G_1616 G 1616

User Manual: G_1616

Open the PDF directly: View PDF PDF.
Page Count: 84

DownloadG_1616 G 1616
Open PDF In BrowserView PDF
GenRad

GR 161 6 Precision
Capacitance Bridge

GR 1621 CapacitanceMeasurement System
Form 1616-0100-00

InstructionManual

WARNING
Dangerous voltages may be present at the terminals of this instrument. Observe all warnings
contained in this manual. Refer all servicing to
qualified personnel.

G R 161 6 Precision
Capacitance Bridge

G R 1621 CapacitanceMeasurement System
Form 1616-0100-00

@GenRad, Inc., 1988
Concord, Massachusetts, U.S.A. 01742
January, 1988

Tableof Contents

Page

Section

CONDENSED OPERATING INSTRUCTIONS

ix

3.4

SPECIFICATIONS . . . . . . . . . . . . . . . . . . . . . ..

x

Section

1 INTRODUCTION
1.1
Purpose
1.2
Description
1.2. 1 General.................
1.2.2 Bridge Circuit
1.2.3 Standards
1.2.4 Oscillator
1.2.5 Detector
1.3
Controls, Indicators, and
Connections . . . . . . . . . . . . . . . . . . .
1.4 Accessories. . . . . . . . . . . . . . . . . . .

1-1
1-1
1-1
1-1
1-2
1-2
1-2

3.5

. 1-3
. 1-3

3.6

2.5

2.6
2.7
2.8

General
2-2
Dimensions
2-3
Environment
2-2
Bench Models
2-2
2.4.1 Cabinet Removal
2-2
2.4.2 Conversion For Rack
Mounting
2-2
Rack Models
2-2
2.5.1 General
2-3
2.5.2 Installation
2-3
2.5.3 Conversion To Bench Use .. 2-3
Power-Line Connection
2-4
Line-Voltage Regulation
2-4
System Connections
2-4
2.8.1 Oscillator. Bridge, and
Detector
2-4
2.8.2 BCD-Capac itanee-Output
Connector
2-4
2.8.3 BCD-Conductance-Output
Connector
2-5
2.8.4 Analog Outputs
2-5

3.7

3.8
3.9

Preliminary Checks
Functional Self-Checks
Phase Adjustment

Balance and Readout
3.5.1
Readout Multiplier
3.5.2
Initial Settings of Lever
Switches

3-3
3-3
3-3
3-4
3-4
. 3-5

3.5.3
Balance Procedures
3.5.4
Final Balance
3.5.5
Readout Correction
3.5.6
Units of Measurement
Parameters of The Unknown
Capacitor . . . . . . . . . . . . . . . . . . . . ..
Series Equivalent
Parameters . . . . . . . . . . . . . .
3.6.2
Dissipation Factor
Frequency
3.7.1
Setting The Frequency
3.7.2
Monitoring Frequency
3.7.3
Locking To a Frequency
Reference . . . . . . . . . . . . . ..

3-5
3-6
3-7
3-7
3-7
3-7
3-8
3-8
3-8
3-8
3-1 0

Voltage Level
3-10
Accuracy
3-10
3.9.1 Accuracy Versus Frequency
and Cx
3-10
3.9.2

Accuracy Versus
Temperature

3-11

3.9.3

Range and Dissipation Factor
Limitations
3-12

3.9.4

Shunt Capacitance To
Ground

3-1 2

3.9.5
Fringing Capacitance
3-13
3.9.6 Conductance Accuracy
3-14
3. 10 Precision
3-15
3.11 External Standards
3-15
3.11.1 Range Extension To 111
Microfarads . . . . . . . . . . . . .. 3-15
3.11 .2 Extension of C Resolution .. 3-16
3.11.3 Externally Determined Accuracy/
Comparisons
3-16

3 OPERATION

3.1
3.2
3.3

Connection of Unknown Capacitor
3.4.1 Three-Terminal Capacitors
3.4.2 Two-Terminal Coaxial
Capacitors

3.6.1

2 INSTALLATION

2.1
2.2
2.3
2.4

Page

3-1
3-1
3-2

iii

Tableof Contents(Cont'd)
Page

Section

3.12

3.13

3.14

3.15

Section

3.11.4 Test-Fixture Compensation. 3-17
3.11. 5 Range Extensions To 11
Microsiemans
3-1 7
Precise Comparisons
3-17
3.12.1 Balance Comparisons
3-17
3.12.2 Direct Substitution
3-17
Non-coaxial 2- Terminal Capacitors . 3-17
3.13.1 Unshielded 2- Terminal
Capacitors
3-18
3.13.2 Shielded 2- Terminal
Capacitors
3-18
Dc Bias
3-20
3.14.1 Normal Bridge Configuration/
Parallel Bias
3-21
3.14.2 Reversed Bridge Configuration/
Series Bias
3-21
3.14.3 Dc In The Ratio Transformer/
Demagnetization
3-22
Reversed Configuration
3-22
3.15.1 Explanation
3-22
3.15.2 Procedure
3-22

4 THEORY
4.1
4.2

4.3

4.4

4.5

Introduction
4-1
Properties of Capacitors
4-1
4.2.1
Basic Components of
Capacitance
4-1
4.2.2
Inductive and Lossy
Components
4-2
4.2.3
Frequency Characteristics .. 4-4
Basic Bridge Circuitry
4-5
4.3.1
Elementary Capacitance
Bridges
4-5
4.3.2
Transformer-Ratio Bridges .4-5
Circuitry of The 1616 Bridge
4-7
4.4.1
Excitation
4-7
4.4.2
Circuit For The Unknown
4-7
4.4.3
Capacitance Standards
4-7
4.4.4
Conductance Standards
4-8
4.4.5
External Standards
4-9
4.4.6
Zero Adjust
4-9
C-Standards Accuracy
4-9
4.5.1
Calibration
4-9
4.5.2
Sealing
4-10
4.5.3
Thermal Lag
4-10

iv

4.6
4.7

Page

G-Standards Accuracy
Ratio Accuracy
4.7.1
Residual Impedances . . . .
4.7.2
Example With 1: 1 Ratio . .
4.7.3
The 10:1 Ratio
4.7.4
The 1:100 Ratio
4.7.5
C Offset Due To Induction
4.7.6
Ground Circuit Impedance

..
..

.
.

4-9
4-11
4-11
4-1 2
4-13
4-13
4-14
4-14

5 SERVICE and DIAGRAMS
5. 1
5.2

GR Field Service
Minimum Performance Standards ..
5.2.1
General
5.2.2
Zero Setting, Offsets, and
Sensitivity

5.2.3
Capacitance Accuracy
5.2.4
Capacitance Ratios
5.2.5
Conductance Accuracy
5.2.6
Conductance Multipliers
5.3
Dissassembly
5.3.1
Knobs
5.3.2
Cabinet Removal
5.3.3
C-Box Removal
5.3.4
The G Box
5.4 . Recalibration and Adjustments
5.4.1
Internal Capacitance
Standards

5.5

5.6

'"
"

5.4.2
5.4.3
5.4.4
5.4.5

Conductance
C301/Setting
Lever-Switch
Maintenance
Switches

Multipliers
Zero C
Stiffness
Note On

Trouble
5.5.1
5.5.2
5.5.3

Analysis
Mechanical Damage
BCD Circuits
Non-Repairable
Subassemblies . . . . . . . . . ..

5.5.4
Typical Parameters
Parts Lists and Diagrams

5-1
5-1
5-1
5-3
5-3
5-3
5-3
5-4
5-5
5-5
5-5
5-5
5-7
5-7
5-8
5-8
5-8
5-8
5-8
5-9
5-9
5-9
5-9
5-9
5-9

APPENDIX A
AN INTERPRETATION of MILLEA'S
METHOD TO ELIMINATE FRINGING FROM TWOTERMINAL MEASUREMENT
A-1

! '

Listof Illustrations

Title

Figure

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

1621 Precision Capacitance-Measurement System
1616 Precision Capacitance Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Elementary Diagram, 1621 Precision Capacitance-Measurement System-GR1616
Bridge with 1316 Oscillator and 1238 Detector
Front-Panel Controls, Indicators, and Connectors of the 1621 Precision CapacitanceMeasurement System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Rear-Panel Controls, Connectors, and Cables

ii
x

Dimensions of the Bench Models
Dimensions of the Rack Models
Bench-Cabinet Installation
Rack-Cabinet Installation
Terminal Identification at BCD CAPACITANCE OUTPUT Connector (A-J13, Rear Panel
of Bridge)
Terminal Identification at BCD CONDUCTANCE OUTPUT Connector (A-J 12, Rear Panel
of Bridge)
DC Meter Outputs Socket on 1238 Detector, Exterior (Rear) View

2-1
2-1
2-2
2-3

Simplified Bridge Diagram 3- Terminal Connection of Unknown Capacitor
Simplified Bridge Diagram 2- Terminal Connection of Unknown Capacitor ....
The 1621 Capacitance-Measurement System Simplified Diagram .....
Equivalent Circuits of the Unknown Capacitor ...
Dissipation Factor Versus Directly Measured Parameters
Specified Capacitance Accuracy of the 1616 Bridge
Worst-Case Measurement Error Due to Temperature (Based on Maximum Temp Coef
of C Stds)
Non-Coaxial 2- Terminal Capacitors and Some Measurement Configurations ...
Circuit For Applying Bias To The Unknown Capacitor-Normal Configuration ..
Reversed Configuration - Bridge Circuit With Oscillator and Detector Interchanged

3-3
3-4
3-5
3-8
3-9
3-11

0

2-7
3-1
3-2
3-3
3-4
3-5
3-6
3-7

4-1
4-2
4-3

•••

0

0

•••

0

0

••

0

0

0

0

0

0

•••••••••••••

0

0

•

0

0

••••

0

0

0

0

•••••••

0

•••••••••

0

0

•••

00

•••

00

••

0

0

0

0

•

0

0

0

•••••••

0

0

0

••

•

0

•••••

0

•

0

0

0

0

•••••

••••

0

0

0

0

••••

0

•

0

••

0

0

••

0

0

•

0

0

0

•••••••••••••

0

0

0

•

0

•••

0

0

0

•

0

0

•••

0

0

0

0

0

••••••••••••••

0

0

•••••••••

0

•••••

0

0

0

0

v

•

0

•

0

••

0

0

•

••••••••••••••••••••

0

••••••••

0

0

••

0

0

•

0

0

0

•••••••••

••••

•

•

•

•

•

•

•

0

0

•

•

•

••

0

••••

•

••

•••

0

0

••

0

0

•••••

0

•••••

•

•

•

•

•

•

0

0

•

•

•

•

0

•••

••••••

0

••••

••••••••

0

0

••

0

•••

0

0

3-12
3-19
3-21
3-22

•••••••

•••••••

••••••••••

0

0

0

0

••••••••

0

0

•••••

•••••••

0

0

0

0

Schematic Diagram of a Capacitor ..
Structure of a 3- Terminal Capacitor With 2 Coaxial Connectors
A Coaxial 2- Terminal Capacitor, It's Structure, Component Capacitance, and
Connection To a Bridge
Capacitor Lumped-Parameter Equivalent Circuits and Vector Diagrams
Capacitance Versus Frequency Characteristic of a Mica Capacitor
Dissipation Factor Versus Frequency Characteristics of a Mica Capacitor
An Elementary Capacitance Measuring Bridge
An Elementary Capacitance Bridge With Transformer Ratio Arms
Circuitry For 3 Methods of Balancing a Transformer-Ratio Capacitance Bridge ....
Simplified Bridge Schematic Diagram
0

•

0

••••

0

0

••••

0

••

0

00

1-3
1-7

2-5
2-6

•••••

0

•••••

0

•

0

••••

0

0

•

•••••••••••••••

0

0

0

1-2

2-5

••••••••••••

•••••••••

0

o.

4-4
4-5
4-6
4-7
4-8
4-9
4-10

0

0

00000.0.0

3-8, 3-9
3-10
3-11

Page

0

••

•

0

0

•

0

0

0

•

0

•

•

•

•

•

4-1
4-1
4-2
4-3
4-4
4-5
4-5
4-6
4-6
4-11

Listof Illustrations
(Cant'd)

Figure

Page

Title

5-1
5-2
5-3
5-4
5-5
5-6
5-7
5-8
5-9

Three-Terminal Conductance Standards
Interior View, Upper Rear, With the C-Box Removed
Interior Bottom View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Detent Mechanism of a Typical Lever Switch, A and B Constructions
Front View; Mechanical Replaceable Parts Identified
Rear View; Mechanical Replaceable Parts
Wiring Diagram of BCD Output Circuits
Conductance Multiplier Board Assembly 1616-4720 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Schematic Diagram of the 1616 Precision Capacitance Bridge . . . . . . . . . . . . . . . . . . . . . . .

5-5
5-6
5-7
5-9
5-11
5-11
5-11
5-12
5-13

A-1

Test Setup For 2- Terminal Measurement, With Fringing Eliminated

A-1

Listof Tables
Table

Page

Title

1-1
1-2
1-3

Front-Panel Controls, Indicators, and Connectors
1-4
Rear-Panel Controls and Connectors
1-6
Accessories Supplied With the 1621 System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 1-9

2-1
2-2
2-3
2-4

Rack-Mounting Cabinets and Brackets for 1621 System
Key to Screw Sizes
Binary-Coded Decimals (BCD)
Readout-Multiplier Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3-1
3-2
3-3

Senses of Phase-Sensitive Meters
Characteristics At Bridge Detector Output Connector
Bias Current Limits For Transformer Saturation

4-1
4-2

Approximate Magnitudes of Residual Parameters
Error Example, Ratio 1:1

5-1
5-2
5-3

Test Equipment
5-2
Capacitance Ratio Checks
5-2
Conductance Accuracy and Multiplier Checks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-4

vi

2-2
2-3
2-5
2-5

. . . . . . . . . . . . . 3-2
3-21
3-22
4-11
4-13

Condensed
Operating
Instructions
WARNING
Dangerous voltages may be present at the terminals of this
instrument.
To reduce the risk of electric shock, turn the
voltage source to "0"
before connecting or disconnecting
device under test.
CAUTION
Be sure the line-voltage switches on Oscillator
and Detector rear panels are properly set for
the available power.

b Set TERM. SELECTOR to upper X1 position unless

C; is larger than 1 n F (1000 pf ). then use X 10 If appropriate, raise C-MAX lever.
c. Set C and G levers to approximate values of C; and
G x (parallel components of "unknown").
d. With GAIN control keep MAGNITUDE indication on
scale Turn SENSITIVITY and TIME CONSTANT cw, if
required to achieve final balance (below) For best resolution, increase Eo RANGE to 150 V

TUNING AND PHASE ADJUSTMENT.
a. Set the controls as follows.
POWER : ON (up) oscillator and detector
FREQUENCY SELECTORS: as desired or 1.01 kHz
Oscillator Eo RANGE: 15V
Eo ADJUST: MAX
Detector TIME CONSTANT: 0.1 s
PUSH BUTTONS: both out
IN-PHASE FINE ADJUST: midrange
QUAD FINE ADJUST: midrange
PHASE SHIFT (inner knob): 180 deg
SENSITIVITY full on (cw)
Bridge TERM SELECTOR: CAL
EXT MULT: OFF

CAUTION
Do not exceed either 350 V rms or 0.16 f volts
(example: 16 V at 100 Hz) in normal configuration.
e. Refine the balance, left-to-right with C levers, bringing
IN PHASE meter toward zero until the G error predominates. Then continue with G levers, bringing QUAD meter
toward zero until the C error predominates. Repeat step e
until balance reaches'the resolution you need.

6

G levers X G Mull: 0/15 000.0 nS X 10C levers: 05 pF 000 fF 000 0 aF
(First three digits masked)
GAIN: 30db (or keep MAGNITUDE on scale)

2-TERMINAL MEASUREMENT.

b. Fine tune oscillator to detector frequency (peak on
MAGNITUDE meter). (Note 3rd osc dial is continuous).
c. Reset C levers to 00 pF 000 fF 500
0 aF
Reset GAIN to 100db
Read just C levers for MAGNITUDE of 20-40
d. Using PHASE SHIFT (large knob) bring QUADRATURE
meter to zero.
Reset C levers for MAGNITUDE of 80-100, reset zero
using PHASE SHIFT first and then QUAD FINE ADJUST.
NOTE
Keep M/\GNITUDE meter on scale for correct
phase indication; other meters may be off scale.
e. Reset C levers to all zeros.
Adjust G levers for MAGNITUDE of 80-100.
Using only IN-PHASE FINE ADJ, set IN PHASE meter to
zero.

3-TERMINAL MEASUREMENT.
a. Connect unknown capacitor between inner conductors (HIGH, LOW) of 3-term C, port, shield to outer
conductor of either or both connectors. Cables are optional; shield at least the LOW one. (Cable capacitance is
excluded from the measurement \

vii

a. Set TERMINAL SELECTOR to right X1 position
unless C, is larger than 1 nF, then use X 100. Raise C-MAX
lever, if appropriate.
b. Set readout to value of fringing capacitance of 2-terminal port, i.e., ** nF *00 pF 115 fF 000.0 aF, 0 j-tS 000.0
nS X 10-6 , if you selected Xl in step a. If XlO or Xl00,
set C proportionally smaller (11.5 fF or 1.15 fF, respectively).
c. Balance the bridge with ZERO ADJUST.
d. Connect unknown capacitor to the 2-TERM C; port
outer shell - ungrounded
to outer conductor of connector, inner (shielded) terminal to inner conductor.
e. Proceed as before - steps, c. d. e of the 3-terminal
instructions.

