints of
the core (in peripheral mode). The application must read the device all endpoints interrupt
(OTG_HS_DAINT) register to determine the exact number of the OUT endpoint on which
the interrupt occurred, and then read the corresponding device OUT Endpoint-x Interrupt
(OTG_HS_DOEPINTx) register to determine the exact cause of the interrupt. The
application must clear the appropriate status bit in the corresponding OTG_HS_DOEPINTx
register to clear this bit.
Note: Only accessible in peripheral mode.
Bit 18 IEPINT: IN endpoint interrupt
The core sets this bit to indicate that an interrupt is pending on one of the IN endpoints of the
core (in peripheral mode). The application must read the device All Endpoints Interrupt
(OTG_HS_DAINT) register to determine the exact number of the IN endpoint on which the
interrupt occurred, and then read the corresponding device IN Endpoint-x interrupt
(OTG_HS_DIEPINTx) register to determine the exact cause of the interrupt. The application
must clear the appropriate status bit in the corresponding OTG_HS_DIEPINTx register to
clear this bit.
Note: Only accessible in peripheral mode.
Bits 17:16 Reserved, must be kept at reset value.
Bit 15 EOPF: End of periodic frame interrupt
Indicates that the period specified in the periodic frame interval field of the device
configuration register (PFIVL bit in OTG_HS_DCFG) has been reached in the current frame.
Note: Only accessible in peripheral mode.
Bit 14 ISOODRP: Isochronous OUT packet dropped interrupt
The core sets this bit when it fails to write an isochronous OUT packet into the RxFIFO
because the RxFIFO does not have enough space to accommodate a maximum size packet
for the isochronous OUT endpoint.
Note: Only accessible in peripheral mode.
Bit 13 ENUMDNE: Enumeration done
The core sets this bit to indicate that speed enumeration is complete. The application must
read the device Status (OTG_HS_DSTS) register to obtain the enumerated speed.
Note: Only accessible in peripheral mode.
Bit 12 USBRST: USB reset
The core sets this bit to indicate that a reset is detected on the USB.
Note: Only accessible in peripheral mode.
Bit 11 USBSUSP: USB suspend
The core sets this bit to indicate that a suspend was detected on the USB. The core enters
the Suspended state when there is no activity on the data lines for a period of 3 ms.
Note: Only accessible in peripheral mode.
Bit 10 ESUSP: Early suspend
The core sets this bit to indicate that an Idle state has been detected on the USB for 3 ms.
Note: Only accessible in peripheral mode.
Bits 9:8

Reserved, must be kept at reset value.

RM0090 Rev 17

1419/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bit 7 GONAKEFF: Global OUT NAK effective
Indicates that the Set global OUT NAK bit in the Device control register (SGONAK bit in
OTG_HS_DCTL), set by the application, has taken effect in the core. This bit can be cleared
by writing the Clear global OUT NAK bit in the Device control register (CGONAK bit in
OTG_HS_DCTL).
Note: Only accessible in peripheral mode.
Bit 6 GINAKEFF: Global IN nonperiodic NAK effective
Indicates that the Set global nonperiodic IN NAK bit in the Device control register (SGINAK
bit in OTG_HS_DCTL), set by the application, has taken effect in the core. That is, the core
has sampled the Global IN NAK bit set by the application. This bit can be cleared by clearing
the Clear global nonperiodic IN NAK bit in the Device control register (CGINAK bit in
OTG_HS_DCTL).
This interrupt does not necessarily mean that a NAK handshake is sent out on the USB. The
STALL bit takes precedence over the NAK bit.
Note: Only accessible in peripheral mode.
Bit 5 NPTXFE: Nonperiodic TxFIFO empty
This interrupt is asserted when the nonperiodic TxFIFO is either half or completely empty,
and there is space in at least one entry to be written to the nonperiodic transmit request
queue. The half or completely empty status is determined by the nonperiodic TxFIFO empty
level bit in the OTG_HS_GAHBCFG register (TXFELVL bit in OTG_HS_GAHBCFG).
Note: Only accessible in host mode.
Bit 4 RXFLVL: RxFIFO nonempty
Indicates that there is at least one packet pending to be read from the RxFIFO.
Note: Accessible in both host and peripheral modes.
Bit 3 SOF: Start of frame
In host mode, the core sets this bit to indicate that an SOF (FS), or Keep-Alive (LS) is
transmitted on the USB. The application must write a 1 to this bit to clear the interrupt.
In peripheral mode, in the core sets this bit to indicate that an SOF token has been received
on the USB. The application can read the Device Status register to get the current frame
number. This interrupt is seen only when the core is operating in FS.
Note: Accessible in both host and peripheral modes.
Bit 2 OTGINT: OTG interrupt
The core sets this bit to indicate an OTG protocol event. The application must read the OTG
Interrupt Status (OTG_HS_GOTGINT) register to determine the exact event that caused this
interrupt. The application must clear the appropriate status bit in the OTG_HS_GOTGINT
register to clear this bit.
Note: Accessible in both host and peripheral modes.
Bit 1 MMIS: Mode mismatch interrupt
The core sets this bit when the application is trying to access:
A host mode register, when the core is operating in peripheral mode
A peripheral mode register, when the core is operating in host mode
The register access is completed on the AHB with an OKAY response, but is ignored by the
core internally and does not affect the operation of the core.
Note: Accessible in both host and peripheral modes.
Bit 0 CMOD: Current mode of operation
Indicates the current mode.
0: Peripheral mode
1: Host mode
Note: Accessible in both host and peripheral modes.

1420/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

OTG_HS interrupt mask register (OTG_HS_GINTMSK)
Address offset: 0x018
Reset value: 0x0000 0000

3

2

1

MMISM

rw

rw

rw

rw

rw

rw

rw

0

Reserved

4

OTGINT

rw

5

SOFM

rw

6

RXFLVLM

rw

7

NPTXFEM

rw

8

GINAKEFFM

rw

9

Reserved

ESUSPM

rw

USBSUSPM

rw

USBRST

rw

ISOODRPM

rw

ENUMDNEM

rw

EOPFM

IEPINT

rw

Reserved

OEPINT

rw

IISOIXFRM

rw

IPXFRM/IISOOXFRM

rw

FSUSPM

PRTIM

rw

Reserved

rw

HCIM

rw

PTXFEM

DISCINT

CIDSCHGM

rw

Reserved

WUIM

SRQIM

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

GONAKEFFM

This register works with the Core interrupt register to interrupt the application. When an
interrupt bit is masked, the interrupt associated with that bit is not generated. However, the
Core Interrupt (OTG_HS_GINTSTS) register bit corresponding to that interrupt is still set.

Bit 31 WUIM: Resume/remote wakeup detected interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both host and peripheral modes.
Bit 30 SRQIM: Session request/new session detected interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both host and peripheral modes.
Bit 29 DISCINT: Disconnect detected interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both host and peripheral modes.
Bit 28 CIDSCHGM: Connector ID status change mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both host and peripheral modes.
Bit 27 Reserved, must be kept at reset value.
Bit 26 PTXFEM: Periodic TxFIFO empty mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in host mode.
Bit 25 HCIM: Host channels interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in host mode.
Bit 24 PRTIM: Host port interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in host mode.
Bit 23 Reserved, must be kept at reset value.

RM0090 Rev 17

1421/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bit 22 FSUSPM: Data fetch suspended mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in peripheral mode.
Bit 21 IPXFRM: Incomplete periodic transfer mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in host mode.
IISOOXFRM: Incomplete isochronous OUT transfer mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in peripheral mode.
Bit 20 IISOIXFRM: Incomplete isochronous IN transfer mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in peripheral mode.
Bit 19 OEPINT: OUT endpoints interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in peripheral mode.
Bit 18 IEPINT: IN endpoints interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in peripheral mode.
Bits 17:16 Reserved, must be kept at reset value.
Bit 15 EOPFM: End of periodic frame interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in peripheral mode.
Bit 14 ISOODRPM: Isochronous OUT packet dropped interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in peripheral mode.
Bit 13 ENUMDNEM: Enumeration done mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in peripheral mode.
Bit 12 USBRST: USB reset mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in peripheral mode.
Bit 11 USBSUSPM: USB suspend mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in peripheral mode.

1422/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

Bit 10 ESUSPM: Early suspend mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in peripheral mode.
Bits 9:8 Reserved, must be kept at reset value.
Bit 7 GONAKEFFM: Global OUT NAK effective mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in peripheral mode.
Bit 6 GINAKEFFM: Global nonperiodic IN NAK effective mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in peripheral mode.
Bit 5 NPTXFEM: Nonperiodic TxFIFO empty mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both peripheral and host modes.
Bit 4 RXFLVLM: Receive FIFO nonempty mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both peripheral and host modes.
Bit 3 SOFM: Start of frame mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both peripheral and host modes.
Bit 2 OTGINT: OTG interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both peripheral and host modes.
Bit 1 MMISM: Mode mismatch interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both peripheral and host modes.
Bit 0 Reserved, must be kept at reset value.

RM0090 Rev 17

1423/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

OTG_HS Receive status debug read/OTG status read and pop registers
(OTG_HS_GRXSTSR/OTG_HS_GRXSTSP)
Address offset for Read: 0x01C
Address offset for Pop: 0x020
Reset value: 0x0000 0000
A read to the Receive status debug read register returns the contents of the top of the
Receive FIFO. A read to the Receive status read and pop register additionally pops the top
data entry out of the RxFIFO.
The receive status contents must be interpreted differently in host and peripheral modes.
The core ignores the receive status pop/read when the receive FIFO is empty and returns a
value of 0x0000 0000. The application must only pop the Receive Status FIFO when the
Receive FIFO nonempty bit of the Core interrupt register (RXFLVL bit in
OTG_HS_GINTSTS) is asserted.

Host mode:
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

7

6

5

4

2

1

DPID

BCNT

CHNUM

r

r

r

r

Bits 31:21 Reserved, must be kept at reset value.
Bits 20:17 PKTSTS: Packet status
Indicates the status of the received packet
0010: IN data packet received
0011: IN transfer completed (triggers an interrupt)
0101: Data toggle error (triggers an interrupt)
0111: Channel halted (triggers an interrupt)
Others: Reserved
Bits 16:15 DPID: Data PID
Indicates the Data PID of the received packet
00: DATA0
10: DATA1
01: DATA2
11: MDATA
Bits 14:4 BCNT: Byte count
Indicates the byte count of the received IN data packet.
Bits 3:0 CHNUM: Channel number
Indicates the channel number to which the current received packet belongs.

1424/1747

3

PKTSTS

RM0090 Rev 17

0

RM0090

USB on-the-go high-speed (OTG_HS)

Peripheral mode:
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

7

6

5

4

3

2

1

FRMNUM

PKTSTS

DPID

BCNT

EPNUM

r

r

r

r

r

0

Bits 31:25 Reserved, must be kept at reset value.
Bits 24:21 FRMNUM: Frame number
This is the least significant 4 bits of the frame number in which the packet is received on the
USB. This field is supported only when isochronous OUT endpoints are supported.
Bits 20:17 PKTSTS: Packet status
Indicates the status of the received packet
0001: Global OUT NAK (triggers an interrupt)
0010: OUT data packet received
0011: OUT transfer completed (triggers an interrupt)
0100: SETUP transaction completed (triggers an interrupt)
0110: SETUP data packet received
Others: Reserved
Bits 16:15 DPID: Data PID
Indicates the Data PID of the received OUT data packet
00: DATA0
10: DATA1
01: DATA2
11: MDATA
Bits 14:4 BCNT: Byte count
Indicates the byte count of the received data packet.
Bits 3:0 EPNUM: Endpoint number
Indicates the endpoint number to which the current received packet belongs.

OTG_HS Receive FIFO size register (OTG_HS_GRXFSIZ)
Address offset: 0x024
Reset value: 0x0000 0400
The application can program the RAM size that must be allocated to the RxFIFO.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

RXFD

Reserved

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 RXFD: RxFIFO depth
This value is in terms of 32-bit words.
Minimum value is 16
Maximum value is 1024
The power-on reset value of this register is specified as the largest Rx data FIFO depth.

RM0090 Rev 17

1425/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

OTG_HS nonperiodic transmit FIFO size/Endpoint 0 transmit FIFO size
register (OTG_HS_GNPTXFSIZ/OTG_HS_TX0FSIZ)
Address offset: 0x028
Reset value: 0x0000 0200

Host mode:
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

NPTXFD

NPTXFSA

rw

rw

6

5

4

3

2

1

0

2

1

0

Bits 31:16 NPTXFD: Nonperiodic TxFIFO depth
This value is in terms of 32-bit words.
Minimum value is 16
Maximum value is 1024
Bits 15:0 NPTXFSA: Nonperiodic transmit RAM start address
This field contains the memory start address for nonperiodic transmit FIFO RAM.

Peripheral mode:
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

TX0FD

TX0FSA

r/rw

r/rw

6

5

4

3

Bits 31:16 T0XFD: Endpoint 0 TxFIFO depth
This value is in terms of 32-bit words.
Minimum value is 16
Maximum value is 256
Bits 15:0 TX0FSA: Endpoint 0 transmit RAM start address
This field contains the memory start address for Endpoint 0 transmit FIFO RAM.

OTG_HS nonperiodic transmit FIFO/queue status register
(OTG_HS_GNPTXSTS)
Address offset: 0x02C
Reset value: 0x0008 0400
Note:

In peripheral mode, this register is not valid.
This read-only register contains the free space information for the nonperiodic TxFIFO and
the nonperiodic transmit request queue.

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

1426/1747

9

8

7

NPTXQTOP

NPTQXSAV

NPTXFSAV

r

r

r

RM0090 Rev 17

6

5

4

3

2

1

0

RM0090

USB on-the-go high-speed (OTG_HS)

Bit 31 Reserved, must be kept at reset value.
Bits 30:24 NPTXQTOP: Top of the nonperiodic transmit request queue
Entry in the nonperiodic Tx request queue that is currently being processed by the MAC.
Bits [30:27]: Channel/endpoint number
Bits [26:25]:
–
00: IN/OUT token
–
01: Zero-length transmit packet (device IN/host OUT)
–
10: PING/CSPLIT token
–
11: Channel halt command
Bit [24]: Terminate (last entry for selected channel/endpoint)
Bits 23:16 NPTQXSAV: Nonperiodic transmit request queue space available
Indicates the amount of free space available in the nonperiodic transmit request queue.
This queue holds both IN and OUT requests in host mode. Peripheral mode has only IN
requests.
00: Nonperiodic transmit request queue is full
01: dx1 location available
10: dx2 locations available
bxn: dxn locations available (0 ≤ n ≤ dx8)
Others: Reserved
Bits 15:0 NPTXFSAV: Nonperiodic TxFIFO space available
Indicates the amount of free space available in the nonperiodic TxFIFO.
Values are in terms of 32-bit words.
00: Nonperiodic TxFIFO is full
01: dx1 word available
10: dx2 words available
0xn: dxn words available (where 0 ≤ n ≤ dx1024)
Others: Reserved

OTG_HS general core configuration register (OTG_HS_GCCFG)
Address offset: 0x038
Reset value: 0x0000 XXXX

SOFOUTEN

VBUSBSEN

VBUSASEN

I2CPADEN

.PWRDWN

Reserved

NOVBUSSENS

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

rw

rw

rw

rw

rw

rw

RM0090 Rev 17

9

8

7

6

5

4

3

2

1

0

Reserved

1427/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bits 31:22 Reserved, must be kept at reset value.
Bit 21 NOVBUSSENS: VBUS sensing disable option
When this bit is set, VBUS is considered internally to be always at VBUS valid level (5 V). This
option removes the need for a dedicated VBUS pad, and leave this pad free to be used for
other purposes such as a shared functionality. VBUS connection can be remapped on
another general purpose input pad and monitored by software.
This option is only suitable for host-only or device-only applications.
0: VBUS sensing available by hardware
1: VBUS sensing not available by hardware.
Bit 20 SOFOUTEN: SOF output enable
0: SOF pulse not available on PAD
1: SOF pulse available on PAD
Bit 19 VBUSBSEN: Enable the VBUS sensing “B” device
0: VBUS sensing “B” disabled
1: VBUS sensing “B” enabled
Bit 18 VBUSASEN: Enable the VBUS sensing “A” device
0: VBUS sensing “A” disabled
1: VBUS sensing “A” enabled
Bit 17 I2CPADEN: Enable I2C bus connection for the external I2C PHY interface.
0: I2C bus disabled
1: I2C bus enabled
Bit 16 PWRDWN: Power down
Used to activate the transceiver in transmission/reception
0: Power down active
1: Power down deactivated (“Transceiver active”)
Bits 15:0 Reserved, must be kept at reset value.

OTG_HS core ID register (OTG_HS_CID)
Address offset: 0x03C
Reset value:0x0000 1100
This is a register containing the Product ID as reset value.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

PRODUCT_ID
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 PRODUCT_ID: Product ID field
Application-programmable ID field.

OTG_HS Host periodic transmit FIFO size register (OTG_HS_HPTXFSIZ)
Address offset: 0x100
Reset value: 0x0200 0800

1428/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

PTXFD
r

r

r

r

r

r

r

r

r

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

PTXSA
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:16 PTXFD: Host periodic TxFIFO depth
This value is in terms of 32-bit words.
Minimum value is 16
Maximum value is 512
Bits 15:0 PTXSA: Host periodic TxFIFO start address
The power-on reset value of this register is the sum of the largest Rx data FIFO depth and
largest nonperiodic Tx data FIFO depth.

OTG_HS device IN endpoint transmit FIFO size register (OTG_HS_DIEPTXFx)
(x = 1..7, where x is the FIFO_number)
Address offset: 0x104 + 0x04 * (x - 1)
Reset value: 0x02000400
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

INEPTXFD
r

r

r

r

r

r

r

r

r

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

INEPTXSA
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:16 INEPTXFD: IN endpoint TxFIFO depth
This value is in terms of 32-bit words.
Minimum value is 16
Maximum value is 512
The power-on reset value of this register is specified as the largest IN endpoint FIFO
number depth.
Bits 15:0 INEPTXSA: IN endpoint FIFOx transmit RAM start address
This field contains the memory start address for IN endpoint transmit FIFOx. The address
must be aligned with a 32-bit memory location.

35.12.3

Host-mode registers
Bit values in the register descriptions are expressed in binary unless otherwise specified.
Host-mode registers affect the operation of the core in the host mode. Host mode registers
must not be accessed in peripheral mode, as the results are undefined. Host mode registers
can be categorized as follows:

OTG_HS host configuration register (OTG_HS_HCFG)
Address offset: 0x400
Reset value: 0x0000 0000
This register configures the core after power-on. Do not change to this register after
initializing the host.

RM0090 Rev 17

1429/1747
1541

RM0090

9

8

7

6

5

4

3

2

1

r

0
FSLSPCS

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

FSLSS

USB on-the-go high-speed (OTG_HS)

rw

Bits 31:3 Reserved, must be kept at reset value.
Bit 2 FSLSS: FS- and LS-only support
The application uses this bit to control the core’s enumeration speed. Using this bit, the
application can make the core enumerate as an FS host, even if the connected device
supports HS traffic. Do not make changes to this field after initial programming.
0: HS/FS/LS, based on the maximum speed supported by the connected device
1: FS/LS-only, even if the connected device can support HS (read-only)
Bits 1:0 FSLSPCS: FS/LS PHY clock select
When the core is in FS host mode:
01: PHY clock is running at 48 MHz
Others: Reserved
When the core is in LS host mode:
00: Reserved
01: PHY clock is running at 48 MHz.
10: Select 6 MHz PHY clock frequency
11: Reserved
Note: The FSLSPCS bit must be set on a connection event according to the speed of the
connected device. A software reset must be performed after changing this bit.

1430/1747

RM0090 Rev 17

rw

RM0090

USB on-the-go high-speed (OTG_HS)

OTG_HS Host frame interval register (OTG_HS_HFIR)
Address offset: 0x404
Reset value: 0x0000 EA60
This register stores the frame interval information for the current speed to which the
OTG_HS controller has enumerated.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

FRIVL

Reserved

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 FRIVL: Frame interval
The value that the application programs to this field specifies the interval between two
consecutive SOFs (FS), micro-SOFs (HS) or Keep-Alive tokens (LS). This field contains the
number of PHY clocks that constitute the required frame interval. The application can write a
value to this register only after the Port enable bit of the host port control and status register
(PENA bit in OTG_HS_HPRT) has been set. If no value is programmed, the core calculates
the value based on the PHY clock specified in the FS/LS PHY Clock Select field of the Host
configuration register (FSLSPCS in OTG_HS_HCFG):
frame duration × PHY clock frequency
– Frame interval = 1 ms × (FRIVL - 1)
Note: The FRIVL bit can be modified whenever the application needs to change the Frame
interval time.

OTG_HS host frame number/frame time remaining register
(OTG_HS_HFNUM)
Address offset: 0x408
Reset value: 0x0000 3FFF
This register indicates the current frame number. It also indicates the time remaining (in
terms of the number of PHY clocks) in the current frame.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

FTREM
r

r

r

r

r

r

r

r

r

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

FRNUM
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:16 FTREM: Frame time remaining
Indicates the amount of time remaining in the current frame, in terms of PHY clocks. This
field decrements on each PHY clock. When it reaches zero, this field is reloaded with the
value in the Frame interval register and a new SOF is transmitted on the USB.
Bits 15:0 FRNUM: Frame number
This field increments when a new SOF is transmitted on the USB, and is cleared to 0 when
it reaches 0x3FFF.

RM0090 Rev 17

1431/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

OTG_HS_Host periodic transmit FIFO/queue status register
(OTG_HS_HPTXSTS)
Address offset: 0x410
Reset value: 0x0008 0100
This read-only register contains the free space information for the periodic TxFIFO and the
periodic transmit request queue.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
PTXQTOP
r

r

r

r

r

9

PTXQSAV
r

r

r

r

r

r

r

r

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

PTXFSAVL
r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:24 PTXQTOP: Top of the periodic transmit request queue
This indicates the entry in the periodic Tx request queue that is currently being processed by
the MAC.
This register is used for debugging.
Bit [31]: Odd/Even frame
–
0: send in even (micro) frame
–
1: send in odd (micro) frame
Bits [30:27]: Channel/endpoint number
Bits [26:25]: Type
–
00: IN/OUT
–
01: Zero-length packet
–
11: Disable channel command
Bit [24]: Terminate (last entry for the selected channel/endpoint)
Bits 23:16 PTXQSAV: Periodic transmit request queue space available
Indicates the number of free locations available to be written in the periodic transmit request
queue. This queue holds both IN and OUT requests.
00: Periodic transmit request queue is full
01: dx1 location available
10: dx2 locations available
bxn: dxn locations available (0 ≤ dxn ≤ PTXFD)
Others: Reserved
Bits 15:0 PTXFSAVL: Periodic transmit data FIFO space available
Indicates the number of free locations available to be written to in the periodic TxFIFO.
Values are in terms of 32-bit words
0000: Periodic TxFIFO is full
0001: dx1 word available
0010: dx2 words available
bxn: dxn words available (where 0 ≤ dxn ≤ dx512)
Others: Reserved

1432/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

OTG_HS Host all channels interrupt register (OTG_HS_HAINT)
Address offset: 0x414
Reset value: 0x0000 000
When a significant event occurs on a channel, the host all channels interrupt register
interrupts the application using the host channels interrupt bit of the Core interrupt register
(HCINT bit in OTG_HS_GINTSTS). This is shown in Figure 414. There is one interrupt bit
per channel, up to a maximum of 16 bits. Bits in this register are set and cleared when the
application sets and clears bits in the corresponding host channel-x interrupt register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

HAINT
r

r

r

r

r

r

r

r

r

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 HAINT: Channel interrupts
One bit per channel: Bit 0 for Channel 0, bit 15 for Channel 15

OTG_HS host all channels interrupt mask register (OTG_HS_HAINTMSK)
Address offset: 0x418
Reset value: 0x0000 0000
The host all channel interrupt mask register works with the host all channel interrupt register
to interrupt the application when an event occurs on a channel. There is one interrupt mask
bit per channel, up to a maximum of 16 bits.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

HAINTM
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 HAINTM: Channel interrupt mask
0: Masked interrupt
1: Unmasked interrupt
One bit per channel: Bit 0 for channel 0, bit 15 for channel 15

RM0090 Rev 17

1433/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

OTG_HS host port control and status register (OTG_HS_HPRT)
Address offset: 0x440
Reset value: 0x0000 0000
This register is available only in host mode. Currently, the OTG host supports only one port.

rw

rw

rw

rw

r

r

rs

rw

rc_
w1

r

2

1

0
PCSTS

rw

3

PENA

6

PCDET

7

PENCHNG

POCA

rw

4

POCCHNG

r

5

PRES

r

8
PRST

PTCTL

Reserved

9

Reserved

PPWR

PSPD

PLSTS

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

PSUSP

A single register holds USB port-related information such as USB reset, enable, suspend,
resume, connect status, and test mode for each port. It is shown in Figure 414. The rc_w1
bits in this register can trigger an interrupt to the application through the host port interrupt
bit of the core interrupt register (HPRTINT bit in OTG_HS_GINTSTS). On a Port Interrupt,
the application must read this register and clear the bit that caused the interrupt. For the
rc_w1 bits, the application must write a 1 to the bit to clear the interrupt.

rc_ rc_ rc_
w1 w0 w1

r

Bits 31:19 Reserved, must be kept at reset value.
Bits 18:17 PSPD: Port speed
Indicates the speed of the device attached to this port.
00: High speed
01: Full speed
10: Low speed
11: Reserved
Bits 16:13 PTCTL: Port test control
The application writes a nonzero value to this field to put the port into a Test mode, and the
corresponding pattern is signaled on the port.
0000: Test mode disabled
0001: Test_J mode
0010: Test_K mode
0011: Test_SE0_NAK mode
0100: Test_Packet mode
0101: Test_Force_Enable
Others: Reserved
Bit 12 PPWR: Port power
The application uses this field to control power to this port, and the core clears this bit on an
overcurrent condition.
0: Power off
1: Power on
Bits 11:10 PLSTS: Port line status
Indicates the current logic level USB data lines
Bit [10]: Logic level of OTG_HS_FS_DP
Bit [11]: Logic level of OTG_HS_FS_DM
Bit 9 Reserved, must be kept at reset value.

1434/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

Bit 8 PRST: Port reset
When the application sets this bit, a reset sequence is started on this port. The application
must time the reset period and clear this bit after the reset sequence is complete.
0: Port not in reset
1: Port in reset
The application must leave this bit set for a minimum duration of at least 10 ms to start a
reset on the port. The application can leave it set for another 10 ms in addition to the
required minimum duration, before clearing the bit, even though there is no maximum limit
set by the USB standard.
High speed: 50 ms
Full speed/Low speed: 10 ms
Bit 7 PSUSP: Port suspend
The application sets this bit to put this port in Suspend mode. The core only stops sending
SOFs when this is set. To stop the PHY clock, the application must set the Port clock stop
bit, which asserts the suspend input pin of the PHY.
The read value of this bit reflects the current suspend status of the port. This bit is cleared
by the core after a remote wakeup signal is detected or the application sets the Port reset bit
or Port resume bit in this register or the Resume/remote wakeup detected interrupt bit or
Disconnect detected interrupt bit in the Core interrupt register (WKUINT or DISCINT in
OTG_HS_GINTSTS, respectively).
0: Port not in Suspend mode
1: Port in Suspend mode
Bit 6

PRES: Port resume
The application sets this bit to drive resume signaling on the port. The core continues to
drive the resume signal until the application clears this bit.
If the core detects a USB remote wakeup sequence, as indicated by the Port
resume/remote wakeup detected interrupt bit of the Core interrupt register (WKUINT bit in
OTG_HS_GINTSTS), the core starts driving resume signaling without application
intervention and clears this bit when it detects a disconnect condition. The read value of this
bit indicates whether the core is currently driving resume signaling.
0: No resume driven
1: Resume driven

Bit 5 POCCHNG: Port overcurrent change
The core sets this bit when the status of the Port overcurrent active bit (bit 4) in this register
changes.
Bit 4 POCA: Port overcurrent active
Indicates the overcurrent condition of the port.
0: No overcurrent condition
1: Overcurrent condition
Bit 3 PENCHNG: Port enable/disable change
The core sets this bit when the status of the Port enable bit [2] in this register changes.

RM0090 Rev 17

1435/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bit 2 PENA: Port enable
A port is enabled only by the core after a reset sequence, and is disabled by an overcurrent
condition, a disconnect condition, or by the application clearing this bit. The application
cannot set this bit by a register write. It can only clear it to disable the port. This bit does not
trigger any interrupt to the application.
0: Port disabled
1: Port enabled
Bit 1 PCDET: Port connect detected
The core sets this bit when a device connection is detected to trigger an interrupt to the
application using the host port interrupt bit in the Core interrupt register (HPRTINT bit in
OTG_HS_GINTSTS). The application must write a 1 to this bit to clear the interrupt.
Bit 0 PCSTS: Port connect status
0: No device is attached to the port
1: A device is attached to the port

OTG_HS host channel-x characteristics register (OTG_HS_HCCHARx)
(x = 0..11, where x = Channel_number)
Address offset: 0x500 + 0x20 * x
Reset value: 0x0000 0000

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

EPDIR

rw

Reserved

ODDFRM

rs

MC

LSDEV

CHDIS

rs

DAD

EPTYP

CHENA

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

rw

9

8

7

6

EPNUM
rw

rw

rw

5

4

3

2

1

0

rw

rw

rw

rw

rw

MPSIZ
rw

rw

rw

rw

rw

rw

rw

Bit 31 CHENA: Channel enable
This field is set by the application and cleared by the OTG host.
0: Channel disabled
1: Channel enabled
Bit 30 CHDIS: Channel disable
The application sets this bit to stop transmitting/receiving data on a channel, even before
the transfer for that channel is complete. The application must wait for the Channel disabled
interrupt before treating the channel as disabled.
Bit 29 ODDFRM: Odd frame
This field is set (reset) by the application to indicate that the OTG host must perform a
transfer in an odd frame. This field is applicable for only periodic (isochronous and interrupt)
transactions.
0: Even (micro) frame
1: Odd (micro) frame
Bits 28:22 DAD: Device address
This field selects the specific device serving as the data source or sink.

1436/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

Bits 21:20 MC: Multi Count (MC) / Error Count (EC)
– When the split enable bit (SPLITEN) in the host channel-x split control register
(OTG_HS_HCSPLTx) is reset (0), this field indicates to the host the number of transactions
that must be executed per micro-frame for this periodic endpoint. For nonperiodic transfers,
this field specifies the number of packets to be fetched for this channel before the internal
DMA engine changes arbitration.
00: Reserved This field yields undefined results
01: 1 transaction
b10: 2 transactions to be issued for this endpoint per micro-frame
11: 3 transactions to be issued for this endpoint per micro-frame.
– When the SPLITEN bit is set (1) in OTG_HS_HCSPLTx, this field indicates the number of
immediate retries to be performed for a periodic split transaction on transaction errors. This
field must be set to at least 01.
Bits 19:18 EPTYP: Endpoint type
Indicates the transfer type selected.
00: Control
01: Isochronous
10: Bulk
11: Interrupt
Bit 17 LSDEV: Low-speed device
This field is set by the application to indicate that this channel is communicating to a lowspeed device.
Bit 16 Reserved, must be kept at reset value.
Bit 15 EPDIR: Endpoint direction
Indicates whether the transaction is IN or OUT.
0: OUT
1: IN
Bits 14:11 EPNUM: Endpoint number
Indicates the endpoint number on the device serving as the data source or sink.
Bits 10:0 MPSIZ: Maximum packet size
Indicates the maximum packet size of the associated endpoint.

RM0090 Rev 17

1437/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

OTG_HS host channel-x split control register (OTG_HS_HCSPLTx) (x = 0..11,
where x = Channel_number)
Address offset: 0x504 + 0x20 * x
Reset value: 0x0000 0000

rw

rw

rw

rw

9

8

7

6

5

4

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

3

2

1

0

rw

rw

rw

PRTADDR

HUBADDR

XACTPOS

COMPLSPLT

Reserved

SPLITEN

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

rw

Bit 31 SPLITEN: Split enable
The application sets this bit to indicate that this channel is enabled to perform split
transactions.
Bits 30:17

Reserved, must be kept at reset value.

Bit 16 COMPLSPLT: Do complete split
The application sets this bit to request the OTG host to perform a complete split transaction.
Bits 15:14 XACTPOS: Transaction position
This field is used to determine whether to send all, first, middle, or last payloads with each
OUT transaction.
11: All. This is the entire data payload of this transaction (which is less than or equal to 188
bytes)
10: Begin. This is the first data payload of this transaction (which is larger than 188 bytes)
00: Mid. This is the middle payload of this transaction (which is larger than 188 bytes)
01: End. This is the last payload of this transaction (which is larger than 188 bytes)
Bits 13:7 HUBADDR: Hub address
This field holds the device address of the transaction translator’s hub.
Bits 6:0 PRTADDR: Port address
This field is the port number of the recipient transaction translator.

1438/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

OTG_HS host channel-x interrupt register (OTG_HS_HCINTx) (x = 0..11, where
x = Channel_number)
Address offset: 0x508 + 0x20 * x
Reset value: 0x0000 0000

5

FRMOR

BBERR

TXERR

NYET

ACK

Reserved

4

3

2

1

0

CHH

6

XFRC

7

AHBERR

8

NAK

9

STALL

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
DTERR

This register indicates the status of a channel with respect to USB- and AHB-related events.
It is shown in Figure 414. The application must read this register when the host channels
interrupt bit in the Core interrupt register (HCINT bit in OTG_HS_GINTSTS) is set. Before
the application can read this register, it must first read the host all channels interrupt
(OTG_HS_HAINT) register to get the exact channel number for the host channel-x interrupt
register. The application must clear the appropriate bit in this register to clear the
corresponding bits in the OTG_HS_HAINT and OTG_HS_GINTSTS registers.

rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_
w1 w1 w1 w1 w1 w1 w1 w1 w1 w1 w1

Bits 31:11 Reserved, must be kept at reset value.
Bit 10 DTERR: Data toggle error
Bit 9 FRMOR: Frame overrun
Bit 8 BBERR: Babble error
Bit 7 TXERR: Transaction error
Indicates one of the following errors occurred on the USB.
CRC check failure
Timeout
Bit stuff error
False EOP
Bit 6 NYET: Response received interrupt
Bit 5 ACK: ACK response received/transmitted interrupt
Bit 4 NAK: NAK response received interrupt
Bit 3 STALL: STALL response received interrupt
Bit 2 AHBERR: AHB error
This error is generated only in Internal DMA mode when an AHB error occurs during an AHB
read/write operation. The application can read the corresponding DMA channel address
register to get the error address.
Bit 1 CHH: Channel halted
Indicates the transfer completed abnormally either because of any USB transaction error or in
response to disable request by the application.
Bit 0 XFRC: Transfer completed
Transfer completed normally without any errors.

RM0090 Rev 17

1439/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

OTG_HS host channel-x interrupt mask register (OTG_HS_HCINTMSKx)
(x = 0..11, where x = Channel_number)
Address offset: 0x50C + 0x20 * x
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

DTERRM

FRMORM

BBERRM

TXERRM

NYET

ACKM

NAKM

STALLM

AHBERRM

CHHM

XFRCM

This register reflects the mask for each channel status described in the previous section.

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Reserved

Bits 31:11 Reserved, must be kept at reset value.
Bit 10 DTERRM: Data toggle error mask
0: Masked interrupt
1: Unmasked interrupt
Bit 9 FRMORM: Frame overrun mask
0: Masked interrupt
1: Unmasked interrupt
Bit 8 BBERRM: Babble error mask
0: Masked interrupt
1: Unmasked interrupt
Bit 7 TXERRM: Transaction error mask
0: Masked interrupt
1: Unmasked interrupt
Bit 6 NYET: response received interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 5 ACKM: ACK response received/transmitted interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 4 NAKM: NAK response received interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 3 STALLM: STALL response received interrupt mask
0: Masked interrupt
1: Unmasked interrupt

1440/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

Bit 2 AHBERRM: AHB error mask
0: Masked interrupt
1: Unmasked interrupt
Bit 1 CHHM: Channel halted mask
0: Masked interrupt
1: Unmasked interrupt
Bit 0 XFRCM: Transfer completed mask
0: Masked interrupt
1: Unmasked interrupt

OTG_HS host channel-x transfer size register (OTG_HS_HCTSIZx) (x = 0..11,
where x = Channel_number)
Address offset: 0x510 + 0x20 * x
Reset value: 0x0000 0000

DOPING

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

rw

DPID
rw

rw

PKTCNT
rw

rw

rw

rw

rw

rw

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

XFRSIZ
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bit 31 DOPING: Do ping
This bit is used only for OUT transfers. Setting this field to 1 directs the host to do PING
protocol.
Note: Do not set this bit for IN transfers. If this bit is set for IN transfers it disables the channel.
Bits 30:29 DPID: Data PID
The application programs this field with the type of PID to use for the initial transaction. The
host maintains this field for the rest of the transfer.
00: DATA0
01: DATA2
10: DATA1
11: MDATA (noncontrol)/SETUP (control)
Bits 28:19 PKTCNT: Packet count
This field is programmed by the application with the expected number of packets to be
transmitted (OUT) or received (IN).
The host decrements this count on every successful transmission or reception of an OUT/IN
packet. Once this count reaches zero, the application is interrupted to indicate normal
completion.
Bits 18:0 XFRSIZ: Transfer size
For an OUT, this field is the number of data bytes the host sends during the transfer.
For an IN, this field is the buffer size that the application has reserved for the transfer. The
application is expected to program this field as an integer multiple of the maximum packet
size for IN transactions (periodic and nonperiodic).

RM0090 Rev 17

1441/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

OTG_HS host channel-x DMA address register (OTG_HS_HCDMAx) (x = 0..11,
where x = Channel_number)
Address offset: 0x514 + 0x20 * x
Reset value: 0x0000 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

DMAADDR
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:0 DMAADDR: DMA address
This field holds the start address in the external memory from which the data for the endpoint
must be fetched or to which it must be stored. This register is incremented on every AHB
transaction.

35.12.4

Device-mode registers
OTG_HS device configuration register (OTG_HS_DCFG)
Address offset: 0x800
Reset value: 0x0220 0000

rw

rw

rw

rw

6

5

4

rw

rw

rw

rw

rw

3

2

rw

1

0
DSPD

rw

7

NZLSOHSK

8

DAD

Reserved

9

PFIVL

Reserved

Reserved

PERSCHIVL

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

Reserved

This register configures the core in peripheral mode after power-on or after certain control
commands or enumeration. Do not make changes to this register after initial programming.

rw

Bits 31:26 Reserved, must be kept at reset value.
Bits 25:24 PERSCHIVL: Periodic scheduling interval
This field specifies the amount of time the Internal DMA engine must allocate for fetching
periodic IN endpoint data. Based on the number of periodic endpoints, this value must be
specified as 25, 50 or 75% of the (micro)frame.
–
When any periodic endpoints are active, the internal DMA engine allocates the
specified amount of time in fetching periodic IN endpoint data
–
When no periodic endpoint is active, then the internal DMA engine services
nonperiodic endpoints, ignoring this field
–
After the specified time within a (micro)frame, the DMA switches to fetching
nonperiodic endpoints
00: 25% of (micro)frame
01: 50% of (micro)frame
10: 75% of (micro)frame
11: Reserved
Bits 23:13 Reserved, must be kept at reset value.

1442/1747

RM0090 Rev 17

rw

RM0090

USB on-the-go high-speed (OTG_HS)

Bits 12:11 PFIVL: Periodic (micro)frame interval
Indicates the time within a (micro) frame at which the application must be notified using the
end of periodic (micro) frame interrupt. This can be used to determine if all the isochronous
traffic for that frame is complete.
00: 80% of the frame interval
01: 85% of the frame interval
10: 90% of the frame interval
11: 95% of the frame interval
Bits 10:4 DAD: Device address
The application must program this field after every SetAddress control command.
Bit 3 Reserved, must be kept at reset value.
Bit 2 NZLSOHSK: Nonzero-length status OUT handshake
The application can use this field to select the handshake the core sends on receiving a
nonzero-length data packet during the OUT transaction of a control transfer’s Status stage.
1: Send a STALL handshake on a nonzero-length status OUT transaction and do not send
the received OUT packet to the application.
0: Send the received OUT packet to the application (zero-length or nonzero-length) and
send a handshake based on the NAK and STALL bits for the endpoint in the device endpoint
control register.
Bits 1:0 DSPD: Device speed
Indicates the speed at which the application requires the core to enumerate, or the
maximum speed the application can support. However, the actual bus speed is determined
only after the chirp sequence is completed, and is based on the speed of the USB host to
which the core is connected.
00: High speed
01: Full speed using external ULPI PHY
10: Reserved
11: Full speed using internal embedded PHY

RM0090 Rev 17

1443/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

OTG_HS device control register (OTG_HS_DCTL)
Address offset: 0x804

w

w

rw

rw

rw

3

2

1

0

SDIS

w

4

RWUSIG

w

5

GINSTS

SGINAK

rw

6

TCTL

CGINAK

7

SGONAK

8

CGONAK

Reserved

9

POPRGDNE

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

GONSTS

Reset value: 0x0000 0000

r

r

rw

rw

Bits 31:12 Reserved, must be kept at reset value.
Bit 11 POPRGDNE: Power-on programming done
The application uses this bit to indicate that register programming is completed after a
wakeup from power down mode.
Bit 10 CGONAK: Clear global OUT NAK
Writing 1 to this field clears the Global OUT NAK.
Bit 9 SGONAK: Set global OUT NAK
Writing 1 to this field sets the Global OUT NAK.
The application uses this bit to send a NAK handshake on all OUT endpoints.
The application must set the this bit only after making sure that the Global OUT NAK
effective bit in the Core interrupt register (GONAKEFF bit in OTG_HS_GINTSTS) is
cleared.
Bit 8 CGINAK: Clear global IN NAK
Writing 1 to this field clears the Global IN NAK.
Bit 7 SGINAK: Set global IN NAK
Writing 1 to this field sets the Global nonperiodic IN NAK.The application uses this bit to
send a NAK handshake on all nonperiodic IN endpoints.
The application must set this bit only after making sure that the Global IN NAK effective bit
in the Core interrupt register (GINAKEFF bit in OTG_HS_GINTSTS) is cleared.
Bits 6:4 TCTL: Test control
000: Test mode disabled
001: Test_J mode
010: Test_K mode
011: Test_SE0_NAK mode
100: Test_Packet mode
101: Test_Force_Enable
Others: Reserved
Bit 3 GONSTS: Global OUT NAK status
0: A handshake is sent based on the FIFO Status and the NAK and STALL bit settings.
1: No data is written to the RxFIFO, irrespective of space availability. Sends a NAK
handshake on all packets, except on SETUP transactions. All isochronous OUT packets are
dropped.

1444/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

Bit 2 GINSTS: Global IN NAK status
0: A handshake is sent out based on the data availability in the transmit FIFO.
1: A NAK handshake is sent out on all nonperiodic IN endpoints, irrespective of the data
availability in the transmit FIFO.
Bit 1 SDIS: Soft disconnect
The application uses this bit to signal the USB OTG core to perform a soft disconnect. As
long as this bit is set, the host does not see that the device is connected, and the device
does not receive signals on the USB. The core stays in the disconnected state until the
application clears this bit.
0: Normal operation. When this bit is cleared after a soft disconnect, the core generates a
device connect event to the USB host. When the device is reconnected, the USB host
restarts device enumeration.
1: The core generates a device disconnect event to the USB host.
Bit 0 RWUSIG: Remote wakeup signaling
When the application sets this bit, the core initiates remote signaling to wake up the USB
host. The application must set this bit to instruct the core to exit the Suspend state. As
specified in the USB 2.0 specification, the application must clear this bit 1 ms to 15 ms after
setting it.

Table 214 contains the minimum duration (according to device state) for which the Soft
disconnect (SDIS) bit must be set for the USB host to detect a device disconnect. To
accommodate clock jitter, it is recommended that the application add some extra delay to
the specified minimum duration.
Table 214. Minimum duration for soft disconnect
Operating speed

Device state

Minimum duration

High speed

Not Idle or Suspended (Performing transactions)

125 µs

Full speed

Suspended

1 ms + 2.5 µs

Full speed

Idle

2.5 µs

Full speed

Not Idle or Suspended (Performing transactions)

2.5 µs

RM0090 Rev 17

1445/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

OTG_HS device status register (OTG_HS_DSTS)
Address offset: 0x808
Reset value: 0x0000 0010
This register indicates the status of the core with respect to USB-related events. It must be
read on interrupts from the device all interrupts (OTG_HS_DAINT) register.

FNSOF

Reserved
r

r

r

r

r

r

r

r

7

6

5

Reserved
r

r

r

r

r

r

4

3

2

r

1

r

0
SUSPSTS

8

ENUMSPD

9

EERR

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

r

r

Bits 31:22 Reserved, must be kept at reset value.
Bits 21:8 FNSOF: Frame number of the received SOF
Bits 7:4 Reserved, must be kept at reset value.
Bit 3 EERR: Erratic error
The core sets this bit to report any erratic errors.
Due to erratic errors, the OTG_HS controller goes into Suspended state and an interrupt is
generated to the application with Early suspend bit of the Core interrupt register (ESUSP bit
in OTG_HS_GINTSTS). If the early suspend is asserted due to an erratic error, the
application can only perform a soft disconnect recover.
Bits 2:1 ENUMSPD: Enumerated speed
Indicates the speed at which the OTG_HS controller has come up after speed detection
through a chirp sequence.
00: High speed
01: Reserved
10: Reserved
11: Full speed (PHY clock is running at 48 MHz)
Others: reserved
Bit 0 SUSPSTS: Suspend status
In peripheral mode, this bit is set as long as a Suspend condition is detected on the USB.
The core enters the Suspended state when there is no activity on the USB data lines for a
period of 3 ms. The core comes out of the suspend:
– When there is an activity on the USB data lines
– When the application writes to the Remote wakeup signaling bit in the Device control register
(RWUSIG bit in OTG_HS_DCTL).

1446/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

OTG_HS device IN endpoint common interrupt mask register
(OTG_HS_DIEPMSK)
Address offset: 0x810
Reset value: 0x0000 0000

5

4

3

2

1

0

TOM

AHBERRM

EPDM

XFRCM

rw

6

ITTXFEMSK

rw

7

INEPNEM

Reserved

8

Reserved

Reserved

9

TXFURM

NAKM

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

INEPNMM

This register works with each of the Device IN endpoint interrupt (OTG_HS_DIEPINTx)
registers for all endpoints to generate an interrupt per IN endpoint. The IN endpoint interrupt
for a specific status in the OTG_HS_DIEPINTx register can be masked by writing to the
corresponding bit in this register. Status bits are masked by default.

rw

rw

rw

rw

rw

rw

rw

Bits 31:10 Reserved, must be kept at reset value.
Bit 13 NAKM: NAK interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bits 12:9 Reserved, must be kept at reset value.
Bit 8 TXFURM: FIFO underrun mask
0: Masked interrupt
1: Unmasked interrupt
Bit 7 Reserved, must be kept at reset value.
Bit 6 INEPNEM: IN endpoint NAK effective mask
0: Masked interrupt
1: Unmasked interrupt
Bit 5 INEPNMM: IN token received with EP mismatch mask
0: Masked interrupt
1: Unmasked interrupt
Bit 4 ITTXFEMSK: IN token received when TxFIFO empty mask
0: Masked interrupt
1: Unmasked interrupt
Bit 3 TOM: Timeout condition mask (nonisochronous endpoints)
0: Masked interrupt
1: Unmasked interrupt

RM0090 Rev 17

1447/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bit 2 AHBERRM: AHB error mask
0: Masked interrupt
1: Unmasked interrupt
Bit 1 EPDM: Endpoint disabled interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 0 XFRCM: Transfer completed interrupt mask
0: Masked interrupt
1: Unmasked interrupt

OTG_HS device OUT endpoint common interrupt mask register
(OTG_HS_DOEPMSK)
Address offset: 0x814
Reset value: 0x0000 0000

Bits 31:15 Reserved, must be kept at reset value.
Bit 14 NYETMSK: NYET interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 13 NAKMSK: NAK interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 12 BERRM: Babble error interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bits 11:9 Reserved, must be kept at reset value.
Bit 8 OPEM: OUT packet error mask
0: Masked interrupt
1: Unmasked interrupt
Bit 7 Reserved, must be kept at reset value.

1448/1747

RM0090 Rev 17

4

3

2

1

0

STUPM

AHBERRM

EPDM

XFRCM

rw

5

OTEPDM

rw

Reserved

6
B2BSTUP

BERRM

rw

7

STSPHSRXM

NAKMSK

rw

8

Reserved

NYETMSK

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

Reserved

9

OPEM

This register works with each of the Device OUT endpoint interrupt (OTG_HS_DOEPINTx)
registers for all endpoints to generate an interrupt per OUT endpoint. The OUT endpoint
interrupt for a specific status in the OTG_HS_DOEPINTx register can be masked by writing
into the corresponding bit in this register. Status bits are masked by default.

rw

rw

rw

rw

rw

rw

rw

RM0090

USB on-the-go high-speed (OTG_HS)

Bit 6 B2BSTUP: Back-to-back SETUP packets received mask
Applies to control OUT endpoints only.
0: Masked interrupt
1: Unmasked interrupt
Bit 5 STSPHSRXM: Status phase received for control write mask
0: Masked interrupt
1: Unmasked interrupt
Bit 4 OTEPDM: OUT token received when endpoint disabled mask
Applies to control OUT endpoints only.
0: Masked interrupt
1: Unmasked interrupt
Bit 3 STUPM: SETUP phase done mask
Applies to control endpoints only.
0: Masked interrupt
1: Unmasked interrupt
Bit 2 AHBERRM: AHB error mask
0: Masked interrupt
1: Unmasked interrupt
Bit 1 EPDM: Endpoint disabled interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 0 XFRCM: Transfer completed interrupt mask
0: Masked interrupt
1: Unmasked interrupt

OTG_HS device all endpoints interrupt register (OTG_HS_DAINT)
Address offset: 0x818
Reset value: 0x0000 0000
When a significant event occurs on an endpoint, a device all endpoints interrupt register
interrupts the application using the Device OUT endpoints interrupt bit or Device IN
endpoints interrupt bit of the Core interrupt register (OEPINT or IEPINT in
OTG_HS_GINTSTS, respectively). There is one interrupt bit per endpoint, up to a maximum
of 16 bits for OUT endpoints and 16 bits for IN endpoints. For a bidirectional endpoint, the
corresponding IN and OUT interrupt bits are used. Bits in this register are set and cleared
when the application sets and clears bits in the corresponding Device Endpoint-x interrupt
register (OTG_HS_DIEPINTx/OTG_HS_DOEPINTx).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

OEPINT
r

r

r

r

r

r

r

r

r

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

IEPINT
r

r

r

r

r

r

r

r

r

RM0090 Rev 17

r

r

r

r

r

r

r

1449/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bits 31:16 OEPINT: OUT endpoint interrupt bits
One bit per OUT endpoint:
Bit 16 for OUT endpoint 0, bit 31 for OUT endpoint 15
Bits 15:0 IEPINT: IN endpoint interrupt bits
One bit per IN endpoint:
Bit 0 for IN endpoint 0, bit 15 for endpoint 15

OTG_HS all endpoints interrupt mask register (OTG_HS_DAINTMSK)
Address offset: 0x81C
Reset value: 0x0000 0000
The device endpoint interrupt mask register works with the device endpoint interrupt register
to interrupt the application when an event occurs on a device endpoint. However, the device
all endpoints interrupt (OTG_HS_DAINT) register bit corresponding to that interrupt is still
set.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

OEPM
rw

rw

rw

rw

rw

rw

rw

rw

rw

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

IEPM
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:16 OEPM: OUT EP interrupt mask bits
One per OUT endpoint:
Bit 16 for OUT EP 0, bit 18 for OUT EP 3
0: Masked interrupt
1: Unmasked interrupt
Bits 15:0 IEPM: IN EP interrupt mask bits
One bit per IN endpoint:
Bit 0 for IN EP 0, bit 3 for IN EP 3
0: Masked interrupt
1: Unmasked interrupt

OTG_HS device VBUS discharge time register (OTG_HS_DVBUSDIS)
Address offset: 0x0828
Reset value: 0x0000 17D7
This register specifies the VBUS discharge time after VBUS pulsing during SRP.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

VBUSDT
rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 VBUSDT: Device VBUS discharge time
Specifies the VBUS discharge time after VBUS pulsing during SRP. This value equals:
VBUS discharge time in PHY clocks / 1 024
Depending on your VBUS load, this value may need adjusting.

1450/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

OTG_HS device VBUS pulsing time register (OTG_HS_DVBUSPULSE)
Address offset: 0x082C
Reset value: 0x0000 05B8
This register specifies the VBUS pulsing time during SRP.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

DVBUSP

Reserved

rw

rw

rw

rw

rw

rw

rw

Bits 31:12 Reserved, must be kept at reset value.
Bits 11:0 DVBUSP: Device VBUS pulsing time
Specifies the VBUS pulsing time during SRP. This value equals:
VBUS pulsing time in PHY clocks / 1 024

RM0090 Rev 17

1451/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

OTG_HS Device threshold control register (OTG_HS_DTHRCTL)
Address offset: 0x0830

RXTHRLEN

rw

rw

rw

rw

rw

rw

rw

rw

rw

9

8

6

5

4

3

2

rw

rw

rw

TXTHRLEN

Reserved

rw

7

rw

rw

rw

rw

rw

rw

1

0

ISOTHREN

rw

RXTHREN

Reserved

Reserved

ARPEN

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

NONISOTHREN

Reset value: 0x0000 0000

rw

rw

Bits 31:28 Reserved, must be kept at reset value.
Bit 27 ARPEN: Arbiter parking enable
This bit controls internal DMA arbiter parking for IN endpoints. When thresholding is enabled
and this bit is set to one, then the arbiter parks on the IN endpoint for which there is a token
received on the USB. This is done to avoid getting into underrun conditions. By default
parking is enabled.
Bit 26

Reserved, must be kept at reset value.

Bits 25: 17 RXTHRLEN: Receive threshold length
This field specifies the receive thresholding size in words. This field also specifies the
amount of data received on the USB before the core can start transmitting on the AHB. The
threshold length has to be at least eight words. The recommended value for RXTHRLEN is
to be the same as the programmed AHB burst length (HBSTLEN bit in
OTG_HS_GAHBCFG).
Bit 16 RXTHREN: Receive threshold enable
When this bit is set, the core enables thresholding in the receive direction.
Bits 15: 11

Reserved, must be kept at reset value.

Bits 10:2 TXTHRLEN: Transmit threshold length
This field specifies the transmit thresholding size in words. This field specifies the amount of
data in bytes to be in the corresponding endpoint transmit FIFO, before the core can start
transmitting on the USB. The threshold length has to be at least eight words. This field
controls both isochronous and nonisochronous IN endpoint thresholds. The recommended
value for TXTHRLEN is to be the same as the programmed AHB burst length (HBSTLEN bit
in OTG_HS_GAHBCFG).
Bit 1 ISOTHREN: ISO IN endpoint threshold enable
When this bit is set, the core enables thresholding for isochronous IN endpoints.
Bit 0 NONISOTHREN: Nonisochronous IN endpoints threshold enable
When this bit is set, the core enables thresholding for nonisochronous IN endpoints.

1452/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

OTG_HS device IN endpoint FIFO empty interrupt mask register:
(OTG_HS_DIEPEMPMSK)
Address offset: 0x834
Reset value: 0x0000 0000
This register is used to control the IN endpoint FIFO empty interrupt generation
(TXFE_OTG_HS_DIEPINTx).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

1

0

INEPTXFEM

Reserved

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 INEPTXFEM: IN EP Tx FIFO empty interrupt mask bits
These bits act as mask bits for OTG_HS_DIEPINTx.
TXFE interrupt one bit per IN endpoint:
Bit 0 for IN endpoint 0, bit 15 for IN endpoint 15
0: Masked interrupt
1: Unmasked interrupt

OTG_HS device each endpoint interrupt register (OTG_HS_DEACHINT)
Address offset: 0x0838
Reset value: 0x0000 0000

r

9

8

Reserved

7

6

5

4

3

2

IEP1INT

Reserved

OEP1INT

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

r

Reserved

There is one interrupt bit for endpoint 1 IN and one interrupt bit for endpoint 1 OUT.

Bits 31:18 Reserved, must be kept at reset value.
Bit 17 OEP1INT: OUT endpoint 1 interrupt bit
Bits 16:2 Reserved, must be kept at reset value.
Bit 1 IEP1INT: IN endpoint 1interrupt bit
Bit 0

Reserved, must be kept at reset value.

RM0090 Rev 17

1453/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

OTG_HS device each endpoint interrupt register mask
(OTG_HS_DEACHINTMSK)
Address offset: 0x083C
Reset value: 0x0000 0000
There is one interrupt bit for endpoint 1 IN and one interrupt bit for endpoint 1 OUT.
8

7

6

5

4

3

2

1
IEP1INTM

9

Reserved

rw

0
Reserved

Reserved

OEP1INTM

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

rw

Bits 31:18 Reserved, must be kept at reset value.
Bit 17 OEP1INTM: OUT Endpoint 1 interrupt mask bit
Bits 16:2

Reserved, must be kept at reset value.

Bit 1 IEP1INTM: IN Endpoint 1 interrupt mask bit
Bit 0

Reserved, must be kept at reset value.

OTG_HS device each in endpoint-1 interrupt register
(OTG_HS_DIEPEACHMSK1)
Address offset: 0x844

Bits 31:14 Reserved, must be kept at reset value.
Bit 13 NAKM: NAK interrupt mask
0: Masked interrupt
1: unmasked interrupt
Bit 12:10 Reserved, must be kept at reset value.
Bit 9 BIM: BNA interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 8 TXFURM: FIFO underrun mask
0: Masked interrupt
1: Unmasked interrupt
Bit 7 Reserved, must be kept at reset value.

1454/1747

RM0090 Rev 17

5

4

3

2

1

0

ITTXFEMSK

TOM

AHBERRM

EPDM

XFRCM

rw

6
INEPNEM

rw

7
Reserved

8

INEPNMM

rw

9

BIM

Reserved

Reserved

NAKM

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

TXFURM

Reset value: 0x0000 0000

rw

rw

rw

rw

rw

rw

rw

RM0090

USB on-the-go high-speed (OTG_HS)

Bit 6 INEPNEM: IN endpoint NAK effective mask
0: Masked interrupt
1: Unmasked interrupt
Bit 5 INEPNMM: IN token received with EP mismatch mask
0: Masked interrupt
1: Unmasked interrupt
Bit 4 ITTXFEMSK: IN token received when TxFIFO empty mask
0: Masked interrupt
1: Unmasked interrupt
Bit 3 TOM: Timeout condition mask (nonisochronous endpoints)
0: Masked interrupt
1: Unmasked interrupt
Bit 2 AHBERRM: AHB error mask
0: Masked interrupt
1: Unmasked interrupt
Bit 1 EPDM: Endpoint disabled interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 0 XFRCM: Transfer completed interrupt mask
0: Masked interrupt
1: Unmasked interrupt

OTG_HS device each OUT endpoint-1 interrupt register
(OTG_HS_DOEPEACHMSK1)
Address offset: 0x884

5

4

3

2

1

0

ITTXFEMSK

TOM

AHBERRM

EPDM

XFRCM

rw

6
INEPNEM

rw

7

INEPNMM

rw

8

Reserved

rw

9

BIM

BERRM

rw

Reserved

NAKM

Reserved

NYETM

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

TXFURM

Reset value: 0x0000 0000

rw

rw

rw

rw

rw

rw

rw

Bits 31:15 Reserved, must be kept at reset value.
Bit 14 NYETM: NYET interrupt mask
0: Masked interrupt
1: unmasked interrupt
Bit 13 NAKM: NAK interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 12 BERRM: Bubble error interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 11:10 Reserved, must be kept at reset value.

RM0090 Rev 17

1455/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bit 9 BIM: BNA interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 8 OPEM: OUT packet error mask
0: Masked interrupt
1: Unmasked interrupt
Bits 7:3 Reserved, must be kept at reset value.
Bit 2 AHBERRM: AHB error mask
0: Masked interrupt
1: Unmasked interrupt
Bit 1 EPDM: Endpoint disabled interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 0 XFRCM: Transfer completed interrupt mask
0: Masked interrupt
1: Unmasked interrupt

OTG device endpoint-x control register (OTG_HS_DIEPCTLx) (x = 0..7, where
x = Endpoint_number)
Address offset: 0x900 + 0x20 * x
Reset value: 0x0000 0000
The application uses this register to control the behavior of each logical endpoint other than
endpoint 0.

1456/1747

rw

rw

rw

rw/
rs

rw

rw

USBAEP

rw

EONUM/DPID

w

NAKSTS

w

EPTYP

w

Stall

CNAK

w

TXFNUM

Reserved

SNAK

rs

SODDFRM

EPDIS

rs

SD0PID/SEVNFRM

EPENA

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

r

r

rw

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

MPSIZ
Reserved

RM0090 Rev 17

rw

rw

rw

rw

rw

rw

RM0090

USB on-the-go high-speed (OTG_HS)

Bit 31 EPENA: Endpoint enable
The application sets this bit to start transmitting data on an endpoint.
The core clears this bit before setting any of the following interrupts on this endpoint:
– SETUP phase done
– Endpoint disabled
– Transfer completed
Bit 30 EPDIS: Endpoint disable
The application sets this bit to stop transmitting/receiving data on an endpoint, even before
the transfer for that endpoint is complete. The application must wait for the Endpoint
disabled interrupt before treating the endpoint as disabled. The core clears this bit before
setting the Endpoint disabled interrupt. The application must set this bit only if Endpoint
enable is already set for this endpoint.
Bit 29 SODDFRM: Set odd frame
Applies to isochronous IN and OUT endpoints only.
Writing to this field sets the Even/Odd frame (EONUM) field to odd frame.
Bit 28 SD0PID: Set DATA0 PID
Applies to interrupt/bulk IN endpoints only.
Writing to this field sets the endpoint data PID (DPID) field in this register to DATA0.
SEVNFRM: Set even frame
Applies to isochronous IN endpoints only.
Writing to this field sets the Even/Odd frame (EONUM) field to even frame.
Bit 27

SNAK: Set NAK
A write to this bit sets the NAK bit for the endpoint.
Using this bit, the application can control the transmission of NAK handshakes on an
endpoint. The core can also set this bit for OUT endpoints on a Transfer completed interrupt,
or after a SETUP is received on the endpoint.

Bit 26 CNAK: Clear NAK
A write to this bit clears the NAK bit for the endpoint.
Bits 25:22 TXFNUM: TxFIFO number
These bits specify the FIFO number associated with this endpoint. Each active IN endpoint
must be programmed to a separate FIFO number.
This field is valid only for IN endpoints.
Bit 21 STALL: STALL handshake
Applies to noncontrol, nonisochronous IN endpoints only (access type is rw).
The application sets this bit to stall all tokens from the USB host to this endpoint. If a NAK
bit, Global IN NAK, or Global OUT NAK is set along with this bit, the STALL bit takes priority.
Only the application can clear this bit, never the core.
Applies to control endpoints only (access type is rs).
The application can only set this bit, and the core clears it, when a SETUP token is received
for this endpoint. If a NAK bit, Global IN NAK, or Global OUT NAK is set along with this bit,
the STALL bit takes priority. Irrespective of this bit’s setting, the core always responds to
SETUP data packets with an ACK handshake.
Bit 20 Reserved, must be kept at reset value.

RM0090 Rev 17

1457/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bits 19:18 EPTYP: Endpoint type
This is the transfer type supported by this logical endpoint.
00: Control
01: Isochronous
10: Bulk
11: Interrupt
Bit 17 NAKSTS: NAK status
It indicates the following:
0: The core is transmitting nonNAK handshakes based on the FIFO status.
1: The core is transmitting NAK handshakes on this endpoint.
When either the application or the core sets this bit:
For nonisochronous IN endpoints: The core stops transmitting any data on an IN endpoint,
even if there are data available in the TxFIFO.
For isochronous IN endpoints: The core sends out a zero-length data packet, even if there
are data available in the TxFIFO.
Irrespective of this bit’s setting, the core always responds to SETUP data packets with an ACK
handshake.
Bit 16 EONUM: Even/odd frame
Applies to isochronous IN endpoints only.
Indicates the frame number in which the core transmits/receives isochronous data for this
endpoint. The application must program the even/odd frame number in which it intends to
transmit/receive isochronous data for this endpoint using the SEVNFRM and SODDFRM
fields in this register.
0: Even frame
1: Odd frame
DPID: Endpoint data PID
Applies to interrupt/bulk IN endpoints only.
Contains the PID of the packet to be received or transmitted on this endpoint. The
application must program the PID of the first packet to be received or transmitted on this
endpoint, after the endpoint is activated. The application uses the SD0PID register field to
program either DATA0 or DATA1 PID.
0: DATA0
1: DATA1
Bit 15 USBAEP: USB active endpoint
Indicates whether this endpoint is active in the current configuration and interface. The core
clears this bit for all endpoints (other than EP 0) after detecting a USB reset. After receiving
the SetConfiguration and SetInterface commands, the application must program endpoint
registers accordingly and set this bit.
Bits 14:11 Reserved, must be kept at reset value.
Bits 10:0 MPSIZ: Maximum packet size
The application must program this field with the maximum packet size for the current logical
endpoint. This value is in bytes.

1458/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

OTG_HS device control OUT endpoint 0 control register
(OTG_HS_DOEPCTL0)
Address offset: 0xB00
Reset value: 0x0000 8000
This section describes the device control OUT endpoint 0 control register. Nonzero control
endpoints use registers for endpoints 1–15.

r

r

r

USBAEP

rw

Reserved

rs

EPTYP

NAKSTS

w

Stall

w

Reserved

SNPM

CNAK

r

SNAK

EPDIS

w

Reserved

EPENA

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

MPSIZ

Reserved

r

r

r

Bit 31 EPENA: Endpoint enable
The application sets this bit to start transmitting data on endpoint 0.
The core clears this bit before setting any of the following interrupts on this endpoint:
– SETUP phase done
– Endpoint disabled
– Transfer completed
Bit 30 EPDIS: Endpoint disable
The application cannot disable control OUT endpoint 0.
Bits 29:28 Reserved, must be kept at reset value.
Bit 27 SNAK: Set NAK
A write to this bit sets the NAK bit for the endpoint.
Using this bit, the application can control the transmission of NAK handshakes on an
endpoint. The core can also set this bit on a Transfer completed interrupt, or after a SETUP
is received on the endpoint.
Bit 26 CNAK: Clear NAK
A write to this bit clears the NAK bit for the endpoint.
Bits 25:22 Reserved, must be kept at reset value.
Bit 21 STALL: STALL handshake
The application can only set this bit, and the core clears it, when a SETUP token is received
for this endpoint. If a NAK bit or Global OUT NAK is set along with this bit, the STALL bit
takes priority. Irrespective of this bit’s setting, the core always responds to SETUP data
packets with an ACK handshake.
Bit 20 SNPM: Snoop mode
This bit configures the endpoint to Snoop mode. In Snoop mode, the core does not check
the correctness of OUT packets before transferring them to application memory.
Bits 19:18 EPTYP: Endpoint type
Hardcoded to 2’b00 for control.

RM0090 Rev 17

1459/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bit 17 NAKSTS: NAK status
Indicates the following:
0: The core is transmitting nonNAK handshakes based on the FIFO status.
1: The core is transmitting NAK handshakes on this endpoint.
When either the application or the core sets this bit, the core stops receiving data, even if
there is space in the RxFIFO to accommodate the incoming packet. Irrespective of this bit’s
setting, the core always responds to SETUP data packets with an ACK handshake.
Bit 16 Reserved, must be kept at reset value.
Bit 15 USBAEP: USB active endpoint
This bit is always set to 1, indicating that a control endpoint 0 is always active in all
configurations and interfaces.
Bits 14:2 Reserved, must be kept at reset value.
Bits 1:0 MPSIZ: Maximum packet size
The maximum packet size for control OUT endpoint 0 is the same as what is programmed in
control IN endpoint 0.
00: 64 bytes
01: 32 bytes
10: 16 bytes
11: 8 bytes

OTG_HS device endpoint-x control register (OTG_HS_DOEPCTLx) (x = 1..3,
where x = Endpoint_number)
Address offset for OUT endpoints: 0xB00 + 0x20 * x
Reset value: 0x0000 0000
The application uses this register to control the behavior of each logical endpoint other than
endpoint 0.

1460/1747

rw

rw

rw

USBAEP

rw

EONUM/DPID

w

NAKSTS

w

EPTYP

w

Stall

CNAK

w

Reserved

SNPM

SNAK

rs

SODDFRM

EPDIS

rs

SD0PID/SEVNFRM

EPENA

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

r

r

rw

9

8

7

6

4

3

2

1

0

rw

rw

rw

rw

rw

MPSIZ

Reserved

RM0090 Rev 17

5

rw

rw

rw

rw

rw

rw

RM0090

USB on-the-go high-speed (OTG_HS)

Bit 31 EPENA: Endpoint enable
Applies to IN and OUT endpoints.
The application sets this bit to start transmitting data on an endpoint.
The core clears this bit before setting any of the following interrupts on this endpoint:
– SETUP phase done
– Endpoint disabled
– Transfer completed
Bit 30 EPDIS: Endpoint disable
The application sets this bit to stop transmitting/receiving data on an endpoint, even before
the transfer for that endpoint is complete. The application must wait for the Endpoint
disabled interrupt before treating the endpoint as disabled. The core clears this bit before
setting the Endpoint disabled interrupt. The application must set this bit only if Endpoint
enable is already set for this endpoint.
Bit 29 SODDFRM: Set odd frame
Applies to isochronous OUT endpoints only.
Writing to this field sets the Even/Odd frame (EONUM) field to odd frame.
Bit 28 SD0PID: Set DATA0 PID
Applies to interrupt/bulk OUT endpoints only.
Writing to this field sets the endpoint data PID (DPID) field in this register to DATA0.
SEVNFRM: Set even frame
Applies to isochronous OUT endpoints only.
Writing to this field sets the Even/Odd frame (EONUM) field to even frame.
Bit 27 SNAK: Set NAK
A write to this bit sets the NAK bit for the endpoint.
Using this bit, the application can control the transmission of NAK handshakes on an
endpoint. The core can also set this bit for OUT endpoints on a Transfer Completed
interrupt, or after a SETUP is received on the endpoint.
Bit 26 CNAK: Clear NAK
A write to this bit clears the NAK bit for the endpoint.
Bits 25:22 Reserved, must be kept at reset value.
Bit 21 STALL: STALL handshake
Applies to noncontrol, nonisochronous OUT endpoints only (access type is rw).
The application sets this bit to stall all tokens from the USB host to this endpoint. If a NAK
bit, Global IN NAK, or Global OUT NAK is set along with this bit, the STALL bit takes
priority. Only the application can clear this bit, never the core.
Bit 20 SNPM: Snoop mode
This bit configures the endpoint to Snoop mode. In Snoop mode, the core does not check
the correctness of OUT packets before transferring them to application memory.
Bits 19:18 EPTYP: Endpoint type
This is the transfer type supported by this logical endpoint.
00: Control
01: Isochronous
10: Bulk
11: Interrupt

RM0090 Rev 17

1461/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bit 17 NAKSTS: NAK status
Indicates the following:
0: The core is transmitting nonNAK handshakes based on the FIFO status.
1: The core is transmitting NAK handshakes on this endpoint.
When either the application or the core sets this bit:
The core stops receiving any data on an OUT endpoint, even if there is space in the
RxFIFO to accommodate the incoming packet.
Irrespective of this bit’s setting, the core always responds to SETUP data packets with an
ACK handshake.
Bit 16 EONUM: Even/odd frame
Applies to isochronous IN and OUT endpoints only.
Indicates the frame number in which the core transmits/receives isochronous data for this
endpoint. The application must program the even/odd frame number in which it intends to
transmit/receive isochronous data for this endpoint using the SEVNFRM and SODDFRM
fields in this register.
0: Even frame
1: Odd frame
DPID: Endpoint data PID
Applies to interrupt/bulk OUT endpoints only.
Contains the PID of the packet to be received or transmitted on this endpoint. The
application must program the PID of the first packet to be received or transmitted on this
endpoint, after the endpoint is activated. The application uses the SD0PID register field to
program either DATA0 or DATA1 PID.
0: DATA0
1: DATA1
Bit 15 USBAEP: USB active endpoint
Indicates whether this endpoint is active in the current configuration and interface. The core
clears this bit for all endpoints (other than EP 0) after detecting a USB reset. After receiving
the SetConfiguration and SetInterface commands, the application must program endpoint
registers accordingly and set this bit.
Bits 14:11 Reserved, must be kept at reset value.
Bits 10:0 MPSIZ: Maximum packet size
The application must program this field with the maximum packet size for the current logical
endpoint. This value is in bytes.

1462/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

OTG_HS device endpoint-x interrupt register (OTG_HS_DIEPINTx) (x = 0..7,
where x = Endpoint_number)
Address offset: 0x908 + 0x20 * x
Reset value: 0x0000 0080

rc_
w1

2

1

0

XFRC

3

EPDISD

4

AHBERR

r

5

TOC

rc_
w1

6

ITTXFE

7

INEPNE

8

INEPNM

Reserved

9

TXFE

rc_
w1

PKTDRPSTS

Reserved

Reserved

NAK

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

TXFIFOUDRN

This register indicates the status of an endpoint with respect to USB- and AHB-related
events. It is shown in Figure 414. The application must read this register when the IN
endpoints interrupt bit of the Core interrupt register (IEPINT in OTG_HS_GINTSTS) is set.
Before the application can read this register, it must first read the device all endpoints
interrupt (OTG_HS_DAINT) register to get the exact endpoint number for the device
endpoint-x interrupt register. The application must clear the appropriate bit in this register to
clear the corresponding bits in the OTG_HS_DAINT and OTG_HS_GINTSTS registers.

rc_ rc_ rc_ rc_ rc_ rc_ rc_
w1 w1 w1 w1 w1 w1 w1

Bits 31:14 Reserved, must be kept at reset value.
Bit 13 NAK: NAK interrupt
The core generates this interrupt when a NAK is transmitted or received by the device. In
case of isochronous IN endpoints the interrupt gets generated when a zero length packet is
transmitted due to unavailability of data in the Tx FIFO.
Bit 12 Reserved, must be kept at reset value.
Bit 11 PKTDRPSTS: Packet dropped status
This bit indicates to the application that an ISOC OUT packet has been dropped. This bit
does not have an associated mask bit and does not generate an interrupt.
Bits 10:9

Reserved, must be kept at reset value.

Bit 8 TXFIFOUDRN: Transmit Fifo Underrun (TxfifoUndrn) The core generates this interrupt when it
detects a transmit FIFO underrun condition for this endpoint.
Dependency: This interrupt is valid only when Thresholding is enabled
Bit 7 TXFE: Transmit FIFO empty
This interrupt is asserted when the TxFIFO for this endpoint is either half or completely
empty. The half or completely empty status is determined by the TxFIFO empty level bit in
the Core AHB configuration register (TXFELVL bit in OTG_HS_GAHBCFG).
Bit 6 INEPNE: IN endpoint NAK effective
This bit can be cleared when the application clears the IN endpoint NAK by writing to the
CNAK bit in OTG_HS_DIEPCTLx.
This interrupt indicates that the core has sampled the NAK bit set (either by the application
or by the core). The interrupt indicates that the IN endpoint NAK bit set by the application
has taken effect in the core.
This interrupt does not guarantee that a NAK handshake is sent on the USB. A STALL bit
takes priority over a NAK bit.

RM0090 Rev 17

1463/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bit 5 INEPNM: IN token received with EP mismatch
Indicates that the data in the top of the non-periodic TxFIFO belongs to an endpoint other
than the one for which the IN token was received. This interrupt is asserted on the endpoint
for which the IN token was received.
Bit 4 ITTXFE: IN token received when TxFIFO is empty
Applies to nonperiodic IN endpoints only.
Indicates that an IN token was received when the associated TxFIFO (periodic/nonperiodic)
was empty. This interrupt is asserted on the endpoint for which the IN token was received.
Bit 3 TOC: Timeout condition
Applies only to Control IN endpoints.
Indicates that the core has detected a timeout condition on the USB for the last IN token on
this endpoint.
Bit 2 AHBERR: AHB error
This is generated only in internal DMA mode when there is an AHB error during an AHB
read/write. The application can read the corresponding endpoint DMA address register to
get the error address.
Bit 1 EPDISD: Endpoint disabled interrupt
This bit indicates that the endpoint is disabled per the application’s request.
Bit 0 XFRC: Transfer completed interrupt
This field indicates that the programmed transfer is complete on the AHB as well as on the
USB, for this endpoint.

