DEC 10 H5EC D PDP RP10 Disk Pack Synchronizer Maintenance Manual Volume 1

DEC-10-H5EC-D PDP-10 RP10 Disk Pack Synchronizer Maintenance Manual Volume 1 DEC-10-H5EC-D PDP-10 RP10 Disk Pack Synchronizer Maintenance Manual Volume 1

User Manual: DEC-10-H5EC-D PDP-10 RP10 Disk Pack Synchronizer Maintenance Manual Volume 1

Open the PDF directly: View PDF PDF.
Page Count: 104

DownloadDEC-10-H5EC-D PDP-10 RP10 Disk Pack Synchronizer Maintenance Manual Volume 1
Open PDF In BrowserView PDF
Digital Equipment Corporation
Maynard, Massachusetts

PDP-10
Maintenance Manual

DEC-IO-HSEC-D (1)

PDP-10
RP10 DISK PACK
SYNCHRONIZER
MAINTENANCE MANUAL
VOLUME 1

DIGITAL

EQUIPMENT

CORPORATION • MAYNARD, MASSACHUSETTS

1st Edition November 1969
2nd Printing May 1970
2nd Edition November 1970
3rd Edition August 1971

Copyright © 1969, 1970, 1971 by Digital Equipment Corporation

The material in this manual is for information
purposes and is subject to change without
notice.

The following are trademarks of Digital Equipment
Corporation, Maynard, Massachusetts:
DEC
FLIP CHIP
DIGITAL

PDP
FOCAL
COMPUTER LAB

CONTENTS
Page
CHAPTER I GENERAL INFORMATION
l.l

lntroduction

1-1

1.2

Specifications

1-1

1.2.1

Physical

I-I

1.2.2

Environmental

1-2

1.2.3

Electrical

1-2

1.2.4

Performance

1-3

1.3

Interface Req uirements

1-3

1.3.1
1.3.2

RPIO/RPOI/02 Interface

1-3

RPIO/DFIO Interface
RP I 0/1(0 Interface
Disk Pack/Disk Pack Interface
Equipment Supplied
Reference Documents
Glossary

1-5

1.3.3
1.3.4
1.4
1.5
1.6

1-5

1-6
1-8
1-8
1-8

CHAPTER 2 INSTALLATION
2.1

2-1

2.2

General
Unpacking

2.2.1

Special Handling

2-1

2.2.2

Inspection

2-1

2.2.3
2.3
2.4

Power Requirements
Installation Procedure

2-1
2-2

Turn-On and Checkout Procedures

2-4

2-1

CHAPTER 3 OPERATION
3.1

General

3-1

3.2

Controls and Indicators
Operation

3-1

3.3
3.3.1
3.4
3.4.1
3.4.2

Loading, Unloading, and Storage of RPOIP/02P Disk Packs
Programming
RPO 1/02 Disk Pack Characteristics
RPIO Programming

3.4.2.1
3.4.2.2

DATAO Information

3.4.2.3

CONO Information
CONI Information

3.4.2.4

3-6
3-6
3-7
3-7
3-11

3-12
3-16

DATAl Information

3-17
3-18

iii

CONTENTS (Cont)
Page
CHAPTER 4 THEORY OF OPERATION
4.1
4.2
4.3
4.3.1
4.3.2
4.3.3
4.4
4.5
4.6
4.7
4.7.1
4.7.2
4.7.3
4.7.4
4.7.5
4.7.6
4.8
4.8.1
4.8.2
4.8.2.1
4.8.2.2
4.8.2.3
4.8.2.4
4.8.2.5
4.8.2.6
4.8.3
4.8.4
4.9
4.9.1
4.9.2
4.9.3
4.9.4
4.9.5
4.1 0
4.1 0.1
4.11
4.11.1
4.11.2
4.11.3
4.11.4
4.11.5
4.11.6

General
General Block Diagram Discussion
Sequence of Operation
Write Data or Read Data
Write Headers and Data (Format)
Local/Remote
RPIO Recording Technique
RPOI/02 Addressing Method
RP I 0 Register Organization
RP I 0 Registers
Assembly Register
Assembly Register Data Gate
Condition Register
Data Address Register
Longitudinal Parity Register
Shift Register
RPIO Control Circuits
Channel Control
Data Transfer Control
Read Data
Write Data
Write Headers and Data
Clear Attentions
Restore
Seek
I/O Bus Control
RPIO Read Data Separator
RPIO Counters
Sector Counters
Sector Counter Buffer
Sector Counter Control Pulses
Sector Counter Multiplexer
Word Counters
RPIO Comparators
Header Compare and Designation Error
RPIO Data Flow
Starting
Recognizing Headers
Reading Data
Writing Data
Commanding the Drive
Reading Multiple Sectors

4-1
4-1
4-1
4-2
4-4
4-4
4-5
4-5
4-6
4-6
4-6
4-7
4-10
4-13
4-13
4-13
4-13
4-13
4-17
4-17
4-19
4-19
4-19
4-19
4-22
4-22
4-23
4-28
4-29
4-29
4-29
4-31
4-31
4-34
4-34
4-34
4-34
4-37
4-37
4-38
4-38
4-38

iv

CONTENTS (Cont)
Page
4.11.7
4.11. 7.1
4.11.7.2
4.11.8
4.11.8.1
4.11.8.2
4.11.9
4.11.10
4. 11.11

4-39
4-39
4-39
4-39
4-39
4-39
4-39
4-39
4-40

Loading Words
Header Words
Data Words
Transferring Words
Write Header
Write Data
Writing Headers
Generating Parity
Terminating

CHAPTER 5 MAINTENANCE
5.1
5.2
5.2.1
5.2.1.1
5.2.1.2
5.2.1.3
5.2.1.4

I

5.3
5.3.1
5.3.2
5.3.2.1
5.3.3
5.3.3.1
5.3.4
5.3.4.1
5.3.4.2
5.3.4.3
5.3.4.4
5.3.4.5
5.3.4.6
5.3.4.7

Introduction
Preventive Maintenance
Preventiw Maintenance Procedures
Mechanical Checks
Test Equipment Required
Electrical Checks
Electronic Checks
Corrective Maintenance
General Corrective Procedures
Diagnostic Testing
Vibration Tests
Adjustment Procedures
Read Data Separator Calibration
General Troubleshooting Guides
Disk Pack Compatibility
Proper Use of C.E. Packs
Search Errors
Parity Errors
Failure to Restore
Disk Pack Reliability Test Map Limitations
Cleaning

5-1
5-1
5-1
5-1
5-1
5-2
5-3
5-4
5-4
5-5
5-5
5-5
5-5
5-9
5-9
5-9
5-10
5-11
5-11
5-11
5-11

CHAPTER 6 DRAWINGS
6.1

General

6-1

v

ILLUSTRA nONS
Figure No.
1-1
1-2
1-3
1-4
1-5
1-6
2-1
2-2
3-1
3-2
3-3
3-4
3-5
3-6
3-7
3-8
3-9
3-10
3-11
3-12

*

3-13
3-14

*

3-15
4-1
4-2
4-3
4-4
4-5
4-6

*

4-7
4-8
4-9

* 4-10
* 4-11
* 4-12
4-13

* 4-14
4-15
4-16
4-17
4-18
4-19
4-20

Title

Art No.

DEC Type RP 10 Disk Pack Synchronizer
Typical PDP-l 0 System with RPIO Disk Pack Synchronizer
RPI OIDFI O/RPOI /RP02/I/O Interface
RP I O/RPO 1/02 Electrical Interface
RP I 0 Interface Wiring Diagram
RPIO Cable Diagram
RP I 0 Overall Dimensions
RPI0-RPOI/RP02 Signal and Power Connections
RP I 0 Controls and Indicators
RPOI /RP02 Disk Pack Characteristics
RPI0/RPOI/RP02 Data Control Signals
RPOI/RP02 Data Format
DATAO Instruction Word Format
Position Instruction Word
Recalibrate Instruction Word
Read Data Instruction Word
Write Data Instruction Word
Write Header and Data Instruction Word
Clear Attentions Instruction Word
No Op Instruction Word
DA T AI Instruction Word Format
CONO Instruction Word Format
CONI Instruction Word Format
RP 10 General Block Diagram
RP I 0 Active and Inactive Functions
RP I 0 Register Organization
RP I 0 Assembly Register Simplified Block Diagram
RPIO Assembly Register Data Gate, Simplified Block Diagram
RP I 0 Condition Register, Block Diagram
RPIO Data Address Register, Block Diagram
RP I 0 Longitudinal Parity Register, Block Diagram
RP 10 Shift Register, Block Diagram
DTC OP Decoding, Block Diagram
DTC Read Data Command, Block Diagram
DTC Write Data Command, Block Diagram
DTC Write Headers and Data Command, Block Diagram
DTC Clear Attentions Command, Block Diagram
DTC Restore Command, Block Diagram
Tag and Bus Line Strobe Waveforms
DTC Seek Command, Block Diagram
Sources of Jitter in Disk Playback
VFO Startup Waveform
RDS DATA WIND Waveform

*
* Denotes figures revised since the November 1970 printing.

vi

10-0160
10-0162
10-0163
10-0177
10-0161
10-0019
10-0534

10-0165
10-0166
10-0168
10-0170

10-0169
10-0171
10-0167
10-0159
10-0108
10-0507
10-0508
10-0509
10-0511
10-0512
10-0513
10-0514
10-0515
10-0516
10-0517
10-0518
10-0519
10-0520
10-0535
10-0538
10-0537
10-0687

Page
1-1
1-2
1-4
1-4
1-6
1-7
2-4
2-5
3-1
3-8
3-9
3-10
3-12
3-12
3-13
3-13
3-14
3-15
3-15
3-16
3-17
3-17
3-19
4-2
4-3
4-7
4-9
4-10
4-11
4-14
4-15
4-16
4-18
4-18
4-20
4-20
4-21
4-21
4-22
4-22
4-24
4-25
4-25

ILLUSTRATIONS (Cont)
Figure No.

* 4-21
4-22

*

4-23
4-24
* 4-25
4-26
4-27
* 4-28
* 4-29
**5-1
**5-2

Art No.

Title

RPIO Read Data Separator, Block Diagram
RDS Error Flops Waveforms
Optimum Window Generation Waveforms
RPIO Sector Counter, Block Diagram
RPIO Sector Counter Buffer, Block Diagram
RPIO Sector Counter Control Pulses, Simplified Block Diagram
RPIO Sector Counter Multiplexer, Block Diagram
RPIO Word Counters Block Diagram
RPIO Header Compare and Designation Error,
Simplified Block Diagram
RDS Waveform No.1
RDS Waveform No.2

Page

10-0522
10-0523
10-0524
10-0525
10-0526
10-0527
10-0528
10-0529

4-27
4-28
4-28
4-29
4-30
4-31
4-32
4-33

10-0530
10-0688
10-0689

4-35
5-7
5-7

TABLES
Table No.
1-1
1-2
1-3
1-4
1-5
2-1
3-1
4-1
5-1
5-2
5-3
5-4
5-5
6-1

Title

Page

RPOI/02 Features Compatible with RPIO Logic Configuration
RPIOjRPOI and RP02 Interface Chart
RPlO/DFlO Interface Chart
RPIO/I/O Interface Chart
Equipment Supplied
RPOI/RP02 Power Requirements
RPIO Controls and Indicators
Data Transfer Control Commands
Test Equipment Required
Power Supply Output Checks
Critical Delay Modules
DFIO Voltage Margins
RPIO Voltage Margins
RPIO Engineering Drawings

1-3

*Denotes figures revised since the November 1970 printing.
**Denotes figures added since the November 1970 printing.

vii

1-5

1-5
1-6
1-8
2"2
3-2
4-36
5-2

5-2
5-3
5-3
5-4
6-1

Chapter 1
General Information

1.1 INTRODUCTION
The RPlO Disk Pack Synchronizer, shown in Figure 1-1, provides the interface logic between the DFIO Data
Channel and as many as eight of the DEC Type RPO I or RP02 Disk Pack Drives. The RPIO operates under
control of the PDP-I a I/O Bus and, in turn, controls the RPO I or RP02 (see Figure 1-2). These units provide the
PDP-I a computer with on-line auxiliary storage for between 1.28 and 40 million 36-bit words. Direct access to
core memory is obtained through the DFIO Data Channel.
The RPIO is housed in a 19-in. cabinet containing eight module racks.
Indicator and switch panels are located on the top front of the unit.
These comprise monitoring and switch register facilities for data, address,
and control. The unit contains the standard margin check voltage bracket,
+ 10 and -15 V dc power supplies and power control, and built-in cooling
fans.
The RPIO is shipped complete and factory tested including all interconnecting cabling to the PDP-I a and the RPOl/02 Disk Packs. Two
types are available: the RPIO-A for 60-Hz operation, and the RPIO-B
for 50-Hz operation.
1.2 SPECIFICATIONS
Physical, environmental, electrical, and performance specifications for
the RPIO Disk Pack Synchronizer are given in the paragraphs that follow.
1.2.1 Physical
Dimensions
Width
Height
Depth
Weight

= 21 in.
= 69 in.
= 27 in.
= 390 Ib

Service Access Dimensions
Front
Rear

36 in.
36 in.
Figure I-I DEC Type RPIO Disk
Pack Synchronizer

I-I

CONSOLE PAPER
TELETAPE
PRINTER READER
MEMORY
MODULE
(8,192
WORDS)

- ~

MEMORY
MODULE
(16,384
WORDS)

r- ~~

r-<

INPUTI OUTPUT BUS
TO OTH ER PERIPHERALS

PDP-10
CENTRAL
PROCESSOR

r------i-----'-----,
MEMORY
EXT
I
I

I

I

PROT. : ORDE'R : FAST I
AND I CODE I MEMORY:
I
REL. I
I
I
1.. ______ 1 ______ !.. ______ J

:

UP TO
2 62,144
WORDS

1
MEMORY 0
MODULE

PAPER
TAPE
PUNCH

DATA
CHANNEL
OF to

~(CONTROLLER
DISK PACK

r-<

RP10

H

I--<
r- --f~

,
, :
1

1
MEMORY 0
MODULE

:

DISK
PACK

I

DISK
PACK

J

)

RP01 AND lOR RP02

,

UP TO
8

UP TO
8
DEVICES

~

10-01eo

~.~ =MC10

Figure 1-2 Typical PDP-l 0 System with
RP 10 Disk Pack Synchronizer

1.2.2 Environmental
The RPIO operates under normal conditions of humidity, shock, and vibration.
Recommended Ambient
Operating Temperature = 60° to 95°F
Storage Temperature = 40° to 110°F
Humidity = 20% to 80% Relative
Wet Bulb = 78°F (Max)
1.2.3 Electrical
Power Requirement at Line Cord
115 Vac, I-phase, 60 Hz@ 8.0A
Power/Heat Dissipation = 9000W/3000 Btu/Hr
Internal Power - From 4 self-contained Type 728 Power Supplies and one
self-contained Type 844 Power Control.
+10 Vdc
-15 Vdc
U nit Cable Lengths

50 ft (Max)

1-2

1.2.4 Performance
Word Transfer Time
When synchronizing RPO I = 30 JlS
When synchronizing RP02 = 15 JlS
Positioning Time
Track-to-track = 20 ms
Average = 50 ms
Maximum = 80 ms
Latency Time
Average = 12.5 ms
Maximum = 25 ms
1.3 INTERFACE REQUIREMENTS
The RP I 0 Synchronizer is designed to interface with any of the RPO I /02 Disk Pack options listed in Table I-I.
Figure 1-3 is the Interface Block Diagram for the system. A complete description of the RPO I /02 is contained in
the vendor manuals for the Disk Drives. These manuals are supplied with each piece of peripheral equipment.
Table 1-1
RPOl/02 Features Compatible with RPIO
Logic Configuration
Feature

Variations

Input Voltage and Frequency

208/230 Vac
± 10%, 60 Hz ± 1/2 Hz
220/380 Vac
± 10%,50 Hz ±2 Hz

NOTE
All drives in a given system must be driven by
the same voltage and the same frequency.
Recording Density (bit cell
time)

RPOI = 800 ns
RP02 = 400 ns

Number of Drives

RPOI=I-8
RP02 = I - 8
NOTE
May be intermixed.

1.3.1 RPlO/RPOI/02 Interface
There are two cable assemblies through which the RPIO connects to the RPOI or RP02 Disk Pack (see Figures 1-4
and 1-5); these assemblies and their functions are listed in Table 1-2. The electrical specification of the RPI O/RPO 1/02
interface is shown in Figure 1-4.

1-3

MEMORY BUS

DFIO
DATA
CHANNEL

I/O BUS

CHANNEL
BUS

I

I
L_

_J

CHANNEL BUS TO
NEXT DEVICE

KAIO
CENTRAL
PROCESSOR

I/O BUS TO NEXT DEVICE

RPIO
DISK PACK
SYNCH RON I Z E R

SIGNAL
BUS
CABLE

UNIT BUS CABLES

SIGNAL
BUS
CABLES
RPOI/02

RPOI/02

----- ~---.... ------

AC POWER

UP TO
EIGHT
10-01-62

Figure 1-3 RPIO/DFIO/RPOI/RP02/I/O Interface

OF 10
DATA
CHANN E L

WOl2 WOl2

C"I
A B L E f CABLE
t
I A
I B #2

w
-'~#
u

"t::~
o~

-

'"
~

o
~

-

It)

0

'"

PDP-IO
I/O
BUS

II

G796 G796

IX)

C\I

~

";:~
~
'" ~
~

'"

N

CD

CABLE
#3

W250 W250

G796 G796

WB51 W851

W021 W021

'"

it)

It)

IX)

IX)

~

~

it)

It)

IX)

IX)

~

-

'"

~

RPIO
SYNCHRONIZER

IX)

'"

~

~

~ on
IX)

'"on

~

~

~

It)

RP 01/02
DISK
PACKS

on

IX)

IX)

w

w

-''''
~II

.:.Jro

~II

u

u

10-0163

Figure 1-4 RP lO/RPO I /02 Electrical Interface
1-4

Table 1-2
RPI0/RPOI and RP02 Interface Chart
Between
RPI0

Function

RPOI/02

Cable Type

P/R-21

13

Address and Control Bus Lines
Cylinder Address Bus Lines

G796-G796

P/R-22

15

Tag lines, Unit Select, Write Data,
+36 Vdc in, Controlled Ground,
'Attention', Unit Selected, Ready,
On-Line, Index Pulse, Sector Pulse,
File Unsafe, Seek Incomplete, End
of Cylinder, Read Data, Read Only,
Write Current Sense.

G796-G796

The RPIO feeds the Disk Pack with eight multifunction address and control bus lines, three tag lines, a unit select
line, a write data line, sequence out, +5V to terminator, and controlled ground. The Disk Pack feeds the RPIO
with 'attention', ready, on-line, index pulse, sector pulse, file unsafe, seek incomplete, end of cylinder, read data
coax, eight cylinder address lines, read only, and selected unit sector pulse.
1.3.2 RPlO/DFI0 Interface
There are three cable assemblies through which the RPIO connects to the DFIO Data Channel (see Figures 1-5
and 1-6). These assemblies and their functions are listed in Table 1-3.
Table 1-3
RPIO/DFIO Interface Chart
Between
RPIO
DFIO

Function

Cable Type

P/R-1/2

K/L-31/32

Chan Cable No. 1

W250-W012

P/R-ll

K-21

Chan Cable No.2

G796-G796

P/R-12

L-21

Chan Cable No.3

G796-G796

1.3.3 RPIO/I/O Interface
There are two cable assemblies through which the RPIO connects to the PDP-lO I/O Bus (see Figures 1-5 and 1-6).
These cable assemblies are listed in Table 1-4; the signals they carry are also included.

1-5

Table 1-4
RPI0/1/0 Interface Chart
Between
RPIO

Cable Type

Function

I/O Bus

P/R-13/
14/17/18

Cable I

lOB 00-3S

W2S0-WOI2

P/R-IS/
16/19/20

Cable 2

lOB Reset, IOBD DR SPLIT, IC
lOS 03-09, DATAO, DATAl, CONO,
CONI, RDIPLS, RDl DATA, PI 1-7

W2S0-WOI2

1.3.4 Disk Pack/Disk Pack Interface

.

Two cable assemblies run serially between all Disk Pack Units. These cables carry sequence out plus all signal and
control interconnections between the RPlO and the first Disk Pack. The sequence out from the RPlO is relayed
to the next drive when this drive is up to speed and is drawing its running current (not more than SA, RPO I, or 7 A,
RP02).

RPOI

RPIO

RP02

RPIO

10-0177

Figure I-S RPIO Interface Wiring Diagram

1-6

.

DF10 DATA CHANNEL

N

~

~~
.

Z

.

.
w
~

"

;;

;

u

r

W

N

~

~

•

0

•

"

0

1

~

;;

>

~

~

~>

~

! 0~
~

%

u

~
o

z
u

•

~

~

>

~

•
u

~

z

c

8

0

'"
'"

.. ·", ! 8~ E;;: .., "
~

2

>
0

~

>

%

u

u

~

~

Z
0

u

~

~

0

0

·" . : .

c c
u

~

~

>

2

%

z z
z z
%

~

~

>

~

~

0

~

0

~ -'

e : :>c
z

>

~
~

"z "z

~

c

~

~

z
~

>

~
~

0
~

'""
~

z

c z z > z
z > c c ~ c%
c % % ~ U
~

•8

0

u

..
0

~
>
c ~

u

u

u

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ u u

~
0

•z

~

>

"

0

~

~

0

u
u

~

~

0

"

~

z
~

>

~

~

.
.• .
0

~
~

~

0

"
~
~

"

>
~
c

~

~

0

"

~
~

.. .. · .
" .. .. ..
n

N

z ~
~
> ~
z ~ z 0 u > z N ~
u ~ z
> z u
~
z 0 0
>
c ~ c ~ ~
c z Z
>
%
%

•.

8

~

0
~
~

or

> c >
~

u
u

~

u
u

u
u
u

• •
u

u
u

u
u

u
u

..

w

0

u
u

~

%

u

~

%

%

u

u

or

~

N

~

0

~

0

0

0
0

0

0
0

0

0

~

N

~

0
0
z z z z% z%

0

%

u

~ ~ c c ~ c

~

.·.
%

u

~

%

u

u

u

or

~

c

!;

W250

WOIZ

lOB 00- 08
R13(W85,1

10809-17

0

lOB 00-08

PI7(W8511

~

10809-'7

R17(W85t1

N

lOB 18-26

l.A8L!:;"

P14(W851l

G796

~

osec ue

A14(W8511

osec ue
osec ue

R18(W8511

ossc us

4

ossc us

5

ossc us

6

-H'-":.:':::"::'-'#:..:''------H-'
-I+-------H-'
-I+-------H-'

Ie IDS 03(0)-05101

ossc us

7

-Y--------Hof

Ie IDS 05111-09111

OSSC CYL ADR REG 01

W250

lOB 27- 35
Ie 108 RESET

W250

} •••• w.")

IOSD OR SPL IT

RISlw8511

lOB OR SPLIT

...

}

p ••

(we."

Ie IDe DATAO SET

I

)""..."

Ie lOB CONO CLEAR

Ie lOB CONO SET
Ie lOB OATAI
CABLE #2

CABl£IIZ

Ie lOB CONI

RPIO DISK
SYNCHRONIZER

_CAIILE#3-

~I+----+---I
12B-IlJ----G~7'_i.~.'__l

osec CYL ADR REG 64

OSBC SEQUENCE OUT

P22 (We.21!