FURTHER INFORMA TION.
Refer to the Table of Contents.

Specifications
1616

PRECISION

CAPACITANCE

Capacitance
measurement,
3-terminal:
DECADES:
12.
RANGE 01 aF to 1 IlF 00
'to 10 6 F). ACCURACY:' =10
ppm, when most-significant
decade is 1, 10, or 100 pF per
step; otherwise, and at other frequencies,
accuracy is =[50
ppm
(05 + 20 CIl , ) (f.,,)' ppm + (f H . ) aF].
Capacitance,
2-terminal:
Same as above, except as follows.
RANGE: One additional decade, to 10 IlF (10 'to 10 ' F).
Conductance measurement,
3-terminal: DECADES: 5 (virtually
extended to 11 by G multiplier).
RANGE 100 as to 100 p.s
(10 'to 10'u)
ACCURACY'
(01%
1 step in least significant decade).
There is a small reduction in conductance
accuracy at frequencies
other than 1 kHz.
RESIDUAL C
« 0.03 pF).
(across conductance standards):
Conductance,
2-terminal:
Same as above, except as follows:
RANGE One additional decade, to 1000 p-S (10 'to 10 S).

xi.

and

DETECTOR

Connect

to

rear

BNC

con-

Required: OSCILLATOR: GR 1316 recommended.
DETECTOR:
GR 1238 recommended.
The 1616 Bridge is available with
this oscillator and detector as the 1621 Capacitance-Measuring Assembly.
Available:
1316 OSCILLATOR, 1238 DETECTOR and a broad line
of capacitance standards.
Mechanical:
Bench or rack model.
DIMENSIONS (wxhxd).
Bench, 19.75x1381x12.88
in. (502x351x327
rnrn), rack, 19x
1222xlO.56
in. (483x310x268
mrn).
WEIGHT: Bench, 57 Ib
(26 kg) net, 69 Ib (32 kg) shipping; rack, 49 Ib (23 kg) net,
61 Ib (28 kg) shipping.

xi.

Multipliers:
FOR 3-TERM:
X10; FOR 2-TERM:
X10,
X100; affect both C and G. FOR CONDUCTANCE ONLY: Xl,
Effects of these multipliers
X10 , ...
XlO 6 (7 positions).
are included in the specified ranges.
Frequency:

OSCILLATOR
nectors.

BRIDGE

for these conditions:
23()

±

1

C: humidity,

<50%

Frequ81lc/

RH

1 kHZ, except as noted: temperature

See manual for detailed

accuracy analvsis.

t Registered trademark of the Carpenter Steel Co
G900 and G874 - Gilbert Engineering. Glendale. Arizona 85301

10 Hz to 100 kHz.

Standards:
CAPACITANCE
Air dielectric
with TC < +40
ppm/C
and D 10 ppm for 7 lowest decades; lnvar t , air dietectric with TC of 3 1 ppm/C
and D <10 ppm for 3 middle decades; mica dielectric with TC of 20 =10 ppm! 'C and
D <200 ppm for 2 highest decades.
ADJUSTMENTS for all
capacitance
standards available through key-locked door on
panel.
THERMAL LAG: C standards
for first 8 decades
mounted in an insulated compartment
with a thermal time
constant of 6 h (time required for compartment
interior to
reach 63% of ambient change).
CONDUCTANCE: Metal-film
resistors in T networks With small phase angles.
Comparison:
Terminals provided to connect external standard
for comparison measurements;
13-position panel switch multiplies standard by-0.1,
0 .
11.
input: The smaller of 160 f.". or 350 V rrns can be applied to
bridge transformer
at the GENERATOR terminal without
waveform distortion;
500 V rrns max, depending on conductance range, when GENERATOR and DETECTOR connections
interchanged.
interface:
G900 locking coaxial connector on panel to connect 2coaxra I connectors on
unknowns. 'G8 74 locking
panel to connect 3-terminal unknowns and 2 to connect external
ccta,ndcird.DATA OUTPUT: 50- pin and 36- pin type 57 connectors on
rear provide connection to 8-4- 2 -1 weighted BCD contacts (reated at
28 V. 1 AI on each SWitch for capacitance and conductance values

CAPACITANCE

Catalog
Number

Description

1616 Precision

Capacitance

1621 PRECISION
- MEASUREMENT

I nternal Temperature: C standards in bridge, about 1°C above
ambient, for ultimate accuracy. allow 24 hrs to stabilize with ac
power on.
Frequency: 10 Hz to 100 kHz.
Supplied: 1616 Precision Capacitance Bridge, 1316 Oscillator,
1238 Detector,
all necessary
interconnection
cables, and
power cord.
Available: 1404 REFERENCE STANDARD CAPACITORS (10 pF,
100 pF. and 1000 pF) for calibration.
Power: 100 to 125 and 200 to 250 V, 50 to 60 Hz, 51 W.

Bridge

Bench Model
Rack Mudel

1616-9700
1616-9701

SYSTEM

Mechanical:
Bench or rack models.
DIMENSIONS (wxhxd):
Bench, 19.75x24.25x15
in. (502x616x381
rnrn): rack, 19x
20.91x11.44
in. (483x531x291
mrn).
WEIGHT: Bench, 105
Ib (48 kg) net, 140 Ib (64 kg) shipping; rack, 90 Ib (41 kg)
net. 125 Ib (57 kg) shipping.
Description

1621 Precision Capacitance·Measurement

Bench Model, 60-Hz
Rack Model, 60-Hz
Bench Model, 50-Hz
Rack Model, 50-Hz

viii

Catalog Number
System
1621-9701
1621-9702
1621-9703
1621·9704

FREQUENCY
SELECTORS

PHASE
SHIFT

~

QUAD

1238

DETECTOR

"'-

\..:

GAIN

c,

MAGNITUDE

EXT STD

ZERO

1616

BRIDGE

G LEVERS

1621 Precision Capacitance-Measurement

k

System

GenRad
WARRANTY

We warrant that this product Is tree from defects In material and workmanship and,
when properly used. wm perform In accordance with GenRad's applicable published
specttlcations , If WIthin one tt j year after orlglnalshlpment It Is found not to meet
this standard, It will be repaired or at the option of GenRad. replaced at no charge
when returned to a GenRad service facUlty.
CHANGES iN THE PRODUCT NOT APPROVED BY GENRAD
THIS WARRANTY

SHALL VOID

GENRAD SHAll NOT BE liABLE FOR ANY INDIRECT ,SPECIAL, OR CONSEQUENTIAL DAMAGES. EVEN IF NOTICE HAS BEEN GIVEN OF THE POSSIB1UTY OF SUCH DAMAGES
THIS WARRANTY

IS IN LIEU OF ALL OTHER WARRANTIES,

EXPRESS OR

IMPLIED, INCLUDING, BUT NOT LIMITED TO ANY IMPLIED WARRANTY OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE,

SERVICE POLICY
Your local GenRad office or representative will assist you In all matters relating to
product rnamtenance. such as calibration, repair, replacement parts and service
contracts
GenRad policy IS to maintain product repair capability for a period of five 151years

after the unit goes out of production and to make this capability
available at the then prevailing schedule of charges.

ix

Introduction-Section
1

1.1
1.2
1.3
1.4

1-1
1-1
1-3
1-3

PURPOSE.
DESCRIPTION
CONTROLS, INDICATORS, AND CONNECTORS
ACCESSORIES.

1.1 PURPOSE.
The 1621 Capacitance-Measurement Svstern is designed
for the precise measurement of capacitors and capacitance
standards. In the standards laboratory, its high resolution
for capacitance and conductance make th is system well
suited for capacitance standards measurements. Its phaseindicating error meters facilitate rapid balancing. Convenient in-line readout maximizes accuracy of manual data
recording and BCD outputs are provided for automatic data
processing.
The 1621 system measures either 3-terminal or 2-terminal capacitors. The transformer-ratio-arm circu itry of the
bridge assures that 3-terminal measurements can be made
accurately, even in the presence of large capacitances to
ground. For instance, a ground capacitance of 1 pF
produces an error of only 0.03% in the measurement of
1000-pF capacitor. This feature makes the assembly very
useful for in situ measurements of ungrounded circuit
capacitances.
The 1616 Precision Capacitance Bridge, one of the 3
instruments in that system, may be obtained separately.
The bridge will perform as described herein, if used with an
oscilIator and detector equivalent to the G.R 1316 and
1238.

NOTE
This manual describes the 1621 system generally and provides its operating instructions. This
manual also describes in detail the 1616 bridge
only.

A wide range of capacitances can be measured, extending from the resolution limit of 0.1 aF (10- 7 pF) to a
maximum of 10 pf, with internal standards, or farther with
external standards. For 3-terminal unknown capacitors, a

pair of coaxial terminals is provided; for 2-terminal coaxial
"unknowns," a single precision connector facilitates exacting control of fringing effects.
Since an important use of this bridge is the comparison
of capacitance standards, another pair of coaxial terminals
is provided on the bridge to which a 3-terminal reference
standard can be connected and designated EXTERNAL
STANDARD. The other standard is then connected to the
selected UN KNOWN terminals, and the internal standards
are used to complete the balance. If the ratio between the
two standards is close to 0.1, 1., or 10, the accuracy of the
measurement is equal to the accuracy of calibration of the
reference standard, and the precision of comparison is 1
part in 108 (0.01 ppm) of a 10-pF capacitor (or even better
for larger ones).

1.2 DESCRIPTION.

Figure 1-1.

1.2.1 General.

The 1621 Precision Capacitance-Measurement System
consists of the 1616 Precision Capacitance Bridge with the
1316 Oscillator and the 1238 Detector, a complete system
for the precise measurement of capacitance.
Oscillator and detector are mounted above the bridge, in
a pedestal cabinet, as pictured in the front pages (and
Figure 1-2); or the three instruments may be rack mounted.
Connecting cables, supplied, go neatly behind the assembly.
An elementary system diagram is given in Figure 1-1.

1.2.2 Bridge Circuit.

The ratio arms of the bridge are transformer windings,
tapped on the standard side in deci mal steps (-1, 0, 1,
2 ... 9, X) and on the unknown side in decade steps
(X 100, Xl 0,
Separate, fixed-capacitance standards are
used, whose values range in decade steps from 1 aF to 100
nF. This combination of internal standards and transformer

xn

INTRODUCTION

1-1

ratios makes possible the wide measurement range of 1 to
10

14

Loss in the measured capacitor is expressed as parallel
conductance from the resolution Ii mit of 0.1 f S to a
maximum of 1 mS a measurement range of 1 to 10 1 3 The
values of the set of 5 conductance standards are effectively
extended by series resistance standards, in 6 decade steps
(X 1 ... X 10-6 )

READOUT
MULTIPLI ER

UNKNOWN
CAPACITOR

o~LOW

GR 1316
OSCILLATOR

CAPACITANCE

short-circuit

loading.

Auxiliary
outputs are provided, both in-phase and
quadrature, for detector references. These signals are
comparable in quality with the main output. The level of
each aux iIiary output signal is about 1.3 V rrns, driving the
minimum recommended load impedance of 47 kSl. The
phase separation between them is typically 87 to 90°,
except below 50 Hz it may be a few degrees less. (F INE
ADJUST controls on the 1238 Detector panel enable you
to establ ish the desired quadrature phase in the detector.i

Xl
HIGH

you are sure of frequency within ±1%. After warmup the
frequency stability is typically within ±0.001% for a few
minutes. Set the level as desired, up to 1.6 W into a wide
range of load impedances (0.25 Sl to 2.5 kSl) with the help
of the front-panel meter (reading 0.1 to 125 V) and you are
sure of a pure and constant output signal. Its level varies
less then ±2% with tuning; its distortion remains less than
0.4% over 3 decades of frequency and from open to

With the synchronizing circuit you can conveniently
lock this oscillator to a more stable source or provide sync
to a scope, counter, or another oscillator.

LEVER SWITCHES

Figure 1-1. Elementary diagram, 1621 Precision Capacitance-Measurement
System - GR 1616 bridge
with 1316 Oscillator and 1238 Detector. Conductance
circuitry is omitted.

NOTE
For more details about the 1316 Oscillator,
refer to its instruction manual.

1.2.5 Detector.
1.2.3 Standards.
Internal capacitance standards are of 4 types. The 3
most stable capacitors, made with Invar* steel, have
temperature coefficients about 3 ppm per °c, and the next
3 smaller capacitors (also Invar), about 20 ppm per 0c.
Their nitrogen dielectric is hermetically enclosed to assure
independence from effects of changing atmospheric pressure and humidity.
The 2 largest standards, of sealed,
low-loss mica construction, and the 4 smallest, being open
air-dielectric
capacitors, have temperature
coefficients
about 20 ppm per "C.
The set of 6 Invar and 2 mica standard capacitors are
well insulated from the environment, the thermal time
constant being 6 hours. So the bridge is remarkably
insensitive to fluctuations of environmental temperature
caused, for example, by a cycling air conditioner or the
movement of personnel.

The 1238 Detector, also developed for the 1621 system,
complements the osci lIator and bridge with ccnven ience
and sensitivity. Set the frequency dials to match those of
the oscillator, or select the flat response characteristic. Set
the gain as required; you can have full-scale readout for any
bridge error from 70 nV (tuned response) to 400 mV
(FLAT) - a range of 135 dB. Yet the instrument is immune
from damage by signals as large as 200 V, at any gain
setting
Watch the in-phase and quadrature meters as you
balance the bridge, they indicate conveniently whether to
adjust C or G next, and whether to increase or decrease the
weighting. (Convenient phase adjustments enable you to
compensate for any phase shift (0-360°) through bridge,
cables, and filter and to set the 2 phase-detector references
exactly in quadrature.)

1.2.4 Oscillator.

In addition to filter tuning, gain, and phase, front-panel
controls select linear vs 20-dB-compressed response, rejection of power-line frequency components, and meter time
constants from 0.1 to las.

The 1316 Oscillator, developed for the 1621 system, isa
conven ient, stable, powerfu I source. Set the 5 in-line decade
frequency dials anywhere between 10 Hz and 100 kHz, and

• Registered

1-2 INTRODUCTION

trademark

of the Carpenter

Steel Co., Reading,

Pa.

1.4 ACCESSORIES.

NOTE
For more detai Is about the 1238 Detector. refer
to its instruction manual.

1.3 CONTROLS, INDICATORS,

Table 1-3 lists the accessories supplied with the 1621
Precision Capacitance Measurement System. Power cords
are supplied for the 2 instruments that use them.
Rackmounting hardware is also supplied with the system, if it is
the "rack" version. For mounting refer to paragraph 2.5.

AND CONNECTORS.

Figures 1-2 and 1-3 illustrate the instrument system, front
and rear. Tables 1-1 and 1-2 further describe the individual
controls, indicators, and connectors.

7
8
34

9

33-_

10

32
31

30
29

28
27
16

2G

17
18
19

20
21

22
23
Figure 1-2. Front-panel controls, indicators, and connectors
1621 Precision Capacitance-Measurement System.

of the

INTRODUCTION

1-3

Table 1-1 -----------------FRONT-PANEL CONTROLS, INDICATORS, AND CONNECTORS
Fig. 1·2
Item

Name

Description

Function

Oscillator

POWER switch

Toggle switch, down position
OFF.

Turns oscillator on and off.

2

FREQUENCY
selector

Set of 2 rotary switches with
decimal steps, 0 ... 9, and 1
stepless pot with similar
calibration and detent at O.
Illuminated decimal points

Selects and indicates frequency, a 3-digit number.
Decimal-point illumination serves as pilot light.

3

FREQUENCY
range switch

Rotary switch with 4 positions Hz, Hz, kHz, kHz

Selects a frequency range, indicates units, and
controls illuminated decimal point in item 2.

4

OUTPUT VOLTAGE RANGE
switch

Rotary switch with 5 positions 15,50, 150,500,
150.

Selects output-voltage range and indicates fullscale meter range (item 5) Simultaneously
switches the output impedance from 0.25 S1
to 25 kS1 in decade steps.

5

Voltmeter

Ac meter with 0-50 and 0-15
scales; has mechanical zeroadjustment screw.

Indicates output terminal voltage, in ranges
selected by itern 4.

6

OUTPUT
ADJUST

Step less rotary pot with limits
labeled 0 and MAX

Controls output level in the range selected
by item 4.

Detector, Right Side

7

TIME
CONSTANT

Rotary switch with 5 positions 0.1,0.3, 1, 3, 10
SECONDS.

Controls the smoothing (integration) of
detected signals and hence, effectively, the
meter damping.

8

IN-PHASE
meter

Zero-center meter graduated
50-0-50; has mechanical zeroadjustment screw.

Ind ication of one component of bridge error
signal, such as the C component.

9

FINE ADJUST
(IN-PHASE)

Step less rotary pot.

Adjusts phase of item-8 reference so the
quadrature component is rejected (a fine
adj ustrnent).