1464/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

OTG_HS device endpoint-x interrupt register (OTG_HS_DOEPINTx) (x = 0..7,
where x = Endpoint_number)
Address offset: 0xB08 + 0x20 * x
Reset value: 0x0000 0080

rc_ rc_ rc_
w1 w1 w1

rc_
w1

rc_
w1

2

1

0
XFRC

3

EPDISD

5

AHBERR

4

STUP

6
B2BSTUP

7

Reserved

Reserved

8

Reserved

9

OUTPKTERR

BERR

Reserved

NAK

NYET

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

OTEPDIS

This register indicates the status of an endpoint with respect to USB- and AHB-related
events. It is shown in Figure 414. The application must read this register when the OUT
Endpoints Interrupt bit of the Core interrupt register (OEPINT bit in OTG_HS_GINTSTS) is
set. Before the application can read this register, it must first read the device all endpoints
interrupt (OTG_HS_DAINT) register to get the exact endpoint number for the device
Endpoint-x interrupt register. The application must clear the appropriate bit in this register to
clear the corresponding bits in the OTG_HS_DAINT and OTG_HS_GINTSTS registers.

rc_ rc_ rc_ rc_ rc_
w1 w1 w1 w1 w1

Bits 31:15 Reserved, must be kept at reset value.
Bit 14 NYET: NYET interrupt
The core generates this interrupt when a NYET response is transmitted for a
nonisochronous OUT endpoint.
Bit 13 NAK: NAK input
The core generates this interrupt when a NAK is transmitted or received by the device. In
case of isochronous IN endpoints the interrupt gets generated when a zero length packet is
transmitted due to unavailability of data in the Tx FIFO.
Bit 12 BERR: Babble error interrupt
The core generates this interrupt when babble is received for the endpoint.
Bits 11:9 Reserved, must be kept at reset value.
Bit 8 OUTPKTERR: OUT packet error
This interrupt is asserted when the core detects an overflow or a CRC error for an OUT
packet. This interrupt is valid only when thresholding is enabled.
Bit 7 Reserved, must be kept at reset value.
Bit 6 B2BSTUP: Back-to-back SETUP packets received
Applies to Control OUT endpoint only.
This bit indicates that the core has received more than three back-to-back SETUP packets
for this particular endpoint.
Bit 5

Reserved, must be kept at reset value.

Bit 4 OTEPDIS: OUT token received when endpoint disabled
Applies only to control OUT endpoint.
Indicates that an OUT token was received when the endpoint was not yet enabled. This
interrupt is asserted on the endpoint for which the OUT token was received.

RM0090 Rev 17

1465/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bit 3 STUP: SETUP phase done
Applies to control OUT endpoints only.
Indicates that the SETUP phase for the control endpoint is complete and no more back-toback SETUP packets were received for the current control transfer. On this interrupt, the
application can decode the received SETUP data packet.
Bit 2 AHBERR: AHB error
This is generated only in internal DMA mode when there is an AHB error during an AHB
read/write. The application can read the corresponding endpoint DMA address register to
get the error address.
Bit 1 EPDISD: Endpoint disabled interrupt
This bit indicates that the endpoint is disabled per the application’s request.
Bit 0 XFRC: Transfer completed interrupt
This field indicates that the programmed transfer is complete on the AHB as well as on the
USB, for this endpoint.

OTG_HS device IN endpoint 0 transfer size register (OTG_HS_DIEPTSIZ0)
Address offset: 0x910
Reset value: 0x0000 0000
The application must modify this register before enabling endpoint 0. Once endpoint 0 is
enabled using the endpoint enable bit in the device control endpoint 0 control registers
(EPENA in OTG_HS_DIEPCTL0), the core modifies this register. The application can only
read this register once the core has cleared the Endpoint enable bit.
Nonzero endpoints use the registers for endpoints 1–15.
31 30 29 28 27 26 25 24 23 22 21
Reserved

20

19

18 17 16 15 14 13 12 11 10

PKTCNT
rw

Reserved

rw

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

XFRSIZ
rw

rw

rw

rw

Bits 31:21 Reserved, must be kept at reset value.
Bits 20:19 PKTCNT: Packet count
Indicates the total number of USB packets that constitute the Transfer Size amount of data
for endpoint 0.
This field is decremented every time a packet (maximum size or short packet) is read from
the TxFIFO.
Bits 18:7 Reserved, must be kept at reset value.
Bits 6:0 XFRSIZ: Transfer size
Indicates the transfer size in bytes for endpoint 0. The core interrupts the application only
after it has exhausted the transfer size amount of data. The transfer size can be set to the
maximum packet size of the endpoint, to be interrupted at the end of each packet.
The core decrements this field every time a packet from the external memory is written to
the TxFIFO.

1466/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

OTG_HS device OUT endpoint 0 transfer size register (OTG_HS_DOEPTSIZ0)
Address offset: 0xB10
Reset value: 0x0000 0000
The application must modify this register before enabling endpoint 0. Once endpoint 0 is
enabled using the Endpoint enable bit in the device control endpoint 0 control registers
(EPENA bit in OTG_HS_DOEPCTL0), the core modifies this register. The application can
only read this register once the core has cleared the Endpoint enable bit.
Nonzero endpoints use the registers for endpoints 1–15.

STUPC
NT
rw

rw

Reserved

PKTCNT

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

3

2

1

0

rw

rw

XFRSIZ

Reserved

rw

4

rw

rw

rw

rw

rw

Bit 31 Reserved, must be kept at reset value.
Bits 30:29 STUPCNT: SETUP packet count
This field specifies the number of back-to-back SETUP data packets the endpoint can
receive.
01: 1 packet
10: 2 packets
11: 3 packets
Bits 28:20 Reserved, must be kept at reset value.
Bit 19 PKTCNT: Packet count
This field is decremented to zero after a packet is written into the RxFIFO.
Bits 18:7 Reserved, must be kept at reset value.
Bits 6:0 XFRSIZ: Transfer size
Indicates the transfer size in bytes for endpoint 0. The core interrupts the application only
after it has exhausted the transfer size amount of data. The transfer size can be set to the
maximum packet size of the endpoint, to be interrupted at the end of each packet.
The core decrements this field every time a packet is read from the RxFIFO and written to
the external memory.

RM0090 Rev 17

1467/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

OTG_HS device endpoint-x transfer size register (OTG_HS_DIEPTSIZx)
(x = 1..3, where x = Endpoint_number)
Address offset: 0x910 + 0x20 * x
Reset value: 0x0000 0000
The application must modify this register before enabling the endpoint. Once the endpoint is
enabled using the Endpoint enable bit in the device endpoint-x control registers (EPENA bit
in OTG_HS_DIEPCTLx), the core modifies this register. The application can only read this
register once the core has cleared the Endpoint enable bit.

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
MCNT
rw

PKTCNT

rw

rw

rw

rw

rw

rw

rw

9

8

7

6

5

4

3

2

1

0

rw

rw

rw

rw

rw

rw

rw

rw

rw

XFRSIZ
rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

Bit 31 Reserved, must be kept at reset value.
Bits 30:29 MCNT: Multi count
For periodic IN endpoints, this field indicates the number of packets that must be transmitted
per frame on the USB. The core uses this field to calculate the data PID for isochronous IN
endpoints.
01: 1 packet
10: 2 packets
11: 3 packets
Bit 28:19 PKTCNT: Packet count
Indicates the total number of USB packets that constitute the Transfer Size amount of data
for this endpoint.
This field is decremented every time a packet (maximum size or short packet) is read from
the TxFIFO.
Bits 18:0 XFRSIZ: Transfer size
This field contains the transfer size in bytes for the current endpoint. The core only interrupts
the application after it has exhausted the transfer size amount of data. The transfer size can
be set to the maximum packet size of the endpoint, to be interrupted at the end of each
packet.
The core decrements this field every time a packet from the external memory is written to the
TxFIFO.

1468/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

OTG_HS device IN endpoint transmit FIFO status register
(OTG_HS_DTXFSTSx) (x = 0..5, where x = Endpoint_number)
Address offset for IN endpoints: 0x918 + 0x20 + x
This read-only register contains the free space information for the Device IN endpoint
TxFIFO.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

INEPTFSAV

Reserved

r

r

r

r

r

r

r

r

r

31:16 Reserved, must be kept at reset value.
15:0 INEPTFSAV: IN endpoint TxFIFO space avail ()
Indicates the amount of free space available in the Endpoint TxFIFO.
Values are in terms of 32-bit words:
0x0: Endpoint TxFIFO is full
0x1: 1 word available
0x2: 2 words available
0xn: n words available (0 < n < 512)
Others: Reserved

OTG_HS device endpoint-x transfer size register (OTG_HS_DOEPTSIZx)
(x = 1..5, where x = Endpoint_number)
Address offset: 0xB10 + 0x20 * x
Reset value: 0x0000 0000
The application must modify this register before enabling the endpoint. Once the endpoint is
enabled using Endpoint Enable bit of the device endpoint-x control registers (EPENA bit in
OTG_HS_DOEPCTLx), the core modifies this register. The application can only read this
register once the core has cleared the Endpoint enable bit.

Reserved

31

30

29

RXDPID/S
TUPCNT
r/rw

28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
PKTCNT

9

8

7

6

5

4

3

2

1

0

XFRSIZ

r/rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

RM0090 Rev 17

1469/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Bit 31 Reserved, must be kept at reset value.
Bits 30:29 RXDPID: Received data PID (access type is "r")
Applies to isochronous OUT endpoints only.
This is the data PID received in the last packet for this endpoint.
00: DATA0
01: DATA2
10: DATA1
11: MDATA
STUPCNT: SETUP packet count (access type is "rw")
Applies to control OUT Endpoints only.
This field specifies the number of back-to-back SETUP data packets the endpoint can
receive.
01: 1 packet
10: 2 packets
11: 3 packets
Bit 28:19 PKTCNT: Packet count
Indicates the total number of USB packets that constitute the Transfer Size amount of data
for this endpoint.
This field is decremented every time a packet (maximum size or short packet) is written to
the RxFIFO.
Bits 18:0 XFRSIZ: Transfer size
This field contains the transfer size in bytes for the current endpoint. The core only interrupts
the application after it has exhausted the transfer size amount of data. The transfer size can
be set to the maximum packet size of the endpoint, to be interrupted at the end of each
packet.
The core decrements this field every time a packet is read from the RxFIFO and written to
the external memory.

OTG_HS device endpoint-x DMA address register (OTG_HS_DIEPDMAx /
OTG_HS_DOEPDMAx) (x = 1..5, where x = Endpoint_number)
Address offset for IN endpoints: 0x914 + 0x20 * x
Reset value: 0xXXXX XXXX
Address offset for OUT endpoints: 0xB14 + 0x20 * x
Reset value: 0xXXXX XXXX
31

30

29

28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

DMAADDR
rw

rw

rw

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:0 DMAADDR: DMA address
This bit holds the start address of the external memory for storing or fetching endpoint data.
Note: For control endpoints, this field stores control OUT data packets as well as SETUP
transaction data packets. When more than three SETUP packets are received back-toback, the SETUP data packet in the memory is overwritten. This register is incremented
on every AHB transaction. The application can give only a word-aligned address.

1470/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

35.12.5

OTG_HS power and clock gating control register
(OTG_HS_PCGCCTL)
Address offset: 0xE00
Reset value: 0x0000 0000
This register is available in host and peripheral modes.
9

8

7

6

5

4

3

Reserved

2

1

0

rw

STPPCLK

28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

GATEHCLK

29

Reserved

30

PHYSUSP

31

rw rw

Bit 31:5 Reserved, must be kept at reset value.
Bit 4 PHYSUSP: PHY suspended
Indicates that the PHY has been suspended. This bit is updated once the PHY is suspended
after the application has set the STPPCLK bit (bit 0).
Bits 3:2 Reserved, must be kept at reset value.
Bit 1 GATEHCLK: Gate HCLK
The application sets this bit to gate HCLK to modules other than the AHB Slave and Master
and wakeup logic when the USB is suspended or the session is not valid. The application
clears this bit when the USB is resumed or a new session starts.
Bit 0 STPPCLK: Stop PHY clock
The application sets this bit to stop the PHY clock when the USB is suspended, the session
is not valid, or the device is disconnected. The application clears this bit when the USB is
resumed or a new session starts.

35.12.6

OTG_HS register map
The table below gives the USB OTG register map and reset values.

Reset value

RM0090 Rev 17

0

SRQ

0

Reserved

SRQSCS

0
SEDET

HNPRQ
0

0

0

Res.

0

0

0

Reserved

GINT

Reserved

HNGSCS

0

0

SRSSCHG

0

0

HNSSCHG

0

DHNPEN

1

TXFELVL

OTG_HS_GA
HBCFG

0

Reserved

PTXFELVL

0x008

0

HSHNPEN

Reset value

Reserved

0

Reserved

Reserved

OTG_HS_GO
TGINT

DBCT

0x004

CIDSTS

Reset value

ASVLD

Reserved

HNGDET

OTG_HS_GO
TGCTL

BSVLD

0x000

DBCDNE

Register

ADTOCHG

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 215. OTG_HS register map and reset values

0

1471/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Register

0x00C

OTG_HS_GU
SBCFG

CTXPKT

FDMOD

FHMOD

Reset value

0

0

0

OTG_HS_GR
STCTL

AHBIDL

DMAREQ

Reset value

1

0

OTG_HS_GIN
TSTS

WKUINT

SRQINT

DISCINT

CIDSCHG

Reset value

0

0

0

0

OTG_HS_GIN
TMSK

WUIM

DISCINT

CIDSCHGM

Reset value

0

0

0

0

OTG_HS_GR
XSTSR (Host
mode)
OTG_HS_GR
XSTSR
(peripheral
mode)

OTG_HS_GR
XSTSP
(peripheral
mode)
Reset value

0x024

OTG_HS_GR
XFSIZ

PKTSTS
0
FRMNUM

Reserved
0

0

0

0

0

0

0

0

0

0
FRMNUM

Reserved
0

0

0

0

0

0

0

0

0

PHYSEL

FCRST

CSRST

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

NPTXFE

RXFLVL

SOF

OTGINT

MMIS

CMOD

1

0

0

0

0

0

RXFLVLM

SOFM

OTGINT

MMISM

Reserved

0

0

0

0

0

0

0

BCNT
0

0

0

0

0

0

CHNUM
0

0

0

0

0

0

BCNT
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CHNUM
0

0

0

0

0

0

BCNT
0

0

EPNUM

BCNT

DPID
0

GINAKEFF

0

0

NPTXFEM

0

0

GINAKEFFM

0

0

GONAKEFF

0

0

GONAKEFFM

ESUSPM

0

0

Reserved

ESUSP

USBSUSPM

0

0

0

HSRST

1
Reserved

0

Reserved

USBRST

USBSUSP

0

DPID

PKTSTS
0

0

DPID

PKTSTS

Reserved

0

DPID

PKTSTS
0

ENUMDNE

0

0

USBRST

0

0

ENUMDNEM

IEPINT

0

ISOODRP

OEPINT

0

0

ISOODRPM

IISOIXFRM

0

Reserved

0

TOCAL

RXFFLSH

1

EOPF

IEPINT
0

0

0

EPNUM
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

RXFD

Reserved

Reset value

1472/1747

0

EOPFM

OEPINT
0

Reserved

IISOIXFR
0

Reserved

DATAFSUSP

0

IPXFR/INCOMPISOOUT

PRTIM

0

0

FSUSPM

HCIM

0

0

Reset value
0x020

1

TXFNUM

IPXFRM/IISOOXFRM

0

Reserved

HCINT

HPRTINT

0

Reserved

PTXFE
1

Reset value
OTG_HS_GR
XSTSP (Host
mode)

0

0

Reset value
0x01C

0

Reserve
d

TXFFLSH

0

SRPCAP

0

Reserved

0

HNPCAP

ULPIAR

ULPIFSLS

0

PHYLPCS

ULPICSM

0

Reserved

ULPIEVBUSD

0

Reserved

TSDPS

0

ULPIEVBUSI

PTCI

PCCI

ULPIIPD

0

PTXFEM

0x018

0

TRDT

Reserved

Reserved

0x014

Reserved

Reserved

0x010

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Offset

SRQIM

Table 215. OTG_HS register map and reset values (continued)

0

RM0090 Rev 17

0

0

0

0

1

0

0

0

RM0090

USB on-the-go high-speed (OTG_HS)

OTG_HS_GN
PTXFSIZ
(Host mode)
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0x108

OTG_HS_DIE
PTXF2

0x10C

OTG_HS_DIE
PTXF3

0x110

OTG_HS_DIE
PTXF4

Reset value

Reset value

Reset value

Reset value

Reset value

0x400

OTG_HS_HCF
G

0x404

OTG_HS_HFI
R

0x408

OTG_HS_HFN
UM

0x410

OTG_HS_HPT
XSTS

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

NPTXFSAV
0

0

0

0

ADDR

Reserved

1

0

0

0

1

0

0

0

0

REGADDR

0

0

RWDATA

Reserved

PRODUCT_ID
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

PTXFD
0

0

0

0

0

0

1

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

INEPTXSA
0

0

0

0

0

0

0

0

0

0

0

1

INEPTXFD
0

0

INEPTXSA

INEPTXFD
0

0

INEPTXSA

INEPTXFD
0

0

PTXSA

INEPTXFD
0

1

0

0

0

0

INEPTXSA
0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

Reserved

Reset value

Reset value

1

1

1

0

1

0

1

FTREM
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

1

1

PTXQSAV
0

0

Y

0

0

0

0

1

1

0

0

0

0

0

1

1

1

1

1

1

1

Y

Y

Y

Y

Y

Y

FRNUM

PTXQTOP
0

0
FRIVL

Reserved

Reset value

Reset value

FSLSPCS

0x104

OTG_HS_DIE
PTXF1

0

FSLSS

0x100

0

NPTQXSAV

Reserved

0

I2CDEVADR

0

Reserved

NPTXQTOP

0

.PWRDWN

0

OTG_HS_CID
Reset value

0

.I2CPADEN

0

VBUSASEN

0

VBUSBSEN

0

I2CDATSE0

Res.

0

OTG_HS_GC
CFG

OTG_HS_HPT
XFSIZ

0

TX0FSA

Reset value
0x03C

0

SOFOUTEN

0x038

OTG_HS_GI2
CCTL

0

RW

OTG_HS_GN
PTXSTS
Reset value

0x030

0

TX0FD

BSYDNE

Reset value
0x02C

0

OTG_HS_GN
PTXFSIZ
(peripheral
mode)

NPTXFSA

NOVBUSSENS

Reset value
0x028

NPTXFD

ACK

Register

I2CEN

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 215. OTG_HS register map and reset values (continued)

Y

Y

Y

Y

Y

1

1

1

PTXFSAVL
Y

RM0090 Rev 17

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

1473/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Reset value

0

0

0

0

DAD
0

0

0

0

0

0

DAD
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

MC
0

0

0

DAD
0

0

MC

DAD
0

0

MC
0

0

0

MC
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

RM0090 Rev 17

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

EPNUM
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

MPSIZ
0

0

0

0

0

EPNUM
0

0

MPSIZ

EPNUM
0

0

MPSIZ

EPNUM
0

0

MPSIZ

EPNUM
0

0

MPSIZ

EPNUM
0

0

MPSIZ

EPNUM
0

0

MPSIZ

EPNUM
0

0

MPSIZ

EPNUM
0

PCSTS

0

0

0

PENA

0

0

0

PCDET

ODDFRM

Reset value

0

0

POCA

OTG_HS_HC
CHAR8

1474/1747

0

PENCHNG

0

0

MC

0

EPDIR

0

0

0

EPDIR

0

DAD

0

EPDIR

ODDFRM

Reset value

0

EPDIR

OTG_HS_HC
CHAR7

0

EPDIR

0

0

EPDIR

0

0

0

EPDIR

0

0

0

EPDIR

ODDFRM

Reset value

0

MC

0

EPDIR

OTG_HS_HC
CHAR6

0

0

LSDEV

0

DAD

0

Reserved

0

0

LSDEV

0

0

Reserved

ODDFRM

Reset value

0

LSDEV

OTG_HS_HC
CHAR5

0

0

Reserved

0

0

0

LSDEV

ODDFRM

0

0

MC

0

Reserved

CHDIS

0

0

0

LSDEV

CHENA

Reset value

DAD

0

Reserved

OTG_HS_HC
CHAR4

0

LSDEV

0

0

Reserved

ODDFRM

0

0

LSDEV

CHDIS

0

0

0

Reserved

CHENA

Reset value

0

0

0

LSDEV

OTG_HS_HC
CHAR3

0

MC

0

0

Reserved

0

0

0

0

PTCTL

0

LSDEV

ODDFRM

0

DAD

0

EPTYP

CHDIS

0

0

EPTYP

CHENA

Reset value

0

EPTYP

OTG_HS_HC
CHAR2

0

EPTYP

0

0

EPTYP

ODDFRM

0

0

EPTYP

CHDIS

0

0

EPTYP

ODDFRM

CHENA

Reset value

0

MC

EPTYP

CHDIS

OTG_HS_HC
CHAR1

DAD

EPTYP

CHENA

0

CHDIS

0x600

0

CHENA

0x5E0

0

CHDIS

0x5C0

Reset value

CHENA

0x5A0

OTG_HS_HC
CHAR0

CHDIS

0x580

0

CHENA

0x560

0

PRES

PSP
D

Reserved

Reserved

OTG_HS_HPR
T

CHDIS

0x540

0

POCCHNG

0

CHENA

0x520

0

HAINTM

Reset value

0x500

0

Reserved

Reset value

0x440

0

PRST

0x418

OTG_HS_HAI
NTMSK

HAINT

Reserved

PSUSP

OTG_HS_HAI
NT

Reserved

0x414

PLSTS

Register

PPWR

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 215. OTG_HS register map and reset values (continued)

0

0

0

MPSIZ
0

0

0

0

0

0

0

0

RM0090

USB on-the-go high-speed (OTG_HS)

Register

0x620

OTG_HS_HC
CHAR9

CHENA

CHDIS

ODDFRM

Reset value

0

0

0

OTG_HS_HC
CHAR10

CHENA

CHDIS

ODDFRM

Reset value

0

0

0

OTG_HS_HC
CHAR11

CHENA

CHDIS

ODDFRM

Reset value

0

0

0

OTG_HS_HCS
PLT0

Reset value

0

XACTPOS
0

0

0

Reserved

OTG_HS_HCI
NT3

0

Reserved

Reset value

RM0090 Rev 17

XACTPOS

0

COMPLSPLT

SPLITEN

Reset value

Reserved

0

0

0

0

0

0

0

0

NAK

0

0

0

0

0

0

0

0

0

0

0

0

0

0

PRTADDR

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

PRTADDR

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

HUBADDR

0

CHH

0

XFRC

0

AHBERR

0

0

0

0

PRTADDR

0

0

0

0

0

0

0

0

0

0

0
XFRC

0x568

OTG_HS_HCS
PLT3

0

XFRC

0

Reset value

0x564

PRTADDR

HUBADDR

0

0

CHH

0

0

XFRC

0

0

STALL

0

0

NAK

0

0

AHBERR

0

0

STALL

OTG_HS_HCI
NT2

COMPLSPLT

SPLITEN
0

0

STALL

0x548

Reset value

Reserved

0

HUBADDR

Reserved

OTG_HS_HCS
PLT2

0

STALL

0

Reset value

0x544

0

CHH

XACTPOS
0

0

NAK

0

0

ACK

0

0

NYET

0

0

AHBERR

0

HUBADDR

0

0

NAK

OTG_HS_HCI
NT1

COMPLSPLT

SPLITEN
0

0

CHH

0

0

ACK

0x528

Reset value

Reserved

0

AHBERR

0

Reserved

OTG_HS_HCS
PL1

0

MPSIZ

Reset value

0x524

0

ACK

0

0

ACK

0

0

EPNUM
0

0

MPSIZ

TXERR

Reserved

0

0

NYET

0

0

NYET

0

0

NYET

LSDEV

EPTYP
0

0

0

BBERR

0

0

0

TXERR

0

0

0

BBERR

0

EPNUM

0

TXERR

0

0

BBERR

0

0

0

TXERR

0

0

0

BBERR

0

0

0

DTERR

0

MC

0

0

FRMOR

0

0

0

DTERR

DAD

0

0

FRMOR

0

0

DTERR

0

0

FRMOR

0

EPDIR

0

0

EPDIR

0

0

EPDIR

0

MC

0

Reserved

0

0

LSDEV

DAD

0

Reserved

0

LSDEV

0

Reserved

0

EPTYP

0

DTERR

OTG_HS_HCI
NT0

0

MPSIZ

FRMOR

0x508

0

EPNUM

XACTPOS

0x504

0

MC

COMPLSPLT

0x660

DAD

EPTYP

0x640

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Offset

SPLITEN

Table 215. OTG_HS register map and reset values (continued)

0

0

0

0

0

0

0

0

0

0

0

1475/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Register

0x584

OTG_HS_HCS
PLT4

Reset value

0

1476/1747

RM0090 Rev 17

0

ACK

NAK

STALL

0

0

0

0

0

0

0

0

XACTPOS

0

0

0

0

0

0

0

0

0

0

0

NAK

STALL

0

0

0

0

0

0

0

0

XFRC

0

CHH

0

AHBERR

0

0

0

0

PRTADDR

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

XFRC

0

CHH

0

0

0

0

PRTADDR

0

0

0

0

0

0

0

0

0

0

0
XFRC

0

0

PRTADDR

HUBADDR

0

0

STALL

0

CHH
0

XFRC

0

CHH

0

AHBERR

0

NAK

0

XFRC

NAK

STALL

0

ACK

0

AHBERR

ACK

0

CHH

Reset value

0

AHBERR

Reserved

0

0

STALL

0

0

NAK

OTG_HS_HCI
NT8

COMPLSPLT

SPLITEN
0

0

ACK

0x608

Reset value

Reserved

PRTADDR

HUBADDR

0

0

ACK

0

0

NYET

0

0

TXERR

0

Reserved

OTG_HS_HCS
PLT8

0

BBERR

0

Reset value

0x604

0

HUBADDR

XACTPOS
0

0

NYET

0

0

NYET

OTG_HS_HCI
NT7

COMPLSPLT

SPLITEN
0

0

TXERR

0x5E8

Reset value

Reserved

0

NYET

0

Reserved

OTG_HS_HCS
PLT7

0

TXERR

0

Reset value

0x5E4

0

BBERR

0

XACTPOS
0

0

BBERR

0

0

TXERR

OTG_HS_HCI
NT6

COMPLSPLT

SPLITEN
0

0

BBERR

0x5C8

Reset value

Reserved

0

DTERR

0

Reserved

OTG_HS_HCS
PLT6

0

HUBADDR

Reset value

0x5C4

0

FRMOR

0

0

DTERR

OTG_HS_HCI
NT5

0

0

FRMOR

0

XACTPOS

Reset value

Reserved

0

DTERR

0x5A8

OTG_HS_HCS
PLT5

COMPLSPLT

0x5A4

SPLITEN

Reset value

0

NYET

Reserved

0

TXERR

0

BBERR

0

DTERR

0

FRMOR

0

FRMOR

0

PRTADDR

AHBERR

XACTPOS

COMPLSPLT
0

HUBADDR

DTERR

OTG_HS_HCI
NT4

Reserved

FRMOR

0x588

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Offset

SPLITEN

Table 215. OTG_HS register map and reset values (continued)

0

0

0

0

0

0

0

0

0

0

0

RM0090

USB on-the-go high-speed (OTG_HS)

Register

0x624

OTG_HS_HCS
PLT9

Reset value

0

RM0090 Rev 17

XACTPOS

0

0

0

0

PRTADDR

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0
XFRCM

0

0

0

0

0

0

0

0

0

0

0

0
XFRCM

0

CHHM

XFRCM

0

CHHM

0

AHBERRM

0

STALLM

0

NAKM

0

ACKM

XFRCM

0

CHHM

0

AHBERRM

0

STALLM

0

NAKM

0

ACKM

0

NYET

XFRCM

0

CHHM

0

AHBERRM

0

STALLM

0

NAKM

0

ACKM

0

NYET

XFRC

0

CHH

0

AHBERR

0

STALL

0

NAK

0

ACK

0

NYET

0

TXERR

0

BBERR

0

0

TXERRM

0

CHH

0

XFRC

STALL

0

CHH

NAK

0

BBERRM

0

XFRC

NAK

STALL

ACK

0

HUBADDR

0

AHBERR

ACK

NYET

0

AHBERRM

Reset value

0

CHHM

Reserved

0

AHBERRM

OTG_HS_HCI
NTMSK4

0

STALLM

0x58C

0

STALLM

Reset value

0

NAKM

Reserved

0

NAKM

OTG_HS_HCI
NTMSK3

0

ACKM

0x56C

0

ACKM

Reset value

0

NYET

Reserved

0

NYET

OTG_HS_HCI
NTMSK2

0

NYET

0x54C

0

TXERRM

Reset value

PRTADDR

TXERR

0

0

BBERR

0

0

DTERR

0

0

BBERRM

Reserved

0

TXERRM

OTG_HS_HCI
NTMSK1

0

BBERRM

0x52C

0

TXERRM

Reset value

0

BBERRM

Reserved

0

TXERRM

OTG_HS_HCI
NTMSK0

0

BBERRM

0x50C

0

DTERR

Reset value

0

FRMOR

Reserved

0

0

DTERRM

0

0

FRMORM

OTG_HS_HCI
NT11

COMPLSPLT

SPLITEN
0

0

DTERRM

0x668

Reset value

Reserved

0

FRMOR

0

Reserved

OTG_HS_HCS
PLT11

0

HUBADDR

Reset value

0x664

0

FRMORM

0

0

DTERRM

OTG_HS_HCI
NT10

0

0

FRMORM

0

XACTPOS

Reset value

Reserved

0

DTERRM

0x648

OTG_HS_HCS
PLT10

COMPLSPLT

0x644

SPLITEN

Reset value

0

NYET

Reserved

0

TXERR

0

BBERR

0

DTERR

0

FRMOR

0

FRMORM

0

PRTADDR

AHBERR

XACTPOS

COMPLSPLT
0

HUBADDR

DTERRM

OTG_HS_HCI
NT9

Reserved

FRMORM

0x628

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Offset

SPLITEN

Table 215. OTG_HS register map and reset values (continued)

0

0

0

0

0

0

0

0

0

0

0

1477/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

0

OTG_HS_HCT
SIZ1
Reset value

0

OTG_HS_HCT
SIZ2
Reset value

0

OTG_HS_HCT
SIZ3
Reset value

0

1478/1747

DPID
0

0

PKTCNT
0

0

0

0

DPID
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

CHHM

XFRCM

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0
XFRCM

0

CHHM

0

AHBERRM

XFRCM

XFRCM

XFRCM

XFRCM

AHBERRM

CHHM
CHHM

0

CHHM

0

AHBERRM

0

NAKM

0

STALLM

0

ACKM

CHHM

0

AHBERRM

AHBERRM

AHBERRM

NAKM

STALLM

NAKM
NAKM

0

NAKM

0

STALLM

STALLM

STALLM

NYET

ACKM

0

ACKM

ACKM

ACKM

TXERRM

0

NYET

NYET

NYET

BBERRM

TXERRM

TXERRM

BBERRM

BBERRM

DTERRM

FRMORM

DTERRM

FRMORM

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

XFRSIZ
0

0

0

0

0

0

0

0

0

0

0

0

PKTCNT
0

0

XFRSIZ

PKTCNT

DPID
0

0

0

XFRSIZ

PKTCNT

DPID
0

0

0

XFRCM

0x570

DOPING

0x550

Reset value

DOPING

0x530

OTG_HS_HCT
SIZ0

DOPING

0x510

DOPING

Reset value

0

CHHM

Reserved

0

AHBERRM

OTG_HS_HCI
NTMSK11

0

NAKM

0x66C

0

STALLM

Reset value

0

NAKM

Reserved

0

STALLM

OTG_HS_HCI
NTMSK10

0

ACKM

0x64C

0

ACKM

Reset value

0

NYET

Reserved

0

NYET

OTG_HS_HCI
NTMSK9

0

NYET

0x62C

0

TXERRM

Reset value

0

BBERRM

Reserved

0

TXERRM

OTG_HS_HCI
NTMSK8

0

BBERRM

0x60C

0

TXERRM

Reset value

0

BBERRM

Reserved

0

TXERRM

OTG_HS_HCI
NTMSK7

0

BBERRM

0x5EC

0

DTERRM

Reset value

0

FRMORM

Reserved

0

DTERRM

OTG_HS_HCI
NTMSK6

0

FRMORM

0x5CC

0

DTERRM

Reset value

FRMORM

Reserved

DTERRM

OTG_HS_HCI
NTMSK5

FRMORM

0x5AC

DTERRM

Register

FRMORM

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 215. OTG_HS register map and reset values (continued)

0

0

0

XFRSIZ
0

0

0

0

0

0

RM0090 Rev 17

0

0

0

0

0

0

0

0

0

RM0090

USB on-the-go high-speed (OTG_HS)

Register

0x590

OTG_HS_HCT
SIZ4

DOPING

Reset value

0

OTG_HS_HCT
SIZ5

DOPING

Reset value

0

OTG_HS_HCT
SIZ6

DOPING

Reset value

0

OTG_HS_HCT
SIZ7

DOPING

Reset value

0

OTG_HS_HCT
SIZ8

DOPING

Reset value

0

OTG_HS_HCT
SIZ9

DOPING

Reset value

0

OTG_HS_HCT
SIZ10

DOPING

Reset value

0

0x670

OTG_HS_HCT
SIZ11
Reset value

0

0x514

OTG_HS_HC
DMA0

0x524

OTG_HS_HC
DMA1

0x544

OTG_HS_HC
DMA2

0x564

OTG_HS_HC
DMA3

0x584

OTG_HS_HC
DMA4

0x5A4

OTG_HS_HC
DMA5

0x5B0

0x5D0

0x5F0

0x610

0x630

0x650

Reset value

Reset value

Reset value

Reset value

Reset value

Reset value

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Offset

DOPING

Table 215. OTG_HS register map and reset values (continued)

DPID
0

0

PKTCNT
0

0

0

0

DPID
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

XFRSIZ
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

XFRSIZ
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

XFRSIZ
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

XFRSIZ
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

XFRSIZ
0

0

0

0

0

0

0

0

0

0

0

0

PKTCNT
0

0

XFRSIZ

PKTCNT

DPID
0

0

PKTCNT

DPID
0

0

PKTCNT

DPID
0

0

PKTCNT

DPID
0

0

PKTCNT

DPID
0

0

PKTCNT

DPID
0

0

XFRSIZ

0

0

0

XFRSIZ
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMAADDR
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMAADDR
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMAADDR
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMAADDR
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMAADDR
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMAADDR
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

RM0090 Rev 17

0

0

0

1479/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

0x624

OTG_HS_HC
DMA9

0x644

OTG_HS_HC
DMA10

0x664

OTG_HS_HC
DMA11

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMAADDR
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMAADDR
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMAADDR
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Reserved

Reset value

FNSOF

0

0

0

0

0

0

Reserved

0x81C

OTG_HS_DAI
NTMSK

0x828

OTG_HS_DVB
USDIS

Reserved

Reset value

Reset value

Reset value

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

1

0

1

1

1

IEPINT
0

0

0

0

0

0

0

0

0

0

0

0

0

0

OEPM
0

0

0

OEPINT

0

0

IEPM
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

VBUSDT

Reserved

Reset value

1480/1747

0

0
BERRM

0x818

OTG_HS_DAI
NT

0

Reserved

NAKMSK

OTG_HS_DO
EPMSK

0

NYETMSK

Reset value

0x814

Reserved
0

0

XFRCM

OTG_HS_DIE
PMSK

0

0

EPDM

0x810

0

0

TOM

Reset value

0

0

AHBERRM

Reserved

0

0

INEPNMM

OTG_HS_DST
S

NAKM

0x808

0

0

XFRCM

OTG_HS_DCT
L

0

RWUSIG

0

SUSPSTS

Reserved

SDIS

0

ENUMSPD

0

EPDM

0

GINSTS

0

GONSTS

0

EERR

0

STUPM

0

AHBERRM

0

TCTL

1

0

ITTXFEMSK

Reserved

0

OTEPDM

0

STSPHSRXM

0

DAD

0

INEPNEM

0

B2BSTUP

0

Reserved

0

Reserved

0

TXFURM

0

DSPD

DMAADDR

Reset value

0x804

0

DMAADDR

OTG_HS_
DCFG

0x800

0

OPEM

Reset value

0

Reserved

Reset value

0

PFIVL

Reset value

0

Reserved

Reset value

0

PERSCHIVL

Reset value

0

NZLSOHSK

0x604

OTG_HS_HC
DMA8

Reset value

Reserved

0x5E4

OTG_HS_HC
DMA7

SGINAK

DMAADDR

CGINAK

OTG_HS_HC
DMA6

SGONAK

0x5C4

CGONAK

Register

POPRGDNE

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 215. OTG_HS register map and reset values (continued)

0

RM0090 Rev 17

0

0

1

0

1

1

1

1

RM0090

USB on-the-go high-speed (OTG_HS)

0

0

0

0

0

0

0

0

0

0

0

1

1

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

OTG_HS_DEA
CHINTMSK

0

0

0

0

0

0

USBAEP

0

NAKSTS

0

EONUM/DPID

0

EPTYP

0

STALL

0

TXFNUM

Reserved

0x918

CNAK

0

SNAK

0

SODDFRM

EPDIS

Reset value
TG_FS_DTXF
STS0

SD0PID/SEVNFRM

EPENA

0

0

0

0

0

0

0

0

0

USBAEP

0

NAKSTS

0

EONUM/DPID

0

EPTYP

0

Stall

0

TXFNUM

Reserved

0x938

CNAK

0

SNAK

0

SODDFRM

EPDIS

Reset value

SD0PID/SEVNFRM

EPENA

0

TG_FS_DTXF
STS1

EPDM

XFRCM

0

0

0

0

EPDM

0

XFRCM

TOM

AHBERRM

0

AHBERRM

ITTXFEMSK
0

TOM

Reserved

0

0

0

0

0

0

0

MPSIZ

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

INEPTFSAV

Reset value

OTG_HS_DIE
PCTL1

0

0

ITTXFEMSK

0

0

INEPNEM

0

0

INEPNMM

0

Reserved

0x920

0

INEPNMM

NAKM

Reset value

BERRM

Reserved

NYETM

0

OTG_HS_DIE
PCTL0

0

INEPNEM

Reserve
d

Reserved

NAKM

Reserved

0x900

0

Reserved
0

OTG_HS_DO
EPEACHMSK
1

0

0

Reserved

OTG_HS_DIE
PEACHMSK1

0

Reserved
0

Reset value

0x884

0

Reserved

Reset value

0x844

0

0

Reset value

0x83C

0

INEPTXFEM

Reset value
0x838

1

TXTHRLEN

Reserved

Reserved

OTG_HS_DEA
CHINT

1

ISOTHREN

0

0

Reserved Reserved

Reset value

RXTHRLEN

1

BIM

0x834

OTG_HS_DIE
PEMPMSK

Reserved

RXTHREN

OTG_HS_DTH
RCTL

Reserved

0x830

0

ARPEN

Reset value

NONISOTHREN

DVBUSP

Reserved

TXFURM

OTG_HS_DVB
USPULSE

BIM

0x82C

TXFURM

Register

Reserved

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 215. OTG_HS register map and reset values (continued)

0

0

0

0

0

0

0

0

1

0

0

0

MPSIZ

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

INEPTFSAV

Reserved

Reset value

0

RM0090 Rev 17

0

0

0

0

0

1

0

0

0

1481/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Register

0x940

OTG_HS_DIE
PCTL2

EPDIS

Reset value

0

0

0x958

TG_FS_DTXF
STS2

0

0

0

0

0

0

USBAEP

0

NAKSTS

0

EONUM/DPID

0

EPTYP

CNAK

0

Stall

SNAK

0

TXFNUM

Reserved

SODDFRM

SD0PID/SEVNFRM

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Offset

EPENA

Table 215. OTG_HS register map and reset values (continued)

0

0

0

0

0

0

0

0

0

USBAEP

0

NAKSTS

0

EONUM/DPID

0

EPTYP

0

Stall

0

TXFNUM

Reserved

0x978

CNAK

0

SNAK

EPDIS

0

SODDFRM

EPENA

Reset value
TG_FS_DTXF
STS3

SD0PID/SEVNFRM

OTG_HS_DIE
PCTL3

0

0

0

1482/1747

Reserved

0

0

0

USBAEP
USBAEP

NAKSTS

EONUM/DPID

NAKSTS

EPTYP
EPTYP

EONUM/DPID

0

0

0

0

0

0

0

EPT
YP

0

0

USBAEP

0

0

0

USBAEP

0

0

0

0

0

0

RM0090 Rev 17

USBAEP

TXFNUM

0

0

EPTYP

Stall
0

0

NAKSTS

0

0

EONUM/DPID

0

0

NAKSTS

0

TXFNUM

0

0

0

EONUM/DPID

0

0

0

NAKSTS

0

0

0

Reserved

0

0

0

EPTYP

CNAK
CNAK

0

CNAK

SNAK

0
Reserved

SNAK

EPDIS
0

SNAK

SODDFRM

SD0PID/SEVNFRM

0

SODDFRM

0

SD0PID/SEVNFRM

EPDIS

0

0

Reserved

0

0

STALL

0

0

TXFNUM

0

Reserved

Reset value

0

0

STALL

OTG_HS_DO
EPCTL0

0

0

Reserved

0

0

0

STALL

CNAK

Reset value

CNAK

OTG_HS_DIE
PCTL7

0

0

Reserved

SNAK

0

SNAK

Reset value

0

0

SNPM

SODDFRM

SD0PID/SEVNFRM

OTG_HS_DIE
PCTL6

SODDFRM

0

SD0PID/SEVNFRM

Reset value

TXFNUM

STALL

EPDIS

OTG_HS_DIE
PCTL5

EPDIS

EPENA

0

EPENA

0

EPENA

0

EPENA

0

EPDIS

0xB00

0

0

EPENA

0x9E0

0

Reset value

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

MPSIZ

Reserved

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

INEPTFSAV

Reset value

0x9C0

0

Reserved

OTG_HS_DIE
PCTL4

0

INEPTFSAV
0

0x960

0x9A0

0

Reserved

Reset value

0x980

MPSIZ

Reserved

1

0

0

0

0

0

1

0

0

0

MPSIZ

Reserved

0

0

0

0

0

0

0

MPSIZ

Reserved

0

0

0

0

0

0

0

MPSIZ

Reserved

0

0

0

0

0

0

0

MPSIZ

Reserved

0

0

0

0

Reserved

0

0

0

MPSI
Z
0

0

0x968

0x988

OTG_HS_DIE
PINT3

OTG_HS_DIE
PINT4

Reserved

Reset value

Reserved

Reset value

0

RM0090 Rev 17

0

0

0

INEPNM

ITTXFE

TOC

AHBERR

EPDISD

XFRC

0

1

0

0

0

0

0

0

0

TOC

AHBERR

EPDISD

XFRC

0
INEPNE
INEPNM
ITTXFE
TOC
AHBERR
EPDISD
XFRC

0

ITTXFE

0

INEPNM

Reset value

INEPNE

Reserved
TXFIFOUDRN
TXFE
INEPNE
INEPNM
ITTXFE
TOC
AHBERR
EPDISD
XFRC

0

INEPNE

0

TXFE

Reset value
TXFIFOUDRN
TXFE
INEPNE
INEPNM
ITTXFE
TOC
AHBERR
EPDISD
XFRC

0

TXFIFOUDRN

0

0

TXFE

Reserved
0

0

TXFIFOUDRN

Reset value
0

SNPM

NAKSTS
EONUM/DPID
USBAEP

EPTYP

STALL

0

TXFE

Reserved

0

TXFIFOUDRN

0

Reserved

0

Reserved

USBAEP

USBAEP

0

Reserved

NAKSTS
EONUM/DPID

NAKSTS

0

Reserved

0

0

EONUM/DPID

EPTYP

0
EPTYP

Stall
SNPM

Stall
SNPM

0
0

PKTDRPSTS

0
0

PKTDRPSTS

CNAK

0
0

PKTDRPSTS

CNAK

SNAK

0
0

NAK

SNAK

SODDFRM
SD0PID/SEVNFRM

0
0

0

Reserved

SODDFRM
SD0PID/SEVNFRM

EPDIS

0
Reserved
0

0

NAK

EPDIS

0

0

Reserved

OTG_HS_DIE
PINT2
0

0

PKTDRPSTS

0x948
OTG_HS_DIE
PINT1
0
0

PKTDRPSTS

0x928
OTG_HS_DIE
PINT0
0
0

Reserved

0x908
0
Reserved
0

NAK

0
0

BERR

Reset value
0

NAK

OTG_HS_DO
EPCTL3

0xB60
0

Reserved

0
0

NAK

EPENA

Reset value
0

Reserved

Reserved

OTG_HS_DO
EPCTL2

0xB40
CNAK

0

SNAK

EPENA

Reset value
SODDFRM

OTG_HS_DO
EPCTL1
SD0PID/SEVNFRM

0xB20

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Register

EPDIS

Offset

EPENA

RM0090
USB on-the-go high-speed (OTG_HS)

Table 215. OTG_HS register map and reset values (continued)

Reserved
MPSIZ

0

0

0

Reserved

0

Reserved

0

0

0

0

0

MPSIZ

0

0
0
0
0

0
0
0
0

MPSIZ

0
0
0
0
0
0
0
0
0

0
1
0
0
0
0
0
0
0

0
1
0
0
0
0
0
0
0

0
1
0
0
0
0
0
0
0

0

1

0

0

0

0

0

0

0

1483/1747

1541

0xB88

0xBA8

OTG_HS_DO
EPINT4

OTG_HS_DO
EPINT5

1484/1747

Reserved

Reset value

Reserved

Reset value

RM0090 Rev 17

0

0

0

0

0

0

0

0

AHBERR

EPDISD

XFRC

0

0

0

0

EPDISD

XFRC

0

0

AHBERR

0
STUP
AHBERR
EPDISD
XFRC

0
OTEPDIS

STUP
AHBERR
EPDISD
XFRC

Reserved
OTEPDIS

0

STUP

0
Reserved

0
B2BSTUP

STUP
AHBERR
EPDISD
XFRC

Reserved

B2BSTUP

Reserved

OUTPKTERR

OTEPDIS

0

STUP

0

STUP
AHBERR
EPDISD
XFRC

Reserved

ITTXFE
TOC
AHBERR
EPDISD
XFRC

INEPNE
INEPNM

TXFE
Reserved
B2BSTUP

TXFIFOUDRN

OUTPKTERR

OTEPDIS

0

OTEPDIS

0

0

Reserved

0

1

OTEPDIS

0

0

Reserved

0

B2BSTUP

0
0

B2BSTUP

0

0

B2BSTUP

0

Reserved

0

OUTPKTERR

0

Reserved

0

OUTPKTERR

0

Reserved

0

Reserved

TXFE
INEPNE
INEPNM
ITTXFE
TOC
AHBERR
EPDISD
XFRC

Reserved

PKTDRPSTS

TXFIFOUDRN

0

OUTPKTERR

0

PKTDRPSTS

NAK
Reserved

NAK

TXFE
INEPNE
INEPNM
ITTXFE
TOC
AHBERR
EPDISD
XFRC

Reserved

PKTDRPSTS

Reserved

TXFIFOUDRN

0

Reserved

Reset value

Reserved

0

Reserved

NAK

Reserved
Reserved

Reset value

Reserved

NAK
BERR

Reserved

Reserved

BERR

0

OUTPKTERR

Reset value
BERR

Reset value

Reserved

Reserved
BERR

Reset value
NYET

Reserved

Reserved

OTG_HS_DO
EPINT3
Reserved
NAK

Reset value

BERR

0xB68
OTG_HS_DO
EPINT2
Reserved
NYET

Reset value

BERR

0xB48
OTG_HS_DO
EPINT1
Reserved

NAK

0xB28
OTG_HS_DO
EPINT0

NYET

0xB08
OTG_HS_DIE
PINT7

NAK

0x9E8
OTG_HS_DIE
PINT6

NYET

0x9C8

NAK

OTG_HS_DIE
PINT5

NYET

0x9A8

NAK

Register

NYET

Offset
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

USB on-the-go high-speed (OTG_HS)
RM0090

Table 215. OTG_HS register map and reset values (continued)

0
1
0
0
0
0
0
0
0

0
1
0
0
0
0
0
0
0

0
0
0
0
0
0

0
1
0
0
0

0
0
0
0
0

0
0
0
0
0

0
0
0
0
0

0

0

0

0

0

RM0090

USB on-the-go high-speed (OTG_HS)

Reserved

0x934

0x93C

OTG_HS_DIE
PDMAB1

0x954

0x95C

OTG_HS_DIE
PDMAB2

Reset value

Reset value
0x970

OTG_HS_DIE
PTSIZ3
Reset value

0x974

OTG_HS_DIE
PDMA3

0x97C

OTG_HS_DIE
PDMAB3

Reset value

Reset value

0xB10

OTG_HS_DO
EPTSIZ0
Reset value

0xB30

OTG_HS_DO
EPTSIZ1

0xB34

OTG_HS_DO
EPDMA1

Reset value

Reset value

Reserved

0

0

0

0

0

0

EPDISD

XFRC

0

XFRC

0

EPDISD

STUP

AHBERR

0

AHBERR

OTEPDIS

0

STUP

B2BSTUP

Reserved

Reserved
Reserved

B2BSTUP

OUTPKTERR

Reserved

0

0

0

0

XFRSIZ

Reserved

0

PKTCNT
0

0

0

0

0

0

0

0

0

0

XFRSIZ
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

MCN
T
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

PKTCNT
0

0

0

0

0

0

XFRSIZ
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMAADDR
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMABADDR
0
Reserved

Reset value
OTG_HS_DIE
PDMA2

0

0

DMABADDR

0

0

0

0

0

0

MCN
T
0

0

0

0

0

0

0

0

0

0

0

0

0

PKTCNT
0

0

0

0

0

0

XFRSIZ
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMAADDR
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMABADDR
0
Reserved

0x950

0

Reserved

Reset value

0

0

0

DMAADDR

Reset value

OTG_HS_DIE
PTSIZ2

0

0

0

0

0

0

0

STU
PCN
T
0

0

0

0

0

0

0

0

Reserved

0

0

0

0

0

0

0

0

PKTCNT

Reset value
OTG_HS_DIE
PDMA1

MCN
T

XFRSIZ

Reserved

0

RXDPID/
STUPCNT

0x930

0
Reserved

Reset value
OTG_HS_DIE
PTSIZ1

0
PKT
CNT

0

OTEPDIS

Reset value

0

Reserved

0x910

OTG_HS_DIE
PTSIZ0

Reserved

0

OUTPKTERR

OTG_HS_DO
EPINT7

0

Reserved

0xBE8

NAK

Reset value

BERR

Reserved

BERR

OTG_HS_DO
EPINT6

NYET

0xBC8

NAK

Register

NYET

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 215. OTG_HS register map and reset values (continued)

PKTCNT

0

0

0

0

0

0

0

0

0

0

0

0

0

XFRSIZ

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMAADDR
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

RM0090 Rev 17

0

0

0

1485/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Register

0xB3C

OTG_HS_DO
EPDMAB1

DMABADDR
0

0xB50

OTG_HS_DO
EPTSIZ2

0xB54

OTG_HS_DO
EPDMA2

0xB5C

OTG_HS_DO
EPDMAB2

0

OTG_HS_DO
EPTSIZ3

0xB74

OTG_HS_DO
EPDMA3

0xB7C

OTG_HS_DO
EPDMAB3

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

PKTCNT

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

XFRSIZ

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

PKTCNT

0

0

0

0

0

0

XFRSIZ

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

DMAADDR
0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

OTG_HS_PC
GCCTL

0

0

0

0

0

0

0

0

0

0

0

0

Reserved

Reserved

DMABADDR

PHYSUSP

0xE00

0

STPPCLK

0xB70

Reset value

0

GATEHCLK

0

Reset value

0

DMABADDR

Reset value

Reset value

0

DMAADDR

Reserved

Reset value

0

RXDPID/
STUPCNT

Reset value

0

RXDPID/
STUPCNT

Reset value

Reserved

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 215. OTG_HS register map and reset values (continued)

Reset value

Refer to Section 2.3: Memory map for the register boundary addresses.

35.13

OTG_HS programming model

35.13.1

Core initialization
The application must perform the core initialization sequence. If the cable is connected
during power-up, the current mode of operation bit in the Core interrupt register (CMOD bit
in OTG_HS_GINTSTS) reflects the mode. The OTG_HS controller enters host mode when
an “A” plug is connected or peripheral mode when a “B” plug is connected.
This section explains the initialization of the OTG_HS controller after power-on. The
application must follow the initialization sequence irrespective of host or peripheral mode
operation. All core global registers are initialized according to the core’s configuration:

1486/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
1.

2.

3.

Program the following fields in the Global AHB configuration (OTG_HS_GAHBCFG)
register:
–

DMA mode bit

–

AHB burst length field

–

Global interrupt mask bit GINT = 1

–

RxFIFO nonempty (RXFLVL bit in OTG_HS_GINTSTS)

–

Periodic TxFIFO empty level

Program the following fields in OTG_HS_GUSBCFG register:
–

HNP capable bit

–

SRP capable bit

–

FS timeout calibration field

–

USB turnaround time field

The software must unmask the following bits in the GINTMSK register:
OTG interrupt mask
Mode mismatch interrupt mask

4.

35.13.2

The software can read the CMOD bit in OTG_HS_GINTSTS to determine whether the
OTG_HS controller is operating in host or peripheral mode.

Host initialization
To initialize the core as host, the application must perform the following steps:
1.

Program the HPRTINT in GINTMSK to unmask

2.

Program the OTG_HS_HCFG register to select full-speed host

3.

Program the PPWR bit in OTG_HS_HPRT to 1. This drives VBUS on the USB.

4.

Wait for the PCDET interrupt in OTG_HS_HPRT0. This indicates that a device is
connecting to the port.

5.

Program the PRST bit in OTG_HS_HPRT to 1. This starts the reset process.

6.

Wait at least 10 ms for the reset process to complete.

7.

Program the PRST bit in OTG_HS_HPRT to 0.

8.

Wait for the PENCHNG interrupt in OTG_HS_HPRT.

9.

Read the PSPD bit in OTG_HS_HPRT to get the enumerated speed.

10. Program the HFIR register with a value corresponding to the selected PHY clock 1.
11. Program the FSLSPCS field in OTG_FS_HCFG register according to the speed of the
detected device read in step 9. If FSLSPCS has been changed, reset the port.
12. Program the OTG_HS_GRXFSIZ register to select the size of the receive FIFO.
13. Program the OTG_HS_GNPTXFSIZ register to select the size and the start address of
the nonperiodic transmit FIFO for nonperiodic transactions.
14. Program the OTG_HS_HPTXFSIZ register to select the size and start address of the
periodic transmit FIFO for periodic transactions.
To communicate with devices, the system software must initialize and enable at least one
channel.

RM0090 Rev 17

1487/1747
1541

USB on-the-go high-speed (OTG_HS)

35.13.3

RM0090

Device initialization
The application must perform the following steps to initialize the core as a device on powerup or after a mode change from host to device.
1.

2.

Program the following fields in the OTG_HS_DCFG register:
–

Device speed

–

Nonzero-length status OUT handshake

Program the OTG_HS_GINTMSK register to unmask the following interrupts:
–

USB reset

–

Enumeration done

–

Early suspend

–

USB suspend

–

SOF

3.

Program the VBUSBSEN bit in the OTG_HS_GCCFG register to enable VBUS sensing
in “B” peripheral mode and supply the 5 volts across the pull-up resistor on the DP line.

4.

Wait for the USBRST interrupt in OTG_HS_GINTSTS. It indicates that a reset has
been detected on the USB that lasts for about 10 ms on receiving this interrupt.

Wait for the ENUMDNE interrupt in OTG_HS_GINTSTS. This interrupt indicates the end of
reset on the USB. On receiving this interrupt, the application must read the OTG_HS_DSTS
register to determine the enumeration speed and perform the steps listed in Endpoint
initialization on enumeration completion on page 1515.
At this point, the device is ready to accept SOF packets and perform control transfers on
control endpoint 0.

35.13.4

DMA mode
The OTG host uses the AHB master interface to fetch the transmit packet data (AHB to
USB) and receive the data update (USB to AHB). The AHB master uses the programmed
DMA address (HCDMAx register in host mode and DIEPDMAx/DOEPDMAx register in
peripheral mode) to access the data buffers.

35.13.5

Host programming model
Channel initialization
The application must initialize one or more channels before it can communicate with
connected devices. To initialize and enable a channel, the application must perform the
following steps:

1488/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
1.

Program the GINTMSK register to unmask the following:

2.

Channel interrupt
–

Nonperiodic transmit FIFO empty for OUT transactions (applicable for Slave mode
that operates in pipelined transaction-level with the packet count field
programmed with more than one).

–

Nonperiodic transmit FIFO half-empty for OUT transactions (applicable for Slave
mode that operates in pipelined transaction-level with the packet count field
programmed with more than one).

3.

Program the OTG_HS_HAINTMSK register to unmask the selected channels’
interrupts.

4.

Program the OTG_HS_HCINTMSK register to unmask the transaction-related
interrupts of interest given in the host channel interrupt register.

5.

Program the selected channel’s OTG_HS_HCTSIZx register with the total transfer size,
in bytes, and the expected number of packets, including short packets. The application
must program the PID field with the initial data PID (to be used on the first OUT
transaction or to be expected from the first IN transaction).

6.

Program the selected channels in the OTG_HS_HCSPLTx register(s) with the hub and
port addresses (split transactions only).

7.

Program the selected channels in the HCDMAx register(s) with the buffer start address.

8.

Program the OTG_HS_HCCHARx register of the selected channel with the device’s
endpoint characteristics, such as type, speed, direction, and so forth. (The channel can
be enabled by setting the channel enable bit to 1 only when the application is ready to
transmit or receive any packet).

Halting a channel
The application can disable any channel by programming the OTG_HS_HCCHARx register
with the CHDIS and CHENA bits set to 1. This enables the OTG_HS host to flush the posted
requests (if any) and generates a channel halted interrupt. The application must wait for the
CHH interrupt in OTG_HS_HCINTx before reallocating the channel for other transactions.
The OTG_HS host does not interrupt the transaction that has already been started on the
USB.
To disable a channel in DMA mode operation, the application does not need to check for
space in the request queue. The OTG_HS host checks for space to write the disable
request on the disabled channel’s turn during arbitration. Meanwhile, all posted requests are
dropped from the request queue when the CHDIS bit in HCCHARx is set to 1.
Before disabling a channel, the application must ensure that there is at least one free space
available in the nonperiodic request queue (when disabling a nonperiodic channel) or the
periodic request queue (when disabling a periodic channel). The application can simply
flush the posted requests when the Request queue is full (before disabling the channel), by
programming the OTG_HS_HCCHARx register with the CHDIS bit set to 1, and the CHENA
bit cleared to 0.
The application is expected to disable a channel on any of the following conditions:
1.

When an XFRC interrupt in OTG_HS_HCINTx is received during a nonperiodic IN
transfer or high-bandwidth interrupt IN transfer (Slave mode only)

2.

When an STALL, TXERR, BBERR or DTERR interrupt in OTG_HS_HCINTx is
received for an IN or OUT channel (Slave mode only). For high-bandwidth interrupt INs
in Slave mode, once the application has received a DTERR interrupt it must disable the

RM0090 Rev 17

1489/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

channel and wait for a channel halted interrupt. The application must be able to receive
other interrupts (DTERR, NAK, Data, TXERR) for the same channel before receiving
the halt.
3.

When a DISCINT (Disconnect Device) interrupt in OTG_HS_GINTSTS is received.
(The application is expected to disable all enabled channels

4.

When the application aborts a transfer before normal completion.

Ping protocol
When the OTG_HS host operates in high speed, the application must initiate the ping
protocol when communicating with high-speed bulk or control (data and status stage) OUT
endpoints.
The application must initiate the ping protocol when it receives a NAK/NYET/TXERR
interrupt. When the HS_OTG host receives one of the above responses, it does not
continue any transaction for a specific endpoint, drops all posted or fetched OUT requests
(from the request queue), and flushes the corresponding data (from the transmit FIFO).
This is valid in slave mode only. In Slave mode, the application can send a ping token either
by setting the DOPING bit in HCTSIZx before enabling the channel or by just writing the
HCTSIZx register with the DOPING bit set when the channel is already enabled. This
enables the HS_OTG host to write a ping request entry to the request queue. The
application must wait for the response to the ping token (a NAK, ACK, or TXERR interrupt)
before continuing the transaction or sending another ping token. The application can
continue the data transaction only after receiving an ACK from the OUT endpoint for the
requested ping. In DMA mode operation, the application does not need to set the DOPING
bit in HCTSIZx for a NAK/NYET response in case of Bulk/Control OUT. The OTG_HS host
automatically sets the DOPING bit in HCTSIZx, and issues the ping tokens for Bulk/Control
OUT. The HS_OTG host continues sending ping tokens until it receives an ACK, and then
switches automatically to the data transaction.

Operational model
The application must initialize a channel before communicating to the connected device.
This section explains the sequence of operation to be performed for different types of USB
transactions.
•

Writing the transmit FIFO

The OTG_HS host automatically writes an entry (OUT request) to the periodic/nonperiodic
request queue, along with the last word write of a packet. The application must ensure that
at least one free space is available in the periodic/nonperiodic request queue before starting
to write to the transmit FIFO. The application must always write to the transmit FIFO in
words. If the packet size is non-word-aligned, the application must use padding. The
OTG_HS host determines the actual packet size based on the programmed maximum
packet size and transfer size.

1490/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
Figure 416. Transmit FIFO write task
Start

Read GNPTXSTS/
HPTXFSIZ registers for
available FIFO and
queue spaces

W ait for
TXFELVL or PTXFELVL
interrupt in
OTG_FS_GAHBCFG

No

1 MPS
or LPS FIFO space
available?

Yes

Yes

W rite 1 packet
data to
Transmit FIFO

More packets
to send?

No
MPS: Maximum packet size
LPS: Lastt packet
ac et size

Done
ai15673

•

Reading the receive FIFO

The application must ignore all packet statuses other than IN data packet (bx0010).

RM0090 Rev 17

1491/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Figure 417. Receive FIFO read task
Start

No

RXFLVL
interrupt ?

Yes

Unmask RXFLVL
interrupt

Read the received
packet from the
Receive FIFO

Mask RXFLVL
interrupt

Unmask RXFLVL
interrupt

Read
OTG_FS_GRXSTSP

PKTSTS
0b0010?

No
No

Yes
Yes

BCNT > 0?

ai15674

•

Bulk and control OUT/SETUP transactions
A typical bulk or control OUT/SETUP pipelined transaction-level operation is shown in
Figure 418. See channel 1 (ch_1). Two bulk OUT packets are transmitted. A control
SETUP transaction operates in the same way but has only one packet. The
assumptions are:

•

–

The application is attempting to send two maximum-packet-size packets (transfer
size = 1, 024 bytes).

–

The nonperiodic transmit FIFO can hold two packets (128 bytes for FS).

–

The nonperiodic request queue depth = 4.

Normal bulk and control OUT/SETUP operations
The sequence of operations for channel 1 is as follows:

1492/1747

a)

Initialize channel 1

b)

Write the first packet for channel 1

c)

Along with the last word write, the core writes an entry to the nonperiodic request
queue

d)

As soon as the nonperiodic queue becomes nonempty, the core attempts to send
an OUT token in the current frame

e)

Write the second (last) packet for channel 1

f)

The core generates the XFRC interrupt as soon as the last transaction is
completed successfully

g)

In response to the XFRC interrupt, de-allocate the channel for other transfers

h)

Handling nonACK responses

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
Figure 418. Normal bulk/control OUT/SETUP and bulk/control IN transactions - DMA
mode

RM0090 Rev 17

1493/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Figure 419. Normal bulk/control OUT/SETUP and bulk/control IN transactions - Slave
mode

The channel-specific interrupt service routine for bulk and control OUT/SETUP
transactions in Slave mode is shown in the following code samples.
•

Interrupt service routine for bulk/control OUT/SETUP and bulk/control IN
transactions

a) Bulk/Control OUT/SETUP
Unmask (NAK/TXERR/STALL/XFRC)
if (XFRC)
{
Reset Error Count

1494/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
Mask ACK
De-allocate Channel
}
else if (STALL)
{
Transfer Done = 1
Unmask CHH
Disable Channel
}
else if (NAK or TXERR )
{
Rewind Buffer Pointers
Unmask CHH
Disable Channel
if (TXERR)
{
Increment Error Count
Unmask ACK
}
else
{
Reset Error Count
}
}
else if (CHH)
{
Mask CHH
if (Transfer Done or (Error_count == 3))
{
De-allocate Channel
}
else
{
Re-initialize Channel
}
}
else if (ACK)
{
Reset Error Count
Mask ACK
}
The application is expected to write the data packets into the transmit FIFO as and
when the space is available in the transmit FIFO and the Request queue. The
application can make use of the NPTXFE interrupt in OTG_HS_GINTSTS to find the
transmit FIFO space.
b) Bulk/Control IN
Unmask (TXERR/XFRC/BBERR/STALL/DTERR)
if (XFRC)
{
Reset Error Count
Unmask CHH
Disable Channel
RM0090 Rev 17

1495/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Reset Error Count
Mask ACK
}
else if (TXERR or BBERR or STALL)
{
Unmask CHH
Disable Channel
if (TXERR)
{
Increment Error Count
Unmask ACK
}
}
else if (CHH)
{
Mask CHH
if (Transfer Done or (Error_count == 3))
{
De-allocate Channel
}
else
{
Re-initialize Channel
}
}
else if (ACK)
{
Reset Error Count
Mask ACK
}
else if (DTERR)
{
Reset Error Count
}
The application is expected to write the requests as and when the Request queue
space is available and until the XFRC interrupt is received.
•

Bulk and control IN transactions
A typical bulk or control IN pipelined transaction-level operation is shown in Figure 420.
See channel 2 (ch_2). The assumptions are:

1496/1747

–

The application is attempting to receive two maximum-packet-size packets
(transfer size = 1 024 bytes).

–

The receive FIFO can contain at least one maximum-packet-size packet and two
status words per packet (72 bytes for FS).

–

The nonperiodic request queue depth = 4.

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
Figure 420. Bulk/control IN transactions - DMA mode

RM0090 Rev 17

1497/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Figure 421. Bulk/control IN transactions - Slave mode

The sequence of operations is as follows:

1498/1747

a)

Initialize channel 2.

b)

Set the CHENA bit in HCCHAR2 to write an IN request to the nonperiodic request
queue.

c)

The core attempts to send an IN token after completing the current OUT
transaction.

d)

The core generates an RXFLVL interrupt as soon as the received packet is written
to the receive FIFO.

e)

In response to the RXFLVL interrupt, mask the RXFLVL interrupt and read the
received packet status to determine the number of bytes received, then read the

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
receive FIFO accordingly. Following this, unmask the RXFLVL interrupt.
f)

The core generates the RXFLVL interrupt for the transfer completion status entry
in the receive FIFO.

g)

The application must read and ignore the receive packet status when the receive
packet status is not an IN data packet (PKTSTS in GRXSTSR ≠ 0b0010).

h)

The core generates the XFRC interrupt as soon as the receive packet status is
read.

i)

In response to the XFRC interrupt, disable the channel and stop writing the
OTG_HS_HCCHAR2 register for further requests. The core writes a channel
disable request to the nonperiodic request queue as soon as the
OTG_HS_HCCHAR2 register is written.

j)

The core generates the RXFLVL interrupt as soon as the halt status is written to
the receive FIFO.

k)

Read and ignore the receive packet status.

l)

The core generates a CHH interrupt as soon as the halt status is popped from the
receive FIFO.

m) In response to the CHH interrupt, de-allocate the channel for other transfers.
n)
•

Handling nonACK responses

Control transactions in slave mode
Setup, Data, and Status stages of a control transfer must be performed as three
separate transfers. Setup-, Data- or Status-stage OUT transactions are performed
similarly to the bulk OUT transactions explained previously. Data- or Status-stage IN
transactions are performed similarly to the bulk IN transactions explained previously.
For all three stages, the application is expected to set the EPTYP field in
OTG_HS_HCCHAR1 to Control. During the Setup stage, the application is expected to
set the PID field in OTG_HS_HCTSIZ1 to SETUP.

•

Interrupt OUT transactions
A typical interrupt OUT operation in Slave mode is shown in Figure 422. The
assumptions are:
–

The application is attempting to send one packet in every frame (up to 1 maximum
packet size), starting with the odd frame (transfer size = 1 024 bytes)

–

The periodic transmit FIFO can hold one packet (1 KB)

–

Periodic request queue depth = 4

The sequence of operations is as follows:
a)

Initialize and enable channel 1. The application must set the ODDFRM bit in
OTG_HS_HCCHAR1.

b)

Write the first packet for channel 1. For a high-bandwidth interrupt transfer, the
application must write the subsequent packets up to MCNT (maximum number of
packets to be transmitted in the next frame times) before switching to another
channel.

c)

Along with the last word write of each packet, the OTG_HS host writes an entry to
the periodic request queue.

d)

The OTG_HS host attempts to send an OUT token in the next (odd) frame.

e)

The OTG_HS host generates an XFRC interrupt as soon as the last packet is
transmitted successfully.

f)

In response to the XFRC interrupt, reinitialize the channel for the next transfer.

RM0090 Rev 17

1499/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Figure 422. Normal interrupt OUT/IN transactions - DMA mode

1500/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
Figure 423. Normal interrupt OUT/IN transactions - Slave mode

•

Interrupt service routine for interrupt OUT/IN transactions

a) Interrupt OUT
Unmask (NAK/TXERR/STALL/XFRC/FRMOR)
if (XFRC)
{
Reset Error Count
Mask ACK
De-allocate Channel
}
else
if (STALL or FRMOR)
{

RM0090 Rev 17

1501/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Mask ACK
Unmask CHH
Disable Channel
if (STALL)
{
Transfer Done = 1
}
}
else
if (NAK or TXERR)
{
Rewind Buffer Pointers
Reset Error Count
Mask ACK
Unmask CHH
Disable Channel
}
else
if (CHH)
{
Mask CHH
if (Transfer Done or (Error_count == 3))
{
De-allocate Channel
}
else
{
Re-initialize Channel (in next b_interval - 1 Frame)
}
}
else
if (ACK)
{
Reset Error Count
Mask ACK
}
The application is expected to write the data packets into the transmit FIFO when the
space is available in the transmit FIFO and the Request queue up to the count
specified in the MCNT field before switching to another channel. The application uses
the NPTXFE interrupt in OTG_HS_GINTSTS to find the transmit FIFO space.
b) Interrupt IN
Unmask (NAK/TXERR/XFRC/BBERR/STALL/FRMOR/DTERR)
if (XFRC)
{
Reset Error Count
Mask ACK
if (OTG_HS_HCTSIZx.PKTCNT == 0)
{
De-allocate Channel
}
else
{
1502/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
Transfer Done = 1
Unmask CHH
Disable Channel
}
}
else
if (STALL or FRMOR or NAK or DTERR or BBERR)
{
Mask ACK
Unmask CHH
Disable Channel
if (STALL or BBERR)
{
Reset Error Count
Transfer Done = 1
}
else
if (!FRMOR)
{
Reset Error Count
}
}
else
if (TXERR)
{
Increment Error Count
Unmask ACK
Unmask CHH
Disable Channel
}
else
if (CHH)
{
Mask CHH
if (Transfer Done or (Error_count == 3))
{
De-allocate Channel
}
else
Re-initialize Channel (in next b_interval - 1 /Frame)
}
}
else
if (ACK)
{
Reset Error Count
Mask ACK

RM0090 Rev 17

1503/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

}
The application is expected to write the requests for the same channel when the
Request queue space is available up to the count specified in the MCNT field before
switching to another channel (if any).
•

Interrupt IN transactions
The assumptions are:

•

–

The application is attempting to receive one packet (up to 1 maximum packet size)
in every frame, starting with odd (transfer size = 1 024 bytes).

–

The receive FIFO can hold at least one maximum-packet-size packet and two
status words per packet (1 031 bytes).

–

Periodic request queue depth = 4.

Normal interrupt IN operation
The sequence of operations is as follows:

1504/1747

a)

Initialize channel 2. The application must set the ODDFRM bit in
OTG_HS_HCCHAR2.

b)

Set the CHENA bit in OTG_HS_HCCHAR2 to write an IN request to the periodic
request queue. For a high-bandwidth interrupt transfer, the application must write
the OTG_HS_HCCHAR2 register MCNT (maximum number of expected packets
in the next frame times) before switching to another channel.

c)

The OTG_HS host writes an IN request to the periodic request queue for each
OTG_HS_HCCHAR2 register write with the CHENA bit set.

d)

The OTG_HS host attempts to send an IN token in the next (odd) frame.

e)

As soon as the IN packet is received and written to the receive FIFO, the OTG_HS
host generates an RXFLVL interrupt.

f)

In response to the RXFLVL interrupt, read the received packet status to determine
the number of bytes received, then read the receive FIFO accordingly. The
application must mask the RXFLVL interrupt before reading the receive FIFO, and
unmask after reading the entire packet.

g)

The core generates the RXFLVL interrupt for the transfer completion status entry
in the receive FIFO. The application must read and ignore the receive packet
status when the receive packet status is not an IN data packet (PKTSTS in
GRXSTSR ≠ 0b0010).

h)

The core generates an XFRC interrupt as soon as the receive packet status is
read.

i)

In response to the XFRC interrupt, read the PKTCNT field in OTG_HS_HCTSIZ2.
If the PKTCNT bit in OTG_HS_HCTSIZ2 is not equal to 0, disable the channel
before re-initializing the channel for the next transfer, if any). If PKTCNT bit in

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
OTG_HS_HCTSIZ2 = 0, reinitialize the channel for the next transfer. This time, the
application must reset the ODDFRM bit in OTG_HS_HCCHAR2.
•

Isochronous OUT transactions
A typical isochronous OUT operation in Slave mode is shown in Figure 424. The
assumptions are:
–

The application is attempting to send one packet every frame (up to 1 maximum
packet size), starting with an odd frame. (transfer size = 1 024 bytes).

–

The periodic transmit FIFO can hold one packet (1 KB).

–

Periodic request queue depth = 4.

The sequence of operations is as follows:
a)

Initialize and enable channel 1. The application must set the ODDFRM bit in
OTG_HS_HCCHAR1.

b)

Write the first packet for channel 1. For a high-bandwidth isochronous transfer, the
application must write the subsequent packets up to MCNT (maximum number of
packets to be transmitted in the next frame times before switching to another
channel.

c)

Along with the last word write of each packet, the OTG_HS host writes an entry to
the periodic request queue.

d)

The OTG_HS host attempts to send the OUT token in the next frame (odd).

e)

The OTG_HS host generates the XFRC interrupt as soon as the last packet is
transmitted successfully.

f)

In response to the XFRC interrupt, reinitialize the channel for the next transfer.

g)

Handling nonACK responses

RM0090 Rev 17

1505/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Figure 424. Normal isochronous OUT/IN transactions - DMA mode

1506/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
Figure 425. Normal isochronous OUT/IN transactions - Slave mode

•

Interrupt service routine for isochronous OUT/IN transactions

Code sample: Isochronous OUT
Unmask (FRMOR/XFRC)
if (XFRC)
{
De-allocate Channel
}
else
if (FRMOR)
{
Unmask CHH
Disable Channel
}

RM0090 Rev 17

1507/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

else
if (CHH)
{
Mask CHH
De-allocate Channel
}
Code sample: Isochronous IN
Unmask (TXERR/XFRC/FRMOR/BBERR)
if (XFRC or FRMOR)
{
if (XFRC and (OTG_HS_HCTSIZx.PKTCNT == 0))
{
Reset Error Count
De-allocate Channel
}
else
{
Unmask CHH
Disable Channel
}
}
else
if (TXERR or BBERR)
{
Increment Error Count
Unmask CHH
Disable Channel
}
else
if (CHH)
{
Mask CHH
if (Transfer Done or (Error_count == 3))
{
De-allocate Channel
}
else
{
Re-initialize Channel
}
}

1508/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
•

Isochronous IN transactions
The assumptions are:
–

The application is attempting to receive one packet (up to 1 maximum packet size)
in every frame starting with the next odd frame (transfer size = 1 024 bytes).

–

The receive FIFO can hold at least one maximum-packet-size packet and two
status words per packet (1 031 bytes).

–

Periodic request queue depth = 4.