(SHIELD GROUND) ------++-+__~
OSBC SET HEAO TAG LINE ---+l-!-tJ~

(108 ROI PULSE)

osec SET CYL TAG LINE

IIOB RDI OATAI

OSBC CONTROLS TAG LINE

RISlw8511

CXR lOB PII-7

OSSC SEL UNIT READY

:.:
UJ

o

OSBC CYL AOR REG 32
OSBC CYL ADR REG

:.:

~

CY L ADR REG 04

osec CYL ADR REG 16

A19(W8511

Ie IDS OM 1i-09(1)

UJ
(J

OSSC CYL ADR REG 08

Ie lOB OATAO CLEAR

11.
0
11.

OSSC

Ie IDS 03(01-0510)

0

I

ossc cY L ADR REG 02

Ie rOB RESET

0

_A-_ _ _ _ _ _ _G_7 .....:.J

I

PI8(W95' I

lOB 27-35

lOB 18-26

.....

DSRC uR 0

---++-+~

---~++--I

N

o

11.
II::

.....

o

11.
II::

::c,:-_+f--I--I
-----'--='-'-''---1++-+

OS BC S E L UN I T ON 1I N E "-,,"-,--:,,:..

Ie lOB OATAO CLEAR

)... ...

Ie lOB OATAO SfT
Ie IDe CONO CLEAR

, ,

Ie lOB CONO SET
Ie lOB DATAl
IC 10e CONI

osec 5EL INDEX

osec SEL FILE UNSAFE - - - - + + - + - - - 1
OSBC SEL UNIT SK INC

----1++-+

OSSC SEl EO CYL -----++""G7'*."""'.--1
osec SEL SEC PLS

-----4J.="'--I

(loe RDI PULSE)
W250

A20 (W851)

to-OUSt

Figure 1-6 RPIO Cable Diagram

1-7

1.4 EQUIPMENT SUPPLIED
Table 1-5 lists the equipment supplied with the RPIO Disk Pack Synchronizer.
Table 1-5
Equipment Supplied
Quantity
Per
Equipment
1

Description

Part Type No.

Synchronizer

Quantity
Per
Equipment

Description

I

Cable G796 to G796

C-IA-7005469-13

RPIO

Part Type No.

2

Cables W250 to WOl2 D-IA-7005459-12

I

Cable G796 to G796

C-IA-7005469-6

5

Cables W250 to WO 12 D-IA-7005459-24

I

Cable G796 to G796

C-IA-7005469-7

2

Cables W250 to WOl2 D-IA-7005459-15

1

Cable, Grounding

1

Cable W250 to WO 12 D-IA-7005459-8

I set

Spare Modules

2

Cables W250 to WO 12 D-IA-7005459-11

1 set

Diagnostic Tapes

MAINDEC 10-D50B
MAINDEC 10-DSMA
MAINDEC 10-D5NB

2

Cables G796 to G796 C-IA-7005469-10

1 ea.

Maintenance Manual
(2 volumes)

DEC-IO-H5EB-D

1

Cable G796 to G796

1 set

Engineering Drawings

C-IA-7005469-2

1.5 REFERENCE DOCUMENTS
The following documents supplement the information contained in this manual:
PDP-IO System Reference Manual
DFIO Data Channel Maintenance Manual
PDP-1O Site Preparation Guide
PDP-IO Interface Manual
Vendor Manual for RPOljRP02 Disk Pack Drive
This material is available from the nearest DEC Field Office or from:
Digital Equipment Corporation
Direct Mail Department
146 Main Street
Maynard, Massachusetts 01754

1.6 GLOSSARY
The abbreviations used throughout this manual are defined in Volume II of this manual.

1-8

Chapter 2
Installation

2.1 GENERAL
This chapter contains information required for installation of the RPIO Disk Pack Synchronizer. The installation
procedures are not complex because the RPIO is shipped complete and factory tested. When the equipment is
received, all modules are in place and, all intra-bay cabling has been installed. The unit has been tested extensively while operating as part of a standard PDP-I 0 System. Turn-on and checkout procedures are included herein to conflfnl operation of the equipment after it has been installed.
The equipment rack should be located as near as possible to the DFIO Data Channel, preferably next to the DFIO
frame. Bolt the equipment rack to the main rack in four places using the hardware provided. Cable lengths and
the location of the RPOI/RP02 drives are the major factors governing placement of the RPIO.
2.2 UNPACKING
The following paragraphs describe the RP I 0 unpacking procedures.
2.2.1 Special Handling
No special handling procedures are required for the RPIO beyond the normal care afforded any piece of scientific
equipment of comparable size and weight. However, particular care should be exercised in the use of cranes or
hoists to prevent damage to the unit.
2.2.2 Inspection
On receipt of the equipment, inspect it for visible damage such as dents and abrasions that may have occurred in
transit. Inspect the logic modules for foreign matter that may have lodged in them during shipment. Any damage
should be reported immediately to both the Carrier and Digital Equipment Corporation. Check the contents of
the carton with the shipping document and with Table 1-5 of this handbook. Immediately report any omissions
or incorrect parts to Digital Equipment Corporation.
2.2.3 Power Requirements
The RPIO is equipped with self-contained power supplies and power control. The + 10 Vdc and -IS Vdc requirements of the unit are provided by four DEC Type 728 Power Supplies and a DEC Type 844 Power Control.
Modifications for accommodating either 50- or 60-Hz operation are made within the basic unit. These are designated as RPlO-A for 60-Hz and RPIO-B for 50-Hz operation. Power is supplied from an 8A, liS Vac, single-phase
source. The RPO I /02 Drives require a three-phase line. Power requirements per phase are listed in Table 2-1.

2-1

Table 2-1
RPOI/RP02 Power Requirements
Type of Drive

Number of Drives

RPOI

1-3
8A

4-6
16A

7 or 8
23A

RP02

12A

24A

36A

NOTE
Three-phase power is required for from one through any
number of drives. Each unit is a single-phase device but is
connected on a phase rotation basis.
PDP-I 0 Systems operate from three-phase, Y-connected, 4-wire plus earth ground power mains. Standard voltages
are liS /200 Vac ±10%, 60 Hz ±1/2 Hz and 230/400 Vac ±10%, 50Hz ±I /2 Hz. Other common voltages can also
be accommodated. A system earth ground connection must be supplied through the power cords in addition to
the ground bus requirement.
The RPIO is supplied with terminal lugs, a 3-wire cord, and in North America, a 5-wire Hubbell #3521 male plug
(mates with Hubbell #3520). One wall power cord is required for every RPOI/02.
Terminal lugs are suitable for No. IOta No. 18 AWG wire (0.04 to 0.12 in., I to 3 mm).
2.3 INSTALLATION PROCEDURE
To install the RPlO, proceed as follows:
Procedure

Step

Ensure that all power is removed from the PDP-I 0 System and from the
RPO 1/02 Disk Packs.
2

Remove the shipping skid from the bottom of the rack and discard it.

3

Remove the front and rear panel doors, and both side panels from the
RPIO.
NOTE
Overall dimensions are given in Figure 2-1.

4

Set the rack in place next to the OF I 0 Data Channel and lower the floor
pads on both units to prevent rolling.

S

Bolt the RPIO rack (in four places) to the DFIO rack already in place.

6

Install the grounding cable to the bottom horizontal rack member. Tie
the ground lead to the next grounded rack or to a system grounding bus.

7

Install separate grounds (No.4 standard PDP-IO ground bus) from each
drive to the RPIO (star ground system).

8

Connect cables according to the site plan and cabling list supplied with
each installation (see Figure 2-2).

2-2

Procedure

Step
8 (Cont)

NOTE
Inspect all cable connectors for shipping and
handling damage; install them carefully and
securely. All power and signal cables should
be routed 1 ft apart, if possible. Power and
signal cables should cross at a 90° angle.

9

Locate the drives and level each to within 5° .

10

Connect the remote tum-on power cable from the lower REMOTE
CONTROL power outlet on the power control unit of the preceding
rack to the upper REMOTE CONTROL power receptacle of the
RPIO.

11

Connect the Margin Check Remote Control Cable from J 1 of the
preceding Power Connector Assembly Bracket to 12 of the same bracket
on the RPIO.
NOTE
If the RPIO is the last unit in the Margin Check bus,

install the terminating plug in 11 of the RPIO Power
Connector Assembly Bracket.
12

Connect the RPIO power cable, already connected and secured to the
Power Control Unit, to the system source of ac power.

13

Install any I/O bus termination plugs required (refer to DEC-lO-HIFB-D,
PDP-I0 Interface Manual).

14

Replace side panels and panel doors.

15

On the drives, carefully remove tape from heads (installed for shipping
purposes).

16

Inspect each head to ascertain that it is resting on the proper launching
block.

17

Before attempting ON-LINE operation, Head and Servo Alignment must
be checked and, if necessary, adjusted with C.E. Packs and an OFF-LINE
Tester. See considerations in Paragraph 5.3.4. To load a disk pack on
the drive, refer to Paragraph 3.3.1.

18

Using the OFF-LINE Tester perform the following steps on each drive as
described in the Vendor Maintenance Manual:

1.
2.
3.

Adjust voltage and high/low voltage sense.
Adjust positioner as described in Vendor Manual.
Perform head alignment with C.E. Pack after completing
a Thermal Equilibrium Procedure as described in Paragraph
5.3.4.2d.
NOTE
When using the OFF-LINE Tester, do not have
either a cable or terminator block on 'signal out'.

2-3

114--- 21" ---I
------[----

27"

------]

---I

69"

FRONT VIEW

SIDE VIEW
10-0019

Figure 2-1

RPIO Overall Dimensions

2.4 TURN-ON AND CHECKOUT PROCEDURES
When the RPIO has been installed, verify its operation by proceeding as follows:
Step

Procedure
Set all switches on switch panel to the down position. Make certain
that all margin check switches are in the OFF position.

2

Apply power to PDP-IO computer system and RPOI/02 Disk Pack Drives.
The POWER indicator lamp should light.

3

Set LOCAL/OFF/REMOTE switch on the DEC Type 844 Power Control
panel to LOCAL position.

4

Set PWR circuit breaker and the power switch located on Power Control
panel to ON. The cooling fans should begin to operate.

5

Refer to the MAINDEC Diagnostic Routine and RPIO System Test
Procedure to conduct final checkout of RPIO Synchronizer.
NOTE
Run diagnostics and reliability programs using scratch
packs or maintenance packs. Scratch packs used for
reliability testing must be Mapped before they can be
used with Monitor system because the entire surface
is subject to writing by Reliability Test.

2-4

Step

Procedure

6

Refer to Chapter 5 for any adjustments required during, or as a result of,
this installation procedure. Adjustment procedures for the drives are
contained in applicable Vendor Manuals.