10

PHASE SHIFT
(smaller knob)

Rotary switch with 4 oositions: 0°, 90°, 180°, 270°

Selects phase shift in coarse steps, supp lemented
by items 9, 11, 12.

11

PHASE SHIFT
(larger knob)

Step lessrotary control, calibrated -50° to +50°.

Shifts phase-detector references; set so that nearby meters (items 8, 13) respond independently
to C and G error.

12

FINE ADJUST
(QUADRATURE)

Step less rotary pot.

Adjusts phase of item -13 reference so the inphase component is rejected (a fine adjustment).

13

QUADRATURE

Zero-center meter graduated
50-0-50; has mechanical zeroadjustment screw.

Indication of one component of bridge error
signal, such es the G component.

14

SENSITIVITY
control

Step less rotary pot.

Fine gain control; use it to keep IN-PHASE
and QUADRATURE meters reading on scale
(does not affect MAGNITUDE meter)

1-4 INTRODUCTION

Table 1-1 (Cont) ----------------FRONT-PANEL CONTROLS, INDICATORS, AND CONNECTORS
Fig. 1-2
Item

Name

Description

Function

Bridge
15

Ground

Socket for banana plug.

Direct connection to master ground (electrical
midpoint of ratio transformer) and chassis.

16

2-TERMINAL
UNKNOWN
port

G900 precision coaxial
connector.

Connection for 2-terrninal "unknown" capacitor.
Note Neither terminal may be connected directly
to gnd (itern 15); outer = high; inner = low (F ig. 1-1).

17

3-TERMINAL
UNKNOWf\i port

Pair of G874 coaxial
connectors, LOW and HIGH.

Connection for 3-terminal "unknown" capacitor
Note outer shields are tied to master gnd (item 15)

18

EXTERNAL
STANDARD
port

Pair of G874 coaxial
connectors, identified as
HIGH and LOW.

Connection for 3-terrninal external standard capacitor for special measurements, comparisons, or range
extension.

19

ZERO
ADJUST

Step less lO-turn pot.

Capacitance offset adjustment Range a few aF in
CAL or 3-TERM positions of item 20; 3,3, and 50 pF
(respectively) in 2-TE RM Xl, X 10, and X 100 positions

20

TERMINAL
SELECTOR
(READOUT
MUL TIPLlER)

Rotary switch with 6 positions 3 TERMINAL (XlO,
Xl); CAL; 2 TERMINAL
(X 100, X 10, X 1)

Selects which UNKNOWN port (items 16, 17)
connects to the bridge, or neither (CAL position);
grounds the terminals of each port not so connected. Selects the READOUT MULTIPLIER apply it to both C and G.

21

EXT
MUL TIPLIER
switch

Rotary switch with 13 positionsOFF,-Ol,O,O.l
10

Gives any external standard (at item 18) one of 12
weights (including zero) or disconnects that port
from the bridge and grounds both terminals.

22

CONDUCT,L"NCE
rnultiplier lever

Lever switch with 7 positions
identified as Xl .. X 10-6

Gives the conductance standards an add itional
set of multipliers.

23

CONDUCTANCE
standards
levers

Set of 5 lever switches, each
12-position with readout
indicator -1,0,1
... 9, X 1st
digit, JLS next 3 digits, nS,

Determines effective value of internal conductance standards, along with items 20, 22. Note
at G balance, unknown
(readout X CONDUCTANCE multiplier + ext std G X EXT MUL T) X
READOUT MULTIPLIER.

24

CAPACITANCE
standards
levers

Set of 12 lever switches, each
12 position with readout

Determines effective value of internal capacitance
standards, along with item 20 Note at C balance, unknown = (readout + ext std C X EXT
MULT) X READOUT MULTIPLIER.

25

C MAX
switch

Lever switch with 4 positions
Up, no effect, down, shutters
over 1st 3 capacitance digits,
cumulatively in 3 steps.

Allows insertion or removal of first 1,2, or 3
(largest) standard capacitors from bridge Each is
removed when a shutter covers its indicator. Note
removal of large standards not used serves to reduce capacitive loading across detector (for best
sensitivity), setting item 24 to /ero does not.

26

C-standards
tri mrners

Set of 12 screwdriver adjustments hidden behind a small
panel; labels 100 nF .. 1 aF.

To trim or adjust internal C standards if necessary.
Labels indicate nominal (X 1) weight of corresponding standard when its switch is up (readout = X)

indicator
1,0,1
.9, X 1st
2 digits, nF; then in blocks of
3pF,fF,aF.

INTRODUCTION

1-5

Table ,-, (Cont]
FRONT-PANEL CONTROLS, INDICATORS, AND CONNECTORS
Fig. 1-2

Item

27

Name

C-standards
tri rn rners lock

Function

Description

Secures the small panel over item 26 to preserve its

Lock with keys.

adjustments.

Detector, Left Side

28

MAGNITUDE
meter

Meter, calibrated 0 to 100;
has mechanical zero-adjust-

Ind ication of bridge error signal level.

ment screw
GAIN, dB

29

COMPRESSION

30

Step attenuator;

12 positions,

Coarse gain control; use it to keep MAGNITUDE

20 .. 130 dB

meter reading on scale

Push-button switch (push to
engage, push again to release).

Out linear response, full gain In compressed response, 20-dB-larger signal can be handled with
meters on scale

31

LINE
REJECTION

Push-button switch (push to
engage, push again to reo
lease)

Out normal. In 40-dB attenuation of line-frequency
component in bridge error signal (Circuit can be
adapted for 60 or 50 Hz)

32

Frequencyrange

switcn

Rotary switch with 5 positions FLAT, Hz, Hz, kHz,
kHz.

Selects broad-band characteristic or frequency
range of tuned response, indicating the units and
controlling the decimal point in item 34.

33

POWER
switch

Toggle switch, up ON; down
OFF.

Turns detector on and off

34

FREOUENCY
selector

Set of 3 rotary switches with
decimal steps, O. 10. IIlumi
nated decimal points

Selects and indicates frequency to which detector
is tuned (unless item 32 says F LA T). Decimal"[Joint
illumination serves as pilot light

Table '-2
REAR-PANEL CONTROLS AND CONNECTORS
Fig. 1-3

Item
Oscillator

Name

Description

Function

1R

QUADRATURE
REFERENCE
OUTPUT

BNC jack *

Provides a reference output, 90° leading the
"in-phase" reference, at 1.3 V open circuit
(connect to item 21R)

2R

IN·PHASE
REFERENCE
OUTPUT

BNC jack *

Provides the other reference output, at the same
level. Approx in phase with item 4R (connect

'-6 INTRODUCTION

to item 20R)

Table 1-2 (contl-----------------

-----------------

REAR PANEL CONTROLS AND CONNECTORS
Fig. 1·3
Item

Name

Description

Function

3R

EXT SYNC

BNC jack *

Use for synchronization, if desired As an input,
lock range is ±2%/V r ms, up to 10 V As an
output. 0.3 V behind 27 kSL

4R

POWER
OUTPUT

BNC jack *

Main output, up to 16 W max, rnay be 125 V
open circuit or 5A short circuit (connect to
item 13R)

5R

8/10 AMP
fuse

Fuse in extractor-post
holder

Protection against damage from short circuit.

lR

2H

3H

411

Gil

lH

8H

10H

11 H
21R
2011

121l

19R

nil

18R
141l

17R

Figure 1-3. Hear-panel controls, connectors, and cables. Dashed lines indicate cable connections.

INTRODUCTION

1-7

Table 1-2 (Contl----------------REAR-PANEL CONTROLS AND CONNECTORS

----------------Fig. 1-3
Item

Description

Name

Function

6R

Line-voltage
switch

Slide switch (labeled 50-60
Hz); 2 positions 100-125 V,
200-250 V.

Accomodates power supply to either range of
line voltages.

7R

Power plug

3-pin power plug. t

Connects from power line and earth ground.

Detector, Left Side

8R

1/2 AMP
fuse

Fuse in extractor-post
holder

Protection against damage from short circuit.

9R

Line-voltage
switch

Slide switch (labeled 50-60
Hz) 2 positions 100-125 V;
200-250 V.

Accomodates power supply to either range of
line voltages.

lOR

Power plug

3-pin power plug. t

Connects from power line and earth ground.

11R

DC-METER
OUTPUTS

5-pin socket.

Outputs for remote metering; all 3 meter circuits
included.

12R

INPUT SIGNAL

BNC jack.*

Main input to be detected (connect from item 14R).

Bridge and Cables

13R

GENERATOR
INPUT

BNC jack.*

Input port for audio-frequency power to bridge
circuitry. Connect from item 4R.

14R

DETECTOR
OUTPUT

BNC jack."

Output port for bridge error signal (unbalance).
Connect to item 12R.

15R

0776-2020

BNC patch cord

Interconnect items 4R, 13R.6.

16R

0776-2040
8161-5200

BNC patch cord
BNC patch cord (red band)

Connect items 1R to 21 R, 2R to 20R 6.
Connect items 12R to 14R 6.

17R

BCD
CONDUCT ANCE
OUTPUT

36-pin socket

Indicates in BCD code to external instruments the
CONDUCTANCE-readout and CONDUCT ANCEmultiplier values (items 22, 23) and the position of
the TERMINAL SELECTOR (READOUT MUL TIPLlER) switch (item 20).6

18R

BCD
CAPACITANCE
OUTPUT

50-pin socket

Indicates similarly the CAPACITANCE readout, i.e..
the positions of the levers of items 24 and 25.6

Detector, Right Side

19R

AMPLIFIER
OUTPUT

BNC jack*

Output for remote instrumentation; ac voltage.

20R

IN-PHASE
REF INPUT

BNC jack*

Input reference for phase-sensitivedetectors;
required level> I' V rms.

21R

QUADRATURE
REFERENCE
INPUT

BNC jack*

Input like 18R except leading that by 900.

·BNC jack accepts

Amphenol

6Refer

to para. 2.8.

t Refer

to note, pg. 1-12.

1-8 INTRODUCTION

"BNC"

plug or military

connector

no. UG-88/U.

-------------------Table

1·3 -----------------ACCESSORIES SUPPLIED WITH THE 1621 SYSTEM

Name

Power
cord

Description or Function

3-wire AWG number 18 type SVT cable, rated at 7A,
230 V. The connectors, designed for 125-V operation, conform to the Standard for
Grounding Type Attachment Plug Caps and Receptacles, ANSI C73.11-1963.
Length. 7 ft. 2 required (1 each for oscillator and detector).

GR Catalog No.
(Type)

4220-0220

Patch
cords

Shielded cable with BNC plugs; see para. 2.8;
length 15 in. (2 req'd)
length 24 in. (1 req'd)
length 15 in, double shielded, red banded (1 req'd)

0776-2040
0776-2020
8161-5200

Plug

To fit DC METER OUTPUTS socket; pins 5 (Amphenol 126-217)

4220-5401

Cap

Plastic dust cover for G900

0900-7190

connector

INTRODUCTION

1·9

Installation
-Section2

2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8

GENERAL
DIMENSIONS
ENVI RONMENT
BENCH MODELS
RACK MODELS
POWER-LINE CONNECTION
LINE-VOLTAGE REGULATION
SYSTEM CONNECTIONS

2-2
2-2
2-2
2-2
2-2
2-4
2-4
2-4

1616

1621
DIM
A
B
C

o
19.8

FRONT

.,

C

1

I

348.0
381.0
596.9
617.2

mm
mm
mm
mm

(13.7
(15.0
(23.5
(24.3

in.)
in.)
in.)
in.]

246.6
327.7
330.2
350.5

mm
mm
mm
mm

(11.6
(12.9
(13.0
(13.8

in.I
in.)
in.l
in.l

BENCH MODEL

Tn

PANEL

BRIDGE

SYSTEM

END VIEW

D

J

Figure 2-1. Dimensions of the bench models.
~ CLEARANCE FOR REAR CABLES

1621

P~

1I

TOP VIEW
----------

U

DIM

P
Q

R

1
U

19

FRONT

S
T

SYSTEM

368.3
292.1
332.7
40.6
576.6

mm
mm
mm
mm
mm

(14.5 in.]
(11.5 in.l
113.1 in.)
(1.6 in.]
(22.7 in.l

1616
BRIDGE

368.3
269.2
309.9
40.6
309.9

mm
mm
mm
mm
mm

(14.5 in.)
(10.6 ln.)
(12.2 in.)
(1.6 in.)
(12.2 in.)

RACK MODEL

PANEL

~

END VIEW

Figure 2-2. Dimensions of the rack models.

INSTALLATION 2-1

2.1 GENERAL
The 1621 Precision Capacitance Measurement System,
or the 1616 Bridge alone, is available for either bench use
or for installation in an EIA Standard RS-31O 19-in. relay
rack with universal hole spacing. Appropriate cabinet and
hardware sets are available for conversion of a bench model
for rack installation or vice versa.
Locate the instrument for convenience of operation and
in a suitable environment. Avoid blocking the flow of air
through the vents. Some open bench area, to the right or in
front of the bridge, should be provided.
NOTE
If you assemble a 1621 System from separate
instruments,
either
install them in the
4177-2621 cabinet or consult GenRad
about the need for a magnetic shield between
bridge and detector. If you convert a 1621
from bench to rack, transfer the magnetic
shield.

2.2 DIMENSIONS.
The dimensions of bench and relay-rack models of the
system and of the bridge are given in Figures 2-1 and 2-2.

2.3 ENVIRONMENT.
The system is designed to operate in standards laboratories, in which the environment is typically very well
controlled. All specifications are valid over a temperature
range of 22-24°C. Storage range is -20 to +70°C.

2.4 BENCH MODELS.

b. Withdraw
cabinet.

each instrument

out

of

the

2.4.2 Conversionfor Rack Mounting.
To convert a bench instrument for rack mounting,
exchange the cabinet and install appropriate hardware, as
follows
a. Obtain the appropriate Rack-Mounting Cabinets, as
described in Table2-1, from GenRad.
b. Obtain, optionally, a Bracket Set (Table 2-1) for each
cabinet. Brackets are especially recommended for heavy
instruments, which need support from the rear rail of the
rack.

--------Table

2-1---------

RACK-MOUNTING CABINETS AND
BRACKETS FOR 1621 SYSTEM
Quantity

1
1
1
3

Description

Rack-mounting cabinet (for oscillator).
Rack-mounting cabinet (for detector).
Rack-mounting cabinet (for bridge).
Sets of rear-support brackets and
screws (1 for each instru ment).

Part No.

4174-3240
4174-3624
4174-3627
4174-2007

c. Remove the cabinet, as in paragraph 2.4.1.
d. Remove the rear cover from the bench cabinet, with
screws (B, Figure 2-3), for later installation on the rack
cabinet.
e. Proceed with the rack installation, skip to paragraph
2.5.2, step b.

Figure 2-3.

2.4.1 Cabinet Removal.

2.5 RACK MODELS.

To remove the bench-model cabinet, first stand the
system (or instrument) in the normal, horizontal position,
free of all cables, and proceed as follows
a. Remove the 4 dress-panel screws (A) accessible
through holes in the handles of each instru ment.

2.5.1 General.
Each rack model comes completely assembled in a
suitable metal cabinet, which is designed to stay semi-permanently in a rack. Each instrument can be drawn forward
on extending tracks for accesswith support, or (with a lift)

Figure 2-3. Bench-cabinet installation.