The sequence of operations is as follows:

•

a)

Initialize channel 2. The application must set the ODDFRM bit in
OTG_HS_HCCHAR2.

b)

Set the CHENA bit in OTG_HS_HCCHAR2 to write an IN request to the periodic
request queue. For a high-bandwidth isochronous transfer, the application must
write the OTG_HS_HCCHAR2 register MCNT (maximum number of expected
packets in the next frame times) before switching to another channel.

c)

The OTG_HS host writes an IN request to the periodic request queue for each
OTG_HS_HCCHAR2 register write with the CHENA bit set.

d)

The OTG_HS host attempts to send an IN token in the next odd frame.

e)

As soon as the IN packet is received and written to the receive FIFO, the OTG_HS
host generates an RXFLVL interrupt.

f)

In response to the RXFLVL interrupt, read the received packet status to determine
the number of bytes received, then read the receive FIFO accordingly. The
application must mask the RXFLVL interrupt before reading the receive FIFO, and
unmask it after reading the entire packet.

g)

The core generates an RXFLVL interrupt for the transfer completion status entry in
the receive FIFO. This time, the application must read and ignore the receive
packet status when the receive packet status is not an IN data packet (PKTSTS bit
in OTG_HS_GRXSTSR ≠ 0b0010).

h)

The core generates an XFRC interrupt as soon as the receive packet status is
read.

i)

In response to the XFRC interrupt, read the PKTCNT field in OTG_HS_HCTSIZ2.
If PKTCNT≠ 0 in OTG_HS_HCTSIZ2, disable the channel before re-initializing the
channel for the next transfer, if any. If PKTCNT = 0 in OTG_HS_HCTSIZ2,
reinitialize the channel for the next transfer. This time, the application must reset
the ODDFRM bit in OTG_HS_HCCHAR2.

Selecting the queue depth
Choose the periodic and nonperiodic request queue depths carefully to match the
number of periodic/nonperiodic endpoints accessed.
The nonperiodic request queue depth affects the performance of nonperiodic transfers.
The deeper the queue (along with sufficient FIFO size), the more often the core is able
to pipeline nonperiodic transfers. If the queue size is small, the core is able to put in
new requests only when the queue space is freed up.
The core’s periodic request queue depth is critical to perform periodic transfers as
scheduled. Select the periodic queue depth, based on the number of periodic transfers
scheduled in a micro-frame. In Slave mode, however, the application must also take
into account the disable entry that must be put into the queue. So, if there are two
nonhigh-bandwidth periodic endpoints, the periodic request queue depth must be at
least 4. If at least one high-bandwidth endpoint is supported, the queue depth must be

RM0090 Rev 17

1509/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

8. If the periodic request queue depth is smaller than the periodic transfers scheduled
in a micro-frame, a frame overrun condition occurs.
•

Handling babble conditions
OTG_HS controller handles two cases of babble: packet babble and port babble.
Packet babble occurs if the device sends more data than the maximum packet size for
the channel. Port babble occurs if the core continues to receive data from the device at
EOF2 (the end of frame 2, which is very close to SOF).
When OTG_HS controller detects a packet babble, it stops writing data into the Rx
buffer and waits for the end of packet (EOP). When it detects an EOP, it flushes already
written data in the Rx buffer and generates a Babble interrupt to the application.
When OTG_HS controller detects a port babble, it flushes the RxFIFO and disables the
port. The core then generates a Port disabled interrupt (HPRTINT in
OTG_HS_GINTSTS, PENCHNG in OTG_HS_HPRT). On receiving this interrupt, the
application must determine that this is not due to an overcurrent condition (another
cause of the Port Disabled interrupt) by checking POCA in OTG_HS_HPRT, then
perform a soft reset. The core does not send any more tokens after it has detected a
port babble condition.

•

Bulk and control OUT/SETUP transactions in DMA mode
The sequence of operations is as follows:

•

a)

Initialize and enable channel 1 as explained in Section : Channel initialization.

b)

The HS_OTG host starts fetching the first packet as soon as the channel is
enabled. For internal DMA mode, the OTG_HS host uses the programmed DMA
address to fetch the packet.

c)

After fetching the last word of the second (last) packet, the OTG_HS host masks
channel 1 internally for further arbitration.

d)

The HS_OTG host generates a CHH interrupt as soon as the last packet is sent.

e)

In response to the CHH interrupt, de-allocate the channel for other transfers.

NAK and NYET handling with internal DMA
a)

The OTG_HS host sends a bulk OUT transaction.

b)

The device responds with NAK or NYET.

c)

If the application has unmasked NAK or NYET, the core generates the
corresponding interrupt(s) to the application. The application is not required to
service these interrupts, since the core takes care of rewinding the buffer pointers
and re-initializing the Channel without application intervention.

d)

The core automatically issues a ping token.

e)

When the device returns an ACK, the core continues with the transfer. Optionally,
the application can utilize these interrupts, in which case the NAK or NYET
interrupt is masked by the application.

The core does not generate a separate interrupt when NAK or NYET is received by the
host functionality.
•

Bulk and control IN transactions in DMA mode
The sequence of operations is as follows:

1510/1747

a)

Initialize and enable the used channel (channel x) as explained in Section :
Channel initialization.

b)

The OTG_HS host writes an IN request to the request queue as soon as the
channel receives the grant from the arbiter (arbitration is performed in a roundrobin fashion).
RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

•

•

c)

The OTG_HS host starts writing the received data to the system memory as soon
as the last byte is received with no errors.

d)

When the last packet is received, the OTG_HS host sets an internal flag to remove
any extra IN requests from the request queue.

e)

The OTG_HS host flushes the extra requests.

f)

The final request to disable channel x is written to the request queue. At this point,
channel 2 is internally masked for further arbitration.

g)

The OTG_HS host generates the CHH interrupt as soon as the disable request
comes to the top of the queue.

h)

In response to the CHH interrupt, de-allocate the channel for other transfers.

Interrupt OUT transactions in DMA mode
a)

Initialize and enable channel x as explained in Section : Channel initialization.

b)

The OTG_HS host starts fetching the first packet as soon the channel is enabled
and writes the OUT request along with the last word fetch. In high-bandwidth
transfers, the HS_OTG host continues fetching the next packet (up to the value
specified in the MC field) before switching to the next channel.

c)

The OTG_HS host attempts to send the OUT token at the beginning of the next
odd frame/micro-frame.

d)

After successfully transmitting the packet, the OTG_HS host generates a CHH
interrupt.

e)

In response to the CHH interrupt, reinitialize the channel for the next transfer.

Interrupt IN transactions in DMA mode
The sequence of operations (channelx) is as follows:

•

•

a)

Initialize and enable channel x as explained in Section : Channel initialization.

b)

The OTG_HS host writes an IN request to the request queue as soon as the
channel x gets the grant from the arbiter (round-robin with fairness). In highbandwidth transfers, the OTG_HS host writes consecutive writes up to MC times.

c)

The OTG_HS host attempts to send an IN token at the beginning of the next (odd)
frame/micro-frame.

d)

As soon the packet is received and written to the receive FIFO, the OTG_HS host
generates a CHH interrupt.

e)

In response to the CHH interrupt, reinitialize the channel for the next transfer.

Isochronous OUT transactions in DMA mode
a)

Initialize and enable channel x as explained in Section : Channel initialization.

b)

The OTG_HS host starts fetching the first packet as soon as the channel is
enabled, and writes the OUT request along with the last word fetch. In highbandwidth transfers, the OTG_HS host continues fetching the next packet (up to
the value specified in the MC field) before switching to the next channel.

c)

The OTG_HS host attempts to send an OUT token at the beginning of the next
(odd) frame/micro-frame.

d)

After successfully transmitting the packet, the HS_OTG host generates a CHH
interrupt.

e)

In response to the CHH interrupt, reinitialize the channel for the next transfer.

Isochronous IN transactions in DMA mode
The sequence of operations ((channel x) is as follows:

RM0090 Rev 17

1511/1747
1541

USB on-the-go high-speed (OTG_HS)

•

RM0090

a)

Initialize and enable channel x as explained in Section : Channel initialization.

b)

The OTG_HS host writes an IN request to the request queue as soon as the
channel x gets the grant from the arbiter (round-robin with fairness). In highbandwidth transfers, the OTG_HS host performs consecutive write operations up
to MC times.

c)

The OTG_HS host attempts to send an IN token at the beginning of the next (odd)
frame/micro-frame.

d)

As soon the packet is received and written to the receive FIFO, the OTG_HS host
generates a CHH interrupt.

e)

In response to the CHH interrupt, reinitialize the channel for the next transfer.

Bulk and control OUT/SETUP split transactions in DMA mode
The sequence of operations in (channel x) is as follows:

•

a)

Initialize and enable channel x for start split as explained in Section : Channel
initialization.

b)

The OTG_HS host starts fetching the first packet as soon the channel is enabled
and writes the OUT request along with the last word fetch.

c)

After successfully transmitting start split, the OTG_HS host generates the CHH
interrupt.

d)

In response to the CHH interrupt, set the COMPLSPLT bit in HCSPLT1 to send the
complete split.

e)

After successfully transmitting complete split, the OTG_HS host generates the
CHH interrupt.

f)

In response to the CHH interrupt, de-allocate the channel.

Bulk/Control IN split transactions in DMA mode
The sequence of operations (channel x) is as follows:

•

a)

Initialize and enable channel x as explained in Section : Channel initialization.

b)

The OTG_HS host writes the start split request to the nonperiodic request after
getting the grant from the arbiter. The OTG_HS host masks the channel x
internally for the arbitration after writing the request.

c)

As soon as the IN token is transmitted, the OTG_HS host generates the CHH
interrupt.

d)

In response to the CHH interrupt, set the COMPLSPLT bit in HCSPLT2 and reenable the channel to send the complete split token. This unmasks channel x for
arbitration.

e)

The OTG_HS host writes the complete split request to the nonperiodic request
after receiving the grant from the arbiter.

f)

The OTG_HS host starts writing the packet to the system memory after receiving
the packet successfully.

g)

As soon as the received packet is written to the system memory, the OTG_HS
host generates a CHH interrupt.

h)

In response to the CHH interrupt, de-allocate the channel.

Interrupt OUT split transactions in DMA mode
The sequence of operations in (channel x) is as follows:

1512/1747

a)

Initialize and enable channel 1 for start split as explained in Section : Channel
initialization. The application must set the ODDFRM bit in HCCHAR1.

b)

The HS_OTG host starts reading the packet.
RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

•

c)

The HS_OTG host attempts to send the start split transaction.

d)

After successfully transmitting the start split, the OTG_HS host generates the
CHH interrupt.

e)

In response to the CHH interrupt, set the COMPLSPLT bit in HCSPLT1 to send the
complete split.

f)

After successfully completing the complete split transaction, the OTG_HS host
generates the CHH interrupt.

g)

In response to CHH interrupt, de-allocate the channel.

Interrupt IN split transactions in DMA mode
The sequence of operations in (channel x) is as follows:

•

a)

Initialize and enable channel x for start split as explained in Section : Channel
initialization.

b)

The OTG_HS host writes an IN request to the request queue as soon as channel
x receives the grant from the arbiter.

c)

The OTG_HS host attempts to send the start split IN token at the beginning of the
next odd micro-frame.

d)

The OTG_HS host generates the CHH interrupt after successfully transmitting the
start split IN token.

e)

In response to the CHH interrupt, set the COMPLSPLT bit in HCSPLT2 to send the
complete split.

f)

As soon as the packet is received successfully, the OTG_HS host starts writing the
data to the system memory.

g)

The OTG_HS host generates the CHH interrupt after transferring the received
data to the system memory.

h)

In response to the CHH interrupt, de-allocate or reinitialize the channel for the next
start split.

Isochronous OUT split transactions in DMA mode
The sequence of operations (channel x) is as follows:

•

a)

Initialize and enable channel x for start split (begin) as explained in Section :
Channel initialization. The application must set the ODDFRM bit in HCCHAR1.
Program the MPS field.

b)

The HS_OTG host starts reading the packet.

c)

After successfully transmitting the start split (begin), the HS_OTG host generates
the CHH interrupt.

d)

In response to the CHH interrupt, reinitialize the registers to send the start split
(end).

e)

After successfully transmitting the start split (end), the OTG_HS host generates a
CHH interrupt.

f)

In response to the CHH interrupt, de-allocate the channel.

Isochronous IN split transactions in DMA mode
The sequence of operations (channel x) is as follows:
a)

Initialize and enable channel x for start split as explained in Section : Channel
initialization.

b)

The OTG_HS host writes an IN request to the request queue as soon as channel
x receives the grant from the arbiter.

RM0090 Rev 17

1513/1747
1541

USB on-the-go high-speed (OTG_HS)

35.13.6

RM0090

c)

The OTG_HS host attempts to send the start split IN token at the beginning of the
next odd micro-frame.

d)

The OTG_HS host generates the CHH interrupt after successfully transmitting the
start split IN token.

e)

In response to the CHH interrupt, set the COMPLSPLT bit in HCSPLT2 to send the
complete split.

f)

As soon as the packet is received successfully, the OTG_HS host starts writing the
data to the system memory.

g)

The OTG_HS host generates the CHH interrupt after transferring the received
data to the system memory. In response to the CHH interrupt, de-allocate the
channel or reinitialize the channel for the next start split.

Device programming model
Endpoint initialization on USB reset
1.

Set the NAK bit for all OUT endpoints
–

2.

3.

4.

–

INEP0 = 1 in OTG_HS_DAINTMSK (control 0 IN endpoint)

–

OUTEP0 = 1 in OTG_HS_DAINTMSK (control 0 OUT endpoint)

–

STUP = 1 in DOEPMSK

–

XFRC = 1 in DOEPMSK

–

XFRC = 1 in DIEPMSK

–

TOC = 1 in DIEPMSK

Set up the Data FIFO RAM for each of the FIFOs
–

Program the OTG_HS_GRXFSIZ register, to be able to receive control OUT data
and setup data. If thresholding is not enabled, at a minimum, this must be equal to
1 max packet size of control endpoint 0 + 2 words (for the status of the control
OUT data packet) + 10 words (for setup packets).

–

Program the OTG_HS_TX0FSIZ register (depending on the FIFO number
chosen) to be able to transmit control IN data. At a minimum, this must be equal to
1 max packet size of control endpoint 0.

Program the following fields in the endpoint-specific registers for control OUT endpoint
0 to receive a SETUP packet
–

5.

SNAK = 1 in OTG_HS_DOEPCTLx (for all OUT endpoints)

Unmask the following interrupt bits

STUPCNT = 3 in OTG_HS_DOEPTSIZ0 (to receive up to 3 back-to-back SETUP
packets)

In DMA mode, the DOEPDMA0 register should have a valid memory address to store
any SETUP packets received.

At this point, all initialization required to receive SETUP packets is done.

1514/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

Endpoint initialization on enumeration completion
1.

On the Enumeration Done interrupt (ENUMDNE in OTG_HS_GINTSTS), read the
OTG_HS_DSTS register to determine the enumeration speed.

2.

Program the MPSIZ field in OTG_HS_DIEPCTL0 to set the maximum packet size. This
step configures control endpoint 0. The maximum packet size for a control endpoint
depends on the enumeration speed.

3.

In DMA mode, program the DOEPCTL0 register to enable control OUT endpoint 0, to
receive a SETUP packet.
–

EPENA bit in DOEPCTL0 = 1

At this point, the device is ready to receive SOF packets and is configured to perform control
transfers on control endpoint 0.

Endpoint initialization on SetAddress command
This section describes what the application must do when it receives a SetAddress
command in a SETUP packet.
1.

Program the OTG_HS_DCFG register with the device address received in the
SetAddress command

1.

Program the core to send out a status IN packet

Endpoint initialization on SetConfiguration/SetInterface command
This section describes what the application must do when it receives a SetConfiguration or
SetInterface command in a SETUP packet.
1.

When a SetConfiguration command is received, the application must program the
endpoint registers to configure them with the characteristics of the valid endpoints in
the new configuration.

2.

When a SetInterface command is received, the application must program the endpoint
registers of the endpoints affected by this command.

3.

Some endpoints that were active in the prior configuration or alternate setting are not
valid in the new configuration or alternate setting. These invalid endpoints must be
deactivated.

4.

Unmask the interrupt for each active endpoint and mask the interrupts for all inactive
endpoints in the OTG_HS_DAINTMSK register.

5.

Set up the Data FIFO RAM for each FIFO.

6.

After all required endpoints are configured; the application must program the core to
send a status IN packet.

At this point, the device core is configured to receive and transmit any type of data packet.

RM0090 Rev 17

1515/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Endpoint activation
This section describes the steps required to activate a device endpoint or to configure an
existing device endpoint to a new type.
1.

2.

Program the characteristics of the required endpoint into the following fields of the
OTG_HS_DIEPCTLx register (for IN or bidirectional endpoints) or the
OTG_HS_DOEPCTLx register (for OUT or bidirectional endpoints).
–

Maximum packet size

–

USB active endpoint = 1

–

Endpoint start data toggle (for interrupt and bulk endpoints)

–

Endpoint type

–

TxFIFO number

Once the endpoint is activated, the core starts decoding the tokens addressed to that
endpoint and sends out a valid handshake for each valid token received for the
endpoint.

Endpoint deactivation
This section describes the steps required to deactivate an existing endpoint.
1.

In the endpoint to be deactivated, clear the USB active endpoint bit in the
OTG_HS_DIEPCTLx register (for IN or bidirectional endpoints) or the
OTG_HS_DOEPCTLx register (for OUT or bidirectional endpoints).

2.

Once the endpoint is deactivated, the core ignores tokens addressed to that endpoint,
which results in a timeout on the USB.

Note:

The application must meet the following conditions to set up the device core to handle
traffic:
NPTXFEM and RXFLVLM in GINTMSK must be cleared.

35.13.7

Operational model
SETUP and OUT data transfers
This section describes the internal data flow and application-level operations during data
OUT transfers and SETUP transactions.
•

Packet read

This section describes how to read packets (OUT data and SETUP packets) from the
receive FIFO in Slave mode.

1516/1747

1.

On catching an RXFLVL interrupt (OTG_HS_GINTSTS register), the application must
read the Receive status pop register (OTG_HS_GRXSTSP).

2.

The application can mask the RXFLVL interrupt (in OTG_HS_GINTSTS) by writing to
RXFLVL = 0 (in GINTMSK), until it has read the packet from the receive FIFO.

3.

If the received packet’s byte count is not 0, the byte count amount of data is popped
from the receive Data FIFO and stored in memory. If the received packet byte count is
0, no data is popped from the receive data FIFO.

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
4.

The receive FIFO packet status readout indicates one of the following:
a)

Global OUT NAK pattern:
PKTSTS = Global OUT NAK, BCNT = 0x000, EPNUM = Don’t Care (0x0),
DPID = Don’t Care (0b00).
These data indicate that the global OUT NAK bit has taken effect.

b)

SETUP packet pattern:
PKTSTS = SETUP, BCNT = 0x008, EPNUM = Control EP Num, DPID = D0.
These data indicate that a SETUP packet for the specified endpoint is now
available for reading from the receive FIFO.

c)

Setup stage done pattern:
PKTSTS = Setup Stage Done, BCNT = 0x0, EPNUM = Control EP Num,
DPID = Don’t Care (0b00).
These data indicate that the Setup stage for the specified endpoint has completed
and the Data stage has started. After this entry is popped from the receive FIFO,
the core asserts a Setup interrupt on the specified control OUT endpoint.

d)

Data OUT packet pattern:
PKTSTS = DataOUT, BCNT = size of the received data OUT packet (0 ≤ BCNT
≤ 1 024), EPNUM = EPNUM on which the packet was received, DPID = Actual
Data PID.

e)

Data transfer completed pattern:
PKTSTS = Data OUT Transfer Done, BCNT = 0x0, EPNUM = OUT EP Num
on which the data transfer is complete, DPID = Don’t Care (0b00).
These data indicate that an OUT data transfer for the specified OUT endpoint has
completed. After this entry is popped from the receive FIFO, the core asserts a
Transfer Completed interrupt on the specified OUT endpoint.

5.

After the data payload is popped from the receive FIFO, the RXFLVL interrupt
(OTG_HS_GINTSTS) must be unmasked.

6.

Steps 1–5 are repeated every time the application detects assertion of the interrupt line
due to RXFLVL in OTG_HS_GINTSTS. Reading an empty receive FIFO can result in
undefined core behavior.

Figure 426 provides a flowchart of the above procedure.

RM0090 Rev 17

1517/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Figure 426. Receive FIFO packet read in slave mode

wait until RXFLVL in OTG_FS_GINTSTSG

rd_data = rd_reg (OTG_FS_GRXSTSP);

Y

rd_data.BCNT = 0

rcv_out_pkt ()

N

packet
store in
memory

mem[0:dword_cnt-1] =
rd_rxfifo(rd_data.EPNUM,
dword_cnt )

dword_cnt =
BCNT[11:2]
C
+
(BCNT[1] | BCNT[1])

ai15677

•

SETUP transactions

This section describes how the core handles SETUP packets and the application’s
sequence for handling SETUP transactions.
•

Application requirements

1.

To receive a SETUP packet, the STUPCNT field (OTG_HS_DOEPTSIZx) in a control
OUT endpoint must be programmed to a nonzero value. When the application
programs the STUPCNT field to a nonzero value, the core receives SETUP packets
and writes them to the receive FIFO, irrespective of the NAK status and EPENA bit
setting in OTG_HS_DOEPCTLx. The STUPCNT field is decremented every time the
control endpoint receives a SETUP packet. If the STUPCNT field is not programmed to
a proper value before receiving a SETUP packet, the core still receives the SETUP
packet and decrements the STUPCNT field, but the application may not be able to
determine the correct number of SETUP packets received in the Setup stage of a
control transfer.
–

2.

1518/1747

STUPCNT = 3 in OTG_HS_DOEPTSIZx

The application must always allocate some extra space in the Receive data FIFO, to be
able to receive up to three SETUP packets on a control endpoint.
–

The space to be reserved is 10 words. Three words are required for the first
SETUP packet, 1 word is required for the Setup stage done word and 6 words are
required to store two extra SETUP packets among all control endpoints.

–

3 words per SETUP packet are required to store 8 bytes of SETUP data and 4
bytes of SETUP status (Setup packet pattern). The core reserves this space in the
receive data.

–

FIFO to write SETUP data only, and never uses this space for data packets.

3.

The application must read the 2 words of the SETUP packet from the receive FIFO.

4.

The application must read and discard the Setup stage done word from the receive
FIFO.

•

Internal data flow

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
5.

When a SETUP packet is received, the core writes the received data to the receive
FIFO, without checking for available space in the receive FIFO and irrespective of the
endpoint’s NAK and STALL bit settings.
–

6.

7.

The core internally sets the IN NAK and OUT NAK bits for the control IN/OUT
endpoints on which the SETUP packet was received.

For every SETUP packet received on the USB, 3 words of data are written to the
receive FIFO, and the STUPCNT field is decremented by 1.
–

The first word contains control information used internally by the core

–

The second word contains the first 4 bytes of the SETUP command

–

The third word contains the last 4 bytes of the SETUP command

When the Setup stage changes to a Data IN/OUT stage, the core writes an entry
(Setup stage done word) to the receive FIFO, indicating the completion of the Setup
stage.

8.

On the AHB side, SETUP packets are emptied by the application.

9.

When the application pops the Setup stage done word from the receive FIFO, the core
interrupts the application with an STUP interrupt (OTG_HS_DOEPINTx), indicating it
can process the received SETUP packet.
–

•
1.

The core clears the endpoint enable bit for control OUT endpoints.

Application programming sequence
Program the OTG_HS_DOEPTSIZx register.
–

STUPCNT = 3

2.

Wait for the RXFLVL interrupt (OTG_HS_GINTSTS) and empty the data packets from
the receive FIFO.

3.

Assertion of the STUP interrupt (OTG_HS_DOEPINTx) marks a successful completion
of the SETUP Data Transfer.
–

On this interrupt, the application must read the OTG_HS_DOEPTSIZx register to
determine the number of SETUP packets received and process the last received
SETUP packet.

RM0090 Rev 17

1519/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Figure 427. Processing a SETUP packet
Wait for STUP in OTG_FS_DOEPINTx

rem_supcnt =
rd_reg(DOEPTSIZx)

setup_cmd[31:0] = mem[4 – 2 * rem_supcnt]
setup_cmd[63:32] = mem[5 – 2 * rem_supcnt]

Find setup cmd type

Read

ctrl-rd/wr/2 stage

Write

2-stage
setup_np_in_pkt
Data IN phase

setup_np_in_pkt
Status IN phase

rcv_out_pkt
Data OUT phase
ai15678

•

Handling more than three back-to-back SETUP packets

Per the USB 2.0 specification, normally, during a SETUP packet error, a host does not send
more than three back-to-back SETUP packets to the same endpoint. However, the USB 2.0
specification does not limit the number of back-to-back SETUP packets a host can send to
the same endpoint. When this condition occurs, the OTG_HS controller generates an
interrupt (B2BSTUP in OTG_HS_DOEPINTx).
•

Setting the global OUT NAK

Internal data flow:

1520/1747

1.

When the application sets the Global OUT NAK (SGONAK bit in OTG_HS_DCTL), the
core stops writing data, except SETUP packets, to the receive FIFO. Irrespective of the
space availability in the receive FIFO, nonisochronous OUT tokens receive a NAK
handshake response, and the core ignores isochronous OUT data packets

2.

The core writes the Global OUT NAK pattern to the receive FIFO. The application must
reserve enough receive FIFO space to write this data pattern.

3.

When the application pops the Global OUT NAK pattern word from the receive FIFO,
the core sets the GONAKEFF interrupt (OTG_HS_GINTSTS).

4.

Once the application detects this interrupt, it can assume that the core is in Global OUT
NAK mode. The application can clear this interrupt by clearing the SGONAK bit in
OTG_HS_DCTL.

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
Application programming sequence:
1.

To stop receiving any kind of data in the receive FIFO, the application must set the
Global OUT NAK bit by programming the following field:
–

SGONAK = 1 in OTG_HS_DCTL

2.

Wait for the assertion of the GONAKEFF interrupt in OTG_HS_GINTSTS. When
asserted, this interrupt indicates that the core has stopped receiving any type of data
except SETUP packets.

3.

The application can receive valid OUT packets after it has set SGONAK in
OTG_HS_DCTL and before the core asserts the GONAKEFF interrupt
(OTG_HS_GINTSTS).

4.

The application can temporarily mask this interrupt by writing to the GINAKEFFM bit in
GINTMSK.
–

5.

Whenever the application is ready to exit the Global OUT NAK mode, it must clear the
SGONAK bit in OTG_HS_DCTL. This also clears the GONAKEFF interrupt
(OTG_HS_GINTSTS).
–

6.

OTG_HS_DCTL = 1 in CGONAK

If the application has masked this interrupt earlier, it must be unmasked as follows:
–

•

GINAKEFFM = 0 in GINTMSK

GINAKEFFM = 1 in GINTMSK

Disabling an OUT endpoint

The application must use this sequence to disable an OUT endpoint that it has enabled.
Application programming sequence:
1.

Before disabling any OUT endpoint, the application must enable Global OUT NAK
mode in the core.
–

2.
3.

4.

5.

SGONAK = 1 in OTG_HS_DCTL

Wait for the GONAKEFF interrupt (OTG_HS_GINTSTS)
Disable the required OUT endpoint by programming the following fields:
–

EPDIS = 1 in OTG_HS_DOEPCTLx

–

SNAK = 1 in OTG_HS_DOEPCTLx

Wait for the EPDISD interrupt (OTG_HS_DOEPINTx), which indicates that the OUT
endpoint is completely disabled. When the EPDISD interrupt is asserted, the core also
clears the following bits:
–

EPDIS = 0 in OTG_HS_DOEPCTLx

–

EPENA = 0 in OTG_HS_DOEPCTLx

The application must clear the Global OUT NAK bit to start receiving data from other
nondisabled OUT endpoints.
–

SGONAK = 0 in OTG_HS_DCTL

RM0090 Rev 17

1521/1747
1541

USB on-the-go high-speed (OTG_HS)
•

RM0090

Generic non-isochronous OUT data transfers

This section describes a regular nonisochronous OUT data transfer (control, bulk, or
interrupt).
Application requirements:
1.

Before setting up an OUT transfer, the application must allocate a buffer in the memory
to accommodate all data to be received as part of the OUT transfer.

2.

For OUT transfers, the transfer size field in the endpoint’s transfer size register must be
a multiple of the maximum packet size of the endpoint, adjusted to the word boundary.

3.

–

transfer size[EPNUM] = n × (MPSIZ[EPNUM] + 4 – (MPSIZ[EPNUM] mod 4))

–

packet count[EPNUM] = n

–

n>0

On any OUT endpoint interrupt, the application must read the endpoint’s transfer size
register to calculate the size of the payload in the memory. The received payload size
can be less than the programmed transfer size.
–

Payload size in memory = application programmed initial transfer size – core
updated final transfer size

–

Number of USB packets in which this payload was received = application
programmed initial packet count – core updated final packet count

Internal data flow:

1522/1747

1.

The application must set the transfer size and packet count fields in the endpointspecific registers, clear the NAK bit, and enable the endpoint to receive the data.

2.

Once the NAK bit is cleared, the core starts receiving data and writes it to the receive
FIFO, as long as there is space in the receive FIFO. For every data packet received on
the USB, the data packet and its status are written to the receive FIFO. Every packet
(maximum packet size or short packet) written to the receive FIFO decrements the
packet count field for that endpoint by 1.
–

OUT data packets received with bad data CRC are flushed from the receive FIFO
automatically.

–

After sending an ACK for the packet on the USB, the core discards
nonisochronous OUT data packets that the host, which cannot detect the ACK, resends. The application does not detect multiple back-to-back data OUT packets
on the same endpoint with the same data PID. In this case the packet count is not
decremented.

–

If there is no space in the receive FIFO, isochronous or nonisochronous data
packets are ignored and not written to the receive FIFO. Additionally,
nonisochronous OUT tokens receive a NAK handshake reply.

–

In all the above three cases, the packet count is not decremented because no data
are written to the receive FIFO.

3.

When the packet count becomes 0 or when a short packet is received on the endpoint,
the NAK bit for that endpoint is set. Once the NAK bit is set, the isochronous or
nonisochronous data packets are ignored and not written to the receive FIFO, and
nonisochronous OUT tokens receive a NAK handshake reply.

4.

After the data are written to the receive FIFO, the application reads the data from the
receive FIFO and writes it to external memory, one packet at a time per endpoint.

5.

At the end of every packet write on the AHB to external memory, the transfer size for
the endpoint is decremented by the size of the written packet.

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
6.

7.

The OUT data transfer completed pattern for an OUT endpoint is written to the receive
FIFO on one of the following conditions:
–

The transfer size is 0 and the packet count is 0

–

The last OUT data packet written to the receive FIFO is a short packet
(0 ≤ packet size < maximum packet size)

When either the application pops this entry (OUT data transfer completed), a transfer
completed interrupt is generated for the endpoint and the endpoint enable is cleared.

Application programming sequence:
1.

Program the OTG_HS_DOEPTSIZx register for the transfer size and the
corresponding packet count.

2.

Program the OTG_HS_DOEPCTLx register with the endpoint characteristics, and set
the EPENA and CNAK bits.

3.

–

EPENA = 1 in OTG_HS_DOEPCTLx

–

CNAK = 1 in OTG_HS_DOEPCTLx

Wait for the RXFLVL interrupt (in OTG_HS_GINTSTS) and empty the data packets
from the receive FIFO.
–

This step can be repeated many times, depending on the transfer size.

4.

Asserting the XFRC interrupt (OTG_HS_DOEPINTx) marks a successful completion of
the nonisochronous OUT data transfer.

5.

Read the OTG_HS_DOEPTSIZx register to determine the size of the received data
payload.

•

Generic isochronous OUT data transfer

This section describes a regular isochronous OUT data transfer.
Application requirements:
1.

All the application requirements for nonisochronous OUT data transfers also apply to
isochronous OUT data transfers.

2.

For isochronous OUT data transfers, the transfer size and packet count fields must
always be set to the number of maximum-packet-size packets that can be received in a
single frame and no more. Isochronous OUT data transfers cannot span more than 1
frame.

3.

The application must read all isochronous OUT data packets from the receive FIFO
(data and status) before the end of the periodic frame (EOPF interrupt in
OTG_HS_GINTSTS).

4.

To receive data in the following frame, an isochronous OUT endpoint must be enabled
after the EOPF (OTG_HS_GINTSTS) and before the SOF (OTG_HS_GINTSTS).

Internal data flow:
1.

The internal data flow for isochronous OUT endpoints is the same as that for
nonisochronous OUT endpoints, but for a few differences.

2.

When an isochronous OUT endpoint is enabled by setting the Endpoint Enable and
clearing the NAK bits, the Even/Odd frame bit must also be set appropriately. The core
receives data on an isochronous OUT endpoint in a particular frame only if the
following condition is met:
–

3.

EONUM (in OTG_HS_DOEPCTLx) = SOFFN[0] (in OTG_HS_DSTS)

When the application completely reads an isochronous OUT data packet (data and
status) from the receive FIFO, the core updates the RXDPID field in

RM0090 Rev 17

1523/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

OTG_HS_DOEPTSIZx with the data PID of the last isochronous OUT data packet read
from the receive FIFO.
Application programming sequence:
1.

Program the OTG_HS_DOEPTSIZx register for the transfer size and the
corresponding packet count

2.

Program the OTG_HS_DOEPCTLx register with the endpoint characteristics and set
the Endpoint Enable, ClearNAK, and Even/Odd frame bits.

3.

–

EPENA = 1

–

CNAK = 1

–

EONUM = (0: Even/1: Odd)

In Slave mode, wait for the RXFLVL interrupt (in OTG_HS_GINTSTS) and empty the
data packets from the receive FIFO
–

This step can be repeated many times, depending on the transfer size.

4.

The assertion of the XFRC interrupt (in OTG_HS_DOEPINTx) marks the completion of
the isochronous OUT data transfer. This interrupt does not necessarily mean that the
data in memory are good.

5.

This interrupt cannot always be detected for isochronous OUT transfers. Instead, the
application can detect the IISOOXFRM interrupt in OTG_HS_GINTSTS.

6.

Read the OTG_HS_DOEPTSIZx register to determine the size of the received transfer
and to determine the validity of the data received in the frame. The application must
treat the data received in memory as valid only if one of the following conditions is met:
–

RXDPID = D0 (in OTG_HS_DOEPTSIZx) and the number of USB packets in
which this payload was received = 1

–

RXDPID = D1 (in OTG_HS_DOEPTSIZx) and the number of USB packets in
which this payload was received = 2

–

RXDPID = D2 (in OTG_HS_DOEPTSIZx) and the number of USB packets in
which this payload was received = 3
The number of USB packets in which this payload was received =
Application programmed initial packet count – Core updated final packet count

The application can discard invalid data packets.
•

Incomplete isochronous OUT data transfers

This section describes the application programming sequence when isochronous OUT data
packets are dropped inside the core.
Internal data flow:
1.

2.

1524/1747

For isochronous OUT endpoints, the XFRC interrupt (in OTG_HS_DOEPINTx) may not
always be asserted. If the core drops isochronous OUT data packets, the application
could fail to detect the XFRC interrupt (OTG_HS_DOEPINTx) under the following
circumstances:
–

When the receive FIFO cannot accommodate the complete ISO OUT data packet,
the core drops the received ISO OUT data

–

When the isochronous OUT data packet is received with CRC errors

–

When the isochronous OUT token received by the core is corrupted

–

When the application is very slow in reading the data from the receive FIFO

When the core detects an end of periodic frame before transfer completion to all
isochronous OUT endpoints, it asserts the incomplete Isochronous OUT data interrupt

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
(IISOOXFRM in OTG_HS_GINTSTS), indicating that an XFRC interrupt (in
OTG_HS_DOEPINTx) is not asserted on at least one of the isochronous OUT
endpoints. At this point, the endpoint with the incomplete transfer remains enabled, but
no active transfers remain in progress on this endpoint on the USB.
Application programming sequence:
1.

Asserting the IISOOXFRM interrupt (OTG_HS_GINTSTS) indicates that in the current
frame, at least one isochronous OUT endpoint has an incomplete transfer.

2.

If this occurs because isochronous OUT data is not completely emptied from the
endpoint, the application must ensure that the application empties all isochronous OUT
data (data and status) from the receive FIFO before proceeding.
–

3.

When all data are emptied from the receive FIFO, the application can detect the
XFRC interrupt (OTG_HS_DOEPINTx). In this case, the application must reenable the endpoint to receive isochronous OUT data in the next frame.

When it receives an IISOOXFRM interrupt (in OTG_HS_GINTSTS), the application
must read the control registers of all isochronous OUT endpoints
(OTG_HS_DOEPCTLx) to determine which endpoints had an incomplete transfer in
the current micro-frame. An endpoint transfer is incomplete if both the following
conditions are met:
–

EONUM bit (in OTG_HS_DOEPCTLx) = SOFFN[0] (in OTG_HS_DSTS)

–

EPENA = 1 (in OTG_HS_DOEPCTLx)

4.

The previous step must be performed before the SOF interrupt (in
OTG_HS_GINTSTS) is detected, to ensure that the current frame number is not
changed.

5.

For isochronous OUT endpoints with incomplete transfers, the application must discard
the data in the memory and disable the endpoint by setting the EPDIS bit in
OTG_HS_DOEPCTLx.

6.

Wait for the EPDIS interrupt (in OTG_HS_DOEPINTx) and enable the endpoint to
receive new data in the next frame.
–

•

Because the core can take some time to disable the endpoint, the application may
not be able to receive the data in the next frame after receiving bad isochronous
data.

Stalling a nonisochronous OUT endpoint

This section describes how the application can stall a nonisochronous endpoint.
1.

Put the core in the Global OUT NAK mode.

2.

Disable the required endpoint
–

When disabling the endpoint, instead of setting the SNAK bit in
OTG_HS_DOEPCTL, set STALL = 1 (in OTG_HS_DOEPCTL).
The STALL bit always takes precedence over the NAK bit.

3.

When the application is ready to end the STALL handshake for the endpoint, the
STALL bit (in OTG_HS_DOEPCTLx) must be cleared.

4.

If the application is setting or clearing a STALL for an endpoint due to a
SetFeature.Endpoint Halt or ClearFeature.Endpoint Halt command, the STALL bit must
be set or cleared before the application sets up the Status stage transfer on the control
endpoint.

RM0090 Rev 17

1525/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

Examples
This section describes and depicts some fundamental transfer types and scenarios.
•

Slave mode bulk OUT transaction

Figure 428 depicts the reception of a single Bulk OUT Data packet from the USB to the AHB
and describes the events involved in the process.
Figure 428. Slave mode bulk OUT transaction
Host

USB

Application

Device

init_ out_ ep
1

2

XFRSIZ = 512 bytes
PKTCNT = 1

wr_reg (DOEPTSIZx)

O UT

wr_reg(D OEPCTLx)

3

EPENA= 1
CNAK = 1

512 bytes
4
AC K

5

6

xact _1
RXFLVL iintr
T L x.N A
K
=
1
PKTCN
T0

D OE P C

XFRSIZ
=0
r
OU T
NA K

7

idle until intr

rcv_out _pkt()

XF
int r RC

8

On new xfer
or RxFIFO
not empty

idle until intr

ai15679

After a SetConfiguration/SetInterface command, the application initializes all OUT endpoints
by setting CNAK = 1 and EPENA = 1 (in OTG_HS_DOEPCTLx), and setting a suitable
XFRSIZ and PKTCNT in the OTG_HS_DOEPTSIZx register.

1526/1747

1.

Host attempts to send data (OUT token) to an endpoint.

2.

When the core receives the OUT token on the USB, it stores the packet in the RxFIFO
because space is available there.

3.

After writing the complete packet in the RxFIFO, the core then asserts the RXFLVL
interrupt (in OTG_HS_GINTSTS).

4.

On receiving the PKTCNT number of USB packets, the core internally sets the NAK bit
for this endpoint to prevent it from receiving any more packets.

5.

The application processes the interrupt and reads the data from the RxFIFO.

6.

When the application has read all the data (equivalent to XFRSIZ), the core generates
an XFRC interrupt (in OTG_HS_DOEPINTx).

7.

The application processes the interrupt and uses the setting of the XFRC interrupt bit
(in OTG_HS_DOEPINTx) to determine that the intended transfer is complete.

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)

IN data transfers
•

Packet write

This section describes how the application writes data packets to the endpoint FIFO in Slave
mode when dedicated transmit FIFOs are enabled.
1.

2.

The application can either choose the polling or the interrupt mode.
–

In polling mode, the application monitors the status of the endpoint transmit data
FIFO by reading the OTG_HS_DTXFSTSx register, to determine if there is
enough space in the data FIFO.

–

In interrupt mode, the application waits for the TXFE interrupt (in
OTG_HS_DIEPINTx) and then reads the OTG_HS_DTXFSTSx register, to
determine if there is enough space in the data FIFO.

–

To write a single nonzero length data packet, there must be space to write the
entire packet in the data FIFO.

–

To write zero length packet, the application must not look at the FIFO space.

Using one of the above mentioned methods, when the application determines that
there is enough space to write a transmit packet, the application must first write into the
endpoint control register, before writing the data into the data FIFO. Typically, the
application, must do a read modify write on the OTG_HS_DIEPCTLx register to avoid
modifying the contents of the register, except for setting the Endpoint Enable bit.

The application can write multiple packets for the same endpoint into the transmit FIFO, if
space is available. For periodic IN endpoints, the application must write packets only for one
micro-frame. It can write packets for the next periodic transaction only after getting transfer
complete for the previous transaction.
•

Setting IN endpoint NAK

Internal data flow:
1.

2.

When the application sets the IN NAK for a particular endpoint, the core stops
transmitting data on the endpoint, irrespective of data availability in the endpoint’s
transmit FIFO.
Nonisochronous IN tokens receive a NAK handshake reply
–

Isochronous IN tokens receive a zero-data-length packet reply

3.

The core asserts the INEPNE (IN endpoint NAK effective) interrupt in
OTG_HS_DIEPINTx in response to the SNAK bit in OTG_HS_DIEPCTLx.

4.

Once this interrupt is seen by the application, the application can assume that the
endpoint is in IN NAK mode. This interrupt can be cleared by the application by setting
the CNAK bit in OTG_HS_DIEPCTLx.

Application programming sequence:

RM0090 Rev 17

1527/1747
1541

USB on-the-go high-speed (OTG_HS)
1.

RM0090

To stop transmitting any data on a particular IN endpoint, the application must set the
IN NAK bit. To set this bit, the following field must be programmed.
–

SNAK = 1 in OTG_HS_DIEPCTLx

2.

Wait for assertion of the INEPNE interrupt in OTG_HS_DIEPINTx. This interrupt
indicates that the core has stopped transmitting data on the endpoint.

3.

The core can transmit valid IN data on the endpoint after the application has set the
NAK bit, but before the assertion of the NAK Effective interrupt.

4.

The application can mask this interrupt temporarily by writing to the INEPNEM bit in
DIEPMSK.
–

5.

To exit Endpoint NAK mode, the application must clear the NAK status bit (NAKSTS) in
OTG_HS_DIEPCTLx. This also clears the INEPNE interrupt (in OTG_HS_DIEPINTx).
–

6.

CNAK = 1 in OTG_HS_DIEPCTLx

If the application masked this interrupt earlier, it must be unmasked as follows:
–

•

INEPNEM = 0 in DIEPMSK

INEPNEM = 1 in DIEPMSK

IN endpoint disable

Use the following sequence to disable a specific IN endpoint that has been previously
enabled.
Application programming sequence:
1.

The application must stop writing data on the AHB for the IN endpoint to be disabled.

2.

The application must set the endpoint in NAK mode.
–

SNAK = 1 in OTG_HS_DIEPCTLx

3.

Wait for the INEPNE interrupt in OTG_HS_DIEPINTx.

4.

Set the following bits in the OTG_HS_DIEPCTLx register for the endpoint that must be
disabled.

5.

–

EPDIS = 1 in OTG_HS_DIEPCTLx

–

SNAK = 1 in OTG_HS_DIEPCTLx

Assertion of the EPDISD interrupt in OTG_HS_DIEPINTx indicates that the core has
completely disabled the specified endpoint. Along with the assertion of the interrupt, the
core also clears the following bits:
–

EPENA = 0 in OTG_HS_DIEPCTLx

–

EPDIS = 0 in OTG_HS_DIEPCTLx

6.

The application must read the OTG_HS_DIEPTSIZx register for the periodic IN EP, to
calculate how much data on the endpoint were transmitted on the USB.

7.

The application must flush the data in the Endpoint transmit FIFO, by setting the
following fields in the OTG_HS_GRSTCTL register:
–

TXFNUM (in OTG_HS_GRSTCTL) = Endpoint transmit FIFO number

–

TXFFLSH in (OTG_HS_GRSTCTL) = 1

The application must poll the OTG_HS_GRSTCTL register, until the TXFFLSH bit is cleared
by the core, which indicates the end of flush operation. To transmit new data on this
endpoint, the application can re-enable the endpoint at a later point.
•

Generic nonperiodic IN data transfers

Application requirements:

1528/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
1.

Before setting up an IN transfer, the application must ensure that all data to be
transmitted as part of the IN transfer are part of a single buffer.

2.

For IN transfers, the Transfer Size field in the Endpoint Transfer Size register denotes a
payload that constitutes multiple maximum-packet-size packets and a single short
packet. This short packet is transmitted at the end of the transfer.
–

To transmit a few maximum-packet-size packets and a short packet at the end of
the transfer:
Transfer size[EPNUM] = x × MPSIZ[EPNUM] + sp
If (sp > 0), then packet count[EPNUM] = x + 1.
Otherwise, packet count[EPNUM] = x

–

To transmit a single zero-length data packet:
Transfer size[EPNUM] = 0
Packet count[EPNUM] = 1

–

To transmit a few maximum-packet-size packets and a zero-length data packet at
the end of the transfer, the application must split the transfer into two parts. The
first sends maximum-packet-size data packets and the second sends the zerolength data packet alone.
First transfer: transfer size[EPNUM] = x × MPSIZ[epnum]; packet count = n;
Second transfer: transfer size[EPNUM] = 0; packet count = 1;

3.

Once an endpoint is enabled for data transfers, the core updates the Transfer size
register. At the end of the IN transfer, the application must read the Transfer size
register to determine how much data posted in the transmit FIFO have already been
sent on the USB.

4.

Data fetched into transmit FIFO = Application-programmed initial transfer size – coreupdated final transfer size
–

Data transmitted on USB = (application-programmed initial packet count – Core
updated final packet count) × MPSIZ[EPNUM]

–

Data yet to be transmitted on USB = (Application-programmed initial transfer size
– data transmitted on USB)

Internal data flow:
1.

The application must set the transfer size and packet count fields in the endpointspecific registers and enable the endpoint to transmit the data.

2.

The application must also write the required data to the transmit FIFO for the endpoint.

3.

Every time a packet is written into the transmit FIFO by the application, the transfer size
for that endpoint is decremented by the packet size. The data is fetched from the
memory by the application, until the transfer size for the endpoint becomes 0. After
writing the data into the FIFO, the “number of packets in FIFO” count is incremented
(this is a 3-bit count, internally maintained by the core for each IN endpoint transmit
FIFO. The maximum number of packets maintained by the core at any time in an IN
endpoint FIFO is eight). For zero-length packets, a separate flag is set for each FIFO,
without any data in the FIFO.

4.

Once the data are written to the transmit FIFO, the core reads them out upon receiving
an IN token. For every nonisochronous IN data packet transmitted with an ACK

RM0090 Rev 17

1529/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

handshake, the packet count for the endpoint is decremented by one, until the packet
count is zero. The packet count is not decremented on a timeout.
5.

For zero length packets (indicated by an internal zero length flag), the core sends out a
zero-length packet for the IN token and decrements the packet count field.

6.

If there are no data in the FIFO for a received IN token and the packet count field for
that endpoint is zero, the core generates an “IN token received when TxFIFO is empty”
(ITTXFE) Interrupt for the endpoint, provided that the endpoint NAK bit is not set. The
core responds with a NAK handshake for nonisochronous endpoints on the USB.

7.

The core internally rewinds the FIFO pointers and no timeout interrupt is generated.

8.

When the transfer size is 0 and the packet count is 0, the transfer complete (XFRC)
interrupt for the endpoint is generated and the endpoint enable is cleared.

Application programming sequence:
1.

Program the OTG_HS_DIEPTSIZx register with the transfer size and corresponding
packet count.

2.

Program the OTG_HS_DIEPCTLx register with the endpoint characteristics and set the
CNAK and EPENA (Endpoint Enable) bits.

3.

When transmitting nonzero length data packet, the application must poll the
OTG_HS_DTXFSTSx register (where x is the FIFO number associated with that
endpoint) to determine whether there is enough space in the data FIFO. The
application can optionally use TXFE (in OTG_HS_DIEPINTx) before writing the data.

•

Generic periodic IN data transfers

This section describes a typical periodic IN data transfer.
Application requirements:
1.

Application requirements 1, 2, 3, and 4 of Generic nonperiodic IN data transfers on
page 1528 also apply to periodic IN data transfers, except for a slight modification of
requirement 2.
–

1530/1747

The application can only transmit multiples of maximum-packet-size data packets
or multiples of maximum-packet-size packets, plus a short packet at the end. To

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
transmit a few maximum-packet-size packets and a short packet at the end of the
transfer, the following conditions must be met:
transfer size[EPNUM] = x × MPSIZ[EPNUM] + sp
(where x is an integer ≥ 0, and 0 ≤ sp < MPSIZ[EPNUM])
If (sp > 0), packet count[EPNUM] = x + 1
Otherwise, packet count[EPNUM] = x;
MCNT[EPNUM] = packet count[EPNUM]
–

The application cannot transmit a zero-length data packet at the end of a transfer.
It can transmit a single zero-length data packet by itself. To transmit a single zerolength data packet:

–

transfer size[EPNUM] = 0
packet count[EPNUM] = 1
MCNT[EPNUM] = packet count[EPNUM]

2.

3.

4.

The application can only schedule data transfers one frame at a time.
–

(MCNT – 1) × MPSIZ ≤ XFERSIZ ≤ MCNT × MPSIZ

–

PKTCNT = MCNT (in OTG_HS_DIEPTSIZx)

–

If XFERSIZ < MCNT × MPSIZ, the last data packet of the transfer is a short
packet.

–

Note that: MCNT is in OTG_HS_DIEPTSIZx, MPSIZ is in OTG_HS_DIEPCTLx,
PKTCNT is in OTG_HS_DIEPTSIZx and XFERSIZ is in OTG_HS_DIEPTSIZx

The complete data to be transmitted in the frame must be written into the transmit FIFO
by the application, before the IN token is received. Even when 1 word of the data to be
transmitted per frame is missing in the transmit FIFO when the IN token is received, the
core behaves as when the FIFO is empty. When the transmit FIFO is empty:
–

A zero data length packet would be transmitted on the USB for isochronous IN
endpoints

–

A NAK handshake would be transmitted on the USB for interrupt IN endpoints

For a high-bandwidth IN endpoint with three packets in a frame, the application can
program the endpoint FIFO size to be 2 × max_pkt_size and have the third packet
loaded in after the first packet has been transmitted on the USB.

Internal data flow:
1.

The application must set the transfer size and packet count fields in the endpointspecific registers and enable the endpoint to transmit the data.

2.

The application must also write the required data to the associated transmit FIFO for
the endpoint.

3.

Every time the application writes a packet to the transmit FIFO, the transfer size for that
endpoint is decremented by the packet size. The data are fetched from application
memory until the transfer size for the endpoint becomes 0.

4.

When an IN token is received for a periodic endpoint, the core transmits the data in the
FIFO, if available. If the complete data payload (complete packet, in dedicated FIFO

RM0090 Rev 17

1531/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

mode) for the frame is not present in the FIFO, then the core generates an IN token
received when TxFIFO empty interrupt for the endpoint.

5.

6.

–

A zero-length data packet is transmitted on the USB for isochronous IN endpoints

–

A NAK handshake is transmitted on the USB for interrupt IN endpoints

The packet count for the endpoint is decremented by 1 under the following conditions:
–

For isochronous endpoints, when a zero- or nonzero-length data packet is
transmitted

–

For interrupt endpoints, when an ACK handshake is transmitted

–

When the transfer size and packet count are both 0, the transfer completed
interrupt for the endpoint is generated and the endpoint enable is cleared.

At the “Periodic frame Interval” (controlled by PFIVL in OTG_HS_DCFG), when the
core finds nonempty any of the isochronous IN endpoint FIFOs scheduled for the
current frame nonempty, the core generates an IISOIXFR interrupt in
OTG_HS_GINTSTS.

Application programming sequence:
1.

Program the OTG_HS_DIEPCTLx register with the endpoint characteristics and set the
CNAK and EPENA bits.

2.

Write the data to be transmitted in the next frame to the transmit FIFO.

3.

Asserting the ITTXFE interrupt (in OTG_HS_DIEPINTx) indicates that the application
has not yet written all data to be transmitted to the transmit FIFO.

4.

If the interrupt endpoint is already enabled when this interrupt is detected, ignore the
interrupt. If it is not enabled, enable the endpoint so that the data can be transmitted on
the next IN token attempt.

5.

Asserting the XFRC interrupt (in OTG_HS_DIEPINTx) with no ITTXFE interrupt in
OTG_HS_DIEPINTx indicates the successful completion of an isochronous IN transfer.
A read to the OTG_HS_DIEPTSIZx register must give transfer size = 0 and packet
count = 0, indicating all data were transmitted on the USB.

6.

Asserting the XFRC interrupt (in OTG_HS_DIEPINTx), with or without the ITTXFE
interrupt (in OTG_HS_DIEPINTx), indicates the successful completion of an interrupt
IN transfer. A read to the OTG_HS_DIEPTSIZx register must give transfer size = 0 and
packet count = 0, indicating all data were transmitted on the USB.

7.

Asserting the incomplete isochronous IN transfer (IISOIXFR) interrupt in
OTG_HS_GINTSTS with none of the aforementioned interrupts indicates the core did
not receive at least 1 periodic IN token in the current frame.

•

Incomplete isochronous IN data transfers

This section describes what the application must do on an incomplete isochronous IN data
transfer.
Internal data flow:
1.

1532/1747

An isochronous IN transfer is treated as incomplete in one of the following conditions:
a)

The core receives a corrupted isochronous IN token on at least one isochronous
IN endpoint. In this case, the application detects an incomplete isochronous IN
transfer interrupt (IISOIXFR in OTG_HS_GINTSTS).

b)

The application is slow to write the complete data payload to the transmit FIFO
and an IN token is received before the complete data payload is written to the
FIFO. In this case, the application detects an IN token received when TxFIFO
empty interrupt in OTG_HS_DIEPINTx. The application can ignore this interrupt,

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
as it eventually results in an incomplete isochronous IN transfer interrupt
(IISOIXFR in OTG_HS_GINTSTS) at the end of periodic frame.
The core transmits a zero-length data packet on the USB in response to the
received IN token.
2.

The application must stop writing the data payload to the transmit FIFO as soon as
possible.

3.

The application must set the NAK bit and the disable bit for the endpoint.

4.

The core disables the endpoint, clears the disable bit, and asserts the Endpoint Disable
interrupt for the endpoint.

Application programming sequence
1.

The application can ignore the IN token received when TxFIFO empty interrupt in
OTG_HS_DIEPINTx on any isochronous IN endpoint, as it eventually results in an
incomplete isochronous IN transfer interrupt (in OTG_HS_GINTSTS).

2.

Assertion of the incomplete isochronous IN transfer interrupt (in OTG_HS_GINTSTS)
indicates an incomplete isochronous IN transfer on at least one of the isochronous IN
endpoints.

3.

The application must read the Endpoint Control register for all isochronous IN
endpoints to detect endpoints with incomplete IN data transfers.

4.

The application must stop writing data to the Periodic Transmit FIFOs associated with
these endpoints on the AHB.

5.

Program the following fields in the OTG_HS_DIEPCTLx register to disable the
endpoint:

6.

–

SNAK = 1 in OTG_HS_DIEPCTLx

–

EPDIS = 1 in OTG_HS_DIEPCTLx

The assertion of the Endpoint Disabled interrupt in OTG_HS_DIEPINTx indicates that
the core has disabled the endpoint.
–

•

At this point, the application must flush the data in the associated transmit FIFO or
overwrite the existing data in the FIFO by enabling the endpoint for a new transfer
in the next micro-frame. To flush the data, the application must use the
OTG_HS_GRSTCTL register.

Stalling nonisochronous IN endpoints

This section describes how the application can stall a nonisochronous endpoint.
Application programming sequence:

RM0090 Rev 17

1533/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090

1.

Disable the IN endpoint to be stalled. Set the STALL bit as well.

2.

EPDIS = 1 in OTG_HS_DIEPCTLx, when the endpoint is already enabled
–

STALL = 1 in OTG_HS_DIEPCTLx

–

The STALL bit always takes precedence over the NAK bit

3.

Assertion of the Endpoint Disabled interrupt (in OTG_HS_DIEPINTx) indicates to the
application that the core has disabled the specified endpoint.

4.

The application must flush the nonperiodic or periodic transmit FIFO, depending on the
endpoint type. In case of a nonperiodic endpoint, the application must re-enable the
other nonperiodic endpoints that do not need to be stalled, to transmit data.

5.

Whenever the application is ready to end the STALL handshake for the endpoint, the
STALL bit must be cleared in OTG_HS_DIEPCTLx.

6.

If the application sets or clears a STALL bit for an endpoint due to a
SetFeature.Endpoint Halt command or ClearFeature.Endpoint Halt command, the
STALL bit must be set or cleared before the application sets up the Status stage
transfer on the control endpoint.

Special case: stalling the control OUT endpoint
The core must stall IN/OUT tokens if, during the data stage of a control transfer, the host
sends more IN/OUT tokens than are specified in the SETUP packet. In this case, the
application must enable the ITTXFE interrupt in OTG_HS_DIEPINTx and the OTEPDIS
interrupt in OTG_HS_DOEPINTx during the data stage of the control transfer, after the core
has transferred the amount of data specified in the SETUP packet. Then, when the
application receives this interrupt, it must set the STALL bit in the corresponding endpoint
control register, and clear this interrupt.

35.13.8

Worst case response time
When the OTG_HS controller acts as a device, there is a worst case response time for any
tokens that follow an isochronous OUT. This worst case response time depends on the AHB
clock frequency.
The core registers are in the AHB domain, and the core does not accept another token
before updating these register values. The worst case is for any token following an
isochronous OUT, because for an isochronous transaction, there is no handshake and the
next token could come sooner. This worst case value is 7 PHY clocks when the AHB clock
is the same as the PHY clock. When the AHB clock is faster, this value is smaller.
If this worst case condition occurs, the core responds to bulk/interrupt tokens with a NAK
and drops isochronous and SETUP tokens. The host interprets this as a timeout condition
for SETUP and retries the SETUP packet. For isochronous transfers, the Incomplete
isochronous IN transfer interrupt (IISOIXFR) and Incomplete isochronous OUT transfer
interrupt (IISOOXFR) inform the application that isochronous IN/OUT packets were
dropped.

Choosing the value of TRDT in OTG_HS_GUSBCFG
The value in TRDT (OTG_HS_GUSBCFG) is the time it takes for the MAC, in terms of PHY
clocks after it has received an IN token, to get the FIFO status, and thus the first data from
the PFC block. This time involves the synchronization delay between the PHY and AHB
clocks. The worst case delay for this is when the AHB clock is the same as the PHY clock.
In this case, the delay is 5 clocks.

1534/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
Once the MAC receives an IN token, this information (token received) is synchronized to the
AHB clock by the PFC (the PFC runs on the AHB clock). The PFC then reads the data from
the SPRAM and writes them into the dual clock source buffer. The MAC then reads the data
out of the source buffer (4 deep).
If the AHB is running at a higher frequency than the PHY, the application can use a smaller
value for TRDT (in OTG_HS_GUSBCFG).
Figure 429 has the following signals:
•

tkn_rcvd: Token received information from MAC to PFC

•

dynced_tkn_rcvd: Doubled sync tkn_rcvd, from PCLK to HCLK domain

•

spr_read: Read to SPRAM

•

spr_addr: Address to SPRAM

•

spr_rdata: Read data from SPRAM

•

srcbuf_push: Push to the source buffer

•

srcbuf_rdata: Read data from the source buffer. Data seen by MAC

Refer to Table 213: TRDT values for the values of TRDT versus AHB clock frequency.
Figure 429. TRDT max timing case
0ns

1

50ns

2

3

100ns

4

150ns

5

6

200ns

7

8

HCLK

PCLK

tkn_rcvd

dsynced_tkn_rcvd

spr_read

spr_addr

A1

D1

spr_rdata

srcbuf_push

srcbuf_rdata

D1

5 Clocks
ai15680

RM0090 Rev 17

1535/1747
1541

USB on-the-go high-speed (OTG_HS)

35.13.9

RM0090

OTG programming model
The OTG_HS controller is an OTG device supporting HNP and SRP. When the core is
connected to an “A” plug, it is referred to as an A-device. When the core is connected to a
“B” plug it is referred to as a B-device. In host mode, the OTG_HS controller turns off VBUS
to conserve power. SRP is a method by which the B-device signals the A-device to turn on
VBUS power. A device must perform both data-line pulsing and VBUS pulsing, but a host can
detect either data-line pulsing or VBUS pulsing for SRP. HNP is a method by which the Bdevice negotiates and switches to host role. In Negotiated mode after HNP, the B-device
suspends the bus and reverts to the device role.

A-device session request protocol
The application must set the SRP-capable bit in the Core USB configuration register. This
enables the OTG_HS controller to detect SRP as an A-device.
Figure 430. A-device SRP
Suspend

6
1

DRV_VBUS

5

2
VBUS_VALID

VBUS pulsing

A_VALID

3

OTG_HS_FS_DP

OTG_HS_FS_DM

4
Data line pulsing

7
Connect

Low
ai15681b

1. DRV_VBUS = VBUS drive signal to the PHY
VBUS_VALID = VBUS valid signal from PHY
A_VALID = A-device VBUS level signal to PHY
DP = Data plus line
DM = Data minus line

1536/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
1.

To save power, the application suspends and turns off port power when the bus is idle
by writing the port suspend and port power bits in the host port control and status
register.

2.

PHY indicates port power off by deasserting the VBUS_VALID signal.

3.

The device must detect SE0 for at least 2 ms to start SRP when VBUS power is off.

4.

To initiate SRP, the device turns on its data line pull-up resistor for 5 to 10 ms. The
OTG_HS controller detects data-line pulsing.

5.

The device drives VBUS above the A-device session valid (2.0 V minimum) for VBUS
pulsing.
The OTG_HS controller interrupts the application on detecting SRP. The Session
request detected bit is set in Global interrupt status register (SRQINT set in
OTG_HS_GINTSTS).

6.

The application must service the Session request detected interrupt and turn on the
port power bit by writing the port power bit in the host port control and status register.
The PHY indicates port power-on by asserting the VBUS_VALID signal.

7.

When the USB is powered, the device connects, completing the SRP process.

B-device session request protocol
The application must set the SRP-capable bit in the Core USB configuration register. This
enables the OTG_HS controller to initiate SRP as a B-device. SRP is a means by which the
OTG_HS controller can request a new session from the host.
Figure 431. B-device SRP
6

Suspend
1

VBUS_VALID

2

B_VALID

3
DISCHRG_VBUS
4
SESS_END
5

8
Data line pulsing

Connect

OTG_HS_FS_DP

OTG_HS_FS_DM

Low
7
VBUS pulsing

CHRG_VBUS

ai1568b2

1. VBUS_VALID = VBUS valid signal from PHY
B_VALID = B-device valid session to PHY
DISCHRG_VBUS = discharge signal to PHY
SESS_END = session end signal to PHY
CHRG_VBUS = charge VBUS signal to PHY
DP = Data plus line
DM = Data minus line

RM0090 Rev 17

1537/1747
1541

USB on-the-go high-speed (OTG_HS)
1.

RM0090

To save power, the host suspends and turns off port power when the bus is idle.
The OTG_HS controller sets the early suspend bit in the Core interrupt register after 3
ms of bus idleness. Following this, the OTG_HS controller sets the USB suspend bit in
the Core interrupt register.
The OTG_HS controller informs the PHY to discharge VBUS.

2.

The PHY indicates the session’s end to the device. This is the initial condition for SRP.
The OTG_HS controller requires 2 ms of SE0 before initiating SRP.
For a USB 1.1 full-speed serial transceiver, the application must wait until VBUS
discharges to 0.2 V after BSVLD (in OTG_HS_GOTGCTL) is deasserted. This
discharge time can be obtained from the transceiver vendor and varies from one
transceiver to another.

3.

The USB OTG core informs the PHY to speed up VBUS discharge.

4.

The application initiates SRP by writing the session request bit in the OTG Control and
status register. The OTG_HS controller perform data-line pulsing followed by VBUS
pulsing.

5.

The host detects SRP from either the data-line or VBUS pulsing, and turns on VBUS.
The PHY indicates VBUS power-on to the device.

6.

The OTG_HS controller performs VBUS pulsing.
The host starts a new session by turning on VBUS, indicating SRP success. The
OTG_HS controller interrupts the application by setting the session request success
status change bit in the OTG interrupt status register. The application reads the session
request success bit in the OTG control and status register.

7.

When the USB is powered, the OTG_HS controller connects, completing the SRP
process.

A-device host negotiation protocol
HNP switches the USB host role from the A-device to the B-device. The application must set
the HNP-capable bit in the Core USB configuration register to enable the OTG_HS
controller to perform HNP as an A-device.
Figure 432. A-device HNP
1
OTG core

Host

Device

Suspend 2
DP

4
3

Host
6

5
Reset

DM

Traffic

8
7

Connect

Traffic

DPPULLDOWN

DMPULLDOWN
ai15683b

1. DPPULLDOWN = signal from core to PHY to enable/disable the pull-down on the DP line inside the PHY.
DMPULLDOWN = signal from core to PHY to enable/disable the pull-down on the DM line inside the PHY.

1538/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
1.

The OTG_HS controller sends the B-device a SetFeature b_hnp_enable descriptor to
enable HNP support. The B-device’s ACK response indicates that the B-device
supports HNP. The application must set host Set HNP Enable bit in the OTG Control
and status register to indicate to the OTG_HS controller that the B-device supports
HNP.

2.

When it has finished using the bus, the application suspends by writing the Port
suspend bit in the host port control and status register.

3.

When the B-device observes a USB suspend, it disconnects, indicating the initial
condition for HNP. The B-device initiates HNP only when it must switch to the host role;
otherwise, the bus continues to be suspended.
The OTG_HS controller sets the host negotiation detected interrupt in the OTG
interrupt status register, indicating the start of HNP.
The OTG_HS controller deasserts the DM pull down and DM pull down in the PHY to
indicate a device role. The PHY enables the OTG_HS_DP pull-up resistor to indicate a
connect for B-device.
The application must read the current mode bit in the OTG Control and status register
to determine peripheral mode operation.

4.

The B-device detects the connection, issues a USB reset, and enumerates the
OTG_HS controller for data traffic.

5.

The B-device continues the host role, initiating traffic, and suspends the bus when
done.
The OTG_HS controller sets the early suspend bit in the Core interrupt register after 3
ms of bus idleness. Following this, the OTG_HS controller sets the USB Suspend bit in
the Core interrupt register.

6.

In Negotiated mode, the OTG_HS controller detects the suspend, disconnects, and
switches back to the host role. The OTG_HS controller asserts the DM pull down and
DM pull down in the PHY to indicate its assumption of the host role.

7.

The OTG_HS controller sets the Connector ID status change interrupt in the OTG
Interrupt Status register. The application must read the connector ID status in the OTG
Control and Status register to determine the OTG_HS controller operation as an Adevice. This indicates the completion of HNP to the application. The application must
read the Current mode bit in the OTG control and status register to determine host
mode operation.

8.

The B-device connects, completing the HNP process.

B-device host negotiation protocol
HNP switches the USB host role from B-device to A-device. The application must set the
HNP-capable bit in the Core USB configuration register to enable the OTG_HS controller to
perform HNP as a B-device.

RM0090 Rev 17

1539/1747
1541

USB on-the-go high-speed (OTG_HS)

RM0090
Figure 433. B-device HNP

1
OTG core

Device

Host

Suspend 2
DP

4
3

Device
6

5
Reset

DM

Traffic

8
7

Connect

Traffic

DPPULLDOWN

DMPULLDOWN
ai15684b

1. DPPULLDOWN = signal from core to PHY to enable/disable the pull-down on the DP line inside the PHY.
DMPULLDOWN = signal from core to PHY to enable/disable the pull-down on the DM line inside the PHY.

1.

The A-device sends the SetFeature b_hnp_enable descriptor to enable HNP support.
The OTG_HS controller’s ACK response indicates that it supports HNP. The
application must set the Device HNP enable bit in the OTG Control and status register
to indicate HNP support.
The application sets the HNP request bit in the OTG Control and status register to
indicate to the OTG_HS controller to initiate HNP.

2.

When it has finished using the bus, the A-device suspends by writing the Port suspend
bit in the host port control and status register.
The OTG_HS controller sets the Early suspend bit in the Core interrupt register after 3
ms of bus idleness. Following this, the OTG_HS controller sets the USB suspend bit in
the Core interrupt register.
The OTG_HS controller disconnects and the A-device detects SE0 on the bus,
indicating HNP. The OTG_HS controller asserts the DP pull down and DM pull down in
the PHY to indicate its assumption of the host role.
The A-device responds by activating its OTG_HS_DP pull-up resistor within 3 ms of
detecting SE0. The OTG_HS controller detects this as a connect.
The OTG_HS controller sets the host negotiation success status change interrupt in
the OTG Interrupt status register, indicating the HNP status. The application must read
the host negotiation success bit in the OTG Control and status register to determine

1540/1747

RM0090 Rev 17

RM0090

USB on-the-go high-speed (OTG_HS)
host negotiation success. The application must read the current Mode bit in the Core
interrupt register (OTG_HS_GINTSTS) to determine host mode operation.
3.

The application sets the reset bit (PRST in OTG_HS_HPRT) and the OTG_HS
controller issues a USB reset and enumerates the A-device for data traffic.

4.

The OTG_HS controller continues the host role of initiating traffic, and when done,
suspends the bus by writing the Port suspend bit in the host port control and status
register.

5.

In Negotiated mode, when the A-device detects a suspend, it disconnects and switches
back to the host role. The OTG_HS controller deasserts the DP pull down and DM pull
down in the PHY to indicate the assumption of the device role.

6.

The application must read the current mode bit in the Core interrupt
(OTG_HS_GINTSTS) register to determine the host mode operation.

7.

The OTG_HS controller connects, completing the HNP process.

RM0090 Rev 17

1541/1747
1541

Flexible static memory controller (FSMC)

36

RM0090

Flexible static memory controller (FSMC)
This section applies to the whole STM32F40x/41x family only.

36.1

FSMC main features
The FSMC block is able to interface with synchronous and asynchronous memories and
16-bit PC memory cards. Its main purpose is to:
•

Translate the AHB transactions into the appropriate external device protocol

•

Meet the access timing requirements of the external devices

All external memories share the addresses, data and control signals with the controller.
Each external device is accessed by means of a unique chip select. The FSMC performs
only one access at a time to an external device.
The FSMC has the following main features:
•

Interfaces with static memory-mapped devices including:
–

Static random access memory (SRAM)

–

NOR Flash memory/OneNAND Flash memory

–

PSRAM (4 memory banks)

•

Two banks of NAND Flash with ECC hardware that checks up to 8 Kbytes of data

•

16-bit PC Card compatible devices

•

Supports burst mode access to synchronous devices (NOR Flash and PSRAM)

•

8- or 16-bit wide databus

•

Independent chip select control for each memory bank

•

Independent configuration for each memory bank

•

Programmable timings to support a wide range of devices, in particular:
–

Programmable wait states (up to 15)

–

Programmable bus turnaround cycles (up to 15)

–

Programmable output enable and write enable delays (up to 15)

–

Independent read and write timings and protocol, so as to support the widest
variety of memories and timings

•

Write enable and byte lane select outputs for use with PSRAM and SRAM devices

•

Translation of 32-bit wide AHB transactions into consecutive 16-bit or 8-bit accesses to
external 16-bit or 8-bit devices

•

A Write FIFO, 2-word long (16-word long for STM32F42x and STM32F43x), each word
is 32 bits wide, only stores data and not the address. Therefore, this FIFO only buffers
AHB write burst transactions. This makes it possible to write to slow memories and free
the AHB quickly for other operations. Only one burst at a time is buffered: if a new AHB
burst or single transaction occurs while an operation is in progress, the FIFO is
drained. The FSMC will insert wait states until the current memory access is complete.

•

External asynchronous wait control

The FSMC registers that define the external device type and associated characteristics are
usually set at boot time and do not change until the next reset or power-up. However, it is
possible to change the settings at any time.

1542/1747

RM0090 Rev 17

RM0090

36.2

Flexible static memory controller (FSMC)

Block diagram
The FSMC consists of four main blocks:
•

The AHB interface (including the FSMC configuration registers)

•

The NOR Flash/PSRAM controller

•

The NAND Flash/PC Card controller

•

The external device interface

The block diagram is shown in Figure 434.
Figure 434. FSMC block diagram
&3-# INTERRUPT TO .6)#

&ROM CLOCK
CONTROLLER

./2032!MEMORY
CONTROLLER

!(" BUS

(#,+

#ONFIGURATION
REGISTERS

.!.$0# #ARD
MEMORY
CONTROLLER

&3-#?.%;=
&3-#?.,OR .!$6
&3-#?.",;=
&3-#?#,+

./2032!SIGNALS

&3-#?!;=
&3-#?$;=
&3-#?./%
&3-#?.7%
&3-#?.7!)4

3HARED
SIGNALS

&3-#?.#%;=
&3-#?).4;=

.!.$
SIGNALS

&3-#?).42
&3-#?.#%?
&3-#?.#%?
&3-#?.)/2$
&3-#?.)/72

0# #ARD
SIGNALS

&3-#?.2%'
&3-#?#$

AIB

36.3

AHB interface
The AHB slave interface enables internal CPUs and other bus master peripherals to access
the external static memories.
AHB transactions are translated into the external device protocol. In particular, if the
selected external memory is 16 or 8 bits wide, 32-bit wide transactions on the AHB are split
into consecutive 16- or 8-bit accesses. The FSMC Chip Select (FSMC_NEx) does not

RM0090 Rev 17

1543/1747
1599

Flexible static memory controller (FSMC)

RM0090

toggle between consecutive accesses except when performing accesses in mode D with the
extended mode enabled.
The FSMC generates an AHB error in the following conditions:
•

When reading or writing to an FSMC bank which is not enabled

•

When reading or writing to the NOR Flash bank while the FACCEN bit is reset in the
FSMC_BCRx register.

•

When reading or writing to the PC Card banks while the input pin FSMC_CD (Card
Presence Detection) is low.

The effect of this AHB error depends on the AHB master which has attempted the R/W
access:
•

If it is the Cortex®-M4 with FPU CPU, a hard fault interrupt is generated

•

If is a DMA, a DMA transfer error is generated and the corresponding DMA channel is
automatically disabled.

The AHB clock (HCLK) is the reference clock for the FSMC.

36.3.1

Supported memories and transactions
General transaction rules
The requested AHB transaction data size can be 8-, 16- or 32-bit wide whereas the
accessed external device has a fixed data width. This may lead to inconsistent transfers.
Therefore, some simple transaction rules must be followed:
•

AHB transaction size and memory data size are equal
There is no issue in this case.

•

AHB transaction size is greater than the memory size
In this case, the FSMC splits the AHB transaction into smaller consecutive memory
accesses in order to meet the external data width.

•

AHB transaction size is smaller than the memory size
Asynchronous transfers may or not be consistent depending on the type of external
device.
–

Asynchronous accesses to devices that have the byte select feature (SRAM,
ROM, PSRAM).
a) FSMC allows write transactions accessing the right data through its byte lanes
NBL[1:0]
b) Read transactions are allowed. All memory bytes are read and the useless
ones are discarded. The NBL[1:0] are kept low during read transactions.

–

1544/1747

Asynchronous accesses to devices that do not have the byte select feature (NOR
and NAND Flash 16-bit).
This situation occurs when a byte access is requested to a 16-bit wide Flash
memory. Clearly, the device cannot be accessed in byte mode (only 16-bit words
can be read from/written to the Flash memory) therefore:

a)

Write transactions are not allowed

b)

Read transactions are allowed. All memory bytes are read and the useless ones
are discarded. The NBL[1:0] are set to 0 during read transactions.

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)

Configuration registers
The FSMC can be configured using a register set. See Section 36.5.6, for a detailed
description of the NOR Flash/PSRAM control registers. See Section 36.6.8, for a detailed
description of the NAND Flash/PC Card registers.

36.4

External device address mapping
From the FSMC point of view, the external memory is divided into 4 fixed-size banks of
256 Mbytes each (Refer to Figure 435):
•

Bank 1 used to address up to 4 NOR Flash or PSRAM memory devices. This bank is
split into 4 NOR/PSRAM subbanks with 4 dedicated Chip Selects, as follows:
–

Bank 1 - NOR/PSRAM 1

–

Bank 1 - NOR/PSRAM 2

–

Bank 1 - NOR/PSRAM 3

–

Bank 1 - NOR/PSRAM 4

•

Banks 2 and 3 used to address NAND Flash devices (1 device per bank)

•

Bank 4 used to address a PC Card device

For each bank the type of memory to be used is user-defined in the Configuration register.
Figure 435. FSMC memory banks
Address

Banks

Supported memory type

6000 0000h
Bank 1

NOR / PSRAM

4 × 64 MB
6FF F FFF Fh
7000 0000h
Bank 2
4 × 64 MB
7FF F FFF Fh
NAND Flash
8000 0000h

Bank 3
4 × 64 MB

8FF F FFF Fh
9000 0000h
Bank 4
PC Card
4 × 64 MB
9FF F FFF Fh
ai14719

36.4.1

NOR/PSRAM address mapping
HADDR[27:26] bits are used to select one of the four memory banks as shown in Table 216.