RP10


\
I
_I I I I I I
ONE R PO 1 TRACK

0

I

2

3

~~~TORS
TRACK

4

""O,~

/

132 WORDS

~

HEADER

-/..

~TRAILER

DATA FIELD

'tJ
o

SYNCHRONIZATION

DATA WORD

o

DATA WORDS
1-126

DATA WORD
127

CONTENTS
TAIL
LPR
(POSTAMBLEI

10- 0166

Figure 3-4 RPOI/RP02 Data Format

Data formats for information recorded on the disks are arranged as shown in Figure 3-4. These data formats
consist of a twenty-word header field on the RPO I (thirty-two on RP02), a data field of 132 words, followed by
a two-word trailer.
The first 18 words in the RPO 1 header field (30 words in RP02) consist of zeros to allow the read circuitry to
synchronize with the data. The 19th (31 st on RP02) word contains an octal one (000000000001). The last
header field word comprises the addressing field.
The first three words in the data field function as a READ DISENABLE/WRITE ENABLE field. The next data
field word is used for synchronization, and contains an octal one (00000000000 I). These words are followed
by 128 36-bit words, each followed by its own 37th parity bit.
The first word in the trailer field contains the contents of the longitudinal parity register, and the last word (tail
or post amble) contains a second LPR.
NOTE
When executing a Write Header command, the lrrst 24 words in
an RPOI or the lrrst 36 words in an RP02 are written without
parity. Therefore, both bits must be in data and the lrrst bit of
the word following the header is interpreted as the parity bit of
the header when read back.
Addressing is effected by the 18-bit partitioned binary number, defined as word two in the header. The partitioning has been selected to provide sequential access to the entire memory with a minimum of positioning operations.
The Exec area (Figure 3-4) is determined by the executive program (monitor, operating system) and is ignored by
the interface, except for parity checking.
The cylinder address, an unsigned binary number which functions as the track address (positioner), contains eight
bits to describe the 313 8 tracks. The outer track cylinder is represented by 000 8 , and the inner track cylinder is
represented by 312 8 .
The surface address, a five-bit, unsigned binary number, selects one of 10 disk surfaces for the RPO I (codes 008
to 118 are legal), and selects one of 20 disk surfaces for the RP02 (codes 00 8 to 23 8 are legal). Other codes
are illegal (nonexistent).
The sector address, a five-bit, unsigned binary number, selects the proper sector on a track. One of five sectors is
selected on the RPO I (codes 00 8 to 048 are legal), and one of 10 sectors on the RP02 (00 8 to 118 are legal). Other
codes are illegal (nonexistent).
In the RPO I /02 there is one record per sector (see Figure 3-4). The sector mark defines the beginning of a record.
All records are fixed length with a data field of 128 36-bit words plus parity. A header field associated with each
record is used to identify the record and ensure that cylinder, surface, and sector selection is correct before writing
or reading a record.
3.4.2 RPIO Programming
The RPIO modes of operation are position (seek cylinder), recalibrate, read data, write data, write header and
data, and clear attentions. Channel instructions must be placed in memory prior to initiating a data transfer in
the synchronizer.

3-11

3.4.2.1 DATAO Information (see Figure 3-5) - A DATAO 250 instruction either performs a positioning operation or initiates a data transfer, if the synchronizer is not already busy. The RPIO is capable of executing the
six instructions mentioned in the preceding paragraph. The instructions are described below (DATAO 254 is
used if a second RP lOis installed in a PDP-I 0 System).
Position Instruction (seek cylinder) - The Position instruction word, shown in Figure 3-6, is
divided into four fields:

Q.

( 1) Op Code: The Op Code is 4 8 ,

(2) Drive: The Drive field indicates the disk drive addressed. All codes are legal if the drives

exist.
(3)

Cylinder: The Cylinder field specifies the cylinder sought. Codes 000 8 through 3128

are legal. Others are nonexistent.
(4) Bits 14-35: These bits are spares.

Execution of a position instruction requires 3 to 5 IlS of synchronizer time and 20 to 150 ms
of drive time. Positioning is controlled by a difference count of the present address register
and the cylinder address register that is loaded by this instruction. When a zero difference
count (compare) is reached, the heads are positioned.

o

23

13 14

56

5 BITS

8 BITS

\3 BITS 13 BITS 1

23 24 25

18 19

1
0 1 2 0 1 2 0 1 2 3 4 5 6 7 o I 23 4
\:.........-I '---..",--J ~ ~
OP
DRIVE
CYLINDER
SURFACE
OR SPARES
OR SPARES

I

o

5 BITS
I 234

'-----'
SECTOR
OR SPARES

26

27

I

34 35
8 BITS

B:TI B:TJ BIT

76543210
~
ICWA OR
ATTENTIONS

I
BIT

WRITE EVE~OR
PARITY INTO SPARE
MEMORY

LONGITUDINAL}
CHANNEL DATA
DISAB LE PARITY
ERROR STOP
DISK WORD

Figure 3-5 DATAO Instruction Word Format

o

2 3

1314

5 6

22 BITS
OP

DRIVE

CYLINDER

i

Figure 3-6 Position Instruction Word
b.

Recalibrate Instruction - the Recalibrate instruction word, shown in Figure 3-7, is divided into three
fields:
(J) Op Code: The Op Code is 78'
(2) Drive: The Drive field indicates the disk drive addressed. All codes are legal if the drives exist.
(3)

Bits 6-35: These bits are spares.

3-12

Execution of a recalibrate instruction moves the positioner ,to cylinder zero by force. Positioning is not controlled by the difference count from the current address register, which may be
wrong.

c.

Read Data Instruction - The Read Data instruction word, shown in Figure 3-8, is divided into
ten fields:
( 1)

Op Code: The Op Code is 0 8 ,

(2)

Drive: The Drive field indicates the disk drive addressed. All codes are legal if the drives
exist.

(3)

Cylinder: The Cylinder field specifies the cylinder on which the data to be read resides.
Codes 000 8 through 3128 are legal. Other cylinders are nonexistent.

NOTE
The heads must have been previously positioned; this field
is for verification only.
(4) Surface: The Surface field indicates the recording surface on which the data to be read
resides. Codes 008 through 118 are legal on RPO 1. Codes 00 8 through 23 8 are legal on
RP02. All others are nonexistent.
(5) Sector: The Sector field indicates the sector on which the data to be read resides. Codes
00 8 through 048 are legal on RPO 1. Codes 00 8 through 118 are legal on RP02. All others
are nonexistent.
(6) Bit 24: This bit is set to 1 to disable longitudinal parity error stop. If not disabled, the
discovery of a longitudinal parity error stops data transmission and sets the longitudinal
parity error bit.
(7)

Bit 25: This bit is a spare.

(8)

Bit 26: This bit is set to 1 to disable the disk word parity error stop. If not disabled, the
discovery of a disk word parity error stops data transmission and sets the disk word parity
error bit.

(9) ICWA: The IeWA field specifies bits 27-34 of the Initial Control Word Address. All binary
values above 178 are legal.

NOTE
The channel cannot access the fast registers.
(10) Bit 35: This bit, when set, causes the channel to write even parity into memory on all data

words read from the disk.
The Read Data instruction causes data to be read until either the channel shuts down or the end
of the cylinder is reached.
o

2 3

56

35
30 BITS

~

OP

DRIVE

SPARES
(0-0170

Figure 3-7 Recalibrate Instruction Word

BBITS
CYLINDER

Figure 3-8 Read Data Instruction Word
3-13

d.

Write Data Instruction - The Write Data instruction word, shown in Figure 3-9, is divided
into ten fields:

( 1) Op Code: The Op Code is 18 .
(2) Drive: The Drive field indicates the disk drive addressed. All codes are legal if the drives
exist.
(3) Cylinder: The Cylinder field specifies the cylinder on which the data is to be written.
Codes 000 8 through 3128 are legal. Other cylinders are nonexistent.

NOTE
The heads must have been previously positioned. This field
is for verification only.

(4) Surface: The Surface field indicates the recording surface on which the data is to be
written. Codes 00 8 through 118 are legal on RPOI. Codes 00 8 through 23 8 are legal
on RP02. All others are nonexistent.
(5) Sector: The Sector field indicates the sector on which data is to be written. Codes 00 8
through 048 are legal on RPO I. Codes 008 through 118 are legal on RP02. All others
are nonexistent.
(6) Bit 24: This bit is a spare.
(7)

Bit 25: This bit is set to I to disable channel data word parity error stop. If not disabled,
the discovery of a channel data word parity error stops data transmission and sets the
channel data word parity error bit.

(8)

Bit 26: This bit is a spare.

(9) lCWA: The ICWA field specifies bits 27-34 of the initial control word address. All binary

values above 178 are legal.

NOTE
The channel cannot access the fast registers.
( 10) Bit 35: This bit is a spare.
This instruction causes data to be written until either the channel shuts down or until the end
of the cylinder is reached.

NOTE
If the last (or only) sector to be written is shorter than
128 words, the remainder of the field is fdIed with words
"aU zeros."

Figure 3-9 Write Data Instruction Word

3-14

e.

Write Header and Data Instruction - The Write Header and Data instruction word, shown in
Figure 3-10, is divided into nine fields:
(1)

Op Code: The Op Code is 38'

(2) Drive: The Drive field indicates the disk drive addressed. All codes are legal if the drives
exist.
(3)

Surface: The Surface field indicates the recording surface on which the headers and data
are to be written. Codes 008 through 118 are legal on RPOI. Codes 00 8 through 23 8
are legal on RP02. All others are nonexistent.

(4) Bits 19-24: These bits are spares.
(5)

Bit 25: This bit is set to 1 to disable channel data word parity error stop. If not disabled,
the discovery of a channel data word parity error stops data transmission and sets the
channel data word parity error bit.

(6)

Bit 26: This bit is a spare.

(7) ICWA: The ICWA field specifies bits 27-34 of the initial control word address. All binary
values above 178 are legal.

NOTE
The channel cannot access fast registers.
(8)

Bit 35: This bit is a spare.

This instruction formats the disk. The RPIO searches for an index mark, then writes header and
data fields as shown in Figure 3-4. The programmer supplies headers and data as described in
Paragraph 3.4.1. The synchronizer writes one track of headers and data and adds the data word
parity bits and the trailer field.

f.

Clear Attentions Instruction - The Clear Attentions instruction word, shown in Figure 3-11, is
divided into five fields:
(1)

Op Code: The Op Code is 58'

(2) Drive: The Drive field indicates the disk drive addressed. All codes are legal if the drives
exist.
(3)

Bits 6-26: These bits are spares.

(4)

Bits 27-34: These bits clear attentions 0-7, respectively.

(5)

Bit 35: This bit is a spare.

a alTS
CYLINDER

Figure 3-10 Write Header and Data Instruction Word

o

2 3

56

ttl
OP

DRIVE

2627
21 BITS

3435
BBITS

~

"

SPARES

ATTENTIONS

Figure 3-11 Clear Attentions Instruction Word

3-15

' [ SPARE

g.

No Op Instruction - The No Op instruction word, shown in Figure 3-12, is divided into three
fields:

(1)

Op Code: The Op Code is 6 8 ,

(2) Drive: The Drive field indicates the disk drive addressed. All codes are legal if the drives
exist.
(3) Bits 6-35: These bits are spares.
o

2 3

35

56
30 BITS

SPARES

Figure 3-12 No Op Instruction Word

3.4.2.2 DATAl Information - A DATAl 250/254 Instruction supplies the programmer with information concerning the disk and cylinder selected, as defined below and illustrated in Figure 3-13 (250 for RPOI or 254 for
RP02):

a.

Disk Selected - This field indicates the disk most recently selected by a legal DATAO. lOB 0-2

b.

Cylinder Address Register - This field displays the contents of the drive's cylinder address
register. lOB 3-10

c.

Seek Incomplete - This bit indicates that the drive was unable to complete the most recent
position command addressed to it. The error recovery procedure is: issue a recalibrate command,
which clears this error condition. lOB II

d.

On Cylinder - This bit indicates that the heads are positioned on the cylinder specified by the
cylinder address register bits of this instruction word, and that data transfer may proceed. lOB 12

e.

On Line - This bit indicates that]) the ENABLE/DISABLE switch is in the ENABLE position,
2) a disk pack is mounted, 3) tl!le dust cover is closed, 4) the spindle is up to operating speed,
and 5) the heads are mounted on the disk. IO B 13

f.

File Unsafe - This bit indicates that either 1) a head is selected when not on a cylinder, 2) more
than one head has been selected simultaneously, 3) read and write were selected at the same time,
4) write and/or erase were selectedwhenthe file was not ready, 5) write current was sensed when
write was not selected, 6) write was selected with no write current, 7) erase current was sensed
without erase being selected, or 8) write was selected without erase current. lOB 14
NOTE
The interface provides hardware to prevent any of these
conditions from being caused by programming errors.
To clear this bit, depress the STOP and START buttons.
A full power-down/power-up sequence, requiring over
one minute, must follow.

g.

No Such Drive - When this bit is true, it indicates that the drive addressed does not exist. In
this case, the other bits of this instruction word are meaningless. lOB 15

h.

Read Only - When this bit is true, it indicates that the drive's READ-WRITE/READ ONLY
switch is in the READ ONLY position. lOB 16

i.

Write Header Lockout - This bit, when set, prevents headers from being written. lOB 17

3-16

j.

Sector Counter - These bits indicate the state of this disk's sector counter. The sector counter
is cleared by an Index Pulse and counts sector pulses. lOB 18-22

k.

Attentions - These indicate which drives are requesting an interrupt due to Seek Incomplete or
Seek Complete. lOB 27-34

3.4.2.3 CONO Information - A CONO command is illustrated in Figure 3-14 and described below:

a.

Clear: These bits perform the following clear operations;
(1)

Bit 20: When this bit is set to aI, it clears the power supply failure (PS FAIL) flag

(2)

Bit 21: When this bit is set to aI, it clears the search error (SRCH) flag.

(3)

Bit 22: When this bit is set to aI, it clears the overrun (OVERRUN) flag.

(4)

Bit 23: When this bit is set to aI, it clears the no such memory location (NXM) flag.

(5)

Bit 24: When this bit is set to aI, it clears the Channel Control Word Parity Error

only if the power supply fail conditions have been corrected.

(CCPE), the Disk Sector Parity Error (DSPE), the Disk Word Parity Error (DISK WDPE),
and the Channel Data Parity Error (CDPE) flags.
(6)

Bit 25: This bit is a spare.

(7)

Bit 26: When this bit is set to aI, it clears the Illegal Write (ILL WR) flag.

(8)

Bit 27: When this bit is set to a 1, it clears the Illegal Command While Busy (ILL COM) flag.

(9)

Bit 28: When this bit is set to a 1, it clears the Sector Designation Error (SEC DE) flag.

(10) Bit 29: When this bit is set to aI, it clears the Surface Designation Error (SURF DE) flag.

o

1 23

1 11
BIT BIT BIT

8 BITS

~

DISK
SELECTED

CYLINDER
ADDRESS
REGISTER

WRITE HEADER LOCKOUT
READ ONLY
NO SUCH DRIVE
FILE UNSAFE
ON LI NE
ON CYLINDER
SEEK INCOMPLETE
10-0169

Figure 3-13 DATAl Instruction Word Format (Revised)
o
20 BITS

SPARES
STOP
WRITE CHAN"IEL
CONTROL INTO
MEMORY/CLEAR
CONTROL WORD
TRANSFER COMPLETE
10-0171

Figure 3-14 CONO Instruction Word Format

3-17

b.

Write Channel Control Word into Memory: This bit causes the data channel to store the current
contents of the data address register and the control word address register into memory location
B + I where B (an even number) is the Initial Channel Control Word Address. The corresponding
CONI bit signifies completion of the operation. Any channel termination causes the control word
to be written, but CONI bit 30 will only be set when the operation was requested by the synchronizer
via CONO. This bit, when it is set to a I, clears the Control Word Transfer Complete (CWX: CaMP)
flag. lOB 30

c.

This bit stops transmission immediately. Part of last word and last sector transmitted is in determinant. lOB 31

d.

Clear: This bit clears the Done (DONE) flag. IOB32

e.

PI 0, I, 2: (PI Assignment) These bits assign a priority interrupt channel to the synchronizer. lOB 33-35

3.4.2.4 CONI Information - A CONI command is illustrated in Figure 3-15 and described in the following
paragraphs:

I

a.

Channel Control Word Parity Error - This bit indicates that the Channel Control Word contained a
parity error. This error condition cannot be disabled. The error flag is cleared by a CONO with
bit 24 set to a 1. lOB 14

b.

Disk Sector Parity Error - This bit indicates that the checksum bits, at the end of a sector which
has been read by the most recent read data command, do not yield a modulo-two sum of 1. Unless disabled, this error terminates data transmission. The error flag is cleared by a CON a with
bit 24 set to a 1. lOB 15

c.

Channel Data Word Parity Error - This bit indicates that the word read from memory during
execution of the most recent write data or write header and data instruction contained incorrect
parity. Unless disabled, this error terminates data transmission. If the instruction was write data,
the remainder of the sector is filled with Os and correct parity is added. If the instruction was
write header and data, the results are indeterminate. lOB 16

d.

Disk Word Parity Error - This bit indicates that the word read from the disk during execution of
the most recent read data instruction contained incorrect parity. Unless disabled, this error
terminates data transmission. lOB 17

e.

Search Complete - This bit indicates that the search for the sector to be read or written has
been completed. lOB 18

f.

End of Cylinder - When this bit is set to I, it indicates that the most recently executed read data,
write data, or write header and data instruction attempted to transfer too much data and, as a
result, incremented the head address register past the last surface on the disk. On the RPO I, the
register is incremented to 128' On the RP02, the register is incremented to 24 8 , This bit is cleared
by the next data transfer instruction to the same disk drive. lOB 19

g.

Power Supply Failure - This bit indicates that the interface power supply voltages are out of
tolerance. When this occurs, the interface terminates at the end of the current sector and turns
on both the done bit and the power supply failure bit. The power supply failure bit cannot be
cleared until the error condition has been corrected. lOB 20

h.

Search Error - This bit indicates that the cylinder, surface, and sector fields of the most recently
executed read data or write data instruction could not be matched (compared) with corresponding
fields of any sector on the addressed track. This could occur because either 1) there was a bad spot
in the header field, 2) the instruction contained the wrong fields, 3) the program did not previously
position the heads to the proper cylinder, 4) the disk drive positioning mechanism is in need of adjustment, 5) the PDP-IO System has failed, or 6) a header was read incorrectly after search complete
was set during a multiple sector transfer. lOB 21

3-18

o

13

14 BITS

SPARES

14

15

16

'7

18

19

20

21

1
1
1
1
1
1
1
1
BIT BIT BIT BIT BIT BIT BIT BIT

22

23

24

25

26

27

28

29

30

31

32

1
1
1
1
1
1
1
1
1
1
BIT BIT BIT BIT BIT BIT BIT BIT BIT BIT

35
14 BITS

l ,~

PI
BUSY
CONTROL WORD TRANSFER COMPLETE
SURFACE DESIGNATION ERROR
SECTOR DESIGNATION ERROR
ILLEGAL COMMAND WHILE BUSY
ILLEGAL WRITE
DISK NOT READY
PARITY ERROR (14-17)
NO SUCH MEMORY LOCATION
CHANNEL TOO SLOW
SEARCH ERROR
POWER SUPPLY FAILURE
END OF CYLINDER
SEARCH COMPLETE
DISK WORD PARITY ERROR
CHANNEL DATA WORD PARITY ERROR
DISK SECTOR PARITY ERROR
CHANNEL CONTROL WORD PARITY ERROR

l

10-0167

Figure 3-15 CONI Instruction Word Format (Revised)

i.

Channel Too Slow - This bit indicates that the channel failed to respond to a request for a data
word, or failed to accept a data word fast enough during a data transfer. If the instruction was
a read data, data transmission is terminated. If the instruction was a write data, the remainder
of the sector is filled with Os and proper parity is added. If the instruction was a write header
and data, the results are indeterminate. lOB 22

j.

No Such Memory Location - This bit indicates that the channel attempted to access a nonexistent
memory location. lOB 23

k.

Parity Error (14-17) - This bit (24) is true when anyone of bits 14 through 17 are true. Clearing this bit clears bits 14 through 17. lOB 24

I.

Disk Not Ready - lOB 25

m.

Illegal Write - This bit indicates that a write header and data or write data instruction was received
which addressed a disk drive whose READ-WRITE/READ ONLY switch was in the READ ONLY
position, or a WRITE was in process when sector pulse was received. lOB 26

n.

Illegal Command While Busy - A DATAO to the RPIO was issued while BUSY flag was on. The
second command is ignored. lOB 27

o.

Sector Designation Error - This bit indicates that the sector addressed by the most recent read
data or write data instruction does not exist (greater than 048 on the RP01 and greater than 118
on the RP02). lOB 28

p.

Surface Designation Error - This bit indicates that the surface addressed by the most recent read
data, write data,. or write header and data instruction does not exist (greater than 118 on the RPOI
and greater than 23 8 on the RP02). lOB 29

q.

Control Word Transfer Complete - This bit indicates that the channel control word was written into
memory only when that request was made by the synchronizer via CONO. lOB 30

3-19

r.

Busy - This bit indicates that the synchronizer is currently executing a data transfer command. lOB 31

s.

PI Enable - This bit indicates that the PI condition has been enabled. It is the result of (DONE
VATTENTIONS)I\-BUSY. lOB 32

t.

PI 0, 1, 2 - (PI Assignment) These bits indicate the priority interrupt channel currently assigned
to the RPIO. lOB 33-35

3-20

Chapter 4
Theory of Operation

4.1 GENERAL
This chapter discusses the principles of operation of the RPIO Disk Pack Synchronizer. A general block diagram
discussion, a brief sequence of operation, and functional descriptions of the various operational circuits used in
the synchronizer are also included.
4.2 GENERAL BLOCK DIAGRAM DISCUSSION
The RPIO Disk Pack Synchronizer is shown in general block diagram form in Figure 4-1. The RPIO acts as an
intermediary between the I/O Bus, the DFIO Data Channel, and the RPOI, or RP02 Disk Pack Drive. For
purposes of this discussion, a knowledge of the PDP-I 0 I/O Bus, the DF I 0 timing diagrams, and the operation
of the RPO I /02 Disk Pack Drives is assumed.
The PDP-I 0 Processor communicates with the RPIO Disk Pack Synchrcrmizer via four basic I/O instructions;
]) Conditions Out (CONO) which transmits the control word to the RPIO control register thereby specifying the
desired operation, 2) Conditions In (CONI) which transmits the contents of the RPIO control and/or status
register to the processor, 3) Data Out (DATAO) which transmits processor data to the RPIO data register, and
4) Data In (DATAl) which transmits the contents of the RPIO status and drive address registers to the processor.
The RPIO communicates with the RPOI or RP02 Disk Pack via the 35 tines shown in Figure 4-1. There are eight
address and control lines; three tag lines; and one line each for sequence out, unit select, controlled ground, and
write data. These lines carry information from the synchronizer to the disk. The disk communicates with the
synchronizer over eight cylinder address lines; and one line each for attention, unit selected, ready, on-line, index
pulse, sector pulse, file unsafe, seek incomplete, end-of-cylinder, read data coax, heads extended, and read only.
In the PDP-l 0 System, several RPIO Disk Pack Synchronizers can be attached to one data channel. As shown
in Figure 4-2, the RPIO will gate nothing onto the channel bus until it receives the CHANNEL BUSY signal
from the data channel. While inactive, the RPIO relays CHANNEL BUSY, CHANNEL START, and SAWRITE
in the appropriate directions on the channel bus. On receipt of - CHANNEL BUSY, it asserts its own CHANNEL
START and SAWRITE, and becomes active.
4.3 SEQUENCE OF OPERATION
In these discussions, reference is made to the flow charts and logic diagrams listed in Chapter 6 and contained in
Volume 2 of this manual. These are general descriptions intended to clarify some of the more obscure points in
the diagrams. They do not trace each signal on a gate-by-gate basis.

4-1

ATTENTION 1 LINE
UNIT SELECTED 1 LINE
READY 1 LINE
DATA 0

ON-LINE 1 LINE
INDEX PULSE 1 LINE
SECTOR PULSE ll1NE

I

FILE UNSAFE 1 LINE
DATAl

SEQUENCE OUT 1 LINE
ADDRESS e.CONTROL 8 LINES

RP10
DISK PACK
SYNCHRONIZER

PDP-l0
PROCESSOR
CONO

RPO 1
OR
RP02
DISK
PACK

UNIT SELECT 1 LINE
WRITE DATA 1 LINE
TAG 3 LINES
CONTROLLED GND

~
~

OTHER PACKS
35 LINES

1 LINE

SEEK INCOMPLETE 1 LINE
END OF CYLINDER 1 LINE
CONI

READ DATA COAX 1 LINE

I

HEADS EXTENDED 1 LINE
CYLINDER ADDRESS 8 LINES
READ ONLY 1 LINE

10- Or5g

Figure 4-1

RP 10 General Block Diagram

The RPlO receives commands (instructions) via DATAO 250 (254 for a second RPlO in a PDP-l 0 System). The
Flow Chart FCS (Start) describes the receipt and decoding of instructions. Note that !BC INITIAL CLEAR does
not clear SR or SR BUF registers. If the command is a No Op or Clear Attentions, execution is completed. If it
is a Restore (recalibrate), or Seek, the RPIO exits through S2 to execute the command. If the command received
is a Data Transfer, the RPIO exits through S5 to clear and set the head register in the drive, and through S4 to
start the data channel. When the RPIO receives control of the data channel, it exits from FCCC through CC2 to
begin data transfer.
4.3.1 Write Data or Read Data

I

I

If the command is Write Data or Read Data, DTC SEARCH is set and the RPI 0 begins searching for the sector
addressed by the DAR CYL, DAR SURF, and DAR SEC registers. The 10-Jls delay DTC SECTOR DLY allows
the drive's head register to be incremented in a multiple track operation before setting DTC READ REFUSE.
(The term REFUSE denotes useless information and not the act of rejection.)
DTC READ HEADER is set by the first I-bit read off the disk after the 350-Jls delay for DTC SEARCH SYNC
times out. The delay is included so that DTC READ HEADER cannot be set by a I which might be read before
the READ DATA SEPARATOR is in sync. When DTC READ HEADER is set, the next 36 bits are read into the
LPR register. These bits and the 37th are exclusive ORed into the DTC PARITY flop. When the DTC PAR CON