2-2 INSTALLATION

forward,

o

~~~

TRACK
/

~INSTRUMENT

B

A
.17-5

Figure 2-4. Rack-cabinet installation.

withdrawn completely. The cabinets listed in Table 2-1 are
all included with a rack-model 1621 system, together with
screws. Table 2-2 lists the screw sizes for reference.

2-2--------

---------'Table

KEY TO SCREW SIZES
Ref Figs.
Description

2-3; 2-4

A
8
C
E

No. Length,
thds/in.
inch

Dress-panel screws with washers

10-32

Thread-cutting screws
Thread-cutting screws
Thread-forming screws

10-32

Figure 2-4.

2.5.2 Installation.
Directions follow

10-32
8-32

.56
.50
.50
.19

for mounting the cabinet in a rack and

installing the instrument on its tracks
a, Remove 4 dress-panel screws (Al and slide the
instrument out of the cabinet until the tracks are fully
extended Continue pulling the instrument forward until
motion along the tracks is stopped. At this juncture, tilt the
front of the instrument up slightly and continue withdrawal, past the stoos. unti I it is free.
b. Insert the rack cabinet wherever desired in the rack
be sure it's level - and fasten it with 4 screws (C) to the
front rails
c. If the rack contains a rear support rail, use brackets
(0) to support the cabi net with the rear rai Is; open-slotted
screw holes allow positioning.
d. Use the set of slots in the sides of the cabinet that
allow alignment of the open-slotted holes in the brackets
with threaded holes in the rail. The long flange should extend to the rear.
e. Insert screws (E) from inside the cabinet, through the
slotted holes and drive them into the holes in the long
flange of the bracket. Each side takes 2.

NOTE
Start the screws in the appropriate holes off the
rack, to make the threading easier.
f. Pass screws (8) through brackets and screw 2 into
each rear rail. (Details may be varied to suit particular
situations. )
g. To install the instrument, first set its rear edge in the
cabinet front open Ing. SIide the instrument back, mak ing
sure that the rear and the upper front slide blocks engage
the tracks. (Stops prevent further insertion.1
h. Pull the instrument forward with the tracks, keeping a
hand on each side (fingers underneath). Slide the instrument back about Y, in. along both tracks, past the stops, by
pressing down on the tracks (with thumbs) while tilting the
front of the instrument up slightly.
i. Push the instrument back into the rack, checking for
smooth operation of the tracks and slide blocks.

NOTE
The instrument is now readi Iy accessible for
behind-the-panel adjustments. It slides in and
out freely on extendi ng tracks.

2.5.3 Conversion to Bench Use.
To convert
exchange the
a.Obtain a
1621 system.
from GenRad.

a rack-mounting instrument for bench use.
cabinet. as follows:
for the
Bench Cabinet. part no. 4177-2621
or 4172-4106
for the 1616 bridge alone.

b. Remove the instrument from the rack cabinet, after
removing the panel screws (A, Figure 2-4). (When free
motion along the tracks is stopped, tilt the front of the
instrument up slightly to clear the stops.)
c. Slide the instrument into the bench cabinet.
d. Fasten instrument to cabinet using dress-panel screws
(A, Figure 2-3).
e. Transfer the rear cover, with screws (8), from rack
cabinet to bench cabinet.

INSTALLATION 2-3

2.6 POWER-LINE CONNECTION.
Power requirement for the 1621 system is 51 W at
100-to- 125 or 200- to- 250 V, 50- to-60 Hz Make connection
as follows
a. Set the line-voltage switches on the rear panels of
oscillator and detector (Figure 1-3) to correspond with the
available power-line voltage. Use a small screwdriver to slide
the switch.
b. Connect the external power line to each power plug
using the power cords supplied or equivalent, 3-conductor
cords (para. 1-4)
The fuses should have the current ratings shown on the
rear panels (Figure 1-3) regardless of which line-voltage
range is chosen in step a.

2.7 LINE-VOLTAGE REGULATION.
The accuracy of measurements accomplished with precision electronic test equipment operated from ac line
sources can often be seriously degraded by fluctuations in
primary input power l.inevoltaqe variations of ±15% are
commonly encountered, even in laboratory environments.
Although most modern electronic instruments incorporate
some degree of regulation, possible power-source problems
should be considered for every instrumentation setup. The
use of line-voltage regulators between power lines and the
test equipment is recommended as the only sure way to
rule out the effects on measurernent data of variations in
Iine voltage.

2.8 SYSTEM CONNECTIONS.

Figure 1-3.

2.8.1 Oscillator, Bridge, and Detector.
Make the 4 essential connections among the instruments,
using the 4 BNC patch cords supplied with the 1621 system
(refer to Table 1-3) as follows
a. Test power to the bridge 1316 POWER OUTPUT to
1616 GENERATOR INPUT.
b. Unbalance signal to the detector: 1616 DETECTOR
OUTPUT to 1238 INPUT SIGNAL. (Red-banded cable).
c. Reference siganls to the detector: 1316 REFERENCE
OUTPUTS to 1238 REFERENCE INPUTS (one cable for IN
PHASE, one for QUADRATURE).

punch coupler, or comparator Use a 50-pin plug, Amphenol PIN 57-30500 (or equivalent) and cable such as Alpha
No. 1181/50, which has AWG No. 22 stranded wires.* For
pin identification, refer to Figure 2-5.
Notice that the output data is provided by SWitch closures
only. The switches in the bridge are rated for up to 0.5 A at
110 V (ac) with resistance loads. If, however. you want for
example a logical "1" (the on state of each data bit) to be
represented by +5 V with respect to the system ground (pin
30) then use an external +5-V power supply and connect it
from ground to VREF (pin 25). Logical "0" is an open circuit.
The capacitance readout is available in binary-coded
decimal form, the code being "1-2-4-8" as detailed in Table
2-3. The body of the table contains only binary numbers,
composed of bits 0 and 1. Notice the extension of the usual
BCD table, to include X (ten) and negative 1.

As an example, suppose the bridge readout is 396 pF.
Figure 2-5 shows us that pins 20, 19,43, 17, 16 and 40 are
"1", while pins 45, 44, 18, 42, 41, and 15 are "0". This
kind of data is commonly accepted by printers.
However, suppose the bridge readout is 4(-1) 6 pF
the same capacitance I Figure 2-5 and Table 2-3 tell us that
pins 44,43,18,17,16
and 40 are "1" while pins 20,19,
45,42,41 and 15 are "0".
A sufficiently sophisticated system will make a computation and print 396 pF (for example). A simple system
utilizes a 12-character printer (including -1 and X as well
as the usual 0
. 9) and will print out just what the bridge
readout shows. If the system must use a printer that does
not recognize -1 and X a logic circuit may be fabricated, to
detect the occurrence of either -1 or X anywhere in the
readout, and trigger an alarm such as a buzzer or a change
in color of the printout (Table 2-3 shows that when both
2-weight and 8-weight binary signals of anyone decimal
digit are" 1", there is such an occurrence. In our example,
the pins involved are numbered 18 and 43.)
Either provide the system complexity needed to handle
data containing -1 and X without ambiguity, or make sure
that the operator removes them from his final balance
adjustment.
Operating instructions
in this manual are
written for the latter case.

NOTE
The capacitance data must be scaled up by 1,
10, or 100 (the READOUT MULTIPLIER)
code for which is given below.

2.8.2 BCD-Capacitance-Output Connector. Figures 1-3,2-5.
Make connections from the BCD CAPACIT ANCE OUT
PUT socket at the rear of the bridge, if you want to record
or process the C-measurement data with a printer, card-

2-4 INSTALLATION

'Alpha

Wire Corp.,

Elizabeth,

N.J.

pF
20

IF
2

200

The READOUT MULTIPLIER
is one of 3 values, Xl,
X 10, X 100. However, the BCD data has 6 possible values,
corresponding to the 6 positions of the TE RM INA L
SELECTOR switch, as shown in Table 2-4.
For example: a binary 101 (decimal 5) means the
READOUT MUL TlPLIER is X10 and (incidentally)
the
measurement is being made via the 2- TERMINAL port.

OF

20

40
80

8

800

~~~~
pF

IF

of

161618

Figure 2-5. Terminal identification at BCD CAPACITANCE OUTPUT connector (A.J13, rear panel of
bridge). Arrow toward connector = input; away = output. These weights apply when the TERMINAL
SELECTOR switch is set to READOUT MULTIpLER = Xl (either of 2 settings). This is a rear view,
exterior of socket.

The BCD output is zero for any internal standard that
has been removed from the bridge by the C MAX switch,
regardless of the position of the lever switch normally
controlling that standard.

------Table

If your printer has decimal points, drive them from the
READOUT MULTIPLIER as follows (for printout in pF)
pins 33, 34, 35 drive the points following columns 5,6,7,
respectively Printing of a 2nd point indicates 2 terminal Xl
or X 10; all 3 points indicate CALIBRATION

--------

Table 2-4-------READOUT-MULTIPLIER
4

0
0
1
1
1
1

Decimal
(Equiv.)

Weight
2

0
1
1

1
0
1

1
2
7

0
0
1

0
1
0

4
5
6

CODE

TERMINAL-SELECTOR
Switch Position

XlO
Xl

3-term ina I
3-term inal
CAL
2-terrninal
X 100
2-terminal
XlO
2-terminal
Xl

2-3-----

BINARY-CODED DECIMALS (BCD)
Signal Weight (Binary)
8
4
2
1

Digit
(Decimal)

1
0
0

0
0
0

1
0
0

1
0
1

-1
0
1

0
0
0

0
0
1

1
1
0

0
1
0

2
3
4

0
0
0

1
1
1

0
1
1

1
0
1

5
6
7

1
1
1

0
0
0

0
0
1

0
1
0

9

8

X,

2.8.3 BCD-Conductance-Output Connector.Figures 1-3,2-6.
Make connection, similarly, from the BCD CONDUCTANCE OUTPUT socket, using a 36-pin plug, Amphenol
PIN 57-30360 (or equivalent) For pin identification refer
to Figure 2-6. In addition to the basic conductance-readout
data, here is multiplier data.
The conductance
multiplier
is one of 7 values,
1
10-6 Its exponent (magnitude only, expressed in
BCD code) appears at pins 13, 14, and 30. For example:
these 3 pins at "0" state means zero exponent, i.e. the G
multiplier is 1.

Figure 2-6. Terminal identification
at BCD CONDUCTANCE OUTPUT connector (A-J12, rear panel
of
bridge).
Arrow
toward
connector = input;
away = output. The indicated conductance weights
apply directly when the effective multiplier is 1, for
example:
CONDUCTANCE
multiplier = 10'2 and
READOUT MUL TlPLlER = Xl00. This is a rear view,
exterior of socket.

2.8.4 Analog Outputs.

Figures 1-3,2-7.

AMPLIFIER
OUTPUT. Use a BNC patch cord to
connect this signal to remote monitoring or recording
equipment if desired. This ac signal is proportional to the
MAGNITUDE meter deflection, and is 4 V rms at full scale.
DC METER OUTPUTS Use the 5-pin plug supplied (see
Table 1-3) and cable suited to your system, if you wish to
have remote indication of the 1238 front-panel meter

INSTALLATION

2-5

deflections. A suitable cable is Alpha No. 1175, which
contains AWG No. 22 wires.
Pins are designated as shown by Figure 2-7 Only pin H
of these circuits may be grounded. A is +, B is - for the
MAGNITUDE meter circuit (6 V corresponds to full-scale
deflection) D is +, H is - for the IN PHASE meter circu it;
E is +, H is - for the QUADRATURE meter circuit. For
each of these 2 circuits, the level is 1 V for full-scale
deflection of the corresponding front-panel meter, when
the SENSITIVITY
control is ccw (minimum)
However,

2-6 INSTALLATION

because that control affects the front-panel meters, not the
DC METER OUTPUT voltage, the voltage is relatively
lower when that control is cv«.
.m _

~-~ ::1
0--

e• -

._-

A

+}

MAGNITUDE

B-

D

+

IN PHASE

QUADRATURE

Figure 2-7. DC METER OUTPUTS
Detector, exterior (rear) view.

socket on 1238

Operation
- SectionJ
3.1 PRELIMINARY CHECKS
3.2 FUNCTIONAL SELF-CHECK
3.3 PHASE ADJUSTMENT
. .
3.4 CONNECTION OF UNKNOWN CAPACITOR.
3.5 BALANCE AND READOUT
. . . . . .
3.6 PARAMETERS OF THE UNKNOWN CAPACITOR
3.7 FREQUENCY
. .
3.8 VOLTAGE LEVEL.
3.9 ACCURACY.
. .
3.10 PRECISION
3.11 EXTERNAL STANDARDS
3.12 PRECISE COMPARISONS.
3.13 NON-COAXIAL 2-TERMINAL CAPACITORS
3.14DCBIAS.
. . . . . . . .
3.15 REVERSED CONFIGURATION.
. .

3-1
3-1
3-2
3-3
3-4
3-7
3-8
3-10
3-10
3-14
3-15
3-17
3-17
3-20
3-22

&
WARNING
Dangerousvoltagesmay be presentat the terminals of this instrument.
Refer to specificwarningscontained in this section.
NOTE
The following instructions apply literally to the
1621 system. If your 1616 bridge is connected
to an oscillator and/or detector other than the
GR 1316 and 1238, interpret the corresponding
parts of this section appropriately.

&
CAUTION
Do not connect a power cord until each linevoltageswitch has been set properly.

3.1 PRELIMINARY

CHECKS.

Refer to paragraph 1.3 for figures illustrating the controls, indicators, and connectors and for tabulation of their
functional descriptions. The recommended initial operating
procedure follows:
a. Check that the line-voltage switches on the rear panels
of oscillator and detector are positioned according to the
available power-line voltage (either 100-125 Vor 200-250
V). To slide these switches, use the tip of a
small
screwdriver
b. The LINE REJECTION filter in the 1238 Detector
has been set for 50 or 60 Hz by GR (it can be purchased
either way) or by a user. If you need to verify that it

matches your power-line frequency (or reset it), refer to
the 1238 Instruction Manual.
c. Check that the 4 BNC patch cords are in place at the
rear, connecting oscillator, detector, and bridge as described
in paragraph 2.8.
d. Check that 4 meters read zero (2 of them at midscale). If necessary, adjust each with a small screwdriver, at
the recessedscrew just below the meter.
e. Set the OUTPUT VOLTAGE RANGE to 15 V. (Higher voltages should be used only when they are needed,
usually to facilitate high precision in measurements.)

&
WARNING
Beware of hazardous voltages: ~ 350 V at
3-TERM UNKNOWN HI or EXT STD HI;
~ 35 V at 2-TERM UNKNOWN HI (outer
shell), while oscillator is set to a high level.

f. Connect the power plugs (rear panel) to a suitable
power line, using the power cords supplied. Flip the POWER switch UP (front panel) on each, oscillator and detector.
Verify that a decimal point is illuminated in each FREQUENCY selector.
3.2 FUNCTIONAL SELF-CHECK.
Adjust oscillator and detector to the same frequency and
make a balance to check for system operation, as follows:

OPERATION

3-1

a. Set the front-panel controls as listed (from upper left
to lower right, Figure 1-2)
FREQUENCY
101 kHz (Oscillator)
OUTPUT VOLTAGE RANGE - 150
OUTPUT ADJUST - MAX
FREQUENCY - 101 kHz (Detector)
TIME CONSTANT 1s
LINE REJEr:TION - push button out
COMPRESSION - push button out
GAIN - 30 dB
SENSITIVITY
ccw (minimum)
PHASE SHI FT - 180°

a

FINE ADJUST - rnidrange (both controls)
EXT MULTIPLIER - OFF (Bridge)
TERMINAL SE LEC, UR - CAL
ZERO ADJUST -- ccw
C MAX - down (Asterisks below represent 3 closed
shutters)
CAPACITANCE
* * nF *05 pF 000 fF 0000 aF
CONDUCTANCF
a fLS 0000 nS X 1(f6
b Reset the GAIN If necessary to make the MAGNITUDE meter read near midscale (20 to 80)
Tune the detector to match the oscillator, as indicated
by a peak reading of the MAGNITUDE
meter, to the
nearest step Fine tune with the oscillator 3rd dial
d Reset the CAPACIT ANCE to ** nF HOO pF 000 fF
0000 aF Verify that MAGNITUDE drops to zero
e. Reset the GAIN to 100 dB. Observe that a noticeable
unbalance occurs when the 1 fF lever is moved to 1 or -1

QUADRATURE
meter reads 25 or more (in either direc
tion) and both read on-scale
c. With the PHASE SH I FT controls, bring the QUADRATURE meter approximately to zero. If the range of the
larger knob is not sufficient, turn the smaller knob to an
adjacent position.
d. Now the I N PHASE meter should read upscale, to the
right (because C readout is greater than C unknown) Reverse the direction of the meter deflection, if necessary, by
turning the smaller PHASE SHIFT knob 2 clicks
e See that both FIN E ADJUST knobs are set to amid
position (dot L) Increase the C unbalance until the MAG·
NITUDE meter reads 8099. Set SEf'>JSITIVITY to max,
cwo Set the PHASE SHIFT for zero on the QUADRATURE
meter. If zero cannot be obtained wi th the PHASE SH 1FT,
use the QUADRATURE
FINE ADJUST.*
t Reestablish the null reading and refine it, If necessary,
uSing the C and G levers The MAGNITUDE reading should
not exceed 2, or % division
g Unbalance the bridge by raisinq CONDUCTANCE
lever switches, achievinq about the same magnitude of
unbalance as before
80-99 (Try usrnq the 1- fLS lever)
Now the QUADRATURE
meter should read upscale, to
the right (G readout is greater than G unknown) Trim the
IN PHASE FINE ADJUST onlv to make the adjacent meter
read zero, or a uiinu num (Do not expect zero at all
possib Ie G unbalances at high !req )
h. If, In step g, the meter reads down scale, interchange
the IN PHASE and QUADRATURE
connections at the
REFERENCE INPUTS behind the 1238 Detector Then
repeat steps a through g, above.

3.3 PHASE ADJUSTMENT.
In order to enjov the convenience of the phase-sensitive
indicators, adjust the phase shifters as described below. The
adjustment is then useful as long as the frequency is unchanged, the detector FREQUENCY
remains tuned for
maximum response, and the CONDUCTANCE multiplier
is unchanged Readj ustrnent of phase shitt may be made
while the bridge IS berng balanced, refer to paragraph 35
NOTE
The goal IS to minimize the response of one
phase sensitive meter to changes in C, the other
to changes in G The following procedure,
though not ideal for every measurement condi
tion, ISadequate in general
a Reestablish the null reading Oil the iv'lAGNITUDE
meter, as in para 32 However, set the CONDUCTANCE
multiplier so its exponent IS 5, --4, -3,
2, for a fre
quency of a 1, 1, 10, or 100 k Hz . respectively Then switch
GA IN to 60 dB and unbalance the bridge by raising CA
PACITANCE lever switches only, making the MAGNITUDE
meter read about 20 (For the first attempt, use the
10 fF lever)
b. With the SENSITIVITY control, set the signal level in
the phase-sensitive detectors so that either I N PHASE or

3-2 OPERATION

---------

Table 3-1
SENSES OF PHASE-SENSITIVE

REFERENCE
connections

METERS

PHASE-SHI FT controls set to
Meter

0°

90°

180

0

270

0

Normal

IN PHASE

C+-

G+-

C-+

G-+

Normal

QUADRATURE

G+-

C-+

G-+

C+-

Crossed

IN PHASE

G*'-

C+-

G-+

C-+

Crossed

QUADFiATUFiF

C+-

(;-+

C-+

G+-

We have assumed that it is preferable for the IN PHASE
meter to respond upscale (to the right) for C, > C x' and the
QUADRATURE
meter, similarly, for Gs G x However, if
you prefer to associate "I N PHASE" with "conductance",
interchange the REF ERE NCE connections at the rear panel
(cross the patch cords) If you prefer the opposite sense of
response, set PHASE SHIFT near 0° instead of 180° Table
3- 1 shows the 8 possible combinations.