RM0090 Rev 17

1545/1747
1599

Flexible static memory controller (FSMC)

RM0090

Table 216. NOR/PSRAM bank selection
HADDR[27:26](1)

Selected bank

00

Bank 1 - NOR/PSRAM 1

01

Bank 1 - NOR/PSRAM 2

10

Bank 1 - NOR/PSRAM 3

11

Bank 1 - NOR/PSRAM 4

1. HADDR are internal AHB address lines that are translated to external memory.

HADDR[25:0] contain the external memory address. Since HADDR is a byte address
whereas the memory is addressed in words, the address actually issued to the memory
varies according to the memory data width, as shown in the following table.
Table 217. External memory address
Memory

width(1)

Data address issued to the memory

Maximum memory capacity (bits)

8-bit

HADDR[25:0]

64 Mbyte x 8 = 512 Mbit

16-bit

HADDR[25:1] >> 1

64 Mbyte/2 x 16 = 512 Mbit

1. In case of a 16-bit external memory width, the FSMC will internally use HADDR[25:1] to generate the
address for external memory FSMC_A[24:0].
Whatever the external memory width (16-bit or 8-bit), FSMC_A[0] should be connected to external memory
address A[0].

Wrap support for NOR Flash/PSRAM
Wrap burst mode for synchronous memories is not supported. The memories must be
configured in linear burst mode of undefined length.

36.4.2

NAND/PC Card address mapping
In this case, three banks are available, each of them divided into memory spaces as
indicated in Table 218.
Table 218. Memory mapping and timing registers

1546/1747

Start address

End address

0x9C00 0000

0x9FFF FFFF

FSMC Bank

Memory space

Timing register

I/O

FSMC_PIO4 (0xB0)

0x9800 0000

0x9BFF FFFF Bank 4 - PC card

Attribute

FSMC_PATT4 (0xAC)

0x9000 0000

0x93FF FFFF

Common

FSMC_PMEM4 (0xA8)

0x8800 0000

0x8BFF FFFF

Attribute

FSMC_PATT3 (0x8C)

0x8000 0000

0x83FF FFFF

Common

FSMC_PMEM3 (0x88)

0x7800 0000

0x7BFF FFFF

Attribute

FSMC_PATT2 (0x6C)

0x7000 0000

0x73FF FFFF

Common

FSMC_PMEM2 (0x68)

Bank 3 - NAND Flash

Bank 2- NAND Flash

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)
For NAND Flash memory, the common and attribute memory spaces are subdivided into
three sections (see in Table 219 below) located in the lower 256 Kbytes:
•

Data section (first 64 Kbytes in the common/attribute memory space)

•

Command section (second 64 Kbytes in the common / attribute memory space)

•

Address section (next 128 Kbytes in the common / attribute memory space)
Table 219. NAND bank selections
Section name

HADDR[17:16]

Address range

Address section

1X

0x020000-0x03FFFF

Command section

01

0x010000-0x01FFFF

Data section

00

0x000000-0x0FFFF

The application software uses the 3 sections to access the NAND Flash memory:
•

To send a command to NAND Flash memory: the software must write the command
value to any memory location in the command section.

•

To specify the NAND Flash address that must be read or written: the software
must write the address value to any memory location in the address section. Since an
address can be 4 or 5 bytes long (depending on the actual memory size), several
consecutive writes to the address section are needed to specify the full address.

•

To read or write data: the software reads or writes the data value from or to any
memory location in the data section.

Since the NAND Flash memory automatically increments addresses, there is no need to
increment the address of the data section to access consecutive memory locations.

36.5

NOR Flash/PSRAM controller
The FSMC generates the appropriate signal timings to drive the following types of
memories:
•

•

•

Asynchronous SRAM and ROM
–

8-bit

–

16-bit

–

32-bit

PSRAM (Cellular RAM)
–

Asynchronous mode

–

Burst mode for synchronous accesses

–

Multiplexed or nonmultiplexed

NOR Flash
–

Asynchronous mode

–

Burst mode for synchronous accesses

–

Multiplexed or nonmultiplexed

The FSMC outputs a unique chip select signal NE[4:1] per bank. All the other signals
(addresses, data and control) are shared.

RM0090 Rev 17

1547/1747
1599

Flexible static memory controller (FSMC)

RM0090

For synchronous accesses, the FSMC issues the clock (CLK) to the selected external
device only during the read/write transactions. This clock is a submultiple of the HCLK clock.
The size of each bank is fixed and equal to 64 Mbytes.
Each bank is configured by means of dedicated registers (see Section 36.5.6).
The programmable memory parameters include access timings (see Table 220) and support
for wait management (for PSRAM and NOR Flash accessed in burst mode).
Table 220. Programmable NOR/PSRAM access parameters
Parameter

36.5.1

Function

Access mode

Unit

Min.

Max.

Address
setup

Duration of the address
setup phase

Asynchronous

AHB clock cycle
(HCLK)

0

15

Address hold

Duration of the address hold Asynchronous,
phase
muxed I/Os

AHB clock cycle
(HCLK)

1

15

Data setup

Duration of the data setup
phase

Asynchronous

AHB clock cycle
(HCLK)

1

256

Bus turn

Duration of the bus
turnaround phase

Asynchronous and
AHB clock cycle
synchronous
(HCLK)
read/write

0

15

Clock divide
ratio

Number of AHB clock cycles
(HCLK) to build one memory Synchronous
clock cycle (CLK)

AHB clock cycle
(HCLK)

2

16

Data latency

Number of clock cycles to
issue to the memory before
the first data of the burst

Memory clock
cycle (CLK)

2

17

Synchronous

External memory interface signals
Table 221, Table 222 and Table 223 list the signals that are typically used to interface NOR
Flash, SRAM and PSRAM.

Note:

Prefix “N”. specifies the associated signal as active low.

NOR Flash, nonmultiplexed I/Os
Table 221. Nonmultiplexed I/O NOR Flash
FSMC signal name

1548/1747

I/O

Function

CLK

O

Clock (for synchronous access)

A[25:0]

O

Address bus

D[15:0]

I/O

Bidirectional data bus

NE[x]

O

Chip select, x = 1..4

NOE

O

Output enable

NWE

O

Write enable

NL(=NADV)

O

Latch enable (this signal is called address
valid, NADV, by some NOR Flash devices)

NWAIT

I

NOR Flash wait input signal to the FSMC

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)
NOR Flash memories are addressed in 16-bit words. The maximum capacity is 512 Mbit (26
address lines).

NOR Flash, multiplexed I/Os
Table 222. Multiplexed I/O NOR Flash
FSMC signal name

I/O

Function

CLK

O

Clock (for synchronous access)

A[25:16]

O

Address bus

AD[15:0]

I/O

16-bit multiplexed, bidirectional address/data bus

NE[x]

O

Chip select, x = 1..4

NOE

O

Output enable

NWE

O

Write enable

NL(=NADV)

O

Latch enable (this signal is called address valid, NADV, by some NOR
Flash devices)

NWAIT

I

NOR Flash wait input signal to the FSMC

NOR-Flash memories are addressed in 16-bit words. The maximum capacity is 512 Mbit
(26 address lines).

PSRAM/SRAM, nonmultiplexed I/Os
Table 223. Nonmultiplexed I/Os PSRAM/SRAM
FSMC signal name

I/O

Function

CLK

O

Clock (only for PSRAM synchronous access)

A[25:0]

O

Address bus

D[15:0]

I/O

Data bidirectional bus

NE[x]

O

Chip select, x = 1..4 (called NCE by PSRAM (Cellular RAM i.e. CRAM))

NOE

O

Output enable

NWE

O

Write enable

NL(= NADV)

O

Address valid only for PSRAM input (memory signal name: NADV)

NWAIT

I

PSRAM wait input signal to the FSMC

NBL[1]

O

Upper byte enable (memory signal name: NUB)

NBL[0]

O

Lowed byte enable (memory signal name: NLB)

PSRAM memories are addressed in 16-bit words. The maximum capacity is 512 Mbit (26
address lines).
PSRAM, multiplexed I/Os

RM0090 Rev 17

1549/1747
1599

Flexible static memory controller (FSMC)

RM0090

Table 224. Multiplexed I/O PSRAM
FSMC signal name

I/O

Function

CLK

O

Clock (for synchronous access)

A[25:16]

O

Address bus

AD[15:0]

I/O

16-bit multiplexed, bidirectional address/data bus

NE[x]

O

Chip select, x = 1..4 (called NCE by PSRAM (Cellular RAM i.e. CRAM))

NOE

O

Output enable

NWE

O

Write enable

NL(= NADV)

O

Address valid PSRAM input (memory signal name: NADV)

NWAIT

I

PSRAM wait input signal to the FSMC

NBL[1]

O

Upper byte enable (memory signal name: NUB)

NBL[0]

O

Lowed byte enable (memory signal name: NLB)

PSRAM memories are addressed in 16-bit words. The maximum capacity is 512 Mbit (26
address lines).

36.5.2

Supported memories and transactions
Table 225 below displays an example of the supported devices, access modes and
transactions when the memory data bus is 16-bit for NOR, PSRAM and SRAM.
Transactions not allowed (or not supported) by the FSMC in this example appear in gray.

Table 225. NOR Flash/PSRAM controller: example of supported memories and transactions
Device

NOR Flash
(muxed I/Os and
nonmuxed I/Os)

1550/1747

R/W

AHB
data
size

Memory
data size

Allowed/
not
allowed

Asynchronous

R

8

16

Y

-

Asynchronous

W

8

16

N

-

Asynchronous

R

16

16

Y

-

Asynchronous

W

16

16

Y

-

Asynchronous

R

32

16

Y

Split into two FSMC accesses

Asynchronous

W

32

16

Y

Split into two FSMC accesses

Asynchronous page

R

-

16

N

Mode is not supported

Synchronous

R

8

16

N

-

Synchronous

R

16

16

Y

-

Synchronous

R

32

16

Y

-

Mode

RM0090 Rev 17

Comments

RM0090

Flexible static memory controller (FSMC)

Table 225. NOR Flash/PSRAM controller: example of supported memories and transactions
Device

PSRAM
(multiplexed and
nonmultiplexed
I/Os)

SRAM and ROM

36.5.3

R/W

AHB
data
size

Memory
data size

Allowed/
not
allowed

Asynchronous

R

8

16

Y

-

Asynchronous

W

8

16

Y

Use of byte lanes NBL[1:0]

Asynchronous

R

16

16

Y

-

Asynchronous

W

16

16

Y

-

Asynchronous

R

32

16

Y

Split into two FSMC accesses

Asynchronous

W

32

16

Y

Split into two FSMC accesses

Asynchronous page

R

-

16

N

Mode is not supported

Synchronous

R

8

16

N

-

Synchronous

R

16

16

Y

-

Synchronous

R

32

16

Y

-

Synchronous

W

8

16

Y

Use of byte lanes NBL[1:0]

Synchronous

W

16 / 32

16

Y

-

Asynchronous

R

8 / 16

16

Y

-

Asynchronous

W

8 / 16

16

Y

Use of byte lanes NBL[1:0]

Asynchronous

R

32

16

Y

Split into two FSMC accesses

Asynchronous

W

32

16

Y

Split into two FSMC accesses.
Use of byte lanes NBL[1:0]

Mode

Comments

General timing rules
Signals synchronization
•

All controller output signals change on the rising edge of the internal clock (HCLK)

•

In synchronous mode (read or write), all output signals change on the rising edge of
HCLK. Whatever the CLKDIV value, all outputs change as follows:
–

NOEL/NWEL/ NEL/NADVL/ NADVH /NBLL/ Address valid outputs change on the
falling edge of FSMC_CLK clock.

–

NOEH/ NWEH / NEH/ NOEH/NBLH/ Address invalid outputs change on the rising
edge of FSMC_CLK clock.

RM0090 Rev 17

1551/1747
1599

Flexible static memory controller (FSMC)

36.5.4

RM0090

NOR Flash/PSRAM controller asynchronous transactions
Asynchronous static memories (NOR Flash memory, PSRAM, SRAM)
•

Signals are synchronized by the internal clock HCLK. This clock is not issued to the
memory

•

The FSMC always samples the data before de-asserting the NOE signals. This
guarantees that the memory data-hold timing constraint is met (chip enable high to
data transition, usually 0 ns min.)

•

If the extended mode is enabled (EXTMOD bit is set in the FSMC_BCRx register), up
to four extended modes (A, B, C and D) are available. It is possible to mix A, B, C and
D modes for read and write operations. For example, read operation can be performed
in mode A and write in mode B.

•

If the extended mode is disabled (EXTMOD bit is reset in the FSMC_BCRx register),
the FSMC can operate in Mode1 or Mode2 as follows:
–

Mode 1 is the default mode when SRAM/PSRAM memory type is selected
(MTYP[0:1] = 0x0 or 0x01 in the FSMC_BCRx register)

–

Mode 2 is the default mode when NOR memory type is selected (MTYP[0:1] =
0x10 in the FSMC_BCRx register).

Mode 1 - SRAM/PSRAM (CRAM)
The next figures show the read and write transactions for the supported modes followed by
the required configuration of FSMC _BCRx, and FSMC_BTRx/FSMC_BWTRx registers.
Figure 436. Mode1 read accesses
-EMORY TRANSACTION
!;=

.",;=

.%X

./%
.7%

(IGH

DATA DRIVEN
BY MEMORY

$;=
!$$3%4
(#,+ CYCLES

$!4!34
(#,+ CYCLES
AI

1. NBL[1:0] are driven low during read access.

1552/1747

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)
Figure 437. Mode1 write accesses
-EMORY TRANSACTION
!;=

.",;=

.%X

./%
(#,+
.7%

DATA DRIVEN BY &3-#

$;=
!$$3%4
(#,+ CYCLES

$!4!34 
(#,+ CYCLES
AI

The one HCLK cycle at the end of the write transaction helps guarantee the address and
data hold time after the NWE rising edge. Due to the presence of this one HCLK cycle, the
DATAST value must be greater than zero (DATAST > 0).
Table 226. FSMC_BCRx bit fields
Bit number
31-20

Bit name

Value to set

Reserved

0x000

CBURSTRW

0x0 (no effect on asynchronous mode)

CPSIZE

0x0 (no effect on asynchronous mode)

15

ASYNCWAIT

Set to 1 if the memory supports this feature. Otherwise keep at 0.

14

EXTMOD

0x0

13

WAITEN

0x0 (no effect on asynchronous mode)

12

WREN

As needed

11

WAITCFG

Don’t care

10

WRAPMOD

0x0

9

WAITPOL

Meaningful only if bit 15 is 1

8

BURSTEN

0x0

7

Reserved

0x1

6

FACCEN

Don’t care

5-4

MWID

As needed

3-2

MTYP[0:1]

As needed, exclude 0x2 (NOR Flash)

19
18:16

RM0090 Rev 17

1553/1747
1599

Flexible static memory controller (FSMC)

RM0090

Table 226. FSMC_BCRx bit fields (continued)
Bit number

Bit name

Value to set

1

MUXE

0x0

0

MBKEN

0x1

Table 227. FSMC_BTRx bit fields
Bit number

Bit name

Value to set

31:30

Reserved

0x0

29-28

ACCMOD

Don’t care

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST+1 HCLK cycles for
write accesses, DATAST HCLK cycles for read accesses).

7-4

ADDHLD

Don’t care

3-0

ADDSET[3:0]

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the first access phase (ADDSET HCLK cycles).
Minimum value for ADDSET is 0.

Mode A - SRAM/PSRAM (CRAM) OE toggling
Figure 438. ModeA read accesses
-EMORY TRANSACTION
!;=

.",;=

.%X

./%
.7%

(IGH

DATA DRIVEN
BY MEMORY

$;=
!$$3%4
(#,+ CYCLES

$!4!34
(#,+ CYCLES
AI

1. NBL[1:0] are driven low during read access.

1554/1747

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)
Figure 439. ModeA write accesses
0HPRU\WUDQVDFWLRQ
$>@

1%/>@

1([

12(
+&/.
1:(

GDWDGULYHQE\)60&

'>@
$''6(7
+&/.F\FOHV

'$7$67
+&/.F\FOHV
DL

The differences compared with mode1 are the toggling of NOE and the independent read
and write timings.
Table 228. FSMC_BCRx bit fields
Bit
number

Bit name

31-20

Reserved

19

CBURSTRW

0x0 (no effect on asynchronous mode)

18:16

CPSIZE

0x0 (no effect on asynchronous mode)

15

ASYNCWAIT

14

EXTMOD

0x1

13

WAITEN

0x0 (no effect on asynchronous mode)

12

WREN

As needed

11

WAITCFG

Don’t care

10

WRAPMOD

9

WAITPOL

Meaningful only if bit 15 is 1

8

BURSTEN

0x0

7

Reserved

0x1

6

FACCEN

Don’t care

5-4

MWID

As needed

3-2

MTYP[0:1]

Value to set
0x000

Set to 1 if the memory supports this feature. Otherwise keep at
0.

0x0

As needed, exclude 0x2 (NOR Flash)

RM0090 Rev 17

1555/1747
1599

Flexible static memory controller (FSMC)

RM0090

Table 228. FSMC_BCRx bit fields (continued)
Bit
number

Bit name

1

MUXEN

0x0

0

MBKEN

0x1

Value to set

Table 229. FSMC_BTRx bit fields
Bit
number

Bit name

31:30

Reserved

0x0

29-28

ACCMOD

0x0

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST HCLK cycles) for
read accesses.

7-4

ADDHLD

Don’t care

3-0

ADDSET[3:0]

Value to set

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the first access phase (ADDSET HCLK cycles) for read
accesses.
Minimum value for ADDSET is 0.

Table 230. FSMC_BWTRx bit fields

1556/1747

Bit
number

Bit name

31:30

Reserved

0x0

29-28

ACCMOD

0x0

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST+1 HCLK cycles for
write accesses,

7-4

ADDHLD

Don’t care

3-0

ADDSET[3:0]

Value to set

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the first access phase (ADDSET HCLK cycles) for write
accesses.
Minimum value for ADDSET is 0.

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)

Mode 2/B - NOR Flash
Figure 440. Mode2 and mode B read accesses
-EMORY TRANSACTION
!;=

.!$6

.%X

./%
.7%

(IGH

DATA DRIVEN
BY MEMORY

$;=
!$$3%4
(#,+ CYCLES

$!4!34
(#,+ CYCLES
AI

Figure 441. Mode2 write accesses
-EMORY TRANSACTION
!;=

.!$6

.%X

./%
(#,+
.7%

$;=

DATA DRIVEN BY &3-#
!$$3%4
(#,+ CYCLES

$!4!34 
(#,+ CYCLES
AI

RM0090 Rev 17

1557/1747
1599

Flexible static memory controller (FSMC)

RM0090

Figure 442. Mode B write accesses
-EMORY TRANSACTION
!;=

.!$6

.%X

./%
(#,+
.7%

$;=

DATA DRIVEN BY &3-#
!$$3%4
(#,+ CYCLES

$!4!34 
(#,+ CYCLES
AI

The differences with mode1 are the toggling of NWE and the independent read and write
timings when extended mode is set (Mode B).
Table 231. FSMC_BCRx bit fields

1558/1747

Bit
number

Bit name

31-20

Reserved

19

CBURSTRW

0x0 (no effect on asynchronous mode)

18:16

Reserved

0x0 (no effect on asynchronous mode)

15

ASYNCWAIT

14

EXTMOD

0x1 for mode B, 0x0 for mode 2

13

WAITEN

0x0 (no effect on asynchronous mode)

12

WREN

As needed

11

WAITCFG

Don’t care

10

WRAPMOD

9

WAITPOL

Meaningful only if bit 15 is 1

8

BURSTEN

0x0

7

Reserved

0x1

6

FACCEN

0x1

5-4

MWID

Value to set
0x000

Set to 1 if the memory supports this feature. Otherwise keep at
0.

0x0

As needed

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)
Table 231. FSMC_BCRx bit fields (continued)
Bit
number

Bit name

3-2

MTYP[0:1]

1

MUXEN

0x0

0

MBKEN

0x1

Value to set
0x2 (NOR Flash memory)

Table 232. FSMC_BTRx bit fields
Bit
number

Bit name

31:30

Reserved

0x0

29-28

ACCMOD

0x1

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST HCLK cycles) for
read accesses.

7-4

ADDHLD

Don’t care

3-0

ADDSET[3:0]

Value to set

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the first access phase (ADDSET HCLK cycles) for read
accesses.
Minimum value for ADDSET is 0.

Table 233. FSMC_BWTRx bit fields

Note:

Bit
number

Bit name

31:30

Reserved

0x0

29-28

ACCMOD

0x1

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST+1 HCLK cycles for
write accesses,

7-4

ADDHLD

Don’t care

3-0

ADDSET[3:0]

Value to set

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the first access phase (ADDSET HCLK cycles) for write
accesses.
Minimum value for ADDSET is 0.

The FSMC_BWTRx register is valid only if extended mode is set (mode B), otherwise all its
content is don’t care.

RM0090 Rev 17

1559/1747
1599

Flexible static memory controller (FSMC)

RM0090

Mode C - NOR Flash - OE toggling
Figure 443. Mode C read accesses
-EMORY TRANSACTION
!;=

.!$6

.%X

./%
.7%

(IGH

DATA DRIVEN
BY MEMORY

$;=
!$$3%4
(#,+ CYCLES

$!4!34
(#,+ CYCLES
AI

Figure 444. Mode C write accesses
-EMORY TRANSACTION
!;=

.!$6

.%X

./%
(#,+
.7%

$;=

DATA DRIVEN BY &3-#
!$$3%4
(#,+ CYCLES

$!4!34 
(#,+ CYCLES
AI

The differences compared with mode1 are the toggling of NOE and the independent read
and write timings.

1560/1747

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)
Table 234. FSMC_BCRx bit fields
Bit No.

Bit name

Value to set

31-20

Reserved

19

CBURSTRW

0x0 (no effect on asynchronous mode)

18:16

CPSIZE

0x0 (no effect on asynchronous mode)

15

ASYNCWAIT

14

EXTMOD

0x1

13

WAITEN

0x0 (no effect on asynchronous mode)

12

WREN

As needed

11

WAITCFG

Don’t care

10

WRAPMOD

9

WAITPOL

Meaningful only if bit 15 is 1

8

BURSTEN

0x0

7

Reserved

0x1

6

FACCEN

0x1

5-4

MWID

3-2

MTYP[0:1]

1

MUXEN

0x0

0

MBKEN

0x1

0x000

Set to 1 if the memory supports this feature. Otherwise keep
at 0.

0x0

As needed
0x2 (NOR Flash memory)

Table 235. FSMC_BTRx bit fields
Bit
number

Bit name

31:30

Reserved

0x0

29-28

ACCMOD

0x2

27-24

DATLAT

0x0

23-20

CLKDIV

0x0

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST HCLK cycles) for
read accesses.

7-4

ADDHLD

Don’t care

3-0

ADDSET[3:0]

Value to set

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the first access phase (ADDSET HCLK cycles) for read
accesses.
Minimum value for ADDSET is 0.

RM0090 Rev 17

1561/1747
1599

Flexible static memory controller (FSMC)

RM0090

Table 236. FSMC_BWTRx bit fields
Bit
number

Bit name

31:30

Reserved

0x0

29-28

ACCMOD

0x2

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST+1 HCLK cycles for
write accesses,

7-4

ADDHLD

Don’t care

3-0

ADDSET[3:0]

Value to set

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the first access phase (ADDSET HCLK cycles) for write
accesses.
Minimum value for ADDSET is 0.

Mode D - asynchronous access with extended address
Figure 445. Mode D read accesses
-EMORY TRANSACTION
!;=

.!$6

.%X

./%
.7%

(IGH

DATA DRIVEN
BY MEMORY

$;=
!$$3%4
(#,+ CYCLES

1562/1747

!$$(,$
(#,+ CYCLES

RM0090 Rev 17

$!4!34
(#,+ CYCLES
AI

RM0090

Flexible static memory controller (FSMC)
Figure 446. Mode D write accesses

The differences with mode1 are the toggling of NOE that goes on toggling after NADV
changes and the independent read and write timings.
Table 237. FSMC_BCRx bit fields
Bit No.

Bit name

Value to set

31-20

Reserved

19

CBURSTRW

0x0 (no effect on asynchronous mode)

18:16

CPSIZE

0x0 (no effect on asynchronous mode)

15

ASYNCWAIT

14

EXTMOD

0x1

13

WAITEN

0x0 (no effect on asynchronous mode)

12

WREN

As needed

11

WAITCFG

Don’t care

10

WRAPMOD

9

WAITPOL

Meaningful only if bit 15 is 1

8

BURSTEN

0x0

7

Reserved

0x1

6

FACCEN

Set according to memory support

5-4

MWID

As needed

3-2

MTYP[0:1]

As needed

0x000

Set to 1 if the memory supports this feature. Otherwise keep
at 0.

0x0

RM0090 Rev 17

1563/1747
1599

Flexible static memory controller (FSMC)

RM0090

Table 237. FSMC_BCRx bit fields (continued)
Bit No.

Bit name

Value to set

1

MUXEN

0x0

0

MBKEN

0x1

Table 238. FSMC_BTRx bit fields
Bit No.

Bit name

Value to set

31:30

Reserved

0x0

29-28

ACCMOD

0x3

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST HCLK cycles) for
read accesses.

7-4

ADDHLD

Duration of the middle phase of the read access (ADDHLD HCLK
cycles)

3-0

ADDSET[3:0]

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the first access phase (ADDSET HCLK cycles) for read
accesses. Minimum value for ADDSET is 0.

Table 239. FSMC_BWTRx bit fields

1564/1747

Bit No.

Bit name

Value to set

31:30

Reserved

0x0

29-28

ACCMOD

0x3

27-24

DATLAT

0x0

23-20

CLKDIV

0x0

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST+1 HCLK cycles) for
write accesses

7-4

ADDHLD

Duration of the middle phase of the write access (ADDHLD HCLK
cycles)

3-0

ADDSET[3:0]

Duration of the first access phase (ADDSET+1 HCLK cycles) for
write accesses. Minimum value for ADDSET is 0.

Time between NEx high to NEx low (BUSTURN HCLK)

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)

Muxed mode - multiplexed asynchronous access to NOR Flash memory
Figure 447. Multiplexed read accesses

Figure 448. Multiplexed write accesses
-EMORY TRANSACTION
!;=

.!$6

.%X

./%
(#,+
.7%

!$;=

,OWER ADDRESS
!$$3%4
(#,+ CYCLES

!$$(,$
(#,+ CYCLES

DATA DRIVEN BY &3-#
$!4!34 
(#,+ CYCLES
AI

The difference with mode D is the drive of the lower address byte(s) on the databus.

RM0090 Rev 17

1565/1747
1599

Flexible static memory controller (FSMC)

RM0090

Table 240. FSMC_BCRx bit fields
Bit No.

Bit name

Value to set

31-21

Reserved

19

CBURSTRW

0x0 (no effect on asynchronous mode)

18:16

CPSIZE

0x0 (no effect on asynchronous mode)

15

ASYNCWAIT

14

EXTMOD

0x0

13

WAITEN

0x0 (no effect on asynchronous mode)

12

WREN

As needed

11

WAITCFG

Don’t care

10

WRAPMOD

9

WAITPOL

Meaningful only if bit 15 is 1

8

BURSTEN

0x0

7

Reserved

0x1

6

FACCEN

0x1

5-4

MWID

3-2

MTYP[0:1]

1

MUXEN

0x1

0

MBKEN

0x1

0x000

Set to 1 if the memory supports this feature. Otherwise keep at
0.

0x0

As needed
0x2 (NOR Flash memory)

Table 241. FSMC_BTRx bit fields
Bit No.

Bit name

Value to set

31:30

Reserved

0x0

29-28

ACCMOD

0x0

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST HCLK cycles for
read accesses and DATAST+1 HCLK cycles for write accesses).

7-4

ADDHLD

Duration of the middle phase of the access (ADDHLD HCLK cycles).

3-0

ADDSET[3:0]

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the first access phase (ADDSET HCLK cycles).
Minimum value for ADDSET is 1.

WAIT management in asynchronous accesses
If the asynchronous memory asserts a WAIT signal to indicate that it is not yet ready to
accept or to provide data, the ASYNCWAIT bit has to be set in FSMC_BCRx register.

1566/1747

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)
If the WAIT signal is active (high or low depending on the WAITPOL bit), the second access
phase (Data setup phase) programmed by the DATAST bits, is extended until WAIT
becomes inactive. Unlike the data setup phase, the first access phases (Address setup and
Address hold phases), programmed by the ADDSET[3:0] and ADDHLD bits, are not WAIT
sensitive and so they are not prolonged.
The data setup phase (DATAST in the FSMC_BTRx register) must be programmed so that
WAIT can be detected 4 HCLK cycles before the end of memory transaction. The following
cases must be considered:
1.

DATAST in FSMC_BTRx register) Memory asserts the WAIT signal aligned to
NOE/NWE which toggles:
DATAST ≥ ( 4 × HCLK ) + max_wait_assertion_time

2.

Memory asserts the WAIT signal aligned to NEx (or NOE/NWE not toggling):
if
max_wait_assertion_time > address_phase + hold_phase
then

DATAST ≥ ( 4 × HCLK ) + ( max_wait_assertion_time

– address_phase – hold_phase )

otherwise
DATAST ≥ 4 × HCLK
where max_wait_assertion_time is the maximum time taken by the memory to assert
the WAIT signal once NEx/NOE/NWE is low.
Figure 449 and Figure 450 show the number of HCLK clock cycles that are added to the
memory access after WAIT is released by the asynchronous memory (independently of the
above cases).

RM0090 Rev 17

1567/1747
1599

Flexible static memory controller (FSMC)

RM0090

Figure 449. Asynchronous wait during a read access
0HPRU\WUDQVDFWLRQ

$>@
DGGUHVVSKDVH

GDWDVHWXSSKDVH

1([

1:$,7

GRQ¶WFDUH

GRQ¶WFDUH

12(

GDWDGULYHQ
E\PHPRU\

'>@

+&/.
DLE

1. NWAIT polarity depends on WAITPOL bit setting in FSMC_BCRx register.

Figure 450. Asynchronous wait during a write access
-EMORY TRANSACTION
!;=
ADDRESS PHASE

DATA SETUP PHASE

.%X

.7!)4

DONT CARE

DONT CARE
(#,+

.7%

$;=

DATA DRIVEN BY &3-#
(#,+
AIC

1. NWAIT polarity depends on WAITPOL bit setting in FSMC_BCRx register.

1568/1747

RM0090 Rev 17

RM0090

36.5.5

Flexible static memory controller (FSMC)

Synchronous transactions
The memory clock, CLK, is a submultiple of HCLK according to the value of parameter
CLKDIV.
NOR Flash memories specify a minimum time from NADV assertion to CLK high. To meet
this constraint, the FSMC does not issue the clock to the memory during the first internal
clock cycle of the synchronous access (before NADV assertion). This guarantees that the
rising edge of the memory clock occurs in the middle of the NADV low pulse.

Data latency versus NOR Flash latency
The data latency is the number of cycles to wait before sampling the data. The DATLAT
value must be consistent with the latency value specified in the NOR Flash configuration
register. The FSMC does not include the clock cycle when NADV is low in the data latency
count.
Caution:

Some NOR Flash memories include the NADV Low cycle in the data latency count, so the
exact relation between the NOR Flash latency and the FMSC DATLAT parameter can be
either of:
•

NOR Flash latency = (DATLAT + 2) CLK clock cycles

•

NOR Flash latency = (DATLAT + 3) CLK clock cycles

Some recent memories assert NWAIT during the latency phase. In such cases DATLAT can
be set to its minimum value. As a result, the FSMC samples the data and waits long enough
to evaluate if the data are valid. Thus the FSMC detects when the memory exits latency and
real data are taken.
Other memories do not assert NWAIT during latency. In this case the latency must be set
correctly for both the FSMC and the memory, otherwise invalid data are mistaken for good
data, or valid data are lost in the initial phase of the memory access.

Single-burst transfer
When the selected bank is configured in burst mode for synchronous accesses, if for
example an AHB single-burst transaction is requested on 16-bit memories, the FSMC
performs a burst transaction of length 1 (if the AHB transfer is 16-bit), or length 2 (if the AHB
transfer is 32-bit) and de-assert the chip select signal when the last data is strobed.
Clearly, such a transfer is not the most efficient in terms of cycles (compared to an
asynchronous read). Nevertheless, a random asynchronous access would first require to reprogram the memory access mode, which would altogether last longer.

Cross boundary page for Cellular RAM 1.5
Cellular RAM 1.5 does not allow burst access to cross the page boundary. The FSMC
controller allows to split automatically the burst access when the memory page size is
reached by configuring the CPSIZE bits in the FSMC_BCR1 register following the memory
page size.

Wait management
For synchronous NOR Flash memories, NWAIT is evaluated after the programmed latency
period, (DATLAT+2) CLK clock cycles.

RM0090 Rev 17

1569/1747
1599

Flexible static memory controller (FSMC)

RM0090

If NWAIT is sensed active (low level when WAITPOL = 0, high level when WAITPOL = 1),
wait states are inserted until NWAIT is sensed inactive (high level when WAITPOL = 0, low
level when WAITPOL = 1).
When NWAIT is inactive, the data is considered valid either immediately (bit WAITCFG = 1)
or on the next clock edge (bit WAITCFG = 0).
During wait-state insertion via the NWAIT signal, the controller continues to send clock
pulses to the memory, keeping the chip select and output enable signals valid, and does not
consider the data valid.
There are two timing configurations for the NOR Flash NWAIT signal in burst mode:
•

Flash memory asserts the NWAIT signal one data cycle before the wait state (default
after reset)

•

Flash memory asserts the NWAIT signal during the wait state

These two NOR Flash wait state configurations are supported by the FSMC, individually for
each chip select, thanks to the WAITCFG bit in the FSMC_BCRx registers (x = 0..3).
Figure 451. Wait configurations
-EMORY TRANSACTION  BURST OF  HALF WORDS
(#,+
#,+
!;=

ADDR;=

.!$6
.7!)4
7!)4#&'  
.7!)4
7!)4#&'  
INSERTED WAIT STATE
!$;=

AADDR;=

DATA

DATA

DATA
AIB

1570/1747

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)
Figure 452. Synchronous multiplexed read mode - NOR, PSRAM (CRAM)
0HPRU\WUDQVDFWLRQ EXUVWRIKDOIZRUGV
+&/.

&/.

$>@

DGGU>@

1([

12(

1:(

+LJK

1$'9
1:$,7
:$,7&)*

'$7/$7
&/.F\FOHV
$'>@

$GGU>@

LQVHUWHGZDLWVWDWH
GDWD

FORFN FORFN
F\FOH
F\FOH

GDWD

'DWDVWUREHV

GDWD

GDWD

'DWDVWUREHV

DLI

1. Byte lane outputs BL are not shown; for NOR access, they are held high, and, for PSRAM (CRAM) access,
they are held low.
2. NWAIT polarity is set to 0.

Table 242. FSMC_BCRx bit fields
Bit No.

Bit name

Value to set

31-20

Reserved

19

CBURSTRW

No effect on synchronous read

18-16

CPSIZE

As needed (0x1 for CRAM 1.5)

15

ASCYCWAIT

0x0

14

EXTMOD

0x0

13

WAITEN

Set to 1 if the memory supports this feature, otherwise keep at 0.

12

WREN

no effect on synchronous read

11

WAITCFG

to be set according to memory

10

WRAPMOD

9

WAITPOL

to be set according to memory

8

BURSTEN

0x1

7

Reserved

0x1

6

FACCEN

Set according to memory support (NOR Flash memory)

5-4

MWID

0x000

0x0

As needed

RM0090 Rev 17

1571/1747
1599

Flexible static memory controller (FSMC)

RM0090

Table 242. FSMC_BCRx bit fields (continued)
Bit No.

Bit name

Value to set

3-2

MTYP[0:1]

0x1 or 0x2

1

MUXEN

As needed

0

MBKEN

0x1

Table 243. FSMC_BTRx bit fields

1572/1747

Bit No.

Bit name

Value to set

31:30

Reserved

0x0

29:28

ACCMOD

0x0

27-24

DATLAT

Data latency

23-20

CLKDIV

0x0 to get CLK = HCLK (not supported)
0x1 to get CLK = 2 × HCLK
..

19-16

BUSTURN

15-8

DATAST

Don’t care

7-4

ADDHLD

Don’t care

3-0

ADDSET[3:0]

Don’t care

Time between NEx high to NEx low (BUSTURN HCLK)

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)
Figure 453. Synchronous multiplexed write mode - PSRAM (CRAM)
-EMORY TRANSACTION  BURST OF  HALF WORDS

(#,+

#,+

!;=

ADDR;=

.%X
(I :
./%

.7%

.!$6

.7!)4
7!)4#&'  
$!4,!4 
#,+ CYCLES
!$;=

!DDR;=

INSERTED WAIT STATE
DATA

DATA

 CLOCK  CLOCK

AIF

1. Memory must issue NWAIT signal one cycle in advance, accordingly WAITCFG must be programmed to 0.
2. NWAIT polarity is set to 0.
3. Byte Lane (NBL) outputs are not shown, they are held low while NEx is active.

Table 244. FSMC_BCRx bit fields
Bit No.

Bit name

Value to set

31-20

Reserved

19

CBURSTRW

18-16

CPSIZE

15

ASCYCWAIT

0x0

14

EXTMOD

0x0

13

WAITEN

Set to 1 if the memory supports this feature, otherwise keep at 0.

12

WREN

0x1

11

WAITCFG

0x0

10

WRAPMOD

0x0

0x000
0x1
As needed (0x1 for CRAM 1.5)

RM0090 Rev 17

1573/1747
1599

Flexible static memory controller (FSMC)

RM0090

Table 244. FSMC_BCRx bit fields (continued)
Bit No.

Bit name

Value to set

9

WAITPOL

to be set according to memory

8

BURSTEN

no effect on synchronous write

7

Reserved

0x1

6

FACCEN

Set according to memory support

5-4

MWID

3-2

MTYP[0:1]

1

MUXEN

As needed

0

MBKEN

0x1

As needed
0x1

Table 245. FSMC_BTRx bit fields

1574/1747

Bit No.

Bit name

Value to set

31:30

Reserved

0x0

29:28

ACCMOD

0x0

27-24

DATLAT

Data latency

23-20

CLKDIV

0x0 to get CLK = HCLK (not supported)
0x1 to get CLK = 2 × HCLK

19-16

BUSTURN

15-8

DATAST

Don’t care

7-4

ADDHLD

Don’t care

3-0

ADDSET[3:0]

Don’t care

Time between NEx high to NEx low (BUSTURN HCLK)

RM0090 Rev 17

RM0090

36.5.6

Flexible static memory controller (FSMC)

NOR/PSRAM control registers
The NOR/PSRAM control registers have to be accessed by words (32 bits).

SRAM/NOR-Flash chip-select control registers 1..4 (FSMC_BCR1..4)
Address offset: 0xA000 0000 + 8 * (x – 1), x = 1...4
Reset value: 0x0000 30DB for Bank1 and 0x0000 30D2 for Bank 2 to 4

BURSTEN

rw

rw

rw

rw

rw

rw

4

3

rw

2

rw

rw

rw

1

0
MBKEN

WAITPOL

rw

5

MUXEN

WRAPMOD

rw

6

MTYP[1:0]

WREN

WAITCFG

rw

7

MWID[1:0]

8

FACCEN

9

Reserved

14 13 12 11 10
WAITEN

rw

CPSIZE[2:0]

15

EXTMOD

Reserved

CBURSTRW

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

ASCYCWAIT

This register contains the control information of each memory bank, used for SRAMs,
PSRAM and NOR Flash memories.

rw

rw

Bits 31: 20 Reserved, must be kept at reset value.
Bit 19 CBURSTRW: Write burst enable.
For Cellular RAM (PSRAM) memories, this bit enables the synchronous burst protocol
during write operations. The enable bit for synchronous read accesses is the BURSTEN
bit in the FSMC_BCRx register.
0: Write operations are always performed in asynchronous mode
1: Write operations are performed in synchronous mode.
Bits 18: 16 CPSIZE[2:0]: CRAM page size.
These are used for Cellular RAM 1.5 which does not allow burst access to cross the
address boundaries between pages. When these bits are configured, the FSMC
controller splits automatically the burst access when the memory page size is reached
(refer to memory datasheet for page size).
000: No burst split when crossing page boundary (default after reset)
001: 128 bytes
010: 256 bytes
011: 512 bytes
100: 1024 bytes
Others: reserved.
Bit 15 ASYNCWAIT: Wait signal during asynchronous transfers
This bit enables/disables the FSMC to use the wait signal even during an asynchronous
protocol.
0: NWAIT signal is not taken into account when running an asynchronous protocol
(default after reset)
1: NWAIT signal is taken into account when running an asynchronous protocol

RM0090 Rev 17

1575/1747
1599

Flexible static memory controller (FSMC)

RM0090

Bit 14 EXTMOD: Extended mode enable.
This bit enables the FSMC to program the write timings for non-multiplexed
asynchronous accesses inside the FSMC_BWTR register, thus resulting in different
timings for read and write operations.
0: values inside FSMC_BWTR register are not taken into account (default after reset)
1: values inside FSMC_BWTR register are taken into account
Note: When the extended mode is disabled, the FSMC can operate in Mode1 or Mode2
as follows:
–
Mode 1 is the default mode when the SRAM/PSRAM memory type is selected
(MTYP [0:1]=0x0 or 0x01)
–
Mode 2 is the default mode when the NOR memory type is selected
(MTYP [0:1]= 0x10).
Bit 13 WAITEN: Wait enable bit.
This bit enables/disables wait-state insertion via the NWAIT signal when accessing the
Flash memory in synchronous mode.
0: NWAIT signal is disabled (its level not taken into account, no wait state inserted after
the programmed Flash latency period)
1: NWAIT signal is enabled (its level is taken into account after the programmed Flash
latency period to insert wait states if asserted) (default after reset)
Bit 12 WREN: Write enable bit.
This bit indicates whether write operations are enabled/disabled in the bank by the
FSMC:
0: Write operations are disabled in the bank by the FSMC, an AHB error is reported,
1: Write operations are enabled for the bank by the FSMC (default after reset).
Bit 11 WAITCFG: Wait timing configuration.
The NWAIT signal indicates whether the data from the memory are valid or if a wait state
must be inserted when accessing the Flash memory in synchronous mode. This
configuration bit determines if NWAIT is asserted by the memory one clock cycle before
the wait state or during the wait state:
0: NWAIT signal is active one data cycle before wait state (default after reset),
1: NWAIT signal is active during wait state (not used for PRAM).
Bit 10 WRAPMOD: Wrapped burst mode support.
Defines whether the controller will or not split an AHB burst wrap access into two linear
accesses. Valid only when accessing memories in burst mode
0: Direct wrapped burst is not enabled (default after reset),
1: Direct wrapped burst is enabled.
Note: This bit has no effect as the CPU and DMA cannot generate wrapping burst
transfers.
Bit 9 WAITPOL: Wait signal polarity bit.
Defines the polarity of the wait signal from memory. Valid only when accessing the
memory in burst mode:
0: NWAIT active low (default after reset),
1: NWAIT active high.
Bit 8 BURSTEN: Burst enable bit.
This bit enables/disables synchronous accesses during read operations. It is valid only
for synchronous memories operating in burst mode:
0: Burst mode disabled (default after reset). Read accesses are performed in
asynchronous mode.
1: Burst mode enable. Read accesses are performed in synchronous mode.

1576/1747

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)

Bit 7 Reserved, must be kept at reset value.
Bit 6 FACCEN: Flash access enable
Enables NOR Flash memory access operations.
0: Corresponding NOR Flash memory access is disabled
1: Corresponding NOR Flash memory access is enabled (default after reset)
Bits 5:4 MWID[1:0]: Memory databus width.
Defines the external memory device width, valid for all type of memories.
00: 8 bits,
01: 16 bits (default after reset),
10: reserved, do not use,
11: reserved, do not use.
Bits 3:2 MTYP[1:0]: Memory type.
Defines the type of external memory attached to the corresponding memory bank:
00: SRAM (default after reset for Bank 2...4)
01: PSRAM (CRAM)
10: NOR Flash/OneNAND Flash (default after reset for Bank 1)
11: reserved
Bit 1 MUXEN: Address/data multiplexing enable bit.
When this bit is set, the address and data values are multiplexed on the databus, valid
only with NOR and PSRAM memories:
0: Address/Data nonmultiplexed
1: Address/Data multiplexed on databus (default after reset)
Bit 0 MBKEN: Memory bank enable bit.
Enables the memory bank. After reset Bank1 is enabled, all others are disabled.
Accessing a disabled bank causes an ERROR on AHB bus.
0: Corresponding memory bank is disabled
1: Corresponding memory bank is enabled

RM0090 Rev 17

1577/1747
1599

Flexible static memory controller (FSMC)

RM0090

SRAM/NOR-Flash chip-select timing registers 1..4 (FSMC_BTR1..4)
Address offset: 0xA000 0000 + 0x04 + 8 * (x – 1), x = 1..4
Reset value: 0x0FFF FFFF
FSMC_BTRx bits are written by software to add a delay at the end of a read /write
transaction. This delay allows matching the minimum time between consecutive
transactions (tEHEL from NEx high to FSMC_NEx low) and the maximum time required by
the memory to free the data bus after a read access (tEHQZ).
This register contains the control information of each memory bank, used for SRAMs,
PSRAM and NOR Flash memories.If the EXTMOD bit is set in the FSMC_BCRx register,
then this register is partitioned for write and read access, that is, 2 registers are available:
one to configure read accesses (this register) and one to configure write accesses
(FSMC_BWTRx registers).

rw

rw

rw

rw

Bits 31:30

rw

rw

rw

rw

rw

rw

rw

rw

rw

8

7

6

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

5

4

3

2

rw

rw

rw

rw

rw

1

0

rw

rw

ADDSET[3:0]

ADDHLD[3:0]

9

DATAST[7:0]

BUSTURN[3:0]

CLKDIV[3:0]

DATLAT[3:0]

ACCMOD[1:0]

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

Reserved, must be kept at reset value.

Bits 29:28 ACCMOD[1:0]: Access mode
Specifies the asynchronous access modes as shown in the timing diagrams. These bits are
taken into account only when the EXTMOD bit in the FSMC_BCRx register is 1.
00: access mode A
01: access mode B
10: access mode C
11: access mode D
Bits 27:24 DATLAT[3:0]: Data latency for synchronous memory (see note below bit description table)
For synchronous accesses with read/write burst mode enabled (BURSTEN / CBURSTRW bits
set), this field defines the number of memory clock cycles (+2) to issue to the memory before
reading/writing the first data. This timing parameter is not expressed in HCLK periods, but in
FSMC_CLK periods. For asynchronous accesses, this value is don't care.
0000: Data latency of 2 CLK clock cycles for first burst access
1111: Data latency of 17 CLK clock cycles for first burst access (default value after reset)
Bits 23:20 CLKDIV[3:0]: Clock divide ratio (for FSMC_CLK signal)
Defines the period of FSMC_CLK clock output signal, expressed in number of HCLK cycles:
0000: Reserved
0001: FSMC_CLK period = 2 × HCLK periods
0010: FSMC_CLK period = 3 × HCLK periods
1111: FSMC_CLK period = 16 × HCLK periods (default value after reset)
In asynchronous NOR Flash, SRAM or PSRAM accesses, this value is don’t care.

1578/1747

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)

Bits 19:16 BUSTURN[3:0]: Bus turnaround phase duration
These bits are written by software to add a delay at the end of a write-to-read (and read-to
write) transaction. The programmed bus turnaround delay is inserted between an
asynchronous read (muxed or D mode) or a write transaction and any other
asynchronous/synchronous read or write to/from a static bank (for a read operation, the bank
can be the same or a different one; for a write operation, the bank can be different except in r
muxed or D mode).
In some cases, the bus turnaround delay is fixed, whatever the programmed BUSTURN
values:
– No bus turnaround delay is inserted between two consecutive asynchronous write transfers
to the same static memory bank except in muxed and D mode.
– A bus turnaround delay of 1 FSMC clock cycle is inserted between:
–
Two consecutive asynchronous read transfers to the same static memory bank
except for muxed and D modes.
–
An asynchronous read to an asynchronous or synchronous write to any static bank
or dynamic bank except for muxed and D modes.
–
An asynchronous (modes 1, 2, A, B or C) read and a read operation from another
static bank.
– A bus turnaround delay of 2 FSMC clock cycles is inserted between:
–
Two consecutive synchronous write accesses (in burst or single mode) to the same
bank
–
A synchronous write (burst or single) access and an asynchronous write or read
transfer to or from static memory bank (the bank can be the same or different in case
of a read operation).
–
Two consecutive synchronous read accesses (in burst or single mode) followed by a
any synchronous/asynchronous read or write from/to another static memory bank.
– A bus turnaround delay of 3 FSMC clock cycles is inserted between:
–
Two consecutive synchronous write operations (in burst or single mode) to different
static banks.
–
A synchronous write access (in burst or single mode) and a synchronous read
access from the same or to a different bank.
0000: BUSTURN phase duration = 0 HCLK clock cycle added
...
1111: BUSTURN phase duration = 15 × HCLK clock cycles (default value after reset)

RM0090 Rev 17

1579/1747
1599

Flexible static memory controller (FSMC)

RM0090

Bits 15:8 DATAST[7:0]: Data-phase duration
These bits are written by software to define the duration of the data phase (refer to Figure 436
to Figure 448), used in asynchronous accesses:
0000 0000: Reserved
0000 0001: DATAST phase duration = 1 × HCLK clock cycles
0000 0010: DATAST phase duration = 2 × HCLK clock cycles
...
1111 1111: DATAST phase duration = 255 × HCLK clock cycles (default value after reset)
For each memory type and access mode data-phase duration, refer to the respective figure
(Figure 436 to Figure 448).
Example: Mode1, write access, DATAST=1: Data-phase duration= DATAST+1 = 2 HCLK clock
cycles.
Note: In synchronous accesses, this value is don't care.
Bits 7:4 ADDHLD[3:0]: Address-hold phase duration
These bits are written by software to define the duration of the address hold phase (refer to
Figure 445 to Figure 448), used in mode D and multiplexed accesses:
0000: Reserved
0001: ADDHLD phase duration =1 × HCLK clock cycle
0010: ADDHLD phase duration = 2 × HCLK clock cycle
...
1111: ADDHLD phase duration = 15 × HCLK clock cycles (default value after reset)
For each access mode address-hold phase duration, refer to the respective figure (Figure 445
to Figure 448).
Note: In synchronous accesses, this value is not used, the address hold phase is always 1
memory clock period duration.
Bits 3:0 ADDSET[3:0]: Address setup phase duration
These bits are written by software to define the duration of the address setup phase (refer to
Figure 436 to Figure 448), used in SRAMs, ROMs and asynchronous NOR Flash and PSRAM
accesses:
0000: ADDSET phase duration = 0 × HCLK clock cycle
...
1111: ADDSET phase duration = 15 × HCLK clock cycles (default value after reset)
For each access mode address setup phase duration, refer to the respective figure (refer to
Figure 436 to Figure 448).
Note: In synchronous NOR Flash and PSRAM accesses, this value is don’t care.

Note:

1580/1747

PSRAMs (CRAMs) have a variable latency due to internal refresh. Therefore these
memories issue the NWAIT signal during the whole latency phase to prolong the latency as
needed.
With PSRAMs (CRAMs) the DATLAT field must be set to 0, so that the FSMC exits its
latency phase soon and starts sampling NWAIT from memory, then starts to read or write
when the memory is ready.
This method can be used also with the latest generation of synchronous Flash memories
that issue the NWAIT signal, unlike older Flash memories (check the datasheet of the
specific Flash memory being used).

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)

SRAM/NOR-Flash write timing registers 1..4 (FSMC_BWTR1..4)
Address offset: 0xA000 0000 + 0x104 + 8 * (x – 1), x = 1...4
Reset value: 0x0FFF FFFF
This register contains the control information of each memory bank, used for SRAMs,
PSRAMs and NOR Flash memories. This register is active for write asynchronous access
only when the EXTMOD bit is set in the FSMC_BCRx register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Res.

ACCM
OD[2:0]
rw

rw

Bits 31:30

Reserved

BUSTURN[3:0]
rw

rw

rw

rw

9

8

DATAST[7:0]
rw

rw

rw

rw

rw

rw

7

6

5

4

ADDHLD[3:0]
rw

rw

rw

rw

rw

rw

3

2

1

0

ADDSET[3:0]
rw

rw

rw

rw

Reserved, must be kept at reset value.

Bits 29:28 ACCMOD[2:0]: Access mode.
Specifies the asynchronous access modes as shown in the next timing diagrams.These bits are
taken into account only when the EXTMOD bit in the FSMC_BCRx register is 1.
00: access mode A
01: access mode B
10: access mode C
11: access mode D
Bits 27:20

Reserved, must be kept at reset value.

Bits 19:16 BUSTURN[3:0]: Bus turnaround phase duration
The programmed bus turnaround delay is inserted between a an asynchronous write transfer and
any other asynchronous/synchronous read or write transfer to/from a static bank (for a read
operation, the bank can be the same or a different one; for a write operation, the bank can be
different except in r muxed or D mode).
In some cases, the bus turnaround delay is fixed, whatever the programmed BUSTURN values:
– No bus turnaround delay is inserted between two consecutive asynchronous write transfers to the
same static memory bank except in muxed and D mode.
– A bus turnaround delay of 2 FSMC clock cycles is inserted between:
–
Two consecutive synchronous write accesses (in burst or single mode) to the same bank.
–
A synchronous write transfer (in burst or single mode) and an asynchronous write or read
transfer to/from static a memory bank.
– A bus turnaround delay of 3 FSMC clock cycles is inserted between:
–
Two consecutive synchronous write accesses (in burst or single mode) to different static
banks.
–
A synchronous write transfer (in burst or single mode) and a synchronous read from the
same or from a different bank.
0000: BUSTURN phase duration = 0 HCLK clock cycle added
...
1111: BUSTURN phase duration = 15 HCLK clock cycles added (default value after reset)

RM0090 Rev 17

1581/1747
1599

Flexible static memory controller (FSMC)

RM0090

Bits 15:8 DATAST[7:0]: Data-phase duration.
These bits are written by software to define the duration of the data phase (refer to Figure 436 to
Figure 448), used in asynchronous SRAM, PSRAM and NOR Flash memory accesses:
0000 0000: Reserved
0000 0001: DATAST phase duration = 1 × HCLK clock cycles
0000 0010: DATAST phase duration = 2 × HCLK clock cycles
...
1111 1111: DATAST phase duration = 255 × HCLK clock cycles (default value after reset)
Note: In synchronous accesses, this value is don't care.
Bits 7:4 ADDHLD[3:0]: Address-hold phase duration.
These bits are written by software to define the duration of the address hold phase (refer to
Figure 445 to Figure 448), used in asynchronous multiplexed accesses:
0000: Reserved
0001: ADDHLD phase duration = 1 × HCLK clock cycle
0010: ADDHLD phase duration = 2 × HCLK clock cycle
...
1111: ADDHLD phase duration = 15 × HCLK clock cycles (default value after reset)
Note: In synchronous NOR Flash accesses, this value is not used, the address hold phase is always
1 Flash clock period duration.
Bits 3:0 ADDSET[3:0]: Address setup phase duration.
These bits are written by software to define the duration of the address setup phase in HCLK cycles
(refer to Figure 445 to Figure 448), used in asynchronous accessed:
0000: ADDSET phase duration = 0 × HCLK clock cycle
...
1111: ADDSET phase duration = 15 × HCLK clock cycles (default value after reset)
Note: In synchronous NOR Flash and PSRAM accesses, this value is don’t care.

36.6

NAND Flash/PC Card controller
The FSMC generates the appropriate signal timings to drive the following types of device:
•

•

NAND Flash
–

8-bit

–

16-bit

16-bit PC Card compatible devices

The NAND/PC Card controller can control three external banks. Bank 2 and bank 3 support
NAND Flash devices. Bank 4 supports PC Card devices.
Each bank is configured by means of dedicated registers (Section 36.6.8). The
programmable memory parameters include access timings (shown in Table 246) and ECC
configuration.

1582/1747

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)
Table 246. Programmable NAND/PC Card access parameters
Parameter

36.6.1

Function

Access mode

Unit

Min. Max.

Memory setup
time

Number of clock cycles (HCLK)
to set up the address before the
command assertion

Read/Write

AHB clock cycle
(HCLK)

1

255

Memory wait

Minimum duration (HCLK clock
Read/Write
cycles) of the command assertion

AHB clock cycle
(HCLK)

2

256

Memory hold

Number of clock cycles (HCLK)
to hold the address (and the data
Read/Write
in case of a write access) after
the command de-assertion

AHB clock cycle
(HCLK)

1

254

Memory
databus high-Z

Number of clock cycles (HCLK)
during which the databus is kept
in high-Z state after the start of a
write access

AHB clock cycle
(HCLK)

0

255

Write

External memory interface signals
The following tables list the signals that are typically used to interface NAND Flash and PC
Card.

Note:

Prefix “N”. specifies the associated signal as active low.

8-bit NAND Flash
t

Table 247. 8-bit NAND Flash
FSMC signal name

I/O

Function

A[17]

O

NAND Flash address latch enable (ALE) signal

A[16]

O

NAND Flash command latch enable (CLE) signal

D[7:0]

I/O

8-bit multiplexed, bidirectional address/data bus

NCE[x]

O

Chip select, x = 2, 3

NOE(= NRE)

O

Output enable (memory signal name: read enable, NRE)

NWE

O

Write enable

NWAIT/INT[3:2]

I

NAND Flash ready/busy input signal to the FSMC

There is no theoretical capacity limitation as the FSMC can manage as many address
cycles as needed.

RM0090 Rev 17

1583/1747
1599

Flexible static memory controller (FSMC)

RM0090

16-bit NAND Flash
Table 248. 16-bit NAND Flash
FSMC signal name

I/O

Function

A[17]

O

NAND Flash address latch enable (ALE) signal

A[16]

O

NAND Flash command latch enable (CLE) signal

D[15:0]

I/O

16-bit multiplexed, bidirectional address/data bus

NCE[x]

O

Chip select, x = 2, 3

NOE(= NRE)

O

Output enable (memory signal name: read enable, NRE)

NWE

O

Write enable

NWAIT/INT[3:2]

I

NAND Flash ready/busy input signal to the FSMC

There is no theoretical capacity limitation as the FSMC can manage as many address
cycles as needed.

16-bit PC Card
Table 249. 16-bit PC Card

1584/1747

FSMC signal name

I/O

Function

A[10:0]

O

Address bus

NIORD

O

Output enable for I/O space

NIOWR

O

Write enable for I/O space

NREG

O

Register signal indicating if access is in Common or Attribute space

D[15:0]

I/O

Bidirectional databus

NCE4_1

O

Chip select 1

NCE4_2

O

Chip select 2 (indicates if access is 16-bit or 8-bit)

NOE

O

Output enable in Common and in Attribute space

NWE

O

Write enable in Common and in Attribute space

NWAIT

I

PC Card wait input signal to the FSMC (memory signal name
IORDY)

INTR

I

PC Card interrupt to the FSMC (only for PC Cards that can generate
an interrupt)

CD

I

PC Card presence detection. Active high. If an access is performed
to the PC Card banks while CD is low, an AHB error is generated.
Refer to Section 36.3: AHB interface

RM0090 Rev 17

RM0090

36.6.2

Flexible static memory controller (FSMC)

NAND Flash / PC Card supported memories and transactions
Table 250 below shows the supported devices, access modes and transactions.
Transactions not allowed (or not supported) by the NAND Flash / PC Card controller appear
in gray.
Table 250. Supported memories and transactions
Device

NAND 8-bit

NAND 16-bit

36.6.3

Mode

R/W

AHB
Memory
Allowed/
data size data size not allowed

Comments

Asynchronous R

8

8

Y

Asynchronous W

8

8

Y

Asynchronous R

16

8

Y

Split into 2 FSMC accesses

Asynchronous W

16

8

Y

Split into 2 FSMC accesses

Asynchronous R

32

8

Y

Split into 4 FSMC accesses

Asynchronous W

32

8

Y

Split into 4 FSMC accesses

Asynchronous R

8

16

Y

Asynchronous W

8

16

N

Asynchronous R

16

16

Y

Asynchronous W

16

16

Y

Asynchronous R

32

16

Y

Split into 2 FSMC accesses

Asynchronous W

32

16

Y

Split into 2 FSMC accesses

Timing diagrams for NAND and PC Card
Each PC Card/CompactFlash and NAND Flash memory bank is managed through a set of
registers:
•

Control register: FSMC_PCRx

•

Interrupt status register: FSMC_SRx

•

ECC register: FSMC_ECCRx

•

Timing register for Common memory space: FSMC_PMEMx

•

Timing register for Attribute memory space: FSMC_PATTx

•

Timing register for I/O space: FSMC_PIOx

Each timing configuration register contains three parameters used to define number of
HCLK cycles for the three phases of any PC Card/CompactFlash or NAND Flash access,
plus one parameter that defines the timing for starting driving the databus in the case of a
write. Figure 454 shows the timing parameter definitions for common memory accesses,
knowing that Attribute and I/O (only for PC Card) memory space access timings are similar.

RM0090 Rev 17

1585/1747
1599

Flexible static memory controller (FSMC)

RM0090

Figure 454. NAND/PC Card controller timing for common memory access

(#,+

!;=

.#%X
.2%'
.)/7
.)/2

(IGH

-%-X3%4



-%-X7!)4



-%-X(/,$



.7%
./%
-%-X():
WRITE?DATA

READ?DATA

6ALID

IB

1. NOE remains high (inactive) during write access. NWE remains high (inactive) during read access.
2. For write accesses, the hold phase delay is (MEMHOLD) x HCLK cycles, while it is (MEMHOLD + 2) x
HCLK cycles for read accesses.

36.6.4

NAND Flash operations
The command latch enable (CLE) and address latch enable (ALE) signals of the NAND
Flash device are driven by some address signals of the FSMC controller. This means that to
send a command or an address to the NAND Flash memory, the CPU has to perform a write
to a certain address in its memory space.
A typical page read operation from the NAND Flash device is as follows:

1586/1747

1.

Program and enable the corresponding memory bank by configuring the FSMC_PCRx
and FSMC_PMEMx (and for some devices, FSMC_PATTx, see Section 36.6.5)
registers according to the characteristics of the NAND Flash (PWID bits for the databus
width of the NAND Flash, PTYP = 1, PWAITEN = 0 or 1 as needed, see Common
memory space timing register 2..4 (FSMC_PMEM2..4) for timing configuration).

2.

The CPU performs a byte write in the common memory space, with data byte equal to
one Flash command byte (for example 0x00 for Samsung NAND Flash devices). The
CLE input of the NAND Flash is active during the write strobe (low pulse on NWE), thus
the written byte is interpreted as a command by the NAND Flash. Once the command
is latched by the NAND Flash device, it does not need to be written for the following
page read operations.

3.

The CPU can send the start address (STARTAD) for a read operation by writing the
required bytes (for example four bytes or three for smaller capacity devices),
STARTAD[7:0], STARTAD[15:8], STARTAD[23:16] and finally STARTAD[25:24] for
64 Mb x 8 bit NAND Flash) in the common memory or attribute space. The ALE input of
the NAND Flash device is active during the write strobe (low pulse on NWE), thus the

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)
written bytes are interpreted as the start address for read operations. Using the
attribute memory space makes it possible to use a different timing configuration of the
FSMC, which can be used to implement the prewait functionality needed by some
NAND Flash memories (see details in Section 36.6.5).

36.6.5

4.

The controller waits for the NAND Flash to be ready (R/NB signal high) to become
active, before starting a new access (to same or another memory bank). While waiting,
the controller maintains the NCE signal active (low).

5.

The CPU can then perform byte read operations in the common memory space to read
the NAND Flash page (data field + Spare field) byte by byte.

6.

The next NAND Flash page can be read without any CPU command or address write
operation, in three different ways:
–

by simply performing the operation described in step 5

–

a new random address can be accessed by restarting the operation at step 3

–

a new command can be sent to the NAND Flash device by restarting at step 2

NAND Flash prewait functionality
Some NAND Flash devices require that, after writing the last part of the address, the
controller wait for the R/NB signal to go low as shown in Figure 455.
Figure 455. Access to non ‘CE don’t care’ NAND-Flash
.#% MUST STAY LOW
.#%

#,%

!,%

.7%
(IGH
./%
T2
)/;=

X

! !

! !

! !

! !

T7"
2."









AIB

1. CPU wrote byte 0x00 at address 0x7001 0000.
2. CPU wrote byte A7-A0 at address 0x7002 0000.
3. CPU wrote byte A15-A8 at address 0x7002 0000.
4. CPU wrote byte A23-A16 at address 0x7002 0000.
5. CPU wrote byte A25-A24 at address 0x7802 0000: FSMC performs a write access using FSMC_PATT2
timing definition, where ATTHOLD ≥ 7 (providing that (7+1) × HCLK = 112 ns > tWB max). This guarantees
that NCE remains low until R/NB goes low and high again (only requested for NAND Flash memories
where NCE is not don’t care).

RM0090 Rev 17

1587/1747
1599

Flexible static memory controller (FSMC)

RM0090

When this functionality is needed, it can be guaranteed by programming the MEMHOLD
value to meet the tWB timing. However CPU read accesses to the NAND Flash memory has
a hold delay of (MEMHOLD + 2) x HCLK cycles, while CPU write accesses have a hold
delay of (MEMHOLD) x HCLK cycles.
To overcome this timing constraint, the attribute memory space can be used by
programming its timing register with an ATTHOLD value that meets the tWB timing, and
leaving the MEMHOLD value at its minimum. Then, the CPU must use the common memory
space for all NAND Flash read and write accesses, except when writing the last address
byte to the NAND Flash device, where the CPU must write to the attribute memory space.

36.6.6

Computation of the error correction code (ECC)
in NAND Flash memory
The FSMC PC-Card controller includes two error correction code computation hardware
blocks, one per memory bank. They are used to reduce the host CPU workload when
processing the error correction code by software in the system.
These two registers are identical and associated with bank 2 and bank 3, respectively. As a
consequence, no hardware ECC computation is available for memories connected to bank
4.
The error correction code (ECC) algorithm implemented in the FSMC can perform 1-bit error
correction and 2-bit error detection per 256, 512, 1 024, 2 048, 4 096 or 8 192 bytes read
from or written to NAND Flash memory. It is based on the Hamming coding algorithm and
consists in calculating the row and column parity.
The ECC modules monitor the NAND Flash databus and read/write signals (NCE and NWE)
each time the NAND Flash memory bank is active.
The functional operations are:
•

When access to NAND Flash is made to bank 2 or bank 3, the data present on the
D[15:0] bus is latched and used for ECC computation.

•

When access to NAND Flash occurs at any other address, the ECC logic is idle, and
does not perform any operation. Thus, write operations for defining commands or
addresses to NAND Flash are not taken into account for ECC computation.

Once the desired number of bytes has been read from/written to the NAND Flash by the
host CPU, the FSMC_ECCR2/3 registers must be read in order to retrieve the computed
value. Once read, they should be cleared by resetting the ECCEN bit to zero. To compute a
new data block, the ECCEN bit must be set to one in the FSMC_PCR2/3 registers.
To perform an ECC computation:

1588/1747

RM0090 Rev 17

RM0090

36.6.7

Flexible static memory controller (FSMC)
1.

Enable the ECCEN bit in the FSMC_PCR2/3 register.

2.

Write data to the NAND Flash memory page. While the NAND page is written, the ECC
block computes the ECC value.

3.

Read the ECC value available in the FSMC_ECCR2/3 register and store it in a
variable.

4.

Clear the ECCEN bit and then enable it in the FSMC_PCR2/3 register before reading
back the written data from the NAND page. While the NAND page is read, the ECC
block computes the ECC value.

5.

Read the new ECC value available in the FSMC_ECCR2/3 register.

6.

If the two ECC values are the same, no correction is required, otherwise there is an
ECC error and the software correction routine returns information on whether the error
can be corrected or not.

PC Card/CompactFlash operations
Address spaces and memory accesses
The FSMC supports Compact Flash storage or PC Cards in Memory Mode and I/O Mode
(True IDE mode is not supported).
The Compact Flash storage and PC Cards are made of 3 memory spaces:
•

Common Memory Space

•

Attribute Space

•

I/O Memory Space

The nCE2 and nCE1 pins (FSMC_NCE4_2 and FSMC_NCE4_1 respectively) select the
card and indicate whether a byte or a word operation is being performed: nCE2 accesses
the odd byte on D15-8 and nCE1 accesses the even byte on D7-0 if A0=0 or the odd byte on
D7-0 if A0=1. The full word is accessed on D15-0 if both nCE2 and nCE1 are low.
The memory space is selected by asserting low nOE for read accesses or nWE for write
accesses, combined with the low assertion of nCE2/nCE1 and nREG.
•

If pin nREG=1 during the memory access, the common memory space is selected

•

If pin nREG=0 during the memory access, the attribute memory space is selected

The I/O Space is selected by asserting low nIORD for read accesses or nIOWR for write
accesses [instead of nOE/nWE for memory Space], combined with nCE2/nCE1. Note that
nREG must also be asserted low during accesses to I/O Space.
Three type of accesses are allowed for a 16-bit PC Card:
•

Accesses to Common Memory Space for data storage can be either 8-bit accesses at
even addresses or 16 bit AHB accesses.
Note that 8-bit accesses at odd addresses are not supported and will not lead to the
low assertion of nCE2. A 32-bit AHB request is translated into two 16-bit memory
accesses.

•

Accesses to Attribute Memory Space where the PC Card stores configuration
information are limited to 8-bit AHB accesses at even addresses.
Note that a 16-bit AHB access will be converted into a single 8-bit memory transfer:
nCE1 will be asserted low, nCE2 will be asserted high and only the even Byte on D7D0 will be valid. Instead a 32-bit AHB access will be converted into two 8-bit memory

RM0090 Rev 17

1589/1747
1599

Flexible static memory controller (FSMC)

RM0090

transfers at even addresses: nCE1 will be asserted low, NCE2 will be asserted high
and only the even bytes will be valid.
•

Accesses to I/O Space can be performed either through AHB 8-bit or 16-bit accesses.

nCE2

nCE1

nREG

nOE/nWE

nIORD /nIOWR

A10

A9

A7-1

A0

Table 251. 16-bit PC-Card signals and access type

1

0

1

0

1

X

X

X-X

X

0

1

1

0

1

X

X

X-X

X

0

0

1

0

1

X

X

X-X

0

X

0

0

0

1

0

1

X-X

0

X

0

0

0

1

0

0

X-X

0

1

0

0

0

1

X

X

X-X

1

0

1

0

0

1

X

X

X-X

x

1

0

0

1

0

X

X

X-X

1

0

0

1

0

X

X

1

0

0

1

0

X

1

0

0

1

0

0

0

0

1

0

0

0

0

1

0

1

Space

Access Type

Allowed/not
Allowed

Read/Write byte on D7-D0

YES

Read/Write byte on D15-D8

Not supported

Read/Write word on D15-D0

YES

Read or Write Configuration
Registers

YES

Read or Write CIS (Card
Information Structure)

YES

Invalid Read or Write (odd
address)

YES

Invalid Read or Write (odd
address)

YES

0

Read Even Byte on D7-0

YES

X-X

1

Read Odd Byte on D7-0

YES

X

X-X

0

Write Even Byte on D7-0

YES

X

X

X-X

1

Write Odd Byte on D7-0

YES

0

X

X

X-X

0

Read Word on D15-0

YES

1

0

X

X

X-X

0

Write word on D15-0

YES

0

1

0

X

X

X-X

X

Read Odd Byte on D15-8

Not supported

0

1

0

X

X

X-X

X

Write Odd Byte on D15-8

Not supported

Common
Memory
Space

Attribute
Space

Attribute
Space

I/O space

The FSMC Bank 4 gives access to those 3 memory spaces as described in Section 36.4.2:
NAND/PC Card address mapping and Table 218: Memory mapping and timing registers.

Wait Feature
The CompactFlash Storage or PC Card may request the FSMC to extend the length of the
access phase programmed by MEMWAITx/ATTWAITx/IOWAITx bits, asserting the nWAIT
signal after nOE/nWE or nIORD/nIOWR activation if the wait feature is enabled through the
PWAITEN bit in the FSMC_PCRx register. In order to detect the nWAIT assertion correctly,
the MEMWAITx/ATTWAITx/IOWAITx bits must be programmed as follows:

1590/1747

RM0090 Rev 17

RM0090

Flexible static memory controller (FSMC)
xxWAITx >= 4 + max_wait_assertion_time/HCLK
Where max_wait_assertion_time is the maximum time taken by NWAIT to go low once
nOE/nWE or nIORD/nIOWR is low.
After the de-assertion of nWAIT, the FSMC extends the WAIT phase for 4 HCLK clock
cycles.

36.6.8

NAND Flash/PC Card control registers
The NAND Flash/PC Card control registers have to be accessed by words (32 bits).

PC Card/NAND Flash control registers 2..4 (FSMC_PCR2..4)
Address offset: 0xA0000000 + 0x40 + 0x20 * (x – 1), x = 2..4

rw

rw

rw

rw

rw

rw

rw

rw

rw

Res.

rw

5

4

rw

rw

3

2

1

rw

rw

rw

0
Reserved

rw

TCLR[2:0]

6

PBKEN

rw

TAR[2:0]

7

PWAITEN

ECCPS[2:0]

8

PTYP

Reserved

9

PWID[1:0]

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

ECCEN

Reset value: 0x0000 0018

Bits 31:20 Reserved, must be kept at reset value.
Bits 19:17 ECCPS[2:0]: ECC page size.
Defines the page size for the extended ECC:
000: 256 bytes
001: 512 bytes
010: 1024 bytes
011: 2048 bytes
100: 4096 bytes
101: 8192 bytes
Bits 16:13 TAR[2:0]: ALE to RE delay.
Sets time from ALE low to RE low in number of AHB clock cycles (HCLK).
Time is: t_ar = (TAR + SET + 2) × THCLK where THCLK is the HCLK clock period
0000: 1 HCLK cycle (default)
1111: 16 HCLK cycles
Note: SET is MEMSET or ATTSET according to the addressed space.
Bits 12:9 TCLR[2:0]: CLE to RE delay.
Sets time from CLE low to RE low in number of AHB clock cycles (HCLK).
Time is t_clr = (TCLR + SET + 2) × THCLK where THCLK is the HCLK clock period
0000: 1 HCLK cycle (default)
1111: 16 HCLK cycles
Note: SET is MEMSET or ATTSET according to the addressed space.
Bits 8:7 Reserved, must be kept at reset value.
Bit 6 ECCEN: ECC computation logic enable bit
0: ECC logic is disabled and reset (default after reset),
1: ECC logic is enabled.

RM0090 Rev 17

1591/1747
1599

Flexible static memory controller (FSMC)

RM0090

Bits 5:4 PWID[1:0]: Databus width.
Defines the external memory device width.
00: 8 bits
01: 16 bits (default after reset). This value is mandatory for PC Cards.
10: reserved, do not use
11: reserved, do not use
Bit 3 PTYP: Memory type.
Defines the type of device attached to the corresponding memory bank:
0: PC Card, CompactFlash, CF+ or PCMCIA
1: NAND Flash (default after reset)
Bit 2 PBKEN: PC Card/NAND Flash memory bank enable bit.
Enables the memory bank. Accessing a disabled memory bank causes an ERROR on AHB
bus
0: Corresponding memory bank is disabled (default after reset)
1: Corresponding memory bank is enabled
Bit 1 PWAITEN: Wait feature enable bit.
Enables the Wait feature for the PC Card/NAND Flash memory bank:
0: disabled
1: enabled
Note: For a PC Card, when the wait feature is enabled, the MEMWAITx/ATTWAITx/IOWAITx
bits must be programmed to a value as follows:
xxWAITx ≥ 4 + max_wait_assertion_time/HCLK
Where max_wait_assertion_time is the maximum time taken by NWAIT to go low once
nOE/nWE or nIORD/nIOWR is low.
Bit 0

Reserved, must be kept at reset value.

FIFO status and interrupt register 2..4 (FSMC_SR2..4)
Address offset: 0xA000 0000 + 0x44 + 0x20 * (x-1), x = 2..4
Reset value: 0x0000 0040

1592/1747

RM0090 Rev 17

6

5

4

3

2

1

0

IFS

ILS

IRS

7

ILEN

8

IREN

9

IFEN

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

FEMPT

This register contains information about FIFO status and interrupt. The FSMC has a FIFO
that is used when writing to memories to store up to16 words of data from the AHB.
This is used to quickly write to the AHB and free it for transactions to peripherals other than
the FSMC, while the FSMC is draining its FIFO into the memory. This register has one of its
bits that indicates the status of the FIFO, for ECC purposes.
The ECC is calculated while the data are written to the memory, so in order to read the
correct ECC the software must wait until the FIFO is empty.

r

rw

rw

rw

rw

rw

rw

RM0090

Flexible static memory controller (FSMC)

Bits 31:7

Reserved, must be kept at reset value.