flop turns off, a 600-ns delay is fired that allows time for the LPR to shift the last character in, and for all signals
to settle.
4-2

I

When DTC SRCH COMP STROB times out, the state of the DTC PARITY flop and the HCDE HEADER
COMPARE signals are sampled. If they are correct, they set DTC SEARCH COMP and clear DTC SEARCH.
HCDE HEADER COMPARE is true when the contents of the low-order 18 bits of the LPR match the contents
of DAR CYL, DAR SURF, and DAR SEC registers. On a multiple sector read or write, the RPIO checks the
cylinder and the parity bits, but does not check sector and surface bits.

TO OTHER
[/0 DEVICES

rl_._
CHANNEL
BUSY
CHANNEL
START
SAWRITE
ACTIVE
lIO DEVICE

uu
INACTIVE

1/0 DEVICE

TO OTHER
I/O DEVICES

INACTIVE
RPIO

CHANNEL BUS
TO OTHER
1/0 DEVICES

.---_L.....,CHAN

TO OTHER

B U f Y h I/O DEVICES

CHANNEL
START
SAWRITE
INACTIVE

INACTIVE

I/O DEVICE

I/O DEVICE

ACTIVE
RP10
10-0108

Figure 4-2 RPIO Active and Inactive Functions

If the operation is write data, the RP I 0 exits from FCSC to FCRW. The 3-J..I.s delay prevents the write current

from destroying the last few bits of the header. Four 36-bit words (consisting of Os) are then written without
parity. Following this, the DTC PAR CONT flop is set and a 1 is written as a trigger for subsequent read operations. This field is written during every write to ensure that the Read Data Separator is lined up with data. A
lack of alignment can occur for two reasons; 1) the gap is discontinuous, and therefore the signal is unpredictable,
or 2) the speed of the drive on which the header was written differs from that of the drive on which data was written.

4-3

After DTC WRITE DATA is set, the RPIO exits to the FCM2 subroutine to write the data. It returns when the
I 28-word word counter overflows and the WDC WORD COUNT EQ 00 becomes true. At this time, DTC WRITE
LPR is set and DTC WRITE DATA is cleared. The RPIO again exits to the FCM2 subroutine to write the contents of the LPR on the disk twice. The second LPR is written as guard bits to ensure that the last bits of the
LPR are not destroyed when write current is removed. The DTC SET WRITE is then cleared and the tunnel
erase heads must be turned off after write is removed or FILE UNSAFE will occur. The tunnel erase is used to
narrow the width of the data track as written by the read/write poles. This prevents crosstalk between tracks.
At this point, DTC GAP is set. If the channel has terminated during the write operation, or if there has been an
OVERRUN, or CHAN DATA PE, the operation is shut down. Otherwise, the RPIO waits for the next DSBC SEL
UNIT SEC PLS and continues on FCSC.
If the operation is read data, the RPIO exits through SC2. The Read Data Separator is enabled by DTC BETA

GAP, during which it resyncs in preparation for a read data operation. DTC BETA GAP is cleared in 44 JlS for
an RP02 or in 88 JlS for an RPOI. At that time, if DTC SEARCH COMP is cleared, it is an indication that the
RPIO has not completed its seek operation. At this time, DTC READ GAP, which was set by the trailing edge
of DTC SRCH COMP STROBE, is cleared and the RPIO waits for the next DSBC SEL UNIT SEC PLS to try
again. If DTC SEARCH COMP has been set by DTC SRCH COMP STROBE, however, the RPIO waits for the
first I-bit from the disk, which signals the start of the data area. When the first I-bit is received, DTC READ
DATA is set, clearing DTC READ GAP and the RPIO exits from FCRW to FCMI subroutine. It returns when
128 words have been read, and the WDC WORD COUNT register overflows to O. DTC READ LPR is then set,
and DTC READ DATA is cleared.
After reading one more word, sector parity is checked. If any bit of the LPR register is zero, CXR DISK SEC PE
is set. In either case, DTC GAP is set, and the RPIO waits for the next DSBC SEL UNIT SEC PLS and then
continues on FCSC.

4.32 Write Headers and Data (Format)
If the command is write headers and data (Format), the RPIO exits from FCSC to FCWH where it waits, first for
a DSBC SEL INDEX pulse and then for a DSBC SELECTED SECTOR. After a delay, it sets DTC WRITE HEADER, and exits to the FCM2 subroutine to write 36 words (24 in RPOI) without parity. The RPIO then exits to
FCRW, sets DTC WRITE DATA, clears DTC WRITE HEADER, and returns to FCM2 to write 128 words with
parity. When the WDC WORD COUNT register overflows to zero, the RPIO returns to FCWH, sets DTC WRITE
LPR, clears DTC WRITE DATA, and exits to FCM2 to write the LPR on the disk twice.

NOTE
The second writing of the contents of the LPR is a convenient
way to write gu~d bits.
After the second LPR is written, the RPIO returns to FCWH, clears DTC WRITE LPR, keeps the erase heads
turned on for 20 JlS, sets DTC GAP and waits for the next DSBC SELECTED SECTOR.

4.3.3 Local/Remote
Local and Remote mode operation are straightforward and can be followed directly on FCL. Major differences
in the two modes are listed below:

a.

Local Mode
( 1) The DAR CYL register is tied to the CYLINDER switches.
(2) CC ACTIVE BUF is held off so that the data channel is not started.
(3) The AR is loaded from the 'pATA switches rather than from the AR CHNDOO-35
Channel Data Lines.
4-4

(4) The Initial Control Word Address is not loaded.
(5) The RPIO is logically disconnected from the PDP-lO I/O bus (it will not respond
to its device code and will not set its DONE and PI Enable signals).
(6) IBC SET DAR ETC does not fire, therefore, the status registers are not set by
information present on the I/O bus.
b.

Remote Mode
(J)

DTC FINISH clears IBC BUSY, sets IBC DONE, and interrupts the PDP-lO on the
assigned PI Channel; whereas in Local Mode, DTC FINISH clears IBC BUSY and,
if the SWP STOP switch is not depressed, the entire operation is repeated after a
100-J,Ls delay.

4.4 RPIO RECORDING TECHNIQUE
Disk pack systems are susceptible to a phenomenon termed "pulse crowding." This phenomenon can occur when
a series of I s are recorded on a track. As the pulse density increases, the location of the recorded I s crowd each
other with the result that they may actually reside either ahead of, or in back of, the location in which they were
recorded. The resultant locations can shift in either direction; consequently, a means is provided in the RPIO to
read through a precisely-timed "window", enabling original timing to be recovered.
The RPIO uses a double-frequency, nonreturn-to-zero (NRZ) recording technique. A S-MHz clock signal is divided to produce two 2.S-MHz signals with a 180 degree phase relationship (see DWG RPIO-O-RDS). When
writing, the leading 2.S-MHz signal continuously records I bits on the disk surface, while the trailing signal samples the data to produce data bits. If the data to be written is continuous Os, a 2.5-MHz signciJ. is recorded on the
disk surface. If the data to be written is continuous Is, then a S.O-MHz signal is recorded. Therefore, for any
given data cell, the recorded frequency is either 5 MHz for a I data bit, or 2.5 MHz for a 0 data bit.
To recover data recorded in this manner, the Os rate frequency component must be removed, which is done by
using a phase-locked oscillator (RDS) similar to a sample-and-hold circuit. The RDS samples the Os rate pattern
in the preamble of each record, phase-locking on this pattern, and then maintains phase through the record, making only minor corrections with the use of the Os rate component of each data cell. The RDS then removes the
Os rate component with which it is familiar, leaving only the recovered data.
The Os rate component (2.5 MHz for RP02 or 1.25 MHz for RPOI) removed from the raw data stored on the
disk is also used. This signal is the main source of timing in the RPIO Synchronizer. The RPO I /02 has no clock
track; this system provides self-clocking.
4.5 RPOI/02 ADDRESSING METHOD
The RPO I /02 Disk Pack consists of six (or eleven respectively) evenly-spaced recording platters mounted on a
single shaft. The top- and bottom-most surfaces are not used for recording; instead, the bottom surface has a
metal disk attached to it. The disk contains twenty evenly-spaced notches with an additional notch termed
the Index.
A circuit that is designed to detect these notches, also divides the pulses they produce into either five or ten
equal sectors. These sectors are addressed by a five-bit register which is part of the Data Address Register (see
DWG RPIO-O-DAR). The sector addresses are coded 008 through 048 when using an RPOI drive, and 008
through 118 when using an RP02 drive. All other codes are illegal and, if given, result in the appropriate interrupts.

4-5

A separate read/write head is provided for each of the inner recording surfaces. These heads are mounted in
parallel and in vertical alignment to each other and are attached to a common head tower. The heads are selected
by an additional five-bit register in the Data Address Register designated DAR SURFOO-04. Head addresses are
coded 00 8 through 118 for the RPOI drives, and 00 8 through 23 8 for RP02 drives. All other codes are illegal
and, if given, result in the appropriate interrupts.
The position of all heads, vertically aligned with respect to the vertical axis that passes through the center of all
surfaces, is called a cylinder. Head positioning is controlled by a linear positioning motor and detenting mechanism that is designed to stop the heads at anyone of 203 different cylinder locations. These cylinders are coded
00 8 through 3128 from the outer-most cylinder to the inner-most cylinder, respectively. Cylinders are addressed
by an eight-bit register called DAR CYLO 1-128 in the Data Address Register.
The intersection of a cylinder, head, and sector address defines a unique sector that is the smallest addressable
unit in the system. Each sector has a header word that uniquely defines that sector.
4.6 RPIO REGISTER ORGANIZATION
The register organization of the RPI 0 is given in Figure 4-3. The various signals which control the flow of data
originate in the Data Transfer Control (DTC). Serial information from or to the disk is assembled or dispersed by
the six-bit Shift Register. The data is then processed in parallel, through the Shift Register Buffer, and is transferred to or from the Assembly Register and Longitudinal Parity Register in six-bit bytes.
In read mode, DTC READ BIT translates information bit-by-bit as it is read from the disk. Each bit enters the
shift register at SR5 and is shifted through the register by DTC SHIFT SR. When the register is filled, DTC LD
BUFFER FM SR transfers the six-bit byte to the SR BUF where it loads AR30-35 at a DTC SHIFT AR time. It
also exclusive ORs the byte into LPR 30-35 by DTC LOAD LPR. Subsequent bytes are then assembled by DTC
SHIFT AR, DTC LOAD LPR, and DTC SHIFT LPR. At the end of six byte transfers, the AR contains the word
read, and the LPR contains the parity for that word.
In write mode, the first six bits of a 36-bit word, which has been previously assembled in the AR, is loaded by a
Is transfer into the SR by DTC LD SR FM AR and into SR BUF by DTC LD BUFFER FM AR. While these six
bits are being serially transferred to the disk by DTC SHIFT SR, they are also exclusive ORed into LPR30-35 by
DTC LOAD LPR. At the end of the six transfers, the contents of the AR are written on the disk, and the LPR
contains the parity for that word. Thus, the SR must be cleared before loading. It is cleared by WRITE ENABLE,
which feeds Os into the SR during each shift. Only five shifts are used during a write; therefore, the state of SRO
is indeterminate during a load from the AR, and it is jam transfemed.
When DTC LD SR FM LPR is issued, the contents of the LPR are written on the disk through the Shift Register
by DTC SHIFT SR.
4.7 RPI0 REGISTERS
4.7.1 Assembly Register
The RPIO Assembly Register is shown in Figure 4-4 and in DWG RPIO-O-AR. The register is arranged in groups
of six flops, each group individually feeding the next lower-order group of six. Input is at AR30-35 by ARD3035 H or ~ARD30-35 HANDed with DTC SHIFT AR H. Each flop can also be set by its individual panel switch
(SWP DATA SWOO-35 H) if ANDed with ~IBC LOAD TEST H. The setting of each flop, when ANDed with CC
DATA STROBE L, results in the AR CHND XX L version of that bit. When AR CHND XX Lis ANDed with
DTC DATA RCVR EN L, it direct sets the flop from which it originated.

4-6

CD

CD DTC SHIFT AR
® DTC SHIFT LPR
® DTC LOAD LPR

CD

@

LD BUFFER
FM AR

® DTC LD S R F MAR
(1) DTC LD

®

SR FM LPR

® DTC SHIFT SR

CD

®

DTC LD BUFFER
FM SR

® DTC

CD

®

®

CD
30

31

32

33

34

CD

CD
-

35

AR 30-35
LOADED FROM
SR BUF 0-5 IN "READ"

® DI~~
,
"

SR 0-5
LOADED FROM
AR 0-5 I N "WRITE"
I0

-O~07

Figure 4-3 RP I 0 Register Organization

4.7.2 Assembly Register Data Gate
The RPlO Assembly Register Data Gate is shown in Figure 4-5 and DWG RPlO-O-ARD. The data gate connects
the Assembly Register output and the Shift Register Buffer, and functions as an input switcher for the AR. The
register consists of four sets of six AND gates and two sets of six OR gates; One set of OR gates feeds the set
inputs of AR30-35 with ARD30-35 H. The other set of OR gates feeds ~ARD30-35 to the AR reset inputs.
Input to the AR Data Gate is either from the Shift Register Buffer (SR BUFO-5) or the output of the Assembly
Register (AROO-05). Inputs from the AR are enabled by the write condition (DTC WRITE ENABLE L) while
inputs from the Shift Register Buffer are enabled by the read condition (DTC READ ENABLE L).

4-7

CC

DATA
STROBE
L

-lBC

DTC
SHIFT
AR H

LOAD
TEST
H

,..-.------

DTC
DATA RCVR
EN L

DTC
CLEAR
ARH

I

•
I

I

I

I

I

•

~
ARD 30 H

-ARD 30 H

SWP
DATA__
SW 30
H

SWP DATA SW 25-28 H
SWP DATA SW 18-23 H

--+

-ARD 3b34 H

-4.

SWP DATA SW 12-17 H

--+

SWP DATA SW 01-04 H

-r
-

L

r;:-L.:.:..

I-1
SWP
DATA
SW 24
H

AR
30

--- rt

AR CHND 30L
IN

~
r;:-

I--

- -- -- - -- -t

~

-

1
AR
24

,

f.-

-~- - - - --

l
- - --- -+ - - - - -- -- -~

I--

r----I

-ARD 35 H

SW~
DAT.
SW3
H

r;;
-L.:...
r-;

l..:.

Jr

G~

f--I

AR
35
0
'----

SWP
DATA
SW 29
H

AR

II
I---'

CH~D 35 L
IN

.rt;"
L..:.:..

.r--;\
l."::"

~

I

L

I
I
I
I

f.1
AR
29
0

'----

-- l
~

AR CHND 29 L
IN

I
I
I

I

TO AR 23

TO AR 23

-.

---II

ct

II

r;-

SWP
DATA
SW 00
H

AR
06

kh-

r;-

0

t
I
I
I

~

-

~-

--

r
I
I
I

!
FROM AR 17

I

FROM ARI7

AR
OO

I---

~

~-

t

- -,

I
I

.-L

I

I
I
I

AR 01-04
SAME

I
I
I

I
I

- -- - -- - - -l - -- _J

r---'

II

I--

~

I--1

I

SWP
DATA
SW05
H

AR

"

0

L...--

AR CHND 07-10 L
IN

T

I

I
I
I
+
-

AR CHND 01-04 L
IN

T

I

I

I
I

1

0

-L....:.:...

I---

-- - -

I""--

-L:-

AR CHND 06 L
IN
AR CHND 00 L
IN

~

~

[Z

~

1

---

I

I
L ______ -.J

I

:

AR07-IO
SAME

I
I

I

CC

DATA
STROBE
L

OTC
SHIFT
AR H

I

I
I

I

II

FROM AR 12

I

I

I

I
I

I
I
I

I
I

t
I

AR 25-28
SAME

I
I

[Z

TO AR 18

AR 18-23 a AR 12-17
SAME

I

AR 31-34
SAME
I

FROM AR 12

TO AR 18

I
I
I
I
I

I
I

I

I

AR CHND 24 L
IN

J

I
I

.....--:

ARD 35 H

I

0

L....:.:.

I
I
I

-

-L

dJ

0

I

ARD 31-34 H

SWP DATA SW 7-10 H

J"

~

~

r

SWP DATA SW 31-34 H

IA
-t.::..

I
I
I
I
I

II

I

AR iI

LOAD
TEST H

•

I

r--

CLE~P

I

I

-/I
-

-IBC

DTC
DATA RCVR
EN l

DTC

-,

~

II

t--

~
-rt;"
L..:.:..

r--;\
-L.:...

kh-

~

I

AR CHND 18-23 L
a 12-17 L
IN

AR CHND 25-28 L
IN
AR CHND 31-34 L
IN

AR CHND" L
IN
AR CHND 05 L
IN

9

10- 0508

AR
05
0
'----

Figure 4-4 RPIO Assembly Register
Simplified Block Diagram
4-9

OTC
WRITE
ENABLE L

OTC
READ
ENABLE L

AROO
(0) L
AR 00
(n L

I
I
I
I
I

THRU

I

'V ARO
30H
ARO
30H

SR BUF
0(0) L

I

THRU
I

I
I

SR BUF
0(1) L

I

I

I

THRU

AR 05
(0) L

I
I
I

I

I

THRU

I

THRU

AR 05
(1) L

I
THRU

I
I
'VARO
35 H

I
I

SR BUF
5 (0) L

I
I
SR BUF
5 (1) L

I
ARO
35H

Figure 4-5 RPIO Assembly Register Data Gate,
Simplified Block Diagram

4.7.3 Condition Register
The RPIO Condition Register is shown in Figure 4-6 and in DWG RPIO-O-CXR. This register monitors various
conditions in the controller, raises indications when these conditions are in error, and provides a means of clearing the conditions or controlling them remotely. In the condition register, all flops are cleared by IBC INITIAL
CLEAR except the three error stop disabling flops (DSPE, CDPE, DWPE), which are cleared by IBC GEN CLR
(1) H, and the three priority interrupt flops (PI0-2), which are cleared by IBC CON CLR H.
During a CONO command, the CXR PIO-2 flops are conditioned by the state of IOBD 33-35 HANDed with
IBC CONSET H. When the outputs of PI0-2 are decoded, they determine the priority interrupt channel assigned
to the RPIO (CXR lOB PIl-7 H). During a CONI command, the priority interrupt channel currently assigned is
indicated. The decoder is enabled by CXR PI ENABLE H, which is the combination of either IBC DONE (1) H
or any DTC ATTNO-7 (1) HANDed with IBC BUSY (0) Land IBC LOCAL (0) L. IOBD 35 H, when set during
a Read Data instruction and ANDed with IBC SET DAR ETC A, H, sets CXR WRITE EVEN PAR flop. This
flop causes the channel to write even parity into memory on all data words read from the disk.
The output CXR PARITY ER L is the OR of CXR CHAN DATA PE( I) H, CXR CHAN CONT PE( I) H, and
CXR DISK SEC PE(1) H. The disk sector parity error flop is set by CXR SEC PARITY ERROR H 600 ns after
DTC READ LPR goes false, thus indicating that the checksum bits at the end of a sector (which has been read
by the most recent Read Data command) do not yield a modulo-two sum of I.

4-10

CC
STRT
SCH
L

IC NO SUCH
MEM L

lOBO 24 H

DTC
DATA
STROBE
L

IBC
INITIAL
CLEAR
CH

IBC
CON SET
H

IBC
SET
DAR ETC
A H

IBC
INITIAL
CLEAR
CH

CC
ILLEGAL
WRITE
H

IBC
GEN
CLEAR
(I) H

IBC
CON SET
H

IBC
BUSY
(1) H

/\

CXR '
NON
EX
MEM
0

lOBO 23 H
IC DATA
PARITY
ERROR L

IBC
CONSET
H

DTC PAR
CONT
(1) L
DTC
READ
DATA
(1) L

CXR'
DIS
DSPE
STOP
0

DTC
PARITY
(0) L

CXR '
DISK
WD
PE O

/\
lOBO 24 H
CXR '
CHAN
DATA
PE
0

lOBO
25 H
CXR
PARITY
ER L

IC CON
PAR
ERROR L

1
CXR
DIS
COPE
STOP
0

'" CC
PREVENT
TEST H
lOBO
26H

IBC BUSY
(1) H
IBC DEVICE
SELECTED L
lBC DATAO
SET L

lOBO
26 H
1
CXR
CHAN
CONT
PE
0

LPR
00-35
(0)

CXR
SEC
PAR ITY
ERROR H

DTC
OVERRUN
H

CXR
01 SK
sEc
PE

1
CXR
OVERRUN
0

IOBD
22 H

0
IC WR
CON WD
COMP L

DTC READ
LPR (0) H
I

lOBO 30 H
IBC CON
CLR H

CXR
CW
XFER
COMP
0

C

C)(R DISK
NOT ROY H

V 1--.....- _ C)(R DISK
NOT ROY L

IBC POWER CLEAR H - - - - - - - - - ,
PS
FAI L
MODULE

DTC ATTN
0-7(1)H
lOBO
35H

IBC CONSET H - - - - - .

/\

1

CXR
WRITE
EVEN
PAR
0

IBC DONE (1) H

lOBO 20 H -

CXR
PS
FAIL

o

......._OJ

IBC BUSY (0) L
DTC SEARCH
(1) L

IBC LOCAL (0) L
lOBO 33 H

3

"'Dsec SEL UNIT READY H
"'Dsec SEL UNIT ON LINE H
DSBC SEL FILE UNSAFE H _

CXR PI ENABLE H
CXR PI ENABLE L
CXR lOB PI 7 H

DSBC SEL
INDEX L

CXR
SEARCH
ERO
0

lOBO
21H

CXR lOB PI 6 H
lOBO 34 H
CXR lOB PI 5 H
CXRIOBPI4H
CXRIOBPl3H
IOBD 35 H

/\

CXR
SEARCH
ERI
0

CXR rOB PI 2 H
CXR lOB PI 1 H

CXR
SEARCH
ER2

f\

0

Figure 4-6 RPIO Condition Register, Block Diagram

4- I I

The channel control parity error flop is direct set by IC CON PAR ERROR Land CC STRTSCH L. It is cleared
during a CONO by IOBD 24 H, which also clears the channel data parity error flop.
The CXR CHAN DATA PE flop is direct set when IC DATA PARITY ERROR Lis ANDed with CC STRTSCH
H. IC DATA PARITY ERROR L is true during a CONI when either Channel Data Word Parity Error, Channel
Control Word Parity Error, Disk Word Parity Error, or Disk Sector Parity Error are true. The flop is cleared
during a CO NO by the combination of IBC CON SET Hand IOBD 24 H, which, in turn, clears the parity error
conditions (lOBD 14-17 of that same word). The CXR DISK WD PE flop is cleared when IOBD 24 H is true
and is direct set on a DTC DATA STROBE L by ANDing DTC PAR CONT(1) L with DTC READ DATA (1) L
and DTC PARITY (0) L.
The CXR NON EX MEM flop is direct set by IC NO SUCH MEM and CC STRTSCH L. During a CONI, this flop
indicates that the channel attempted to access a nonexistent memory location. During a CONO, the flop is reset
by the combination of IOBD 23 Hand IBC CONSET H.
The CXR CW XFER COMP flop is direct set by IC WR CON WD COMPL. This condition causes the channel to
store the current contents of the data address register and the control word address register in memory location
B + 1, where B (an even number) is the Initial Channel Control Word Address. The corresponding CONI bit,
IOBD 30, indicates that the operation is complete. Any channel termination causes the control word to be
written, but CONI bit 30 is only set when the operation was requested during a CONO.
The Search Error flops (CXR SEARCH ERO-2) count DSBC SEL INDEX L pulses from the disk presently in
search mode (DTC SEARCH (1) L). During a CONI, they indicate that the cylinder, surface, and sector fields of
the most recently executed read data or write data instruction could not be compared with corresponding fields
of any sector on the addressed track. If SRCH COMP was set and a header (other than the first header in a multiple sector operation) was read incorrectly, bit 2 will be set. If IOBD 21 H is set during a CONO, it is ANDed
with IBC CONSET H to reset all three flops, thereby clearing the SRCH flag.

I

There are three flops that can be set during a DA TAO instruction to disable certain parity error stops to data
transmission. IOBD 24 H, when set during a Read Data instruction, is ANDed with IBC SET DAR ETC A H
to set CXR DIS DSPE STOP. Under this condition, a longitudinal parity error sets the longitudinal parity error
bit, but will not terminate data transmission. When IOBD 25 H is set during a Write Data instruction or a Write
Header and Data instruction, it prevents a channel data word parity error from stopping transmission. IOBD 25 H
is ANDed with IBC SET DAR ETC A H to set the CXR DIS CDPE STOP flop. If IOBD 26 H is set during a Read
Data instruction, it will disable the disk word parity error stop in a similar manner by setting the CXR DIS DWPE
STOP flop.

I

The CXR ILLEGAL WRITE flop is set by CC ILLEGAL WRITE L when either IBC BUSY is true or CC PREVENT
TEST is false. During a CONI, this indicates that a Write Header and Data or Write Data instruction addressed a
disk drive whose READ-WRITE/READ ONLY switch was in the READ ONLY position, or it indicates that a
write operation was in process when a sector pulse was received. This flop can be reset during a CONO by setting
IOBD 26 H at IBC CONSET H time, thereby clearing the ILL WR flag.
The CXR ILL COM WL BUSY flop is set during IBC DATAO SET L by IBC DEVICE SELECTED Land IBC
BUSY (1) H, thus indicating that a DATAO was issued while the BUSY flag was up. This causes the second
command to be ignored. It is reset during a CONO (IBC CONSET H) by IOBD 27 H, which clears the ILL COM
flag.
CXR OVERRUN is set by DTC OVERRUN H to indicate, during a CONI, that the channel failed to respond to
a request for a data word, or failed to accept a data word fast enough during a data transfer. During a CONO,
IOBD 22 HANDed with IBC CONSET H clears the OVERRUN flag by resetting this flop.

4-12

Whenever the interface power supply voltages are out of tolerance, the PS FAIL module will direct set the CXR
PS FAIL flop and turn on DONE causing the transfer to terminate at the end of the current sector. This flop
cannot be reset until power supply voltages are within tolerance. At that time, 10BD 20 HANDed with IBC
CON SET H clears the PS FAIL flop.
CXR DISK NOT RDY is an output of the condition register. This signal results from either DSBC FILE UNSAFE,
-DSBC SEL UNIT READY H, or ..... DSBC SEL UNIT ON LINE H.
4.7.4 Data Address Register
The RPIO Data Address Register is shown in Figure 4-7 and in DWG RPIO-O-DAR. This register defines the
entire data address and consists of four sub-registers to describe the drive selected (3 bits), the cylinder desired
(8 bits), the surface on which the data resides (5 bits), and the specific sector to be located (5 bits). The setting
of the flops is effected either by switch panel controls which direct set the appropriate flops during local mode
operation (lBC LOCAL (I) L), or by the individual states of 10BD bits 03-23 received during a DAT AO and
enabled by IBC SET DAR ETC A/B H. All flops are cleared by IBC INITIAL CLEAR A/B H.
4.7.5 Longitudinal Parity Register
The RPIO Longitudinal Parity Register is shown in simplified form in Figure 4-8 and in DWG RPIO-O-LPR. Data
from SR BUF00-05 is exclusive ORed and applied to LPR30-35 by DTC LOAD LPR H. On receipt of each DTC
SHiFf LPR B or C, the contents of each column of flops are shifted to the next lower order set of six flops.
LPROO-05 are tied back to LPR30-35 through AND/OR gating. All flops are cleared by DTC CLEAR LPR H.
4.7.6 Shift Register
The RPIO Shift Register is shown in simplified form in Figure 4-9 and in DWG RPIO-O-SR. The circuit comprises
a six-bit shift register interconnected to a six-bit shift register buffer. Input is at SR5.
SR BUFOO-05 flops are all direct set by the OR of SRO-5 and AROO-05, each ANDed by its DTC enabling signal.
The signals designated as SRO-5(L)L and SR BUFOO-05(L) L feed individual lights on the indicator panel.
4.8 RPI0 Control Circuits
4.8.1 Channel Control
The RPIO Channel Control is shown in DWG RPIO-O-CC. The circuit serves as an interface with the DFIO Data
Channel. In order for the RPIO to gain access to the data channel, it must transmit CHANNEL START to the
channel and it must receive CHANNEL BUSY from the channel. If the synchronizer is not actively engaged with
the data channel, it relays the CHANNEL START and CHANNEL BUSY to another I/O device and is prevented
from generating its own CHANNEL START.

I

Either the RPI 0 or the DFIO can terminate operation. If the DFIO terminates operation, it removes CHANNEL
BUSY, and the RPIO negates CHANNEL START. Ifthe RPIO terminates operation, it removes CHANNEL
START, and the DFIO negates CHANNEL BUSY. At this time, the RPIO inhibits the generation or relaying of
CHANNEL START for 500 ns, a requirement imposed by the DFIO.
In operation, the RPIO requests access by generating CHANNEL START and is answered by CHANNEL BUSY
from the channel. This acknowledges that the RPIO has access. The RPIO then generates DEVICE PULSE, which
transfers the ICWA to the channel where it is processed and evaluated. When this is completed, data transfer begins,
the direction of which is determined by SAWRITE.

4-13

IBC
LOCAL
(1)

L

SWP DRIVE 0 L

---10

lOBO 04 H

SWP DRIVE

2 L

---10

1

,,1

~

lOBO 03 H

SWP DRIVE 1 L

IBC
INITIAL CLEAR
BH

IBC
IBC
SET
INITIAL
DAR ETC CLEAR
AH
AH

~
-2J

IJ\l
p

lOBO 05 H

8-

1

lOBO 06 H

DAR
DRIVE
00
0

DAR SU
07 L

-

06 L

'--

8-

05 L

1

BINARY
TO
OCTAL
CONV.

DAR
DRIVE
01
0

C~ti2B L

--

~~t 64

--.--.

03 L
02 L

1
DAR
DRIVE
02
0

OIL

-SWP
CYL 64 L

~~t 32H

10BD19H

SWP SEC

- pr 8-

.-P

10BO 20 H

.-

SWP SEC

02 L

~

lOBO 21 H

SWP SEC 03 L
lOBO 22 H

SWP SEC
lOBO 23 H

-

8-

.--r;-,
L:J

lOBO 09 H

lOBO 14 H

f--1

~

1
DAR
SEC
02
0
'--

8-

P

SWP
SURF 01

L---IO

~

lOBO 15 H

'-1

SWP
SURF 02 L - -

~

lOBO 16 H