>

• Phase adjustment
is now
f and 9 are refinements.

adequate

for most

measurements.

Steps

3.4 CONNECTION
3.4.1

OF UNKNOWN

Three-terminal Capacitors.

CAPACITOR
3-TERMINAL,
SHIELDED
UNKNOWN CAPACITOR

Figure 3-1.

As the simplified daigram shows, the 3-terminal
measurement evaluates the direct capacitance C (and
conductance) between HIGH and LOW bridge ter?ninals.
Shielded cables or patch cords (G874-422A or 874-R22LA)
ma be used to connect these terminals to a capacitances to
ground (C ' C ' including cable capacitances) are excluded
.
Is
hs
f rom the measurement. but If large enough they may affect
accuracy.

Enclosure. Check that the capacitor to be measured is
enclosed in a shield, for connection to bridge ground. Not
only is the shield necessary for precision measurements, but
it must be mechanically fixed (permanently or repeatably)
for Cx to be defined. Physically, as the shield gets larger
and farther from the plates of Cx, the value of Cx becomes
larger and less dependent on spacing of the shield. However,
that dependency never vanishes.
Use the 3-terminal component mount G874-X, or (if that
is too small) a metal box, if an enclosure must be supplied -,

Connectors. Check that the "unknown" capacitor has two
coaxial connectors for convenient connection to the bridge. If
the capacitor is relatively small and light weight and is
equipped with G874 connectors spaced 1.25 in. apart
(center-to-center), patch cords are not necessary. The
capacitor can then be plugged directly into the bridge.
However, shielded patch cords are convenient. their use
allows the capacitor being measured to sit on a bench surface
or at a remote location. The parameters of such cables are
negligible in a 3-terminal measurement, except for noise
introduced into the detector when they are moved.
Provide connectors if necessary, such as those listed in
Table 1-5. Connect the capacitor to the 3-TERMINAL
UNKNOWN ports.
Reversal. The measured direct capacitance is not
changed in most three-terminal measurements when the
connections to the capacitor are reversed, i.e.. bridge HIGH
to capacitor L instead of bridge HIGH to capacitor H
Shielding. Keep both HIGH and LOW connections shielded as a general rule. At least one must be shielded or the
capacitance between connecting terminals and wires becomes part of the measurement.
The HIGH connection may be made without shielding, If
you prefer. This terminal, being at the high-voltage but
low-i mpedance output of the transformer, is not sensitive
to pickup from external sources and seldom needs to be
shielded. There is voltage from the HIGH terminal to
ground. Much capacitance or conductance connected from
this terminal to ground reduces the transformer output
voltage and introduces possible errors into the measured
direct capacitance and conductance. In general, keep such
shunt C below 200 pF (see para. 3.9)
Keep the LOW connection shielded. This terminal, being
at the low-voltage but high-impedance input to the detector, is very sensitive to noise and signal pickup from ex-

LOW

....
OSCILLATOR

LOW

Figure 3-1. Simplified bridge diagram - 3-terminal
connection of unknown capacitor.

ternal sources and must be completely shielded for lowcapacitance or any precision measurements. There is no
voltage from LOW to ground when the bridge is balanced.
Much capacitance or conductance from the LOW terminal
to ground shunts the detector, and by reducing sensitivity
limits the precision of your measurements. In general, keep
such shunt C below 200 pF.
Capacitance of the recommended
R22A (or 874-R22LA) is 90 pF.

patch cord,

3.4.2 Two-terminal Coaxial Capacitors.

G874-

Figure 3-2.

As the diagram shows, a 2-terminal measurement evaluates the capacitance Cx (and conductance) between the
inner conductor and the shell of a coax ial structure. The
inner conductor connects to LOW bridge terminal (which
must be shielded for precision measurements), the shell to
HIGH. Although the shell serves as a shield, do not connect
it to ground. Capacitance to ground Ch g is excluded from
the measurement.

&
WARNING
Beware of possibly hazardous voltage on shell
of "unknown" capacitor whenever generator is
set to a high level.

If necessary. use of a coaxial adaptor (or special test
fixture) between the single G900 connector on the bridge
panel and the capacitor being tested. The capacitance of any
such adaptor is included in the measurement. However, the
ZERO ADJUST control can be used to compensate the bridge
- up to 3 pF - making the CAPACITANCE READOUT directreading for any unknown capacitor connected to the adaptor,
as described below.

OPERATION

3-3

Gx

ex

2-TERMINAL COAXIAL
UNKNOWNCAPACITOR

LOW

L-----11I

OSCILLATOR

LOW

Figure 3-2. Simplified bridge diagram - 2-terminal
connection of unknown capacitor.

NOTE
Non-coaxial 2-terminal capacitors can also be
measured; refer to para. 3.13.

Enclosure. Check that the capacitor to be measured is so
constructed that the shell shields the inner terminal. and that
both are isolated from ground.
Use the 2-terminal
component mount G874-ML if an enclosure must be used.
Connector. Check that the capacitor has G900 or G890
connector for convenient connection to the bridge. If not,
provide the latter type. or whatever fits your special test fixture
(see below].
Zero Adjustment. Compensate the bridge for capacitance of its 2-terminal connector and - if one is used - the
attached short coaxial line, as follows:
a. Check that the front-panel controls are set as of the
completion of functional checks and phase adjustment,
paragraphs 3.2 and 3.3. (Frequency is optional; see para.
3.7.)
b. Set the TERMINAL SELECTOR to 2 TERMINAL
Xl, X 10, or X 100 for measurements up to 1 nF, 10 nF,
or 10 pF, respectively.
c. Set CAPACITANCE lever switches to the value of the
fringing capacitance of the connector that will receive the
unknown capacitor. In the simple case that the READOUT
M U LTIP LIE R is Xl, no adaptor is needed, and accuracy of
±8 fF is sufficient, select CAPACITANCE and CONDUCTANCE readouts as follows

** nF *00 pF 115 fF 000.0 aF and 0 fJ-S000.0 nS X

1(J6

If the READOUT MULTIPLIER was set to XlO or X 100,
use proportionally smaller settings (11 fF 500 aF, or 1 fF
150 aF, respectively). For greater accuracy, refer to para
3.9.5.
d. Adjust the ZERO ADJUST control for minimum
reading on the IN PHASE meter. Turn the GAIN control to

3-4 OPERATION

90 dB, or as far as is appropriate to get a good indication
for this adjustment.
The adjustment is now valid for a particular terminal
capacitance and READOUT MULTIPLIER. If you change
either of these (and leave the TERMINAL SELECTOR on
one of the three 2-TERMINAL positions) repeat the zero
adjustment.
e. Connect the capacitor to the 2-TERMINAL
UNKNOWN port.
Special Test Fixtures. The details of special fixtures are
beyond the scope of this manual. However, these comments
apply:
1. The ZERO ADJUST range is sufficient to compensate
up to about 3 pF with READOUT MULTIPLIER set at
Xl or Xl0, 56 pF at Xl00. Therefore, bridge readings
will generally need correction (by subtraction) unless
you provide an external capacitor for test-fixture compensation. Refer to para. 3.11.
2. Highly precise measurements are possible; refer to
"direct substitution", para. 3.12.
3. Mount the test fixture directly on the 2-TERMINAL
UN KNOWN port or use rigid coaxial lines, elbows, etc.
(Flexible cables change capacitance with position. If you
use them, your precision is only about ±0.1 pF.)
4. Support the fixture, if it's on a long, rigid line, to
avoid damage to the panel-mounted connector. Lock
each connector so it will not shift position.
5. Consider also the possibility of connecting your fixture to the 3-TERMINAL
UNKNOWN port; thereby
tak ing advantage of the fact that various cable and
ground capacitances are excluded from the measureruent,
NOTE
Refer also to NON-COAXIAL
CAPACITORS, para. 3.13.

3.5 BALANCE AND READOUT.

2-TERMINAL

Figure3-3.

This is the actual measurement process. Preceding paragraphs explain how to tune oscillator and detector, make
phase adjustments, and connect the capacitor to be measured. Balance the bridge, as follows, and you have the
measurement in both visual and digital forms. The standards for measurement are contained in the bridge.

3.5.1 Readout Multiplier.
Set the TERMINAL SELECTOR to the appropriate Xl
position (2 or 3 terminal) unless the capacitor being measured is larger than 1 nF (1000 pF). Then set the selector to
the highest READOUT MULTIPLIER (Xl0for3-terminal,
Xl00 for 2-terminal measurements).
The purpose is to utilize one of the three "pf " lever
switches for the most significant digit, if possible, thus
maximizing the accuracy of your measurement.

Cx

~

EXT MULTIPLIER

Gx

-z-

.2

UNKNOWN

TERMINAL SELECTOR
READOUT MULTIPLIER

-zr

0-+-

.40-+

BRIDGE, GR 1616

Figure 3-3. The 1621 Capacitance-Measurement System - simplified diagram. *Note: Effective values are
shown for all G standards and for C standards smaller than 1 pF.

3.5.2 Initial Settingsof lever Switches.
If the capacitor value is known, even approximately, set
that value into the CAPACITANCE readout, and set the
doubtful digits at the right to 5's.
For example, if the capacitor is about 12 pF, set the
bridge to read 12 pF 555 fF 555.5 aF.
C MAX. Move the C MAX lever down, covering as many
as 3 of the unused digits to the left of the desired readout.
Covering as many as possible of the zeroes at the left of the
final readout helps to maximize the precision of your
measurement, by removing stray capacitance which otherwise loads the ratio transformers slightly.

CONDUCTANCE. For a typical measurement, set the
CONDUCTANCE levers initially to the zero readout
J.LS
0000 nS X 10-6 . However, if conductance is known approximately, set the levers to that approximation.
Unknown, A Black Box. If the capacitance or conductance settings cannot be set initially to the right order of
magnitude, the MAGNITUDE meter will read off-scale.
Temporarily, use the following aids to obtain a preliminary

a

balance
a. Push the COMPRESSION button in.

b. Turn OUTPUT VOLTAGE RANGE and OUTPUT
ADJUST ccw as may be required to bring the MAGNITUDE meter on-scale.
c. Explore large ranges of C and G lever-switch settings,
leaving most of the levers at zero so that the ones being
moved are always most-significant digits.
d. Watch the phase-sensitive meters (even though they
may be pinned and not properly phase related). A reversal
of sense (pointer crossing midscale) is an indication of
passingnear a rough balance.

3.5.3 BalanceProcedures.

Figures1-2, 3-3.

GAIN. Set the GAIN control initially so the MAGNITUDE meter reads in the upper half scale. Later, as balance
is improved, turn the control cw to maintain useful readings
on the other 2 meters.
SENSITIVITY. Set the SENSITIVITY control initially
at midrange. Later, as balance is improved, adjust for a large
on-scale reading on one or both of the IN PHASE and
QUADRATURE meters.

OPERATION

3-5

NOTE
Keep the MAGNITUDE meter on scale, otherwise the phase sensitive detectors may be over
loaded. However, the IN-PHASE and QUADRA TU RE meters may be pinned without loss
of phase sense indication.

TIME CONSTANT. Set the TIME CONSTANT initially
to 0.1 s. for immediate response to your balancing operation.
Later, as balance is improved and the GA IN control has
to be set quite high, turn to larger time constants. A noisy
reading (meter pointer jumping) can be expected at high
gain if the time constant is too low. So increase it to calm
jittery IN-PHASE and QUADRA TU RE meters, allow time
enough for each reading, and you can achieve a highly
precise balance.
Lever Switches. Decide which levers to adjust first by
looking at the PHASE-SENSITIVE DETECTOR meters If
the IN PHASE meter reads farther from zero, adjust CAPACITANCE first if QUADRATURE,
adjust CONDUCTANCE.
CAPACITANCE balance. Refine the balance as follows,
left-to-right
a Start with the most significant digit that is in doubt.
Preferably, every CAPACITANCE digit to the right of it
should be set to 5.
b. Push the lever down if the IN PHASE meter reads up
scale; pull up if it reads down scale. Continue until the
meter pointer crosses zero.
c. Leave that lever in the position closest to balance, i.e..
making the IN PHASE meter point closest to zero, on
either side.
d. Adjust the next lever to the right, similarly, if there is
sufficient meter deflection to be a guide. (I f the suggestion
in step a is used, you will probably not have to readjust the
lever you set in step c.)
e. Increase the GAIN as much as necessary to provide
that deflection, but not enough to deflect the QUADRATURE meter off scale.
f. Continue refining the capacitive balance until the
QUADRATURE
indication becomes a limitation, then refine the CON DUCT ANCE balance.
CONDUCTANCE balance. If an approximately correct
conductance setting has been initiated, refine the conductance balance from left to right, similarly to the capacitance
balance. Minimize the reading of the QUADRATURE
meter.
However, if the CONDUCT ANCE levers have been set to
zero, as described above, proceed as follows
a. Raise the CONDUCTANCE levers from 0 to 5, starting at the left of the multiplier and continuing left until
either the QUADRATURE
meter pointer crosses zero (if
so, skip to step c) or all these levers have been raised.

3-6 OPERATION

If the meter points to the right, drop the
significant lever to -1, so the readout is -11J.S 555.5
1(f6 , and skip to step c.
If the meter points to the left or zero, drop the
significant lever to + 1, so the readout is 1 1J.S555.5

mostnS X
mostnS X

1(f6.

b. Raise the conductance multiplier lever until the meter
crosses zero and stop at the setting nearest balance. l f this
lever is set to X1 or x io '. PHASE SHIFT will have to be
reset, as described below.
c. Refine the conductance balance as outlined above (for
capacitance) starting with the left-most (1J.Sj lever and
proceeding to the right.

3.5.4 Final Balance.
Alternate
between CONDUCTANCE
and CAPACIT ANCE balances, as described above, improving whichever
is worse until the other becomes a limitation.
Conclusion. Stop when the desired precision has been
achieved. The bridge is probably capable of greater resolution than you need.
Voltage. If there seems to be a gain or noise limitation,
even though GAIN is 130 dB and TIME CONSTANT is 10
s, raise the oscillator OUTPUT VOLTAGE
However, do
not exceed the level of 350 V rms. If the frequency is less
than 2200 Hz, do not exceed 0.16 f volts, where f is
frequency in Hz.

&
WARNING
While generator is set to a high level, beware of
hazardous voltages: ~ 350 V across 3-TERM
UNKNOWN or EXT STD; ~ 35 V across
2-TERM UNKNOWN (outer shell high).

Phase Shift / Fine Adjust. While making a precise measurement of a low-loss capacitor, check as follows
a. As the 4th significant digit of C is being selected,
verify that the QUADRATUR E meter responds negligibly
(compared to the IN-PHASE meter) for Clever changes
b. Correct the phase if necessary with the QUADRA
TURE FINE ADJUST control
c. When balancing conductance, make the corresponding
check (and adjustment) of the other phase.
NOTE
A slight change in tuning of either oscillator or
detector will affect phase, even though the
change has negligible effect on magnitude of
response.

Phase Shift / Reset. If the capacitor is lossy, the phaseshift settings of paragraph 3.3 are unsuitable. Reset as
follows
a. Make an initial balance, watching the MAGNITUDE

meter, until the 4th significant digit of C or G has been
chosen (whichever predominates).
b. Temporarily turn SENSITIVITY and GAIN controls
ccw enough to bring both phase-sensitive meters within %
division of zero (scale value of 2.5). Set both FINE ADJUST controls to mid range (pointer up)
c. Unbalance the bridge by raising C levers (or G if it
predominates) enough to make large deflections on these
meters.
d. If C predominates, adjust PHASE SHIFT for zero
QUADRA TU RE and upscale IN-PHASE meter readings.
Raise the GAIN and SENSITIVITY settings as is appropriate; refine the phase with the QUADRATURE FINE ADJUST control (as described under Phase Shift / Fine Adjust).

If G predominates, adjust similarly, for zero IN-PHASE
and upscale QUADRA TU RE meter readings.
e. Return to the initial-balance condition, then unbalance G (or C) The meter that read upscale in step d should
not deflect. Trim the nearest FINE ADJUST control for
this condition.
NOTE
Be sure the EXT MULTIPLIER switch is OFF
(not simply zero) whenever the EXTERNAL
STANDARD terminals are unused.
Even though those terminals are open, their stray capacitance may introduce error into the measurement of capacitors below 1 fF and the unshielded LOW terminal may pick
up enough "noise" to have a detrimental effect on precision.
NOTE
Keep the MAGNITUDE meter reading on scale.
If the MAGNITUDE meter is pinned, distortion in the
detector may cause a spurious phase shift. However, the IN
PHASE and QUADRATURE
meters may point off scale
without loss of sense information.

output data (rear panel) that data is now in standard form.
Refer to para. 2.8.