Bit 6

FEMPT: FIFO empty.
Read-only bit that provides the status of the FIFO
0: FIFO not empty
1: FIFO empty

Bit 5

IFEN: Interrupt falling edge detection enable bit
0: Interrupt falling edge detection request disabled
1: Interrupt falling edge detection request enabled

Bit 4

ILEN: Interrupt high-level detection enable bit
0: Interrupt high-level detection request disabled
1: Interrupt high-level detection request enabled

Bit 3

IREN: Interrupt rising edge detection enable bit
0: Interrupt rising edge detection request disabled
1: Interrupt rising edge detection request enabled

Bit 2

IFS: Interrupt falling edge status
The flag is set by hardware and reset by software.
0: No interrupt falling edge occurred
1: Interrupt falling edge occurred

Note: This bit is set by programming it to 1 by software.
Bit 1

ILS: Interrupt high-level status
The flag is set by hardware and reset by software.
0: No Interrupt high-level occurred
1: Interrupt high-level occurred

Bit 0

IRS: Interrupt rising edge status
The flag is set by hardware and reset by software.
0: No interrupt rising edge occurred
1: Interrupt rising edge occurred

Note: This bit is set by programming it to 1 by software.

Common memory space timing register 2..4 (FSMC_PMEM2..4)
Address offset: Address: 0xA000 0000 + 0x48 + 0x20 * (x – 1), x = 2..4
Reset value: 0xFCFC FCFC
Each FSMC_PMEMx (x = 2..4) read/write register contains the timing information for PC
Card or NAND Flash memory bank x, used for access to the common memory space of the
16-bit PC Card/CompactFlash, or to access the NAND Flash for command, address write
access and data read/write access.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
MEMHIZ[7:0]
rw

rw

rw

rw

rw

rw

MEMHOLD[7:0]
rw

rw

rw

rw

rw

rw

rw

9

8

7

6

5

rw

rw

rw

rw

rw

MEMWAIT[7:0]

rw

rw

rw

rw

rw

rw

RM0090 Rev 17

rw

rw

rw

4

3

2

1

0

rw

rw

MEMSET[7:0]
rw

rw

rw

1593/1747
1599

Flexible static memory controller (FSMC)

RM0090

Bits 31:24 MEMHIZx[7:0]: Common memory x databus HiZ time
Defines the number of HCLK clock cycles during which the databus is kept in HiZ after the
start of a PC Card/NAND Flash write access to common memory space on socket x. Only
valid for write transaction:
0000 0000: 1 HCLK cycle
1111 1110: 255 HCLK cycles
1111 1111: Reserved
Bits 23:16 MEMHOLDx[7:0]: Common memory x hold time
For NAND Flash read accesses to the common memory space, these bits define the
number of (HCLK+2) clock cycles during which the address is held after the command is
deasserted (NWE, NOE).
For NAND Flash write accesses to the common memory space, these bits define the
number of HCLK clock cycles during which the data are held after the command is
deasserted (NWE, NOE).
0000 0000: Reserved
0000 0001: 1 HCLK cycle for write accesses, 3 HCLK cycles for read accesses
1111 1110: 254 HCLK cycle for write accesses, 256 HCLK cycles for read accesses
1111 1111: Reserved
Bits 15:8 MEMWAITx[7:0]: Common memory x wait time
Defines the minimum number of HCLK (+1) clock cycles to assert the command (NWE,
NOE), for PC Card/NAND Flash read or write access to common memory space on socket
x. The duration for command assertion is extended if the wait signal (NWAIT) is active (low)
at the end of the programmed value of HCLK:
0000 0000: Reserved
0000 0001: 2 HCLK cycles (+ wait cycle introduced by deasserting NWAIT)
1111 1110: 255 HCLK cycles (+ wait cycle introduced by deasserting NWAIT)
1111 1111: Reserved.
Bits 7:0 MEMSETx[7:0]: Common memory x setup time
Defines the number of HCLK () clock cycles to set up the address before the command
assertion (NWE, NOE), for PC Card/NAND Flash read or write access to common memory
space on socket x:
0000 0000: 1 HCLK cycle
1111 1110: 255 HCLK cycles
1111 1111: Reserved

Attribute memory space timing registers 2..4 (FSMC_PATT2..4)
Address offset: 0xA000 0000 + 0x4C + 0x20 * (x – 1), x = 2..4
Reset value: 0xFCFC FCFC
Each FSMC_PATTx (x = 2..4) read/write register contains the timing information for PC
Card/CompactFlash or NAND Flash memory bank x. It is used for 8-bit accesses to the
attribute memory space of the PC Card/CompactFlash or to access the NAND Flash for the
last address write access if the timing must differ from that of previous accesses (for
Ready/Busy management, refer to Section 36.6.5: NAND Flash prewait functionality).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
ATTHIZ[7:0]
rw

rw

rw

1594/1747

rw

rw

rw

ATTHOLD[7:0]
rw

rw

rw

rw

rw

rw

rw

9

8

7

6

ATTWAIT[7:0]

rw

rw

rw

rw

rw

rw

RM0090 Rev 17

rw

rw

rw

5

4

3

2

1

0

rw

rw

ATTSET[7:0]
rw

rw

rw

rw

rw

rw

rw

rw

RM0090

Flexible static memory controller (FSMC)

Bits 31:24 ATTHIZ[7:0]: Attribute memory x databus HiZ time
Defines the number of HCLK clock cycles during which the databus is kept in HiZ after the
start of a PC CARD/NAND Flash write access to attribute memory space on socket x. Only
valid for write transaction:
0000 0000: 0 HCLK cycle
1111 1110: 255 HCLK cycles
1111 1111: Reserved.
Bits 23:16 ATTHOLD[7:0]: Attribute memory x hold time
For PC Card/NAND Flash read accesses to attribute memory space on socket x, these bits
define the number of HCLK clock cycles (HCLK +2) clock cycles during which the address is
held after the command is deasserted (NWE, NOE).
For PC Card/NAND Flash write accesses to attribute memory space on socket x, these bits
define the number of HCLK clock cycles during which the data are held after the command
is deasserted (NWE, NOE).
0000 0000: reserved
0000 0001: 1 HCLK cycle for write access, 3 HCLK cycles for read accesses
1111 1110: 254 HCLK cycle for write access, 256 HCLK cycles for read accesses
1111 1111: Reserved
Bits 15:8 ATTWAIT[7:0]: Attribute memory x wait time
Defines the minimum number of HCLK (+1) clock cycles to assert the command (NWE,
NOE), for PC Card/NAND Flash read or write access to attribute memory space on socket x.
The duration for command assertion is extended if the wait signal (NWAIT) is active (low) at
the end of the programmed value of HCLK:
0000 0000: Reserved
0000 0001: 2 HCLK cycles (+ wait cycle introduced by deassertion of NWAIT)
1111 1111: 255 HCLK cycles (+ wait cycle introduced by deasserting NWAIT)
1111 1111: Reserved.
Bits 7:0 ATTSET[7:0]: Attribute memory x setup time
Defines the number of HCLK (+1) clock cycles to set up address before the command
assertion (NWE, NOE), for PC CARD/NAND Flash read or write access to attribute memory
space on socket x:
0000 0000: 1 HCLK cycle
1111 1110: 255 HCLK cycles
1111 1111: Reserved.

I/O space timing register 4 (FSMC_PIO4)
Address offset: 0xA000 0000 + 0xB0
Reset value: 0xFCFCFCFC
The FSMC_PIO4 read/write registers contain the timing information used to gain access to
the I/O space of the 16-bit PC Card/CompactFlash.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
IOHIZ[7:0]
rw

rw

rw

rw

rw

IOHOLD[7:0]
rw

rw

rw

rw

rw

rw

rw

rw

9

8

7

6

5

IOWAIT[7:0]

rw

rw

rw

rw

rw

rw

RM0090 Rev 17

rw

rw

rw

4

3

2

1

0

rw

rw

rw

IOSET[7:0]
rw

rw

rw

rw

rw

rw

rw

1595/1747
1599

Flexible static memory controller (FSMC)

RM0090

Bits 31:24 IOHIZ[7:0]: I/O x databus HiZ time
Defines the number of HCLK clock cycles during which the databus is kept in HiZ after the start
of a PC Card write access to I/O space on socket x. Only valid for write transaction:
0000 0000: 0 HCLK cycle
1111 1111: 255 HCLK cycles (default value after reset)
Bits 23:16 IOHOLD[7:0]: I/O x hold time
Defines the number of HCLK clock cycles to hold address (and data for write access) after the
command deassertion (NWE, NOE), for PC Card read or write access to I/O space on socket
x:
0000 0000: reserved
0000 0001: 1 HCLK cycle
1111 1111: 255 HCLK cycles (default value after reset)
Bits 15:8 IOWAIT[7:0]: I/O x wait time
Defines the minimum number of HCLK (+1) clock cycles to assert the command (SMNWE,
SMNOE), for PC Card read or write access to I/O space on socket x. The duration for
command assertion is extended if the wait signal (NWAIT) is active (low) at the end of the
programmed value of HCLK:
0000 0000: reserved, do not use this value
0000 0001: 2 HCLK cycles (+ wait cycle introduced by deassertion of NWAIT)
1111 1111: 256 HCLK cycles (+ wait cycle introduced by the Card deasserting NWAIT)
(default value after reset)
Bits 7:0 IOSET[7:0]: I/O x setup time
Defines the number of HCLK (+1) clock cycles to set up the address before the command
assertion (NWE, NOE), for PC Card read or write access to I/O space on socket x:
0000 0000: 1 HCLK cycle
1111 1111: 256 HCLK cycles (default value after reset)

ECC result registers 2/3 (FSMC_ECCR2/3)
Address offset: 0xA000 0000 + 0x54 + 0x20 * (x – 1), x = 2 or 3
Reset value: 0x0000 0000
These registers contain the current error correction code value computed by the ECC
computation modules of the FSMC controller (one module per NAND Flash memory bank).
When the CPU reads the data from a NAND Flash memory page at the correct address
(refer to Section 36.6.6: Computation of the error correction code (ECC) in NAND Flash
memory), the data read from or written to the NAND Flash are processed automatically by
ECC computation module. At the end of X bytes read (according to the ECCPS field in the
FSMC_PCRx registers), the CPU must read the computed ECC value from the
FSMC_ECCx registers, and then verify whether these computed parity data are the same
as the parity value recorded in the spare area, to determine whether a page is valid, and, to
correct it if applicable. The FSMC_ECCRx registers should be cleared after being read by
setting the ECCEN bit to zero. For computing a new data block, the ECCEN bit must be set
to one.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
ECCx[31:0]
r

1596/1747

RM0090 Rev 17

9

8

7

6

5

4

3

2

1

0

RM0090

Flexible static memory controller (FSMC)

Bits 31:0 ECCx[31:0]: ECC result
This field provides the value computed by the ECC computation logic. Table 252 hereafter
describes the contents of these bit fields.

Table 252. ECC result relevant bits
ECCPS[2:0]

Page size in bytes

ECC bits

000

256

ECC[21:0]

001

512

ECC[23:0]

010

1024

ECC[25:0]

011

2048

ECC[27:0]

100

4096

ECC[29:0]

101

8192

ECC[31:0]

RM0090 Rev 17

1597/1747
1599

0008
FSMC_BCR2
Reserved

0010
FSMC_BCR3
Reserved

0018
FSMC_BCR4
Reserved

0004
FSMC_BTR1
Res.
ACCMOD[1:0]

DATLAT[3:0]

CLKDIV[3:0]

000C
FSMC_BTR2
Res.
ACCMOD[1:0]

DATLAT[3:0]

CLKDIV[3:0]

0014
FSMC_BTR3
Res.
ACCMOD[1:0]

DATLAT[3:0]

CLKDIV[3:0]

001C
FSMC_BTR4
Res.
DATLAT[3:0]

CLKDIV[3:0]

0104

FSMC_BWTR
1

Res.

ACC
MOD
[1:0]

Res.

010C

FSMC_BWTR
2

Res.

ACC
MOD
[1:0]

Res.

1598/1747

RM0090 Rev 17

DATAST[7:0]

ADDHLD[3:0]

ADDSET[3:0]

DATAST[7:0]

ADDHLD[3:0]

ADDSET[3:0]

DATAST[7:0]

ADDHLD[3:0]

ADDSET[3:0]

DATAST[7:0]

ADDHLD[3:0]

ADDSET[3:0]

DATAST[7:0]

ADDHLD[3:0]

ADDSET[3:0]

DATAST[7:0]

ADDHLD[3:0]

BUSTURN[3:0] BUSTURN[3:0] BUSTURN[3:0] BUSTURN[3:0] BUSTURN[3:0] BUSTURN[3:0]

Reserved

CPSIZE[2:0]

CPSIZE[2:0]

CPSIZE[2:0]

WREN
WAITCFG
WRAPMOD
WAITPOL
BURSTEN
Reserved
FACCEN
MWID[1:0]

WREN
WAITCFG
WRAPMOD
WAITPOL
BURSTEN
Reserved
FACCEN
MWID[1:0]

MUXEN
MBKEN

MUXEN
MBKEN

MTYP[0:1]

WAITEN

WAITEN

MTYP[0:1]

EXTMOD

EXTMOD

MBKEN

MUXEN

MTYP[0:1]

MWID[1:0]

FACCEN

Reserved

BURSTEN

WAITPOL

WRAPMOD

WAITCFG

WREN

WAITEN

EXTMOD

MBKEN

MUXEN

MTYP[0:1]

MWID[1:0]

FACCEN

Reserved

BURSTEN

WAITPOL

WRAPMOD

WAITCFG

WREN

WAITEN

EXTMOD

ASYNCWAIT ASYNCWAIT ASYNCWAIT ASYNCWAIT

CPSIZE[2:0]

CBURSTRW

FSMC_BCR1

CBURSTRW

0000

CBURSTRW

Register

CBURSTRW

Offset

ACCMOD[1:0]

0

1

2

36.6.9

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3

Flexible static memory controller (FSMC)
RM0090

FSMC register map
The following table summarizes the FSMC registers.
Table 253. FSMC register map

ADDSET[3:0]

RM0090

Flexible static memory controller (FSMC)

FSMC_BWTR
3

Res.

ACC
MOD
[1:0]

Res.

011C

FSMC_BWTR
4

Res.

ACC
MOD
[1:0]

Res.

0xA000
0060

FSMC_PCR2

Reserved

0xA000
0080

FSMC_PCR3

Reserved

0xA000
00A0

FSMC_PCR4

Reserved

0xA000
0064

FSMC_SR2

Reserved

0xA000
0084

FSMC_SR3

Reserved

0xA000
00A4

FSMC_SR4

Reserved

0xA000
0068

FSMC_PMEM
2

MEMHIZ[7:0]

MEMHOLD[7:0]

MEMWAIT[7:0]

MEMSET[7:0]

0xA000
0088

FSMC_PMEM
3

MEMHIZ[7:0]

MEMHOLD[7:0]

MEMWAIT[7:0]

MEMSET[7:0]

0xA000
00A8

FSMC_PMEM
4

MEMHIZ[7:0]

MEMHOLD[7:0]

MEMWAIT[7:0]

MEMSET[7:0]

0xA000
006C

FSMC_PATT2

ATTHIZ[7:0]

ATTHOLD[7:0]

ATTWAIT[7:0]

ATTSET[7:0]

0xA000
008C

FSMC_PATT3

ATTHIZ[7:0]

ATTHOLD[7:0]

ATTWAIT[7:0]

ATTSET[7:0]

0xA000
00AC

FSMC_PATT4

ATTHIZ[7:0]

ATTHOLD[7:0]

ATTWAIT[7:0]

ATTSET[7:0]

0xA000
00B0

FSMC_PIO4

IOHIZ[7:0]

IOHOLD[7:0]

IOWAIT[7:0]

IOSET[7:0]

0xA000
0074

FSMC_ECCR2

ECC[31:0]

0xA000
0094

FSMC_ECCR3

ECC[31:0]

ADDSET[3:0]

ADDHLD[3:0]

IRS
IRS

IRS

Reserved

Reserved

Reserved

ADDSET[3:0]
PBKEN

PWAITEN
PWAITEN

PBKEN
PBKEN

PTYP
PTYP
PTYP

PWAITEN
ILS
ILS
ILS

IFS
IFS
IFS

ILEN
ILEN

IREN

ILEN

IREN

IREN

PWID[1:0]
PWID[1:0]

ECCEN
ECCEN
ECCEN

PWID[1:0]
IFEN
IFEN
IFEN

TCLR[2:0]
TCLR[2:0]
TCLR[2:0]

Res.

FEMPT FEMPT FEMPT

TAR[2:0]
TAR[2:0]

Res.

TAR[2:0]

DATAST[7:0]

ADDHLD[3:0]

DATAST[7:0]

BUSTURN[3:0] BUSTURN[3:0]
ECCPS[2:0] ECCPS[2:0] ECCPS[2:0]

Res.

0

0114

2

Register

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3

Offset

1

Table 253. FSMC register map (continued)

Refer to Table 1 on page 64 for the register boundary addresses.

RM0090 Rev 17

1599/1747
1599

Flexible memory controller (FMC)

37

RM0090

Flexible memory controller (FMC)
The Flexible memory controller (FMC) includes three memory controllers:
•

The NOR/PSRAM memory controller

•

The NAND/PC Card memory controller

•

The Synchronous DRAM (SDRAM/Mobile LPSDR SDRAM) controller

This section applies to STM32F42xxx and STM32F43xxx only.

37.1

FMC main features
The FMC functional block makes the interface with synchronous and asynchronous static
memories, SDRAM memories, and 16-bit PC memory cards. Its main purposes are:
•

to translate AHB transactions into the appropriate external device protocol

•

to meet the access time requirements of the external memory devices

All external memories share the addresses, data and control signals with the controller.
Each external device is accessed by means of a unique Chip Select. The FMC performs
only one access at a time to an external device.
The main features of the FMC controller are the following:
•

1600/1747

Interface with static-memory mapped devices including:
–

Static random access memory (SRAM)

–

NOR Flash memory/OneNAND Flash memory

–

PSRAM (4 memory banks)

–

16-bit PC Card compatible devices

–

Two banks of NAND Flash memory with ECC hardware to check up to 8 Kbytes of
data

•

Interface with synchronous DRAM (SDRAM/Mobile LPSDR SDRAM) memories

•

Burst mode support for faster access to synchronous devices such as NOR Flash
memory, PSRAM and SDRAM)

•

Programmable continuous clock output for asynchronous and synchronous accesses

•

8-,16- or 32-bit wide data bus

•

Independent Chip Select control for each memory bank

•

Independent configuration for each memory bank

•

Write enable and byte lane select outputs for use with PSRAM, SRAM and SDRAM
devices

•

External asynchronous wait control

•

Write Data FIFO with 16 x33-bit depth

•

Write Address FIFO with 16x30-bit depth

•

Cacheable Read FIFO with 6 x32-bit depth (6 x14-bit address tag) for SDRAM
controller.

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
The FMC embeds two Write FIFOs: a Write Data FIFO with a 16x33-bit depth and a Write
Address FIFO with a 16x30-bit depth.
•

The Write Data FIFO stores the AHB data to be written to the memory (up to 32 bits)
plus one bit for the AHB transfer (burst or not sequential mode)

•

The Write Address FIFO stores the AHB address (up to 28 bits) plus the AHB data size
(up to 2 bits). When operating in burst mode, only the start address is stored except
when crossing a page boundary (for PSRAM and SDRAM). In this case, the AHB burst
is broken into two FIFO entries.

At startup the FMC pins must be configured by the user application. The FMC I/O pins which
are not used by the application can be used for other purposes.
The FMC registers that define the external device type and associated characteristics are
usually set at boot time and do not change until the next reset or power-up. However, the
settings can be changed at any time.

37.2

Block diagram
The FMC consists of five main blocks:
•

The AHB interface (including the FMC configuration registers)

•

The NOR Flash/PSRAM/SRAM controller

•

The NAND Flash/PC Card controller

•

The SDRAM controller

•

The external device interface

The block diagram is shown in Figure 456.

RM0090 Rev 17

1601/1747
1680

Flexible memory controller (FMC)

RM0090

Figure 456. FMC block diagram
)0&LQWHUUXSWVWR19,&

)0&B1/ RU1$'9
)URPFORFN
FRQWUROOHU

125365$0
PHPRU\
FRQWUROOHU

+&/.

)0&B&/.

)0&B1%/>@
)0&B$>@
)0&B'> @
&RQILJXUDWLRQ
UHJLVWHUV

1$1'3&&DUG
PHPRU\
FRQWUROOHU

125365$0
VLJQDOV

65$0365$06'5$0
6KDUHGVLJQDOV
6KDUHG
VLJQDOV

)0&B1(>@
)0&B12(
)0&B1:(
)0&B1:$,7

125365$065$0
6KDUHGVLJQDOV

)0&B1&(>@

1$1'
VLJQDOV

)0&B,17>@
)0&B,175
)0&B1&(B
)0&B1&(B
)0&B1,25'
)0&B1,2:5

3&&DUG
VLJQDOV

)0&B15(*
)0&B&'
)0&B6'&/.
)0&B6'1:(
6'5$0
FRQWUROOHU

)0&B6'&.(>@
)0&B6'1(>@
)0&B15$6
)0&B1&$6

6'5$0
VLJQDOV

069

37.3

AHB interface
The AHB slave interface allows internal CPUs and other bus master peripherals to access
the external memories.
AHB transactions are translated into the external device protocol. In particular, if the
selected external memory is 16- or 8-bit wide, 32-bit wide transactions on the AHB are split
into consecutive 16- or 8-bit accesses. The FMC Chip Select (FMC_NEx) does not toggle
between consecutive accesses except when performing accesses in mode D with the
extended mode enabled.

1602/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
The FMC generates an AHB error in the following conditions:
•

When reading or writing to an FMC bank (Bank 1 to 4) which is not enabled.

•

When reading or writing to the NOR Flash bank while the FACCEN bit is reset in the
FMC_BCRx register.

•

When reading or writing to the PC Card banks while the FMC_CD input pin (Card
Presence Detection) is low.

•

When writing to a write protected SDRAM bank (WP bit set in the SDRAM_SDCRx
register).

•

When the SDRAM address range is violated (access to reserved address range)

The effect of an AHB error depends on the AHB master which has attempted the R/W
access:
•

If the access has been attempted by the Cortex®-M4 with FPU CPU, a hard fault
interrupt is generated.

•

If the access has been performed by a DMA controller, a DMA transfer error is
generated and the corresponding DMA channel is automatically disabled.

The AHB clock (HCLK) is the reference clock for the FMC.

37.3.1

Supported memories and transactions
General transaction rules
The requested AHB transaction data size can be 8-, 16- or 32-bit wide whereas the
accessed external device has a fixed data width. This may lead to inconsistent transfers.
Therefore, some simple transaction rules must be followed:
•

AHB transaction size and memory data size are equal
There is no issue in this case.

•

AHB transaction size is greater than the memory size:
In this case, the FMC splits the AHB transaction into smaller consecutive memory
accesses to meet the external data width. The FMC Chip Select (FMC_NEx) does not
toggle between the consecutive accesses.

•

AHB transaction size is smaller than the memory size:
The transfer may or not be consistent depending on the type of external device:
–

Accesses to devices that have the byte select feature (SRAM, ROM, PSRAM,
SDRAM)
In this case, the FMC allows read/write transactions and accesses the right data
through its byte lanes BL[3:0].
byte to be written are addressed by NBL[3:0].
All memory byte are read (NBL[3:0] are driven low during read transaction) and
the useless ones are discarded.

–

Accesses to devices that do not have the byte select feature (16-bit NOR and
NAND Flash memories)
This situation occurs when a byte access is requested to a 16-bit wide Flash
memory. Since the device cannot be accessed in byte mode (only 16-bit words
can be read/written from/to the Flash memory), Write transactions and Read
transactions are allowed (the controller reads the entire 16-bit memory word and
uses only the required byte).

RM0090 Rev 17

1603/1747
1680

Flexible memory controller (FMC)

RM0090

Configuration registers
The FMC can be configured through a set of registers. Refer to Section 37.5.6, for a
detailed description of the NOR Flash/PSRAM controller registers. Refer to Section 37.6.8,
for a detailed description of the NAND Flash/PC Card registers and to Section 37.7.5 for a
detailed description of the SDRAM controller registers.

37.4

External device address mapping
From the FMC point of view, the external memory is divided into 6 fixed-size banks of
256 Mbyte each (see Figure 457):
•

Bank 1 used to address up to 4 NOR Flash memory or PSRAM devices. This bank is
split into 4 NOR/PSRAM subbanks with 4 dedicated Chip Selects, as follows:
–

Bank 1 - NOR/PSRAM 1

–

Bank 1 - NOR/PSRAM 2

–

Bank 1 - NOR/PSRAM 3

–

Bank 1 - NOR/PSRAM 4

•

Banks 2 and 3 used to address NAND Flash memory devices (1 device per bank)

•

Bank 4 used to address a PC Card

•

Bank 5 and 6 used to address SDRAM devices (1 device per bank).

For each bank the type of memory to be used can be configured by the user application
through the Configuration register.

1604/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
Figure 457. FMC memory banks
!DDRESS

"ANK

3UPPORTED MEMORY TYPE

X 

"ANK 
 X  -"

./2032!-32!-

X&&& &&&&
X 

"ANK 
 X  -"
X&&& &&&&

.!.$ &LASH MEMORY

X 

"ANK 
 X  -"
X&&& &&&&
X 

"ANK 
 X  -"

0# #ARD

X&&& &&&&
X# 

3$2!- "ANK 
 X  -"
3$2!-

X#&&& &&&&
X$ 

3$2!- "ANK 
 X  -"
X$&&& &&&&
-36

37.4.1

NOR/PSRAM address mapping
HADDR[27:26] bits are used to select one of the four memory banks as shown in Table 254.
Table 254. NOR/PSRAM bank selection
HADDR[27:26](1)

Selected bank

00

Bank 1 - NOR/PSRAM 1

01

Bank 1 - NOR/PSRAM 2

10

Bank 1 - NOR/PSRAM 3

11

Bank 1 - NOR/PSRAM 4

1. HADDR are internal AHB address lines that are translated to external memory.

RM0090 Rev 17

1605/1747
1680

Flexible memory controller (FMC)

RM0090

The HADDR[25:0] bits contain the external memory address. Since HADDR is a byte
address whereas the memory is addressed at word level, the address actually issued to the
memory varies according to the memory data width, as shown in the following table.
Table 255. NOR/PSRAM External memory address
Memory width(1)

Data address issued to the memory

Maximum memory capacity (bits)

8-bit

HADDR[25:0]

64 Mbyte x 8 = 512 Mbit

16-bit

HADDR[25:1] >> 1

64 Mbyte/2 x 16 = 512 Mbit

32-bit

HADDR[25:2] >> 2

64 Mbyte/4 x 32 = 512 Mbit

1. In case of a 16-bit external memory width, the FMC will internally use HADDR[25:1] to generate the
address for external memory FMC_A[24:0]. In case of a 32-bit memory width, the FMC will internally use
HADDR[25:2] to generate the external address.
Whatever the external memory width, FMC_A[0] should be connected to external memory address A[0].

Wrap support for NOR Flash/PSRAM
Wrap burst mode for synchronous memories is not supported. The memories must be
configured in linear burst mode of undefined length.

37.4.2

NAND Flash memory/PC Card address mapping
In this case, three banks are available, each of them being divided into memory areas as
indicated in Table 256.
Table 256. NAND/PC Card memory mapping and timing registers
Start address

End address

FMC bank

0x9C00 0000

0x9FFF FFFF

0x9800 0000

0x9BFF FFFF Bank 4 - PC card

Attribute

FMC_PATT4 (0xAC)

0x9000 0000

0x93FF FFFF

Common

FMC_PMEM4 (0xA8)

0x8800 0000

0x8BFF FFFF

Attribute

FMC_PATT3 (0x8C)

0x8000 0000

0x83FF FFFF

Common

FMC_PMEM3 (0x88)

0x7800 0000

0x7BFF FFFF

Attribute

FMC_PATT2 (0x6C)

0x7000 0000

0x73FF FFFF

Common

FMC_PMEM2 (0x68)

Bank 3 - NAND Flash

Bank 2- NAND Flash

Memory space

Timing register

I/O

FMC_PIO4 (0xB0)

For NAND Flash memory, the common and attribute memory spaces are subdivided into
three sections (see in Table 257 below) located in the lower 256 Kbytes:

1606/1747

•

Data section (first 64 Kbytes in the common/attribute memory space)

•

Command section (second 64 Kbytes in the common / attribute memory space)

•

Address section (next 128 Kbytes in the common / attribute memory space)

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
Table 257. NAND bank selection
Section name

HADDR[17:16]

Address range

Address section

1X

0x020000-0x03FFFF

Command section

01

0x010000-0x01FFFF

Data section

00

0x000000-0x0FFFF

The application software uses the 3 sections to access the NAND Flash memory:
•

To send a command to NAND Flash memory, the software must write the command
value to any memory location in the command section.

•

To specify the NAND Flash address that must be read or written, the software
must write the address value to any memory location in the address section. Since an
address can be 4 or 5 byte long (depending on the actual memory size), several
consecutive write operations to the address section are required to specify the full
address.

•

To read or write data, the software reads or writes the data from/to any memory
location in the data section.

Since the NAND Flash memory automatically increments addresses, there is no need to
increment the address of the data section to access consecutive memory locations.

37.4.3

SDRAM address mapping
The HADDR[28] bit (internal AHB address line 28) is used to select one of the two memory
banks as indicated in Table 258.
Table 258. SDRAM bank selection
HADDR[28]

Selected bank

Control register

Timing register

0

SDRAM Bank1

FMC_SDCR1

FMC_SDTR1

1

SDRAM Bank2

FMC_SDCR2

FMC_SDTR2

The following table shows SDRAM mapping for an 13-bit row ,a 11-bit column and 4 internal
bank configurations.
Table 259. SDRAM address mapping
Memory width(1)

Internal bank

Row address

Column
address(2)

Maximum
memory capacity
(Mbyte)

8-bit

HADDR[25:24]

HADDR[23:11]

HADDR[10:0]

64 Mbyte:
4 x 8K x 2K

16-bit

HADDR[26:25]

HADDR[24:12]

HADDR[11:1]

128 Mbyte:
4 x 8K x 2K x 2

32-bit

HADDR[27:26]

HADDR[25:13]

HADDR[12:2]

256 Mbyte:
4 x 8K x 2K x 4

RM0090 Rev 17

1607/1747
1680

Flexible memory controller (FMC)

RM0090

1. When interfacing with a 16-bit memory, the FMC internally uses the HADDR[11:1] internal AHB address
lines to generate the external address. When interfacing with a 32-bit memory, the FMC internally uses
HADDR[12:2] lines to generate the external address. Whatever the memory width, FMC_A[0] has to be
connected to the external memory address A[0].
2. The AutoPrecharge is not supported. FMC_A[10] must be connected to the external memory address
A[10] but it will be always driven ‘low’.

The HADDR[27:0] bits are translated to external SDRAM address depending on the
SDRAM controller configuration:
•

Data size:8, 16 or 32 bits

•

Row size:11, 12 or 13 bits

•

Column size: 8, 9, 10 or 11 bits

•

Number of internal banks: two or four internal banks

Table 260 to Table shows the SDRAM address mapping versus the SDRAM controller
configuration.
)

Row size
configuration

Table 260. SDRAM address mapping with 8-bit data bus width(1)(2)
HADDR(AHB Internal Address Lines)
27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Bank
[1:0]

Res.

Bank
[1:0]

Res.

11-bit row size
configuration

Bank
[1:0]

Bank
[1:0]

Res.

Bank
[1:0]

Res.

Row[11:0]
Row[11:0]
Row[11:0]
Row[11:0]

Bank
[1:0]
Bank
[1:0]

Res.
Res.

Row[10:0]

Bank
[1:0]

Res.

Res.

13-bit row size
configuration

Row[10:0]

Bank
[1:0]

Res.

Res.

Row[10:0]

Bank
[1:0]

Res.
Res.

12-bit row size
configuration

Row[10:0]

Bank
[1:0]
Bank
[1:0]

Row[12:0]
Row[12:0]
Row[12:0]
Row[12:0]

Column[7:0]
Column[8:0]
Column[9:0]
Column[10:0]
Column[7:0]
Column[8:0]
Column[9:0]
Column[10:0]
Column[7:0]
Column[8:0]
Column[9:0]
Column[10:0]

1. BANK[1:0] are the Bank Address BA[1:0]. When only 2 internal banks are used, BA1 must always be set to ‘0’.
2. Access to Reserved (Res.) address range generates an AHB error.

1608/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
)

Table 261. SDRAM address mapping with 16-bit data bus width(1)(2)

Bank
Row[10:0]
[1:0]
Bank
Res.
Row[10:0]
11-bit row size
[1:0]
Bank
configuration
Res.
Row[10:0]
[1:0]
Bank
Res.
Row[10:0]
[1:0]
Bank
Res.
Row[11:0]
[1:0]
Bank
Res.
Row[11:0]
12-bit row size
[1:0]
Bank
configuration
Res.
Row[11:0]
[1:0]
Bank
Res.
Row[11:0]
[1:0]
Bank
Res.
Row[12:0]
[1:0]
Bank
Res.
Row[12:0]
13-bit row size
[1:0]
Bank
configuration
Res.
Row[12:0]
[1:0]
Re Bank
Row[12:0]
s. [1:0]
Res.

0

11

10
9
8
7
6
5
4
3
2
1

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

Row size
Configuration

27

HADDR(AHB address Lines)

BM0(3)

Column[7:0]
Column[8:0]

BM0

Column[9:0]

BM0

Column[10:0]

BM0

Column[7:0]

BM0

Column[8:0]

BM0

Column[9:0]

BM0

Column[10:0]

BM0

Column[7:0]

BM0

Column[8:0]

BM0

Column[9:0]

BM0

Column[10:0]

BM0

1. BANK[1:0] are the Bank Address BA[1:0]. When only 2 internal banks are used, BA1 must always be set to ‘0’.
2. Access to Reserved space (Res.) generates an AHB error.
3. BM0: is the byte mask for 16-bit access.

Table 262. SDRAM address mapping with 32-bit data bus width(1)(2)

Bank
[1:0]
Bank
Res.
[1:0]
Bank
Res.
[1:0]
Bank
Res.
[1:0]
Bank
Res.
[1:0]
Bank
Res.
[1:0]
Bank
Res.
[1:0]
Bank
Res.
[1:0]
Res.

11-bit row size
configuration

12-bit row size
configuration

Row[10:0]
Row[10:0]
Row[10:0]
Row[10:0]
Row[11:0]
Row[11:0]
Row[11:0]
Row[11:0]

RM0090 Rev 17

Column[7:0]
Column[8:0]
Column[9:0]
Column[10:0]
Column[7:0]
Column[8:0]
Column[9:0]
Column[10:0]

0

1

10
9
8
7
6
5
4
3
2

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

Row size
configuration

27

HADDR(AHB address Lines)

BM[1:0
](3)
BM[1:0
BM[1:0
BM[1:0
BM[1:0
BM[1:0
BM[1:0
BM[1:0

1609/1747
1680

Flexible memory controller (FMC)

RM0090

Table 262. SDRAM address mapping with 32-bit data bus width(1)(2) (continued)

Bank
[1:0]
Bank
Res.
13-bit row size
[1:0]
configuration Res. Bank
[1:0]
Bank
[1:0]
Res.

Row[12:0]
Row[12:0]
Row[12:0]

Column[7:0]
Column[8:0]
Column[9:0]
Column[10:0]

0

1

11

Row[12:0]

10
9
8
7
6
5
4
3
2

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

HADDR(AHB address Lines)
Row size
configuration

BM[1:0
BM[1:0
BM[1:0
BM[1:0

1. BANK[1:0] are the Bank Address BA[1:0]. When only 2 internal banks are used, BA1 must always be set to ‘0’.
2. Access to Reserved space (Res.) generates an AHB error.
3. BM[1:0]: is the byte mask for 32-bit access.

37.5

NOR Flash/PSRAM controller
The FMC generates the appropriate signal timings to drive the following types of memories:
•

•

•

Asynchronous SRAM and ROM
–

8 bits

–

16 bits

–

32 bits

PSRAM (Cellular RAM)
–

Asynchronous mode

–

Burst mode for synchronous accesses

–

Multiplexed or non-multiplexed

NOR Flash memory
–

Asynchronous mode

–

Burst mode for synchronous accesses

–

Multiplexed or non-multiplexed

The FMC outputs a unique Chip Select signal, NE[4:1], per bank. All the other signals
(addresses, data and control) are shared.
The FMC supports a wide range of devices through a programmable timings among which:
•

Programmable wait states (up to 15)

•

Programmable bus turnaround cycles (up to 15)

•

Programmable output enable and write enable delays (up to 15)

•

Independent read and write timings and protocol to support the widest variety of
memories and timings

•

Programmable continuous clock (FMC_CLK) output.

The FMC Clock (FMC_CLK) is a submultiple of the HCLK clock. It can be delivered to the
selected external device either during synchronous accesses only or during asynchronous

1610/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
and synchronous accesses depending on the CCKEN bit configuration in the FMC_BCR1
register:
•

If the CCLKEN bit is reset, the FMC generates the clock (CLK) only during
synchronous accesses (Read/write transactions).

•

If the CCLKEN bit is set, the FMC generates a continuous clock during asynchronous
and synchronous accesses. To generate the FMC_CLK continuous clock, Bank 1 must
be configured in synchronous mode (see Section 37.5.6: NOR/PSRAM controller
registers). Since the same clock is used for all synchronous memories, when a
continuous output clock is generated and synchronous accesses are performed, the
AHB data size has to be the same as the memory data width (MWID) otherwise the
FMC_CLK frequency will be changed depending on AHB data transaction (refer to
Section 37.5.5: Synchronous transactions for FMC_CLK divider ratio formula).

The size of each bank is fixed and equal to 64 Mbyte. Each bank is configured through
dedicated registers (see Section 37.5.6: NOR/PSRAM controller registers).
The programmable memory parameters include access times (see Table 263) and support
for wait management (for PSRAM and NOR Flash accessed in burst mode).
Table 263. Programmable NOR/PSRAM access parameters

37.5.1

Parameter

Function

Access mode

Unit

Min.

Max.

Address
setup

Duration of the address
setup phase

Asynchronous

AHB clock cycle
(HCLK)

0

15

Address hold

Duration of the address hold
phase

Asynchronous,
muxed I/Os

AHB clock cycle
(HCLK)

1

15

Data setup

Duration of the data setup
phase

Asynchronous

AHB clock cycle
(HCLK)

1

256

Bust turn

Duration of the bus
turnaround phase

Asynchronous and
AHB clock cycle
synchronous
(HCLK)
read/write

0

15

Clock divide
ratio

Number of AHB clock cycles
(HCLK) to build one memory
clock cycle (CLK)

Synchronous

AHB clock cycle
(HCLK)

2

16

Data latency

Number of clock cycles to
issue to the memory before
the first data of the burst

Synchronous

Memory clock
cycle (CLK)

2

17

External memory interface signals
Table 264, Table 265 and Table 266 list the signals that are typically used to interface with
NOR Flash memory, SRAM and PSRAM.

Note:

The prefix “N” identifies the signals which are active low.

RM0090 Rev 17

1611/1747
1680

Flexible memory controller (FMC)

RM0090

NOR Flash memory, non-multiplexed I/Os
Table 264. Non-multiplexed I/O NOR Flash memory
FMC signal name

I/O

Function

CLK

O

Clock (for synchronous access)

A[25:0]

O

Address bus

D[31:0]

I/O

Bidirectional data bus

NE[x]

O

Chip Select, x = 1..4

NOE

O

Output enable

NWE

O

Write enable

NL(=NADV)

O

Latch enable (this signal is called address
valid, NADV, by some NOR Flash devices)

NWAIT

I

NOR Flash wait input signal to the FMC

The maximum capacity is 512 Mbits (26 address lines).

NOR Flash memory, 16-bit multiplexed I/Os
Table 265. 16-bit multiplexed I/O NOR Flash memory
FMC signal name

I/O

Function

CLK

O

Clock (for synchronous access)

A[25:16]

O

Address bus

AD[15:0]

I/O

16-bit multiplexed, bidirectional address/data bus (the 16-bit address
A[15:0] and data D[15:0] are multiplexed on the databus)

NE[x]

O

Chip Select, x = 1..4

NOE

O

Output enable

NWE

O

Write enable

NL(=NADV)

O

Latch enable (this signal is called address valid, NADV, by some NOR
Flash devices)

NWAIT

I

NOR Flash wait input signal to the FMC

The maximum capacity is 512 Mbits.

PSRAM/SRAM, non-multiplexed I/Os
Table 266. Non-multiplexed I/Os PSRAM/SRAM

1612/1747

FMC signal name

I/O

Function

CLK

O

Clock (only for PSRAM synchronous access)

A[25:0]

O

Address bus

D[31:0]

I/O

Data bidirectional bus

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
Table 266. Non-multiplexed I/Os PSRAM/SRAM (continued)
FMC signal name

I/O

Function

NE[x]

O

Chip Select, x = 1..4 (called NCE by PSRAM (Cellular RAM i.e.
CRAM))

NOE

O

Output enable

NWE

O

Write enable

NL(= NADV)

O

Address valid only for PSRAM input (memory signal name: NADV)

NWAIT

I

PSRAM wait input signal to the FMC

NBL[3]

O

Byte3 Upper byte enable (memory signal name: NUB)

NBL[2]

O

Byte2 Lowed byte enable (memory signal name: NLB)

NBL[1]

O

Byte1 Upper byte enable (memory signal name: NLB)

NBL[0]

O

Byte0 Lower byte enable (memory signal name: NLB)

The maximum capacity is 512 Mbits.

PSRAM, 16-bit multiplexed I/Os
Table 267. 16-Bit multiplexed I/O PSRAM
FMC signal name

I/O

Function

CLK

O

Clock (for synchronous access)

A[25:16]

O

Address bus

AD[15:0]

I/O

16-bit multiplexed, bidirectional address/data bus (the 16-bit address
A[15:0] and data D[15:0] are multiplexed on the databus)

NE[x]

O

Chip Select, x = 1..4 (called NCE by PSRAM (Cellular RAM i.e.
CRAM))

NOE

O

Output enable

NWE

O

Write enable

NL(= NADV)

O

Address valid PSRAM input (memory signal name: NADV)

NWAIT

I

PSRAM wait input signal to the FMC

NBL[1]

O

Upper byte enable (memory signal name: NUB)

NBL[0]

O

Lowed byte enable (memory signal name: NLB)

The maximum capacity is 512 Mbits (26 address lines).

37.5.2

Supported memories and transactions
Table 268 below shows an example of the supported devices, access modes and
transactions when the memory data bus is 16-bit wide for NOR Flash memory, PSRAM and
SRAM. The transactions not allowed (or not supported) by the FMC are shown in gray in
this example.

RM0090 Rev 17

1613/1747
1680

Flexible memory controller (FMC)

RM0090

Table 268. NOR Flash/PSRAM: Example of supported memories and transactions
Device

NOR Flash
(muxed I/Os
and nonmuxed
I/Os)

PSRAM
(multiplexed
I/Os and nonmultiplexed
I/Os)

SRAM and
ROM

1614/1747

Mode

R/W

AHB
data
size

Memory
data size

Allowed/
not
allowed

Asynchronous

R

8

16

Y

Asynchronous

W

8

16

N

Asynchronous

R

16

16

Y

Asynchronous

W

16

16

Y

Asynchronous

R

32

16

Y

Split into 2 FMC accesses

Asynchronous

W

32

16

Y

Split into 2 FMC accesses

Asynchronous
page

R

-

16

N

Mode is not supported

Synchronous

R

8

16

N

Synchronous

R

16

16

Y

Synchronous

R

32

16

Y

Asynchronous

R

8

16

Y

Asynchronous

W

8

16

Y

Asynchronous

R

16

16

Y

Asynchronous

W

16

16

Y

Asynchronous

R

32

16

Y

Split into 2 FMC accesses

Asynchronous

W

32

16

Y

Split into 2 FMC accesses

Asynchronous
page

R

-

16

N

Mode is not supported

Synchronous

R

8

16

N

Synchronous

R

16

16

Y

Synchronous

R

32

16

Y

Synchronous

W

8

16

Y

Synchronous

W

16/32

16

Y

Asynchronous

R

8 / 16

16

Y

Asynchronous

W

8 / 16

16

Y

Use of byte lanes NBL[1:0]

Asynchronous

R

32

16

Y

Split into 2 FMC accesses

Asynchronous

W

32

16

Y

Split into 2 FMC accesses
Use of byte lanes NBL[1:0]

RM0090 Rev 17

Comments

Use of byte lanes NBL[1:0]

Use of byte lanes NBL[1:0]

RM0090

37.5.3

Flexible memory controller (FMC)

General timing rules
Signals synchronization

37.5.4

•

All controller output signals change on the rising edge of the internal clock (HCLK)

•

In synchronous mode (read or write), all output signals change on the rising edge of
HCLK. Whatever the CLKDIV value, all outputs change as follows:
–

NOEL/NWEL/ NEL/NADVL/ NADVH /NBLL/ Address valid outputs change on the
falling edge of FMC_CLK clock.

–

NOEH/ NWEH / NEH/ NOEH/NBLH/ Address invalid outputs change on the rising
edge of FMC_CLK clock.

NOR Flash/PSRAM controller asynchronous transactions
Asynchronous static memories (NOR Flash, PSRAM, SRAM)
•

Signals are synchronized by the internal clock HCLK. This clock is not issued to the
memory

•

The FMC always samples the data before de-asserting the NOE signal. This
guarantees that the memory data hold timing constraint is met (minimum Chip Enable
high to data transition is usually 0 ns)

•

If the extended mode is enabled (EXTMOD bit is set in the FMC_BCRx register), up to
four extended modes (A, B, C and D) are available. It is possible to mix A, B, C and D
modes for read and write operations. For example, read operation can be performed in
mode A and write in mode B.

•

If the extended mode is disabled (EXTMOD bit is reset in the FMC_BCRx register), the
FMC can operate in Mode1 or Mode2 as follows:
–

Mode 1 is the default mode when SRAM/PSRAM memory type is selected
(MTYP[1:0] = 0x0 or 0x01 in the FMC_BCRx register)

–

Mode 2 is the default mode when NOR memory type is selected
(MTYP[1:0] = 0x10 in the FMC_BCRx register).

RM0090 Rev 17

1615/1747
1680

Flexible memory controller (FMC)

RM0090

Mode 1 - SRAM/PSRAM (CRAM)
The next figures show the read and write transactions for the supported modes followed by
the required configuration of FMC _BCRx, and FMC_BTRx/FMC_BWTRx registers.
Figure 458. Mode1 read access waveforms
-EMORY TRANSACTION
!;=

.",;=

.%X

./%
.7%

(IGH

DATA DRIVEN
BY MEMORY

$;=
!$$3%4
(#,+ CYCLES

$!4!34
(#,+ CYCLES
-36

Figure 459. Mode1 write access waveforms
-EMORY TRANSACTION
!;=

.",;=

.%X

./%
(#,+
.7%

$;=

DATA DRIVEN BY &3-#
!$$3%4
(#,+ CYCLES

$!4!34 
(#,+ CYCLES
-36

1616/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
The one HCLK cycle at the end of the write transaction helps guarantee the address and
data hold time after the NWE rising edge. Due to the presence of this HCLK cycle, the
DATAST value must be greater than zero (DATAST > 0).
Table 269. FMC_BCRx bit fields
Bit
number

Bit name

31-21

Reserved

0x000

20

CCLKEN

As needed

19

CBURSTRW

0x0 (no effect in asynchronous mode)

18:16

CPSIZE

0x0 (no effect in asynchronous mode)

15

ASYNCWAIT

14

EXTMOD

0x0

13

WAITEN

0x0 (no effect in asynchronous mode)

12

WREN

As needed

11

WAITCFG

Don’t care

10

WRAPMOD

9

WAITPOL

Meaningful only if bit 15 is 1

8

BURSTEN

0x0

7

Reserved

0x1

6

FACCEN

Don’t care

5-4

MWID

As needed

3-2

MTYP[1:0]

1

MUXE

0x0

0

MBKEN

0x1

Value to set

Set to 1 if the memory supports this feature. Otherwise keep at
0.

0x0

As needed, exclude 0x2 (NOR Flash memory)

Table 270. FMC_BTRx bit fields
Bit
number

Bit name

31:30

Reserved

0x0

29-28

ACCMOD

Don’t care

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Value to set

Time between NEx high to NEx low (BUSTURN HCLK)
Duration of the second access phase (DATAST+1 HCLK cycles for
write accesses, DATAST HCLK cycles for read accesses).

RM0090 Rev 17

1617/1747
1680

Flexible memory controller (FMC)

RM0090

Table 270. FMC_BTRx bit fields (continued)
Bit
number

Bit name

7-4

ADDHLD

Don’t care

3-0

ADDSET

Duration of the first access phase (ADDSET HCLK cycles).
Minimum value for ADDSET is 0.

Value to set

Mode A - SRAM/PSRAM (CRAM) OE toggling
Figure 460. ModeA read access waveforms
-EMORY TRANSACTION
!;=

.",;=

.%X

./%
.7%

(IGH

DATA DRIVEN
BY MEMORY

$;=
!$$3%4
(#,+ CYCLES

$!4!34
(#,+ CYCLES
-36

1. NBL[3:0] are driven low during the read access

1618/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
Figure 461. ModeA write access waveforms
DĞŵŽƌLJƚƌĂŶƐĂĐƚŝŽŶ
΀Ϯϱ͗Ϭ΁

E>΀ϯ͗Ϭ΁

Edž

EK
ϭ,><
Et

΀ϯϭ͗Ϭ΁

ĚĂƚĂĚƌŝǀĞŶďLJ&^D
^d
,><ĐLJĐůĞƐ

;d^dнϭͿ
,><ĐLJĐůĞƐ
D^ϯϬϰϱϱsϭ

The differences compared with mode1 are the toggling of NOE and the independent read
and write timings.
Table 271. FMC_BCRx bit fields
Bit
number

Bit name

31-21

Reserved

0x000

20

CCLKEN

As needed

19

CBURSTRW

0x0 (no effect in asynchronous mode)

18:16

CPSIZE

0x0 (no effect in asynchronous mode)

15

ASYNCWAIT

14

EXTMOD

0x1

13

WAITEN

0x0 (no effect in asynchronous mode)

12

WREN

As needed

11

WAITCFG

Don’t care

10

WRAPMOD

9

WAITPOL

Meaningful only if bit 15 is 1

8

BURSTEN

0x0

7

Reserved

0x1

6

FACCEN

Don’t care

Value to set

Set to 1 if the memory supports this feature. Otherwise keep at
0.

0x0

RM0090 Rev 17

1619/1747
1680

Flexible memory controller (FMC)

RM0090

Table 271. FMC_BCRx bit fields (continued)
Bit
number

Bit name

5-4

MWID

3-2

MTYP[1:0]

1

MUXEN

0x0

0

MBKEN

0x1

Value to set
As needed
As needed, exclude 0x2 (NOR Flash memory)

Table 272. FMC_BTRx bit fields
Bit
number

Bit name

31:30

Reserved

0x0

29-28

ACCMOD

0x0

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST HCLK cycles) for read
accesses.

7-4

ADDHLD

Don’t care

3-0

Value to set

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the first access phase (ADDSET HCLK cycles) for read
ADDSET[3:0] accesses.
Minimum value for ADDSET is 0.

Table 273. FMC_BWTRx bit fields
Bit
number

Bit name

31:30

Reserved

0x0

29-28

ACCMOD

0x0

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST HCLK cycles) for write
accesses.

7-4

ADDHLD

Don’t care

3-0

1620/1747

Value to set

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the first access phase (ADDSET HCLK cycles) for write
ADDSET[3:0] accesses.
Minimum value for ADDSET is 0.

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)

Mode 2/B - NOR Flash
Figure 462. Mode2 and mode B read access waveforms
-EMORY TRANSACTION
!;=

.!$6

.%X

./%
.7%

(IGH

DATA DRIVEN
BY MEMORY

$;=
!$$3%4
(#,+ CYCLES

$!4!34
(#,+ CYCLES
-36

1. NBL[3:0] are driven low during the read access

Figure 463. Mode2 write access waveforms
-EMORY TRANSACTION
!;=

.!$6

.%X

./%
(#,+
.7%

$;=

DATA DRIVEN BY &3-#
!$$3%4
(#,+ CYCLES

$!4!34 
(#,+ CYCLES
-36

RM0090 Rev 17

1621/1747
1680

Flexible memory controller (FMC)

RM0090

Figure 464. ModeB write access waveforms
-EMORY TRANSACTION
!;=

.!$6

.%X

./%
(#,+
.7%

$;=

DATA DRIVEN BY &3-#
!$$3%4
(#,+ CYCLES

$!4!34 
(#,+ CYCLES
-36

The differences with mode1 are the toggling of NWE and the independent read and write
timings when extended mode is set (Mode B).
Table 274. FMC_BCRx bit fields

1622/1747

Bit
number

Bit name

31-21

Reserved

0x000

20

CCLKEN

As needed

19

CBURSTRW

0x0 (no effect in asynchronous mode)

18:16

CPSIZE

0x0 (no effect in asynchronous mode)

15

ASYNCWAIT

14

EXTMOD

0x1 for mode B, 0x0 for mode 2

13

WAITEN

0x0 (no effect in asynchronous mode)

12

WREN

As needed

11

WAITCFG

Don’t care

10

WRAPMOD

9

WAITPOL

Meaningful only if bit 15 is 1

8

BURSTEN

0x0

7

Reserved

0x1

6

FACCEN

0x1

Value to set

Set to 1 if the memory supports this feature. Otherwise keep at
0.

0x0

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
Table 274. FMC_BCRx bit fields (continued)
Bit
number

Bit name

5-4

MWID

3-2

MTYP[1:0]

1

MUXEN

0x0

0

MBKEN

0x1

Value to set
As needed
0x2 (NOR Flash memory)

Table 275. FMC_BTRx bit fields
Bit number

Bit name

Value to set

31-30

Reserved

0x0

29-28

ACCMOD

0x1 if extended mode is set

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Duration of the access second phase (DATAST HCLK cycles) for
read accesses.

7-4

ADDHLD

Don’t care

3-0

ADDSET[3:0]

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the access first phase (ADDSET HCLK cycles) for read
accesses. Minimum value for ADDSET is 0.

Table 276. FMC_BWTRx bit fields

Note:

Bit number

Bit name

Value to set

31-30

Reserved

0x0

29-28

ACCMOD

0x1 if extended mode is set

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Duration of the access second phase (DATAST HCLK cycles) for
write accesses.

7-4

ADDHLD

Don’t care

3-0

ADDSET[3:0]

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the access first phase (ADDSET HCLK cycles) for write
accesses. Minimum value for ADDSET is 0.

The FMC_BWTRx register is valid only if the extended mode is set (mode B), otherwise its
content is don’t care.

RM0090 Rev 17

1623/1747
1680

Flexible memory controller (FMC)

RM0090

Mode C - NOR Flash - OE toggling
Figure 465. ModeC read access waveforms
-EMORY TRANSACTION
!;=

.!$6

.%X

./%
.7%

(IGH

DATA DRIVEN
BY MEMORY

$;=
!$$3%4
(#,+ CYCLES

$!4!34
(#,+ CYCLES
-36

Figure 466. ModeC write access waveforms
-EMORY TRANSACTION
!;=

.!$6

.%X

./%
(#,+
.7%

$;=

DATA DRIVEN BY &3-#
!$$3%4
(#,+ CYCLES

$!4!34 
(#,+ CYCLES
-36

The differences compared with mode1 are the toggling of NOE and the independent read
and write timings.

1624/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
Table 277. FMC_BCRx bit fields
Bit No.

Bit name

Value to set

31-21

Reserved

0x000

20

CCLKEN

As needed

19

CBURSTRW

0x0 (no effect in asynchronous mode)

18:16

Reserved

0x0 (no effect in asynchronous mode)

15

ASYNCWAIT

14

EXTMOD

0x1

13

WAITEN

0x0 (no effect in asynchronous mode)

12

WREN

As needed

11

WAITCFG

Don’t care

10

WRAPMOD

9

WAITPOL

Meaningful only if bit 15 is 1

8

BURSTEN

0x0

7

Reserved

0x1

6

FACCEN

0x1

5-4

MWID

3-2

MTYP[1:0]

1

MUXEN

0x0

0

MBKEN

0x1

Set to 1 if the memory supports this feature. Otherwise keep at 0.

0x0

As needed
0x02 (NOR Flash memory)

Table 278. FMC_BTRx bit fields
Bit No.

Bit name

Value to set

31:30

Reserved

0x0

29-28

ACCMOD

0x2

27-24

DATLAT

0x0

23-20

CLKDIV

0x0

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST HCLK cycles) for
read accesses.

7-4

ADDHLD

Don’t care

3-0

ADDSET[3:0]

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the first access phase (ADDSET HCLK cycles) for read
accesses. Minimum value for ADDSET is 0.

RM0090 Rev 17

1625/1747
1680

Flexible memory controller (FMC)

RM0090
Table 279. FMC_BWTRx bit fields

Bit No.

Bit name

Value to set

31:30

Reserved

0x0

29-28

ACCMOD

0x2

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST HCLK cycles) for
write accesses.

7-4

ADDHLD

Don’t care

3-0

ADDSET[3:0]

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the first access phase (ADDSET HCLK cycles) for write
accesses. Minimum value for ADDSET is 0.

Mode D - asynchronous access with extended address
Figure 467. ModeD read access waveforms
-EMORY TRANSACTION
!;=

.!$6

.%X

./%
.7%

(IGH

DATA DRIVEN
BY MEMORY

$;=
!$$3%4
(#,+ CYCLES

1626/1747

!$$(,$
(#,+ CYCLES

RM0090 Rev 17

$!4!34
(#,+ CYCLES
-36

RM0090

Flexible memory controller (FMC)
Figure 468. ModeD write access waveforms
DĞŵŽƌLJƚƌĂŶƐĂĐƚŝŽŶ
΀Ϯϱ͗Ϭ΁

Es

Edž

EK
ϭ,><
Et

΀ϯϭ͗Ϭ΁

ĚĂƚĂĚƌŝǀĞŶďLJ&^D
^d
,><ĐLJĐůĞƐ

,>
,><ĐLJĐůĞƐ

;d^dнϭͿ
,><ĐLJĐůĞƐ
D^ϯϬϰϲϮsϭ

The differences with mode1 are the toggling of NOE that goes on toggling after NADV
changes and the independent read and write timings.
Table 280. FMC_BCRx bit fields
Bit No.

Bit name

Value to set

31-21

Reserved

0x000

20

CCLKEN

As needed

19

CBURSTRW

0x0 (no effect in asynchronous mode)

18:16

CPSIZE

0x0 (no effect in asynchronous mode)

15

ASYNCWAIT

14

EXTMOD

0x1

13

WAITEN

0x0 (no effect in asynchronous mode)

12

WREN

As needed

11

WAITCFG

Don’t care

10

WRAPMOD

9

WAITPOL

Meaningful only if bit 15 is 1

8

BURSTEN

0x0

7

Reserved

0x1

6

FACCEN

Set according to memory support

5-4

MWID

Set to 1 if the memory supports this feature. Otherwise keep
at 0.

0x0

As needed

RM0090 Rev 17

1627/1747
1680

Flexible memory controller (FMC)

RM0090

Table 280. FMC_BCRx bit fields (continued)
Bit No.

Bit name

Value to set

3-2

MTYP[1:0]

1

MUXEN

0x0

0

MBKEN

0x1

As needed

Table 281. FMC_BTRx bit fields
Bit No.

Bit name

Value to set

31:30

Reserved

0x0

29-28

ACCMOD

0x3

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST HCLK cycles) for
read accesses.

7-4

ADDHLD

Duration of the middle phase of the read access (ADDHLD HCLK
cycles)

3-0

ADDSET[3:0]

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the first access phase (ADDSET HCLK cycles) for read
accesses. Minimum value for ADDSET is 1.

Table 282. FMC_BWTRx bit fields

1628/1747

Bit No.

Bit name

Value to set

31:30

Reserved

0x0

29-28

ACCMOD

0x3

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST + 1 HCLK cycles) for
write accesses.

7-4

ADDHLD

Duration of the middle phase of the write access (ADDHLD HCLK
cycles)

3-0

ADDSET[3:0]

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the first access phase (ADDSET HCLK cycles) for write
accesses. Minimum value for ADDSET is 1.

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)

Muxed mode - multiplexed asynchronous access to NOR Flash memory
Figure 469. Muxed read access waveforms

Figure 470. Muxed write access waveforms
-EMORY TRANSACTION
!;=

.!$6

.%X

./%
(#,+
.7%

!$;=

,OWER ADDRESS
!$$3%4
(#,+ CYCLES

!$$(,$
(#,+ CYCLES

DATA DRIVEN BY &3-#
$!4!34 
(#,+ CYCLES
AI

The difference with mode D is the drive of the lower address byte(s) on the data bus.

RM0090 Rev 17

1629/1747
1680

Flexible memory controller (FMC)

RM0090
Table 283. FMC_BCRx bit fields

Bit No.

Bit name

Value to set

31-21

Reserved

0x000

20

CCLKEN

As needed

19

CBURSTRW

0x0 (no effect in asynchronous mode)

18:16

CPSIZE

0x0 (no effect in asynchronous mode)

15

ASYNCWAIT

14

EXTMOD

0x0

13

WAITEN

0x0 (no effect in asynchronous mode)

12

WREN

As needed

11

WAITCFG

Don’t care

10

WRAPMOD

9

WAITPOL

Meaningful only if bit 15 is 1

8

BURSTEN

0x0

7

Reserved

0x1

6

FACCEN

0x1

5-4

MWID

3-2

MTYP[1:0]

1

MUXEN

0x1

0

MBKEN

0x1

Set to 1 if the memory supports this feature. Otherwise keep at
0.

0x0

As needed
0x2 (NOR Flash memory)

Table 284. FMC_BTRx bit fields
Bit No.

Bit name

Value to set

31:30

Reserved

0x0

29-28

ACCMOD

0x0

27-24

DATLAT

Don’t care

23-20

CLKDIV

Don’t care

19-16

BUSTURN

15-8

DATAST

Duration of the second access phase (DATAST HCLK cycles for
read accesses and DATAST+1 HCLK cycles for write accesses).

7-4

ADDHLD

Duration of the middle phase of the access (ADDHLD HCLK cycles).

3-0

ADDSET[3:0]

Time between NEx high to NEx low (BUSTURN HCLK)

Duration of the first access phase (ADDSET HCLK cycles).
Minimum value for ADDSET is 1.

WAIT management in asynchronous accesses
If the asynchronous memory asserts the WAIT signal to indicate that it is not yet ready to
accept or to provide data, the ASYNCWAIT bit has to be set in FMC_BCRx register.

1630/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
If the WAIT signal is active (high or low depending on the WAITPOL bit), the second access
phase (Data setup phase), programmed by the DATAST bits, is extended until WAIT
becomes inactive. Unlike the data setup phase, the first access phases (Address setup and
Address hold phases), programmed by the ADDSET[3:0] and ADDHLD bits, are not WAIT
sensitive and so they are not prolonged.
The data setup phase must be programmed so that WAIT can be detected 4 HCLK cycles
before the end of the memory transaction. The following cases must be considered:
1.

The memory asserts the WAIT signal aligned to NOE/NWE which toggles:
DATAST ≥ ( 4 × HCLK ) + max_wait_assertion_time

2.

The memory asserts the WAIT signal aligned to NEx (or NOE/NWE not toggling):
if
max_wait_assertion_time > address_phase + hold_phase
then:

DATAST ≥ ( 4 × HCLK ) + ( max_wait_assertion_time

– address_phase – hold_phase )

otherwise
DATAST ≥ 4 × HCLK
where max_wait_assertion_time is the maximum time taken by the memory to assert
the WAIT signal once NEx/NOE/NWE is low.
Figure 471 and Figure 472 show the number of HCLK clock cycles that are added to the
memory access phase after WAIT is released by the asynchronous memory (independently
of the above cases).

Figure 471. Asynchronous wait during a read access waveforms
0HPRU\WUDQVDFWLRQ

$>@

DGGUHVVSKDVH

GDWDVHWXSSKDVH

1([

.7!)4

DONT CARE

DONT CARE

12(

GDWDGULYHQE\PHPRU\

'>@

+&/.

069

1. NWAIT polarity depends on WAITPOL bit setting in FMC_BCRx register.

RM0090 Rev 17

1631/1747
1680

Flexible memory controller (FMC)

RM0090

Figure 472. Asynchronous wait during a write access waveforms

0HPRU\WUDQVDFWLRQ

$>@
DGGUHVVSKDVH

GDWDVHWXSSKDVH

1([

.7!)4

DONT CARE

DONT CARE
+&/.

1:(

'>@

GDWDGULYHQE\)60&
+&/.
069

1. NWAIT polarity depends on WAITPOL bit setting in FMC_BCRx register.

37.5.5

Synchronous transactions
The memory clock, FMC_CLK, is a submultiple of HCLK. It depends on the value of
CLKDIV and the MWID/ AHB data size, following the formula given below:
FMC_CLK divider ratio = max (CLKDIV + 1,MWID ( AHB data size ))
If MWID is 16 or 8 bits, the FMC_CLK divider ratio is always defined by the programmed
CLKDIV value.
If MWID is 32 bits, the FMC_CLK divider ratio depends also on AHB data size.
Example:
•

If CLKDIV=1, MWID=32 bits, AHB data size=8 bits, FMC_CLK=HCLK/4.

•

If CLKDIV=1, MWID=16 bits, AHB data size=8 bits, FMC_CLK=HCLK/2.

NOR Flash memories specify a minimum time from NADV assertion to CLK high. To meet
this constraint, the FMC does not issue the clock to the memory during the first internal
clock cycle of the synchronous access (before NADV assertion). This guarantees that the
rising edge of the memory clock occurs in the middle of the NADV low pulse.

Data latency versus NOR memory latency
The data latency is the number of cycles to wait before sampling the data. The DATLAT
value must be consistent with the latency value specified in the NOR Flash configuration

1632/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
register. The FMC does not include the clock cycle when NADV is low in the data latency
count.

Caution:

Some NOR Flash memories include the NADV Low cycle in the data latency count, so that
the exact relation between the NOR Flash latency and the FMC DATLAT parameter can be
either:
•

NOR Flash latency = (DATLAT + 2) CLK clock cycles

•

or NOR Flash latency = (DATLAT + 3) CLK clock cycles

Some recent memories assert NWAIT during the latency phase. In such cases DATLAT can
be set to its minimum value. As a result, the FMC samples the data and waits long enough
to evaluate if the data are valid. Thus the FMC detects when the memory exits latency and
real data are processed.
Other memories do not assert NWAIT during latency. In this case the latency must be set
correctly for both the FMC and the memory, otherwise invalid data are mistaken for good
data, or valid data are lost in the initial phase of the memory access.

Single-burst transfer
When the selected bank is configured in burst mode for synchronous accesses, if for
example an AHB single-burst transaction is requested on 16-bit memories, the FMC
performs a burst transaction of length 1 (if the AHB transfer is 16 bits), or length 2 (if the
AHB transfer is 32 bits) and de-assert the Chip Select signal when the last data is strobed.
Such transfers are not the most efficient in terms of cycles compared to asynchronous read
operations. Nevertheless, a random asynchronous access would first require to re-program
the memory access mode, which would altogether last longer.

Cross boundary page for Cellular RAM 1.5
Cellular RAM 1.5 does not allow burst access to cross the page boundary. The FMC
controller allows to split automatically the burst access when the memory page size is
reached by configuring the CPSIZE bits in the FMC_BCR1 register following the memory
page size.

Wait management
For synchronous NOR Flash memories, NWAIT is evaluated after the programmed latency
period, which corresponds to (DATLAT+2) CLK clock cycles.
If NWAIT is active (low level when WAITPOL = 0, high level when WAITPOL = 1), wait
states are inserted until NWAIT is inactive (high level when WAITPOL = 0, low level when
WAITPOL = 1).
When NWAIT is inactive, the data is considered valid either immediately (bit WAITCFG = 1)
or on the next clock edge (bit WAITCFG = 0).
During wait-state insertion via the NWAIT signal, the controller continues to send clock
pulses to the memory, keeping the Chip Select and output enable signals valid. It does not
consider the data as valid.
In burst mode, there are two timing configurations for the NOR Flash NWAIT signal:
•

The Flash memory asserts the NWAIT signal one data cycle before the wait state
(default after reset).

•

The Flash memory asserts the NWAIT signal during the wait state

RM0090 Rev 17

1633/1747
1680

Flexible memory controller (FMC)

RM0090

The FMC supports both NOR Flash wait state configurations, for each Chip Select, thanks
to the WAITCFG bit in the FMC_BCRx registers (x = 0..3).
Figure 473. Wait configuration waveforms
0HPRU\WUDQVDFWLRQ EXUVWRIKDOIZRUGV
+&/.
&/.
DGGU>@

$>@
1$'9
1:$,7
:$,7&)* 
1:$,7
:$,7&)* 

LQVHUWHGZDLWVWDWH
DGGU>@

$'>@

GDWD

GDWD

GDWD
DLF

Figure 474. Synchronous multiplexed read mode waveforms - NOR, PSRAM (CRAM)
0HPRU\WUDQVDFWLRQ EXUVWRIKDOIZRUGV
+&/.

&/.

$>@

DGGU>@

1([

12(

1:(

+LJK

1$'9
1:$,7
:$,7&)*

'$7/$7
&/.F\FOHV
$'>@

$GGU>@

LQVHUWHGZDLWVWDWH
GDWD

FORFN FORFN
F\FOH
F\FOH

GDWD

'DWDVWUREHV

GDWD

GDWD

'DWDVWUREHV

DLI

1. Byte lane outputs BL are not shown; for NOR access, they are held high, and, for PSRAM (CRAM) access,

1634/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
they are held low.

Table 285. FMC_BCRx bit fields
Bit No.

Bit name

Value to set

31-21

Reserved

0x000

20

CCLKEN

As needed

19

CBURSTRW

No effect on synchronous read

18-16

CPSIZE

As needed (0x1 for CRAM 1.5)

15

ASYNCWAIT

0x0

14

EXTMOD

0x0

13

WAITEN

to be set to 1 if the memory supports this feature, to be kept at 0
otherwise

12

WREN

no effect on synchronous read

11

WAITCFG

to be set according to memory

10

WRAPMOD

9

WAITPOL

to be set according to memory

8

BURSTEN

0x1

7

Reserved

0x1

6

FACCEN

Set according to memory support (NOR Flash memory)

5-4

MWID

As needed

3-2

MTYP[1:0]

0x1 or 0x2

1

MUXEN

As needed

0

MBKEN

0x1

0x0

Table 286. FMC_BTRx bit fields
Bit No.

Bit name

Value to set

31:30

Reserved

0x0

29:28

ACCMOD

0x0

27-24

DATLAT

Data latency

27-24

DATLAT

Data latency

23-20

CLKDIV

0x0 to get CLK = HCLK (Not supported)
0x1 to get CLK = 2 × HCLK
..

19-16

BUSTURN

15-8

DATAST

Don’t care

7-4

ADDHLD

Don’t care

3-0

ADDSET[3:0]

Don’t care

Time between NEx high to NEx low (BUSTURN HCLK)

RM0090 Rev 17

1635/1747
1680

Flexible memory controller (FMC)

RM0090

Figure 475. Synchronous multiplexed write mode waveforms - PSRAM (CRAM)
-EMORY TRANSACTION  BURST OF  HALF WORDS
(#,+

#,+

!;=

ADDR;=

.%X
(I :
./%

.7%

.!$6

.7!)4
7!)4#&'  
$!4,!4 
#,+ CYCLES
!$;=

!DDR;=

INSERTED WAIT STATE
DATA

DATA

 CLOCK  CLOCK

AIF

1. The memory must issue NWAIT signal one cycle in advance, accordingly WAITCFG must be programmed
to 0.
2. Byte Lane (NBL) outputs are not shown, they are held low while NEx is active.

Table 287. FMC_BCRx bit fields

1636/1747

Bit No.

Bit name

Value to set

31-20

Reserved

0x000

20

CCLKEN

As needed

19

CBURSTRW

18-16

CPSIZE

15

ASYNCWAIT

0x0

14

EXTMOD

0x0

13

WAITEN

to be set to 1 if the memory supports this feature, to be kept at 0
otherwise.

12

WREN

0x1
As needed (0x1 for CRAM 1.5)

0x1

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
Table 287. FMC_BCRx bit fields (continued)
Bit No.

Bit name

Value to set

11

WAITCFG

0x0

10

WRAPMOD

0x0

9

WAITPOL

to be set according to memory

8

BURSTEN

no effect on synchronous write

7

Reserved

0x1

6

FACCEN

Set according to memory support

5-4

MWID

3-2

MTYP[1:0]

1

MUXEN

As needed

0

MBKEN

0x1

As needed
0x1

Table 288. FMC_BTRx bit fields
Bit No.