~~~F

03 L

lOBO 17 H

---10

~

8

DAR
SEC
03

'--

8-

F--1

SWP
SURF 04 L

1
DAR
SURF
00
0

I--

8-

lOBO lB H

DAR
SEC
04
0

I
DAR
SURF
01

I---

P

SWP
CYL OB H
10BD 10 H

-SWP
CYL OB L

~~~04 H

-

tr;"t
p
J-;l

--

~.

~

P

t"f;1

p

- fA1
-

P
~
j....:..:...I

8-

8-

f-----..J
1
DAR
SURF
02

-SWP
CYL 04 L

~

SWP
CYL 02 H

~

10B012 H

DAR
SURF
03

-SWP
CYL 02 L

0

I--

8-

--

I,\l
P

~

.....:..:.J

I

~

1

SWP
CYL 01 H
lOBO 13 H

- I,\lI,\l
-

P

p

DAR
SURF
04

~

"--

t";l

-

lOBO II H

I.l

-~

-SWP
CYL 16 L

F---

0

0

04 L

~

1,:1

DAR
SEC
01

f---

8-

1;1

DAR
SEC
00
0

0

I;:l

~

1

L

SWP
CYL 16 H

~

f---

I;:l

01 L

IBC SET DAR
ETC B

f---

P

lOBO OB H

C~~P32

SWP
SURF OOL -

1;1

P

..........I;l

00 L

l...--

SWP SEC 00 L

~

04 L

f---

IBC
INITIAL
CLEAR
CH

H

r;,

10BD 07 H

-

8-

p" I

SWPCYL 12B.H--+--Oo

~

-SWP
CYL 01 L

--

f-;;:l

8-

1
DAR
CYL
12B
0

I

I---

8-

1
DAR
CYL
64
0

I

~

~

8-

I
DAR
CYL
32
0

I

I-~

~

1
DAR
CYL
16
0

I

I---

8I

~

f-1
DAR
CYL
OB
0

I---

8-

F-----l
1
DAR
CYL
04
0

I -

8I

8-

I--1
DAR
CYL
02
0

I--~

1
DAR
CYL
01

~

L.J

Figure 4-7 RPIO Data Address Register, Block Diagram (Revised)

4-14

~

10-0511

DTC CLEAR LPR

DTC SHIFT LPR

DTC LOAD LPR

---+---,
~ ~+-------------~----------------------------~----+------,

SR aUF 00 (1 ) H

TO
} LPR
24

---+-...

LPR
00

o
~ ~+-----------+-------~----------------~~~----~------~

SR aUF 01 (1) H

_.J

-.r-

SR aUF 02 (1) H - '

SR aUF 03 (1) H

SR aUF 04 (1 ) H

LPR

LPR 31-34

II

SAME

----L _

---+-...

06~29

LpR 01-04
SAME

SAME

II
I

_--.J

-I
1

SR aUF 05 (I) H

-I

~---._

LPR
35

FROM{
LPR

01----+-

11

o

~ ~--------------------------------~
10-0512

Figure 4-8 RPlO Longitudinal Parity Register, Block Diagram

4-15

DTC

DTC
LOAD

SHIFT

SR FM

SRL

LPR L

DTC
LOAD
SR FM
AR L

AR05 (t) L

rr;t

LPR05 (0) L
OTC READ BIT (t) L

----0

DTC WRITE ENABLE L

~

DTC READ BIT (0) L

AI

~

...:.:J

DTC
LOAD
BUF FM
AR L

DTC
LOAD
BUF FM
SR L

";l

~~

..:.:..J

SR5 (LI L

~

SR BUF05 III L
I
SR
BUF
05

SR5

$-~

~

~

AR05(0)L

L.J

AR04 (I) L

tAl
~

LPR04 (0) L

~
~

~

0
L--

I;l
f-:.:-J

~&r---;-

SR41LIL

~

SR BUF 04 (Ll L
I
SR
BUF
04

SR4

G:

0
'---

i;0

AR04 (0) L

LPROI- 03 (0) L
AROI-03 (t) L

Iro
I

~-

~

-- - - - - - - - - - - - -r- -

SR3 A

FROM
L ( OI

~

-- --

FROM
SRI (1)

- - -- - - - - - - - -- - - -r- -

rt;L

LPROO (0) L

L-J

~

L.:.J

~&----;-

-- --

SR BUFOI-03ILl L

~

SROILIL

tJP

-y-~

DTC SR JAM H

t"f:l
L.J

~

Figure 4-9 RPlO Shift Register, Block Diagram

4-16

---

--.J

5RI-3 (LI L

t-;J

SRO

AROO (01 L

"I

-r

TO
SR3 A

(SAME AS A BOVE I

AROO (11 L

"

0
L--

SR BUFOO III L
I
SR
BUF
00
0

----;

IO~

0513

Synchronization of transfers between the Data Channel and the RPIO is achieved by a CHANNEL PULSE issued
by the channel and DEVICE PULSE generated by the RPIO. Regardless of the direction of transfer, these pulses
alternate on each word transferred and perform a bookkeeping function on that transfer.
When the channel has a word ready for transfer during a write, it issues a CHANNEL PULSE to the RPIO implying that it may take the word. When the RPIO has taken that word, it sends a DEVICE PULSE to the channel
indicating that it is ready for the next word. When the next word is ready, the Data Channel issues another
CHANNEL PULSE; when that word is transferred, the RPIO issues another DEVICE PULSE. This dialogue continues until the desired number of words have been transferred.
When the channel is ready to receive a word from the RPIO during a read, it issues a CHANNEL PULSE to the
RPIO indicating its readiness. When the RPIO places the word on line, it sends a DEVICE PULSE to the channel
indicating that it may take that word. When transfer is complete, the Data Channel issues another CHANNEL
PULSE; when that word is on line, the RPIO issues another DEVICE PULSE. This dialogue continues until the
desired number of words have been transferred.
The CHANNEL and DEVICE pulses alternate under normal operation; this fact is used as a bookkeeping arrangement on transfer. If any pulse occurs twice without the other, an error or overrun is indicated.
At termination, the data channel sets IDLE and removes the CHANNEL BUSY which results in the resetting of
the CC ACTIVE flop in the RPIO.
4.8.2 Data Transfer Control
The RPIO Data Transfer Control is shown in DWG RPIO-O-DTC. This circuit functions as a traffic manager; it
determines the mode of operation of the synchronizer and the direction in which information will flow between
Disk and Processor.
As shown in Figure 4-10, the DTC OP CODE 00-02 flops are cleared by IBC INITIAL CLEAR A H. These OP
CODE flops can be either set by IOBDOO-02 H in combination with IBC SET DAR ETC B H, or direct set by
SWP OPO-2 L when ANDed with IBC LOAD TEST C L. When the outputs are decoded, they result in one of
various operation commands to the device.
4.8.2.1 Read Data - If DTC READ DATA COM H is decoded, it sets the DTC SEARCH flop if ANDed with CC
STRTSCH H and enables the DTC READ GAP flop to be set after the header is read when triggered by the leading edge of -DTC SRCH COMP STRB H (see Figure 4-11). Note that each flop pertai!1S to a portion of the disk
format.

I

I

a.

When DTC SEARCH flop sets, it ANDs with DTC SEARCH SYNC H to set DTC READ
REFUSE (DTC SEARCH SYNC is DTC SECTOR DL Y H + 350 J.Ls). When the READ
REFUSE flop sets, it conditions the setting of DTC READ HEADER flop which is set
on receipt of DTC READ CLOCK PULSE L if DTC READ BIT (I) L is present and if
the READ HEADER flop is reset at that time.

b.

When DTC READ GAP flop.,sets, it ANDs with DTC BETA GAP H (the OR of DTC SRCH
COMP STROB H and either DTC RPOI H or DTC RP02 H) and DTC SET READ DATA H
to set DTC READ DATA flop. The ORed delays (88 J.LS for RPOI and 44 J.LS for RP02) are
adjusted for the relative gap widths at nominal rotating speeds for each disk type. When
DTC SET RD/WR LPR H is present, DTC READ DATA (I) L sets DTC READ LPR flop
when WDC WORD COUNT EQ 00 L is asserted, thereby signalling the end of data. When
the LPR flop sets, it resets the DTC READ DATA flop.

4-17

IBC LOAD
TEST C L

SWP OP 0 L

IBC INITIAL
CLEAR A H

1\

lOBO 00 H

I
oTC
OP
CODE
000

oTC READ DATA COM H

OTC WRITE DATA COM H

SWP OP 1 L

lOBO 01 H

DTC READ IMAGE COM H

1\

1
OTC
OP
CODE
01 0

SWP OP 2 L

oTC WRITE HoR COM H
DECODER
OTC SEEK COM H

II

OTC CLEAR ATTN H

lOBO 02 H

1
OTC
OP
CODE
02
0

IBC SET
DAR ETC
BH

oTC NO OP H

OTC RESTORE H

-=-

IO·O~14

Figure 4-10 DTC OP Decoding, Block Diagram (Revised)

oTC
READ DATA
COM H

CC STRTSCH H

oTC
SEARCH
SYNC H

oTC
READ CLOCK
PULSE L

1\
OTC
SEARCH

oTC
READ
REFUSE

0

0

OTC
READ
HEADER
0

oTC READ
BIT (1) L
-OTC SRCH
COMP STRB H

1\
oTC
READ
GAP

1\
OTC
READ
DATA

0

oTC
READ
LPR

0

OTC
BETA
GAP H

0

oTC
SET
RO/WR
LPR H

lo-oele

Figure 4-11

DTC Read Data Command, Block Diagram (Revised)

4-18

4.8.2.2 Write Data - If a DTC WRITE DATA COM H is decoded, it sets the DTC SEARCH flop if ANDed
with CC STRTSCH H, and enables setting of the WRITE GAP flop after the header is read (see Fi!ure 4-12).
NOTE
Each flop pertains to a portion of the disk format; the
synchronizer must read a header before writing data.

a.

When DTC SEARCH flop sets, it sets READ REFUSE flop and, in turn, sets READ HEADER
flop in the same manner as in a read data operation. Assertion of DTC CLR READ HDR H resets DTC READ HEADER flop. After a 3-J.I.s delay, DTC READ HEADER (0) H sets WRITE
GAP (the delay prevents the header parity bit from being destroyed).

b.

When WRITE GAP flops sets, it conditions the setting of DTC WRITE DATA flop. WDC DATA
WD CT04 (I) LANDed with the condition DTC PAR CONT-+O, sets the DTC WRITE DATA
flop to indicate that a sync bit has just been written. DTC WRITE DATA (1) H resets DTC
WRITE GAP flop and conditions the setting of DTC WRITE LPR. When DTC SET RD/WR LPR H
goes High, WRITE LPR will set, thereby clearing DTC WRITE DATA flop.

I
I

2.8.2.3 Write Headers and Data - If a DTC WRITE HDR COM H is decoded (see Figure 4-13), it conditions
DTC W H INDEX flop which sets when DSBC SEL INDEX goes High. In addition, DTC WRITE HDR COM H
enables DTC WRITE DATA flop to set on a DTC WRITE CLOCK DLY L when either (a) WDC DATA WD CTOI
and 04 are on a I and an RP02 is in use (-DTC RPOI L), or (b) WDC DATA WD CT02 and 03 are on a I and an
RPOI is in use (DTC RPOI L).

a.

When DTC W H INDEX flop sets (it can only be set if the DTC W H LOCKOUT switch is
disabled), it ANDs with DTC SECTOR DL Y H to set DTC WRITE HEADER flop. (DTC
SECTOR DLY is a 10-J.I.s delay on DTC SELECTED SECTOR II /\ IBC BUSY (1) H to
allow setting of Head Register in multiple sector reads and writes.) DTC WRITE HEADER
(1) H then asserts DTC WRITE ENABLE L. If CC ACTIVE is false at this time, it asserts
DTC OVERRUN H and also ANDs with IBC LOCAL (0) L to generate DTC SHIFT CT 0
PLS H/L. In addition, it generates DTC SET WRITE Hand ORs with DTC WRITE DATA
(1) H to yield DTC WRITE HDR /\ DATAL which conditions the setting of DTC PREVENT
TERM flop.

b.

When DTC WRITE DATA flop sets, it resets DTC WRITE HEADER and sets DTC WRITE
LPR flop when WDC WORD COUNT EQ 00 L is asserted. The setting of WRITE LPR flop
resets DTC WRITE DATA. The DTC WRITE LPR flop resets when WDC DATA WD CT06
(I) Hand DTC PAR CONT (0) H coincide.

4.8.2.4 Oear Attentions - If lOBO 00, 01, and 02 are decoded as 101 (the CLEAR ATTENTIONS command),

1

IBC SET DAR ETC B produces DTC CLEAR ATTN. When combined with IOBD 27-34, this resets any of the
DTC ATTN flip-flops (see Figure 4-14). Note that if an IBC CLEAR ATTN H or an IC CROBAR H is received,
all DTC ATTN flops will be cleared simultaneously. These flops are set by DTC ATTN 00-07 H, representing up
to eight possible drives. The DTC ATTN signals are received as MPX ATTN 00-07 L from the multiplexer and
are inverted prior to being applied to their respective flops.
4.8.2.5 Restore - If DTC RESTORE H is decoded, it ANDs with IBC SET DAR ETC DL Y H to generate the
DTC TAG and BUS LINE strobes (see Figure 4-15). In addition, DTC RESTORE H is ANDed with DTC BUS
LINE STROBE H and is sent to the disk as DSBC UB 6 L.

4-19

>-t

SECTOR --<350 SEC
PULSE
,.
DTC WRITE
DATA COM H

C C STRTSCH H

DTC READ
CLOCK PULSE L

DTC SEARCH
SYNC H

1\

DTC PAR
CONT (0) H

DTC SET
RD/WR LPR H

DTC SEARCH
COMP (I) L

1
DTC
WRITE
LPR

o

3j1sec
10-0516

Figure 4-12 DTC Write Data Command, Block Diagram (Revised)
DTC
WRITE
HDR
COM

CC
IBC
ACTIVE LOCAL
(0)
(0)

DTC
SECTOR
DLY
H

H

H

L

1--4--_ DTC OVERRUN H
~-+-_ DTC WRITE ENABLE L

DSBC SEL INDEX H

- - - - - . I 1\

DTC SHIFT CT 0 PLS H

1------<..
DTC
WRITE
HEADER

DTC
WH
INDEX

o

o

1----_

r

EQ 00
L
DTC
PREVENT
TERM

- DTC RPOI L

o

WDC DATA WD
CT 01 (1) L

WDC DATA WD
CT 02 (1) L

DTC WRITE HDR V DATA L

WDC
WORD COUNT

LOCKOUT

WDC DATA WD
CT 04 (1) L

~---. DTC SET WRITE H

DTC
WRITE
LPR

DTC
WRITE
DATA

o

o

WDC DATA WD
CT 03 (1) L

DTC RPOI L

DTC
PAR
CONT
(0)

WDC
DATA
WD
CT

H

06 (1)H

Figure 4-13 DTC Write Headers and
Data Command, Block Diagram
4-20

10-0517

The Tag Lines indicate how information present on the Bus Lines is to be used. The Bus Lines contain either
control information or address information. The Tag and Bus Line Strobe waveforms are shown in Figure 4-16.
Note that on input of either DTC SET HD ADV (I) L, DTC SET HD REG (I) L, DTC CLEAR HD REG (I) L,
DTC SET CYL (1) H, DTC SEEK START (1) L, or DTC RESTORE H as described above, the resultant signal
encounters various delays that control the set and reset times of an S202 Type flop. The resultant Tag and Bus
Line Strobes are sent to the disk drivers shown in DWG RPIO-O-DSBC along with their appropriate disk commands where they are decoded. The Bus Line Strobes condition the gates. Then, in the middle of the Bus Line
Strobe, when the lines have had time to settle, the Tag Lines strobe the information down the line to the disk.
IBC CLEAR ATTN H

DTC
ATTN

00 H

-

-

-

IC

-THRU- -

CROBAR H

DTC
-ATTN

-

07 H

o

o

DTC

CA\~ANR_~r-'--

-

-

- THRU - -

-

-

.............

H

lOBD

IOBD

27

34
10 -0518

Figure 4-14 DTC Clear Attentions Command,
Block Diagram (Revised)

DTC
RESTORE
H

TO OSBC

'V

",J

IBC SET DAR ETC DLY H

1~s

-\.

- OTC TAG LINE
STROBE L

- DTC TAG LINE
STROBE H

1~s

1

DTC BUS LINE
STROBE L

0

OTC BUS LINE
STROBE H

S202
DTC
DTC
DTC
OTC
DTC

SET HD ADV (1) L
SET HD REG (1) L
CLEAR HD REG (1) H
SET CYL (1) L
SEEK START

] V

:l

-t

[

\.

1~s

J-10-0519

Figure 4-15 DTC Restore Command,
Block Diagram
4-21

4.8.2.6 Seek - If a DTC SEEK COM H is decoded, it ANDs with IBC SET DAR ETC DLY to set DTC SET CYL
flop (see Figure 4-17). When DTC SET CYL flop sets, it initiates the BUS LINE STROBEs and conditions the
setting of DTC SEEK START flop by -DTC BUS LINE STROBE H. DTC SEEK START (1) Lis ANDed with
DTC BUS LINE STROBE H in DWG RPIO-O-DSBC where it is sent to the disk as DSBC UB 2 L.

INPUT - _......
OTC BUS liNE STROBE L - - . . . ,

-OTC TAG liNE STROBE l - - - - -......

OTC BUS LINE STROBE H - -.........
-OTC TAG LINE STROBE H - - - - - - ,
to-0520

Figure 4-16 Tag and Bus Line Strobe Waveforms
OTC
SEEK
COM H

IBC SETOAR
ETC OlY l

TO OSBC

-OTC BUS
LINE STROBE H
TAGS
BUS
LINE
STROBE
DELAYS

1 - - - - - TO OSBC
10-0535

Figure 4-17 DTC Seek Command, Block Diagram

4.8.3 I/O Bus Control
The RPIO I/O Bus Control regulates the interchange of data and control signals on the bidirectional I/O bus lines
which carry data and status information, device selection codes, and control signals between the RPIO Synchronizer and the PDP-l 0 Processor (see DWG RPIO-O-IBC).
The device select code bits IC IOS03-09 L are ANDed with IBC LOCAL (0) L to yield IBC DEVICE SELECTED
Hand IBC DEVICE SELECTED L. If the selection bits contain the RPIO device selection code (250 or 254)
and if the RPIO is not in local mode, the RPIO is selected. This is the basic activating level for all RPIO functions.
The performance of all basic instructions depends on the presence of this level.
When the READIN or RESET switch is pressed on the PDP-lO Processor or when a programmed reset is issued,
IC lOB RESET L is generated, which halts all motion in the RPIO and clears all conditions that initiate an interrupt. This condition prevails until the RPIO is reactivated by a CONO command.
4-22

In the RPIO, IC lOB RESET LANDed with IBC LOCAL (0) L, results in IBC POWER CLEAR Hand L which
is used in the following ways in the device;

a.

To direct set IBC GEN CLR.

b.

To generate, after hIs delay, IBC CLEAR ATTN.

c.

To assert IBC CON CLR H which is ANDed with IOBD 32H to set IBC DONE.

d.

To set IBC DONE.

e.

To set IBC STOP.

f.

To reset IBC LOCAL.

The IBC LOCAL START flop is held in the direct reset condition when the switch panel START switch is in the
o condition ("'SWP START L), the switch panel STOP switch has been depressed (lBC STOP (1) L), and IBC
BUSY is on a O. By activating the START switch (SWP START H), the IBC LOCAL START flop will set. As a
result, the IBC STOP flop resets and the IBC GEN CLR flop is set (if -SWP OPOO L is present). IBC GEN CLEAR
(1) L conditions the set gate on IBC LOCAL flop, which can then be direct set by SWP LOCAL SET L.
4.8.4 RPIO Read Data Separator
The Read Data Separator (RDS) is shown in DWG RPlO-O-RDS. This circuit enables the controller to read information from any disk pack that has been previously recorded by an RPIO, irrespective of drive speed differentials
between time of write and time of read, or any other variables that might cause the recorded data to reside on
the disk in an unpredictable manner (provided those variables are within certain predetermined limits). During a
read, the circuit supplies the controller with clock pulses (via Data Window ~ 0) and DATA OUT pulses (indicating that a 1 bit was read). A crystal clock is used to strobe data pulses and clock pulses at the time the data is
recorded; consequently, the average frequency of data read from the disk should be constant. However, in reality it is not. Due to a phenomenon called "bit shift", timing between individual pulses varies somewhat because
of speed variations in drives and a phenomenon inherent to magnetic devices termed "pulse crowding."
Figure 4-18 illustrates the effects of these variables. In "a" of Figure 4-18, the ideal theoretical waveforms are
shown for raw data consisting of all Os, Os and 1s, and all 1s. In "b", the effects of pulse crowding are shown
wherein the location of a pulse just recorded will be displaced by the next recorded pulse. The pulse crowding
effect becomes more pronounced the closer pulses occur to one another. Because of this effect 1) there appears
to be no displacement of clock pulses when recording an all Os data pattern, 2) there is a predictable + and - shift
of clock pulses either side of 1s in an alternate data pattern, and 3) there is an unpredictable variance in clock
pulses when recording an all 1s data pattern.
In "c" of Figure 4-18, the effect of drive speed differentials is shown on an all Os data pattern, first with speed of
write greater than speed of read, and then with the opposite conditions true. In the third waveform of "c", the
effect of pulse crowding has been added to a speed variation.
NOTE
The individual shifts of clocks from their normal positions
vary from pulse to pulse producing an overall variance
called "jitter."

4-23

CLOCK

DATA

CLOCK

DATA

~

RAW DATA 0

S

CLOCK

~

DATA

CLOCK

~

DATA

CLOCK

~

DATA
~

- - - - - - 1 AL --"1-"---1A'----"O-"---JAL--"-1"---..JAL--,,O-.---1AL--.1
- . - - } IDEAL

A RAW DATA 0S/1. _ _ _ _ _J
"1"

"1"

"1"

"I"

"I"

'0'

'0'

'0'

'0'

'0'

WAVEFORMS

RAW DATA 1. _ _ _ _ _J

RAW DATA O.

B

----....J~"---.-I.----.J~ --.-O-,---J~I....--'-I-.--J~"---.-o,----l~"---'~-I'--}
A
~,
A"
n
A
A
--J1,1'--..:..--...J!'!
I:L-...J:I_:-_J
'I':
-:+~f":;I' ~-~:"I'
1Hi-'I'-..j-~~
'!,.
...

1'Tl,

RAW DATA Oslls _ _ _

_

L_---'_~L__.J:I'---___'_...J

_ _ _J..j,

_oJ:

RAW DATA 15

:,--':"',

~-tJ. ~

I

:

--;6:--

~

I

RAW DATA Os

---:

--:ll~

I

~ ~

"Q"

110"

~

r-

I

"a"

~

CROWDING

........

~
I

~:--

~

I

~

"0 II

--.....

"0"

_..J:AI;-:
-.J~L---'--,,-O'-'
....L:_ _ _ _ _.....
....J~L~------..J,A~~----. . . A~~____. ...:!...JAL~
. . ___.! .W<~R ~iFEF~~ENTIALS
':""--"-0'-'

---"o-"-.J:AI.::---,,o-"

RAW DATA Os _ _ _ _ _ _

,

"0"

----:.,-H----,
-.:!::.

--..J~'----"-,---In. .
~ II ~

_.!!.W>::.!!R} ONLY

....:..J:,

"1"
RAW DATA O./ls _ _ _

PULSE-

I

r-

------....L---"o-,,---.J:A~'
,

c

I

II

ONLY

"1"

_....In

... ll.....

,t' t
PC

~'.

~

"0"

"1"

~'--...:---InL-....:W::.:<~R SPEED ~

:

~6~

t

~

6.

;...

+P.c.

VARY' jiTTER + DIRECTION

~
--iVARIANtE:,-

10-0538

Figure 4-18 Sources of Jitter in Disk Playback

There is no guarantee that an individual disk pack will always be read on the same drive on which it was written,
or that the drive speed of write will be exactly the same as the speed of read even on the same drive. Therefore,
a means must be provided that enables the controller to read data in a jitter environment provided the jitter does
not exceed certain specified limits. The reading of this data is accomplished by a timed sampling method called
"windowing" in which the controller is forced to read during a variable period of time after the average occurrence
of each clock pulse. The time period is a function of the deviation of each clock pulse from normal with respect
to the deviation of the previous clock pulse. A one-shot type of separator cannot be used because a single value
measured from one pulse cannot be found which provides a full window for the next and only the next pulse.
However, because bit shift is measured by the time difference (either plus or minus) from the time a bit was recorded, windowing can be achieved if the window generator is aligned with the time each pulse was recorded
rather than on the actual pulse detected and by aligning on the average of the clock bits coming in rather than
on each succeeding bit.

I

This averaging is achieved by the VFO clock circuit comprising: flops RDS ERR A and B, the module B410, and
the analog circuit G589 or G590. In operation, the VFO clock frequency slows down before the first sync
field because no pulses have been coming in (0 frequency). Thus, the VFO clock requires considerable time to
come up to frequency (see Figure 4-19). For this reason, the first sync field in the format must be approximately
350 !1S. The second sync field need not be as long (88 !1S for RPO I and 44 !1S for RP02) because the VFO clock
is up to frequency and needs only to realign itself slightly as a result of the switching discontinuity from reading
the header to writing data.

4-24

SLOW START UP

ON FREQUENCY

BEFORE SYNC FIELD
10-0537

Figure 4-19 VFO Startup Waveform
During each sync field the RDS must use all clock pulses received to achieve synchronization. Window generation must be disabled at these times because misalignment can mask some clock pulses. The signal RDS SYNC
FIELD ENBL is generated during these periods to bypass window selection. This signal is removed after sufficient time has passed to ensure that the window is in alignment. To provide the averaging effect, the data
window is toggled by the VFO clock rather than by incoming clock pulses that contain the jitter. The RDS
ERR A and B flops serve as a phase comparator. When these two flops are in different states, correction voltage
is applied to the VFO clock. The correction voltage, resulting in either faster or slower VFO operation, is determined by the flop set to a 1. Neither flop remains on a 1 for any length of time. When both flops are on ai,
a clear pulse is generated which resets both flops. Thus, the amount of time one flop is on a 1 while the other is
on a 0 represents the time difference between the pulses that set the flops. The correction voltage that is applied
forces the pulses to align.
The two set pulses are generated by RDS DIVIDE 2 --+ 1 and RDS CLOCK DL Y. The RDS DATA WIND and
RDS CLOCK WIND are toggled by the VFO clock through RDS DIVIDE 2, which halves the clock frequency;
therefore, the comparator can control the toggling rate. The data window's edge is aligned to RDS CLOCK OUT
(see Figure 4-20) while the clock window is aligned to RDS DIVIDE 2 --+ I and the data pulse may fall anywhere
within the data window. DTC SET READ L is applied to the clear inputs to ensure that the DTC READ BIT is
cleared when executing a write.
THESE ARE THE TWO

SIGNALS ALIGNED

ROS OATA WIND

RDS CLOCK OUT _ ' - -_ _' - -_ _- ' -_ _- ' - -

RQS DATA

our _ _-'-_ _-'-_ _-'-_ _

Figure 4-20 RDS DATA WIND Waveform

A block diagram of the Read Data Separator is shown in Figure 4-21. The functional elements include a pulse
shaper, a sync separator, a window positioner, a phase comparator, and a separate differential amplifier/filter for
each voltage controlled oscillator (VCO). Two VCO's are provided; one for RPOI drives, the other for RP02
drives. In addition, the RDS contains a frequency divider and an optimizer that loops back to control output
gating of incoming data.
During the preamble or synchronization period, the input from the disk (MPX RAW DATA H) is fed to two
AND gates each of which is enabled by the type of disk drive not in use. If an RP02 is in use, the gate enabled
by ~DTC RPOI ENABLE H is turned on. If an RPOI is used, the opposite is true.

4-25

----,
I
I

-OTC RPOI ENABLE H - ; - - - - - ,

ROS DATA IN L
~---------------------------------------------------------------- ROS ~TA OUT H

MPX RAW DATA H -...-.......

- OTC RP02 ENABLE H - t - - - - - '

.....J

-lr
II
OTC BE TA GAP

OTC SEARCH SYNC H

-r-----i

OPTIMIZATION
ROS