NOTE
There may be occasions when it is desirable or
necessary to leave X in the readout.

If, for example, the bridge balances at ** nF X23 pF
456 fF 785.5 aF (the last 2 or 3 digits may be insignificant!
you have this choice For greatest accuracy and a record as
to which internal standard is significant, leave the X in the
readout Take the responsibility for making sure the recorded data is unambiguous (particularly if a printer is being
driven from the BCD CAPACITANCE OUTPUT) For a
simpler readout, move the C MAX lever and correct the
readout to *1 nF 023 pF 456 fF 785.5 aF. But then the
bridge is almost certainly unbalanced; if you now refine the
balance you are almost certainly reducing accuracy by
depending on a larger, lessaccurate internal standard

3.5.6 Units of Measurement.
The units of each, Cx and Gx, appear on the front panel;
but be sure to apply the multipliers. For example, (Figure
1-2) if the readout is X98 nF and the READOUT MULTI
PLiER indicates XlO, C; is 1098 nF or 1098 fJ.F
The conductance readout has 2 multipliers. For example, if the readout is 5.43 p..SX 10-6 and the READOUT
MULTIPLIER indicates Xl00, c, is 5.43 X 10-4 us or
5.43 X 10- 1 0 S

NOTE
The symbol "S'" represents the Siemens unit,
the unit of conductance (or admittance).
Siemens is equivalent to mho (the reciprocal of
ohm). represented by the symbol U.

3.6 PARAMETERS OF THE UNKNOWN CAPACITOR.
3.5.5 Readout Correction.

Figure 3-3.

If there is a -1 or X in the CAPACITANCE or CONDUCT ANCE numbers displayed above the lever switches, it
is generally recommended to correct the readout as follows
a. Proceed from left to right unless X is adjacent to -1,
in either order. (Correct the right-hand digit first in any
such pair.)
b. Change each -1 to 9 and decrease the preceding digit
byone.
c. Change each X to 0 and increase the preceding digit
by one.
d. Verify that the final balance is still valid.
e. The readout is now correct for manual recording.
Also, if your measurement system makes use of the BCD

3.6.1 SeriesEquivalent Parameters.

Figure 3-4.

This bridge always measures directly the parallel equivalent parameters Cx and G x ' even in case the loss component of the unknown capacitor is in fact entirely in series
(as resistance of the connecting leads). Whether a series
R-C, a parallel R-C, or a more elaborate equivalent circuit
best describes the capacitor you are measuring cannot be
decided on the basis of one measurement. At the very least,
measure at several frequencies; there is further discussion in
Section 4.
Here in this paragraph is a group of formulas and diagrams relating the parallel equivalent parameters to the
series equivalent parameters. Notice that D, Q, and ware
equally valid members in either set of parameters.

OPERATION 3-7

Gx
PARALLEL

SERIES

Figure 3-4. Equivalent circuits of the unknown capacitor.

CSE~ C,[1,( WGC',)}
C,11'0'1
R

G2+W2C2
x

where C
G
R

3.7 FREQUENCY.
3.7.1 Setting the Frequency.

Gx

=----S

x

To make measurements at any frequency in the range 10
Hz-lOG kHz, set the oscillator FREQUENCY selector dials
and range switch so that they read as desired. Their indicators together make a convenient, in-line readout.
The 3rd-d igit selector is a continuous control (although
the readout indicates only 10 positions) From the detent,
at the zero position, this control can be rotated cw, through
the indicated decade range, or ccw for a small negative
extension of that range. Observe the limitations on high
voltage at high frequency as described below.
In brief, use the following procedure to set the frequency of measurements by the 1621 Precision CapacitanceMeasurement System
a. Set the oscillator FREQUENCY controls.
b. Set the detector FREQUENCY controls to match.
Without balancing the bridge, peak the response of the
MAGNITUDE meter by fine tuning of the oscillator.
c. Make the phase adjustment as in paragraph 3.3.

G x(Q2+1)

capacitance in farads
in Siemens
= resistance in ohms
w= 21Tf
f = frequency in hertz.

= conductance

3.6.2 Dissipation Factor.

NOTE
Unless your measurement must be made at exactly 1.00 kHz (for example) set the oscillator a
little above such a round number. Doing so will
simplify the tuning of detector and oscillator to
the same frequency.

Figure 3-5.

Dissipation factor or loss tangent 0 (the reciprocal of
storage factor Q) is defined above. Refer also to para. 4.2.
o is presented in convenient, graphical form in Figure 3-5,
which is used as follows
a. Find C x on the appropriate scale, if the measurement
frequency is one of those given. Otherwise, notice that the
horizontal scale is really the product fC x The higher your
frequency, the farther right your capacitance scale shifts.
(For example, at 3 kHz, imagine the l-kHz scale shifted to
the right nearly half a decade; if your C; = 8 pF, locate 24
pF on the l-kHz scale.) Imagine a vertical line through that
point, representing fC x
b. Find G x on the slanted scale. I magine a slanted line to
represent G x .
c. Find 0 on the vertical scale, directly left of the
intersection of your fC x and G x lines.
d. For greater precision, use the expanded partial chart
(but determine order of magnitude from step c) For still
f C;
higher precision, use the formula 0 = G x / 21T
If the range of bridge measurement is extended by the

3-8 OPERATION

use of an external standard, extrapolate the corresponding
scale. (For example, an external standard of conductance
may be advantageous for measuring large, lossy capacitances at high frequency. Add more slanted lines in the
upper right portion of chart to represent such an extension.)
The shaded square in the main chart and that of the
expanded partial chart represent the same range of data. A
visual impression of the magnitude of this expansion helps
you use the partial chart.

3.7.2

Monitoring

Frequency.

a. Connect a counter to the EXT SYNC connector (rear).
b. Select PERIOD measurement (1 IJ.sTIME BASE) and
10 PERIODS AVERAGED on the counter.
For a frequency near 10 Hz, most or all of the counter readout is
now utilized. The most significant digits will spillover to
the left as you proceed; remember them as they disappear.
c. Increase the PERIODS AVERAGED,
turning the
control cw as far as is necessary to obtain the desired
resolution of measurement.
d. Calculate frequency, the reciprocal of the period just
measured.

10. 6
10- 5
10- 4
10- 3

o

2
10

10. 2

0::

o

10.

10

f-

u

«
u,

'
Q

z
o

~

10.

o

10. 2

en
en

10

'

10
10

2
3

4

10

10

10

100 of

I

I

10

100fF

10

100fF

10

100pF

100aF

I

(f = 120)_

0.1

10 100 of

10

= 10 ) -.. O.I

10

100aF
I
CAPACITANCE

CONDUCTANCE G

2

o

-,

~

I

a:: 8

o
f-

~ 6
u. 5

z
o 4
f-

Ci 3
(f)
(f)

Ci 2

~~
~~

'",

""

-,

I _

.'-'

"

<,
,,"-.

-,

-,

,"-.

<, -, -,

(1=1000)
(1= 120) ~

-,

1

1

2

t
CAPACITANCE

3
2

c,

I
100nF

I
I

10fLF

I

"

14 5

6

3

5 6

4

FOR CERTAIN

-,

" <," -,-,

"-:\~

-,
"<,
-,
"-.
~ ~~ -, -, ~ -, ,,1'\
6'

I
10

-,

-:
-, -,
-, <,

"-

i\.'\

"-.

~

I
I

100nF

<,

'" "'"
<, -, t\.'
I"-.'", ",,,,~
-,
"-"
, """-, "-"
<,

I

I
100pF

10

-, "',

~ -,

.:'\

.

i

10

~

,"

".'"
-,
"
f« ,,"-. ."-

i

100pF

SCALES)

~ -,"0-~~ "-,'\

"

~9

(SLANTED

"-\:'"

106
10 fLF

10 100fF
I
10 100pF
I
10 100nF
C x • AT 4 COMMONLY USED FREQUENCIES

'9"S'6',\:'~

,f

10

100fF

I

j

(f

100nF

i

i

10

10

5

8

,,~

Q

3

4
5
6
8

12

1
8

-,

2

1

I
2

FREQUENCIES
161628

Figure 3-5. Dissipation factor vs directly measured parameters. Above: entire range of bridge with internal standards.
Left: expanded partial chart, used for greater resolution. For
any arbitrary frequency, enter C x scale at correct product fC.
(Example: if 1=60, C x=6 pF, use the scale for "1=120" but
read "C x=3 pF"1.

OPERATION 3-9

3.7.3 Locking to a Frequency Reference.
If the reference frequency is known within ±1%, and the
reference source can drive the 27-kQ EXT SYNC circuit
rms, proceed as follows
with 1-10
a. Connect the reference source to the EXT SYNC jack.
b. Tune the capacitance-measuring system to the reference frequency, setting the phase shift as described before.
However, if the reference frequency is not known to
±1%, or its signal level is < 1 V rms, follow this procedure
a. Using a tee, connect both the frequency reference and
a scope to the EXT SYNC jack.
b. Vary the 1316 Oscillator frequency over a range of
30% or so, observing the scope. It should be possible to see
both varying-frequency and fixed-frequency components in
the waveform. (Synchronize the scope, preferably, to the
reference. )

v

c. Tune the oscillator through lock, a condition in which
the waveform "stands still". Determine the range over
which lock can be maintained, and set the FREQUENCY
dials to the center of that range.
d. Tune the detector and set the phase shift.

3.8 VOLTAGE LEVEL.
Selection. The source voltage is selected by the 2 controls,
OUTPUT
VOLTAGE
RANGE and OUTPUT
ADJUST, on the 1316 Oscillator. Read the front-panel
meter, full scale being the number indicated by the OUTPUT VOLTAGE
RANGE switch. Another source can be
used instead; but always select a level within the limits
described below, and never more than 350 V rrns, max. *
Frequency Dependence. Select a source voltage ~ O.16 f
volts rms, to avoid saturation in the bridge ratio transformer. For example, at frequencies of 0.1, 1.0, 10, & 100
kHz, maximum levels are 16, 160,350, & 350 V rms.

NOTE
Voltages at the unknown capacitor, greater
than this limit, can be obtained by using the
reversed configuration of para 3.15

Large capacitors. You can resolve large C at high frequency using relatively low voltage. Appropriately,
levels
available in the 1621 system are then limited by oscillator
loading. Normal operation is below these limits level < 4 X
10-3 / fC volts rms for OUTPUT VOLTAGE RANGE = 150
V; level < lef 2 / fC for lower ranges Example for 01 flF
at 100 kHz, select 10 V max, on the 15-V range
2-TERMINAL
Xl. Use 20 V rrns, max, at this one
setting of the TERMINAL SELECTOR Otherwise, clipping
circu its designed to protect you from a shock hazard will
cause an error in your measurement.
Resolution. The chief reason for using a high voltage
level is to facilitate balancing the bridge to a high resolution. Generally, make preliminary balances with 15 Vor so.
*500V

Then switch to 50, and finally 150 V, as you refine the
measurement beyond the 6th significant digit. Do not ex-

ceed the voltage limits described above.
'

200

400
600

200

600
400

1000 ppm
800

10

IfF +>-l..I-J.-t.Ll..I-J.-t.Ll..I-J.-t-'-'L.w..-t-'----+-Wu...L+-Wu...Lf-Wu...Lj-l-'-'-'-j-l-'-'-L.i
o
10
20
30
40
STEADY-STATE

TEMPERATURE -

50

DEGREES CELSIUS

Figure 3-7. Worst-case measurement error due to temperature Ibased on maximum temp
coaf of C stdsl. Basic specifications apply in shaded band 123°C ±1°CI. In remainder of
chart the error may add lin the worst easel to the tolerance given in Figure 3-6.

the readout were ** nF 103 pF 456 fF 789 aF, then the
measurement would be 103.4583 pF, +.0011, -.0013pF
NOTE
If the bridge is to be used regularly at a temperature other than 23°C, it can be calibrated for
such use. Refer to para. 5.4. For temperature
rise in 1621 System, refer to Specifications, in
front of manual.
3.9.3 Rangeand Dissipation-Factor Limitations. Figure 3-5.
G-Range Limits. The accuracy of a Cx measurement may
be limited by an extreme of Gx (or vice versa) For example, if both Cx and frequency are very large (near the limits
specified for the bridge) and the capacitor is very lossy, the
largest internal G standard may not be large enough for
balance. Similarly. a very small Cx with very low losses
measured at a low frequency, may require smaller G-standard steps for the desired resolution than the smallest in the

bridge. The simplest way to extend the range is by the use
of external standards. The range limits are represented in
the figure by the ends of C; and Gx scales.
Dissipation Factor. Figure 3-5 covers somewhat more
range of D than can easily be evaluated by this bridge. For
moderate values. all the parameters discussed in paragraph
3.6 can be determined readily. However. at extremes of D
(or QJ.only the predominate admittance can be measured
accurately.
Very low D. At about D ~ l(fs the unknown capacitor
has lossesas low or lower than those of the internal stand-

3-12 OPERATION

ard capacitor. If. for example, D = 10-7 , you cannot measure G; directly by the CONDUCTANCE readout. and so
cannot measure D. However, this situation does not limit
the Cx-measurement accuracy, because any decade of the
CONDUCTANCE standards can be set to 1.
C-Accuracy with High D. BecauseG standards have some
stray C, C accuracy is sacrificed when much G has to be
balanced. The error can be + or -, depending on certain
compensations. Calculate accuracy as follows:

where Do is the ratio of conductance to l-k Hz susceptance
of the G standards. and is 103 or more for this bridge.
Comparison accuracy depends on the G standards that
are changed between measurements. Substituting b.G for G
in the formu la. comp acc = ± b.G / (21T103 CD o).
For example. if C 1 = 12.3456 pF, G 1 = 1.239 X l(fl
f.lS and f = 500 Hz, then D = 3.2 and accuracy is ±1600
ppm. If a similar capacitor has C2 = 12.3123 pF and G 2 =
1.241 X l(fl f.lS, then comparison accuracy is ±13 ppm.
Seealso para. 4.7.
3.9.4 Shunt Capacitancesto Ground.

Figures3-1, 3-2.

These capacitances, Cis' Chs' and Chg are excluded from
the measurement of Cx. However, if large, they can reduce
accuracy and precision (particularly at high frequency).
If you can keep these capacitances below 200 pF each
(unknown capacitor directly connected to bridge terminals)

or 100 pF plus cable capacitance (unknown capacitor
connected by a pair of G874-R22LA
patch cords)
measurement accuracy is unimpaired, even at 100 kHz.

Three-terminal Measurements. If you must measure in
the presence of larger shunt capacitances to ground, calculate a bound on the probable error as follows
C readout = Cx - w
Greadout = G

2

ChsCls Lsg

2

w ChsClsRsg

ground, the error from ground capacitance should be negligible.
Coaxial 2-terminal measurements differ in this aspect
because the LOW bridge terminal is cornpletelv shielded by
the HIGH Capacitance from HIGH to qround does not
affect accuracy except for loading of the ratio transformer
That effect is negligible for C hg < 100 pF or 0.1 C x
(whichever is greater) However, a distinction must be made
between C hg and fringing capacitance; see below.

3.9.5
Notice that the error terms are both negative. Cis is
low-side-to-shield capacitance; Chs' high-side-to-shield. Include cable capacitances, if patch cords are used. RSg and
LSg are, respectively, the resistance and inductance of the
path from a virtual common point in the shield surrounding
the unknown capacitor to the ground point in the heart of
the bridge. Include the Rand L of the patch-cord outer
conductors (2 in parallel, if 2 patch cords are used) in R Sg
and LSg'
As Figure 3-1 shows, our error calculation is based on a
lumped-parameter situation. If cables are used, C, L, and R
are really distributed along the cables, and the corresponding error terms are even less significant than those
calculated above
For example, assume shunt capacitance of 45 pF at each
end of the unknown capacitor, in addition to 90 pF for
each G87 4- R22 LA patch cord. Assume 1 /LH and 10 mD for
for the series-impedance parameters of each cable and its
associated connectors Assume Frequency is 100 kHz. Then:

Lsg = 5 X

Hr7

henry

Rsg = 5 X W- 3 ohm

G error = w 2 Chs Cis RSg = 3.6 X 10-11 Siemens.
If C x = 1 pF, compare this C error with the normal
error, 5 X 1(JIS F. (Refer TO Figure 3-6; measuring 1 pF,
we expect 500o-ppm accuracy.) Notice that most of that
error is due to series inductance in the bridge (see para 4.7)
and is accentuated by any inductance of the cable inner
conductors. At lower frequencies, with very long cables,
the series inductance thus usually contributes more error
than the ground-return inductance.

Two-terminal Measurements. Non-coaxial 2-terminal
measurements can be considered a special case of 3-terminal
measurements in this aspect. If neither terminal has more
than 200 pF to nearby instrument or bench "ground", and
if that ground is connected by low impedance to the bridge

Fringing Capacitance.