Bit name

Value to set

31-30

Reserved

0x0

29:28

ACCMOD

0x0

27-24

DATLAT

Data latency

23-20

CLKDIV

0x0 to get CLK = HCLK (not supported)
0x1 to get CLK = 2 × HCLK

19-16

BUSTURN

15-8

DATAST

Don’t care

7-4

ADDHLD

Don’t care

3-0

ADDSET[3:0]

Don’t care

Time between NEx high to NEx low (BUSTURN HCLK)

RM0090 Rev 17

1637/1747
1680

Flexible memory controller (FMC)

37.5.6

RM0090

NOR/PSRAM controller registers
SRAM/NOR-Flash chip-select control registers 1..4 (FMC_BCR1..4)
Address offset: 8 * (x – 1), x = 1...4
Reset value: 0x0000 30DB for Bank1 and 0x0000 30D2 for Bank 2 to 4

rw

rw

rw

rw

rw

3

rw

2

rw

rw

rw

1

0
MBKEN

BURSTEN

rw

4

MUXEN

WAITPOL

rw

5

MTYP[1:0]

WRAPMOD

rw

6

MWID[1:0]

WREN

rw

7

FACCEN

8

Reserved

9

WAITCFG

rw rw

14 13 12 11 10
WAITEN

CPSIZE[2:0]

15

EXTMOD

CBURSTRW

Reserved

CCLKEN

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

ASCYCWAIT

This register contains the control information of each memory bank, used for SRAMs,
PSRAM and NOR Flash memories.

rw

rw

Bits 31: 21 Reserved, must be kept at reset value
Bit 20 CCLKEN: Continuous Clock Enable.
This bit enables the FMC_CLK clock output to external memory devices.
0: The FMC_CLK is only generated during the synchronous memory access (read/write
transaction). The FMC_CLK clock ratio is specified by the programmed CLKDIV value in the
FMC_BCRx register (default after reset) .
1: The FMC_CLK is generated continuously during asynchronous and synchronous access. The
FMC_CLK clock is activated when the CCLKEN is set.
Note: The CCLKEN bit of the FMC_BCR2..4 registers is don’t care. It is only enabled through the
FMC_BCR1 register. Bank 1 must be configured in synchronous mode to generate the
FMC_CLK continuous clock.
Note: If CCLKEN bit is set, the FMC_CLK clock ratio is specified by CLKDIV value in the
FMC_BTR1 register. CLKDIV in FMC_BWTR1 is don’t care.
Note: If the synchronous mode is used and CCLKEN bit is set, the synchronous memories
connected to other banks than Bank 1 are clocked by the same clock (the CLKDIV value in
the FMC_BTR2..4 and FMC_BWTR2..4 registers for other banks has no effect.)
Bit 19 CBURSTRW: Write burst enable.
For PSRAM (CRAM) operating in burst mode, the bit enables synchronous accesses during write
operations. The enable bit for synchronous read accesses is the BURSTEN bit in the FMC_BCRx
register.
0: Write operations are always performed in asynchronous mode
1: Write operations are performed in synchronous mode.
Bits 18:16 CPSIZE[2:0]: CRAM page size.
These are used for Cellular RAM 1.5 which does not allow burst access to cross the address
boundaries between pages. When these bits are configured, the FMC controller splits automatically
the burst access when the memory page size is reached (refer to memory datasheet for page size).
000: No burst split when crossing page boundary (default after reset)
001: 128 bytes
010: 256 bytes
011: 512 bytes
100: 1024 bytes
Others: reserved

1638/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)

Bit 15 ASYNCWAIT: Wait signal during asynchronous transfers
This bit enables/disables the FMC to use the wait signal even during an asynchronous protocol.
0: NWAIT signal is not taken in to account when running an asynchronous protocol (default after
reset)
1: NWAIT signal is taken in to account when running an asynchronous protocol
Bit 14 EXTMOD: Extended mode enable.
This bit enables the FMC to program the write timings for non-multiplexed asynchronous accesses
inside the FMC_BWTR register, thus resulting in different timings for read and write operations.
0: values inside FMC_BWTR register are not taken into account (default after reset)
1: values inside FMC_BWTR register are taken into account
Note: When the extended mode is disabled, the FMC can operate in Mode1 or Mode2 as follows:
–
Mode 1 is the default mode when the SRAM/PSRAM memory type is selected
(MTYP[1:0] =0x0 or 0x01)
–
Mode 2 is the default mode when the NOR memory type is selected (MTYP[1:0] = 0x10).
Bit 13 WAITEN: Wait enable bit.
This bit enables/disables wait-state insertion via the NWAIT signal when accessing the memory in
synchronous mode.
0: NWAIT signal is disabled (its level not taken into account, no wait state inserted after the
programmed Flash latency period)
1: NWAIT signal is enabled (its level is taken into account after the programmed latency period to
insert wait states if asserted) (default after reset)
Bit 12 WREN: Write enable bit.
This bit indicates whether write operations are enabled/disabled in the bank by the FMC:
0: Write operations are disabled in the bank by the FMC, an AHB error is reported,
1: Write operations are enabled for the bank by the FMC (default after reset).
Bit 11 WAITCFG: Wait timing configuration.
The NWAIT signal indicates whether the data from the memory are valid or if a wait state must be
inserted when accessing the memory in synchronous mode. This configuration bit determines if
NWAIT is asserted by the memory one clock cycle before the wait state or during the wait state:
0: NWAIT signal is active one data cycle before wait state (default after reset),
1: NWAIT signal is active during wait state (not used for PSRAM).
Bit 10 WRAPMOD: Wrapped burst mode support.
Defines whether the controller will or not split an AHB burst wrap access into two linear accesses.
Valid only when accessing memories in burst mode
0: Direct wrapped burst is not enabled (default after reset),
1: Direct wrapped burst is enabled.
Note: This bit has no effect as the CPU and DMA cannot generate wrapping burst transfers.
Bit 9 WAITPOL: Wait signal polarity bit.
Defines the polarity of the wait signal from memory used for either in synchronous or asynchronous
mode:
0: NWAIT active low (default after reset),
1: NWAIT active high.
Bit 8 BURSTEN: Burst enable bit.
This bit enables/disables synchronous accesses during read operations. It is valid only for
synchronous memories operating in burst mode:
0: Burst mode disabled (default after reset). Read accesses are performed in asynchronous mode.
1: Burst mode enable. Read accesses are performed in synchronous mode.
Bit 7 Reserved, must be kept at reset value

RM0090 Rev 17

1639/1747
1680

Flexible memory controller (FMC)

RM0090

Bit 6 FACCEN: Flash access enable
Enables NOR Flash memory access operations.
0: Corresponding NOR Flash memory access is disabled
1: Corresponding NOR Flash memory access is enabled (default after reset)
Bits 5:4 MWID[1:0]: Memory data bus width.
Defines the external memory device width, valid for all type of memories.
00: 8 bits,
01: 16 bits (default after reset),
10: 32 bits,
11: reserved, do not use.
Bits 3:2 MTYP[1:0]: Memory type.
Defines the type of external memory attached to the corresponding memory bank:
00: SRAM (default after reset for Bank 2...4)
01: PSRAM (CRAM)
10: NOR Flash/OneNAND Flash (default after reset for Bank 1)
11: reserved
Bit 1 MUXEN: Address/data multiplexing enable bit.
When this bit is set, the address and data values are multiplexed on the data bus, valid only with
NOR and PSRAM memories:
0: Address/Data nonmultiplexed
1: Address/Data multiplexed on databus (default after reset)
Bit 0 MBKEN: Memory bank enable bit.
Enables the memory bank. After reset Bank1 is enabled, all others are disabled. Accessing a
disabled bank causes an ERROR on AHB bus.
0: Corresponding memory bank is disabled
1: Corresponding memory bank is enabled

SRAM/NOR-Flash chip-select timing registers 1..4 (FMC_BTR1..4)
Address offset: 0x04 + 8 * (x – 1), x = 1..4
Reset value: 0x0FFF FFFF
Reset value: 0x0FFF FFFF
FMC_BTRx bits are written by software to add a delay at the end of a read /write
transaction. This delay allows matching the minimum time between consecutive
transactions (tEHEL from NEx high to FMC_NEx low) and the maximum time required by the
memory to free the data bus after a read access (tEHQZ).
This register contains the control information of each memory bank, used for SRAMs,
PSRAM and NOR Flash memories.If the EXTMOD bit is set in the FMC_BCRx register, then
this register is partitioned for write and read access, that is, 2 registers are available: one to
configure read accesses (this register) and one to configure write accesses (FMC_BWTRx
registers).

1640/1747

RM0090 Rev 17

Flexible memory controller (FMC)

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

8

7

6

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

5

4

3

2

rw

rw

rw

rw

ADDHLD3:0]

9

DATAST7:0]

BUSTURN3:0]

CLKDIV3:0]

DATLAT[3:0]

ACCMOD[1:0]

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

rw

1

0

rw

rw

ADDSET[3:0]

RM0090

Bits 31:30 Reserved, must be kept at reset value
Bits 29:28 ACCMOD[1:0]: Access mode
Specifies the asynchronous access modes as shown in the timing diagrams. These bits are
taken into account only when the EXTMOD bit in the FMC_BCRx register is 1.
00: access mode A
01: access mode B
10: access mode C
11: access mode D
Bits 27:24 DATLAT[3:0]: Data latency for synchronous memory (see note below bit description table)
For synchronous accesses with read/write burst mode enabled (BURSTEN / CBURSTRW bits
set), this field defines the number of memory clock cycles (+2) to issue to the memory before
reading/writing the first data. This timing parameter is not expressed in HCLK periods, but in
FMC_CLK periods. For asynchronous accesses, this value is don't care.
0000: Data latency of 2 CLK clock cycles for first burst access
1111: Data latency of 17 CLK clock cycles for first burst access (default value after reset)
Bits 23:20 CLKDIV[3:0]: Clock divide ratio (for FMC_CLK signal)
Defines the period of FMC_CLK clock output signal, expressed in number of HCLK cycles:
0000: Reserved
0001: FMC_CLK period = 2 × HCLK periods
0010: FMC_CLK period = 3 × HCLK periods
1111: FMC_CLK period = 16 × HCLK periods (default value after reset)
In asynchronous NOR Flash, SRAM or PSRAM accesses, this value is don’t care.
Note: Refer to section 37.5.5: Synchronous transactions for FMC_CLK divider ratio formula)

RM0090 Rev 17

1641/1747
1680

Flexible memory controller (FMC)

RM0090

Bits 19:16 BUSTURN[3:0]: Bus turnaround phase duration
These bits are written by software to add a delay at the end of a write-to-read (and read-towrite) transaction. This delay allows to match the minimum time between consecutive
transactions (tEHEL from NEx high to NEx low) and the maximum time needed by the memory
to free the data bus after a read access (tEHQZ). The programmed bus turnaround delay is
inserted between an asynchronous read (muxed or mode D) or write transaction and any
other asynchronous /synchronous read or write to or from a static bank. The bank can be the
same or different in case of read, in case of write the bank can be different except for muxed
or mode D.
In some cases, whatever the programmed BUSTRUN values, the bus turnaround delay is
fixed as follows:
• The bus turnaround delay is not inserted between two consecutive asynchronous write
transfers to the same static memory bank except for modes muxed and D.
• There is a bus turnaround delay of 1 FMC clock cycle between:
–Two consecutive asynchronous read transfers to the same static memory bank except for
modes muxed and D.
–An asynchronous read to an asynchronous or synchronous write to any static bank or
dynamic bank except for modes muxed and D.
–An asynchronous (modes 1, 2, A, B or C) read and a read from another static bank.
• There is a bus turnaround delay of 2 FMC clock cycle between:
–Two consecutive synchronous writes (burst or single) to the same bank.
–A synchronous write (burst or single) access and an asynchronous write or read transfer to
or from static memory bank (the bank can be the same or different for the case of read.
–Two consecutive synchronous reads (burst or single) followed by any
synchronous/asynchronous read or write from/to another static memory bank.
• There is a bus turnaround delay of 3 FMC clock cycle between:
–Two consecutive synchronous writes (burst or single) to different static bank.
–A synchronous write (burst or single) access and a synchronous read from the same or a
different bank.
0000: BUSTURN phase duration = 0 HCLK clock cycle added
...
1111: BUSTURN phase duration = 15 x HCLK clock cycles added (default value after reset)

1642/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)

Bits 15:8 DATAST[7:0]: Data-phase duration
These bits are written by software to define the duration of the data phase (refer to Figure 458
to Figure 470), used in asynchronous accesses:
0000 0000: Reserved
0000 0001: DATAST phase duration = 1 × HCLK clock cycles
0000 0010: DATAST phase duration = 2 × HCLK clock cycles
...
1111 1111: DATAST phase duration = 255 × HCLK clock cycles (default value after reset)
For each memory type and access mode data-phase duration, please refer to the respective
figure (Figure 458 to Figure 470).
Example: Mode1, write access, DATAST=1: Data-phase duration= DATAST+1 = 2 HCLK clock
cycles.
Note: In synchronous accesses, this value is don’t care.
Bits 7:4 ADDHLD[3:0]: Address-hold phase duration
These bits are written by software to define the duration of the address hold phase (refer to
Figure 467 to Figure 470), used in mode D or multiplexed accesses:
0000: Reserved
0001: ADDHLD phase duration =1 × HCLK clock cycle
0010: ADDHLD phase duration = 2 × HCLK clock cycle
...
1111: ADDHLD phase duration = 15 × HCLK clock cycles (default value after reset)
For each access mode address-hold phase duration, please refer to the respective figure
(Figure 467 to Figure 470).
Note: In synchronous accesses, this value is not used, the address hold phase is always 1
memory clock period duration.
Bits 3:0 ADDSET[3:0]: Address setup phase duration
These bits are written by software to define the duration of the address setup phase (refer to
Figure 458 to Figure 470), used in SRAMs, ROMs and asynchronous NOR Flash and PSRAM
accesses:
0000: ADDSET phase duration = 0 × HCLK clock cycle
...
1111: ADDSET phase duration = 15 × HCLK clock cycles (default value after reset)
For each access mode address setup phase duration, please refer to the respective figure
(refer to Figure 458 to Figure 470).
Note: In synchronous accesses, this value is don’t care.
In Muxed mode or Mode D, the minimum value for ADDSET is 1.

Note:

PSRAMs (CRAMs) have a variable latency due to internal refresh. Therefore these
memories issue the NWAIT signal during the whole latency phase to prolong the latency as
needed.
With PSRAMs (CRAMs) the filled DATLAT must be set to 0, so that the FMC exits its latency
phase soon and starts sampling NWAIT from memory, then starts to read or write when the
memory is ready.
This method can be used also with the latest generation of synchronous Flash memories
that issue the NWAIT signal, unlike older Flash memories (check the datasheet of the
specific Flash memory being used).

RM0090 Rev 17

1643/1747
1680

Flexible memory controller (FMC)

RM0090

SRAM/NOR-Flash write timing registers 1..4 (FMC_BWTR1..4)
Address offset: 0x104 + 8 * (x – 1), x = 1...4
Reset value: 0x0FFF FFFF
This register contains the control information of each memory bank. It is used for SRAMs,
PSRAMs and NOR Flash memories. When the EXTMOD bit is set in the FMC_BCRx
register, then this register is active for write access.

rw

rw

rw

rw

rw

8

7

6

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

5

4

3

2

rw

rw

rw

rw

rw

1

0

rw

rw

ADDSET[3:0]

ADDHLD[3:0]

9

DATAST[7:0]

BUSTURN[3:0]

Reserved

ACCMOD[1:0]

Reserved

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

Bits 31:30 Reserved, must be kept at reset value
Bits 29:28 ACCMOD[1:0]: Access mode.
Specifies the asynchronous access modes as shown in the next timing diagrams.These bits are
taken into account only when the EXTMOD bit in the FMC_BCRx register is 1.
00: access mode A
01: access mode B
10: access mode C
11: access mode D
Bits 27:20 Reserved, must be kept at reset value
Bits 19:16 BUSTURN[3:0]: Bus turnaround phase duration
The programmed bus turnaround delay is inserted between an asynchronous write transfer and
any other asynchronous /synchronous read or write transfer to or from a static bank. The bank can
be the same or different in case of read, in case of write the bank can be different expect for muxed
or mode D.
In some cases, whatever the programmed BUSTRUN values, the bus turnaround delay is fixed as
follows:
• The bus turnaround delay is not inserted between two consecutive asynchronous write transfers to
the same static memory bank except for modes muxed and D.
• There is a bus turnaround delay of 2 FMC clock cycle between:
–Two consecutive synchronous writes (burst or single) to the same bank.
–A synchronous write (burst or single) transfer and an asynchronous write or read transfer to or
from static memory bank.
• There is a bus turnaround delay of 3 FMC clock cycle between:
–Two consecutive synchronous writes (burst or single) to different static bank.
–A synchronous write (burst or single) transfer and a synchronous read from the same or a
different bank.
0000: BUSTURN phase duration = 0 HCLK clock cycle added
...
1111: BUSTURN phase duration = 15 HCLK clock cycles added (default value after reset)

1644/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)

Bits 15:8 DATAST[3:0]: Data-phase duration.
These bits are written by software to define the duration of the data phase (refer toFigure 458 to
Figure 470), used in asynchronous SRAM, PSRAM and NOR Flash memory accesses:
0000 0000: Reserved
0000 0001: DATAST phase duration = 1 × HCLK clock cycles
0000 0010: DATAST phase duration = 2 × HCLK clock cycles
...
1111 1111: DATAST phase duration = 255 × HCLK clock cycles (default value after reset)
Bits 7:4 ADDHLD[3:0]: Address-hold phase duration.
These bits are written by software to define the duration of the address hold phase (refer to
Figure 467 to Figure 470), used in asynchronous multiplexed accesses:
0000: Reserved
0001: ADDHLD phase duration = 1 × HCLK clock cycle
0010: ADDHLD phase duration = 2 × HCLK clock cycle
...
1111: ADDHLD phase duration = 15 × HCLK clock cycles (default value after reset)
Note: In synchronous NOR Flash accesses, this value is not used, the address hold phase is always
1 Flash clock period duration.
Bits 3:0 ADDSET[3:0]: Address setup phase duration.
These bits are written by software to define the duration of the address setup phase in HCLK cycles
(refer to Figure 467 to Figure 470), used in asynchronous accesses:
0000: ADDSET phase duration = 0 × HCLK clock cycle
...
1111: ADDSET phase duration = 15 × HCLK clock cycles (default value after reset)
Note: In synchronous NOR Flash and PSRAM accesses, this value is not used, the address setup
phase is always 1 Flash clock period duration. In muxed mode, the minimum ADDSET value
is 1.

37.6

NAND Flash/PC Card controller
The FMC generates the appropriate signal timings to drive the following types of device:
•

8- and 16-bit NAND Flash memories

•

16-bit PC Card compatible devices

The NAND Flash/PC Card controller can control three external banks, Bank 2, 3 and 4:
•

Bank 2 and Bank 3 support NAND Flash devices

•

Bank 4 supports PC Card devices.

Each bank is configured through dedicated registers (Section 37.6.8). The programmable
memory parameters include access timings (shown in Table 289) and ECC configuration.

RM0090 Rev 17

1645/1747
1680

Flexible memory controller (FMC)

RM0090

Table 289. Programmable NAND Flash/PC Card access parameters

37.6.1

Parameter

Function

Access mode

Unit

Min. Max.

Memory setup
time

Number of clock cycles (HCLK)
required to set up the address
before the command assertion

Read/Write

AHB clock cycle
(HCLK)

1

256

Memory wait

Minimum duration (in HCLK clock
cycles) of the command assertion

Read/Write

AHB clock cycle
(HCLK)

2

255

Memory hold

Number of clock cycles (HCLK)
during which the address must be
held (as well as the data if a write
access is performed) after the
command de-assertion

Read/Write

AHB clock cycle
(HCLK)

1

254

Memory
databus high-Z

Number of clock cycles (HCLK)
during which the data bus is kept
in high-Z state after a write
access has started

Write

AHB clock cycle
(HCLK)

1

255

External memory interface signals
The following tables list the signals that are typically used to interface NAND Flash memory
and PC Card.

Note:

The prefix “N” identifies the signals which are active low.

8-bit NAND Flash memory
t

Table 290. 8-bit NAND Flash
FMC signal name

I/O

Function

A[17]

O

NAND Flash address latch enable (ALE) signal

A[16]

O

NAND Flash command latch enable (CLE) signal

D[7:0]

I/O

8-bit multiplexed, bidirectional address/data bus

NCE[x]

O

Chip Select, x = 2, 3

NOE(= NRE)

O

Output enable (memory signal name: read enable, NRE)

NWE

O

Write enable

NWAIT/INT[3:2]

I

NAND Flash ready/busy input signal to the FMC

Theoretically, there is no capacity limitation as the FMC can manage as many address
cycles as needed.

1646/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)

16-bit NAND Flash memory
Table 291. 16-bit NAND Flash
FMC signal name

I/O

Function

A[17]

O

NAND Flash address latch enable (ALE) signal

A[16]

O

NAND Flash command latch enable (CLE) signal

D[15:0]

I/O

16-bit multiplexed, bidirectional address/data bus

NCE[x]

O

Chip Select, x = 2, 3

NOE(= NRE)

O

Output enable (memory signal name: read enable, NRE)

NWE

O

Write enable

NWAIT/INT[3:2]

I

NAND Flash ready/busy input signal to the FMC

Theoretically, there is no capacity limitation as the FMC can manage as many address
cycles as needed.
Table 292. 16-bit PC Card
FMC signal name

I/O

Function

A[10:0]

O

Address bus

NIORD

O

Output enable for I/O space

NIOWR

O

Write enable for I/O space

NREG

O

Register signal indicating if access is in Common or Attribute space

D[15:0]

I/O

Bidirectional databus

NCE4_1

O

Chip Select 1

NCE4_2

O

Chip Select 2 (indicates if access is 16-bit or 8-bit)

NOE

O

Output enable in Common and in Attribute space

NWE

O

Write enable in Common and in Attribute space

NWAIT

I

PC Card wait input signal to the FMC (memory signal name IORDY)

INTR

I

PC Card interrupt to the FMC (only for PC Cards that can generate
an interrupt)

CD

I

PC Card presence detection. Active high. If an access is performed
to the PC Card banks while CD is low, an AHB error is generated.
Refer to Section 37.3: AHB interface

RM0090 Rev 17

1647/1747
1680

Flexible memory controller (FMC)

37.6.2

RM0090

NAND Flash / PC Card supported memories and transactions
Table 293 shows the supported devices, access modes and transactions. Transactions not
allowed (or not supported) by the NAND Flash / PC Card controller are shown in gray.
Table 293. Supported memories and transactions
Device

NAND 8-bit

NAND 16-bit

37.6.3

Mode

R/W

AHB
Memory
Allowed/
data size data size not allowed

Comments

Asynchronous R

8

8

Y

-

Asynchronous W

8

8

Y

-

Asynchronous R

16

8

Y

Split into 2 FMC accesses

Asynchronous W

16

8

Y

Split into 2 FMC accesses

Asynchronous R

32

8

Y

Split into 4 FMC accesses

Asynchronous W

32

8

Y

Split into 4 FMC accesses

Asynchronous R

8

16

Y

-

Asynchronous W

8

16

N

-

Asynchronous R

16

16

Y

-

Asynchronous W

16

16

Y

-

Asynchronous R

32

16

Y

Split into 2 FMC accesses

Asynchronous W

32

16

Y

Split into 2 FMC accesses

Timing diagrams for NAND Flash memory and PC Card
Each PC Card/CompactFlash and NAND Flash memory bank is managed through a set of
registers:
•

Control register: FMC_PCRx

•

Interrupt status register: FMC_SRx

•

ECC register: FMC_ECCRx

•

Timing register for Common memory space: FMC_PMEMx

•

Timing register for Attribute memory space: FMC_PATTx

•

Timing register for I/O space: FMC_PIOx

Each timing configuration register contains three parameters used to define number of
HCLK cycles for the three phases of any PC Card/CompactFlash or NAND Flash access,
plus one parameter that defines the timing for starting driving the data bus when a write
access is performed. Figure 476 shows the timing parameter definitions for common
memory accesses, knowing that Attribute and I/O (only for PC Card) memory space access
timings are similar.

1648/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
Figure 476. NAND Flash/PC Card controller waveforms for common memory access
+&/.
$>@
1&([
+LJK

15(*
1,2:
1,25
1:(
12(

0(0[6(7


0(0[:$,7

0(0[+2/'


0(0[+,=

ZULWHBGDWD
UHDGBGDWD

9DOLG

069

1. NOE remains high (inactive) during write accesses. NWE remains high (inactive) during read accesses.
2. For write accesses, the hold phase delay is (MEMHOLD) x HCLK cycles, while it is (MEMHOLD + 2) x
HCLK cycles for read accesses.

37.6.4

NAND Flash operations
The command latch enable (CLE) and address latch enable (ALE) signals of the NAND
Flash memory device are driven by address signals from the FMC controller. This means
that to send a command or an address to the NAND Flash memory, the CPU has to perform
a write to a specific address in its memory space.
A typical page read operation from the NAND Flash device requires the following steps:
3.

Program and enable the corresponding memory bank by configuring the FMC_PCRx
and FMC_PMEMx (and for some devices, FMC_PATTx, see Section 37.6.5: NAND
Flash prewait functionality) registers according to the characteristics of the NAND
Flash memory (PWID bits for the data bus width of the NAND Flash, PTYP = 1,
PWAITEN = 0 or 1 as needed, see section Section 37.4.2: NAND Flash memory/PC
Card address mapping for timing configuration).

4.

The CPU performs a byte write to the common memory space, with data byte equal to
one Flash command byte (for example 0x00 for Samsung NAND Flash devices). The
LE input of the NAND Flash memory is active during the write strobe (low pulse on
NWE), thus the written byte is interpreted as a command by the NAND Flash memory.
Once the command is latched by the memory device, it does not need to be written
again for the following page read operations.

5.

The CPU can send the start address (STARTAD) for a read operation by writing four
byte (or three for smaller capacity devices), STARTAD[7:0], STARTAD[16:9],
STARTAD[24:17] and finally STARTAD[25] (for 64 Mb x 8 bit NAND Flash memories) in
the common memory or attribute space. The ALE input of the NAND Flash device is
active during the write strobe (low pulse on NWE), thus the written byte are interpreted
as the start address for read operations. Using the attribute memory space makes it
possible to use a different timing configuration of the FMC, which can be used to
implement the prewait functionality needed by some NAND Flash memories (see

RM0090 Rev 17

1649/1747
1680

Flexible memory controller (FMC)

RM0090

details in Section 37.6.5: NAND Flash prewait functionality).

37.6.5

6.

The controller waits for the NAND Flash memory to be ready (R/NB signal high), before
starting a new access to the same or another memory bank. While waiting, the
controller holds the NCE signal active (low).

7.

The CPU can then perform byte read operations from the common memory space to
read the NAND Flash page (data field + Spare field) byte by byte.

8.

The next NAND Flash page can be read without any CPU command or address write
operation. This can be done in three different ways:
–

by simply performing the operation described in step 5

–

a new random address can be accessed by restarting the operation at step 3

–

a new command can be sent to the NAND Flash device by restarting at step 2

NAND Flash prewait functionality
Some NAND Flash devices require that, after writing the last part of the address, the
controller waits for the R/NB signal to go low. (see Figure 457).
Figure 477. Access to non ‘CE don’t care’ NAND-Flash

1. CPU wrote byte 0x00 at address 0x7001 0000.
2. CPU wrote byte A7~A0 at address 0x7002 0000.
3. CPU wrote byte A16~A9 at address 0x7002 0000.
4. CPU wrote byte A24~A17 at address 0x7002 0000.
5. CPU wrote byte A25 at address 0x7802 0000: FMC performs a write access using FMC_PATT2 timing
definition, where ATTHOLD ≥ 7 (providing that (7+1) × HCLK = 112 ns > tWB max). This guarantees that
NCE remains low until R/NB goes low and high again (only requested for NAND Flash memories where
NCE is not don’t care).

When this functionality is required, it can be ensured by programming the MEMHOLD value
to meet the tWB timing. However CPU read accesses to the NAND Flash memory has a hold

1650/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
delay of (MEMHOLD + 2) x HCLK cycles, while CPU write accesses have a hold delay of
(MEMHOLD) x HCLK cycles.
To cope with this timing constraint, the attribute memory space can be used by
programming its timing register with an ATTHOLD value that meets the tWB timing, and by
keeping the MEMHOLD value at its minimum value. The CPU must then use the common
memory space for all NAND Flash read and write accesses, except when writing the last
address byte to the NAND Flash device, where the CPU must write to the attribute memory
space.

37.6.6

Computation of the error correction code (ECC)
in NAND Flash memory
The FMC PC Card controller includes two error correction code computation hardware
blocks, one per memory bank. They reduce the host CPU workload when processing the
ECC by software.
These two ECC blocks are identical and associated with Bank 2 and Bank 3. As a
consequence, no hardware ECC computation is available for memories connected to Bank
4.
The ECC algorithm implemented in the FMC can perform 1-bit error correction and 2-bit
error detection per 256, 512, 1 024, 2 048, 4 096 or 8 192 byte read or written from/to the
NAND Flash memory. It is based on the Hamming coding algorithm and consists in
calculating the row and column parity.
The ECC modules monitor the NAND Flash data bus and read/write signals (NCE and
NWE) each time the NAND Flash memory bank is active.
The ECC operates as follows:
•

When accessing NAND Flash memory bank 2 or bank 3, the data present on the
D[15:0] bus is latched and used for ECC computation.

•

When accessing any other address in NAND Flash memory, the ECC logic is idle, and
does not perform any operation. As a result, write operations to define commands or
addresses to the NAND Flash memory are not taken into account for ECC
computation.

Once the desired number of byte has been read/written from/to the NAND Flash memory by
the host CPU, the FMC_ECCR2/3 registers must be read to retrieve the computed value.
Once read, they should be cleared by resetting the ECCEN bit to ‘0’. To compute a new data
block, the ECCEN bit must be set to one in the FMC_PCR2/3 registers.
To perform an ECC computation:
1.

Enable the ECCEN bit in the FMC_PCR2/3 register.

2.

Write data to the NAND Flash memory page. While the NAND page is written, the ECC
block computes the ECC value.

3.

Read the ECC value available in the FMC_ECCR2/3 register and store it in a variable.

4.

Clear the ECCEN bit and then enable it in the FMC_PCR2/3 register before reading
back the written data from the NAND page. While the NAND page is read, the ECC
block computes the ECC value.

5.

Read the new ECC value available in the FMC_ECCR2/3 register.

6.

If the two ECC values are the same, no correction is required, otherwise there is an
ECC error and the software correction routine returns information on whether the error
can be corrected or not.
RM0090 Rev 17

1651/1747
1680

Flexible memory controller (FMC)

37.6.7

RM0090

PC Card/CompactFlash operations
Address spaces and memory accesses
The FMC supports CompactFlash devices and PC Cards in Memory mode and I/O mode
(True IDE mode is not supported).
The CompactFlash and PC Cards are made of 3 memory spaces:
•

Common Memory space

•

Attribute space

•

I/O Memory space

The nCE2 and nCE1 pins (FMC_NCE4_2 and FMC_NCE4_1 respectively) select the card
and indicate whether a byte or a word operation is being performed: nCE2 accesses the odd
byte on D15-8 and nCE1 accesses the even byte on D7-0 if A0=0 or the odd byte on D7-0 if
A0=1. The full word is accessed on D15-0 if both nCE2 and nCE1 are low.
The memory space is selected by asserting low nOE for read accesses or nWE for write
accesses, combined with the low assertion of nCE2/nCE1 and nREG.
•

If pin nREG=1 during the memory access, the common memory space is selected

•

If pin nREG=0 during the memory access, the attribute memory space is selected

The I/O space is selected by asserting nIORD space for read accesses or nIOWR for write
accesses [instead of nOE/nWE for memory space], combined with nCE2/nCE1. Note that
nREG must also be asserted low when accessing I/O space.
Three type of accesses are allowed for a 16-bit PC Card:
•

Accesses to Common Memory space for data storage can be either 8-bit accesses at
even addresses or 16-bit AHB accesses.
Note that 8-bit accesses at odd addresses are not supported and nCE2 will not be
driven low. A 32-bit AHB request is translated into two 16-bit memory accesses.

•

Accesses to Attribute Memory space where the PC Card stores configuration
information are limited to 8-bit AHB accesses at even addresses.
Note that a 16-bit AHB access will be converted into a single 8-bit memory transfer:
nCE1 will be asserted low, nCE2 will be asserted high and only the even byte on D7-D0
will be valid. Instead a 32-bit AHB access will be converted into two 8-bit memory
transfers at even addresses: nCE1 will be asserted low, NCE2 will be asserted high
and only the even byte will be valid.

•

Accesses to I/O space can be either 8-bit or 16 bit AHB accesses.

nCE2

nCE1

nREG

nOE/nWE

nIORD

A10

A9

A7-1

A0

Table 294. 16-bit PC-Card signals and access type

1

0

1

0

1

X

X

X-X

X

0

1

1

0

1

X

X

X-X

X

0

0

1

0

1

X

X

X-X

0

1652/1747

Space

Common
Memory
Space

Access type

Allowed/not
Allowed

Read/Write byte on D7-D0

YES

Read/Write byte on D15-D8

Not supported

Read/Write word on D15-D0

YES

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)

nCE2

nCE1

nREG

nOE/nWE

nIORD

A10

A9

A7-1

A0

Table 294. 16-bit PC-Card signals and access type (continued)

X

0

0

0

1

0

1

X-X

0

X

0

0

0

1

0

0

X-X

0

1

0

0

0

1

X

X

X-X

1

0

1

0

0

1

X

X

X-X

x

1

0

0

1

0

X

X

X-X

0

Read Even Byte on D7-0

YES

1

0

0

1

0

X

X

X-X

1

Read Odd Byte on D7-0

YES

1

0

0

1

0

X

X

X-X

0

Write Even Byte on D7-0

YES

1

0

0

1

0

X

X

X-X

1

Write Odd Byte on D7-0

YES

0

0

0

1

0

X

X

X-X

0

Read Word on D15-0

YES

0

0

0

1

0

X

X

X-X

0

Write word on D15-0

YES

0

1

0

1

0

X

X

X-X

X

Read Odd Byte on D15-8

Not supported

0

1

0

1

0

X

X

X-X

X

Write Odd Byte on D15-8

Not supported

Access type

Allowed/not
Allowed

Read or Write Configuration
Registers

YES

Read or Write CIS (Card
Information Structure)

YES

Invalid Read or Write (odd
address)

YES

Invalid Read or Write (odd
address)

YES

Space

Attribute
Space

Attribute
Space

I/O space

FMC Bank 4 gives access to those 3 memory spaces as described in Section 37.4.2: NAND
Flash memory/PC Card address mapping and Table 256: NAND/PC Card memory mapping
and timing registers.

Wait feature
The CompactFlash or PC Card may request the FMC to extend the length of the access
phase programmed by MEMWAITx/ATTWAITx/IOWAITx bits, asserting the nWAIT signal
after nOE/nWE or nIORD/nIOWR activation if the wait feature is enabled through the
PWAITEN bit in the FMC_PCRx register. To detect correctly the nWAIT assertion, the
MEMWAITx/ATTWAITx/IOWAITx bits must be programmed as follows:
max_wait_assertion_time
xxWAITx ≥ 4 + ------------------------------------------------------------------HCLK
where max_wait_assertion_time is the maximum time taken by NWAIT to go low once
nOE/nWE or nIORD/nIOWR is low.
After WAIT de-assertion, the FMC extends the WAIT phase for 4 HCLK clock cycles.

RM0090 Rev 17

1653/1747
1680

Flexible memory controller (FMC)

37.6.8

RM0090

NAND Flash/PC Card controller registers
PC Card/NAND Flash control registers 2..4 (FMC_PCR2..4)
Address offset: 0x40 + 0x20 * (x – 1), x = 2..4

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

4

rw

rw

3

2

1

rw

rw

rw

Bits 31:20 Reserved, must be kept at reset value
Bits 19:17 ECCPS[2:0]: ECC page size.
Defines the page size for the extended ECC:
000: 256 byte
001: 512 byte
010: 1024 byte
011: 2048 byte
100: 4096 byte
101: 8192 byte
Bits 16:13 TAR[3:0]: ALE to RE delay.
Sets time from ALE low to RE low in number of AHB clock cycles (HCLK).
Time is: t_ar = (TAR + SET + 2) × THCLK where THCLK is the HCLK clock period
0000: 1 HCLK cycle (default)
1111: 16 HCLK cycles
Note: SET is MEMSET or ATTSET according to the addressed space.
Bits 12:9 TCLR[3:0]: CLE to RE delay.
Sets time from CLE low to RE low in number of AHB clock cycles (HCLK).
Time is t_clr = (TCLR + SET + 2) × THCLK where THCLK is the HCLK clock period
0000: 1 HCLK cycle (default)
1111: 16 HCLK cycles
Note: SET is MEMSET or ATTSET according to the addressed space.
Bits 8:7 Reserved, must be kept at reset value
Bit 6 ECCEN: ECC computation logic enable bit
0: ECC logic is disabled and reset (default after reset),
1: ECC logic is enabled.
Bits 5:4 PWID[1:0]: Data bus width.
Defines the external memory device width.
00: 8 bits
01: 16 bits (default after reset). This value is mandatory for PC Cards.
10: reserved, do not use
11: reserved, do not use

1654/1747

RM0090 Rev 17

0
Reserved

5
PWID[1:0]

6
ECCEN

Reserved

7

PBKEN

rw

8

PWAITEN

rw

9

TCLR[3:0]

Reserved

TAR[3:0]

ECCPS[2:0]

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

PTYP

Reset value: 0x0000 0018

RM0090

Flexible memory controller (FMC)

Bit 3 PTYP: Memory type.
Defines the type of device attached to the corresponding memory bank:
0: PC Card, CompactFlash, CF+ or PCMCIA
1: NAND Flash (default after reset)
Bit 2 PBKEN: PC Card/NAND Flash memory bank enable bit.
Enables the memory bank. Accessing a disabled memory bank causes an ERROR on AHB
bus
0: Corresponding memory bank is disabled (default after reset)
1: Corresponding memory bank is enabled
Bit 1 PWAITEN: Wait feature enable bit.
Enables the Wait feature for the PC Card/NAND Flash memory bank:
0: disabled
1: enabled
Note: For a PC Card, when the wait feature is enabled, the MEMWAITx/ATTWAITx/IOWAITx
bits must be programmed to a value as follows:
xxWAITx ≥ 4 + max_wait_assertion_time/HCLK
Where max_wait_assertion_time is the maximum time taken by NWAIT to go low once
nOE/nWE or nIORD/nIOWR is low.
Bit 0 Reserved.

FIFO status and interrupt register 2..4 (FMC_SR2..4)
Address offset: 0x44 + 0x20 * (x-1), x = 2..4
Reset value: 0x0000 0040
This register contains information about the FIFO status and interrupt. The FMC features a
FIFO that is used when writing to memories to transfer up to 16 words of data from the AHB.
This is used to quickly write to the FIFO and free the AHB for transactions to peripherals
other than the FMC, while the FMC is draining its FIFO into the memory. One of these
register bits indicates the status of the FIFO, for ECC purposes.

6

5

4

3

2

1

0

IFS

ILS

IRS

7

ILEN

8

IREN

Reserved

9

IFEN

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

FEMPT

The ECC is calculated while the data are written to the memory. To read the correct ECC,
the software must consequently wait until the FIFO is empty.

r

rw

rw

rw

rw

rw

rw

Bits 31:7 Reserved, must be kept at reset value
Bit 6 FEMPT: FIFO empty.
Read-only bit that provides the status of the FIFO
0: FIFO not empty
1: FIFO empty
Bit 5 IFEN: Interrupt falling edge detection enable bit
0: Interrupt falling edge detection request disabled
1: Interrupt falling edge detection request enabled

RM0090 Rev 17

1655/1747
1680

Flexible memory controller (FMC)

RM0090

Bit 4 ILEN: Interrupt high-level detection enable bit
0: Interrupt high-level detection request disabled
1: Interrupt high-level detection request enabled
Bit 3 IREN: Interrupt rising edge detection enable bit
0: Interrupt rising edge detection request disabled
1: Interrupt rising edge detection request enabled
Bit 2 IFS: Interrupt falling edge status
The flag is set by hardware and reset by software.
0: No interrupt falling edge occurred
1: Interrupt falling edge occurred

Note: This bit is set by programming it to 1 by software.
Bit 1 ILS: Interrupt high-level status
The flag is set by hardware and reset by software.
0: No Interrupt high-level occurred
1: Interrupt high-level occurred
Bit 0 IRS: Interrupt rising edge status
The flag is set by hardware and reset by software.
0: No interrupt rising edge occurred
1: Interrupt rising edge occurred

Note: This bit is set by programming it to 1 by software.

Common memory space timing register 2..4 (FMC_PMEM2..4)
Address offset: Address: 0x48 + 0x20 * (x – 1), x = 2..4
Reset value: 0xFCFC FCFC
Each FMC_PMEMx (x = 2..4) read/write register contains the timing information for PC Card
or NAND Flash memory bank x. This information is used to access either the common
memory space of the 16-bit PC Card/CompactFlash, or the NAND Flash for command,
address write access and data read/write access.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
MEMHIZ[7:0]
rw

rw

rw

rw

rw

rw

MEMHOLD[7:0]
rw

rw

rw

rw

rw

rw

rw

9

8

7

6

MEMWAIT[7:0]

rw

rw

rw

rw

rw

rw

rw

rw

rw

5

4

3

2

1

0

rw

rw

MEMSET[7:0]
rw

rw

rw

rw

rw

rw

rw

rw

Bits 31:24 MEMHIZ[7:0]: Common memory x data bus Hi-Z time
Defines the number of HCLK clock cycles during which the data bus is kept Hi-Z after the
start of a PC Card/NAND Flash write access to common memory space on socket x. This is
only valid for write transactions:
0000 0000: 1 HCLK cycle
1111 1110: 255 HCLK cycles
1111 1111: Reserved.

1656/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)

Bits 23:16 MEMHOLD[7:0]: Common memory x hold time
For NAND Flash read accesses to the common memory space, these bits define the
number of (HCLK+2) clock cycles during which the address is held after the command is
deasserted (NWE, NOE).
For NAND Flash write accesses to the common memory space, these bits define the
number of HCLK clock cycles during which the data are held after the command is
deasserted (NWE, NOE).
0000 0000: reserved
0000 0001: 1 HCLK cycle for write accesses, 3 HCLK cycles for read accesses
1111 1110: 254 HCLK cycle for write accesses, 256 HCLK cycles for read accesses
1111 1111: Reserved.
Bits 15:8 MEMWAIT[7:0]: Common memory x wait time
Defines the minimum number of HCLK (+1) clock cycles to assert the command (NWE,
NOE), for PC Card/NAND Flash read or write access to common memory space on socket
x. The duration of command assertion is extended if the wait signal (NWAIT) is active (low)
at the end of the programmed value of HCLK:
0000 0000: reserved
0000 0001: 2 HCLK cycles (+ wait cycle introduced by deasserting NWAIT)
1111 1110: 255 HCLK cycles (+ wait cycle introduced by deasserting NWAIT)
1111 1111: Reserved
Bits 7:0 MEMSET[7:0]: Common memory x setup time
Defines the number of HCLK (+1) clock cycles to set up the address before the command
assertion (NWE, NOE), for PC Card/NAND Flash read or write access to common memory
space on socket x:
0000 0000: 1 HCLK cycle
1111 1110: 255 HCLK cycles
1111 1111: Reserved.

Attribute memory space timing registers 2..4 (FMC_PATT2..4)
Address offset: 0x4C + 0x20 * (x – 1), x = 2..4
Reset value: 0xFCFC FCFC
Each FMC_PATTx (x = 2..4) read/write register contains the timing information for PC
Card/CompactFlash or NAND Flash memory bank x. It is used for 8-bit accesses to the
attribute memory space of the PC Card/CompactFlash or to access the NAND Flash for the
last address write access if the timing must differ from that of previous accesses (for
Ready/Busy management, refer to Section 37.6.5: NAND Flash prewait functionality).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
ATTHIZ[7:0]
rw

rw

rw

rw

rw

rw

ATTHOLD[7:0]
rw

rw

rw

rw

rw

rw

rw

9

8

7

6

5

rw

rw

rw

rw

rw

ATTWAIT[7:0]

rw

rw

rw

rw

rw

rw

rw

rw

rw

4

3

2

1

0

rw

rw

ATTSET[7:0]
rw

rw

rw

Bits 31:24 ATTHIZ[7:0]: Attribute memory x data bus Hi-Z time
Defines the number of HCLK clock cycles during which the data bus is kept in Hi-Z after the
start of a PC CARD/NAND Flash write access to attribute memory space on socket x. Only
valid for write transaction:
0000 0000: 0 HCLK cycle
1111 1110: 255 HCLK cycles
1111 1111: Reserved.

RM0090 Rev 17

1657/1747
1680

Flexible memory controller (FMC)

RM0090

Bits 23:16 ATTHOLD[7:0]: Attribute memory x hold time
For PC Card/NAND Flash read accesses to attribute memory space on socket x, these bits
define the number of HCLK clock cycles (HCLK +2) clock cycles during which the address is
held after the command is deasserted (NWE, NOE).
For PC Card/NAND Flash write accesses to attribute memory space on socket x, these bits
define the number of HCLK clock cycles during which the data are held after the command
is deasserted (NWE, NOE).
0000 0000: Reserved
0000 0001: 1 HCLK cycle for write access, 3 HCLK cycles for read accesses
1111 1110: 254 HCLK cycle for write access, 256 HCLK cycles for read accesses
1111 1111: Reserved.
Bits 15:8 ATTWAIT[7:0]: Attribute memory x wait time
Defines the minimum number of HCLK (+1) clock cycles to assert the command (NWE,
NOE), for PC Card/NAND Flash read or write access to attribute memory space on socket x.
The duration for command assertion is extended if the wait signal (NWAIT) is active (low) at
the end of the programmed value of HCLK:
0000 0000: reserved
0000 0001: 2 HCLK cycles (+ wait cycle introduced by deassertion of NWAIT)
1111 1110: 255 HCLK cycles (+ wait cycle introduced by the card deasserting NWAIT)
1111 1111: Reserved
Bits 7:0 ATTSET[7:0]: Attribute memory x setup time
Defines the number of HCLK (+1) clock cycles to set up address before the command
assertion (NWE, NOE), for PC CARD/NAND Flash read or write access to attribute memory
space on socket x:
0000 0000: 1 HCLK cycle
1111 1110: 255 HCLK cycles
1111 1111: Reserved

I/O space timing register 4 (FMC_PIO4)
Address offset: 0xB0
Reset value: 0xFCFCFCFC
The FMC_PIO4 read/write registers contain the timing information used to access the I/O
space of the 16-bit PC Card/CompactFlash.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
IOHIZ[7:0]
rw

rw

rw

1658/1747

rw

rw

IOHOLD[7:0]
rw

rw

rw

rw

rw

rw

rw

rw

9

8

7

6

5

IOWAIT[7:0]

rw

rw

rw

rw

rw

rw

RM0090 Rev 17

rw

rw

rw

4

3

2

1

0

rw

rw

rw

IOSET[7:0]
rw

rw

rw

rw

rw

rw

rw

RM0090

Flexible memory controller (FMC)

Bits 31:24 IOHIZ[7:0]: I/O x data bus Hi-Z time
Defines the number of HCLK clock cycles during which the data bus is kept in Hi-Z after the
start of a PC Card write access to I/O space on socket x. Only valid for write transaction:
0000 0000: 0 HCLK cycle
1111 1111: 255 HCLK cycles
Bits 23:16 IOHOLD[7:0]: I/O x hold time
Defines the number of HCLK clock cycles during which the address is held (and data for write
access) after the command deassertion (NWE, NOE), for PC Card read or write access to I/O
space on socket x:
0000 0000: reserved
0000 0001: 1 HCLK cycle
1111 1111: 255 HCLK cycles
Bits 15:8 IOWAIT[7:0]: I/O x wait time
Defines the minimum number of HCLK (+1) clock cycles to assert the command (SMNWE,
SMNOE), for PC Card read or write access to I/O space on socket x. The duration for
command assertion is extended if the wait signal (NWAIT) is active (low) at the end of the
programmed value of HCLK:
0000 0000: reserved, do not use this value
0000 0001: 2 HCLK cycles (+ wait cycle introduced by deassertion of NWAIT)
1111 1111: 256 HCLK cycles (+ wait cycle introduced by the Card deasserting NWAIT)
Bits 7:0 IOSET[7:0]: I/O x setup time
Defines the number of HCLK (+1) clock cycles to set up the address before the command
assertion (NWE, NOE), for PC Card read or write access to I/O space on socket x:
0000 0000: 1 HCLK cycle
1111 1111: 256 HCLK cycles

RM0090 Rev 17

1659/1747
1680

Flexible memory controller (FMC)

RM0090

ECC result registers 2/3 (FMC_ECCR2/3)
Address offset: 0x54 + 0x20 * (x – 1), x = 2 or 3
Reset value: 0x0000 0000
These registers contain the current error correction code value computed by the ECC
computation modules of the FMC controller (one module per NAND Flash memory bank).
When the CPU reads the data from a NAND Flash memory page at the correct address
(refer to Section 37.6.6: Computation of the error correction code (ECC) in NAND Flash
memory), the data read/written from/to the NAND Flash memory are processed
automatically by the ECC computation module. When X byte have been read (according to
the ECCPS field in the FMC_PCRx registers), the CPU must read the computed ECC value
from the FMC_ECCx registers. It then verifies if these computed parity data are the same as
the parity value recorded in the spare area, to determine whether a page is valid, and, to
correct it otherwise. The FMC_ECCRx registers should be cleared after being read by
setting the ECCEN bit to ‘0’. To compute a new data block, the ECCEN bit must be set to ’1’.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

ECC[31:0]
r

Bits 31:0 ECC[31:0]: ECC result
This field contains the value computed by the ECC computation logic. Table 295 describes the
contents of these bit fields.

Table 295. ECC result relevant bits

1660/1747

ECCPS[2:0]

Page size in byte

ECC bits

000

256

ECC[21:0]

001

512

ECC[23:0]

010

1024

ECC[25:0]

011

2048

ECC[27:0]

100

4096

ECC[29:0]

101

8192

ECC[31:0]

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)

37.7

SDRAM controller

37.7.1

SDRAM controller main features
The main features of the SDRAM controller are the following:

37.7.2

•

Two SDRAM banks with independent configuration

•

8-bit, 16-bit, 32-bit data bus width

•

13-bits Address Row, 11-bits Address Column, 4 internal banks: 4x16Mx32bit
(256 MB), 4x16Mx16bit (128 MB), 4x16Mx8bit (64 MB)

•

Word, half-word, byte access

•

SDRAM clock can be HCLK/2 or HCLK/3

•

Automatic row and bank boundary management

•

Multibank ping-pong access

•

Programmable timing parameters

•

Automatic Refresh operation with programmable Refresh rate

•

Self-refresh mode

•

Power-down mode

•

SDRAM power-up initialization by software

•

CAS latency of 1,2,3

•

Cacheable Read FIFO with depth of 6 lines x32-bit (6 x14-bit address tag)

SDRAM External memory interface signals
At startup, the SDRAM I/O pins used to interface the FMC SDRAM controller with the
external SDRAM devices must configured by the user application. The SDRAM controller
I/O pins which are not used by the application, can be used for other purposes.
Table 296. SDRAM signals
SDRAM signal

I/O
type

Description

SDCLK

O

SDRAM clock

SDCKE[1:0]

O

SDCKE0: SDRAM Bank 1 Clock Enable
SDCKE1: SDRAM Bank 2 Clock Enable

SDNE[1:0]

O

SDNE0: SDRAM Bank 1 Chip Enable
SDNE1: SDRAM Bank 2 Chip Enable

A[12:0]

O

Address

FMC_A[12:0]

D[31:0]

I/O

Bidirectional data bus

FMC_D[31:0]

BA[1:0]

O

Bank Address

FMC_A[15:14]

NRAS

O

Row Address Strobe

NCAS

O

Column Address Strobe

SDNWE

O

Write Enable

NBL[3:0]

O

Output Byte Mask for write accesses
(memory signal name: DQM[3:0])

RM0090 Rev 17

Alternate function

FMC_NBL[3:0]

1661/1747
1680

Flexible memory controller (FMC)

37.7.3

RM0090

SDRAM controller functional description
All SDRAM controller outputs (signals, address and data) change on the falling edge of the
memory clock (FMC_SDCLK).

SDRAM initialization
The initialization sequence is managed by software. If the two banks are used, the
initialization sequence must be generated simultaneously to Bank 1and Bank 2 by setting
the Target Bank bits CTB1 and CTB2 in the FMC_SDCMR register:
1.

Program the memory device features into the FMC_SDCRx register.The SDRAM clock
frequency, RBURST and RPIPE must be programmed in the FMC_SDCR1 register.

2.

Program the memory device timing into the FMC_SDTRx register. The TRP and TRC
timings must be programmed in the FMC_SDTR1 register.

3.

Set MODE bits to ‘001’ and configure the Target Bank bits (CTB1 and/or CTB2) in the
FMC_SDCMR register to start delivering the clock to the memory (SDCKE is driven
high).

4.

Wait during the prescribed delay period. Typical delay is around 100 μs (refer to the
SDRAM datasheet for the required delay after power-up).

5.

Set MODE bits to ‘010’ and configure the Target Bank bits (CTB1 and/or CTB2) in the
FMC_SDCMR register to issue a “Precharge All” command.

6.

Set MODE bits to ‘011’, and configure the Target Bank bits (CTB1 and/or CTB2) as well
as the number of consecutive Auto-refresh commands (NRFS) in the FMC_SDCMR
register. Refer to the SDRAM datasheet for the number of Auto-refresh commands that
should be issued. Typical number is 8.
Configure the MRD field according to your SDRAM device, set the MODE bits to '100',
and configure the Target Bank bits (CTB1 and/or CTB2) in the FMC_SDCMR register
to issue a "Load Mode Register" command in order to program the SDRAM. In
particular:

7.

8.

a)

The CAS latency must be selected following configured value in FMC_SDCR1/2
registers

b)

The Burst Length (BL) of 1 must be selected by configuring the M[2:0] bits to 000
in the mode register (refer to the SDRAM datasheet). If the Mode Register is not
the same for both SDRAM banks, this step has to be repeated twice, once for
each bank, and the Target Bank bits set accordingly.

Program the refresh rate in the FMC_SDRTR register
The refresh rate corresponds to the delay between refresh cycles. Its value must be
adapted to SDRAM devices.

9.

For mobile SDRAM devices, to program the extended mode register it should be done
once the SDRAM device is initialized: First, a dummy read access should be performed
while BA1=1 and BA=0 (refer to SDRAM address mapping section for BA[1:0] address
mapping) in order to select the extended mode register instead of Load mode register
and then program the needed value.

At this stage the SDRAM device is ready to accept commands. If a system reset occurs
during an ongoing SDRAM access, the data bus might still be driven by the SDRAM device.
Therefor the SDRAM device must be first reinitialized after reset before issuing any new
access by the NOR Flash/PSRAM/SRAM or NAND Flash/PC Card controller.
Note:

1662/1747

If two SDRAM devices are connected to the FMC, all the accesses performed at the same
time to both devices by the Command Mode register (Load Mode Register and Self-refresh

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
commands) are issued using the timing parameters configured for SDRAM Bank 1 (TMRD,
TRAS and TXSR timings) in the FMC_SDTR1 register.

SDRAM controller write cycle
The SDRAM controller accepts single and burst write requests and translates them into
single memory accesses. In both cases, the SDRAM controller keeps track of the active row
for each bank to be able to perform consecutive write accesses to different banks (Multibank
ping-pong access).
Before performing any write access, the SDRAM bank write protection must be disabled by
clearing the WP bit in the FMC_SDCRx register.
Figure 478. Burst write SDRAM access waveforms
75&' 

6'1(

6'&/.
5RZQ &ROD &ROE &ROF &ROG &ROH &ROI &RJ &ROK &ROL &ROM &RON &ROO

$>@

15$6

1&$6
6'1:(

'QD

'$7$>@

'QE 'QF

'QG 'QH 'QI

'QJ 'QK 'QL

'QM

'QN 'QO
069

The SDRAM controller always checks the next access.
•

If the next access is in the same row or in another active row, the write operation is
carried out,

•

if the next access targets another row (not active), the SDRAM controller generates a
precharge command, activates the new row and initiates a write command.

RM0090 Rev 17

1663/1747
1680

Flexible memory controller (FMC)

RM0090

SDRAM controller read cycle
The SDRAM controller accepts single and burst read requests and translates them into
single memory accesses. In both cases, the SDRAM controller keeps track of the active row
in each bank to be able to perform consecutive read accesses in different banks (Multibank
ping-pong access).
Figure 479. Burst read SDRAM access
75&' 

&$6ODWHQF\ 

6'1(

6'&/.
$>@

5RZQ

&ROD &ROE

&ROF

&ROG &ROH &ROI

15$6

1&$6

1:(

'QD

'$7$>@

'QE

'QF

'QG

'QH

'QI
069

The FMC SDRAM controller features a Cacheable read FIFO (6 lines x 32 bits). It is used to
store data read in advance during the CAS latency period and during the RPIPE delay. The
following the formula is applied:

Number of anticipated data = CAS latency + 1 + ( RPIPE delay ) ⁄ 2
The RBURST bit must be set in the FMC_SDCR1 register to anticipate the next read
access.
Example:
•
CAS latency = 3, RPIPE delay = 0: 4 data (not committed) are stored in the FIFO.
•
CAS latency = 3, RPIPE delay = 2: 5 data (not committed) are stored in the FIFO.
The read FIFO features a 14-bit address tag to each line to identify its content: 11 bits for the
column address, 2 bits to select the internal bank and the active row, and 1 bit to select the
SDRAM device
When the end of the row is reached in advance during an AHB burst read, the data read in
advance (not committed) are not stored in the read FIFO. For single read access, data are
correctly stored in the FIFO.

1664/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
Each time a read request occurs, the SDRAM controller checks:
•

If the address matches one of the address tags, data are directly read from the FIFO
and the corresponding address tag/ line content is cleared and the remaining data in
the FIFO are compacted to avoid empty lines.

•

Otherwise, a new read command is issued to the memory and the FIFO is updated with
new data. If the FIFO is full, the older data are lost.
Figure 480. Logic diagram of Read access with RBURST bit set (CAS=2, RPIPE=0)
ST 2EAD ACCESS 2EQUESTED DATA IS NOT IN THE &)&/
&-# 3$2!- #ONTROLLER

!(" -ASTER

READ REQUEST
$ATA 

X

3$2!$EVICE
#!3  

 LINES &)&/

X $ATA 
X $ATA 


!DD 4AG READ &)&/

$ATA STORED IN &)&/
IN ADVANCE DURING
THE #!3 LATENCY PERIOD

ND 2EAD ACCESS  2EQUESTED DATA WAS PREVIOUSLY STORED IN THE &)&/
!DDRESS MATCHES WITH
ONE OF THE ADDRESS TAGS

!(" -ASTER

&-# 3$2!- #ONTROLLER
READ REQUEST
$ATA 

X

3$2!$EVICE
#!3  

 LINES &)&/

X $ATA 
X $ATA 


!DD 4AG READ &)&/

$ATA READ FROM &)&/

-36

During a write access or a Precharge command, the read FIFO is flushed and ready to be
filled with new data.
After the first read request, if the current access was not performed to a row boundary, the
SDRAM controller anticipates the next read access during the CAS latency period and the
RPIPE delay (if configured). This is done by incrementing the memory address. The
following condition must be met:
•

RBURST control bit should be set to ‘1’ in the FMC_SDCR1 register.

RM0090 Rev 17

1665/1747
1680

Flexible memory controller (FMC)

RM0090

The address management depends on the next AHB request:
•

Next AHB request is sequential (AHB Burst)
In this case, the SDRAM controller increments the address.

•

Next AHB request is not sequential
–

If the new read request targets the same row or another active row, the new
address is passed to the memory and the master is stalled for the CAS latency
period, waiting for the new data from memory.

–

If the new read request does not target an active row, the SDRAM controller
generates a Precharge command, activates the new row, and initiates a read
command.

If the RURST is reset, the read FIFO is not used.

Row and bank boundary management
When a read or write access crosses a row boundary, if the next read or write access is
sequential and the current access was performed to a row boundary, the SDRAM controller
executes the following operations:
1.

Precharge of the active row,

2.

Activation of the new row

3.

Start of a read/write command.

At a row boundary, the automatic activation of the next row is supported for all columns and
data bus width configurations.
If necessary, the SDRAM controller inserts additional clock cycles between the following
commands:
•

Between Precharge and Active commands to match TRP parameter (only if the next
access is in a different row in the same bank),

•

Between Active and Read commands to match the TRCD parameter.

These parameters are defined into the FMC_SDTRx register.
Refer to Figure 481 and Figure 482 for read and burst write access crossing a row
boundary.

1666/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
Figure 481. Read access crossing row boundary
42#$   #!3 LATENCY  

420  

3$.%

3$#,+
!;=

2OW N
#OL A

2OW N 

#OL A #OL B

.2!3

.#!3

.7%

$ATA;=

$NA

$N  A
0RECHARGE

!CTIVATE 2OW

2EAD #OMMAND
-36

Figure 482. Write access crossing row boundary
420  

42#$  

3$.%

3$#,+
!;=

#NA

#OLB

2OW N 

#OLA

#OLB

.2!3

.#!3

.7%

$ATA;=

$NA

$N A

$NB
0RECHARGE

!CTIVATE 2OW

$N B

7RITE COMMAND
-36

RM0090 Rev 17

1667/1747
1680

Flexible memory controller (FMC)

RM0090

If the next access is sequential and the current access crosses a bank boundary, the
SDRAM controller activates the first row in the next bank and initiates a new read/write
command. Two cases are possible:

Note:

•

If the current bank is not the last one, the active row in the new bank must be
precharged. At a bank boundary, the automatic activation of the next row is supported
for all rows/columns and data bus width configuration.

•

For 13-bit row address, 11-bit column address, 4 internal banks and bus width 32-bit
SDRAM memories, if the current bank is the last one and the selected SDRAM device
is connected to Bank 1, the SDRAM controller continues to read/write from the second
SDRAM device (assuming it has been initialized):
a)

The SDRAM controller activates the first row (after precharging the active row, if
there is already an active row in the first internal bank, and initiates a new
read/write command.

b)

If the first row is already activated, the SDRAM controller just initiates a read/write
command.

At bank boundary, if the current bank is the last one, the automatic activation of the next row
is supported only when addressing 13-bit rows, 11-bit columns, 4 internal banks and 32-bit
data bus SDRAM devices. Otherwise, the SDRAM address range is violated and an AHB
error is generated.

SDRAM controller refresh cycle
The Auto-refresh command is used to refresh the SDRAM device content. The SDRAM
controller periodically issues auto-refresh commands. An internal counter is loaded with the
COUNT value in the register FMC_SDRTR. This value defines the number of memory clock
cycles between the refresh cycles (refresh rate). When this counter reaches zero, an
internal pulse is generated.
If a memory access is ongoing, the auto-refresh request is delayed. However, if the memory
access and the auto-refresh requests are generated simultaneously, the auto-refresh
request takes precedence.
If the memory access occurs during an auto-refresh operation, the request is buffered and
processed when the auto-refresh is complete.
If a new auto-refresh request occurs while the previous one was not served, the RE
(Refresh Error) bit is set in the Status register. An Interrupt is generated if it has been
enabled (REIE = ‘1’).
If SDRAM lines are not in idle state (not all row are closed), the SDRAM controller generates
a PALL (Precharge ALL) command before the auto-refresh.
If the Auto-refresh command is generated by the FMC_SDCMR Command Mode register
(Mode bits = ‘011’), a PALL command (Mode bits =’ 010’) must be issued first.

1668/1747

RM0090 Rev 17

RM0090

37.7.4

Flexible memory controller (FMC)

Low power modes
Two low power modes are available:
•

Self-refresh mode
The auto-refresh cycles are performed by the SDRAM device itself to retain data
without external clocking.

•

Power-down mode
The auto-refresh cycles are performed by the SDRAM controller.

Self-refresh mode
This mode is selected by setting the MODE bits to ‘101’ and by configuring the Target Bank
bits (CTB1 and/or CTB2) in the FMC_SDCMR register.
The SDRAM clock stops running after a TRAS delay and the internal refresh timer stops
counting only if one of the following conditions is met:
•

A Self-refresh command is issued to both devices

•

One of the devices is not activated (SDRAM bank is not initialized).

Before entering Self-Refresh mode, the SDRAM controller automatically issues a PALL
command.
If the Write data FIFO is not empty, all data are sent to the memory before activating the
Self-refresh mode and the BUSY status flag remains set.
In Self-refresh mode, all SDRAM device inputs become don’t care except for SDCKE which
remains low.
The SDRAM device must remain in Self-refresh mode for a minimum period of time of
TRAS and can remain in Self-refresh mode for an indefinite period beyond that. To
guarantee this minimum period, the BUSY status flag remains high after the Self-refresh
activation during a TRAS delay.
As soon as an SDRAM device is selected, the SDRAM controller generates a sequence of
commands to exit from Self-refresh mode. After the memory access, the selected device
remains in Normal mode.
To exit from Self-refresh, the MODE bits must be set to ‘000’ (Normal mode) and the Target
Bank bits (CTB1 and/or CTB2) must be configured in the FMC_SDCMR register.

RM0090 Rev 17

1669/1747
1680

Flexible memory controller (FMC)

RM0090
Figure 483. Self-refresh mode

4

4

4

4N 

4 

4 

3$#,+

T2!3MIN
3$#+%

#/--!.$

02%#(!2'%

!54/
2%&2%3(

./0

./0

OR #/--!.$
).(%2)4

!54/
2%&2%3(

$/-
$/-,$/-5

! !
! !

!,,
"!.+3

!

$ATA;= (I :
T832

T20
0RECHARGE ALL
ACTIVE BANKS

%XIT 3ELF REFRESH MODE
RESTART REFRESH TIMEBASE
#,+ STABLE PRIOR TO EXISTING
3ELF REFRESH MODE

%NTER 3ELF REFRESH MODE

-36

Power-down mode
This mode is selected by setting the MODE bits to ‘110’ and by configuring the Target Bank
bits (CTB1 and/or CTB2) in the FMC_SDCMR register.

1670/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
Figure 484. Power-down mode

3$#,+

3$#+%

#/--!.$

./0

./0

!#4)6%

T2#$
!LL BANKS IDLE

)NPUT BUFFERS GATED OFF
T2!3

%NTER 0OWER DOWN

%XIT 0OWER DOWN

T2#
-36

If the Write data FIFO is not empty, all data are sent to the memory before activating the
Power-down mode.
As soon as an SDRAM device is selected, the SDRAM controller exits from the Power-down
mode. After the memory access, the selected SDRAM device remains in Normal mode.
During Power-down mode, all SDRAM device input and output buffers are deactivated
except for the SDCKE which remains low.
The SDRAM device cannot remain in Power-down mode longer than the refresh period and
cannot perform the Auto-refresh cycles by itself. Therefore, the SDRAM controller carries
out the refresh operation by executing the operations below:
1.

Exit from Power-down mode and drive the SDCKE high

2.

Generate the PALL command only if a row was active during Power-down mode

3.

Generate the auto-refresh command

4.

Drive SDCKE low again to return to Power-down mode.

To exit from Power-down mode, the MODE bits must be set to ‘000’ (Normal mode) and the
Target Bank bits (CTB1 and/or CTB2) must be configured in the FMC_SDCMR register.

RM0090 Rev 17

1671/1747
1680

Flexible memory controller (FMC)

37.7.5

RM0090

SDRAM controller registers
SDRAM Control registers 1,2 (FMC_SDCR1,2)
Address offset: 0x140+ 4* (x – 1), x = 1,2
Reset value: 0x0000 02D0
This register contains the control parameters for each SDRAM memory bank

rw

rw

rw

rw

rw

rw

rw

rw

rw

4

3

rw

rw

rw

2

1

rw

rw

Bits 31:15 Reserved, must be kept at reset value
Bits 14:13 RPIPE[1:0]: Read pipe
These bits define the delay, in HCLK clock cycles, for reading data after CAS latency.
00: No HCLK clock cycle delay
01: One HCLK clock cycle delay
10: Two HCLK clock cycle delay
11: reserved, do not use
Note: The corresponding bits in the FMC_SDCR2 register are read only.
Bit 12 RBURST: Burst read
This bit enables burst read mode. The SDRAM controller anticipates the next read commands
during the CAS latency and stores data in the Read FIFO.
0: single read requests are not managed as bursts
1: single read requests are always managed as bursts
Note: The corresponding bit in the FMC_SDCR2 register is don’t care.
Bits 11:10 SDCLK[1:0]: SDRAM clock configuration
These bits define the SDRAM clock period for both SDRAM banks and allow disabling the clock
before changing the frequency. In this case the SDRAM must be re-initialized.
00: SDCLK clock disabled
01: Reserved
10: SDCLK period = 2 x HCLK periods
11: SDCLK period = 3 x HCLK periods
Note: The corresponding bits in the FMC_SDCR2 register are read only.
Bit 9 WP: Write protection
This bit enables write mode access to the SDRAM bank.
0: Write accesses allowed
1: Write accesses ignored
Bits 8:7 CAS[1:0]: CAS Latency
This bits sets the SDRAM CAS latency in number of memory clock cycles
00: reserved, do not use.
01: 1 cycle
10: 2 cycles
11: 3 cycles

1672/1747

RM0090 Rev 17

0
NC[1:0]

5

NR[1:0]

6

MWID[1:0]

7

NB

8
CAS[1:0]

9

WP

SDCLK[1:0]

Reserved

RBURST.

RPIPE[1:0]

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

rw

RM0090

Flexible memory controller (FMC)

Bit 6 NB: Number of internal banks
This bit sets the number of internal banks.
0: Two internal Banks
1: Four internal Banks
Bits 5:4 MWID[1:0]: Memory data bus width.
These bits define the memory device width.
00: 8 bits
01: 16 bits
10: 32 bits
11: reserved, do not use.
Bits 3:2 NR[1:0]: Number of row address bits
These bits define the number of bits of a row address.
00: 11 bit
01: 12 bits
10: 13 bits
11: reserved, do not use.
Bits 1:0 NC[1:0]: Number of column address bits
These bits define the number of bits of a column address.
00: 8 bits
01: 9 bits
10: 10 bits
11: 11 bits.

Note:

Before modifying the RBURST or RPIPE settings or disabling the SDCLK clock, the user
must first send a PALL command to make sure ongoing operations are complete.

SDRAM Timing registers 1,2 (FMC_SDTR1,2)
Address offset: 0x148 + 4 * (x – 1), x = 1,2
Reset value: 0x0FFF FFFF
This register contains the timing parameters of each SDRAM bank
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10
Reserved
rw

rw

rw

TRCD[3:0]
rw

rw

rw

rw

TRP[3:0
rw

rw

rw

rw

TWR[3:0
rw

rw

rw

rw

TRC[3:0
rw

rw

rw

rw

9

8

7

TRAS[3:0
rw

rw

rw

rw

6

5

4

3

TXSR[3:0
rw

rw

rw

rw

2

1

0

TMRD[3:0
rw

rw

rw

rw

rw

Bits 31:28 Reserved, must be kept at reset value
Bits 27:24 TRCD[3:0]: Row to column delay
These bits define the delay between the Activate command and a Read/Write command in number of
memory clock cycles.
0000: 1 cycle.
0001: 2 cycles
....
1111: 16 cycles

RM0090 Rev 17

1673/1747
1680

Flexible memory controller (FMC)

RM0090

Bits 23:20 TRP[3:0]: Row precharge delay
These bits define the delay between a Precharge command and another command in number of
memory clock cycles. The TRP timing is only configured in the FMC_SDTR1 register. If two SDRAM
devices are used, the TRP must be programmed with the timing of the slowest device.
0000: 1 cycle
0001: 2 cycles
....
1111: 16 cycles
Note: The corresponding bits in the FMC_SDTR2 register are don’t care.
Bits 19:16 TWR[3:0]: Recovery delay
These bits define the delay between a Write and a Precharge command in number of memory clock
cycles.
0000: 1 cycle
0001: 2 cycles
....
1111: 16 cycles
Note: TWR must be programmed to match the write recovery time (tWR) defined in the SDRAM
datasheet, and to guarantee that:
TWR ≥ TRAS - TRCD and TWR ≥TRC - TRCD - TRP
Example: TRAS= 4 cycles, TRCD= 2 cycles. So, TWR >= 2 cycles. TWR must be
programmed to 0x1.
If two SDRAM devices are used, the FMC_SDTR1 and FMC_SDTR2 must be programmed
with the same TWR timing corresponding to the slowest SDRAM device.
Bits 15:12 TRC[3:0]: Row cycle delay
These bits define the delay between the Refresh command and the Activate command, as well as d
the delay between two consecutive Refresh commands. It is expressed in number of memory clock
cycles. The TRC timing is only configured in the FMC_SDTR1 register. If two SDRAM devices are
used, the TRC must be programmed with the timings of the slowest device.
0000: 1 cycle
0001: 2 cycles
....
1111: 16 cycles
Note: TRC must match the TRC and TRFC (Auto Refresh period) timings defined in the SDRAM
device datasheet.
Note: The corresponding bits in the FMC_SDTR2 register are don’t care.
Bits 11:8 TRAS[3:0]: Self refresh time
These bits define the minimum Self-refresh period in number of memory clock cycles.
0000: 1 cycle
0001: 2 cycles
....
1111: 16 cycles
Bits 7:4 TXSR[3:0]: Exit Self-refresh delay
These bits define the delay from releasing the Self-refresh command to issuing the Activate
command in number of memory clock cycles.
0000: 1 cycle
0001: 2 cycles
....
1111: 16 cycles

1674/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)

Bits 3:0 TMRD[3:0]: Load Mode Register to Active
These bits define the delay between a Load Mode Register command and an Active or Refresh
command in number of memory clock cycles.
0000: 1 cycle
0001: 2 cycles
....
1111: 16 cycles

Note:

If two SDRAM devices are connected, all the accesses performed simultaneously to both
devices by the Command Mode register (Load Mode Register and Self-refresh commands)
are issued using the timing parameters configured for Bank 1 (TMRD, TRAS and TXSR
timings) in the FMC_SDTR1 register.
The TRP and TRC timings are only configured in the FMC_SDTR1 register. If two SDRAM
devices are used, the TRP and TRC timings must be programmed with the timings of the
slowest device.

SDRAM Command Mode register (FMC_SDCMR)
Address offset: 0x150
Reset value: 0x0000 0000

7

rw

rw

rw

rw

rw

rw

rw

rw

rw

rw

5

rw

rw

NRFS[3:0]

Reserved

6

rw

rw

rw

rw

rw

4

3

w

w

2

1

0

MODE[2:0]

8

CTB2

9

MRD[11:0]

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

CTB1

This register contains the command issued when the SDRAM device is accessed. This
register is used to initialize the SDRAM device, and to activate the Self-refresh and the
Power-down modes. As soon as the MODE field is written, the command will be issued only
to one or to both SDRAM banks according to CTB1 and CTB2 command bits. This register
is the same for both SDRAM banks.

w

w

w

Bits 31:22 Reserved, must be kept at reset value
Bits 21:9 MRD[11:0]: Mode Register definition
This 13-bit field defines the SDRAM Mode Register content. The Mode Register is programmed using
the Load Mode Register command.
Bits 8:5 NRFS[3:0]: Number of Auto-refresh
These bits define the number of consecutive Auto-refresh commands issued when MODE = ‘011’.
0000: 1 Auto-refresh cycle
0001: 2 Auto-refresh cycles
....
1110: 15 Auto-refresh cycles
1111: Reserved
Bit 4 CTB1: Command Target Bank 1
This bit indicates whether the command will be issued to SDRAM Bank 1 or not.
0: Command not issued to SDRAM Bank 1
1: Command issued to SDRAM Bank 1

RM0090 Rev 17

1675/1747
1680

Flexible memory controller (FMC)

RM0090

Bit 3 CTB2: Command Target Bank 2
This bit indicates whether the command will be issued to SDRAM Bank 2 or not.
0: Command not issued to SDRAM Bank 2
1: Command issued to SDRAM Bank 2
Bits 2:0 MODE[2:0]: Command mode
These bits define the command issued to the SDRAM device.
000: Normal Mode
001: Clock Configuration Enable
010: PALL (“All Bank Precharge”) command
011: Auto-refresh command
100: Load Mode Register
101: Self-refresh command
110: Power-down command
111: Reserved
Note: When a command is issued, at least one Command Target Bank bit ( CBT1 or CBT2) must be
set. If both banks are used, the commands must be issued to the two banks at the same time
by setting the CBT1 and CBT2 bits.

SDRAM Refresh Timer register (FMC_SDRTR)
Address offset:0x154
Reset value: 0x0000 0000
This register sets the refresh rate in number of SDCLK clock cycles between the refresh
cycles by configuring the Refresh Timer Count value.
Refresh rate = ( SDRAM refresh rate × SDRAM clock frequency ) – 20
SDRAM refresh rate = SDRAM refresh period ⁄ Number of rows

Example
SDRAM refresh rate = 64 ms ⁄ ( 8196rows ) = 7.81μs
where 64 ms is the SDRAM refresh period.
7.81μs × 60MHz = 468.6
The refresh rate must be increased by 20 SDRAM clock cycles (as in the above example) to
obtain a safe margin if an internal refresh request occurs when a read request has been
accepted. It corresponds to a COUNT value of ‘0000111000000’ (448).
This 13-bit field is loaded into a timer which is decremented using the SDRAM clock. This
timer generates a refresh pulse when zero is reached. The COUNT value must be set at
least to 41 SDRAM clock cycles.
As soon as the FMC_SDRTR register is programmed, the timer starts counting. If the value
programmed in the register is ’0’, no refresh is carried out. This register must not be
reprogrammed after the initialization procedure to avoid modifying the refresh rate.
Each time a refresh pulse is generated, this 13-bit COUNT field is reloaded into the counter.

1676/1747

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
If a memory access is in progress, the Auto-refresh request is delayed. However, if the
memory access and Auto-refresh requests are generated simultaneously, the Auto-refresh
takes precedence. If the memory access occurs during a refresh operation, the request is
buffered to be processed when the refresh is complete.
This register common to SDRAM bank 1 and bank 2.
.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15
Reserved

14

13 12 11 10

9

8

7

6

5

4

3

2

1

0

REIE

COUNT[12:0]

CRE

rw

rw rw rw rw rw rw rw rw rw rw rw rw rw

w

Bits 31: 15 Reserved, must be kept at reset value
Bit 14 REIE: RES Interrupt Enable
0: Interrupt is disabled
1: An Interrupt is generated if RE = 1
Bits 13:1 COUNT[12:0]: Refresh Timer Count
This 13-bit field defines the refresh rate of the SDRAM device. It is expressed in number of memory
clock cycles. It must be set at least to 41 SDRAM clock cycles (0x29).
COUNT = (SDRAM refresh rate x SDRAM clock frequency) - 20
SDRAM refresh rate = SDRAM refresh period / Number of rows
Bit 0 CRE: Clear Refresh error flag
This bit is used to clear the Refresh Error Flag (RE) in the Status Register.
0: no effect
1: Refresh Error flag is cleared

Note:

The programmed COUNT value must not be equal to the sum of the following timings:
TWR+TRP+TRC+TRCD+4 memory clock cycles .
SDRAM Status register (FMC_SDSR)
Address offset: 0x158
Reset value: 0x0000 0000
6

5

4

r

3

2

r

r

r

1

0

RE

7

MODES1[1:0]

8

MODES2[1:0]

Reserved

9

BUSY

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

r

r

Bits 31:5 Reserved, must be kept at reset value
Bit 5 BUSY: Busy status
This bit defines the status of the SDRAM controller after a Command Mode request
0: SDRAM Controller is ready to accept a new request
1; SDRAM Controller is not ready to accept a new request
Bits 4:3 MODES2[1:0]: Status Mode for Bank 2
This bit defines the Status Mode of SDRAM Bank 2.
00: Normal Mode
01: Self-refresh mode
10: Power-down mode

RM0090 Rev 17

1677/1747
1680

Flexible memory controller (FMC)

RM0090

Bits 2:1 MODES1[1:0]: Status Mode for Bank 1
This bit defines the Status Mode of SDRAM Bank 1.
00: Normal Mode
01: Self-refresh mode
10: Power-down mode
Bit 0 RE: Refresh error flag
0: No refresh error has been detected
1: A refresh error has been detected
An interrupt is generated if REIE = 1 and RE = 1

37.8

FMC register map
The following table summarizes the FMC registers.
Table 297. FMC register map
Register

0x00

FMC_BCR1

0x08

FMC_BCR2

Reserved

0x10

FMC_BCR3

Reserved

0x18

FMC_BCR4

Reserved

0x04

FMC_BTR1

Res.