~~~~ I-_+-+----i
o

II

"I

-f--------'

I

1
ROS
CLOCK
WI NO "'~+-----i

II

L ______________ .J
OTCSET

a

REAOL---------------------------~------------~~+_------~

r-;INDOWPostrioNI';;G - - - -

OTC RP02ENABLEL

-

-

-

---

--,

I

I

--------------------------------~~~---_r-~I~I-_+_rt_----~---------------------------------~~--_,

II
II
II

"

~~-+-+--- RDS EXT DATA WIND

~__.....- - - - ROS

__

I
I

I

CLOCK OUT H

___
---___
----_---_--___
----_----_---___
----=:.-_-~--_-_--._J

OTC RPOI ENABLEL -;-------------------------------r----~-r~-.~L_--_-----~-_---

I
I

L

I
L

1
RDS
ERR
A

I
I

I
v

____ J

veo
FAST

o r---+---=----,
VCO
SLOW r----------------~

"--

10-0522

-

• Ad just able delays

.....J
RDS CLOCK
DLY H

I

I
L

L

Figure 4-21 RPIO Read Data Separator,
Block Diagram (Revision I)
4-27

I

The RDS CLOCK DLY H is used as a reference signal in setting RDS ERR B, while RDS DIVIDE 2 is used to
set RDS ERR A. Ifboth RDS ERROR flops are set, a signal is generated that resets both flops. As such, they
function as a phase comparator (see Figure 4-22).

OUT OF
PHASE

OF
IN PHASE OUT
PHASE

RDS ERRA

ri. ..JL --1i

RDS ERRS

----1t,

,

-+ J""1.
I

I

10-0525

Figure 4-22 RDS Error Flops Waveforms
The outputs of RDS ERR A and B are fed to two differential amplifiers; each provides control voltage for individual voltage controlled oscillators (VCOs). Only one VCO is enabled at any time by the type of drive in use.
The differential amplifiers are equipped with low-pass fIlters that remove jitter and random frequencies in the
pulse train, thereby effectively averaging the frequency of the incoming bits. Any difference in the set times of
RDS ERR A and RDS ERR B appears as a differential input to the differential amplifiers and causes an output
to the VCO of such a polarity as to correct the error in VCO frequency, thereby synchronizing the VCO to the
incoming clock bits.
The frequency of each VCO is set at the I s rate component of the disk drive it serves. Consequently, VCO SLOW
is set to operate at approximately 2.5 MHz while VCO FAST operates at approximately 5.0 MHz. The output
from the enabled oscillator is used to complement a frequency divider flop (RDS DIVIDE 2) whose Is output,
in turn, is applied to a window gating network (see Figure 4-23).

---JIolL.__..J!clL.__..JIolL.__....

RDS DATA IN _ _

Ic"lL_ __

ROS DATA W I N O - - - - - - - ' " " ' "

DATA CLOCK WIND - - - - - - - - ,

RDS DIVIDE 2 - - - - - - - '

ROS DATA OUT

ROS CLOCK OUT

--------+----'
-------1

Figure 4-23 Optimum Window Generation Waveforms
The RDS DIVIDE 2 (I) pulse train, which is essentially a clock window, is applied to the RDS DATA WIND and
RDS CLOCK WIND flops in complemented fashion through a latching arrangement of gates. The RDS DATA
WIND and RDS CLOCK WIND flops are reset by the separated clock pulse RDS CLOCK OUT H. On the TE of
RDS CLOCK OUT, the data window is opened and the clock window is closed. These conditions prevail until
the RDS DIVIDE 2 once again transitions to the I state. At this time, the data window is closed and the clock
window is opened.

.. .i

When the data window has been generated, the two delay lines associated with RDS CLOCK DLY H are adjusted
to set the data and clock pulses in the center of their respective gating waveforms.
4.9 RPIO COUNTERS
The sector and word counters are described in the following paragraphs.
4-28

4.9.1 Sector Counters
The RP I 0 sector counter consists of eight separate sector counters which operate independently to count sectors
on eight possible Disk Pack Drives (see Figure 4-24 and DWG RPIO-O-SC). Each counter utilizes the raw sector
pulse train from its associated drive; a pulse that is detected by the transducer and from which index pulses have
not been removed. If two pulses less than 350 JiS apart are detected, the logic interprets this as an index pulse or
sector O. This detected index pulse then forms SCC X CLR PLS H, which is used to clear the sector counter. The
SCC X + I pulse is formed from all other sector pulses (other than at index time) and is used to increment the
sector counters.

SCCO +1 PLS H

SCCO CLR

PLSH------~--------------~--------+_------------~---------J

SCCI-S+l PLS H

l

SCCI-S CLR PLS H

1----------------,

I
I

SAME AS ABOVE AND BELOW

1

L..- _ _ _ _ _ _ _ _ _ _ _ _ _ _ ---I

SCC7+1 PLS H

SCC7CLRPLSH------~--------------~--------+_------------~---------J

10-0525

Figure 4-24 RPlO Sector Counter, Block Diagram
4.9.2 Sector Counter Buffer

I

The RPI 0 Sector Counter Buffer is shown in Figure 4-25 and in DWG RPI O-O-SCB. The state of the selected
sector counter is fed into the buffer from the multiplexer by SCB FM SC REG JAM L. This normally occurs
250 ns after the selected sector counter has changed its count. However, when IBC DOD CHis true (this is the
time when the sector counter bits are being strobed onto the I/O bus), the sector counter buffer is prevented
from changing. An SCB FM SC REG JAM L may have been suppressed during that time; therefore, an SCB FM
SC REG JAM L is unconditionally generated on -IBC DOD C H. IBC DATAO SET L is used to update the Sector Counter Buffer with the DATAO that selects the drive. This prevents transfer of the old sector count back
to the processor with a DATAl, when de-selecting one drive and then selecting a new one.

I

4.9.3 Sector Counter Control Pulses
A simplified block diagram of the RP 10 Sector Counter Control Pulses circuit is shown in Figure 4-26. Thedetailed logic appears in DWG RPIO-O-SCC. This circuit comprises eight identical sections of logic. Each section
receives the DTC SEC PLS X L for a particular drive, and produces three control signals. The signal SCC X

4-29

CHANGE COUNT L is the result of inversion, pulse amplification, and reinversion. Each sector pulse is also
delayed by 500 MS, ANDed with the output of the pulse amplifier, and inverted to produce SCCO + I PLS H.
The output of the delay (-SCC X SS) is also ANDed with the inverted output of the pulse amplifier to yield
SCC X CLR PLS H. These outputs are then applied to the sector counter and the sector counter multiplexer.

lBC

DATAO

SET L - - - - . I

SCM CHANGE COUNT L

-IBC DOD C H

SCM BIT 0 L

----_-+---------!

-----------------------+1

1\

SCB
BUF

0
0
-SCM BIT 0 H - - - - - - -

1\

SCM BIT 1 L - - - - - - - - - - - - - -

1\

SCB
BUF

1

0
-SCM BIT I H

SCM BIT 2 L

------------------------+1

1\

- - - - - - - - - - - - - - - - - - - - - - - + 1 1\
SCB
BUF

2

0
-SCM BIT 2 H

---------------------1

1\

SCM BIT 3 L

-----------------------+1

1\

SCB
BUF

3
0
-SCM BIT 3 H

SCM BI T 4 L

- - - - - - - - - - - - - - - - - - - - - - - + 1 1\

-----------------------1

SCB
BUF

4
-SCM BIT 4 H

o

---------------------1
'O-O~26

Figure 4-25

RPIO Sector Counter Buffer,
Block Diagram (Revised)
4-30

OTC SEC PLS 00 L

--~---t----<-t

500".

t-----1I----+--+--------

SCC 0 CHANGE
COUNT L

t----,I----+--------

SCC 0+1 PLS H

'----------1 "
-( SCC 055)

OTC SEC PLS
01-06 L

1---------

I ---------- l

+

-t-

(SAME AS ABOVE)

L ----------

I·

t---1--+---t---------

OTC SEC PLS 07 L

t--f---+--------1--_ _ _ _ _-1

"

t--------

SCC 0 CLR PLS H

SCC 1-6 CHANGE
COUNT L
SCC 1+1-6+1 PLS H
SCC 1-6 CLR PLS H

SCC 7 CHANGE
COUNT L

SCC 7+1 PLS H

SCC 7 CLR PLS H

-(SCC 7 SS)
10-0527

Figure 4-26 RPIO Sector Counter Control Pulses,
Simplified Block Diagram
4.9.4 Sector Counter Multiplexer
The RPIO Sector Counter Multiplexer (DWG RPIO-O-SCM) is shown in block diagram form in Figure 4-27. The
Sector Counter Multiplexer selects the outputs from one of the sector counters contained in the RPIO (eight
counters are provided; one for each drive). Selection is determined by the DAR SEL DRIVE 0-7 L signals - each
DAR SEL DRIVE signal corresponds to a specific disk drive. The DAR SEL DRIVE signals also enable selection
of one of the eight SCC CHANGE COUNT L signals.
4.9.5 Word Counters
The RPIO Word Counters are shown in block diagram form in Figure 4-28 and in DWG RPIO-O-WDC. The circuit
contains two 3-bit ring-tail counters that count the six bits per byte and the six bytes per word being shifted. The
circuit also contains a modulo 128 counter to count the number of words in each data field.
Input is from DTC SHIFT BIT CNT L. This signal is generated in the Data Transfer Control and results from
either DTC READ IMAGE (1) Land DTC READ CLOCK PULSE L; DTC WRITE CLOCK L, DTC WRITE
ENABLE L, and DTC PAR CaNT (0) L; or DTC READ ENABLE L, DTC READ CLOCK PULSE L, and DTC
PAR CaNT (0) L.
Each DTC SHIFT BIT CNT L increments the WDC BIT SHFTCT counter by 1. At the count of 5, the gated inputs to WDC WD CT + I A & Band WDC BYTE CT + 1 become enabled. The next DTC SHIFT BIT CNT sets
the three flip-flops at the same time that the bit shift counter returns to 0, producing DTC BIT SHFTCT EQ O.
Setting WDC BYTE CT +1 produces an increment input to the BYTE SHFTCT counter, thus, causing it to increment by 1. The three + 1 flip-flops are cleared by DTC DATA CLOCK DLY and reset again at each BIT SHFTCT
of 5. If WDC BYTE SHFTCT EQ 5 is true when the +1 flip-flops are set, the WDC DATA WD CT register increments by 1. Each complete cycle of the BYTE SHFTCT register increments the word counter until it reaches
overflow at word 128. This produces WDC WORD COUNT EQ 0 L, which indicates the end of a particular data
field.
4-31

DAR
SEL
DRIVE
OL

SC 0
BITO
(0) L

~

DAR
SEL
DRIVE

IL

DAR
SEL
DRIVE

2L

DAR
SEL
DRIVE
:5L

DAR
SEL
DRIVE
4L

DAR
SEL
DRIVE
5L

DAR
S.EL
DRIVE
7L

!

r,;1.:':'"

SC I
BIT 0
(0) L

v

~
L.:...

SC 2
BITO
(0) L

L.:...

(0) L

"'SCM
...- BIT
0 H

JA1

SC 4
BITO

"L.:...

(0) L

!

~
l_:~

SC 5
BIT 0
(0) L

v

-rt:
-L.:.:....

SC 6
BIT 0
(0) L

(0) L

...

_t....-_~_.Io...._

--~_i.oo--_~--t--

"'SCM
BITI H

I

I

2---.1

(0) L

L"'SCM
BIT 2 H

I

SAME AS ABOVE

SCO-7

i

BIT:5~

(0) L

-

r

.rt\
-L.:.

SC 7
BIT 0

SC 0-7
BIT 1
(O)L

-

f

.r;-

SC :5
BIT 0

SC 0-7
BIT

DAR
SEL
DRIVE
6L

I

SC 0-7
BIT 4

_ _ ---oJI-----

(0) L

"'SCM
BIT :5H
"'SCM
BIT4H

SCC 0
CHANGE
COUNT L
SCC I
CHANGE
COUNT L
SCC 2
CHANGE
COUNT L
SCC :5
CHANGE
COUNT L
SCC 4
CHANGE
COUNT L
SCC 5
CHANGE
COUNT L
SCC 6
CHANGE
COUNT L

"

....

V

"
"

SCM
CHANGE
COUNT H

'\0

"

"

V
10-0528

SCC 7
CHANGE
COUNT L

Figure 4-27 RPIO Sector Counter Multiplexer,
Block Diagram

The bit and byte count flops are cleared by DTC CLEAR SHFTCT H. The word count flops are cleared by WDC
CLEAR DATA WD CT H, which is the ORed result of DTC SELECTED SECTOR, DTC WRITE GAP, or DTC
WRITE HEADER.

4-32

HFTCT H
DTC
SHIFT BIT CNT

woe
-<0

8-

hJ"Il

~.

r"

~

f"

~

N

:;.ti

~

0
.j:>.

W

0-

w

....

...

'---;"
WDC
BIT
SHFT
CTOO
0

BYTE
SHFT

WDC
BIT
SHFT
CT DI
Dt--

A

f--

0

~ ~J

r+

WDC
BIT

A

~

::l

F-;;- -0 IWD00CTor--

L--

f-

II~

f-

tp-

~

i3
,....,

WDC
BYTE
SHFTCT
EQO H

t---

fA

f-

f-

J

woc
WD
CT
+IA

DTC BI
SHFTC
OD (I) I

1

r--

WOC
DATA
WD CT

~

tr

02 0

t--

f-

CT02
0

'-WOC
WOCT+IB
I--

'----

WDC
DATA
WO CT

~

rrn

f--'-

I--

1--.-

A

t---- t - 03 0

I--

A

01

I

woc

WD CT
+IB

01

J

I BYTE
WOCCT
+1

AI

°

I-I-A

I--

J Al

f-

WOC
DATA
WO CT
04 0

A

fA

t-

DTC BI
SHFTC
01 (0) I

r---

r-----OTC WRITE GAP (I) H
DTC WRITE GAP (0) H

--=::l..-.

=:3

I

V

I

~

W

~

i-J

t--

'"Af- r----"j"

f-

DTC SELECTED SECTOR H

tT 'I

~f- ------;-

f

r
III

~
A

I---

A r--- r-----"j

:

DTC SHIF
BIT CNT

I

WDCT
01 0

r--

I--

I--

DTC
DATA ClOC
OlY

o-

A

WDC
BYTE
SHFT

~.

'-'

7f""-

A

:;.ti
CD
'"0-

f-

A

~

~
~
~

r--

WDC

~~

WDC
BYTE
SHFT
CTDbt--

;;!
0

t-II---

7d[n

DATA

8- r-,J
A

0

L---

.....
CD

~

WDC
BIT
SHFTCT
EQO H

WOC
DATA

II

,.. ~ ~J

tp- WD~

(")

7 -0------;L--

~ ~I

r--

WOC BYTE SHFTCT E05 l

:g~T+1A-

BYTE CT+l-

8- ~J

00

:E
0

..& ~~

L-

WOC
DATA
WD CT

~

05 0

t-t--

GJ

1

WOC
DATA
WO CT
~~

WOC CLEAR DATA

OTC WRITE HEADER (0)
10-0529

WDC WORD COUNT EOO l

4.10 RPI0 COMPARATORS
4.10.1 Header Compare and Designation Error
The RPI 0 Header Comparator and the circuit that brings up a designation error are shown in Figure 4-29 and in
DWG RPIO-O-HCDE. Header comparison is indicated by no output (HCDE HEADER COMPARE L) from the
composite AND/OR gating associated with cylinder and surface/sector comparison. The comparison operation is
split in two to provide header protection by continuous monitoring of cylinder address (-HCDE CYL COMP H).
Protection is realized any time a lack of cylinder comparison is combined with the completed search operation
(DTC SEARCH COMP (1) L) to generate HCDE HDR READ ER H. This error condition immediately terminates
operation by setting CXR SEARCH ER 2 on -DTC SRCH COMP STROB H, which fires DTC FINISH and shuts
down the controller.
The states of each cylinder, surface, and sector flop in the data address register are individually ANDed with the
opposite state of equivalent bits in the LPR. The results of each ANDing are ORed to produce an output if any
one bit state does not compare with its counterpart.
Logic is provided to indicate that a drive, not included in the system, has been designated (HCDE DRIVE DES
ERROR H). The SEL DRIVE signals from the data address register are applied to an OR gate in this comparator.
The input to the gate can be jumpered for any illegal drives, thereby producing an error indication if one of these
drives is commanded.
The sector and surface designation error portion of this circuit contains a flop for each error. These flops are set
by decoding the appropriate flops in the data address register. The error flops are reset by ANDing IBC CON
CLR H with 10BD 28 H for sector designation error and 10BD 29 H for surface designation error. Signal DTC
RPOI L is applied to the designation error logic to allow decoding of illegal addresses for either an RPOI or an
RP02 drive. The setting of each flop is conditioned by the ANDing of DTC OP CODEOO (0) Land IBC SET
DAR ETC DLY L. Both designation error flops are cleared by IBC INITIAL CLEAR C H.
4.11

RPIO DATA FLOW

The flow diagrams contained in Volume 2 of this manual are discussed in the following paragraphs.
4.11.1

Starting

On receipt of DATAO 250 or 254, IC lOB DATAO CLEAR generates IBC INITIAL CLEAR if IBC BUSY and
CC ACTIVE are both false. The assertion of IBC INITIAL CLEAR generates CLR IBC GEN CLR if 10BD 00 is
false when IBC LOCAL is false. The INITIAL CLEAR signal also clears ILLEGAL WRITE, ILL COM WL BUSY,
OVERRUN, DISK WD PE, DISK SEC PE, PARITY ERROR, PREVENT TERM, WRITE EVEN PAR, SEARCH
ER 0-2, NON EX MEM, CHAN DATA PE, CHAN CONT PE, and CW XFER COMP in the Condition Register;
the cylinder, surface, sector, and drive registers in the DAR; and the OP CODE register in the DTC. In addition,
when IBC INITIAL CLEAR is asserted, it clears SET CYL, SEEK START, NO SECTOR, INDEX, SET HD ADV,
CLEAR HD REG, SET HD REG, HEAD SET, CHN PLS BUF, and DTC4 CI3M, also in the Data Transfer Control
and it clears IBC DONE and both sector and surface designation errors in the HCDE.
When IBC INITIAL CLEAR is generated, it results in DTC CLEAR A and B which clear ARO-17 and ARI8-35,
respectively. It is used also to generate DTC CLEAR SHFTCT (which, in turn, clears the bit and byte counters)
and to generate the DTC CLEAR PAR PLS which results in DTC CLEAR PARITY and CLR DTC PAR CONT.
With all of the above conditions cleared, receipt of IC lOB DAT AO SET results in IBC DATAO SET. If the
controller is busy, IBC BUSY is set. This signal ANDed with IBC DEVICE SELECTED (true only if IBC LOCAL
is false) generates the ILL COM WL BUSY condition to indicate that a command was given to the control before it
finished a previous command. If the synchronizer is not busy (lBC BUSY (0)), IBC SET DAR ETC A and B is
generated.
4-34

DAR SEC 04 (1) H
lBC
CON
CLR
H

DAR SEC 03 (1) H
DAR SEC 02 (1) L

lBC
INITIAL
CLEAR
CH

DAR SEC 01 (1) H

1--_-+---1----

DAR SEC 00 (1) L

HCDE SEC DES ER H

DAR SEC 01 (1) H
HCDE
SEC
DES
ER

DAR SEC 02 (1) L
DAR SEC 03 (1) L
lOBO 28 H

o

--~----------------------------------r_~ A

DTC RPOI L
DTC OP CODE 00 (0) L
lBC SET DAR ETC DLY L
DAR SURF 01 (1) L
DAR SURF 00 (1) H

DAR SURF 03 (1) H

1-_-1---1------11----- HCDE SURF DES ER H

DAR SURF 00 (1) L

HCDE
SURF
DES
ER

DAR SURF 01 (1) H
DAR SURF 02 (1) H
lOBO 29 H

DAR SEL DRIVE 01-07 H

o

---;.~GJI-------------------------.