Fringing capacitance is significant in coaxial 2-terminal
measurements. (The HIGH and LOW connectors for 3terminal measurements are far enough apart so the fringing
between their inner terminals is only 125 aF.) As explained in
paragraph 4.2, the recommended setting of ZERO ADJUST
makes the balanced-bridge C readout express just the
capacitance C due to fields beyond the reference plane of the
x
G900 connectors, entirely within the "unknown" capacitor
and its connector.
As described in para. 3.4, you may
assume the fringing is 0.11 5 pF and be correct within ± .008
pF, i e., 8 fF. Two methods are suggested for greater
accuracy, as follows.

Standard Capacitor,Use a Coaxial Capacitance Standard
(2 pF) or (1 pF) with calibration specified within ±5 fF, or a
G900-W04 Precision Open-Circuit Terminal (2.670 ± .0067
pF). You may have one calibrated to greater accuracy by the
National Bureau of Standards.
a. Set front-panel controls as of the completion of functional checks and phase adjustment, para. 32 and 3.3, at
the desired frequency, para. 3.7.
b. Install the coaxial capacitance standard and set the
CAPACITANCE lever switches to its value.
c. Set the TERMINAL
SELECTOR to 2 TERMINAL
Xl, XlO, or X100 depending on whether you wish to
measure capacitors up to 1 n F, 10 n F, or 10 J.1.F,respectively. Keep the C MAX switch down, during this procedure
(and during measurement of your unknown capacitors except when their values require large C standards with a
READOUT MULTIPLIER of Xl00)
d. Balance the bridge, using the ZERO ADJUST control
for C and the CONDUCTANCE lever switches for G Turn
the GAIN and SENSITIVITY controls as required for suitable indications of balance.
NOTE
The resolution of ZERO ADJUST settings is
about 0.5 fF for X 1 and X 10, about 5 fF for
Xl00 READOUT MULTIPLIER.
e. If you want to obtain a more exact value for the
fringing of the G900 Connector on your bridge (for use in
para. 3.4) remove the capacitance standard and rebalance the
bridge using the CAPACITANCE lever switches. The

OPERATION 3·13

repeatability is about ±1 fF, verify
above procedure a few times.

this by repeating the

Direct Measurement. Calibrate a coaxial 2-terr'ninal
capacitor against the internal bridge standards. If you prefer.
by the method of Millea'* (Refer to Appendix A.)
This method requires the capacitor to be measured twice.
while attached to a fixture made with a tee. suitable
connectors. and 2 lengths of flexible cable. The fixture is not
readily available; details are left to your ingenuity.
3.9.6 Conductance Accuracy.
The accuracy specification is ±O.1% ± 1 step in the least
significant digit, particularly for G :::=10- 1 ] S However,
that last step need not be a limitation at larger conduc
tance. If you avoid making the first digit of the CONDUC
TANCE readout a zero
G readout = Gx ±O.1% Gx, for Gx;;;;'l cr S p.S

Offset. Measure Go and use the correction, as follows.
a. With EXT MULTIPLIER
OFF, TERMINAL SELEC
TOR at CAL, and readouts initially at zero, balance the
bridge
b. The G readout is G,,; it rnay be + or -, with magnilcr 8 p.S This offset is unaffected by
tude (typically)
ZERO ADJUST
c. For any measurement
(typically if the G multiplier
from the readout, thus

G;

in which Go is significant
is lcr s or lcr 6 ) subtract it

= (G readout ± 1 in 5th window

o':

coefficients
.

are

-700,

-300,

±50, ±50, and

Of course, you can calibrate the bridge at a temperature
other than 23°C, making use of the "±15 ppm" internal
standards or any suitable external standards for references
Refer to para. 5.4.
The internal conductance standards respond to environmental temperature change with a srnall time constant (a
few minutes) in contrast to the large capacitance standards.
C-Range Limits. As discussed above (conversely), G x
accuracy may be lirnited by the available ranqe of internal C

'Millea, Aurel, "Connector
Pair Techniques
for the Accurate Meas·
Journal of
urement
of Two-Terminal
Low-Value
Capacitances,"
Research, 3, of the National Bureau of Standards, Vol 74C, Nos
3&4,

.Julv-Dec , 1970.

3-14 OPERATION

Capacitance to Ground. As explained in the Cxaccuracy
discussion, G, accuracy may be affected by the ground
capacitance and series impedance from the shield around
the unknown device through the cables to the bridge
ground. I f you must use long cables, and particularly if you
need to test at high frequency, use the formula given there
to esti mate the consequent error

Go) ± 01%

Temperature. The temperature effects on conductance
accuracy are nearly negligible. The basic specification of
±1000 ppm (see above) applies over the temperature range
of 23 ±l°C
For further temperature changes, the worst
case obtains when the CONDUCTANCE multiplier is set to
1cr 4 , 1
or 1cr 6 , then the readout at balance varies
inversely with temperature at the rate of 700 ppmtC
For
settings of 10-4 , 10-], 1cr 2 , 1cr 1 , and 1, respectively, the
temperature
±15 ppmtC

standards. For example, if G;
10- 1 ] S ?"d D = 10 at f =
105, the smallest internal C standard is not small enough
for a precise balance. (Stray C in the G standards has to be
cancelled, so we are not able to measure Cx anvwav.) If Gx
= lcr 4 Sand D lcr 2 at f = 120, the largest internal C
standard is not large enough for balance. Use an external
standard, if you want to ex te.id either limitation.
Dissipation Factor. At very large D, the unknown con
ductance may have less capacitance than the internal stand
ard conductance. Then you cannot measure C; directly and
so cannot measure D But (in contrast to small-D limits)
this situation is generally no limitation on C;x measurement
accuracy. C standards can be set negative if necessary for
balance.
At very low D, the accuracy of G, measurements is
limited because a significant part of the conductance in the
standard arm of the bridge consists of losses in the internal
capacitance standards. It is therefore impractical to measure
G; in the region near or below the bottom edge of the
chart Figure 3-5. If the low-loss Cvstandards are used (C; <
1 nF) this G, accuracy limitation is ~'Yr) at D io '. 10e;;" at
D = 10-4 , and about 100 e;;,) at D 10- 5 .

3.10 PRECISION.
The comparison precision of the 1621 system (or the
1616 bridge if set up in an equivalent system) is specified to
be ±01 ppm (one part in 108 ) for low-loss capacitors
between 10 pF and 10 pF This statement means that the
system has the resolution, sensitivity, stability, repeatability, and operating convenience necessary for you to
balance capacitance to ±01 ppm.
Such precision is significant in measuring capacitance
changes, comparing capacitors, adjusting and evaluating
capacitance standards, etc. The precision is available for
measurements described in preceding paragraphs, but far
exceeds the absolute accuracy to which we can guarantee
the many possible combinations of internal standards.
Resolution.
The smallest internal standard step is
l 9
lcr
F, so resolution of ±.01 ppm extends from C x
1cr l 1 F (10 pF) to 1cr 5 F (10 pF) without external standards. An extra decade, at one end or the other of the range,
can be provided by a suitable external standard.

Sensitivity. You can balance the 1621 system to ±.01
ppm because (among other things) you can see aOl ppm
unbalance on the detector. This remarkable sensitivity depends on instrument capabilities, but also on your selection

of a high source voltage, proper tuning of the detector, and
setting of the phase shift, as described before.
Because source voltage must be kept below a limit proportional to frequency, sensitivity is generally adequate for
the above-mentioned resolution only at frequencies above
about 900 Hz. These somewhat arbitrary limits obtain: at
10, 102 , 103 , 104 , and 105 Hz you can resolve to ±.01 ppm
when c,
10-9 , 10-1 0 , 10-1 1 , 10-1 2 , and 10-1 2 F, respectively. (To realize the 1

Stability. You can enjoy the above-mentioned precision
because (also) the internal standards which determine the
most-significant digits of any such precise measurement are
sufficiently stable. They are stable in terms of mechanicai
shock, aging, atmospheric changes, ambient temperature,
and other factors.
Consider temperature in more detail. These standards are
thermally isolated, with a time constant of 6 hours. Assuming you can make a comparison (2 measurements) in 6
oC
minutes, the effect of a l_
ambient temperature change
on that comparison cannot exceed 1/60 of the specified
temperature coefficient. That is 3 ppmtC for C; in the
range 1-1000 pF, so the effect is .05 ppm. Therefore, for
±.Ol-ppm precision under these conditions, be sure the
ambient temperature is regulated within ±0.2°C.
NOTE
Because of the long thermal time delay in the
bridge, it must be held at the desired temperature a long time before precise measurements
can be made Refer to para 4.5 for a quantitative discussion.
For example, a 12...?Cchange, 24 hours before, will
disturb a comparison taking 6 minutes as much as a 0.2°-C
change during the comparison.
Take reasonable care to avoid mechanical shock, regulate
temperature as described above, and you can rely on the
stability of the internal standards, for C; about 1-1000 pF
However, the internal 10 and 100-nF standards, used for C;
in the range of 2-100 nF, may have as much as an order of
magnitude larger temperature coefficient.
To obtain .01 ppm precision using them, make more
rapid comparisons, provide better temperature regulation,
or use an unusually-stable external standard as described
below under the heading of Externally Determined Accuracy.

3.11 EXTERNAL STANDARDS.
The 1616 Precision Capacitance Bridge has considerable
versatility to make special measurements and comparisons
in addition to those already described. These paragraphs
deal with the use of external standards to extend range,

3.11.1 RangeExtensionto 111 [.1F.
Accessories. A 1-JlF. highly stable. external standard. i e..
GR 1409- Y Standard Capacitor and two G874- R33 patch
cords.
Connections. Connect the capacitor to the EXTERNAL
STANDARD port of the bridge, thus capacitor H to bridge
HIGH, capacitor L to bridge LOW, and capacitor G to the
shield of the LOW cable. Connect the unknown capacitor
to its port as usual.
Calibration. Verify the calibration of the external standard with adaptors and cable at the frequency you plan to
use, by the method of para. 3.4 and 3.5, or as follows set
EXT MULTIPLIER to -0.1, TERMINAL SELECTOR to
CAL; balance the bridge; and multiply the CAPACITANCE
readout by 10.
Balance. Measure the unknown capacitor as usual except
that you now have a more-significant C digit than before,
controlled by the EXT MULTIPLIER switch.
Readout. If the external standard is 1 [.1F,with sufficient
accuracy for your measurement, and if the series-inductance error described below is tolerable, interpret the EXT
MULTIPLIER setting multiplied by 10 as an extra digit at
the left of the usual C readout. (Example ext std value
1.00011 [.1F;EXT MUL T 0.6; readout 54 nF 321 pF .
READOUT MU LT X 100; desired accuracy ±0.1 %. Final
value is 65.4 [.1F.)
But if v.ou want greater accuracy, the measurement is
[Lexternal standard C value) (EXT MULTIPLIER)
+
CAPACITANCE readout] (READOUT MULTIPLIER)
(series-inductance correction). In the same example
(1.00011 X 0.6 + .054321) X 100 = 65.4387 [.1F,without
the correction (see below). In this example, that might be
.05 /IF. Unless you can refine that, only 2 decimal places
are valid. Then the final measurement is
65.44-.05 = 65.39 [.1F
Similarly, the final G readout is [(external standard G
+ (CONDUCTANCE readout)
value) (EXT MULTIPLIER)
(CON DUCT ANCE
multiplier)]
(READOUT MUL TIPLlER).

Series-L Correction. The accuracy of measurement of
large capacitance is usually Ii mited by the bridge and lead
inductance in series with the capacitance. The bridge reading of capacitance is greater than the unknown capacitance
C; by a capacitance error tJ,C = ",} C2 Q. If bridge induc-

OPERATION 3-15

tance Q is about 0.3 pH, in series with the UNKNOWN
terminals, you have a minimum error of the order of
+0.002% C}.If (f k H z ) 2 Hence, at 1000 Hz and 100 pf, the
bridge reading might be high by about 0.2%. Refer also to
para. 4.7.

3.11.2 Extension of C Resolution.
Although the resolution of the bridge is exceptionally
fine, you can make it still finer with a sufficiently small
external capacitance standard. Thoroughly shield anything
connected to EXTERNAL STANDARD port, particularly
on the LOW side.

greater accuracy than the internal standards permit, and if
you have suitable external standards, use this procedure.
(For example: the bridge may not yet be stabilized at room
temperature.) Alternatively,
you may wish to measure
against a standard deliberately set a few ppm different from
the internal standards.
Comparisons between external standards of the same
nominal value, or values differing by convenient ratios, like
1, 2, or 10, can be made to great precision (para. 3.10). Use
one as "standard", one as "unknown."
As a check on
possible zero offset, interchange them and compare again.
Accessories. A

Accessories. A well-shielded O.l-aF (i.e., 10-\ 9 F) 3-terminal capacitor; well-shielded cables, such as 874-R22LA.
NOTE
If you make such a capacitor, use button-sized
plates spaced several cm either side of a shield
with a small pinhole in the center.

Zero Adjust. If the measurement is to be made at the
3-TERMINAL port, shield the 3-T LOW connector using an
open-circuit termination; set TERMINAL SELECTOR to
3-T Xl, C MAX lever down, and the entire C readout to
zero. Balance the bridge with ZERO ADJUST and the G
lever switches. [If the measurement is to be 2-terminal,
omit the shield, install a coax ial C standard, set TE RMINAL SELECTOR accordingly, and balance the bridge
with the C readout fixed at the value of the C standard.)
Calibration. Measure the external standard at the EXTERNAL STANDARD port. A balance should obtain with
C readout set to 0.1 aF, EXT MULTIPLIER to 1.0. [If the
measurement is to be 2-terminal, set the C readout 0.1 aF
above the value of the C standard.) Use a high frequency
(10-100 kHz) and high oscillator level. Measurement precision of ±1O% is probably adequate. (Setting the EXT
MULTIPLIER
to 0.9 should noticeably unbalance the
bridge) If possible, adjust the external standard to OlaF.
Balance. Connect the unk nown capacitor to its port.
Measure as usual except that you now have a less-significant
C digit than before, controlled by the EXT MULTIPLIER
switch.
Readout. Interpret

the EXT MUL TIPLI ER indication

multiplied by 10 as an extra digit at the right of the usual C
readout. (Exarnple ext std value 0.1 aF; EXT MU LT 0.6;
readout 234 aF 5; READOUT MUL T X 10. Final value is
2.3456 fF)

3.11.3 Externally Determined Accuracy I Comparisons.
Applications. If you want to make a series of measurements with the convenience of direct readout but with

3-16 OPERATION

suitable reference, i.e.. GR 1404-A,
and 2 of G87 4- R22 LA patch cords.

Connections. External standard capacitor to EXTERNAL STANDARD port; H to HIGH, L to LOW. Unknown
capacitor to its port.
Balance. Measure as usual, except be sure to control the
most significant digit by the EXT MULTIPLIER switch.
Notice which C-Iever switch (call it S) corresponds in magnitude to the EXT MU LT. Set "S" and all at its left to zero.
For good external determination of accuracy, EXT MUL T
should be set high, and one or more levers right of "S", low
(preferably zero).
(Example Ext std 100 pF; "S" is the lO-pF lever,
regardless whether READOUT MU LT is 1X or higher.
Note "S" set to X would represent 100 pf . and so would
EXT MUL T set to 1.0.)
Readout. Read out as usual except that EXT MUL TIPLiER reading multiplied by 10 must be substituted for the
zero of lever switch "S." (Example, as above, also EXT
MUL T 0.8; READOUT MUL T lOX; C readout 123 fF
456 aF. Final value is 801.23456 pF.)
Accuracy. Estimate the. C and G accuracies by adding
these error contributions for each.
1. Error in calibration of the external standard capacitor,
ppm.
2. Error of internal standards as they apply (refer to
para. 3.9) express temporarily as C or G (not ppm) and
multiply by the indicated READOUT MULTIPLIER.
Express result as ppm of the "final value".
3. Error of ratio transformation. Typically
error is <
for ratio

3
10

10
100,

ppm

where the ratio is the product of EXT MU LTI PLI ER
and READOUT MULTIPLIER settings. Ratio of 1 must
be (1.0) (Xl) not (0.1) (XlO)
4. Error of zero offset. Bridge reading at balance with
both ports open but shielded (use a G874-W) Termina-

tion on each LOW connector), EXT MULTIPLIER at 1.0,
and READOUT MULTIPLIER at Xl. This error should be
zero.

3.11.4 Test-Fixture Compensation.
Applications. You may have a test fixture, leads, terminals, etc. which have parameters included in C x and Gx but
which you want to balance separately so that the bridge
readout is only the additional C and G of capacitors connected to the fixture.
Notice that ZERO ADJUST performs just this function for
2-TERMINAL measurements only, for C only, and for a very
limited range, but with great precision.
(Its function,
specifically, is to balance the capacitance of the G900
connector on the bridge.)

mulae for C and G are both given there. The "external
standard G value" is 100 JLS.
NOTE
External standards of C and G can be used
simultaneously by connecting them in parallel.

3.12 PRECISE COMPARISONS.
3.12.1 BalanceComparisons.
Very precise comparisons between 2 capacitors nominally related by convenient ratios, such as 1,2, or 10, can be
made by connecting one as EXTERNAL STANDARD, the
other as UN KNOWN capacitor. The internal standards are
used only to evaluate the difference. Refer to para. 3.11.