0x0C

FMC_BTR2

Res.

0x14

FMC_BTR3

Res.

0x1C

FMC_BTR4

Res.

0x104

FMC_BWTR1

Res.

1678/1747

ACCMOD[1:0 ACCMOD[1:0 ACCMOD[1:0 ACCMOD[1:0 ACCMOD[1:0

MBKEN

MUXEN

MBKEN
MBKEN

MBKEN

MUXEN
MUXEN

MUXEN

MTYP[1:0]
MTYP[1:0]
MTYP[1:0]

MTYP[1:0]

MWID[1:0]
MWID[1:0]
MWID[1:0]

MWID[1:0]

FACCEN

Reserved

FACCEN
FACCEN

FACCEN

Reserved
Reserved

Reserved

BURSTEN

WAITPOL
WAITPOL

BURSTEN

WAITPOL

BURSTEN

WAITPOL

BURSTEN

WRAPMOD WRAPMOD WRAPMOD WRAPMOD

WREN

WAITCFG
WAITCFG
WAITCFG

WAITCFG

WAITEN

WREN

WREN

EXTMOD

WREN

WAITEN
WAITEN
WAITEN

EXTMOD
EXTMOD
EXTMOD

ASYNCWAIT ASYNCWAIT ASYNCWAIT ASYNCWAIT

CPSIZE[2:0] CPSIZE[2:0] CPSIZE[2:0] CPSIZE[2:0]

CCLKEN

Reserved

CBURSTRW CBURSTRW CBURSTRW CBURSTRW

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Offset

DATLAT[3:0]

CLKDIV[3:0] BUSTURN[3:0]

DATAST[7:0]

ADDHLD[3:0] ADDSET[3:0]

DATLAT[3:0]

CLKDIV[3:0] BUSTURN[3:0]

DATAST[7:0]

ADDHLD[3:0] ADDSET[3:0]

DATLAT[3:0]

CLKDIV[3:0] BUSTURN[3:0]

DATAST[7:0]

ADDHLD[3:0] ADDSET[3:0]

DATLAT[3:0]

CLKDIV[3:0] BUSTURN[3:0]

DATAST[7:0]

ADDHLD[3:0] ADDSET[3:0]

DATAST[7:0]

ADDHLD[3:0] ADDSET[3:0]

Res.

BUSTURN[3:0]

RM0090 Rev 17

RM0090

Flexible memory controller (FMC)
Table 297. FMC register map (continued)
Register

0x10C

FMC_BWTR2

Res.

0x104

FMC_BWTR3

Res.

0x10C

FMC_BWTR4

Res.

0x60

FMC_PCR2

Reserved

0x80

FMC_PCR3

Reserved

0xA0

FMC_PCR4

Reserved

0x64

FMC_SR2

Reserved

0x84

FMC_SR3

Reserved

0xA4

FMC_SR4

Reserved

0x68

FMC_PMEM2

ADDHLD[3:0] ADDSET[3:0]

Res.

BUSTURN[3:0]

DATAST[7:0]

ADDHLD[3:0] ADDSET[3:0]

MEMHIZ[7:0]

ILS

Reserved
Reserved
IRS

ILS

IRS

ILS

IRS

PBKEN

PTYP
PTYP

PBKEN
IFS
IFS
IFS

IREN
IREN
IREN

ILEN
ILEN
ILEN

IFEN

ECCEN

MEMWAIT[7:0]

IFEN

Res.

IFEN

Res.

FEMPT FEMPT FEMPT ECCEN

TAR[3:0]
TAR[3:0]
TAR[3:0]

MEMHOLD[7:0]

Res.

Reserved

DATAST[7:0]

PWAITEN PWAITEN PWAITEN

BUSTURN[3:0]

PTYP

Res.

PBKEN

ADDHLD[3:0] ADDSET[3:0]

PWID[1:0] PWID[1:0] PWID[1:0]

DATAST[7:0]

ECCEN

BUSTURN[3:0]

TCLR[3:0] TCLR[3:0] TCLR[3:0]

Res.

ECCPS[2:0] ECCPS[2:0] ECCPS[2:0]

ACCMOD[1:0 ACCMOD[1:0 ACCMOD[1:0

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Offset

MEMSET[7:0]

ATTSET[7:0]

ATTWAIT[7:0]

ATTSET[7:0]

0xAC

FMC_PATT4

ATTHIZ[7:0]

ATTHOLD[7:0]

ATTWAIT[7:0]

ATTSET[7:0]

0xB0

FMC_PIO4

IOHIZ[7:0]

IOHOLD[7:0]

IOWAIT[7:0]

IOSET[7:0]

0x74

FMC_ECCR2

ECC[31:0]

0x94

FMC_ECCR3

ECC[31:0]

0x140

FMC_SDCR_1

0x144

FMC_SDCR_2

0x148

FMC_SDTR1

Reserved

TRCD[3:0]

TRP[3:0]

TWR[3:0]

TRC[3:0]

TRAS[3:0]

TXSR[3:0]

TMRD[3:0]

0x14C

FMC_SDTR2

Reserved

TRCD[3:0]

TRP[3:0]

TWR[3:0]

TRC[3:0]

TRAS[3:0]

TXSR[3:0]

TMRD[3:0]

0x150

FMC_SDCMR

RM0090 Rev 17

NC[1:0]
MODE[2:0]

NB

NR[1:0]
CTB2

Reserved

CTB1

MRD[12:0]

Reserved

NRFS[3:0]

Reserved

NC[1:0]

ATTWAIT[7:0]

ATTHOLD[7:0]

NR[1:0]

ATTHOLD[7:0]

ATTHIZ[7:0]

MWID[1:0] MWID[1:0]

ATTHIZ[7:0]

FMC_PATT3

NB

FMC_PATT2

0x8C

CAS[1:0] CAS[1:0]

0x6C

WP

MEMSET[7:0]
MEMSET[7:0]

WP

MEMWAIT[7:0]
MEMWAIT[7:0]

CLK[1:0]

MEMHOLD[7:0]
MEMHOLD[7:0]

CLK[1:0]

MEMHIZ[7:0]
MEMHIZ[7:0]

RBURST

FMC_PMEM3
FMC_PMEM4

RPIPE[1:0]

0x88
0xA8

1679/1747
1680

Flexible memory controller (FMC)

RM0090

1680/1747

Reserved

RM0090 Rev 17

CRE
RE

FMC_SDSR

MODES1[1:0]

0x158

Reserved

MODES2[1:0]

FMC_SDRTR

BUSY

0x154

COUNT[12:0]

Register

REIE

Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 297. FMC register map (continued)

RM0090

38

Debug support (DBG)

Debug support (DBG)
This section applies to the whole STM32F4xx family, unless otherwise specified.

38.1

Overview
The STM32F4xx are built around a Cortex®-M4 with FPU core which contains hardware
extensions for advanced debugging features. The debug extensions allow the core to be
stopped either on a given instruction fetch (breakpoint) or data access (watchpoint). When
stopped, the core’s internal state and the system’s external state may be examined. Once
examination is complete, the core and the system may be restored and program execution
resumed.
The debug features are used by the debugger host when connecting to and debugging the
STM32F4xx MCUs.
Two interfaces for debug are available:
•

Serial wire

•

JTAG debug port
Figure 485. Block diagram of STM32 MCU and Cortex®-M4 with FPU-level debug
support
34-&XX DEBUG SUPPORT
#ORTEX - DEBUG SUPPORT

*4-3
37$)/
*4$)

.*4234

$#ODE
INTERFACE

3YSTEM
INTERFACE

%XTERNAL PRIVATE
PERIPHERAL BUS 00"
"RIDGE

42!#%37/
4RACE PORT

40)5

!(" !0

)NTERNAL PRIVATE
PERIPHERAL BUS 00"

42!#%#+
42!#%$;=

.6)#

$74

$"'-#5

&0"

ES

*4#+
37#,+

37* $0

TR
IC

*4$/
42!#%37/

$ATA

TE
D

#ORTEX -
CORE

$

I

"US MATRIX

)4-

-36

Note:

The debug features embedded in the Cortex®-M4 with FPU core are a subset of the Arm®
CoreSight Design Kit.

RM0090 Rev 17

1681/1747
1711

Debug support (DBG)

RM0090

The Arm® Cortex®-M4 with FPU core provides integrated on-chip debug support. It is
comprised of:
•

SWJ-DP: Serial wire / JTAG debug port

•

AHP-AP: AHB access port

•

ITM: Instrumentation trace macrocell

•

FPB: Flash patch breakpoint

•

DWT: Data watchpoint trigger

•

TPUI: Trace port unit interface (available on larger packages, where the corresponding
pins are mapped)

•

ETM: Embedded Trace Macrocell (available on larger packages, where the
corresponding pins are mapped)

It also includes debug features dedicated to the STM32F4xx:
•

Flexible debug pinout assignment

•

MCU debug box (support for low-power modes, control over peripheral clocks, etc.)

Note:

For further information on debug functionality supported by the Arm® Cortex®-M4 with FPU
core, refer to the Cortex®-M4 with FPU -r0p1 Technical Reference Manual and to the
CoreSight Design Kit-r0p1 TRM (see Section 38.2: Reference Arm® documentation).

38.2

Reference Arm® documentation
•

Cortex®-M4 with FPU r0p1 Technical Reference Manual (TRM)
(see Related documents on page 1)

38.3

•

Arm® Debug Interface V5

•

Arm® CoreSight Design Kit revision r0p1 Technical Reference Manual

SWJ debug port (serial wire and JTAG)
The core of the STM32F4xx integrates the Serial Wire / JTAG Debug Port (SWJ-DP). It is an
Arm® standard CoreSight debug port that combines a JTAG-DP (5-pin) interface and a SWDP (2-pin) interface.
•

The JTAG Debug Port (JTAG-DP) provides a 5-pin standard JTAG interface to the
AHP-AP port.

•

The Serial Wire Debug Port (SW-DP) provides a 2-pin (clock + data) interface to the
AHP-AP port.

In the SWJ-DP, the two JTAG pins of the SW-DP are multiplexed with some of the five JTAG
pins of the JTAG-DP.

1682/1747

RM0090 Rev 17

RM0090

Debug support (DBG)
Figure 486. SWJ debug port

Figure 486 shows that the asynchronous TRACE output (TRACESWO) is multiplexed with
TDO. This means that the asynchronous trace can only be used with SW-DP, not JTAG-DP.

38.3.1

Mechanism to select the JTAG-DP or the SW-DP
By default, the JTAG-Debug Port is active.
If the debugger host wants to switch to the SW-DP, it must provide a dedicated JTAG
sequence on TMS/TCK (respectively mapped to SWDIO and SWCLK) which disables the
JTAG-DP and enables the SW-DP. This way it is possible to activate the SWDP using only
the SWCLK and SWDIO pins.
This sequence is:

38.4

1.

Send more than 50 TCK cycles with TMS (SWDIO) =1

2.

Send the 16-bit sequence on TMS (SWDIO) = 0111100111100111 (MSB transmitted
first)

3.

Send more than 50 TCK cycles with TMS (SWDIO) =1

Pinout and debug port pins
The STM32F4xx MCUs are available in various packages with different numbers of
available pins. As a result, some functionality (ETM) related to pin availability may differ
between packages.

RM0090 Rev 17

1683/1747
1711

Debug support (DBG)

38.4.1

RM0090

SWJ debug port pins
Five pins are used as outputs from the STM32F4xx for the SWJ-DP as alternate functions of
general-purpose I/Os. These pins are available on all packages.
Table 298. SWJ debug port pins
JTAG debug port

SW debug port

Pin
assignment

SWJ-DP pin name
Type

Description

JTMS/SWDIO

I

JTAG Test Mode Selection

JTCK/SWCLK

I

JTAG Test Clock

JTDI

I

JTDO/TRACESWO
NJTRST

38.4.2

Type
IO

Debug assignment
Serial Wire Data Input/Output

PA13

I

Serial Wire Clock

PA14

JTAG Test Data Input

-

-

PA15

O

JTAG Test Data Output

-

TRACESWO if async trace is
enabled

PB3

I

JTAG Test nReset

-

-

PB4

Flexible SWJ-DP pin assignment
After RESET (SYSRESETn or PORESETn), all five pins used for the SWJ-DP are assigned
as dedicated pins immediately usable by the debugger host (note that the trace outputs are
not assigned except if explicitly programmed by the debugger host).
However, the STM32F4xx MCUs offers the possibility of disabling some or all of the SWJDP ports and so, of releasing the associated pins for general-purpose IO (GPIO) usage. For
more details on how to disable SWJ-DP port pins, please refer to Section 8.3.2: I/O pin
multiplexer and mapping.
Table 299. Flexible SWJ-DP pin assignment
SWJ IO pin assigned
Available debug ports

1684/1747

PB3 /
JTDO

PB4 /
NJTRST
X

Full SWJ (JTAG-DP + SW-DP) - Reset State

X

X

X

X

Full SWJ (JTAG-DP + SW-DP) but without
NJTRST

X

X

X

X

JTAG-DP Disabled and SW-DP Enabled

X

X

JTAG-DP Disabled and SW-DP Disabled

Note:

PA13 / PA14 /
PA15 /
JTMS / JTCK /
JTDI
SWDIO SWCLK

Released

When the APB bridge write buffer is full, it takes one extra APB cycle when writing the
GPIO_AFR register. This is because the deactivation of the JTAGSW pins is done in two
cycles to guarantee a clean level on the nTRST and TCK input signals of the core.
•

Cycle 1: the JTAGSW input signals to the core are tied to 1 or 0 (to 1 for nTRST, TDI
and TMS, to 0 for TCK)

•

Cycle 2: the GPIO controller takes the control signals of the SWJTAG IO pins (like
controls of direction, pull-up/down, Schmitt trigger activation, etc.).

RM0090 Rev 17

RM0090

38.4.3

Debug support (DBG)

Internal pull-up and pull-down on JTAG pins
It is necessary to ensure that the JTAG input pins are not floating since they are directly
connected to flip-flops to control the debug mode features. Special care must be taken with
the SWCLK/TCK pin which is directly connected to the clock of some of these flip-flops.
To avoid any uncontrolled IO levels, the device embeds internal pull-ups and pull-downs on
the JTAG input pins:
•

NJTRST: Internal pull-up

•

JTDI: Internal pull-up

•

JTMS/SWDIO: Internal pull-up

•

TCK/SWCLK: Internal pull-down

Once a JTAG IO is released by the user software, the GPIO controller takes control again.
The reset states of the GPIO control registers put the I/Os in the equivalent state:
•

NJTRST: AF input pull-up

•

JTDI: AF input pull-up

•

JTMS/SWDIO: AF input pull-up

•

JTCK/SWCLK: AF input pull-down

•

JTDO: AF output floating

The software can then use these I/Os as standard GPIOs.
Note:

The JTAG IEEE standard recommends to add pull-ups on TDI, TMS and nTRST but there is
no special recommendation for TCK. However, for JTCK, the device needs an integrated
pull-down.
Having embedded pull-ups and pull-downs removes the need to add external resistors.

RM0090 Rev 17

1685/1747
1711

Debug support (DBG)

38.4.4

RM0090

Using serial wire and releasing the unused debug pins as GPIOs
To use the serial wire DP to release some GPIOs, the user software must change the GPIO
(PA15, PB3 and PB4) configuration mode in the GPIO_MODER register. This releases
PA15, PB3 and PB4 which now become available as GPIOs.
When debugging, the host performs the following actions:

Note:

•

Under system reset, all SWJ pins are assigned (JTAG-DP + SW-DP).

•

Under system reset, the debugger host sends the JTAG sequence to switch from the
JTAG-DP to the SW-DP.

•

Still under system reset, the debugger sets a breakpoint on vector reset.

•

The system reset is released and the Core halts.

•

All the debug communications from this point are done using the SW-DP. The other
JTAG pins can then be reassigned as GPIOs by the user software.

For user software designs, note that:
To release the debug pins, remember that they will be first configured either in input-pull-up
(nTRST, TMS, TDI) or pull-down (TCK) or output tristate (TDO) for a certain duration after
reset until the instant when the user software releases the pins.
When debug pins (JTAG or SW or TRACE) are mapped, changing the corresponding IO pin
configuration in the IOPORT controller has no effect.

38.5

STM32F4xx JTAG TAP connection
The STM32F4xx MCUs integrate two serially connected JTAG TAPs, the boundary scan
TAP (IR is 5-bit wide) and the Cortex®-M4 with FPU TAP (IR is 4-bit wide).
To access the TAP of the Cortex®-M4 with FPU for debug purposes:

Note:

1686/1747

1.

First, it is necessary to shift the BYPASS instruction of the boundary scan TAP.

2.

Then, for each IR shift, the scan chain contains 9 bits (=5+4) and the unused TAP
instruction must be shifted in using the BYPASS instruction.

3.

For each data shift, the unused TAP, which is in BYPASS mode, adds 1 extra data bit in
the data scan chain.

Important: Once Serial-Wire is selected using the dedicated Arm® JTAG sequence, the
boundary scan TAP is automatically disabled (JTMS forced high).

RM0090 Rev 17

RM0090

Debug support (DBG)
Figure 487. JTAG TAP connections

34-&XXX
.*4234
*4-3
37 $0
3ELECTED

4-3 N4234

*4$)

*4$/

4$)

4$/

"OUNDARY SCAN
4!0
)2 IS  BIT WIDE

4-3 N4234

4$)

4$/

#ORTEX - 4!0
)2 IS  BIT WIDE

-36

RM0090 Rev 17

1687/1747
1711

Debug support (DBG)

38.6

RM0090

ID codes and locking mechanism
There are several ID codes inside the STM32F4xx MCUs. ST strongly recommends tools
designers to lock their debuggers using the MCU DEVICE ID code located in the external
PPB memory map at address 0xE0042000.

38.6.1

MCU device ID code
The STM32F4xx MCUs integrate an MCU ID code. This ID identifies the ST MCU partnumber and the die revision. It is part of the DBG_MCU component and is mapped on the
external PPB bus (see Section 38.16 on page 1700). This code is accessible using the
JTAG debug port (four to five pins) or the SW debug port (two pins) or by the user software.
It is even accessible while the MCU is under system reset.
Only the DEV_ID[11:0] should be used for identification by the debugger/programmer tools.

DBGMCU_IDCODE
Address: 0xE004 2000
Only 32-bits access supported. Read-only.
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

r

r

r

r

r

r

r

6

5

4

3

2

1

0

r

r

r

r

r

REV_ID[15:0]
r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

Reserved

DEV_ID[11:0]
r

r

r

r

r

r

r

Bits 31:16 REV_ID[15:0] Revision identifier
This field indicates the revision of the device.
STM32F405xx/07xx and STM32F415xx/17xx devices:
0x1000 = Revision A
0x1001 = Revision Z
0x1003 = Revision 1
0x1007 = Revision 2
0x100F= Revision Y and 4
STM32F42xxx and STM32F43xxx devices:
0x1000 = Revision A
0x1003 = Revision Y
0x1007 = Revision 1
0x2001= Revision 3, 4, 5 and B
0x2007= Revision 4, 5 and B
Bits 15:12 Reserved, must be kept at reset value.
Bits 11:0 DEV_ID[11:0]: Device identifier (STM32F405xx/07xx and STM32F415xx/17xx)
The device ID is 0x413.
Bits 11:0 DEV_ID[11:0]: Device identifier (STM32F42xxx and STM32F43xxx)
The device ID is 0x419

1688/1747

RM0090 Rev 17

RM0090

38.6.2

Debug support (DBG)

Boundary scan TAP
JTAG ID code
The TAP of the STM32F4xx BSC (boundary scan) integrates a JTAG ID code equal to .

38.6.3

•

0x06413041 for STM32F405xx/07xx and STM32F415xx/17xx devices

•

0x06419041 for STM32F42xxx and STM32F43xxx devices

Cortex®-M4 with FPU TAP
The TAP of the Arm® Cortex®-M4 with FPU integrates a JTAG ID code. This ID code is the
Arm® default one and has not been modified. This code is only accessible by the JTAG
Debug Port.
This code is 0x4BA00477 (corresponds to Cortex®-M4 with FPU r0p1, see Section 38.2).

38.6.4

Cortex®-M4 with FPU JEDEC-106 ID code
The Arm® Cortex®-M4 with FPU integrates a JEDEC-106 ID code. It is located in the 4KB
ROM table mapped on the internal PPB bus at address 0xE00FF000_0xE00FFFFF.
This code is accessible by the JTAG Debug Port (4 to 5 pins) or by the SW Debug Port (two
pins) or by the user software.

38.7

JTAG debug port
A standard JTAG state machine is implemented with a 4-bit instruction register (IR) and five
data registers (for full details, refer to the Cortex®-M4 with FPU r0p1 Technical Reference
Manual (TRM), for references, see Section 38.2).
Table 300. JTAG debug port data registers
IR(3:0)

Data register

Details

1111

BYPASS
[1 bit]

-

1110

IDCODE
[32 bits]

ID CODE
0x4BA00477 (Arm® Cortex®-M4 with FPU r0p1 ID Code)

DPACC
[35 bits]

Debug port access register
This initiates a debug port and allows access to a debug port register.
– When transferring data IN:
Bits 34:3 = DATA[31:0] = 32-bit data to transfer for a write request
Bits 2:1 = A[3:2] = 2-bit address of a debug port register.
Bit 0 = RnW = Read request (1) or write request (0).
– When transferring data OUT:
Bits 34:3 = DATA[31:0] = 32-bit data which is read following a read
request
Bits 2:0 = ACK[2:0] = 3-bit Acknowledge:
010 = OK/FAULT
001 = WAIT
OTHER = reserved
Refer to Table 301 for a description of the A[3:2] bits

1010

RM0090 Rev 17

1689/1747
1711

Debug support (DBG)

RM0090
Table 300. JTAG debug port data registers (continued)

IR(3:0)

Data register

Details

1011

APACC
[35 bits]

Access port access register
Initiates an access port and allows access to an access port register.
– When transferring data IN:
Bits 34:3 = DATA[31:0] = 32-bit data to shift in for a write request
Bits 2:1 = A[3:2] = 2-bit address (sub-address AP registers).
Bit 0 = RnW= Read request (1) or write request (0).
– When transferring data OUT:
Bits 34:3 = DATA[31:0] = 32-bit data which is read following a read
request
Bits 2:0 = ACK[2:0] = 3-bit Acknowledge:
010 = OK/FAULT
001 = WAIT
OTHER = reserved
There are many AP Registers (see AHB-AP) addressed as the
combination of:
– The shifted value A[3:2]
– The current value of the DP SELECT register

1000

ABORT
[35 bits]

Abort register
– Bits 31:1 = Reserved
– Bit 0 = DAPABORT: write 1 to generate a DAP abort.

Table 301. 32-bit debug port registers addressed through the shifted value A[3:2]
Address A[3:2] value
0x0

00

Reserved, must be kept at reset value.

01

DP CTRL/STAT register. Used to:
– Request a system or debug power-up
– Configure the transfer operation for AP accesses
– Control the pushed compare and pushed verify operations.
– Read some status flags (overrun, power-up acknowledges)

0x8

10

DP SELECT register: Used to select the current access port and the
active 4-words register window.
– Bits 31:24: APSEL: select the current AP
– Bits 23:8: reserved
– Bits 7:4: APBANKSEL: select the active 4-words register window on the
current AP
– Bits 3:0: reserved

0xC

11

DP RDBUFF register: Used to allow the debugger to get the final result
after a sequence of operations (without requesting new JTAG-DP
operation)

0x4

1690/1747

Description

RM0090 Rev 17

RM0090

Debug support (DBG)

38.8

SW debug port

38.8.1

SW protocol introduction
This synchronous serial protocol uses two pins:
•

SWCLK: clock from host to target

•

SWDIO: bidirectional

The protocol allows two banks of registers (DPACC registers and APACC registers) to be
read and written to.
Bits are transferred LSB-first on the wire.
For SWDIO bidirectional management, the line must be pulled-up on the board (100 KΩ
recommended by Arm®).
Each time the direction of SWDIO changes in the protocol, a turnaround time is inserted
where the line is not driven by the host nor the target. By default, this turnaround time is one
bit time, however this can be adjusted by configuring the SWCLK frequency.

38.8.2

SW protocol sequence
Each sequence consist of three phases:
1.

Packet request (8 bits) transmitted by the host

2.

Acknowledge response (3 bits) transmitted by the target

3.

Data transfer phase (33 bits) transmitted by the host or the target
Table 302. Packet request (8-bits)
Bit

Name

Description

0

Start

Must be “1”

1

APnDP

0: DP Access
1: AP Access

2

RnW

0: Write Request
1: Read Request

4:3

A[3:2]

Address field of the DP or AP registers (refer to Table 301)

5

Parity

Single bit parity of preceding bits

6

Stop

0

7

Park

Not driven by the host. Must be read as “1” by the target because of
the pull-up

Refer to the Cortex®-M4 with FPU r0p1 TRM for a detailed description of DPACC and
APACC registers.
The packet request is always followed by the turnaround time (default 1 bit) where neither
the host nor target drive the line.

RM0090 Rev 17

1691/1747
1711

Debug support (DBG)

RM0090
Table 303. ACK response (3 bits)

Bit
0..2

Name

Description
001: FAULT
010: WAIT
100: OK

ACK

The ACK Response must be followed by a turnaround time only if it is a READ transaction
or if a WAIT or FAULT acknowledge has been received.
Table 304. DATA transfer (33 bits)
Bit
0..31
32

Name

Description

WDATA or RDATA Write or Read data
Parity

Single parity of the 32 data bits

The DATA transfer must be followed by a turnaround time only if it is a READ transaction.

38.8.3

SW-DP state machine (reset, idle states, ID code)
The State Machine of the SW-DP has an internal ID code which identifies the SW-DP. It
follows the JEP-106 standard. This ID code is the default Arm® one and is set to
0x2BA01477 (corresponding to Cortex®-M4 with FPU r0p1).

Note:

Note that the SW-DP state machine is inactive until the target reads this ID code.
•

The SW-DP state machine is in RESET STATE either after power-on reset, or after the
DP has switched from JTAG to SWD or after the line is high for more than 50 cycles

•

The SW-DP state machine is in IDLE STATE if the line is low for at least two cycles
after RESET state.

•

After RESET state, it is mandatory to first enter into an IDLE state AND to perform a
READ access of the DP-SW ID CODE register. Otherwise, the target will issue a
FAULT acknowledge response on another transactions.

Further details of the SW-DP state machine can be found in the Cortex®-M4 with FPU r0p1
TRM and the CoreSight Design Kit r0p1 TRM.

38.8.4

1692/1747

DP and AP read/write accesses
•

Read accesses to the DP are not posted: the target response can be immediate (if
ACK=OK) or can be delayed (if ACK=WAIT).

•

Read accesses to the AP are posted. This means that the result of the access is
returned on the next transfer. If the next access to be done is NOT an AP access, then
the DP-RDBUFF register must be read to obtain the result.
The READOK flag of the DP-CTRL/STAT register is updated on every AP read access
or RDBUFF read request to know if the AP read access was successful.

•

The SW-DP implements a write buffer (for both DP or AP writes), that enables it to
accept a write operation even when other transactions are still outstanding. If the write
buffer is full, the target acknowledge response is “WAIT”. With the exception of

RM0090 Rev 17

RM0090

Debug support (DBG)
IDCODE read or CTRL/STAT read or ABORT write which are accepted even if the write
buffer is full.
•

38.8.5

Because of the asynchronous clock domains SWCLK and HCLK, two extra SWCLK
cycles are needed after a write transaction (after the parity bit) to make the write
effective internally. These cycles should be applied while driving the line low (IDLE
state)
This is particularly important when writing the CTRL/STAT for a power-up request. If the
next transaction (requiring a power-up) occurs immediately, it will fail.

SW-DP registers
Access to these registers are initiated when APnDP=0
Table 305. SW-DP registers
A[3:2]

R/W

CTRLSEL bit
of SELECT
register

Register

00

Read

-

IDCODE

00

Write

-

ABORT

Notes
The manufacturer code is not set to ST
code. 0x2BA01477 (identifies the SW-DP)
-

01

Read/Write

0

DPCTRL/STAT

Purpose is to:
– request a system or debug power-up
– configure the transfer operation for AP
accesses
– control the pushed compare and pushed
verify operations.
– read some status flags (overrun, powerup acknowledges)

01

Read/Write

1

WIRE
CONTROL

Purpose is to configure the physical serial
port protocol (like the duration of the
turnaround time)

10

Read

-

READ
RESEND

Enables recovery of the read data from a
corrupted debugger transfer, without
repeating the original AP transfer.

10

Write

-

SELECT

The purpose is to select the current access
port and the active 4-words register window

READ
BUFFER

This read buffer is useful because AP
accesses are posted (the result of a read AP
request is available on the next AP
transaction).
This read buffer captures data from the AP,
presented as the result of a previous read,
without initiating a new transaction

11

Read/Write

-

RM0090 Rev 17

1693/1747
1711

Debug support (DBG)

38.8.6

RM0090

SW-AP registers
Access to these registers are initiated when APnDP=1
There are many AP Registers (see AHB-AP) addressed as the combination of:

38.9

•

The shifted value A[3:2]

•

The current value of the DP SELECT register

AHB-AP (AHB access port) - valid for both JTAG-DP
and SW-DP
Features:
•

System access is independent of the processor status.

•

Either SW-DP or JTAG-DP accesses AHB-AP.

•

The AHB-AP is an AHB master into the Bus Matrix. Consequently, it can access all the
data buses (Dcode Bus, System Bus, internal and external PPB bus) but the ICode
bus.

•

Bitband transactions are supported.

•

AHB-AP transactions bypass the FPB.

The address of the 32-bits AHP-AP resisters are 6-bits wide (up to 64 words or 256 bytes)
and consists of:
c)

Bits [7:4] = the bits [7:4] APBANKSEL of the DP SELECT register

d)

Bits [3:2] = the 2 address bits of A[3:2] of the 35-bit packet request for SW-DP.

The AHB-AP of the Cortex®-M4 with FPU includes 9 x 32-bits registers:
Table 306. Cortex®-M4 with FPU AHB-AP registers
Address
offset

Register name

Notes

0x00

AHB-AP Control and Status
Word

Configures and controls transfers through the AHB
interface (size, hprot, status on current transfer, address
increment type

0x04

AHB-AP Transfer Address

-

0x0C

AHB-AP Data Read/Write

-

0x10

AHB-AP Banked Data 0

0x14

AHB-AP Banked Data 1

0x18

AHB-AP Banked Data 2

0x1C

AHB-AP Banked Data 3

0xF8

AHB-AP Debug ROM Address Base Address of the debug interface

0xFC

AHB-AP ID Register

Directly maps the 4 aligned data words without rewriting
the Transfer Address Register.

-

Refer to the Cortex®-M4 with FPU r0p1 TRM for further details.

1694/1747

RM0090 Rev 17

RM0090

38.10

Debug support (DBG)

Core debug
Core debug is accessed through the core debug registers. Debug access to these registers
is by means of the Advanced High-performance Bus (AHB-AP) port. The processor can
access these registers directly over the internal Private Peripheral Bus (PPB).
It consists of 4 registers:
Table 307. Core debug registers

Note:

Register

Description

DHCSR

The 32-bit Debug Halting Control and Status Register
This provides status information about the state of the processor enable core debug
halt and step the processor

DCRSR

The 17-bit Debug Core Register Selector Register:
This selects the processor register to transfer data to or from.

DCRDR

The 32-bit Debug Core Register Data Register:
This holds data for reading and writing registers to and from the processor selected
by the DCRSR (Selector) register.

DEMCR

The 32-bit Debug Exception and Monitor Control Register:
This provides Vector Catching and Debug Monitor Control. This register contains a
bit named TRCENA which enable the use of a TRACE.

Important: these registers are not reset by a system reset. They are only reset by a poweron reset.
Refer to the Cortex®-M4 with FPU r0p1 TRM for further details.
To Halt on reset, it is necessary to:
•

enable the bit0 (VC_CORRESET) of the Debug and Exception Monitor Control
Register

•

enable the bit0 (C_DEBUGEN) of the Debug Halting Control and Status Register.

RM0090 Rev 17

1695/1747
1711

Debug support (DBG)

38.11

RM0090

Capability of the debugger host to connect under system
reset
The reset system of the STM32F4xx MCU comprises the following reset sources:
•

POR (power-on reset) which asserts a RESET at each power-up.

•

Internal watchdog reset

•

Software reset

•

External reset

The Cortex®-M4 with FPU differentiates the reset of the debug part (generally
PORRESETn) and the other one (SYSRESETn)
This way, it is possible for the debugger to connect under System Reset, programming the
Core Debug Registers to halt the core when fetching the reset vector. Then the host can
release the system reset and the core will immediately halt without having executed any
instructions. In addition, it is possible to program any debug features under System Reset.
Note:

It is highly recommended for the debugger host to connect (set a breakpoint in the reset
vector) under system reset.

38.12

FPB (Flash patch breakpoint)
The FPB unit:
•

implements hardware breakpoints

•

patches code and data from code space to system space. This feature gives the
possibility to correct software bugs located in the Code Memory Space.

The use of a Software Patch or a Hardware Breakpoint is exclusive.
The FPB consists of:

1696/1747

•

2 literal comparators for matching against literal loads from Code Space and remapping
to a corresponding area in the System Space.

•

6 instruction comparators for matching against instruction fetches from Code Space.
They can be used either to remap to a corresponding area in the System Space or to
generate a Breakpoint Instruction to the core.

RM0090 Rev 17

RM0090

38.13

Debug support (DBG)

DWT (data watchpoint trigger)
The DWT unit consists of four comparators. They are configurable as:
•

a hardware watchpoint or

•

a trigger to an ETM or

•

a PC sampler or

•

a data address sampler

The DWT also provides some means to give some profiling informations. For this, some
counters are accessible to give the number of:
•

Clock cycle

•

Folded instructions

•

Load store unit (LSU) operations

•

Sleep cycles

•

CPI (clock per instructions)

•

Interrupt overhead

38.14

ITM (instrumentation trace macrocell)

38.14.1

General description
The ITM is an application-driven trace source that supports printf style debugging to trace
Operating System (OS) and application events, and emits diagnostic system information.
The ITM emits trace information as packets which can be generated as:
•

Software trace. Software can write directly to the ITM stimulus registers to emit
packets.

•

Hardware trace. The DWT generates these packets, and the ITM emits them.

•

Time stamping. Timestamps are emitted relative to packets. The ITM contains a 21-bit
counter to generate the timestamp. The Cortex®-M4 with FPU clock or the bit clock rate
of the Serial Wire Viewer (SWV) output clocks the counter.

The packets emitted by the ITM are output to the TPIU (Trace Port Interface Unit). The
formatter of the TPIU adds some extra packets (refer to TPIU) and then output the complete
packets sequence to the debugger host.
The bit TRCEN of the Debug Exception and Monitor Control Register must be enabled
before programming or using the ITM.

38.14.2

Time stamp packets, synchronization and overflow packets
Time stamp packets encode time stamp information, generic control and synchronization. It
uses a 21-bit timestamp counter (with possible prescalers) which is reset at each time
stamp packet emission. This counter can be either clocked by the CPU clock or the SWV
clock.
A synchronization packet consists of 6 bytes equal to 0x80_00_00_00_00_00 which is
emitted to the TPIU as 00 00 00 00 00 80 (LSB emitted first).
A synchronization packet is a timestamp packet control. It is emitted at each DWT trigger.

RM0090 Rev 17

1697/1747
1711

Debug support (DBG)

RM0090

For this, the DWT must be configured to trigger the ITM: the bit CYCCNTENA (bit0) of the
DWT Control Register must be set. In addition, the bit2 (SYNCENA) of the ITM Trace
Control Register must be set.
Note:

If the SYNENA bit is not set, the DWT generates Synchronization triggers to the TPIU which
will send only TPIU synchronization packets and not ITM synchronization packets.
An overflow packet consists is a special timestamp packets which indicates that data has
been written but the FIFO was full.
Table 308. Main ITM registers
Address
@E0000FB0

Register
ITM lock access

Details
Write 0xC5ACCE55 to unlock Write Access to the other ITM
registers
Bits 31-24 = Always 0
Bits 23 = Busy
Bits 22-16 = 7-bits ATB ID which identifies the source of the
trace data.
Bits 15-10 = Always 0
Bits 9:8 = TSPrescale = Time Stamp Prescaler
Bits 7-5 = Reserved

@E0000E80

ITM trace control

Bit 4 = SWOENA = Enable SWV behavior (to clock the
timestamp counter by the SWV clock).
Bit 3 = DWTENA: Enable the DWT Stimulus
Bit 2 = SYNCENA: this bit must be to 1 to enable the DWT to
generate synchronization triggers so that the TPIU can then
emit the synchronization packets.
Bit 1 = TSENA (Timestamp Enable)
Bit 0 = ITMENA: Global Enable Bit of the ITM
Bit 3: mask to enable tracing ports31:24

@E0000E40

ITM trace privilege

Bit 2: mask to enable tracing ports23:16
Bit 1: mask to enable tracing ports15:8
Bit 0: mask to enable tracing ports7:0

@E0000E00

ITM trace enable

@E0000000- Stimulus port
E000007C
registers 0-31

1698/1747

Each bit enables the corresponding Stimulus port to generate
trace.
Write the 32-bits data on the selected Stimulus Port (32
available) to be traced out.

RM0090 Rev 17

RM0090

Debug support (DBG)

Example of configuration
To output a simple value to the TPIU:
•

Configure the TPIU and assign TRACE I/Os by configuring the DBGMCU_CR (refer to
Section 38.17.2: TRACE pin assignment and Section 38.16.3: Debug MCU
configuration register)

•

Write 0xC5ACCE55 to the ITM Lock Access Register to unlock the write access to the
ITM registers

•

Write 0x00010005 to the ITM Trace Control Register to enable the ITM with Sync
enabled and an ATB ID different from 0x00

•

Write 0x1 to the ITM Trace Enable Register to enable the Stimulus Port 0

•

Write 0x1 to the ITM Trace Privilege Register to unmask stimulus ports 7:0

•

Write the value to output in the Stimulus Port Register 0: this can be done by software
(using a printf function)

38.15

ETM (Embedded trace macrocell)

38.15.1

ETM general description
The ETM enables the reconstruction of program execution. Data are traced using the Data
Watchpoint and Trace (DWT) component or the Instruction Trace Macrocell (ITM) whereas
instructions are traced using the Embedded Trace Macrocell (ETM).
The ETM transmits information as packets and is triggered by embedded resources. These
resources must be programmed independently and the trigger source is selected using the
Trigger Event Register (0xE0041008). An event could be a simple event (address match
from an address comparator) or a logic equation between 2 events. The trigger source is
one of the fourth comparators of the DWT module, The following events can be monitored:
•

Clock cycle matching

•

Data address matching

For more informations on the trigger resources refer to Section 38.13: DWT (data
watchpoint trigger).
The packets transmitted by the ETM are output to the TPIU (Trace Port Interface Unit). The
formatter of the TPIU adds some extra packets (refer to Section 38.17: TPIU (trace port
interface unit)) and then outputs the complete packet sequence to the debugger host.

38.15.2

ETM signal protocol and packet types
This part is described in the chapter 7 ETMv3 Signal Protocol of the Arm® IHI 0014N
document.

RM0090 Rev 17

1699/1747
1711

Debug support (DBG)

38.15.3

RM0090

Main ETM registers
For more information on registers refer to the chapter 3 of the Arm® IHI 0014N specification.
Table 309. Main ETM registers
Address

Details

0xE0041FB0 ETM Lock Access

Write 0xC5ACCE55 to unlock the write access to the
other ETM registers.

0xE0041000 ETM Control

This register controls the general operation of the ETM,
for instance how tracing is enabled.

0xE0041010 ETM Status

This register provides information about the current
status of the trace and trigger logic.

0xE0041008 ETM Trigger Event

This register defines the event that will control trigger.

0xE004101C

38.15.4

Register

ETM Trace Enable
Control

This register defines which comparator is selected.

0xE0041020 ETM Trace Enable Event

This register defines the trace enabling event.

0xE0041024 ETM Trace Start/Stop

This register defines the traces used by the trigger source
to start and stop the trace, respectively.

ETM configuration example
To output a simple value to the TPIU:

38.16

•

Configure the TPIU and enable the I/IO_TRACEN to assign TRACE I/Os in the
STM32F4xx debug configuration register.

•

Write 0xC5AC CE55 to the ETM Lock Access Register to unlock the write access to the
ITM registers

•

Write 0x0000 1D1E to the ETM control register (configure the trace)

•

Write 0x0000 406F to the ETM Trigger Event register (define the trigger event)

•

Write 0x0000 006F to the ETM Trace Enable Event register (define an event to
start/stop)

•

Write 0x0200 00000x0000 0001 to the ETM Trace Start/stop register (enable the trace)

•

Write 0x0000191E to the ETM Control Register (end of configuration)

MCU debug component (DBGMCU)
The MCU debug component helps the debugger provide support for:

38.16.1

•

Low-power modes

•

Clock control for timers, watchdog, I2C and bxCAN during a breakpoint

•

Control of the trace pins assignment

Debug support for low-power modes
To enter low-power mode, the instruction WFI or WFE must be executed.
The MCU implements several low-power modes which can either deactivate the CPU clock
or reduce the power of the CPU.

1700/1747

RM0090 Rev 17

RM0090

Debug support (DBG)
The core does not allow FCLK or HCLK to be turned off during a debug session. As these
are required for the debugger connection, during a debug, they must remain active. The
MCU integrates special means to allow the user to debug software in low-power modes.
For this, the debugger host must first set some debug configuration registers to change the
low-power mode behavior:

38.16.2

•

In Sleep mode, DBG_SLEEP bit of DBGMCU_CR register must be previously set by
the debugger. This will feed HCLK with the same clock that is provided to FCLK
(system clock previously configured by the software).

•

In Stop mode, the bit DBG_STOP must be previously set by the debugger. This will
enable the internal RC oscillator clock to feed FCLK and HCLK in STOP mode.

Debug support for timers, watchdog, bxCAN and I2C
During a breakpoint, it is necessary to choose how the counter of timers and watchdog
should behave:
•

They can continue to count inside a breakpoint. This is usually required when a PWM is
controlling a motor, for example.

•

They can stop to count inside a breakpoint. This is required for watchdog purposes.

For the bxCAN, the user can choose to block the update of the receive register during a
breakpoint.
For the I2C, the user can choose to block the SMBUS timeout during a breakpoint.
For timers having complementary outputs, when the counter is stopped
(DBG_TIMx_STOP = 1), the outputs are disabled (as if the MOE bit was reset) for safety
purposes.

38.16.3

Debug MCU configuration register
This register allows the configuration of the MCU under DEBUG. This concerns:
•

Low-power mode support

•

Timer and watchdog counter support

•

bxCAN communication support

•

Trace pin assignment

This DBGMCU_CR is mapped on the External PPB bus at address 0xE0042004
It is asynchronously reset by the PORESET (and not the system reset). It can be written by
the debugger under system reset.
If the debugger host does not support these features, it is still possible for the user software
to write to these registers.

DBGMCU_CR register
Address: 0xE004 2004
Only 32-bit access supported
POR Reset: 0x0000 0000 (not reset by system reset)

RM0090 Rev 17

1701/1747
1711

Debug support (DBG)

31

30

29

28

RM0090

27

26

25

24

23

22

21

20

19

18

17

16

6

5

4

3

2

1

0

DBG_
STAND
BY

DBG_
STOP

DBG_
SLEEP

rw

rw

rw

Reserved
15

14

13

12

11

Reserved

10

9

8

7

TRACE_
MODE
[1:0]
rw

rw

TRACE
_IOEN
rw

Reserved

Bits 31:8 Reserved, must be kept at reset value.
Bits 7:5 TRACE_MODE[1:0] and TRACE_IOEN: Trace pin assignment control
– With TRACE_IOEN=0:
TRACE_MODE=xx: TRACE pins not assigned (default state)
– With TRACE_IOEN=1:
–
TRACE_MODE=00: TRACE pin assignment for Asynchronous Mode
–
TRACE_MODE=01: TRACE pin assignment for Synchronous Mode with a
TRACEDATA size of 1
–
TRACE_MODE=10: TRACE pin assignment for Synchronous Mode with a
TRACEDATA size of 2
–
TRACE_MODE=11: TRACE pin assignment for Synchronous Mode with a
TRACEDATA size of 4
Bits 4:3 Reserved, must be kept at reset value.
Bit 2 DBG_STANDBY: Debug Standby mode
0: (FCLK=Off, HCLK=Off) The whole digital part is unpowered.
From software point of view, exiting from Standby is identical than fetching reset vector
(except a few status bit indicated that the MCU is resuming from Standby)
1: (FCLK=On, HCLK=On) In this case, the digital part is not unpowered and FCLK and
HCLK are provided by the internal RC oscillator which remains active. In addition, the MCU
generate a system reset during Standby mode so that exiting from Standby is identical than
fetching from reset
Bit 1 DBG_STOP: Debug Stop mode
0: (FCLK=Off, HCLK=Off) In STOP mode, the clock controller disables all clocks (including
HCLK and FCLK). When exiting from STOP mode, the clock configuration is identical to the
one after RESET (CPU clocked by the 8 MHz internal RC oscillator (HSI)). Consequently,
the software must reprogram the clock controller to enable the PLL, the Xtal, etc.
1: (FCLK=On, HCLK=On) In this case, when entering STOP mode, FCLK and HCLK are
provided by the internal RC oscillator which remains active in STOP mode. When exiting
STOP mode, the software must reprogram the clock controller to enable the PLL, the Xtal,
etc. (in the same way it would do in case of DBG_STOP=0)
Bit 0 DBG_SLEEP: Debug Sleep mode
0: (FCLK=On, HCLK=Off) In Sleep mode, FCLK is clocked by the system clock as
previously configured by the software while HCLK is disabled.
In Sleep mode, the clock controller configuration is not reset and remains in the previously
programmed state. Consequently, when exiting from Sleep mode, the software does not
need to reconfigure the clock controller.
1: (FCLK=On, HCLK=On) In this case, when entering Sleep mode, HCLK is fed by the same
clock that is provided to FCLK (system clock as previously configured by the software).

1702/1747

RM0090 Rev 17

RM0090

Debug support (DBG)

38.16.4

Debug MCU APB1 freeze register (DBGMCU_APB1_FZ)
The DBGMCU_APB1_FZ register is used to configure the MCU under Debug. It concerns
APB1 peripherals. It is mapped on the external PPB bus at address 0xE004 2008.
The register is asynchronously reset by the POR (and not the system reset). It can be
written by the debugger under system reset.
Address : 0xE004 2008
Only 32-bits access are supported.
Power-on reset (POR): 0x0000 0000 (not reset by system reset)

Bits 31:27

rw

rw

rw

8

7

6

5

4

3

2

1

0

DBG_TIM4_STOP

DBG_TIM3_STOP

DBG_TIM2_STOP

16

DBG_TIM5_STOP

17

DBG_TIM6_STOP

18

DBG_I2C1_SMBUS_TIMEOUT

19

DBG_TIM7_STOP

rw

20

DBG_I2C2_SMBUS_TIMEOUT

rw

21

DBG_TIM12_STOP

Reserved

22

DBG_I2C3_SMBUS_TIMEOUT

9

23

DBG_TIM13_STOP

rw

10

rw

rw

rw

rw

rw

rw

rw

rw

rw

Reserved

rw
11

24

DBG_TIM14_STOP

13

Reserved

14

25

12

Reserved

15

26

DBG_CAN1_STOP

27

DBG_CAN2_STOP

28

DBG_RTC_STOP

29

DBG_WWDG_STOP

30

DBG_IWDG_STOP

31

Reserved

Reserved, must be kept at reset value.

Bit 26 DBG_CAN2_STOP: Debug CAN2 stopped when Core is halted
0: Same behavior as in normal mode
1: The CAN2 receive registers are frozen
Bit 25 DBG_CAN1_STOP: Debug CAN2 stopped when Core is halted
0: Same behavior as in normal mode
1: The CAN2 receive registers are frozen
Bit 24

Reserved, must be kept at reset value.

Bit 23 DBG_I2C3_SMBUS_TIMEOUT: SMBUS timeout mode stopped when Core is halted
0: Same behavior as in normal mode
1: The SMBUS timeout is frozen
Bit 22 DBG_I2C2_SMBUS_TIMEOUT: SMBUS timeout mode stopped when Core is halted
0: Same behavior as in normal mode
1: The SMBUS timeout is frozen
Bit 21 DBG_I2C1_SMBUS_TIMEOUT: SMBUS timeout mode stopped when Core is halted
0: Same behavior as in normal mode
1: The SMBUS timeout is frozen

RM0090 Rev 17

1703/1747
1711

Debug support (DBG)

Bit 20:13

RM0090

Reserved, must be kept at reset value.

Bit 12 DBG_IWDG_STOP: Debug independent watchdog stopped when core is halted
0: The independent watchdog counter clock continues even if the core is halted
1: The independent watchdog counter clock is stopped when the core is halted
Bit 11 DBG_WWDG_STOP: Debug Window Watchdog stopped when Core is halted
0: The window watchdog counter clock continues even if the core is halted
1: The window watchdog counter clock is stopped when the core is halted
Bit 10 DBG_RTC_STOP: RTC stopped when Core is halted
0: The RTC counter clock continues even if the core is halted
1: The RTC counter clock is stopped when the core is halted
Bit 9

Reserved, must be kept at reset value.

Bits 8:0 DBG_TIMx_STOP: TIMx counter stopped when core is halted (x=2..7, 12..14)
0: The clock of the involved timer counter is fed even if the core is halted
1: The clock of the involved timer counter is stopped and the outputs are disabled when the
core is halted

38.16.5

Debug MCU APB2 Freeze register (DBGMCU_APB2_FZ)
The DBGMCU_APB2_FZ register is used to configure the MCU under Debug. It concerns
APB2 peripherals.
This register is mapped on the external PPB bus at address 0xE004 200C
It is asynchronously reset by the POR (and not the system reset). It can be written by the
debugger under system reset.
Address: 0xE004 200C
Only 32-bit access is supported.
POR: 0x0000 0000 (not reset by system reset)

31

30

29

28

27

26

25

24

23

22

21

20

19

Reserved
15

14

13

12

11

10

9

18

17

8

7

6

Reserved

5

4

3

rw

rw

rw

2

1

0

DBG_TIM8_ DBG_TIM1_
STOP
STOP
rw

Bits 31:19

16

DBG_TIM11 DBG_TIM10 DBG_TIM9_
_STOP
_STOP
STOP

rw

Reserved, must be kept at reset value.

Bits 18:16 DBG_TIMx_STOP: TIMx counter stopped when core is halted (x=9..11)
0: The clock of the involved timer counter is fed even if the core is halted
1: The clock of the involved timer counter is stopped and the outputs are disabled when the
core is halted

1704/1747

RM0090 Rev 17

RM0090

Debug support (DBG)

Bits 15:

Reserved, must be kept at reset value.

Bit 1 DBG_TIM8_STOP: TIM8 counter stopped when core is halted
0: The clock of the involved timer counter is fed even if the core is halted
1: The clock of the involved timer counter is stopped and the outputs are disabled when the
core is halted
Bit 0 DBG_TIM1_STOP: TIM1 counter stopped when core is halted
0: The clock of the involved timer counter is fed even if the core is halted
1: The clock of the involved timer counter is stopped and the outputs are disabled when the
core is halted

38.17

TPIU (trace port interface unit)

38.17.1

Introduction
The TPIU acts as a bridge between the on-chip trace data from the ITM and the ETM.
The output data stream encapsulates the trace source ID, that is then captured by a trace
port analyzer (TPA).
The core embeds a simple TPIU, especially designed for low-cost debug (consisting of a
special version of the CoreSight TPIU).
Figure 488. TPIU block diagram

RM0090 Rev 17

1705/1747
1711

Debug support (DBG)

38.17.2

RM0090

TRACE pin assignment
•

Asynchronous mode
The asynchronous mode requires 1 extra pin and is available on all packages. It is only
available if using Serial Wire mode (not in JTAG mode).
Table 310. Asynchronous TRACE pin assignment
Trace synchronous mode
TPUI pin name
Type

TRACESWO

•

O

Description
TRACE Async Data Output

STM32F4xx pin
assignment
PB3

Synchronous mode
The synchronous mode requires from 2 to 6 extra pins depending on the data trace
size and is only available in the larger packages. In addition it is available in JTAG
mode and in Serial Wire mode and provides better bandwidth output capabilities than
asynchronous trace.
Table 311. Synchronous TRACE pin assignment
Trace synchronous mode
TPUI pin name
Type

Description

STM32F4xxpin
assignment

TRACECK

O

TRACE Clock

PE2

TRACED[3:0]

O

TRACE Sync Data Outputs
Can be 1, 2 or 4.

PE[6:3]

TPUI TRACE pin assignment
By default, these pins are NOT assigned. They can be assigned by setting the
TRACE_IOEN and TRACE_MODE bits in the MCU Debug component configuration
register. This configuration has to be done by the debugger host.
In addition, the number of pins to assign depends on the trace configuration (asynchronous
or synchronous).
•

Asynchronous mode: 1 extra pin is needed

•

Synchronous mode: from 2 to 5 extra pins are needed depending on the size of the
data trace port register (1, 2 or 4):
–

TRACECK

–

TRACED(0) if port size is configured to 1, 2 or 4

–

TRACED(1) if port size is configured to 2 or 4

–

TRACED(2) if port size is configured to 4

–

TRACED(3) if port size is configured to 4

To assign the TRACE pin, the debugger host must program the bits TRACE_IOEN and
TRACE_MODE[1:0] of the Debug MCU configuration Register (DBGMCU_CR). By default
the TRACE pins are not assigned.
This register is mapped on the external PPB and is reset by the PORESET (and not by the
SYSTEM reset). It can be written by the debugger under SYSTEM reset.

1706/1747

RM0090 Rev 17

RM0090

Debug support (DBG)
Table 312. Flexible TRACE pin assignment

DBGMCU_CR
register

TRACE IO pin assigned

Pins
TRACE assigned for:
PE2/
TRACE
PE3 /
PE4 /
PE5 /
PE6 /
PB3 /JTDO/
_MODE
_IOEN
TRACESWO TRACECK TRACED[0] TRACED[1] TRACED[2] TRACED[3]
[1:0]
0

XX

No Trace
(default state)

1

00

Asynchronous
TRACESWO
Trace

1

01

Synchronous
Trace 1 bit

TRACECK TRACED[0]

1

10

Synchronous
Trace 2 bit

Released (1) TRACECK TRACED[0] TRACED[1]

1

11

Synchronous
Trace 4 bit

Released (1)

-

Released
(usable as GPIO)

-

-

-

-

-

TRACECK TRACED[0] TRACED[1] TRACED[2] TRACED[3]

1. When Serial Wire mode is used, it is released. But when JTAG is used, it is assigned to JTDO.

Note:

By default, the TRACECLKIN input clock of the TPIU is tied to GND. It is assigned to HCLK
two clock cycles after the bit TRACE_IOEN has been set.
The debugger must then program the Trace Mode by writing the PROTOCOL[1:0] bits in the
SPP_R (Selected Pin Protocol) register of the TPIU.
•

PROTOCOL=00: Trace Port Mode (synchronous)

•

PROTOCOL=01 or 10: Serial Wire (Manchester or NRZ) Mode (asynchronous mode).
Default state is 01

It then also configures the TRACE port size by writing the bits [3:0] in the CPSPS_R
(Current Sync Port Size Register) of the TPIU:

38.17.3

•

0x1 for 1 pin (default state)

•

0x2 for 2 pins

•

0x8 for 4 pins

TPUI formatter
The formatter protocol outputs data in 16-byte frames:
•

seven bytes of data

•

eight bytes of mixed-use bytes consisting of:

•

–

1 bit (LSB) to indicate it is a DATA byte (‘0) or an ID byte (‘1).

–

7 bits (MSB) which can be data or change of source ID trace.

one byte of auxiliary bits where each bit corresponds to one of the eight mixed-use
bytes:
–

if the corresponding byte was a data, this bit gives bit0 of the data.

–

if the corresponding byte was an ID change, this bit indicates when that ID change
takes effect.

RM0090 Rev 17

1707/1747
1711

Debug support (DBG)

RM0090

Note:

Refer to the Arm® CoreSight Architecture Specification v1.0 (Arm® IHI 0029B) for further
information

38.17.4

TPUI frame synchronization packets
The TPUI can generate two types of synchronization packets:
•

The Frame Synchronization packet (or Full Word Synchronization packet)
It consists of the word: 0x7F_FF_FF_FF (LSB emitted first). This sequence can not
occur at any other time provided that the ID source code 0x7F has not been used.
It is output periodically between frames.
In continuous mode, the TPA must discard all these frames once a synchronization
frame has been found.

•

The Half-Word Synchronization packet
It consists of the half word: 0x7F_FF (LSB emitted first).
It is output periodically between or within frames.
These packets are only generated in continuous mode and enable the TPA to detect
that the TRACE port is in IDLE mode (no TRACE to be captured). When detected by
the TPA, it must be discarded.

38.17.5

Transmission of the synchronization frame packet
There is no Synchronization Counter register implemented in the TPIU of the core.
Consequently, the synchronization trigger can only be generated by the DWT. Refer to the
registers DWT Control Register (bits SYNCTAP[11:10]) and the DWT Current PC Sampler
Cycle Count Register.
The TPUI Frame synchronization packet (0x7F_FF_FF_FF) is emitted:

38.17.6

•

after each TPIU reset release. This reset is synchronously released with the rising
edge of the TRACECLKIN clock. This means that this packet is transmitted when the
TRACE_IOEN bit in the DBGMCU_CFG register is set. In this case, the word
0x7F_FF_FF_FF is not followed by any formatted packet.

•

at each DWT trigger (assuming DWT has been previously configured). Two cases
occur:
–

If the bit SYNENA of the ITM is reset, only the word 0x7F_FF_FF_FF is emitted
without any formatted stream which follows.

–

If the bit SYNENA of the ITM is set, then the ITM synchronization packets will
follow (0x80_00_00_00_00_00), formatted by the TPUI (trace source ID added).

Synchronous mode
The trace data output size can be configured to 4, 2 or 1 pin: TRACED(3:0)
The output clock is output to the debugger (TRACECK)
Here, TRACECLKIN is driven internally and is connected to HCLK only when TRACE is
used.

Note:

In this synchronous mode, it is not required to provide a stable clock frequency.
The TRACE I/Os (including TRACECK) are driven by the rising edge of TRACLKIN (equal
to HCLK). Consequently, the output frequency of TRACECK is equal to HCLK/2.

1708/1747

RM0090 Rev 17

RM0090

38.17.7

Debug support (DBG)

Asynchronous mode
This is a low cost alternative to output the trace using only 1 pin: this is the asynchronous
output pin TRACESWO. Obviously there is a limited bandwidth.
TRACESWO is multiplexed with JTDO when using the SW-DP pin. This way, this
functionality is available in all STM32F4xx packages.
This asynchronous mode requires a constant frequency for TRACECLKIN. For the standard
UART (NRZ) capture mechanism, 5% accuracy is needed. The Manchester encoded
version is tolerant up to 10%.

38.17.8

TRACECLKIN connection inside the STM32F4xx
In the STM32F4xx, this TRACECLKIN input is internally connected to HCLK. This means
that when in asynchronous trace mode, the application is restricted to use to time frames
where the CPU frequency is stable.

Note:

Important: when using asynchronous trace: it is important to be aware that:
The default clock of the STM32F4xx MCUs is the internal RC oscillator. Its frequency under
reset is different from the one after reset release. This is because the RC calibration is the
default one under system reset and is updated at each system reset release.
Consequently, the trace port analyzer (TPA) should not enable the trace (with the
TRACE_IOEN bit) under system reset, because a Synchronization Frame Packet will be
issued with a different bit time than trace packets which will be transmitted after reset
release.

38.17.9

TPIU registers
The TPIU APB registers can be read and written only if the bit TRCENA of the Debug
Exception and Monitor Control Register (DEMCR) is set. Otherwise, the registers are read
as zero (the output of this bit enables the PCLK of the TPIU).
Table 313. Important TPIU registers

Address

Register

Description
Allows the trace port size to be selected:
Bit 0: Port size = 1
Bit 1: Port size = 2
Bit 2: Port size = 3, not supported
Bit 3: Port Size = 4
Only 1 bit must be set. By default, the port size is one bit. (0x00000001)

0xE0040004

Current port size

0xE00400F0

Allows the Trace Port Protocol to be selected:
Bit1:0=
00: Sync Trace Port Mode
Selected pin protocol
01: Serial Wire Output - manchester (default value)
10: Serial Wire Output - NRZ
11: reserved

RM0090 Rev 17

1709/1747
1711

Debug support (DBG)

RM0090
Table 313. Important TPIU registers (continued)

Address

Register

Description

0xE0040304

Formatter and flush
control

Bits 31-9 = always ‘0
Bit 8 = TrigIn = always ‘1 to indicate that triggers are indicated
Bits 7-4 = always 0
Bits 3-2 = always 0
Bit 1 = EnFCont. In Sync Trace mode (Select_Pin_Protocol register
bit1:0=00), this bit is forced to ‘1: the formatter is automatically enabled in
continuous mode. In asynchronous mode (Select_Pin_Protocol register
bit1:0 <> 00), this bit can be written to activate or not the formatter.
Bit 0 = always 0
The resulting default value is 0x102
Note: In synchronous mode, because the TRACECTL pin is not mapped
outside the chip, the formatter is always enabled in continuous mode -this
way the formatter inserts some control packets to identify the source of the
trace packets).

0xE0040300

Formatter and flush
status

Not used in Cortex®-M4 with FPU, always read as 0x00000008

38.17.10 Example of configuration

1710/1747

•

Set the bit TRCENA in the Debug Exception and Monitor Control Register (DEMCR)

•

Write the TPIU Current Port Size Register to the desired value (default is 0x1 for a 1-bit
port size)

•

Write TPIU Formatter and Flush Control Register to 0x102 (default value)

•

Write the TPIU Select Pin Protocol to select the sync or async mode. Example: 0x2 for
async NRZ mode (UART like)

•

Write the DBGMCU control register to 0x20 (bit IO_TRACEN) to assign TRACE I/Os
for async mode. A TPIU Sync packet is emitted at this time (FF_FF_FF_7F)

•

Configure the ITM and write the ITM Stimulus register to output a value

RM0090 Rev 17

RM0090

38.18

Debug support (DBG)

DBG register map
The following table summarizes the Debug registers.

Reset value

Reserved

0

0

0

0

0

0

X

X

X

X

DBG_STOP

0

X

DBG_SLEEP

Reserved

0

X

0

0

0

DBG_TIM2_STOP

Reserved

X

DBG_STANDBY

0

X

DBG_TIM3_STOP

0

X

Reserved

0

X

DBG_TIM4_STOP

0

0

X

DBG_TIM5_STOP

0

0

X

DBG_TIM6_STOP

DBG_I2C1_SMBUS_TIMEOUT

0

X

DBG_TIM7_STOP

DBG_I2C2_SMBUS_TIMEOUT

0

X

0

0

0

0

0

0

0

0

0

0

Reserved

DBG_TIM1_STOP

DBGMCU_
APB2_FZ

0

X

DBG_TIM8_STOP

0xE004
200C

0

Reserved

Reset value

Reserved

DBG_I2C3_SMBUS_TIMEOUT

DBGMCU_
APB1_FZ

DBG_CAN1_STOP

0xE004
2008

DBG_CAN2_STOP

Reset value

X

TRACE_
MODE
[1:0]
TRACE_

Reserved

X

DBG_TIM12_STOP

X

DBG_TIM13_STOP

X

DBG_TIM14_STOP

X

Reserved

X

DBG_RTC_STOP

X

DBG_IWDG_STOP

X

DBG_TIM8_STOP

X

DBG_I2C2_SMBUS_TIMEOUT

X

DBG_TIM9_STOP

X

DBG_TIM5_STOP

DBGMCU_CR

X

DBG_TIM10_STOP

0xE004
2004

X

DEV_ID

Reserved

DBG_TIM6_STOP

Reset value(1)

REV_ID

DBG_WWDG_STOP

DBGMCU
_IDCODE

DBG_TIM11_STOP

0xE004
2000

Register

DBG_TIM7_STOP

Addr.

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Table 314. DBG register map and reset values

0

0

1. The reset value is product dependent. For more information, refer to Section 38.6.1: MCU device ID code.

RM0090 Rev 17

1711/1747
1711

Device electronic signature

39

RM0090

Device electronic signature
The electronic signature is stored in the Flash memory area. It can be read using the
JTAG/SWD or the CPU. It contains factory-programmed identification data that allow the
user firmware or other external devices to automatically match its interface to the
characteristics of the STM32F4xx microcontrollers.

39.1

Unique device ID register (96 bits)
The unique device identifier is ideally suited:
•

for use as serial numbers (for example USB string serial numbers or other end
applications)

•

for use as security keys in order to increase the security of code in Flash memory while
using and combining this unique ID with software cryptographic primitives and
protocols before programming the internal Flash memory

•

to activate secure boot processes, etc.

The 96-bit unique device identifier provides a reference number which is unique for any
device and in any context. These bits can never be altered by the user.
The 96-bit unique device identifier can also be read in single bytes/half-words/words in
different ways and then be concatenated using a custom algorithm.

Base address: 0x1FFF 7A10
Address offset: 0x00
Read only = 0xXXXX XXXX where X is factory-programmed
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

r

r

UID(31:0)
r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

Bits 31:0 UID(31:0): X and Y coordinates on the wafer.

Address offset: 0x04
Read only = 0xXXXX XXXX where X is factory-programmed
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

r

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

UID(63:48)

UID(47:32)
r

r

1712/1747

r

r

r

r

r

r

r

RM0090 Rev 17

RM0090

Device electronic signature

Bits 31:8 UID(63:40): LOT_NUM[23:0]
Lot number (ASCII encoded).
Bits 7:0 UID(39:32): WAF_NUM[7:0]
Wafer number (ASCII encoded).

Address offset: 0x08
Read only = 0xXXXX XXXX where X is factory-programmed
31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

UID(95:80)
r

r

r

r

r

r

r

15

14

13

12

11

10

9

r

r

r

r

r

r

r

r

r

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

UID(79:64)
r

r

r

r

r

r

r

r

r

Bits 31:0 UID(95:64): LOT_NUM[55:24]
Lot number (ASCII encoded).

39.2

Flash size
Base address: 0x1FFF 7A22
Address offset: 0x00
Read only = 0xXXXX where X is factory-programmed

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

r

r

r

r

r

r

r

r

F_SIZE
r

r

r

r

r

r

r

r

Bits 15:0 F_ID(15:0): Flash memory size
This bitfield indicates the size of the device Flash memory expressed in Kbytes.
As an example, 0x0400 corresponds to 1024 Kbytes.

RM0090 Rev 17

1713/1747
1713

Revision history

40

RM0090

Revision history
Table 315. Document revision history
Date

Version

15-Sep-2011

1

Changes
Initial release.
Updated reference documents and added Table 1: Applicable products on cover
page.
MEMORY:
Updated Section 2: Memory and bus architecture.
PWR:
Updated VDDA and VREF+ decoupling capacitor in Figure 7: Power supply
overview.
Updated case of no external battery in Section 5.1.2: Battery backup domain.
VOSRDY bit changed to read-only in Section 5.4.3: PWR power control/status
register (PWR_CSR).
Removed VDDA in Section 5.2.3: Programmable voltage detector (PVD) and
remove VDDA in PVDO bit description (Section 5.4.3: PWR power control/status
register (PWR_CSR)).

19-Oct-2012

2

RCC:
Updated Figure 20: Simplified diagram of the reset circuit and minimum reset

pulse duration guaranteed by pulse generator restricted to internal reset
sources.

GPIOs:
Updated Section 8.3.1: General-purpose I/O (GPIO).
DMA:
Updated direct mode description in Section 10.2: DMA main features.
Updated direct mode description in Section : Memory-to-peripheral mode, and
Section 10.3.12: FIFO/: Direct mode.
Updated register access in Section 10.5: DMA registers.
Modified Stream2 /Channel 2 in Table 42: DMA1 request mapping.
Added note related to EN bit in Section 10.5.5: DMA stream x configuration register
(DMA_SxCR) (x = 0..7). Updated definition of NDT[15:0] bits in Section 10.5.6:
DMA stream x number of data register (DMA_SxNDTR) (x = 0..7).
Interrupts:
Updated number of maskable interrupts to 82 in Section 12.1.1: NVIC featuress.
Updated Section 12.2: External interrupt/event controller (EXTI).

1714/1747

RM0090 Rev 17

RM0090

Revision history
Table 315. Document revision history (continued)

Date

Version

Changes
ADC:
Changed ADCCLK frequency to 30 MHz in Section 13.5: Channel-wise
programmable sampling timee.
Added recovery from ADC sequence in Section 13.8.1: Using the DMA and
Section 13.8.2: Managing a sequence of conversions without using the DMA.
Updated AWDIE in Section 13.13.2: ADC control register 1 (ADC_CR1). Added
read and write access in Section 13.13: ADC registers.

19-Oct-2012

Advanced control timers (TIM1 and TIM8):
Updated 16-bit prescaler range in Section 17.2: TIM1 and TIM8 main features.
Updated OC1 block diagram in Figure 114: Output stage of capture/compare
channel (channel 1 to 3).
Updated update event generation in Upcounting mode and Downcounting mode in
2
Section 17.3.2: Counter modes and Section 17.3.3: Repetition counter.
(continued)
Updated bits that control the dead-time generation in Section 17.3.11:
Complementary outputs and dead-time insertion.
Updated ways to generate a break in Section 17.3.12: Using the break function.
Changed OCxREF to ETR in the example given in Section 17.3.13: Clearing the
OCxREF signal on an external event and changed
OCREF_CLR to ETRF in Figure 124: Clearing TIMx OCxREF.
Updated configuration for example of counter operation in encoder
interface mode in Section 17.3.16: Encoder interface mode.
Added register access in Section 17.4: TIM1 and TIM8 registers.
Changed definition of ARR[15:0] bits in Section 17.4.12: TIM1 and TIM8 autoreload register (TIMx_ARR).
Updated BKE definition in Section 17.4.18: TIM1 and TIM8 break and dead-time
register (TIMx_BDTR).

RM0090 Rev 17

1715/1747
1741

Revision history

RM0090
Table 315. Document revision history (continued)

Date

19-Oct-2012

Version

2
(continued)

Changes
General purpose timers (TIM2 to TIM5):
Removed all references to “repetition counter”.
Added Figure 134: General-purpose timer block diagram.
Updated 16-bit prescaler range in Section 18.2: TIM2 to TIM5 main features.
External clock mode 2 ETR restricted to TIM2 to TIM4 in
Section 18.3.3: Clock selection and Section 18.3.6: PWM input mode.
Updated Section 18.3.9: PWM mode and Section 18.3.11: Clearing the OCxREF
signal on an external event.
Updated Figure 174: Master/Slave timer example to change ITR1 to
ITR0.
Updated read and write access to registers in Section 18.4: TIM2 to TIM5
registerss.
Restored bits 15 to 8 of TIMx_SMCR as well as Table 98: TIMx internal trigger
connection in Section 14.4.3.
Removed note 1 related to OC1M bits in Section 18.4.13: TIMx capture/compare
register 1 (TIMx_CCR1).
Updated TIMx_CCER bit description for TIM2 to TIM5 in
Section 18.4.9: TIMx capture/compare enable register (TIMx_CCER).
General purpose timers (TIM9 to TIM14):
Updated 16-bit prescaler range in Section 19.2.1: TIM9/TIM12 main features and
Section 19.2.2: TIM10/TIM11 and TIM13/TIM14 main features.
Updated Figure 181: General-purpose timer block diagram (TIM10/11/13/14)) to
remove TRGO trigger controller output.
Added register access in Section 19.4: TIM9 and TIM12 registers
and Section 19.5: TIM10/11/13/14 registers.
Basic timers (TIM6 and TIM7):
Removed all references to “repetition counter”.
Updated 16-bit prescaler range in Section 20.2: TIM6 and TIM7 main features.
HASH:
Updated Section 25.3.1: Duration of the processing.
RNG:
Updated Section 24.1: RNG introduction.

1716/1747

RM0090 Rev 17

RM0090

Revision history
Table 315. Document revision history (continued)

Date

Version

Changes
RTC:
Updated Figure 237: RTC block diagram.
Added formula to compute fck_apre in Figure 26.3.1: Clock and prescalers.
Updated Section 26.3.9: RTC reference clock detection.
Updated Section : RTC register write protection.
Added RTC_SSR shadow register in Section 26.3.6: Reading the calendar.
Updated description of DC[4:0] bits in Section 26.6.7: RTC calibration register
(RTC_CALIBR).
Renamed RTC_BKxR into RTC_BKPxR in Table 121: RTC register map and reset
values.
Added power-on reset value and changed reset value to system
reset value in Section 26.6.11: RTC sub second register (RTC_SSR).
Updated definition of ALARMOUTTYPE in Section 26.6.17: RTC tamper and
alternate function configuration register (RTC_TAFCR).

19-Oct-2012

I2C:
Modified Section 27.3.8: DMA requests.
Updated bit 14 description in Section 27.6.3: I2C Own address register 1
(I2C_OAR1)).
2
Updated definition of PE bit and note related to SWRST bit; moved
(continued) note related to STOP bit to the whole register in Section 27.6.1: I2C Control register
1 (I2C_CR1).
USART:
Section 30.6.6: Control register 3 (USART_CR3)): removed notes
related to UART5 in DMAT and DMAR description.
Updated TTable 142: Error calculation for programmed baud rates at fPCLK = 42
MHz or fPCLK = 84 Hz, oversampling by 16 and
Table 143: Error calculation for programmed baud rates at fPCLK = 42 MHz or
fPCLK = 84 MHz, oversampling by 8.
SPI/I2S:
Updated Section 28.1: SPI introduction.
Changed I2S simplex communication/mode to half-duplex communication/mode.
Updated flags in reception/transmission modes in Section 28.2.2: I2S features.
Added Frame error flag in Table 128: I2S interrupt requests.
Added register access in Section 28.5: SPI and I2S registers.
Updated ERRIE definition in Section 28.5.2: SPI control register 2 (SPI_CR2).
Renamed TIFRFE to FRE and definition updated in Section 28.5.3: SPI status
register (SPI_SR).

RM0090 Rev 17

1717/1747
1741

Revision history

RM0090
Table 315. Document revision history (continued)

Date

Version

Changes
SDIO:
Updated value and description for bits [45:40] and [7:1] in Table 176: R4 response.
Updated value at bits [45:40] in Table 178: R5 response.
CAN:
Updated Figure 335: Dual CAN block diagram.
Modified definition of CAN2SB bits in Section : CAN filter master register
(CAN_FMR).
Added register access in Section 32.9: CAN registers
ETHERNET:
Updated standard for precision networked clock synchronization in Section 33.1:
Ethernet introduction and Section 33.2.1: MAC core features.
Updated CR bit definition in Section : Ethernet MAC MII address register
(ETH_MACMIIAR).
Replace RTPR by PM bit in Table 192: Source address filtering.

19-Oct-2012

USB OTG FS
Updated remote wakeup signaling bit and the resume
2
interrupt in Section : Suspended state.
(continued)
Added peripheral register access in Section 34.16: OTG_FS control and status
registerss.
Updated INEPTXSA description in OTG_FS_DIEPTXFx.
Changed PHYSEL from bit 7 to bit 6 of the OTG_FS_GUSBCFG
register.
USB OTG HS
Updated remote wakeup signaling bit and the resume
interrupt in Section : Suspended state.
Added peripheral register access in Section 35.12: OTG_HS control and status
registers.
Updated OTG_HS_CID reset value.
Updated INEPTXSA description in OTG_HS_DIEPTXFx.
Updated FSLSPCS for LS host mode, added PHYSEL in Section : OTG_HS host
configuration register (OTG_HS_HCFG).
Renamed PHYSEL into PHSEL and changed from bit 7 to bit 6 of
the OTG_HS_GUSBCFG register.
Updated OTG_HS_DIEPEACHMSK1 and OTG_HS_DOEPEACHMSK1 reset
values.

1718/1747

RM0090 Rev 17

RM0090

Revision history
Table 315. Document revision history (continued)

Date

19-Oct-2012

Version

Changes

FSMC:
Updated step b) in Section 36.3.1: Supported memories and transactions.
Updated Table 196: FSMC_BTRx bit fields.
Changed Clock divide ration min in Table 246: Programmable NAND/PC Card
access parameters.
Updated case of synchronous accesses in Section 36.5: NOR Flash/PSRAM
controller.
Changed minimum value for ADDSET to 0 in Table 203, Table 206, Table 207,
Table 209, and Table 210.
Move note from Figure 437: Mode1 write accesses and Figure 436: Mode1 read
accesses. Move note from Figure 439: ModeA write accesses to Figure 438:
ModeA read accesses.
Updated Section : WAIT management in asynchronous accesses.
Added register access in Section 36.5.6: NOR/PSRAM control registers and
Section 36.6.2: NAND Flash / PC Card supported memories and transactions.
Removed caution note in Section 36.6.1: External memory interface signalss.
Updated Table 249: 16-bit PC Card.
2
(continued) Updated step 3 in Section 36.6.4: NAND Flash operations.
Updated Figure 455: Access to non ‘CE don’t care’ NAND-Flash and note below in
Section 36.6.5: NAND Flash prewait functionality.
Updated access to I/O Space in Section 36.6.7: PC Card/CompactFlash
operationss. Updated Table 251: 16-bit PC-Card signals and access type. Updated
BUSTURN bit definition in Section : SRAM/NOR-Flash chip-select timing registers
1..4 (FSMC_BTR1..4)). Changed bits 16 to 19 to BUSTURN in Section :
SRAM/NOR-Flash write timing registers 1..4 (FSMC_BWTR1..4)
DEBUG:
Updated Section 38.4.3: Internal pull-up and pull-down on JTAG pins.
Electronic signature
Updated Section 39: Device electronic signature introduction.
Updated REV_ID[15:0] to add revision Z in Section 39.1: Unique device ID register
(96 bits).
Updated address and example in Section 39.2: Flash size.