HCDE DRIVE DES ERROR H

DAR CYL 01-128 (0) L
LPR 18-25 (1) L

1----------.--------_ '" HCDE CYL
DAR CYL 01-128 (1) L

COM P H

1------.. HCDE HDR READ ER H

LPR 18-25 (0) L
DAR SURF 00-04 (0) L
'" HCDE SURF/SECT COMP H
LPR 26-30 (1) L
DTC
PARITY
(O)H

DAR SURF 00-04 (1) L
LPR 26-30 (0) L
DAR SEC 00- 04 (0) L

DTC
SEARCH
COMP
(1) L

LPR 31-35 (1) L

DAR SEC 00-04 (I) L

10-0530

LPR 31-35 (0) L

Figure 4-29 RPIO Header Compare and Designation Error,
Simplified Block Diagram (Revised)
4-35

The A version of this signal loads the drive and cylinder registers in the DAR and sets disable stops in the condition register for Disk Sector Parity Error (if IOBD 24 is set), Channel Data Parity Error (if IOBD 25 is set), or
Disk Word Parity Error (if IOBD 26 is set), and SET CXR WRITE EVEN PAR (if IOBD 35 is set).
The B version of IBC SET DAR ETC sets the Surface, Sector, and Op Code Registers in the DAR. The Op codes
are directed to the Data Transfer Control where they are decoded into the commands listed in Table 4-1.
Table 4-1
Data Transfer Control Commands
DTC OP CODE Bits

DTCCommand

0

1

2

0

0

0

Read Data

0

0

I

Write Data

0

I

0

(Not Used)

0

I

I

Write Hdr

I

0

0

Seek

I

0

I

Clear Attn

I

1

0

NoOp

I

I

I

Restore (Recalibrate)

If the decoded command is NO OP or CLEAR ATTN, the commands are executed. If the command is RESTORE
or SEEK and the controller is in Local Mode, IBC LOCAL START (DWG RPIO-O-FCL) will assert CLR IBC SET
DAR ETC DLY. If the RPIO is not in Local Mode (IBC LOCAL (0», IBC BUSY (0) and IBC DEVICE SELECTED
will assert CLR IBC SET DAR ETC DLY, which is combined with IBC LOCAL (0) to generate SET IBC LOAD
ICWA. This ICWA signal then loads the Initial Control Word Address contained in IOBD 27-34 into bits 27-34
of the Assembly Register.
After a 500 ns delay, SET IBC SET DAR ETC DLY generates CLR IBC LOAD ICWA and with DTC CLR ATTN
clears the appropriate ATTN 0-7, as specified by IOBD 27-34. If a seek command is decoded, SET IBC SET DAR
ETC DL Y sets DTC SET CYL which, in turn, starts the Seek operation. If a restore command is given, SET IBC
SET DAR ETC DLY sets DTC BUS LINE STROBE directly, thus enabling DSBC UB 6 to execute the command.
When a data transfer command is received (OP Code 00 on a 0), SET IBC SET DAR ETC DLY triggers CC
PREVENT TEST. If HCDE SEC DES ER is asserted at this time, 1-+ HCDE SEC DES ER results. If HCDE
SURF DES ER is asserted at this time, 1-+ HCDE SURF DES ER results. Either of these conditions raises DTC
FINISH, thereby terminating operation.
When CC PREVENT TEST fires, it starts the I J1S CC PREVENT TEST TIMEOUT. At the end of this delay, the
conditions for CC PREVENT START are checked. The channel will not start (CC PREVENT START) if a drive
designation error has been detected (HCDE DRIVE DES ERROR), the power supply voltage has dropped (CXR
PS FAIL (1», a sector designation error has been indicated (HCDE SEC DES ER), the disk is not ready (CXR
DISK NOT RDY), or a surface designation error has been detected (HCDE SURF DES ER). In addition, the
channel will not start if the drive is in read only mode (DSBC SEL UNIT RD ONLY) and either of the two write
conditions has been commanded (DTC or CODE 00(0) " DTC or CODE 02 (l

».

4-36

If CC PREVENT START is true, IBC DONE and IBC GEN CLR are both set. This condition generates CXR
PI ENABLE if IBC BUSY and IBC LOCAL are both false and clears DIS DSPE STOP, DIS CDPE STOP, and
DISABLE DWPE STOP in the condition register. IfCC PREVENT START is not true, IBC BUSY and CC
ACTIVE BUF are set. This condition sets CC ACTIVE after 40 ns, if IC CHAN STARTED is on a 0 (proceed on
DWG RP 1O-O-FCCC to start the channel). In addition, the combination of IBC BUSY (1) and CC ACTIVE BUF
(1) sets DTC CLEAR HD REG if DTC OP CODE 0 is false, thereby clearing and setting the head register in the
drive as shown in DWG RPIO-O-FCDC.
If CC ILLEGAL WRITE is asserted at the time of CC PREVENT TEST TIMEOUT, CXR ILLEGAL WRITE is set.
4.11.2 Recognizing Headers
CC ACTIVE BUFF sets as a result of CC PREVENT TEST Hand -CC PREVENT START to indicate that the
RPIO is ready to assume control of the data channel as soon as the channel is free. If CC CHAN STARTED is
on a 0 (indicating that no device further down the daisy chain is accessing the channel), CC ACTIVE sets. This
condition results in CC CHAN START OUT and, when the channel is ready, it sends IC CHAN BUSY IN to the
RPIO. At this point, the RPIO has control of the data channel and asserts CC STRT SCH.
Assertion of CC STRT SCH sets DTC SEARCH and the RPIO then waits for the fIrst sec.tor pulse. On receipt of
the sector pulse, DTC SECTOR DLY H is triggered. This delay allows the drive head register to increment during
multiple surface data transfers (see Paragraph 4.11.6). The delay then triggers a 350-#lS delay (DTC SEARCH
SYNC) which sets READ REFUSE. When READ REFUSE sets, it turns on the read circuitry (see Paragraph
4.11.5). The separator's window is bypassed during this time, but after the delay times out, DTC READ HEADER
is conditioned for setting by detection of a 1 bit and the window bypass is removed.
The 36 bits following the 1 bit are read into the LPR (see Paragraph 4.11.7), and when a full word has been loaded,
the Byte Counter and Bit Counter return to O. At this time, DTC PAR CaNT sets indicating that the parity bit is
about to be read. The clock pulse following the parity bit clears PAR CaNT (see Paragraph 4.11.10), and when
parity control clears, DTC SRCH CMPR STRB is triggered. This strobe has a 600 ns delay to allow for gate delays
throughout the compare circuitry. When it times out, DTC SRCH CaMP is set if no parity error occurred and the
headers compare. The headers are checked via HCDE COMPARE signal which determines if the address bit is a 1
and the data bit it 0, or the address bit is a 0 and the data bit a 1. If either of these conditions is true for any bit
between 18 and 35, HCDE HEADER CaMP is false. The data bits in the LPR are checked against the address bits
in the DAR. If they compare, DTC SRCH CaMP sets, and DTC SEARCH clears. If they do not compare, DTC
SEARCH remains set. An R302 delay is fIred unconditionally ~88 #lS for RPOI or 44 #lS for RP02) producing
BETA GAP. The BETA GAP bypasses window selection in the separator to allow resynchronization of the separator. This is necessary due to the discontinuity between the header portion, which is written once, and the data
portion, which can be written any time on any drive. At this point, the flow changes to either a read or a write.
4.11.3 Reading Data
When DTC SRCH CaMP STRB times out (after the header has been read), it sets READ GAP. However, if DTC
SRCH CaMP did not set, DTC READ GAP clears when BETA GAP times out and the controller waits for the
next sector pulse. If DTC SRCH CaMP did set, READ GAP remains set, and the detection of the fIrst 1 bit
then sets READ DATA. After 200 8 words are read with parity (see Paragraph 4.11.7), DTC READ LPR sets,
and the next 37 bits are exclusive ORed into LPR 0-35 and LPR 36 (see Paragraph 4.11.10). When DTC PAR
CaNT switches to a 0, indicating that the 37th bit has been read, DTC READ LPR clears thereby setting DTC
GAP and triggering a 600-ns delay. The delay, in turn, sets CXR DISK SEC P E if any LPR bit is not on a 1.

4-37

4.11.4 Writing Data
After SEARCH COMP has been set, the transition of DTC READ HEADER to 0 sets DTC WRITE GAP after a
delay of 3 [JS. This delay ensures that the first bits written will not destroy the last few bits of the header. When
WRITE GAP sets, the controller writes Os onto the disk without parity. After four words of Os have been written,
WRITE DATA is set, and parity control switches to a 1. Parity is then written for the last sync word (this parity
bit is the 1 bit that sets READ DATA during a read operation). After 2008 words of data are written (see Paragraph 4.11.8), DTC WRITE LPR sets, and the contents of the LPR are written on the disk. The controller then
writes a second LPR as guard bits and, after this has been done, DTC GAP sets. The DTC ERASE flop, which
controls erase current in the drive, is turned off 20 [JS after writing has stopped to allow data that was written to
pass under the erase heads before they are turned off.
4.11.5 Commanding the Drive
Drive commands and data are sent out on the same eight bus lines. Three tag lines indicate whether the information on these lines is data or commands. The information on the bus lines is strobed into the drive by bringing
up the tag signal in the middle of the command. The only exceptions to this are read and write commands. When
either of these commands is issued, both the bus line and tag line remain true throughout the time that read or
write is true. These commands are executed by making SET WRITE or SET READ true by ORing the various
write flops or read flops.
NOTE
All read or write flops are cleared by the next flop transition
to a 1 state. This ensures that SET WRITE or SET READ do
not momentarily go false if one flop should be slow.
4.11.6 Reading Multiple Sectors
Multiple sector reads are performed only in on-line mode (not local). During on-line operations, normal termi-

I

nations are performed via the channel going -BUSY. If the channel stays active after transferring a sector, the
RPl O continues reading. When DTC SRCH COMP STRB H is fired, if the cylinder portion of the succeeding
header was read incorrectly, CXR SRCH ER 2 is set. This causes DTC FINISH to fire and shut down the
controller.

If the last sector read was sector 9 (4 in RPO 1), the controller moves down one surface. An advance head command to the drive, which increments the drive's head register by 1, accomplishes this move. The controller
enables to determine which sector it is reading or writing after the first sector; therefore, the program must keep
track of surfaces and sectors. After reading sector 9 (4 in RPO 1), an index pulse occurs that sets DTC INDEX.
The next sector pulse then sets DTC ADV HEAD, which increments the drive's head register. If reading is initiated on sector 9 (4 in RPO 1), the index pulse occurs before DTC SRCH COMP sets and DTC INDEX is set by
DTC SRCH COMPo
NOTE
The index pulse occurs within 500 ns of the sector pulse for
sector 9 (4 in RP01), and not sector zero.

If the heads are advanced when the drive is on the last surface, the drive sends back EOC (End of Cylinder),
which terminates operation.

4-38

4.11. 7 Loading Words
4.11.7.1 Header Words - When DTC READ HEADER sets, header bits are shifted into the SR, and the Bit
Counter is incremented (the bit counter is a counter that returns to 0 after six counts). When the Bit Counter
returns to zero, the SR is loaded into the LPR (it is actually an exclusive-OR operation, but the LPR has already
been cleared). The Byte Counter is incremented each time the Bit Counter returns to O. When the Byte Counter
returns to 0, the Word Counter is incremented. When the Byte and Bit Counters both return to 0, a full header
word has been read.
4.11.7.2 Data Words - Data words are loaded in the same fashion as header words except that they are loaded
into the AR as well as the LPR. Each time the AR is full, the RPIO issues a device pulse indicating to the channel
that data is on the lines. The channel responds with a channel pulse when it is ready to receive another word.
DTC SUPR BUF LD is used to prevent the buffer from being loaded during the fIrst byte of data, because the
Byte Counter is on a zero. Each parity bit is exclusive ORed into LPR 36.
4.11.8 Transferring Words
4.11.8.1 Write Header - The header words are loaded into the SR and SR BUF directly from the AR. The Bit,
Byte, and Word Counters function exactly as they do in reading.
4.11.8.2 Write Data - Each data word is loaded into the AR by a channel pulse, only after a device pulse has
been issued to the data channel indicating that the RPlO is ready to receive that word. The LPR is loaded only
from the SR BUF.
4.11.9 Writing Headers
The WRITE HEADER command starts with sector zero and writes around the pack. It is impossible to write any
sector without first having written sector zero. This is mainly a software function. The controller starts writing
on the sector following index and writes 36-bit words continuously until the second sync fIeld has been written,
at which time the WRITE DATA mode is entered. The controller writes without parity; if parity were used, the
sync fIeld would contain Is. Thus, because the header is checked for correct parity, the first bit of the first word
following the header word appears as the parity bit. The software must calculate parity and supply the correct
bit. This operation continues for each sector.
4.11.10 Generating Parity
The PARITY CONTROL flop toggles to a 1 each time 36 bits have been read, except during the sync zones, which
are indicated by READ REFUSE (1), READ GAP (1), WRITE HEADER (1), or WRITE GAP (1). The PARITY
flop is toggled for each 1 bit read or written. If this flop is on a 0 at the end of 36 bits (PARITY CONTROL -+ 1),
a 1 is written.
During read, the parity bit toggles both the PARITY flop and the LPR 36 flop. This operation is required because
LPR 36 (parity bit of LPR word) is read differently than it is written. LPR 36 is written as the parity of the LPR
word but is read as the exclusive OR of the parity bits of all the data words.

4-39

4.11.11 Terminating
All terminations occur when DTC FINISH is generated. Gate conditions that fire DTC FINISH are listed below
by triggers. Each gate indicates a different condition.
CXR SEARCH ERR (1) - Indicates that three index pulses have been received since
attempting to read a header and, therefore, the header can not be found or read.
CXR ILL WRITE (1) - Indicates that a write operation was commanded on a writelocked drive.
CXR DISK NOT RD Y - Indicates that the controller tried to communicate with a
drive that was not ready. A file unsafe condition will cause this termination.
CXR DISK SEC P E - Indicates that a disk sector parity error occurred, and that the
stop disable was not set.
CXR DISK WD P E - Indicates that a disk word parity error occurred, and that the
stop disable was not set.
IBC CONSET - Indicates a programmed halt.
HCDE SURF DES ER (1) - Indicates that the data transfer command specified a
nonexistent surface.
DTC GAP (1 )"(CC ACTIVE (O)VCC INHIBIT (1 M;XR PS FAIL) - The combination
of these conditions terminates all write operations and normal read operations. DTC
GAP prevents termination until the controller is between sectors. It terminates if
CC ACTIVE (0) (data transfer complete), CC INHIBIT (1) (overrun or channel data
parity error), or CXR PS FAIL (1) (power supply failure) are true. The presence of
CC PREVENT TERM 0 prevents termination during a write until the next sector pulse
is received. This operation allows erase to stay on longer than write.
DSBC SEL UNIT EO CYL - Indicates that a data transfer did not terminate via the
data channel before the drive head register was incremented past its last surface.

CC ACTIVE (O)"CHN PULS BUF (0) - Indicates that channel access was lost and
gained prior to receiving the first channel pulse.
HCDE SEC DES ER (1) - Indicates that the read command specified a nonexistent
sector.

4-40

Chapter 5
Maintenance

5.1 INTRODUCTION
Maintenance of the RPIO conforms to the accepted maintenance procedures of all electronic equipment presently
in use, i.e., an optimum amount of preventive procedures, performed on a routine schedule, can eliminate many
costly equipment breakdowns and can forecast failures before they occur. When a specific item does fail, the
design of the equipment allows for quick replacement of modular elements, thus restoring the main equipment
to service in a minimum of time. A design objective of the RPI 0 Disk Pack Synchronizer is to provide a dependable and relatively maintenance-free assembly. This chapter contains both preventive and corrective procedures.
5.2 PREVENTIVE MAINTENANCE
Preventive maintenance consists of tasks performed at periodic intervals to ensure proper equipment operation
and minimum unscheduled downtime. These tasks consist of visual inspection, operational checks, adjustment,
and replacement of marginal components.
The preventive maintenance schedule depends on the environmental and operating conditions that exist at the
installation site. Under normal environmental and work-load conditions, the recommended preventive maintenance schedule consists of inspection and cleaning every 600 hours of operation or every 4 months, whichever
occurs first. However, relatively extreme conditions of temperature, humidity, dust, and/or abnormally heavy
work loads demand more frequent maintenance.
5.2.1 Preventive Maintenance Procedures
Preventive maintenance procedures for the RPOI and RP02 Disk Pack Drives are not included in this manual. For
these procedures refer to the Vendor maintenance manual supplied with the equipment.
5.2.1.1 Mechanical Checks - Inspect the RPIO periodically as follows:
a.

Visually inspect the unit for general condition.

b.

Clean the interior and exterior of the rack using a vacuum cleaner or clean cloth moistened in
nonflammable solvent.

c.

Inspect all wiring and cables for cuts, breaks, frays, deterioration, kinks, strain, and mechanical
security. Tape, solder, or replace any defective wiring or cable covering.

5.2.1.2 Test Equipment Required - Maintenance activities for the RPI 0 require the standard test equipment
and special materials listed in Table 5-1, in addition to standard hand tools, cleaners, test cables and probes.
Special test equipment required for any adjustments are given as part of the adjustment procedures.

5-1

Table 5-1
Test Equipment Required
Equipment

Manufacturer

Designation

Multimeter

Triplett or Simpson

Model 630-NA or 260

Oscilloscope

Tektronix

Type 453

Clip-on Current Probe

Tektronix

Type P6016

XIO Probe

Tektronix

P6008

Recessed tip, 0.065 in. for wirewrap terminals

Tektronix

206-052

Hand Unwrapping Tool

Gardner-Denver

500130

Hand-Operated Wire-Wrap
Tool with a 26263 bit for 24
AWG Wire and 18840 Sleeve

Gardner-Denver

14HIC

Module Extender

DEC

Type W980

Diagnostic Self-Test Routine

DEC

MAINDEC-IO-D5MA
MAINDEC-IO-D5NB
MAINDEC-IO-D50B

5.2.1.3 Electrical Checks - Perform the power supply output checks described in Table 5-2. Use a multimeter
to make the output voltage measurements with the normal load connected. Use an oscilloscope to measure the
ripple content on all dc outputs of the supply. Voltage measurements should be made at the logic racks. The
+10 and -15 Vdc power supplies are not adjustable; therefore, if any output voltage or ripple content is not within specifications, consider the power supply defective and initiate troubleshooting procedures.
Table 5-2
Power Supply Output Checks
Measurement Terminals
at Power Supply Output

Acceplable Output
Range (V)

Nominal Output
(Vdc)

Maximum Output
Ripple (mVrms)

Red (+) to Black (-)

+10

+9.5 to 11.0

800

Black (+) to Blue (-)

-15

-14.5 to -16.0

100

NOTE
This power supply is the Type 728 (728A for 50 Hz) located
at the bottom of the RPIO rack.

5-2

5.2.1.4 Electronic Cheeks - Perform the following electronic checks.

a.

Critical Delays - Check the critical delays in the RPI0 every six months to determine that
they are still within specified tolerances. These delays are listed in Table 5-3, also included
in Table 5-3 is a list of allowable tolerances for each module. Replace or adjust any modules
that are found to be out of tolerance.

b.

Margins - Check the margins for the RPIO and its companion DFIO every six months. Refer
to the Margin Check Paragraph in the PDP-I 0 KAI0 Central Processor Maintenance Manual,
Volume I, and to Tables 5-4 and 5-5.
NOTE
Before margining the RPIO, remove power and remove
the W505 (PS FAIL) Module at location M02. Replace
the module after margining.

Table 5-3
Critical Delay Modules
Drawing No.

Module Type

Signal Name

Location

Nominal Value

Tolerance

DTC TAG LINE
STROBE

F30E-M

I ItS

±20%

DTC-3

R302}
R302

IBC-2

R302

H25E-M

500ns

±20%

IBC-2

R302

IBC SET DAR ETC
DLY
IBC LOAD TEST

B20N-V

21lS

±20%

DTC-3

Table 5-4
DFIO Voltage Margins
Voltage

-15V

+lOV
Rack Row
A
B
tttc
tD
E
ttF
ttH
J
K
L

Low

High

Low

High

2.5
2.5
3.0
3.5
2.5
6.0
6.0
2.5
2.5
2.5

17.5
17.5
17.5
17.5
17.5
17.5
17.5
17.5
17.5
17.5

-18
-18
-18
-18
-18
-18
-18
-18
-18
-18

-12
-12
-12
-12
-12
-12
-12
-12
-12
-12

tThe B130 Modules limit the + lOY margins of panel D.
ttThe B68S Modules limit the +10V margins ofpane1s F and H.
tttThe R303 Module for NON-EXMEM limits the +10V margins of panel C.

5-3

Table 5-5
RPIO Voltage Margins
Voltage
+10V
Rack Row

A
B
C
D
E
F
H

tJ
K
L

tM
ttN
ttP
ttR
ttts
tttT

-15V

Low

High

Low

High

2.5
2.5
2.5
2.5
2.5
2.5
2.5
6.0
2.5
2.5
6.0
6.5
6.5
6.5
7.5
7.5

17.5
17.5
17.5
17.5
17.5
17.5
17.5
17.5
17.5
17.5
17.5
14.0
14.0
14.0
15.0
15.0

-18
-18
-18
-18
-18
-18
-18
-18
-18
-18
-18
-18
-18
-18
-18
-18

-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12

tB68S limits +10V margins of panel J.
ttWS91 and W692 limit +10V margins of panels N, P, and R.
tttB410 limits +10V margins of panels S and T.

5.3 CORRECTIVE MAINTENANCE
The logic description provided in this manual permits the use of standard troubleshooting techniques for isolating
the trouble quickly and efficiently. For economical maintenance under most conditions, replace the inoperative
module with a spare module and return the defective module to DEC for repair or replacement.
5.3.1 General Corrective Procedures
Before beginning troubleshooting procedures, make certain that the processor portion of the PDP-lO, the DFI0,
and I/O Interface are operating properly. Refer to the specific maintenance manual to determine the status. Also
examine the maintenance log to determine if the fault has occurred before and note what steps were taken to
correct the condition. Note also if a cyclic condition has occurred. Visually inspect the physical and electrical
security of all cables, connectors, modules, and wiring. Check the indicator lamps for operation and their glass
covers for cleanliness. In particular, check the security of ground connections between racks. Faulty grounds
can produce a variety of faults.

5-4

5.3.2 Diagnostic Testing
DEC provides special diagnostic programs (MAINDECs) to assist in localizing faults within the equipment.
Functionally, the programs fall into two categories: test and reliability. Test programs isolate genuine go/no-go
type hardware failures that are easily recognizable, while the reliability programs isolate failures that are more
difficult to detect because they are marginal in nature and/or occur infrequently or sporadically. The family of
test programs are written so that, when run successively, they test the equipment beginning with small portions
of the hardware and gradually expand until they involve the entire machine. To accomplish this, they are built
around instructions and portions of instructions whose demands on equipment capabilities progress from simple
transfers and skips to the most involved data manipulations and computations. As portions of the system are
proven operable, they become available to succeeding tests for use in checking out unproven portions of the
machine.
To perform a diagnostic routine on the RPIO alone, run tests MAINDEC-IO-DSMA and MAINDEC-IO-DSNB. If
the RPlO is operating correctly, both tests should run without failure.
5.3.2.1 Vibration Tests - Many malfunctions can be located by performing a voltage margin check while running
diagnostics. The RP 10 should be margined while running D5MA or D50B. The associated DF I 0 should be
margined while running D5NB. In addition, a vibration test may be performed while margining the RPlO and
DFIO together.
5.3.3 Adjustment Procedures
The following adjustment procedures must be performed when it has been determined by diagnostics and troubleshooting techniques that a specific element is at fault but does not need replacement.
5.3.3.1 Read Data Separator Calibration - If parity errors are occurring from random drives, surfaces, or bits,
the read data separator (RDS) is probably at fault. This failure can be caused by a bad module or by poor alignment. Ifit is determined that a module is at fault, the module should be replaced with one known to be good;
otherwise, the following procedure should be performed. The purpose of this calibration procedure is to adjust
the variable elements of the RDS so that data being read from the disk pack drive is correctly separated into data
bits and clock bits under all combinations of bit shift and speed variations from the drive, and worst case circuit
conditions in the RDS.
All delays used in the portion of the circuitry required for operation from RP02 drives are also active when RPOI
drives are being used. Therefore, all adjustments for operation with an RP02 are always required, even if RPOI
drives alone are to be used.
The calibration procedure is carried out using the crystal clock in the controller, and a drive is not required in
order to set up the read data separator. The jumpers connected in steps 2 and 3 of the procedure ensure that all
necessary signals are present.
Equipment required for this calibration procedure is as follows:
a.

Dual-trace oscilloscope, Tektronix Type 453, Tektronix Type 454, or equivalent.

b.

Oscilloscope probes fitted with short (i.e., 3-in. or less) grounding leads.

c.

Jumper leads, 5 red jumpers (4-in.) and I yellow/white jumper (24-in.).

5-S

NOTE
The differential delay between probes should be less than
2 ns. This should be checked by connecting both probes
to a convenient waveform (e.g., T22 H), setting the oscilloscope time base to 5 ns/div. (10 ns/div., if using 453) and
measuring the delay between the channels at the -1.5V level.
The following calibration procedure must be strictly adhered to. Make all adjustments as precisely as possible.
Step
1
2

Procedure
Remove Module Type B152 at location 129.
Connect the following pins to ground:
a. E32 (DTC RPOI L)
b. CIOF (DTC READ HEADER (I) H)

c. L22N (DTC WRITE HEADER (l) H)
d. L24N (DTC READ CLOCK PULSE L)

e. SI51 (Provides RDS SYNC FIELD ENB L)
3

Connect K24U (CRYSTAL CLOCK OUTPUT, 400 ns period) to S251 (MPX
RAW DATA H).

4

Set oscilloscope to trigger internally from channel I; set vertical sensitivity to 1V/div.
and time base to 50 ns/div. negative trigger.

5

Connect channell to T22H and adjust lower potentiometer on module B41 0 in
location T22 until negative-going pulse width on oscilloscope is 50 ns measured
at -1.5V points.

6

Connect S25E (MPX RAW DATA H) to ground, place channell on T29V, set
for positive trigger, and adjust module B312 at location T29 until width of the
pulse to ground is 50 ns measured at -1.5V points. Remove ground on S25E.

7

Connect T22V to ground.

8

Connect channel I to T22H. Set time base for 100 ns/div. Adjust upper potentiometer on module B410 in location T22 until pulses occur at intervals of 200 ns.
Remove ground on T22V.

9

Connect channell to S24L and channel 2 to S24V. Adjust B312 at location S24
until the delay between S24L and S24V is 80 ns.

10

Connect channel I (only channel I is needed) to S28P and adjust B312 at
location T25 until the square-wave seen at S28P is 200 ns to -3V and 200 ns
to ground.

II

Set oscilloscope to I V/div. Set to negative trigger, internally from channel I.
Set the time base to 10 ns/div. for A453 and 5 ns/div. for A454.

12

Connect channel I to S32M and channel 2 to T30H. Adjust B312 in location
S24 until the falling edge of channel I waveform is aligned with the rising edge
of ch31J.nel 2. Measure at -1.5V (see Figure 5-1).

5-6

_------GND

CHI - - _

-+--+--+-IE--I--+-+-- t. 5 V
CH2

10- 0 . . .

Figure 5-1

RDS Waveform No.1
Procedure

Step
13

Set time base 20 ns/div. and connect channell to M30S. Adjust calibrate control
on time base until the pulses are exactly 10 divisions apart on the oscilloscope
face measured at -1.5V points. (If the pulses cannot be adjusted to 10 div apart,
repeat using 50 ns/div. range setting.)
NOTE
The time base is now calibrated on this range to be 20 nsf
div. Do not move the calibrate control or the range (time/
div) control for the remainder of the RP02 procedure
unless instructed to do so.

14

With channel 1 connected to T24D (RDS ERR A (0) H), and channel 2 connected
to T24N (RDS ERR B (0) H), and time base set for 20 ns/div, adjust the upper
potentiometer on module B410 in location T22 until the waveforms seen on
channel I and channel 2 are exactly the same. ("Exactly", in this context means
within the limits of oscilloscope resolution and operator ability.)

15

Connect channell to S32M (RDS CLOCK WIND (0) L), set for negative trigger,
connect channel 2 to S32L and adjust B312 module at location T2s until the
delay between the negative-going edge of channell and channel 2 is equal to
105 ns measured at 1.5V point (see Figure 5-2).

CHI

I
I
I

I
I
I

I+--TI-....
~I

TI=I05ns RP02
TI=I92ns RPOI
10-0"9

Figure 5-2 RDS Waveform No.2

16

Remove the following jumper to ground:
E32D

5-7

Step

Procedure

17

If the RPOI is not to be used with the system, remove the following jumpers:

a. CIOF to ground

b. L22N to ground
c. L24N to ground
d. S15J to ground
e. K24U to S25J

Replace the module Type B152 at location J29.
NOTE
This completes the calibration of the RP02 section of
the RDS.
18

lethe RPOI is to be used with the system, proceed as follows:
a. Jumper point E32H to ground.

b. Check that the following points are connected:
CIOF to ground
L22N to ground
L24N to ground
SI5J to ground
K24U to S25J

c. Check that the module Type BI52 is removed from location J29.
19

Set oscilloscope to trigger internally from channell, vertical sensitivity set to
1 V/div, and time base on 50 ns negative trigger.

20

Connect channell to S22H and adjust lower potentiometer on module B41 0
in location S22 until pulse width on oscilloscope is 50 ns measured at the
-1.5V points.

21

Connect channell to S32L (RDS DATA IN L), connect channel 2 to T27N and
. adjust module B312 in location T26 until delay between channell and channel
2 is 250 ns.

22

Connect S22V to ground.

23

Connect channell to S22H. Set oscilloscope time base to 200 ns/div. Adjust
upper potentiometer on module Type B410 in location S22 until pulses occur
at 400 ns intervals. Remove S22V from ground.

24

Connect channell to S28P and adjust B312 at location T26 until the square-wave
seen at S28P is 400 ns to -3V and 400 ns to ground.

25

Set time base on oscilloscope to 50 ns/div. Connect channell to M30S and
adjust calibrate control on the oscilloscope time base until the pulses are exactly
10 divisions apart on the oscilloscope face measured at the -1.5V points.
NOTE
The time base is now calibrated on this range to be
40 ns/div. Do not move the calibrate control or the
range (time/div.) control for the remainder of the
RPOI procedure unless instructed to do so.
5-8

Step

Procedure

26

Connect channell to T24D and channel 2 to T24N. Set time base for
100 ns/div. and adjust the upper potentiometer on module B410 in location
S22 until the waveforms seen on channell and channel 2 are exactly the same.

27

Connect channell to S32M (RDS CLOCK WIND (0) L) set for negative
trigger. Connect channel 2 to S32L and adjust B312 module in location
T26 until the negative-going edges of channel I and channel 2 are 192 ns
apart. Measure at the -1.5V point (see Figure 5-2).
NOTE
This completes the calibration of the RPOI system.

28

Remove the following jumpers:

a. E32H to ground
b. CIOF to ground
c. L22N to ground
d. L24N to ground
e. S15J to ground
f, K24U to S25J

29

Replace module Type B152 in location 129.

The following adjustments are performed in the preceding calibration procedure:
Step 5 adjusts the output pulse from B410 to a width of 50 ns.
Step 6 adjusts the delay line in T29 so that incoming pulses, which are 70 ns ± 15 ns, are standardiied
to a 50 ns pulse width before being fed to the remainder of the RDS circuitry.
Step 7 adjusts the variable input resistor of the B41 0 clock to a setting at which a phase lock condition
can be obtained in the phase lock loop. This phase lock loop consists of the B410 voltage controlled
oscillator, phase error comparator flip-flops (called RD RDS ERR A and RDS ERR B), and a differential
error amplifier and filter GS89 or G590.
Step 10 adjusts the delay which is used to set the data and clock pulses correctly with respect to their
gating waveforms. The gating waveforms are generated by the flip-flops S28 and T32, and ensure that
all data bits are gated out of S32H (RDS DATA OUT H), and all clock bits are gated out of S32N.
Step 14 adjusts the width of the gating waveform which is used to gate the data bits and clock bits.
Step IS finally adjusts the setting of the variable input resistor in B41 0 so that the phase lock loop
achieves its lock with a phase error of zero degrees between the internal oscillator (B41O) and the
incoming serial stream of data and clock bits.
Step 20 adjusts the width of the output pulse from B410 to 50 ns.
Step 22 adjusts the variable input resistor of B410 to a setting which allows a phase lock condition to
be established between the B4l 0 internal oscillator and the incoming clock and data bits.

5-9

Step 26 provides a final adjustment to the variable input resistor of B410 to establish the optimum
phase lock condition between the incoming serial stream of data and clock bits and the B 140 internal
oscillator.
Step 27 adjusts the delay used to set the data bits and clock bits correctly to their respective gating
waveforms.
5.3.4 General Troubleshooting Guides
The following paragraphs contain suggested methods of correcting recurring problems. Also, refer to DEC Field
Service Manuals and DEC Tech Tips published periodically.
5.3.4.1 Disk Pack Compatibility - To determine if two disk packs are compatible, run CTEST and swap packs
as directed. If any errors occur, the drives are considered to be incompatible.
NOTE
Before running compatibility tests, the drive should have
passed all diagnostics with a pack formatted on that drive.
If the drives are found to be incompatible, they should be aligned as described in the Vendor Maintenance Manual.
CAUTION
C.E. (Customer Engineering) Packs are not indestructable.
Their specially recorded track can be destroyed rendering
the pack unusable. Extreme care should be exercised in
their use. C.E. Packs should only be used when aligning
heads or checking head alignment. They should never be
used for any other troubleshooting or alignment procedures.
5.3.4.2 Proper Use of C.E. Packs - The following procedures should protect and may extend indefinitely the
usable life of C.E. Packs. For both RPO I and RP02:
Procedure

Step

Do not connect the tester until all power has been removed from the drive.
CAUTION
Before cables are removed, all power must fast be removed
from ail drives (via SI) as well as from the RPlO. Failure to
do so can result in damage to the steering diodes in the power
sequencing network.
2

Set the drive to READ ONLY condition.
CAUTION
Always disable write amp by switching to READ ONLY before
using C.E. Packs to insure safety of specially recorded elliptical
tracks.

3

Apply power to drive the S I, then apply power to tester.

4

Press UNIT DESELECT on tester.
NOTE
This step is necessary to prevent accidental writing over the
test track, which would effectively destroy its usefulness.

5-10

For RPOI only:
Step

Procedure
Remove HEAD SELECT (switch 4 on tester).

2

Remove all head plugs.

3

Insert head alignment plug into desired head socket.

4

Mount the C.E. Pack on the drive.
CAUTION
Whenever a different head is to be tested, or a head plug removed or inserted, always remove HEAD SELECf level
(switch 4 on the tester). It is also recommended to position
off cylinder 73 when repositioning the Head Alignment
Plug.

For RP02 only:
Step

Procedure
Note that the Head Alignment Plug is not used with the RP02.

2

It is not necessary to remove a head plug while the C.E. Pack is on the
drive.

Head and Pack Thermal Equilibrium:
Procedure

Step

Before heads are adjusted, thermal equilibrium procedures should be
followed. The temperature stabilization cycle, which assures standard
operating temperature during head alignment, consists of fIrst running
the drive with C.E. Pack installed and all covers on for one hour and
15 minutes, then running the drive with C.E. Pack installed and the
two top covers off for an additional 20 minutes before attempting
alignment.
2

Use only the special tools for maintenance (Le., beryllium screwdriver,
torque wrench, etc).
NOTE
All RPO 1 and RP02 Vendor Manuals are serialized and contain specifIc information pertaining to that drive (i.e., ECO's,
updated prints, etc). Each manual must be maintained as
Engineering Documentation.

When a C.E. pack is used on a Memorex drive, the READY signal will not occur. Therefore, jump B05 test point
three to ground. This indicates to the drive logic than an index pack is being used, instead of a sectored pack.
5.3.4.3 Search Errors - Search Errors can be caused by a variety of conditions. A known formatted pack should
be run. If search errors occur on only specifIc surfaces of a drive, the read logic should be checked. If search
errors occur from all drives, the RDS operation should be checked. To do this, sync off 109M (DTC SEARCH)
going high and view L21 N (DTC READ BIT). This bit should be on a 1 for approximately 100 IlS for RP02s and
for 250 IlS for RPO 1s. If DTC READ BIT is not on a 0 after 100 IlS (250 IlS for RPO I), or if it never transitions
to a I, the separator is at fault. However, if READ BIT performs correctly the Header Compare, LPR, Header
Read Control, or Shift Register is at fault.
5-11

5.3.4.4 Parity Errors - If parity errors occur from one drive, the read or write logic in that drive should be
checked. If all parity errors occur from word 0, the RDS Sync Field Enable is probably not being produced. If
errors do not exhibit a repetitive pattern, the RDS is at fault. If an AR bit is faulty, this is particularly easy to
see and scope in local mode. Remember that the controller tries to read headers three times before terminating;
therefore, the separator's operation can be quite marginal before search errors occur.
5.3.4.5 Failure to Restore - If the Memorex drive fails to restore, the logic can be checked by moving the
actuator by hand. Proceed as follows:
Step

Procedure
Connect the off-line tester.

2

Remove power from bobbin.

3

Install a jumper to prevent T2 from becoming true (Seek Incomplete).

4

Wiggle head assembly at full forward stop. This gives a forward velocity
signal that is needed to complete the restore sequence. The code 202
should now be seen in the cylinder address lights (force count).

5

Move the coil backwards by hand. Observe that the address register counts
down.

6

When address reaches 0, the detent should pick.
NOTE
The address might not reach 0 because the pull may not be
smooth causing a tooth to be counted more than once.

5.3.4.6 Disk Pack Reliability Test Map Limitations - The MAP routine in MAINDEC 10-D5OB cannot guarantee detection of all weak or bad spots on the disk pack surface. The best procedure for developing an initial list
of questionable or faulty disk pack sectors is to note all failures relating to data errors during acceptance and to
enter the MAP data manually using the manual entry feature of D50B's MAP routine. The MAP routine will
write the map data onto the appropriate sectors of the pack for Monitor reference.
5.3.4.7 Oeaning - For trouble-free operation take the same environmental care as is indicated for tapes (refer
to Paragraph 6.10 in the PDP-l 0 System Reference Manual). In particular, the heads and the packs should be
kept clean.
Materials and equipment needed for cleaning are as follows.

a.

Lint-free wipers (cloth or paper), such as Kimwipes Type 900-S, Stock No. 3415 (about 8 x 5 in.)

b.

Isopropyl alcohol, at least 90%, such as Merck or NF (99% by weight) or Lilly (91% by volume)
NOTE
Store alcohol in its original container or in a glass jar.

c.

Q-tips and pipe cleaners

d.

Wooden tongue depressors, 6 x % in.
NOTE
Do not use plastic depressors (the alcohol softens them).

e.

A high-intensity light or other strong light source

t;

A piece of white cardboard or stiff paper (8Y2 x 11 in.).
5-12

Inspect the heads for dirt accumulation at least twice monthly (weekly for around-the-clock operation). When
necessary, clean them carefuIly with a Q-tip soaked in alcohol, and clean out the two holes in each head with a
pipe cleaner, also soaked in alcohol. Keep all cabinet and pack filters clean and fresh. Dirty heads can be caused
by poor head flight due to poor air flow through the pack or cabinet. Always replace a single head that collects
dirt while others in the same drive remain clean.
Do not clean all of the packs at an installation just for the sake of cleaning them. If a pack is subject to random
errors, or if vital information that is not duplicated elsewhere cannot be retrieved, and the problem is not alleviated
by cleaning the heads, then clean the pack. In any event, depending on environmental conditions, test all packs
every few months by cleaning a couple of surfaces at random in each; clean all surfaces in any pack in which dirt
is discovered.
To clean a pack, mount it on a drive from which the upper case panels have been removed or on a free-standing
spindle mechanism that allows the pack to be turned by hand with the cover removed. Place the light with the
white cardboard as a background so that plenty of light shines into the pack. Follow this procedure for cleaning
each surface.
Step

Procedure
Position the light so the surface is clearly visible.

2

Wrap a fresh wiper around a depressor with wiper extending 3/8 in. beyond
the wood at one end.

3

Soak one side of the wrapped depressor with the alcohol (be sure to wet the
full width of the depressor).

4

Spin the pack by hand at 40-60 rpm (use the flat top surface of the plastic
bezel on the top of the pack).

5

With the pack spinning, insert the prepared depressor with the protected end
toward the center, and press the wet side against the surface. Maintain the
pack spin while applying about 5- 10 pounds pressure against the surface; wet
the surface across the full width of the tracks. The pressure may be lightened
as the surface dries, but be sure to keep the wiper on the surface with the
pack spinning until the surface is completely dry and has a high gloss. Keep the
pack spinning while removing the depressor, and check the wiper for dirt.

6

Inspect the surface carefuIly for scratches. If a scratch corresponds in position
to a bad sector as determined by the program, then further cleaning is unlikely to make the sector usable.
WARNING
Do not attempt to use the drive motor when cleaning a pack.
Always turn drive power off and spin the pack manually.
Do not clean a pack if either the pack or the alcohol is below
40°F (otherwise water vapor might condense on the surface).

5-13

Chapter 6
Drawings

6.1 GENERAL
The RPIO engineering drawings and an accompanying signal glossary are included in Volume 2 of this maintenance
manual. A list of engineering drawings is provided in Table 6-1.
Table 6-1
RPI0 Engineering Drawings
Number

Title

A-ML-RPlO-A

RPIO Master Drawing List

D-DI-RPIO-O-l

RPIO Drawing Index List

D-UA-RPIO-O-O

RPIO Unit Assembly

A-PL-RP 10-0-0
A-PL-RPIO-O-MC

RPIO Parts List
Parts List Disk File Interface

A-PL-7006194-0-0

Wired Assembly RPIO

A-PL-7006195-0-0

Cable Set RPIO
Flow Chart Asynchronous and Sector Counters

D-FD-RPIO-O-FCA
D-FD-RPIO-O-FCCC
D-FD-RPIO-O-FCCI
D-FD-RPIO-O-FCDC
D-FD-RPIO-O-FCE
D-FD-RPIO-O-FCL
D-FD-RPIO-O-FCMI
D-FD-RPlO-0-FCM2
D-FD-RPIO-O-FCRI
D-FD-RPIO-O-FCRW
D-FD-RPiO-O-FCS
D-FD-RPIO-O-FCSC
D-FD-RPIO-O-FCWH

Flow Chart Channel Control
Flow Chart Channel Interface
Flow Chart Disk Control
Flow Chart End
Flow Chart Local
Data Transfer Control Micro I
Data Transfer Control Micro 2
Flow Chart Register Interconnections
Flow Chart Read/Write
Flow Chart Start
Flow Chart Search
Flow Chart WR Headers and Data

D-BS-RPIO-O-ARD

Assembly Register
Assembly Register Data Gate

D-BS-RP 1O-O-CC

Channel Control

D-BS-RPIO-O-CXR

Condition Register

D-BS-RPIO-O-DAR

Da ta Address Register

D-BS-RPIO-O-DSBC

Disk Signal Bus Connectors

D-BS-RPI O-O-DTC

Data Transfer Control

D-BS-RP I O-O-AR

6-1

Table 6-1 (Cont)
RPI0 Engineering Drawings
Number

Title

D-BS-RPIO-O-HCDE

Header Compare and Designation Error

D-BS-RPIO-O-IBC

I/O Bus Control
Interconnecting Cables

D-IC-RPIO-O-IC
D-IC-RPIO-O-INDC
D-BS-RPIO-O-IOB
D-BS-RPIO-O-IOBD

Indicators
lOB Transmitter

D-BS-RPIO-O-LPR

lOB Data Receiver
Longitudinal Parity Register

D-BS-RPIO-O-MPX

Multiplexer

D-BS-RPIO-O-RDS

Read Data Separator

D-BS-RPIO-O-RDS

Read Data Separator Calibrations

D-BS-RPIO-O-SC

Sector Counter

D-BS-RPIO-O-SCB

Sector Counter Buffer

D-BS-RPIO-O-SCCC

Sector Counter Control Pulses

D-BS-RPIO-O-SCM

Sector Counter Multiplexer

D-BS-RPIO-O-SR

Shift Register

D-IC-RPIO-O-SWP

Switch Panel

D-CL-RPIO-O-TERM

Pulse and Level Terminations

D-BS-RPIO-O-WDC

Word Counters

D-MU-RPIO-O-AD

Module Utilization A-D

D-MU-RPIO-O-EJ

Module Utilization E-J

D-MU-RPIO-O-KN

Module Utilization K-N

D-MU-RPIO-O-PT

Module Utilization P-T

D-IC-RPIO-O-3
C-AD-7006 I 95-0-0

Wiring Power DC and AC
Wired Assembly (RPIO)
Cable Set (RPIO)

D-SP-RPIO-O-CEPT

CE Pack Tracks

D-AD-7006194-0-0

6-2

READER'S COMMENTS

RPIO DISK PACK SYNCHRONIZER
MAINTENANCE MANUAL
DEC-IO-HSEC-D (1)

Your comments and suggestions will help us in our continuous effort to improve the quality and usefulness of
our publications.
What is your general reaction to this manual? In your judgement is it complete. accurate. well organized. well
written. etc~ Is it easy to usery _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __

What features are most useful?

What faults do you find with the manual?

Does this manual satisfy the need you think it was intended to satisfy?
Does it satisfy your needs?
Why? _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __

Would you please indicate any factual errors you have found. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __

Please describe your position.
Name _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Organization _ _ _ _ _ _ _ _ _ _ _ _ _ __
Street _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Department _ _ _ _ _ _ _ _ _ _ _ _ _ __
City _ _ _ _ _ _ _ _ _ _ _ _ State _ _ _ _ _ _ _ _ _ _ _ _ __

Zip or Country _ _ __

- - - - - - -

--

-

- - -

- - - - - - - - -

-

- - - -

Fold Here - - - - - - - - - - - - - - - - - - - -

- - Do Not Tear - Fold Here and Staple -

- -

-

- -

-

-

-

-

- - -

FIRST CLASS
PERMIT NO 33
MAYNARD. MASS.
BUSINESS REPLY MAIL
NO POSTAGE STAMP NECESSARY IF MAILED IN THE UNITED STAHS

Postage will be paid by:

mamaama
Digital Equipment Corporation
Technical Documentation Department
146 Main Street (A & M Building)
Maynard, Massachusetts 01754

-

Digital Equipment Corporation
Maynard, Massachusetts

printed in U.S.A.



Source Exif Data:
File Type                       : PDF
File Type Extension             : pdf
MIME Type                       : application/pdf
PDF Version                     : 1.4
Linearized                      : Yes
XMP Toolkit                     : Adobe XMP Core 4.2.1-c041 52.342996, 2008/05/07-21:37:19
Producer                        : Adobe Acrobat 9.0 Paper Capture Plug-in
Modify Date                     : 2017:08:04 02:12:34-07:00
Create Date                     : 2002:01:16 15:50:11+10:00
Metadata Date                   : 2017:08:04 02:12:34-07:00
Creator Tool                    : Acrobat 5.0 Image Conversion Plug-in for Windows
Format                          : application/pdf
Document ID                     : uuid:22f19489-ba66-0346-95a0-8d0b8be9acf5
Instance ID                     : uuid:97839c1c-6979-2647-898a-d823a53ca451
Page Layout                     : SinglePage
Page Count                      : 104
Creator                         : Acrobat 5.0 Image Conversion Plug-in for Windows
EXIF Metadata provided by EXIF.tools

Navigation menu