RM0090 Rev 17

1719/1747
1741

Revision history

RM0090
Table 315. Document revision history (continued)

Date

13-Nov-2012

1720/1747

Version

3

Changes
Added STM32F42x and STM32F43x devices.
Removed reference du Flash programming manual on cover page. Added
Section 2.3.2: Flash memory overview and Section 3: Embedded Flash memory
interface.
Change RTC_50Hz into RTC_REFIN in Section 8.3.2: I/O pin multiplexer and
mapping. Modified RTC alternate function naming in Section 8: General-purpose
I/Os (GPIO) and Section 26: Real-time clock (RTC).
Updated max. input frequency in Section 26.3.1: Clock and prescalers.
Changed bit access type from ‘rw’ to ‘w’ and bit description updated in
Section 10.5.3: DMA low interrupt flag clear register (DMA_LIFCR) and
Section 10.5.4: DMA high interrupt flag clear register (DMA_HIFCR).
Updated Figure 18: Frequency measurement with TIM5 in Input capture mode.
Updated Section : Signals synchronization in Section 36: Flexible static memory
controller (FSMC)
Section 34: USB on-the-go full-speed (OTG_FS): updated Section Figure 389.:
USB host-only connection, Section : VBUS valid, and Section : Host detection of a
peripheral connection.
Section 35: USB on-the-go high-speed (OTG_HS): updated Section : VBUS valid,
and Section : Detection of peripheral connection by the host.

RM0090 Rev 17

RM0090

Revision history
Table 315. Document revision history (continued)

Date

Version

Changes
Updated Section 2: Memory and bus architecture.
Updated Figure 1: System architecture for STM32F405xx/07xx and
STM32F415xx/17xx devices, and Figure 1: System architecture for
STM32F405xx/07xx and STM32F415xx/17xx devices. Updated Table 4: Memory
mapping vs. Boot mode/physical remap. Updated Figure 5: Sequential 32-bit
instruction execution. removed note 1 from Table 12: Program/erase parallelism.
PWR:
Updated Figure 7: Power supply overview.
Updated Section 5.1.3: Voltage regulator.
Added ADCDC1 bit in Section 5.5.1: PWR power control register (PWR_CR) for
STM32F42xxx and STM32F43xxx.

19-Feb-2013

4

SYSCFG:
Added ADCxDC2 bit in Section 8.2.3: SYSCFG peripheral mode configuration
register (SYSCFG_PMC) for STM32F42xxx and STM32F43xxx.
ADC:
Updated Section 13.9.3: Interleaved mode, Section 13.9.4: Alternate trigger mode,
and Section 13.9.5: Combined regular/injected simultaneous mode to describe
case of interrupted conversion.
Updated Section : Temperature sensor, VREFINT and VBAT internal channels,
Section 13.10: Temperature sensor, and Section 13.11: Battery charge monitoring.
RTC:
Updated BKP[31:0] bit description in Section 26.6.20: RTC backup registers
(RTC_BKPxR).
I2C:
Updated Section 27.3.5: Programmable noise filter.

RM0090 Rev 17

1721/1747
1741

Revision history

RM0090
Table 315. Document revision history (continued)

Date

19-Feb-2013

Version

Changes

FSMC:
Updated write FIFO size in Section 36.1: FSMC main features.
Updated Figure 434: FSMC block diagram.
Updated Section 36.5.4: NOR Flash/PSRAM controller asynchronous transactions.
Modified differences between Mode B and mode 1 in Section : Mode 2/B - NOR
Flash.
Modified differences between Mode C and mode 1 in Section : Mode C - NOR
Flash - OE toggling.
Modified differences between Mode D and mode 1 in Section : Mode D asynchronous access with extended address.
4
Updated
NWAIT signal in Figure 449: Asynchronous wait during a read access,
(continued)
Figure 450: Asynchronous wait during a write access, Figure 451: Wait
configurations, Figure 452: Synchronous multiplexed read mode - NOR, PSRAM
(CRAM), and Figure 453: Synchronous multiplexed write mode - PSRAM (CRAM).
Updated Table 195 to Table 214.
Updated Section : SRAM/NOR-Flash chip-select control registers 1..4
(FSMC_BCR1..4).
DEBUG
Updated Figure 485: Block diagram of STM32 MCU and Cortex®-M4 with FPUlevel debug support.

1722/1747

RM0090 Rev 17

RM0090

Revision history
Table 315. Document revision history (continued)

Date

Version

Changes
Added STM32F429xx and STM32F439xx part numbers.
Replaced FSMC by FMC added Chrom-ART Accelerator, LCD-TFT and SAI
interface.
Updated Figure 2: System architecture for STM32F42xxx and STM32F43xxx
devices.
PWR:
Updated Section 5.2.2: Brownout reset (BOR).
Added note related to CSS enabling in Entering Stop mode sections in
Section 5.3.4: Stop mode (STM32F405xx/07xx and STM32F415xx/17xx) and
Section 5.3.5: Stop mode (STM32F42xxx and STM32F43xxx). Updated Stop mode
entry in Table 27 and Table 29.
Updated WUF bit defienition in PWR_CSR registers. Changed CWUF and CSBF
access type to ‘w’ in PWR_CR register.
RCC: Updated LSEBYP bit definition in RCC_BDCR register.
GPIOs:
Updated description of OSPEEDR bits. Removed frequency value in description of
OSPEEDR bits.Corrected typos: "IDRy[15:0]" replaced with "IDRy" in "GPIOx_IDR"
register, "ODRy[15:0]" replaced with "ODRy" in "GPIOx_ODR" register and
"OTy[1:0]" replaced with "OTy" in "GPIOx_OTYPER" register.

15-Sep-2013

5
DCMI: Updated Section 15.4: DCMI clocks.
IWDG: Corrected Figure 213: Independent watchdog block diagram.
RTC:
Replaced all occurrences of “power-on reset” with “backup domain reset”. Added
caution note under Table 121: RTC register map and reset values. Changed SHPF
bit type to ‘r’ in Section 26.6.4: RTC initialization and status register (RTC_ISR)..
SPI: Updated definition of ERRIE bit in Section 28.5.2: SPI control register 2
(SPI_CR2).
UART:
Updated Section 30.3.8: LIN (local interconnection network) mode.
Removed note in Section 30.3.13: Continuous communication using DMA.
ETHERNET:
Modified ETH_MACA0HR (and ETH_DMABMR reset values.
Updated definitions of TSTS bit in ETH_MACSR, and TSTTR in ETH_PTPTSSR.

RM0090 Rev 17

1723/1747
1741

Revision history

RM0090
Table 315. Document revision history (continued)

Date

Version

Changes
USB OTG-FS:
Removed note related to VDD range limitation below Figure 387: OTG A-B device
connection and Figure 388: USB peripheral-only connection.

15-Sep-2013

FSMC:
Updated Table 229, Table 232, Table 235, Table 239.
Replaced all occurences of DATALAT by DATLAT and SRAM/CRAM by
SRAM/PSRAM in the whole section.
Updated Section 36.1: FSMC main features. Changed bits 27 to 20 of
FSMC_BWTR1..4 to reserved.
Updated Section 36.6.7: PC Card/CompactFlash operations.
Updated WREN bit in Table 231, Table 232, Table 233, Table 236, Table 239,
5
Table 242, Table 245, and Table 249.
(continued) Updated Section 36.5.4: NOR Flash/PSRAM controller asynchronous transactions,
Section : SRAM/NOR-Flash chip-select control registers 1..4 (FSMC_BCR1..4),
Section : SRAM/NOR-Flash chip-select timing registers 1..4 (FSMC_BTR1..4) and
Section : SRAM/NOR-Flash write timing registers 1..4 (FSMC_BWTR1..4).
Updated definition of PWID in Section : PC Card/NAND Flash control registers 2..4
(FSMC_PCR2..4).
FMC:
Updated TRDC definition in Section : SDRAM Timing registers 1,2
(FMC_SDTR1,2).
DEBUG: updated Figure 487: JTAG TAP connections.

1724/1747

RM0090 Rev 17

RM0090

Revision history
Table 315. Document revision history (continued)

Date

Version

Changes
Added note related to over-drive mode unavailable in 1.8 to 2.1 V VDD range

in Section 3.5.1: Relation between CPU clock frequency and Flash
memory read time.
Updated maximum CPU frequency in Section 3.5.2: Adaptive real-time
memory accelerator (ART Accelerator™).
PWR:
Updated Run mode/ over-drive mode in Section 5.1.4: Voltage regulator for
STM32F42xxx and STM32F43xxx.
RCC for STM32F42/43xx:
Changed APB1/2 and AHB maximum frequencies.xw
GPIOs:
Updated Figure 27: Selecting an alternate function on STM32F42xxx and
STM32F43xxx.

03-Feb-2014

6

DMA:
Updated Section 10.3.7: Pointer incrementation and Section 10.3.11: Single and
burst transfers..
INTERRUPTS AND EVENTS:
Updated Table 62: Vector table for STM32F42xxx and STM32F43xxx.
ADC:
Updated Section 13.3.10: Discontinuous mode/Section : Regular group.
DCMI:
Updated Section 15.5.2: DCMI physical interface.
LTDC:
Updated resolution in note below Figure 82: LCD-TFT Synchronous timings.
TIM1 and 8:
Added note related to IC1F in Section 17.4.7: TIM1 and TIM8 capture/compare
mode register 1 (TIMx_CCMR1).
TIM2 to 5:
Updated note related to IC1F in Section 18.4.7: TIMx capture/compare mode
register 1 (TIMx_CCMR1).

RM0090 Rev 17

1725/1747
1741

Revision history

RM0090
Table 315. Document revision history (continued)

Date

Version

Changes
TIM9 to 14:
Updated note related to IC1F in Section 19.5.5: TIM10/11/13/14 capture/compare
mode register 1 (TIMx_CCMR1).
RTC:
Updated Section 26.3.11: RTC smooth digital calibration.
Changed ALRBIE to ALRBE (bit 9) in Section 26.6.3: RTC control register
(RTC_CR).
I2C:
Introduced Sm (standard mode) and Fm (fast mode) acronyms.
FSMC:
Updated BUSTURN definition in Table 245: FSMC_BTRx bit fields.

03-Feb-2014

1726/1747

6
(continued) FMC:
Added Mobile LPSDR SDRAM.
Updated Section : SDRAM initialization and Section : SDRAM controller read cycle
and Figure 476: NAND Flash/PC Card controller waveforms for common memory
access.
Updated Section : SRAM/NOR-Flash chip-select control registers 1..4
(FMC_BCR1..4), Section : SRAM/NOR-Flash chip-select timing registers 1..4
(FMC_BTR1..4), Section : SRAM/NOR-Flash write timing registers 1..4
(FMC_BWTR1..4), Section : SDRAM Timing registers 1,2 (FMC_SDTR1,2) and
Section : SDRAM Refresh Timer register (FMC_SDRTR).
Removed mention “default valeur after reset” in Section : Common memory space
timing register 2..4 (FMC_PMEM2..4), Section : Attribute memory space timing
registers 2..4 (FMC_PATT2..4), and Section : I/O space timing register 4
(FMC_PIO4).
Updated BUSTURN definition in Table 288: FMC_BTRx bit fields.
Updated REV_ID bits in Section 38.6.1: MCU device ID code..

RM0090 Rev 17

RM0090

Revision history
Table 315. Document revision history (continued)

Date

Version

Changes
Embedded Flash memory interface:
Updated Section : Physical remap in STM32F42xxx and STM32F43xxx. Updated
bank 2 selection in Section 2.4: Boot configuration. Updated notes related to MERx
and SER bits in Section : Mass Erase. Updated Section 3.7.5: Proprietary code
readout protection (PCROP). Updated FLASH_OPTCR register reset value for
STM32F42/43xx in Section 3.9.10: Flash option control register (FLASH_OPTCR)
for STM32F42xxx and STM32F43xxx and Section 3.9.11: Flash option control
register (FLASH_OPTCR1) for STM32F42xxx and STM32F43xxx.
RCC (STM32F42/43xx):
Updated PPLN caution note in Section 6.3.2: RCC PLL configuration register
(RCC_PLLCFGR)
SYSCFG
Updated MEM_MODE in Section 9.3.1: SYSCFG memory remap register
(SYSCFG_MEMRMP)
LTDC:
Changed resolution do XGA (1024x768) in Section 16.2: LTDC main features,
Section 16.4.1: LTDC Global configuration parameters, and updated
Section 16.7.3: LTDC Active Width Configuration Register (LTDC_AWCR).

15-May-2014

7

RTC
Added note in Section 26.3.14: Calibration clock output.
TIMER 1/8:
Removed note related to IC1F bits in Section 17.4.7: TIM1 and TIM8
capture/compare mode register 1 (TIMx_CCMR1),
TIM2 to 5:
Replaced IC2S by CC2S.
Updated Figure 161: Output stage of capture/compare channel (channel 1).
Removed note related to IC1F bits in Section 18.4.7: TIMx capture/compare mode
register 1 (TIMx_CCMR1).
TIM9 to 14:
Removed note related to IC1F bits in Section 19.5.5: TIM10/11/13/14
capture/compare mode register 1 (TIMx_CCMR1).
USB OTG-HS:
Updated DSPD definition in Section : OTG_HS device configuration register
(OTG_HS_DCFG).
FSMC
Updated DATLAT bits definition in Section : SRAM/NOR-Flash chip-select timing
registers 1..4 (FSMC_BTR1..4).

RM0090 Rev 17

1727/1747
1741

Revision history

RM0090
Table 315. Document revision history (continued)

Date

15-May-2014

Version

Changes

FMC
Updated Figure 474: Synchronous multiplexed read mode waveforms - NOR,
PSRAM (CRAM). Updated DATLAT bits definition in Section : SRAM/NOR-Flash
chip-select timing registers 1..4 (FMC_BTR1..4).
7
(continued) Updated FMC_BWTRx register address offsets in Table 297: FMC register map.
DEBUG
Added revision code ‘3’ in Section : DBGMCU_IDCODE.

1728/1747

RM0090 Rev 17

RM0090

Revision history
Table 315. Document revision history (continued)

Date

Version

Changes
Memory and bus architecture:
Updated Table 3: Memory mapping vs. Boot mode/physical remap in
STM32F405xx/07xx and STM32F415xx/17xx and Table 4: Memory mapping vs.
Boot mode/physical remap in STM32F42xxx and STM32F43xxx.
RCC (STM32F40/41xx) and RCC (STM32F42/43xx):
Removed all references to Flash programming manual. Changed
RCC_AHB1LPENR, RCC_APB1LPENR, RCC_APB2LPENR, RCC_PLLI2SCFGR
and RCC_APB2LPENR reset values.
Updated access type to “r” for bits 24 to 31 in RCC_CSR.
GPIOs:
Updated Figure 27: Selecting an alternate function on STM32F42xxx and
STM32F43xxx.
IWDG
Update note in Table 107: Min/max IWDG timeout period (in ms) at 32 kHz (LSI).
CRYPTO and HASH
Removed STM32F405/407xx and STM32F42xx from the whole sections.

14-Oct-2014

8

Removed STM32F405/407xx and STM32F42xx from the whole section.
TIM10/11/13/14
Added TIMx_DIER description in Section 19.5: TIM10/11/13/14 registers.
ETHERNET:
Updated Table 187: Clock range.
USB OTG FS:
Removed TRDT formula in Section 34.17.7: Worst case response time and added
Table 203: TRDT values.
USB OTG HS:
Removed TRDT formula in Section 35.13.8: Worst case response time and added
Table 213: TRDT values.
FSMC:
Updated EXTMOD definition in Section : SRAM/NOR-Flash chip-select control
registers 1..4 (FSMC_BCR1..4).
Updated ADDSET definition in Section : SRAM/NOR-Flash chip-select timing
registers 1..4 (FSMC_BTR1..4) and Section : SRAM/NOR-Flash write timing
registers 1..4 (FSMC_BWTR1..4).

RM0090 Rev 17

1729/1747
1741

Revision history

RM0090
Table 315. Document revision history (continued)

Date

14-Oct-2014

1730/1747

Version

Changes

FMC:
Modified step 7 in Section : SDRAM initialization.
Modified SDRAM refresh rate equations and example in Section : SDRAM Refresh
Timer register (FMC_SDRTR) and updated definition of COUNT bits.
8
(continued) Updated EXTMOD definition in Section : SRAM/NOR-Flash chip-select control
registers 1..4 (FMC_BCR1..4).
Updated ADDSET definition in Section : SRAM/NOR-Flash chip-select timing
registers 1..4 (FMC_BTR1..4) and Section : SRAM/NOR-Flash write timing
registers 1..4 (FMC_BWTR1..4).

RM0090 Rev 17

RM0090

Revision history
Table 315. Document revision history (continued)

Date

Version

Changes
PWR:
Updated Section 5.1.2: Battery backup domain.
Updated Table 23: Low-power mode summary to add Return from ISR as entry
condition.
Added Section : Entering low-power mode and Section : Exiting low-power mode.
Updated Section : Entering Sleep mode, Section : Exiting Sleep mode, Table 24:
Sleep-now entry and exit and Table 25: Sleep-on-exit entry and exit.
Updated Section : Entering Stop mode (for STM32F405xx/07xx and
STM32F415xx/17xx), Section : Exiting Stop mode (for STM32F405xx/07xx and
STM32F415xx/17xx) and Table 27: Stop mode entry and exit (for
STM32F405xx/07xx and STM32F415xx/17xx). Updated Section : Entering Stop
mode (STM32F42xxx and STM32F43xxx), Section : Exiting Stop mode
(STM32F42xxx and STM32F43xxx) and Table 29: Stop mode entry and exit
(STM32F42xxx and STM32F43xxx).
Updated Section : Entering Standby mode, Section : Exiting Standby mode and
Table 30: Standby mode entry and exit.
RCC:
Updated bits 24 to 31 access type in Section 7.3.21: RCC clock control & status
register (RCC_CSR).

16-Mar-2015

9

GPIOs:
Added port A reset value in Section 8.4.3: GPIO port output speed register
(GPIOx_OSPEEDR) (x = A..I/J/K).
DMA:
Update FTH[1:0] description in Section 10.5.10: DMA stream x FIFO control
register (DMA_SxFCR) (x = 0..7).
TIM2/5:
Register format changed to 32 bits instead of 16 in Section 18.4.10: TIMx counter
(TIMx_CNT) and Section 18.4.12: TIMx auto-reload register (TIMx_ARR).
TIM9 to 14:
Updated Table 101: TIMx internal trigger connection
WWDG:
Updated Figure 214: Watchdog block diagram and Section 22.4: How to program
the watchdog timeout.
Updated Figure 215: Window watchdog timing diagram
RNG:
Replaced PLL48CLK by RNG_CLK in the whole section.

RM0090 Rev 17

1731/1747
1741

Revision history

RM0090
Table 315. Document revision history (continued)

Date

Version

Changes
I2C2:
Updated FREQ[5:0] description in Section 27.6.2: I2C Control register 2 (I2C_CR2).
USART:
Removed note related to RXNEIE in Section : Reception using DMA
FSMC:
Updated Figure 474: Synchronous multiplexed read mode waveforms - NOR,
PSRAM (CRAM).

16-Mar-2015

USB OTG FS
9
Updated Table 203: TRDT values
(continued)
FMC
Updated FMC_NL in Figure 456: FMC block diagram.
Updated ‘Memory wait’ and ‘Memory data bus high-z’ parameters in Table 289:
Programmable NAND Flash/PC Card access parameters.
Updated Section : Common memory space timing register 2..4 (FMC_PMEM2..4).
Updated Figure 476: NAND Flash/PC Card controller waveforms for common
memory access.
DEBUG:
Updated REV_ID[15:0) and JTAG ID code in Section 38.6.1: MCU device ID code
and Section 38.6.2: Boundary scan TAP, respectively

1732/1747

RM0090 Rev 17

RM0090

Revision history
Table 315. Document revision history (continued)

Date

Version

Changes
Embedded Flash memory interface
– Updated Section 3.7.5: Proprietary code readout protection (PCROP),
Power controller (PWR)
– Added the last sentence in Subsection: Entering low-power mode of Section 5.3:
Low-power modes,
– Added the bullet points about the interrupt in mode entry in Table 24: Sleep-now
entry and exit, Table 25: Sleep-on-exit entry and exit, Table 27: Stop mode entry
and exit (for STM32F405xx/07xx and STM32F415xx/17xx), Table 29: Stop mode
entry and exit (STM32F42xxx and STM32F43xxx)
– Added the last point to Mode entry, on return from ISR in Table 30: Standby mode
entry and exit,
– Added the note in Section: Entering sleep mode in Section 5.3.3: Sleep mode.
General-purpose I/Os (GPIO)
– Updated OSPEED[1:0] definition of GPIOx_OSPEEDR register in Section 8.4.3:
GPIO port output speed register (GPIOx_OSPEEDR) (x = A..I/J/K)

28-Jul-2015

10

LCD-TFT Controller (LTDC)
– Corrected the bit field for WHSTPOS in the second bullet point in Section:
Window in Section 16.4.2: Layer programmable parameters.
Advanced-control timers (TIM1&TIM8)
– Added the note in Section 17.3.20: Timer synchronization,
– Updated ETF[3:0] description in Section 17.4.3: TIM1 and TIM8 slave mode
control register (TIMx_SMCR),
– Updated IC1F[3:0] description in Section 17.4.7: TIM1 and TIM8
capture/compare mode register 1 (TIMx_CCMR1),
– Added the note to MMS2 bit description in Section 17.4.8: TIM1 and TIM8
capture/compare mode register 2 (TIMx_CCMR2),
– Added the note to SMS[2:0] bit description in Section 17.4.3: TIM1 and TIM8
slave mode control register (TIMx_SMCR).
General-purpose timers (TIM2 to TIM5)
– Added the note in Section 18.3.15: Timer synchronization,
– Updated SMS[2:0] description in Section 18.4.3: TIMx slave mode control
register (TIMx_SMCR),
– Added the note to MMS2 bit description in Section 18.4.2: TIMx control register 2
(TIMx_CR2),
– Added the note to SMS[2:0] bit description in Section 18.4.3: TIMx slave mode
control register (TIMx_SMCR).

RM0090 Rev 17

1733/1747
1741

Revision history

RM0090
Table 315. Document revision history (continued)

Date

Version

Changes
General-purpose timers (TIM9 to TIM14)
– Added the note in Section 19.3.12: Timer synchronization (TIM9/12),
– Added the note to MMS2 bit description,
– Added the note to SMS[2:0] bit description in Section 19.4.2: TIM9/12 slave
mode control register (TIMx_SMCR).
Window watchdog (WWDG)
– Updated.Figure 214: Watchdog block diagram
Controller area network (bxCAN)
– Replaced tCAN with tq,

28-Jul-2015

1734/1747

Flexible static memory controller (FSMC)
– Added the paragraph about Cross boundary page for Cellular RAM 1.5 in
Section 36.5.5: Synchronous transactions,
10
(Continued) – Updated MEMHIZx, MEMHOLDx, MEMSETx bit field descriptions for
FSMC_PME2..4 register in Section 36.5.5: Synchronous transactions,
– Updated ATTSET, ATTHOLD, ATTHIZ bit field descriptions for FSMC_PATT2..4
register in Section 36.5.5: Synchronous transactions,
– Updated IRS and IFS bit descriptions for FMC_SR2..4 in Section 36.5.5:
Synchronous transactions,
– Renamed ADDSET as ADDSET[3:0] and MTYP as MTYP[1:0],
– Addition of CPSIZE in FSMC_BCRx bit fields in Table 226: FSMC_BCRx bit
fields, Table 228: FSMC_BCRx bit fields, Table 231: FSMC_BCRx bit fields,
Table 234: FSMC_BCRx bit fields, Table 237: FSMC_BCRx bit fields, Table 240:
FSMC_BCRx bit fields, Table 242: FSMC_BCRx bit fields,
– Added CPIZE[2:0] in FMC_BCR1...4 registers in ,Section 36.5.6: NOR/PSRAM
control registers Section NOR/PSRAM control re
– Added CPSIZE[2:0] for FMC_BCRx registers in Section 36.6.9: FSMC register
map.

RM0090 Rev 17

RM0090

Revision history
Table 315. Document revision history (continued)

Date

28-Jul-2015

Version

Changes

Flexible memory controller (FMC)
– Added the paragraph about Cross boundary page for Cellular RAM 1.5 in
Section 37.5.5: Synchronous transactions,
– Updated BUSTURN bit field description for FMC_BTR1..4 register in
Section 37.5.6: NOR/PSRAM controller registers,
– Updated MEMHIZx, MEMHOLDx, MEMSETx bit field descriptions for
FMC_PME2..4 register in Section 37.6.8: NAND Flash/PC Card controller
registers,
– Updated ATTSET, ATTHOLD, ATTHIZ bit field descriptions for FMC_PATT2..4
register in Section 37.6.8: NAND Flash/PC Card controller registers,
– Updated IRS and IFS bit descriptions for FMC_SR2..4 in Section 37.6.8: NAND
Flash/PC Card controller registers,
10
(Continued) – Updated the section SDRAM initialization with the last item in the numbered list in
Section 37.7.5: SDRAM controller registers,
– Renamed ADDSET as ADDSET[3:0] and MTYP as MTYP[1:0],
– Addition of CPSIZE in Table 269: FMC_BCRx bit fields, Table 271: FMC_BCRx
bit fields, Table 274: FMC_BCRx bit fields, Table 277: FMC_BCRx bit fields,
Table 280: FMC_BCRx bit fields, Table 283: FMC_BCRx bit fields, Table 285:
FMC_BCRx bit fields, Table 287: FMC_BCRx bit fields,
– Added the paragraph about Cross boundary page for Cellular RAM 1.5 in
Section 37.5.5: Synchronous transactions,
– Added CPIZE[2:0] in FMC_BCR1...4 registers in Section 37.5.6: NOR/PSRAM
controller registers,
– Added CPSIZE[2:0] for FMC_BCRx registers in Section 37.8: FMC register map.

RM0090 Rev 17

1735/1747
1741

Revision history

RM0090
Table 315. Document revision history (continued)

Date

Version

Changes
Reset and clock controller (RCC)
Updated STM32F405/407/415/417xx Figure 21: Clock tree.
Updated
General purpuse I/O (GPIOs)
Changed definition of OSPEEDR bits in Section 8.4.3: GPIO port output speed
register (GPIOx_OSPEEDR) (x = A..I/J/K).
LCD-TFT display controller (LTDC):
Changed LRDC_IER into LTDC_IER in Section 16.5: LTDC interrupts.
Updated AHBP[11:0], AAV[11:0 and TOTALW[11:0 in Table 92: LTDC register map
and reset values.
Controller area network (bxCAN):
Updated Section 32.3.4: Acceptance filters and Section 32.7.4: Identifier filtering.

20-Oct-2015

11

Flexible static memory controller (FSMC)
Updated BUSTURN description in Section : SRAM/NOR-Flash write timing
registers 1..4 (FSMC_BWTR1..4) and Section : SRAM/NOR-Flash chip-select
timing registers 1..4 (FSMC_BTR1..4)
Updated note related to IRS and IFS bits in Section : FIFO status and interrupt
register 2..4 (FSMC_SR2..4).
Flexible memory controller (FMC)
Updated paragraph related to the cacheable read FIFO in Section : SDRAM
controller read cycle.
Updated BUSTURN description in Section : SRAM/NOR-Flash write timing
registers 1..4 (FMC_BWTR1..4) and Section : SRAM/NOR-Flash chip-select timing
registers 1..4 (FMC_BTR1..4).
Updated note related to IRS and IFS bits in Section : FIFO status and interrupt
register 2..4 (FMC_SR2..4).
Real-time clock (RTC2)
Updated WUCKSEL prescaler input in Figure 237: RTC block diagram.
Updated 3rd step in Section : Programming the wakeup timer.
Updated WUTWF bit definition in Section 26.6.4: RTC initialization and status
register (RTC_ISR).

1736/1747

RM0090 Rev 17

RM0090

Revision history
Table 315. Document revision history (continued)

Date

Version

Changes
Embedded Flash memory interface
Removed note related to boot from Bank 2 in Section 2.4: Boot configuration.
Updated notes in Section 3.7.3: Read protection (RDP).
Changed number of LATENCY bits in Section 3.9.2: Flash access control register
(FLASH_ACR) for STM32F42xxx and STM32F43xxx
In Table 9: 1 Mbyte dual bank Flash memory organization (STM32F42xxx and
STM32F43xxx): updated sector 19 size and option bytes (bank 2) address range.
Power control (PWR)
Removed reference to low-power mode in Section 5.1.4: Voltage regulator for
STM32F42xxx and STM32F43xxx, Section : Entering Stop mode (STM32F42xxx
and STM32F43xxx) and Section : Exiting Stop mode (STM32F42xxx and
STM32F43xxx).
Analog-to-digital converter (ADC)
Added note related to ADC_HTR and ADC_LTR register programming in
Section 13.13.7: ADC watchdog higher threshold register (ADC_HTR) and
Section 13.13.8: ADC watchdog lower threshold register (ADC_LTR).

17-May-2016

12

Chrom-Art Accelerator™ controller (DMA2D)
Updated Section 11.3.12: DMA2D transfer control (start, suspend, abort and
completion).
Section 11.5.8: DMA2D foreground PFC control register (DMA2D_FGPFCCR):
updated START bit access type
Section 11.5.10: DMA2D background PFC control register (DMA2D_BGPFCCR):
updated START bit access and description.
LCD-TFT controller (LTDC)
Updated Section 16.3.2: LTDC reset and clocks.
Modified LCD_DE description in Table 89: LCD-TFT pins and signal interface.
Modified Section 16.7.15: LTDC Layerx Window Horizontal Position Configuration
Register (LTDC_LxWHPCR) (where x=1..2) and Section 16.7.16: LTDC Layerx
Window Vertical Position Configuration Register (LTDC_LxWVPCR) (where
x=1..2).
General-purpose timers (TIM2 to TIM5)
Updated Section 18.4.11: TIMx prescaler (TIMx_PSC).
General-purpose timers (TIM9 to TIM14)
Added OPM bit in Section 19.5.1: TIM10/11/13/14 control register 1 (TIMx_CR1).
Updated Section 19.4.9: TIM9/12 prescaler (TIMx_PSC) and Section 19.5.8:
TIM10/11/13/14 prescaler (TIMx_PSC).

RM0090 Rev 17

1737/1747
1741

Revision history

RM0090
Table 315. Document revision history (continued)

Date

Version

Changes
General-purpose timers (TIM6 and TIM7)
Updated Section 20.4.7: TIM6 and TIM7 prescaler (TIMx_PSC).
Real-time clock (RTC)
Updated conditions for running under System reset in Section 26.3.7: Resetting the
RTC.
Updated Section 26.3.14: Calibration clock output.
Added note related to TSE in Section 26.6.3: RTC control register (RTC_CR).
Updated caution note related to TAMP1TRG in Section 26.6.17: RTC tamper and
alternate function configuration register (RTC_TAFCR) register.
Universal synchronous asynchronous receiver transmitter (USART)
Replaced all occurrences of nCTS by CTS, nRTS by RTS and SCLK by CK.

17-May-2016

Flexible static memory controller (FSMC)
Updated Section 36.3: AHB interface.
Added note related to the hold phase delay below Figure 454: NAND/PC Card
controller timing for common memory access.
Updated Section 36.6.5: NAND Flash prewait functionality.
12
(continued) Updated BUSTURN description in Section : SRAM/NOR-Flash chip-select timing
registers 1..4 (FSMC_BTR1..4).
Updated MEMHOLDx in Section : Common memory space timing register 2..4
(FSMC_PMEM2..4) and ATTHOLD in Section : Attribute memory space timing
registers 2..4 (FSMC_PATT2..4).
Flexible memory controller (FMC)
Updated Section 37.3: AHB interface.
Added note related to the hold phase delay below Figure 476: NAND Flash/PC
Card controller waveforms for common memory access.
Updated Section 37.6.5: NAND Flash prewait functionality.
Updated BUSTURN description in Section : SRAM/NOR-Flash chip-select timing
registers 1..4 (FMC_BTR1..4).
Updated MEMHOLDx in Section : Common memory space timing register 2..4
(FMC_PMEM2..4) and ATTHOLD in Section : Attribute memory space timing
registers 2..4 (FMC_PATT2..4).
Debug (DBG)
Updated value to be programmed to the ETM Trace Start/stop register to enable
the trace in Section 38.15.4: ETM configuration example.

1738/1747

RM0090 Rev 17

RM0090

Revision history
Table 315. Document revision history (continued)

Date

Version

Changes
Analog-to-digital converter (ADC)
Updated DMA mode 1 and DMA mode 3 description in Section 13.9: Multi ADC
mode.
LCD-TFT controller
Updated values to be programmed to LTDC_SSCR in Section : Example of
Synchronous timings configuration
Updated Section 16.4.2: Layer programmable parameters/Windowing.
Advanced-control timers (TIM1 and TIM8)
Updated Section 17.3.21: Debug mode.
Extended Section 17.4.20: TIM1 and TIM8 DMA address for full transfer
(TIMx_DMAR) to 32 bits.
Updated Table 95: Output control bits for complementary OCx and OCxN channels
with break feature output state for MOE = 0.
Updated TIM1 and TIM8 auto-reload register (TIMx_ARR) reset value.
Updated TIMx_CCR1/2/3/4 description when CC1 channel is configured as inputs
and changed bit access type to rw/ro.

20-Sep-2016

13

General-purpose timers (TIM2 to TIM5)
Updated TIMx auto-reload register (TIMx_ARR) reset value.
Updated TIMx_CCR1/2/3/4 description when CC1 channel is configured as inputs
and changed bit access type to rw/ro.
General-purpose timers (TIM9 to TIM14)
Updated TIM9/12 auto-reload register (TIMx_ARR) and TIM10/11/13/14 autoreload register (TIMx_ARR) reset value.
Updated TIMx_CCR1 description when CC1 channel is configured as inputs and
changed bit access type to rw/ro.
Basic timers (TIM6 to TIM7)
Updated TIM6 and TIM7 auto-reload register (TIMx_ARR).
Secure digital input/output interface (SDIO)
Updated Section 31.1: SDIO main features up to 50 MHz.
Updated Section 31.3: SDIO functional description SDIO_CK description.
Updated note removing 48 MHz in Section 31.9.1: SDIO power control register
(SDIO_POWER), Section 31.9.2: SDI clock control register (SDIO_CLKCR),
Section 31.9.4: SDIO command register (SDIO_CMD) and Section 31.9.9: SDIO
data control register (SDIO_DCTRL).

RM0090 Rev 17

1739/1747
1741

Revision history

RM0090
Table 315. Document revision history (continued)

Date

Version

Changes

20-Sep-2016

FMC
Update BUSTURN bit description in Section : SRAM/NOR-Flash chip-select timing
registers 1..4 (FMC_BTR1..4) and Section : SRAM/NOR-Flash write timing
registers 1..4 (FMC_BWTR1..4).
Debug support
13
Specified behavior of timers with complementary outputs in Section 38.16.2: Debug
(continued) support for timers, watchdog, bxCAN and I2C.
Updated DBG_TIMx_STOP bit description in Section 38.16.4: Debug MCU APB1
freeze register (DBGMCU_APB1_FZ) and Section 38.16.4: Debug MCU APB1
freeze register (DBGMCU_APB1_FZ).
Electronic signature
Updated Section 39.1: Unique device ID register (96 bits).

21-Apr-2017

14

Updated:
– Section 5.5.2: PWR power control/status register (PWR_CSR) for STM32F42xxx
and STM32F43xxx
– Section 6.3.14: RCC APB2 peripheral clock enable register (RCC_APB2ENR)
– Section 14.3.5: DAC output voltage
– Section 38.6.1: MCU device ID code
– Figure 237: RTC block diagram
Deleted:
– Section 7.3.15: RCC APB2 peripheral clock enable register(RCC_APB2ENR)

15

Updated:
– Section 3.9.10: Flash option control register (FLASH_OPTCR) for STM32F42xxx
and STM32F43xxx
– OTG_FS USB configuration register (OTG_FS_GUSBCFG)
– Table 142: Error calculation for programmed baud rates at fPCLK = 42 MHz or
fPCLK = 84 Hz, oversampling by 16
– Table 143: Error calculation for programmed baud rates at fPCLK = 42 MHz or
fPCLK = 84 MHz, oversampling by 8

18-Jul-2017

1740/1747

RM0090 Rev 17

RM0090

Revision history
Table 315. Document revision history (continued)

Date

23-Apr-2018

07-Jun-2018

Version

Changes

16

Updated:
– Section 30.6.1: Status register (USART_SR)
– Section 34.16.4: Device-mode registers
– Section 34.17.6: Operational model
– Section 35.12.4: Device-mode registers
– Section 34: USB on-the-go full-speed (OTG_FS)
– Table 199: Host-mode control and status registers (CSRs)
– Table 205: OTG_FS register map and reset values
– Table 210: Device-mode control and status registers
– Table 215: OTG_HS register map and reset values
Added:
– Figure 412: SOF trigger output to TIM2 ITR1 connection
RXOLNY register changed from SPI_CR2 to SPI_CR1 in Section 28.3.4:
Configuring the SPI for half-duplex communication and Unidirectional receive-only
procedure (BIDIMODE=0 and RXONLY=1)

17

Updated:
– Figure 16: Clock tree
– Figure 21: Clock tree
– Figure 27: Selecting an alternate function on STM32F42xxx and STM32F43xxx
– Table 61: Vector table for STM32F405xx/07xx and STM32F415xx/17xx
– Table 62: Vector table for STM32F42xxx and STM32F43xxx
– Section 29.17.5: SAI xInterrupt mask register (SAI_xIM) where x is A or B
– Section 38.6.1: MCU device ID code

RM0090 Rev 17

1741/1747
1741

Index

RM0090

Index
A
ADC_CCR . . . . . . . . . . . . . . . . . . . . . . . . . . .427
ADC_CDR . . . . . . . . . . . . . . . . . . . . . . . . . . .430
ADC_CR1 . . . . . . . . . . . . . . . . . . . . . . . . . . .416
ADC_CR2 . . . . . . . . . . . . . . . . . . . . . . . . . . .418
ADC_CSR . . . . . . . . . . . . . . . . . . . . . . . . . . .426
ADC_DR . . . . . . . . . . . . . . . . . . . . . . . . . . . .425
ADC_HTR . . . . . . . . . . . . . . . . . . . . . . . . . . .421
ADC_JDRx . . . . . . . . . . . . . . . . . . . . . . . . . . .425
ADC_JOFRx . . . . . . . . . . . . . . . . . . . . . . . . .421
ADC_JSQR . . . . . . . . . . . . . . . . . . . . . . . . . .424
ADC_LTR . . . . . . . . . . . . . . . . . . . . . . . . . . . .422
ADC_SMPR1 . . . . . . . . . . . . . . . . . . . . . . . . .420
ADC_SMPR2 . . . . . . . . . . . . . . . . . . . . . . . . .420
ADC_SQR1 . . . . . . . . . . . . . . . . . . . . . . . . . .422
ADC_SQR2 . . . . . . . . . . . . . . . . . . . . . . . . . .423
ADC_SQR3 . . . . . . . . . . . . . . . . . . . . . . . . . .423
ADC_SR . . . . . . . . . . . . . . . . . . . . . . . . . . . . .415

C
CAN_BTR . . . . . . . . . . . . . . . . . . . . . . . . . .1105
CAN_ESR . . . . . . . . . . . . . . . . . . . . . . . . . .1104
CAN_FA1R . . . . . . . . . . . . . . . . . . . . . . . . .1115
CAN_FFA1R . . . . . . . . . . . . . . . . . . . . . . . .1115
CAN_FiRx . . . . . . . . . . . . . . . . . . . . . . . . . .1116
CAN_FM1R . . . . . . . . . . . . . . . . . . . . . . . . .1114
CAN_FMR . . . . . . . . . . . . . . . . . . . . . . . . . .1113
CAN_FS1R . . . . . . . . . . . . . . . . . . . . . . . . .1114
CAN_IER . . . . . . . . . . . . . . . . . . . . . . . . . . .1102
CAN_MCR . . . . . . . . . . . . . . . . . . . . . . . . . .1096
CAN_MSR . . . . . . . . . . . . . . . . . . . . . . . . . .1098
CAN_RDHxR . . . . . . . . . . . . . . . . . . . . . . . .1112
CAN_RDLxR . . . . . . . . . . . . . . . . . . . . . . . .1112
CAN_RDTxR . . . . . . . . . . . . . . . . . . . . . . . .1111
CAN_RF0R . . . . . . . . . . . . . . . . . . . . . . . . .1101
CAN_RF1R . . . . . . . . . . . . . . . . . . . . . . . . .1102
CAN_RIxR . . . . . . . . . . . . . . . . . . . . . . . . . .1110
CAN_TDHxR . . . . . . . . . . . . . . . . . . . . . . . .1109
CAN_TDLxR . . . . . . . . . . . . . . . . . . . . . . . .1109
CAN_TDTxR . . . . . . . . . . . . . . . . . . . . . . . .1108
CAN_TIxR . . . . . . . . . . . . . . . . . . . . . . . . . .1107
CAN_TSR . . . . . . . . . . . . . . . . . . . . . . . . . .1099
CRC_DR . . . . . . . . . . . . . . . . . . . . . . . . . . . .114
CRC_IDR . . . . . . . . . . . . . . . . . . . . . . . . . . . .114
CRYP_CR . . . . . . . . . . . . . . . . . . . . . . .748, 750
CRYP_DIN . . . . . . . . . . . . . . . . . . . . . . . . . . .754
CRYP_DMACR . . . . . . . . . . . . . . . . . . . . . . .756

1742/1747

CRYP_DOUT . . . . . . . . . . . . . . . . . . . . . . . . 755
CRYP_IMSCR . . . . . . . . . . . . . . . . . . . . . . . . 756
CRYP_IV0LR . . . . . . . . . . . . . . . . . . . . . . . . 760
CRYP_IV0RR . . . . . . . . . . . . . . . . . . . . . . . . 760
CRYP_IV1LR . . . . . . . . . . . . . . . . . . . . . . . . 761
CRYP_IV1RR . . . . . . . . . . . . . . . . . . . . . . . . 761
CRYP_K0LR . . . . . . . . . . . . . . . . . . . . . . . . . 758
CRYP_K0RR . . . . . . . . . . . . . . . . . . . . . . . . . 758
CRYP_K1LR . . . . . . . . . . . . . . . . . . . . . . . . . 759
CRYP_K1RR . . . . . . . . . . . . . . . . . . . . . . . . . 759
CRYP_K2LR . . . . . . . . . . . . . . . . . . . . . . . . . 759
CRYP_K2RR . . . . . . . . . . . . . . . . . . . . . . . . . 759
CRYP_K3LR . . . . . . . . . . . . . . . . . . . . . . . . . 759
CRYP_K3RR . . . . . . . . . . . . . . . . . . . . . . . . . 760
CRYP_MISR . . . . . . . . . . . . . . . . . . . . . . . . . 757
CRYP_RISR . . . . . . . . . . . . . . . . . . . . . . . . . 757
CRYP_SR . . . . . . . . . . . . . . . . . . . . . . . . . . . 753

D
DAC_CR . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
DAC_DHR12L1 . . . . . . . . . . . . . . . . . . . . . . . 449
DAC_DHR12L2 . . . . . . . . . . . . . . . . . . . . . . . 450
DAC_DHR12LD . . . . . . . . . . . . . . . . . . . . . . 451
DAC_DHR12R1 . . . . . . . . . . . . . . . . . . . . . . 448
DAC_DHR12R2 . . . . . . . . . . . . . . . . . . . . . . 450
DAC_DHR12RD . . . . . . . . . . . . . . . . . . . . . . 451
DAC_DHR8R1 . . . . . . . . . . . . . . . . . . . . . . . 449
DAC_DHR8R2 . . . . . . . . . . . . . . . . . . . . . . . 450
DAC_DHR8RD . . . . . . . . . . . . . . . . . . . . . . . 452
DAC_DOR1 . . . . . . . . . . . . . . . . . . . . . . . . . . 452
DAC_DOR2 . . . . . . . . . . . . . . . . . . . . . . . . . . 452
DAC_SR . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
DAC_SWTRIGR . . . . . . . . . . . . . . . . . . . . . . 448
DBGMCU_APB1_FZ . . . . . . . . . . . . . . . . . . 1703
DBGMCU_APB2_FZ . . . . . . . . . . . . . . . . . . 1704
DBGMCU_CR . . . . . . . . . . . . . . . . . . . . . . . 1701
DBGMCU_IDCODE . . . . . . . . . . . . . . . . . . 1688
DCMI_CR . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
DCMI_CWSIZE . . . . . . . . . . . . . . . . . . . . . . . 477
DCMI_CWSTRT . . . . . . . . . . . . . . . . . . . . . . 477
DCMI_DR . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
DCMI_ESCR . . . . . . . . . . . . . . . . . . . . . . . . . 475
DCMI_ESUR . . . . . . . . . . . . . . . . . . . . . . . . . 476
DCMI_ICR . . . . . . . . . . . . . . . . . . . . . . . . . . . 474
DCMI_IER . . . . . . . . . . . . . . . . . . . . . . . . . . . 472
DCMI_MIS . . . . . . . . . . . . . . . . . . . . . . . . . . . 473
DCMI_RIS . . . . . . . . . . . . . . . . . . . . . . . . . . . 471
DCMI_SR . . . . . . . . . . . . . . . . . . . . . . . . . . . 470

RM0090 Rev 17

RM0090

Index

DMA_HIFCR . . . . . . . . . . . . . . . . . . . . . . . . .327
DMA_HISR . . . . . . . . . . . . . . . . . . . . . . . . . . .326
DMA_LIFCR . . . . . . . . . . . . . . . . . . . . . . . . . .327
DMA_LISR . . . . . . . . . . . . . . . . . . . . . . . . . . .325
DMA_SxCR . . . . . . . . . . . . . . . . . . . . . . . . . .328
DMA_SxFCR . . . . . . . . . . . . . . . . . . . . . . . . .333
DMA_SxM0AR . . . . . . . . . . . . . . . . . . . . . . . .332
DMA_SxM1AR . . . . . . . . . . . . . . . . . . . . . . . .332
DMA_SxNDTR . . . . . . . . . . . . . . . . . . . . . . . .331
DMA_SxPAR . . . . . . . . . . . . . . . . . . . . . . . . .332

E
ETH_DMABMR . . . . . . . . . . . . . . . . . . . . . .1223
ETH_DMACHRBAR . . . . . . . . . . . . . . . . . . .1237
ETH_DMACHRDR . . . . . . . . . . . . . . . . . . . .1236
ETH_DMACHTBAR . . . . . . . . . . . . . . . . . . .1236
ETH_DMACHTDR . . . . . . . . . . . . . . . . . . . .1236
ETH_DMAIER . . . . . . . . . . . . . . . . . . . . . . .1233
ETH_DMAMFBOCR . . . . . . . . . . . . . . . . . .1235
ETH_DMAOMR . . . . . . . . . . . . . . . . . . . . . .1229
ETH_DMARDLAR . . . . . . . . . . . . . . . . . . . .1225
ETH_DMARPDR . . . . . . . . . . . . . . . . . . . . .1225
ETH_DMARSWTR . . . . . . . . . . . . . . . . . . . .1235
ETH_DMASR . . . . . . . . . . . . . . . . . . . . . . . .1226
ETH_DMATDLAR . . . . . . . . . . . . . . . . . . . .1226
ETH_DMATPDR . . . . . . . . . . . . . . . . . . . . .1224
ETH_MACA0HR . . . . . . . . . . . . . . . . . . . . .1205
ETH_MACA0LR . . . . . . . . . . . . . . . . . . . . . .1206
ETH_MACA1HR . . . . . . . . . . . . . . . . . . . . .1206
ETH_MACA1LR . . . . . . . . . . . . . . . . . . . . . .1207
ETH_MACA2HR . . . . . . . . . . . . . . . . . . . . .1207
ETH_MACA2LR . . . . . . . . . . . . . . . . . . . . . .1208
ETH_MACA3HR . . . . . . . . . . . . . . . . . . . . .1209
ETH_MACA3LR . . . . . . . . . . . . . . . . . . . . . .1209
ETH_MACCR . . . . . . . . . . . . . . . . . . . . . . . .1191
ETH_MACDBGR . . . . . . . . . . . . . . . . . . . . .1202
ETH_MACFCR . . . . . . . . . . . . . . . . . . . . . . .1198
ETH_MACFFR . . . . . . . . . . . . . . . . . . . . . . .1194
ETH_MACHTHR . . . . . . . . . . . . . . . . . . . . .1195
ETH_MACHTLR . . . . . . . . . . . . . . . . . . . . . .1196
ETH_MACIMR . . . . . . . . . . . . . . . . . . . . . . .1205
ETH_MACMIIAR . . . . . . . . . . . . . . . . . . . . .1196
ETH_MACMIIDR . . . . . . . . . . . . . . . . . . . . .1197
ETH_MACPMTCSR . . . . . . . . . . . . . . . . . . .1201
ETH_MACRWUFFR . . . . . . . . . . . . . . . . . .1200
ETH_MACSR . . . . . . . . . . . . . . . . . . . . . . . .1204
ETH_MACVLANTR . . . . . . . . . . . . . . . . . . .1199
ETH_MMCCR . . . . . . . . . . . . . . . . . . . . . . .1210
ETH_MMCRFAECR . . . . . . . . . . . . . . . . . . .1215
ETH_MMCRFCECR . . . . . . . . . . . . . . . . . .1214
ETH_MMCRGUFCR . . . . . . . . . . . . . . . . . .1215

ETH_MMCRIMR . . . . . . . . . . . . . . . . . . . . . 1212
ETH_MMCRIR . . . . . . . . . . . . . . . . . . . . . . 1210
ETH_MMCTGFCR . . . . . . . . . . . . . . . . . . . 1214
ETH_MMCTGFMSCCR . . . . . . . . . . . . . . . 1214
ETH_MMCTGFSCCR . . . . . . . . . . . . . . . . . 1213
ETH_MMCTIMR . . . . . . . . . . . . . . . . . . . . . 1213
ETH_MMCTIR . . . . . . . . . . . . . . . . . . . . . . . 1211
ETH_PTPPPSCR . . . . . . . . . . . . . . . . . . . . 1222
ETH_PTPSSIR . . . . . . . . . . . . . . . . . . . . . . 1218
ETH_PTPTSAR . . . . . . . . . . . . . . . . . . . . . . 1220
ETH_PTPTSCR . . . . . . . . . . . . . . . . . . . . . 1215
ETH_PTPTSHR . . . . . . . . . . . . . . . . . . . . . 1218
ETH_PTPTSHUR . . . . . . . . . . . . . . . . . . . . 1219
ETH_PTPTSLR . . . . . . . . . . . . . . . . . . . . . . 1219
ETH_PTPTSLUR . . . . . . . . . . . . . . . . . . . . 1220
ETH_PTPTSSR . . . . . . . . . . . . . . . . . . . . . . 1221
ETH_PTPTTHR . . . . . . . . . . . . . . . . . . . . . . 1221
ETH_PTPTTLR . . . . . . . . . . . . . . . . . . . . . . 1221
EXTI_EMR . . . . . . . . . . . . . . . . . . . . . . . . . . 384
EXTI_FTSR . . . . . . . . . . . . . . . . . . . . . . . . . . 385
EXTI_IMR . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
EXTI_PR . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
EXTI_RTSR . . . . . . . . . . . . . . . . . . . . . . . . . . 385
EXTI_SWIER . . . . . . . . . . . . . . . . . . . . . . . . . 386

F
FLITF_FCR . . . . . . . . . . . . . . . . . . . . . . 103, 105
FLITF_FKEYR . . . . . . . . . . . . . . . . . . . . . . . . 100
FLITF_FOPTCR . . . . . . . . . . . . . . 106, 108, 110
FLITF_FOPTKEYR . . . . . . . . . . . . . . . . . . . . 100
FLITF_FSR . . . . . . . . . . . . . . . . . . . . . . 101-102
FSMC_BCR1..4 . . . . . . . . . . . . . . . . . 1575, 1638
FSMC_BTR1..4 . . . . . . . . . . . . . . . . . 1578, 1640
FSMC_BWTR1..4 . . . . . . . . . . . . . . . 1581, 1644
FSMC_PCR2..4 . . . . . . . . . . . . . . . . . . . . . . 1591
FSMC_PMEM2..4 . . . . . . . . . . . . . . . . . . . . 1593
FSMC_SR2..4 . . . . . . . . . . . . . . . . . . . . . . . 1592

G
GPIOx_AFRH . . . . . . . . . . . . . . . . . . . . . . . . 286
GPIOx_AFRL . . . . . . . . . . . . . . . . . . . . . . . . 285
GPIOx_BSRR . . . . . . . . . . . . . . . . . . . . . . . . 284
GPIOx_IDR . . . . . . . . . . . . . . . . . . . . . . . . . . 283
GPIOx_LCKR . . . . . . . . . . . . . . . . . . . . . . . . 284
GPIOx_MODER . . . . . . . . . . . . . . . . . . . . . . 281
GPIOx_ODR . . . . . . . . . . . . . . . . . . . . . . . . . 283
GPIOx_OSPEEDR . . . . . . . . . . . . . . . . . . . . 282
GPIOx_OTYPER . . . . . . . . . . . . . . . . . . . . . . 281
GPIOx_PUPDR . . . . . . . . . . . . . . . . . . . . . . . 282

RM0090 Rev 17

1743/1747

Index

RM0090

H
HASH_CR . . . . . . . . . . . . . . . . . . . . . . .782, 785
HASH_CSRx . . . . . . . . . . . . . . . . . . . . . . . . .794
HASH_DIN . . . . . . . . . . . . . . . . . . . . . . . . . . .788
HASH_HR0 . . . . . . . . . . . . . . . . . . . . . . . . . .790
HASH_HR1 . . . . . . . . . . . . . . . . . . . . . . 790-791
HASH_HR2 . . . . . . . . . . . . . . . . . . . . . . 790-791
HASH_HR3 . . . . . . . . . . . . . . . . . . . . . . . . . .791
HASH_HR4 . . . . . . . . . . . . . . . . . . . . . . . . . .791
HASH_IMR . . . . . . . . . . . . . . . . . . . . . . . . . . .792
HASH_SR . . . . . . . . . . . . . . . . . . . . . . . . . . .793
HASH_STR . . . . . . . . . . . . . . . . . . . . . . . . . .789

I
I2C_CCR . . . . . . . . . . . . . . . . . . . . . . . . . . . .870
I2C_CR1 . . . . . . . . . . . . . . . . . . . . . . . . . . . .860
I2C_CR2 . . . . . . . . . . . . . . . . . . . . . . . . . . . .862
I2C_DR . . . . . . . . . . . . . . . . . . . . . . . . . . . . .865
I2C_OAR1 . . . . . . . . . . . . . . . . . . . . . . . . . . .864
I2C_OAR2 . . . . . . . . . . . . . . . . . . . . . . . . . . .864
I2C_SR1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .865
I2C_SR2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .868
I2C_TRISE . . . . . . . . . . . . . . . . . . . . . . . . . . .871
IWDG_KR . . . . . . . . . . . . . . . . . . . . . . . . . . .710
IWDG_PR . . . . . . . . . . . . . . . . . . . . . . . . . . .710
IWDG_RLR . . . . . . . . . . . . . . . . . . . . . . . . . .711
IWDG_SR . . . . . . . . . . . . . . . . . . . . . . . . . . .711

O
OTG_FS_CID . . . . . . . . . . . . . . . . . . . . . . . .1291
OTG_FS_DAINT . . . . . . . . . . . . . . . . . . . . .1308
OTG_FS_DAINTMSK . . . . . . . . . . . . . . . . .1309
OTG_FS_DCFG . . . . . . . . . . . . . . . . . . . . . .1303
OTG_FS_DCTL . . . . . . . . . . . . . . . . . . . . . .1304
OTG_FS_DIEPCTL0 . . . . . . . . . . . . . . . . . .1310
OTG_FS_DIEPCTLx . . . . . . . . . . . . . . . . . .1312
OTG_FS_DIEPEMPMSK . . . . . . . . . . . . . . .1310
OTG_FS_DIEPINTx . . . . . . . . . . . . . . . . . . .1319
OTG_FS_DIEPMSK . . . . . . . . . . . . . . . . . . .1306
OTG_FS_DIEPTSIZ0 . . . . . . . . . . . . . . . . . .1321
OTG_FS_DIEPTSIZx . . . . . . . . . . . . . . . . . .1324
OTG_FS_DIEPTXF0 . . . . . . . . . . . . . . . . . .1289
OTG_FS_DIEPTXFx . . . . . . . . . . . . . . . . . .1292
OTG_FS_DOEPCTL0 . . . . . . . . . . . . . . . . .1315
OTG_FS_DOEPCTLx . . . . . . . . . . . . . . . . .1316
OTG_FS_DOEPINTx . . . . . . . . . . . . . . . . . .1320
OTG_FS_DOEPMSK . . . . . . . . . . . . . . . . . .1307
OTG_FS_DOEPTSIZ0 . . . . . . . . . . . . . . . . .1323
OTG_FS_DOEPTSIZx . . . . . . . . . . . . . . . . .1325
OTG_FS_DSTS . . . . . . . . . . . . . . . . . . . . . .1305

1744/1747

OTG_FS_DTXFSTSx . . . . . . . . . . . . . . . . . 1325
OTG_FS_DVBUSDIS . . . . . . . . . . . . . . . . . 1309
OTG_FS_DVBUSPULSE . . . . . . . . . . . . . . 1309
OTG_FS_GAHBCFG . . . . . . . . . . . . . . . . . 1275
OTG_FS_GCCFG . . . . . . . . . . . . . . . . . . . . 1290
OTG_FS_GINTMSK . . . . . . . . . . . . . . . . . . 1284
OTG_FS_GINTSTS . . . . . . . . . . . . . . . . . . 1280
OTG_FS_GOTGCTL . . . . . . . . . . . . . . . . . . 1272
OTG_FS_GOTGINT . . . . . . . . . . . . . . . . . . 1273
OTG_FS_GRSTCTL . . . . . . . . . . . . . . . . . . 1278
OTG_FS_GRXFSIZ . . . . . . . . . . . . . . . . . . 1288
OTG_FS_GRXSTSP . . . . . . . . . . . . . . . . . . 1287
OTG_FS_GRXSTSR . . . . . . . . . . . . . . . . . . 1287
OTG_FS_GUSBCFG . . . . . . . . . . . . . . . . . 1276
OTG_FS_HAINT . . . . . . . . . . . . . . . . . . . . . 1295
OTG_FS_HAINTMSK . . . . . . . . . . . . . . . . . 1296
OTG_FS_HCCHARx . . . . . . . . . . . . . . . . . . 1299
OTG_FS_HCFG . . . . . . . . . . . . . . . . . . . . . 1293
OTG_FS_HCINTMSKx . . . . . . . . . . . . . . . . 1301
OTG_FS_HCINTx . . . . . . . . . . . . . . . . . . . . 1300
OTG_FS_HCTSIZx . . . . . . . . . . . . . . . . . . . 1302
OTG_FS_HFIR . . . . . . . . . . . . . . . . . . . . . . 1293
OTG_FS_HFNUM . . . . . . . . . . . . . . . . . . . . 1294
OTG_FS_HNPTXFSIZ . . . . . . . . . . . . . . . . 1289
OTG_FS_HNPTXSTS . . . . . . . . . . . . . . . . . 1289
OTG_FS_HPRT . . . . . . . . . . . . . . . . . . . . . 1296
OTG_FS_HPTXFSIZ . . . . . . . . . . . . . . . . . . 1292
OTG_FS_HPTXSTS . . . . . . . . . . . . . . . . . . 1294
OTG_FS_PCGCCTL . . . . . . . . . . . . . . . . . . 1326
OTG_HS_CID . . . . . . . . . . . . . . . . . . . . . . . 1428
OTG_HS_DAINT . . . . . . . . . . . . . . . . . . . . . 1449
OTG_HS_DAINTMSK . . . . . . . . . . . . . . . . . 1450
OTG_HS_DCFG . . . . . . . . . . . . . . . . . . . . . 1442
OTG_HS_DCTL . . . . . . . . . . . . . . . . . . . . . 1444
OTG_HS_DEACHINT . . . . . . . . . . . . . . . . . 1453
OTG_HS_DEACHINTMSK . . . . . . . . . . . . . 1454
OTG_HS_DIEPCTLx . . . . . . . . . . . . . . . . . . 1456
OTG_HS_DIEPDMAx . . . . . . . . . . . . . . . . . 1470
OTG_HS_DIEPEACHMSK1 . . . . . . . . . . . . 1454
OTG_HS_DIEPEMPMSK . . . . . . . . . . . . . . 1453
OTG_HS_DIEPINTx . . . . . . . . . . . . . . . . . . 1463
OTG_HS_DIEPMSK . . . . . . . . . . . . . . . . . . 1447
OTG_HS_DIEPTSIZ0 . . . . . . . . . . . . . . . . . 1466
OTG_HS_DIEPTSIZx . . . . . . . . . . . . . . . . . 1468
OTG_HS_DIEPTXFx . . . . . . . . . . . . . . . . . . 1429
OTG_HS_DOEPCTL0 . . . . . . . . . . . . . . . . . 1459
OTG_HS_DOEPCTLx . . . . . . . . . . . . . . . . . 1460
OTG_HS_DOEPDMAx . . . . . . . . . . . . . . . . 1470
OTG_HS_DOEPEACHMSK1 . . . . . . . . . . . 1455
OTG_HS_DOEPINTx . . . . . . . . . . . . . . . . . 1465
OTG_HS_DOEPMSK . . . . . . . . . . . . . . . . . 1448
OTG_HS_DOEPTSIZ0 . . . . . . . . . . . . . . . . 1467

RM0090 Rev 17

RM0090

Index

OTG_HS_DOEPTSIZx . . . . . . . . . . . . . . . . .1469
OTG_HS_DSTS . . . . . . . . . . . . . . . . . . . . . .1446
OTG_HS_DTHRCTL . . . . . . . . . . . . . . . . . .1452
OTG_HS_DTXFSTSx . . . . . . . . . . . . . . . . .1469
OTG_HS_DVBUSDIS . . . . . . . . . . . . . . . . .1450
OTG_HS_DVBUSPULSE . . . . . . . . . . . . . .1451
OTG_HS_GAHBCFG . . . . . . . . . . . . . . . . . .1410
OTG_HS_GCCFG . . . . . . . . . . . . . . . . . . . .1427
OTG_HS_GINTMSK . . . . . . . . . . . . . . . . . .1421
OTG_HS_GINTSTS . . . . . . . . . . . . . . . . . . .1417
OTG_HS_GNPTXFSIZ . . . . . . . . . . . . . . . .1426
OTG_HS_GNPTXSTS . . . . . . . . . . . . . . . . .1426
OTG_HS_GOTGCTL . . . . . . . . . . . . . . . . . .1407
OTG_HS_GOTGINT . . . . . . . . . . . . . . . . . .1408
OTG_HS_GRSTCTL . . . . . . . . . . . . . . . . . .1414
OTG_HS_GRXFSIZ . . . . . . . . . . . . . . . . . . .1425
OTG_HS_GRXSTSP . . . . . . . . . . . . . . . . . .1424
OTG_HS_GRXSTSR . . . . . . . . . . . . . . . . . .1424
OTG_HS_GUSBCFG . . . . . . . . . . . . . . . . . .1411
OTG_HS_HAINT . . . . . . . . . . . . . . . . . . . . .1433
OTG_HS_HAINTMSK . . . . . . . . . . . . . . . . .1433
OTG_HS_HCCHARx . . . . . . . . . . . . . . . . . .1436
OTG_HS_HCDMAx . . . . . . . . . . . . . . . . . . .1442
OTG_HS_HCFG . . . . . . . . . . . . . . . . . . . . .1429
OTG_HS_HCINTMSKx . . . . . . . . . . . . . . . .1440
OTG_HS_HCINTx . . . . . . . . . . . . . . . . . . . .1439
OTG_HS_HCSPLTx . . . . . . . . . . . . . . . . . .1438
OTG_HS_HCTSIZx . . . . . . . . . . . . . . . . . . .1441
OTG_HS_HFIR . . . . . . . . . . . . . . . . . . . . . .1431
OTG_HS_HFNUM . . . . . . . . . . . . . . . . . . . .1431
OTG_HS_HPRT . . . . . . . . . . . . . . . . . . . . . .1434
OTG_HS_HPTXFSIZ . . . . . . . . . . . . . . . . . .1428
OTG_HS_HPTXSTS . . . . . . . . . . . . . . . . . .1432
OTG_HS_PCGCCTL . . . . . . . . . . . . . . . . . .1471
OTG_HS_TX0FSIZ . . . . . . . . . . . . . . . . . . .1426

P
PWR_CR . . . . . . . . . . . . . . . . . . . . . . . .141, 144
PWR_CSR . . . . . . . . . . . . . . . . . . . . . . .142, 147

R
RCC_AHB1ENR . . . . . . . . . . . . . . . . . .180,
RCC_AHB1LPENR . . . . . . . . . . . . . . . .189,
RCC_AHB1RSTR . . . . . . . . . . . . . . . . .170,
RCC_AHB2ENR . . . . . . . . . . . . . . . . . .182,
RCC_AHB2LPENR . . . . . . . . . . . . . . . .192,
RCC_AHB2RSTR . . . . . . . . . . . . . . . . .173,
RCC_AHB3ENR . . . . . . . . . . . . . . . . . .183,
RCC_AHB3LPENR . . . . . . . . . . . . . . . .193,
RCC_AHB3RSTR . . . . . . . . . . . . . . . . .174,
RCC_APB1ENR . . . . . . . . . . . . . . . . . . .183,

242
250
233
244
252
236
245
253
237
245

RCC_APB1LPENR . . . . . . . . . . . . . . . . 193, 254
RCC_APB1RSTR . . . . . . . . . . . . . . . . . 174, 237
RCC_APB2ENR . . . . . . . . . . . . . . . . . . 187, 248
RCC_APB2LPENR . . . . . . . . . . . . . . . . 197, 257
RCC_APB2RSTR . . . . . . . . . . . . . . . . . 178, 240
RCC_BDCR . . . . . . . . . . . . . . . . . . . . . 199, 259
RCC_CFGR . . . . . . . . . . . . . . . . . . . . . 165, 228
RCC_CIR . . . . . . . . . . . . . . . . . . . . . . . 167, 230
RCC_CR . . . . . . . . . . . . . . . . . . . . . . . . 161, 224
RCC_CSR . . . . . . . . . . . . . . . . . . . . . . . 200, 260
RCC_PLLCFGR . . . . . .163, 203, 206, 226, 263
RCC_SSCGR . . . . . . . . . . . . . . . . . . . . 202, 262
RNG_CR . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769
RNG_DR . . . . . . . . . . . . . . . . . . . . . . . . . . . . 770
RNG_SR . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769
RTC_ALRMAR . . . . . . . . . . . . . . . . . . . . . . . 825
RTC_ALRMBR . . . . . . . . . . . . . . . . . . . . . . . 826
RTC_ALRMBSSR . . . . . . . . . . . . . . . . . . . . . 835
RTC_BKxR . . . . . . . . . . . . . . . . . . . . . . . . . . 836
RTC_CALIBR . . . . . . . . . . . . . . . . . . . . . . . . 824
RTC_CALR . . . . . . . . . . . . . . . . . . . . . . . . . . 830
RTC_CR . . . . . . . . . . . . . . . . . . . . . . . . . . . . 818
RTC_DR . . . . . . . . . . . . . . . . . . . . . . . . . . . . 817
RTC_ISR . . . . . . . . . . . . . . . . . . . . . . . . . . . . 820
RTC_PRER . . . . . . . . . . . . . . . . . . . . . . . . . . 823
RTC_SHIFTR . . . . . . . . . . . . . . . . . . . . . . . . 828
RTC_SSR . . . . . . . . . . . . . . . . . . . . . . . . . . . 827
RTC_TR . . . . . . . . . . . . . . . . . . . . . . . . . . . . 816
RTC_TSDR . . . . . . . . . . . . . . . . . . . . . . . . . . 829
RTC_TSSSR . . . . . . . . . . . . . . . . . . . . . . . . . 830
RTC_TSTR . . . . . . . . . . . . . . . . . . . . . . . . . . 828
RTC_WPR . . . . . . . . . . . . . . . . . . . . . . . . . . . 827
RTC_WUTR . . . . . . . . . . . . . . . . . . . . . . . . . 823

S
SDIO_CLKCR . . . . . . . . . . . . . . . . . . . . . . . 1061
SDIO_DCOUNT . . . . . . . . . . . . . . . . . . . . . 1067
SDIO_DCTRL . . . . . . . . . . . . . . . . . . . . . . . 1066
SDIO_DLEN . . . . . . . . . . . . . . . . . . . . . . . . 1065
SDIO_DTIMER . . . . . . . . . . . . . . . . . . . . . . 1064
SDIO_FIFO . . . . . . . . . . . . . . . . . . . . . . . . . 1074
SDIO_FIFOCNT . . . . . . . . . . . . . . . . . . . . . 1073
SDIO_ICR . . . . . . . . . . . . . . . . . . . . . . . . . . 1069
SDIO_MASK . . . . . . . . . . . . . . . . . . . . . . . . 1071
SDIO_POWER . . . . . . . . . . . . . . . . . . . . . . 1060
SDIO_RESPCMD . . . . . . . . . . . . . . . . . . . . 1063
SDIO_RESPx . . . . . . . . . . . . . . . . . . . . . . . 1064
SDIO_STA . . . . . . . . . . . . . . . . . . . . . . . . . . 1068
SPI_CR1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 916
SPI_CR2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 918
SPI_CRCPR . . . . . . . . . . . . . . . . . . . . . . . . . 920

RM0090 Rev 17

1745/1747

Index

RM0090

SPI_DR . . . . . . . . . . . . . . . . . . . . . . . . . . . . .920
SPI_I2SCFGR . . . . . . . . . . . . . . . . . . . . . . . .922
SPI_I2SPR . . . . . . . . . . . . . . . . . . . . . . . . . . .923
SPI_RXCRCR . . . . . . . . . . . . . . . . . . . . . . . .921
SPI_SR . . . . . . . . . . . . . . . . . . . . . . . . . . . . .919
SPI_TXCRCR . . . . . . . . . . . . . . . . . . . . . . . .921
SYSCFG_EXTICR1 . . . . . . . . . . . . . . . .291, 297
SYSCFG_EXTICR2 . . . . . . . . . . . . . . . .291, 298
SYSCFG_EXTICR3 . . . . . . . . . . . . . . . .292, 298
SYSCFG_EXTICR4 . . . . . . . . . . . . . . . .293, 299
SYSCFG_MEMRMP . . . . . . . . . . . . . . .289, 294

T
TIM2_OR . . . . . . . . . . . . . . . . . . . . . . . . . . . .647
TIM5_OR . . . . . . . . . . . . . . . . . . . . . . . . . . . .647
TIMx_ARR . . . . . . . . . . . . . . 642, 682, 692, 706
TIMx_BDTR . . . . . . . . . . . . . . . . . . . . . . . . . .583
TIMx_CCER . . . . . . . . . . . . . 576, 640, 681, 691
TIMx_CCMR1 . . . . . . . . . . . 572, 636, 678, 688
TIMx_CCMR2 . . . . . . . . . . . . . . . . . . . .575, 639
TIMx_CCR1 . . . . . . . . . . . . . 581, 643, 683, 693
TIMx_CCR2 . . . . . . . . . . . . . . . . . 582, 643, 683
TIMx_CCR3 . . . . . . . . . . . . . . . . . . . . . .582, 644
TIMx_CCR4 . . . . . . . . . . . . . . . . . . . . . .583, 644
TIMx_CNT . . . . . . . . . . 580, 642, 682, 692, 705
TIMx_CR1 . . . . . . . . . . 561, 627, 672, 686, 702
TIMx_CR2 . . . . . . . . . . . . . . . . . . 562, 629, 704
TIMx_DCR . . . . . . . . . . . . . . . . . . . . . . .585, 645
TIMx_DIER . . . . . . . . . . 567, 632, 674, 687, 704
TIMx_DMAR . . . . . . . . . . . . . . . . . . . . . .586, 646
TIMx_EGR . . . . . . . . . . 570, 635, 677, 688, 705
TIMx_PSC . . . . . . . . . . 580, 642, 682, 692, 706
TIMx_RCR . . . . . . . . . . . . . . . . . . . . . . . . . . .581
TIMx_SMCR . . . . . . . . . . . . . . . . . 565, 630, 673
TIMx_SR . . . . . . . . . . . 569, 633, 676, 687, 705

U
USART_BRR . . . . . . . . . . . . . . . . . . . . . . . .1010
USART_CR1 . . . . . . . . . . . . . . . . . . . . . . . .1010
USART_CR2 . . . . . . . . . . . . . . . . . . . . . . . .1013
USART_CR3 . . . . . . . . . . . . . . . . . . . . . . . .1014
USART_DR . . . . . . . . . . . . . . . . . . . . . . . . .1010
USART_GTPR . . . . . . . . . . . . . . . . . . . . . . .1017
USART_SR . . . . . . . . . . . . . . . . . . . . . . . . .1007

W
WWDG_CFR . . . . . . . . . . . . . . . . . . . . . . . . .718
WWDG_CR . . . . . . . . . . . . . . . . . . . . . . . . . .717
WWDG_SR . . . . . . . . . . . . . . . . . . . . . . . . . .718

1746/1747

RM0090 Rev 17

RM0090

IMPORTANT NOTICE – PLEASE READ CAREFULLY
STMicroelectronics NV and its subsidiaries (“ST”) reserve the right to make changes, corrections, enhancements, modifications, and
improvements to ST products and/or to this document at any time without notice. Purchasers should obtain the latest relevant information on
ST products before placing orders. ST products are sold pursuant to ST’s terms and conditions of sale in place at the time of order
acknowledgement.
Purchasers are solely responsible for the choice, selection, and use of ST products and ST assumes no liability for application assistance or
the design of Purchasers’ products.
No license, express or implied, to any intellectual property right is granted by ST herein.
Resale of ST products with provisions different from the information set forth herein shall void any warranty granted by ST for such product.
ST and the ST logo are trademarks of ST. All other product or service names are the property of their respective owners.
Information in this document supersedes and replaces information previously supplied in any prior versions of this document.
© 2018 STMicroelectronics – All rights reserved

RM0090 Rev 17

1747/1747
1747



---

Source Exif Data:

File Type                       : PDF
File Type Extension             : pdf
MIME Type                       : application/pdf
PDF Version                     : 1.6
Linearized                      : Yes
Alternate Name                  : RM0090
Author                          : STMICROELECTRONICS
Classification                  : Unclassified
Create Date                     : 2018:06:07 13:59:05Z
Doc ID                          : RM0090
Document Type                   : Reference manual
Modify Date                     : 2018:08:31 09:44:05+09:00
Revision                        : 17
Subject                         : This reference manual targets application developers. It provides complete information on how to use the STM32F405xx/07xx, STM32F415xx/17xx, STM32F42xxx and STM32F43xxx microcontroller memory and peripherals.
Has XFA                         : No
XMP Toolkit                     : Adobe XMP Core 5.6-c015 84.159810, 2016/09/10-02:41:30
Format                          : application/pdf
Creator                         : STMICROELECTRONICS
Description                     : This reference manual targets application developers. It provides complete information on how to use the STM32F405xx/07xx, STM32F415xx/17xx, STM32F42xxx and STM32F43xxx microcontroller memory and peripherals.
Title                           : STM32F405/415, STM32F407/417, STM32F427/437 and STM32F429/439 advanced Arm®-based 32-bit MCUs - Reference manual
Creator Tool                    : C2 v4.2.0220 build 670 - c2_rendition_config : Techlit_Active
Metadata Date                   : 2018:08:31 09:44:05+09:00
Producer                        : Acrobat Elements 10.0.0 (Windows); modified using iText 2.1.7 by 1T3XT
Document ID                     : uuid:c46d8dff-eacf-4c1b-8bd4-8187dbf7e2dd
Instance ID                     : uuid:c3732f72-b711-4d9f-87be-0f62435c46f6
Page Layout                     : SinglePage
Page Mode                       : UseNone
Page Count                      : 1747

EXIF Metadata provided by EXIF.tools