3.12.2 Direct Substitution.
Accessories.
Stable, adjustable, 3-terminal shielded,
precision capacitor and (optionally) a conductance, i.e..
GR1422-CB capacitor, and two G874-R33 patch cords.
(Adjustable conductance standards, though not usually
needed, may be connected in parallel.]
Connections. Connect the test fixture you plan to use to
the appropriate UNKNOWN port; the compensating capacitor (and conductance, if any) to the EXTERNAL STANDARD port. Set the READOUT MUL TIPLIER as you intend
to keep it for subsequent measurements.
Adjustment. With the C and G lever switches set to zero
and EXT MULTIPLIER on 1.0 (or a lower setting) balance
the bridge by adjustment of the external standard. If that is
only capacitive, manipulate the CONDUCTANCE lever
switches as usual.
Notice that is is quite possible to make satisfactory testfixture compensation without reaching a perfect balance with
the external standard. For example, suppose you need to
compensate about 20 pF: READOUT MULT: Xl: available
capacitor range: 10 to 110 pF. Set EXT MULT to 0.2. This
capacitor has a resolution better than 4 fF, i e., < 1 fF,
referred to the test fixture. If that resolution is sufficient
simply verify that you have achieved it. (Refine the balance
using C lever switches and verify that appreciably less than
±0.5 fF is required of them.)
If, in that example. resolution must be lOaF, you need
another external variable standard capacitor in parallel with a
suitable fixed capacitor.

3.11.5 Range Extension to 11 mS.
Accessories. A standard 10-k!1 resistor and two G874R33 patch cords (or one cord and G874- MB adaptor).
General Procedure. Refer to the analogous paragraph
3.11.1, which concerns C rather than G. The readout for-

Even more precise comparisons between 2 nominally
equal capacitors can be made by connecting them sequentially as the UNKNOWN capacitor. External standards may
be used.
Balance Technique. Make use of (-1) and X settings of
lever switches if doing so will permit more of the most-significant digits to remain unchanged. Example: if capacitor
A measures 12345.0012+ pF, and B something more like
12344.9998+ pF, then make the readouts 12 nF 345 pF
001 fF 255 aF and 12 nF 345 pF 00(-1) fF 855 aF
respectively.
Repeat measurements A, B, A, B, etc. to eliminate any
affects of temperature changes, connector reliability, etc. In
this example, then, B is 0014 pF, i.e., 0.12 ppm, lessthan
A.
Precision. The only limit on how precisely a pair of
capacitors can be matched is the resolution of the bridge.
The errors listed in para. 3.11.3 do not apply; absolute
"farad" value is of secondary importance. If A and B are so
similar that each can be balanced repeatedly with the same
9 most-significant digits, and if the 9th digit is significant
(changing it upsets the balance) then you may be sure that
A and B are equal (for the moment) within ±.001 ppm
Generally, comparison precision is specified at ±.01 ppm
for the 1621 system. Refer to para. 3.10.
Terminal Capacitance. Terminal, fringing, and test-fixture capacitances cancel in direct-substitution
measurements, a fact that makes this method convenient for many
kinds of 2-terminal measurements, even when great precision is not required. Refer to para. 3.13.

3.13 NON-COAXIAL 2-TERM. CAPACITORS. Figure 3- 8.
Measurement of non-coaxial 2-terminal capacitors can be
precise and repeatable only to the extent that the various
stray admittances can be brought under control. Some can
be excluded from the measurement, some need to be included. Some are unavoidably influenced by the method of

OPERATION 3-17

connection to the bridge, and so the measurement is valid
only if such details are specified.
The capacitor terminals are represented in a general way.
If they are a pair of binding posts at standard spacing or a
coaxial connector and if the bridge UN KNOWN port can be
adapted to mate with these terminals, the method of connection can be specified reasonably well.

3.13.1 Unshielded, 2-Terminal Capacitors.

Figure 3-8a.

Diagrammed is a very simple capacitor, with an internal
capacitance C 10. C4 0 represents capacitance between the
terminals; C2 0 and C 3 0 capacitances to the nearest conductor in the environment (we assume for simplicity, there is
only one such conductor of importance)
Intrinsic Value. As it stands alone, the capacitance of
diagram a is
C20 C30
C 10 + C40 + C
C'
20

+

30

Connections. If the capacitor has a G874 connector or
1.75-inch-spaced binding posts. use a G900-0874.
or
G900- 09 adaptor. respectively. A length of cable or rigid
coaxial line can be used to locate the point of attachement
away from the bridge. Other capacitors (in general) will
require special adaptors or test fixtures.
Measured Values. When you connect that capacitor to
the 2-TERMINAL port of the bridge, all the strays change.
Diagram b shows a wire used to make connection at one
terminal. Even if you properly adapt the bridge port so the
capacitor mates without such a wire, C4 0 is now different,
i.e.. C4 1 . C2 1 is certainly larger than C2 0 · C3 0 may be
practically unchanged. C5 0 is capacitance from that "nearby conductor" to bridge ground.
Because C2 1 , C 3 1 , and C5 0 form a Y network with a
midpoint that we don't care about we can substitute the
equivalent 6. (Use a Y-6 or T-1Ttransformation.) Two parts
of the 6 are excluded from the measurement.* The third
part, across the bridge LOW and HIGH terminals, is

C2 1 C31
C2 35 =.
C .
C21 + C31 + 50
Certainly, if we make C5 0 large enough with respect to C2 1
and C3 1 , C2 3 5 will approach zero. If C5 0 is small enough,
C2 3 5 approaches Cj , C 3 1 /(C 2 1 +C 3 d ·
As represented by diagram b. the measured value is

It is usually preferable to make the third term zero
rather than try to make it approximate the third term in
the "intrinsic value." If you do the former, by grounding
any nearby conductors (diagram c) the measured value is:
C 1 0+C
"Because the 1616
ground as guard.

circuitry

3-18 OPERATION

4 1·

is basically

a 3-termlnal

bridge

with

Correction for Cable, Connector, etc. There are 2 distinct methods of removing adaptor capacitance from your
measurement (results differ). For each method, there are 2
techniques (results are identical).
Open-Connector Method. Be sure there is nothing but air
at the connector or fixture to which the capacitor will be
connected. The 2 equivalent techniques are:
1. Use an external standard and/or ZERO ADJUST to
balance the bridge with zero CAPACITANCE readout
(ref. para. 3.11). Now when you measure an unknown
capacitor, the readout is the corrected measurement.
Alternatively.
2. Balance the bridge and record the readout Co F ' the
capacitance of adaptor, cable, and fringing. Now when
you measure an unknown capacitor, correct your measurement as follows: "readout" minus Co F ' (Refer to
para. 3.9.5 for more about fringing.)
Substitution Method. Connect a standard capacitor Cs
with the same terminal configuration as the unknown to
the test point. The 2 techniques are
1. Use an external standard and/or ZERO ADJUST to
balance the bridge with CAPACITANCE readout set to
Cs' Now when you measure an unknown capacitor, the
readout is the corrected measurement. Alternatively
2. Balance the bridge and record the readout (Cs+C o )
Refer to para. 3.12. Now when you measure an unknown capacitor, correct your measurement as follows
"readout" plus Cs minus (Cs+Co )'
NOTE
If the standard used in the substitution method
was calibrated to eliminate fringing and C4 0 ,
the corrected measurement is C 10.

Use the 3-Terminal Port (Figure 3-8. d). If you wish to
measure. as directly as possible. the internal capacitance C .
10
make a separate. shielded connection to each terminal. Make
connections (if the capacitor has a pair of binding posts) with
a pair of G777-03 adaptors and a pair of G874-R22LA
cables. Connect only the shielded plug of each G77 7- 03 to
the capacitor. Ground any nearby conductor as shown. If the
shields extend far enough. C42 will certainly be less than C40'
The measured value is:

NOTE
As a rule, this bridge readout is valid without
correction for cable and adaptor capacitances,
although either method described above may be
used for greatest accuracy.

3.13.2 Shielded 2-terminal Capacitors.

Figure 3-9, e.

These capacitors are enclosed in a conducting shield:
examples are GR 1409-Y.
or 1422-0.
Some may
alternatively be measured as 3-terminal capacitors (refer to

CASE

leI

lal

:

}

i

/

I

I

I
I

/

I

I

::;:<..
7\ c"

I
J

,....--'"""ilt----..,-

I

~C21
I

I

I

~C50

I
I

I

I

I

I

I
I

'-'1E-/-"-:::
~----if~~:

--/

I
I

60

"

I

--L

I

.,C

C70

-h-,E---

-'-

WIRE

C81

"-

CAS
lOW

I
I

!

H

l

BRIDGE 2-TERMINAl

I

H

H

l

H

2-TERMINAl

PORT

PORT OF BRIDGE

1

fI

[b]

C 70

---iE--

"

I/
I\,

C60

C~/1\-:-- t--/

!'--j:----

"

".-

r-

t----,HIGH

lOW

'i(---- r-::V

"82

CA SE

WIRE)

H

l

L
-=l

H

BRIDGE 2-TERMINAl

PORT

H

BRIDGE 3- TERMINAL

PORT

(gl

Icl

C 70

r-

"
;x./--iE-----~ ........

"-

:...--'

42

"-

II

---1E----

C6 0

"-

HIGH

LOW

CASE

:--

"'if ---r-::-:.....--

C83

10

C93

l-

H

T-GAP
H

Idl

BRIDGE 3-TERMINAl

Figure 3-8.

PORT

Ihl

BRIDGE 3- TERMINAL PORT

Figure 3-9.

Non-coaxial 2-terminal capacitors and some measurement configurations.

OPERATION 3-19

para. 3-4). Others may have one terminal already connected
to the shield.
Intrinsic Value. As it stands alone, the capacitance of
diagram e is

However, we are specifically interested in 2-terminal
value, with one terminal connected to the case. Assuming
the right-hand terminal is the one, then the intrinsic capacitance reduces to.

Connections. Connect the LOW terminal to the GND or
shield terminal using a link. 938- L, if it fits. Connect the 2
capacitor terminals to the 2- TERMINAL UNKNOWN port of
the bridge as follows: capacitor LOW to bridge outer (HIGH):
capacitor HIGH to bridge inner (LOW). If the capacitor
terminals are 1. 75-inch-spaced binding posts. use G777-03
and G900- 08 74 adaptors and, if desired, a shielded cable
(G874- R22 LA). Alternatively, use only the link and a G90009 adaptor.
Measured Value. When you connect that capacitor to the
bridge at least 2 of the strays change. C8 0 is probably
reduced (by shielding) to C8 1 ; C9 0 changes completely
(now that LOW and GND are connected) to C9 1 .
As shown by Figure 3-9, f, the measured value is

NOTE
Correct the bridge readout for cable and adaptor capacitance, as described before.
Use the 3- Terminal Port, Simply (F igure 3-9, g) If you
wish to make virtually the same measurement without the
need to correct for cable and adaptor capacitances, proceed
this way
a Connect LOW to shield or GND of the capacitor with
atink as before.
b. Install a G874- MB adaptor at the bridge 3- TERMINAL
HIGH connector. Tie that to the capacitor LOW with a wire or
plain patch cord.
c. Connect the G77 7- 03 adaptor shielded plug only to
the capacitor HIGH terminal. Tie that to the bridge 3TERMINAL LOW connector using a shielded cable. G874R22LA. Be sure the shield of the G777-03 adaptor does not
contact the shield surrounding the capacitor. Make a washer
of paper or plastic(.003 to .015 in. thick) for this purpose.
d. Your measured value (diagram g) is

3-20 OPERATION

(where C8 2 is comparable to C8 1 + C9 1 of diagram f.)
Use the 3-Term. Port, with Correction(Figure 3-9, h).
Some capacitors are specified in terms of the capacitance
added to a given bridge port (Adapted, if necessary, to suit
the capacitors). For example the 2-terminal capacitances of
certain GR 1422 capacitors are so defined. A typical
procedure follows:
a. Install a 938-L link on the capacitor-case (GND)
binding post, connecting also to the adjacent LOW post, if
any.
b. Install a G777-03 adaptor on the capacitor - insulated
plug at capacitor HIGH. shell connected plug at the adjacent
bind post with the link.
c. Install a G874-MB
Coupling Probe as an adaptor on
top of the G777-03JJnserew the binding post just installed,
far enough to determine the gap as follows
d. Install anotherG777-03(shielded plug only) on top of
the stack. With it firmly seated, verify that a 1/16-in. gap (2
mm or less) exists between the shields.
NOTE
This gap is the demarkation between part of
the adaptor stack included in C9 3 and part that
is not. To stabilize the gap, wrap assembled
adaptors tightly with electrical tape

e. Connect the top adaptor to the bridge 3-TERMINAL
UNKNOWN, LOW using a shielded cable, G874- R22 LA.
f. Install another G87 4- MB adaptor at the 3- TERMINAL
UNKNOWN, HIGH connector. Tie this, using a patch cord
or wire, to the link in step a. (If the capacitor has only 2
binding posts, fasten to the link with an aligator clip;
otherwise, use the 3rd binding post.)
g. Temporarily unplug the first G777-03 adaptor from the
capacitor binding posts and support it nearby with the
inner, insulated plug exposed. Tie the shell-connected plug
(uninsulated) to the capacitor-case binding post (bridge
HIGH) as before.
h. Compensate the bridge by the open-connector method (technique 1 or 2) described above. This step eliminates
C9 3 (diagram h) from your measurement.
i. Reinstall the first G777-03 adaptor as in step b. Measure
the unknown capacitor as usual.
J. Your measured value (diagram h) is

3.14 DC BIAS.
Dc bias voltage may be applied in either of two ways to
a capacitor that is being measured on the bridge.

l~

&
CAUTION
Do not apply voltage at the bridge DETECTOR
OUTPUT connector in excess of E M A X in
Table 3-2, or the G standards may be damaged

e. Connect the unit holding R B to the tee. and that to
the DFTECTOR OUTPUT connector of the bridqe iiSlflq
an adaptor (GB740BPA)

A recommended power supply is the G R 1265-A.

&
WARNING
• To minimize electrical shock hazard, limit
bias to 60 V.
• Bias voltage is present at connectors, test
fixtures and on capacitors under test.
• Capacitors remain charged after measurement.
• Do not leave instrument unattended with bias
applied.
• To remove the risk of electric shock, turn
the voltage source to "0" before connecting
or disconnecting the device under test.

...L

+

L

l~---------+--__
-,

Figure 3·10. Circuit for applying bias to the unknown
capacitor - normal configuration.
Notes: • Resistor in
G874-X Insertion Unit; G874-02
Adaptor provides
binding posts for power supply. .. G874- T tee with
G874-QBJA and G874-QBPA Adaptors to BNC connectors.

It should be noted that the instrument is rated to accept
bias voltages up to 500 V rms or de. However, to minimize the risk of electrical shock hazards, the use of biasing
voltages of less than 60 V is highly recommended.
3.14.1 Normal Bridge Configuration / Parallel Bias.Fig. 3·10.
To apply bias In parallel with the detector
a Connect the bridge as usual, but add the de bias
supply In parallel with the deter-tor. as shown in the circuit
of F iqure 310.

b. Shield all leads connected to the high side of the
detector. Use a G87 4- T tee connector for convenient parallel
connections and a G87 4- OBJA Adaptor
If your detector is not the GR 1238, connect a series
capacitor between detector input and bias supply to block
bias voltage train the input stage. (This capacitor IS built
into the detector input stage of the 1238)
d. Connect a series resistor R B between the high lead in
the tee and the bias supply to prevent the low impedance of
the bias supply from shorting the detector input A resis
ranee of about 100 kn is recommendod Lower resistance
reduces the detector sensitivity; higher resistance reduces
the de voltage across the unknown capacitor since

1. Either polari tv of de bias can be appl ied. Choose the
polarity required by the capacitor being tested. Observe the
r.urr ent limitations described below.
g Calculate the bias from a known E B , using the formula in step d.

3.14.2 Reversed Bridge Configuration / Series Bias. Fig. 3-11.
Compared to the normal configuration,
this method
allows higher oscillator voltage to be applied at low frequencies (unless the CONDUCTANCE
multiplier must be set
very high), has lower sensitivity; and permits direct bias
measurement In the reversed configuration, you apply bias
in series with the oscillator; use the following procedure
a Connect the br idge "reversed"; refer to para 3.15
b. Connect the de bias supply or battery in series with
the asci Ilator.
Observe the current limitations

described below.

----------Table
Refer to Table 3-2 for the effective bridge resistance R
at
BO
the DETECTOR OUTPUT connector. Install the resistor In a
shield such as a G874-X Insertion Unit

3-2--------CHARACTERISTICS AT BRIDGE
DETECTOR-OUTPUT CONNECTOR

CONDUCTANCE

Multiplier

1
NOTE
Use a choke in place of R B for high ac impedance, when low dcvoltage
drop is needed.
Shield the choke from both rnagnetic and electric fields.

R BO

70 V

9kn

10-1

210

89

1(J2

500

810

10- 3

1Cf4, 1O'5 , or 1Cf6

~)OO

500

8Mn

80

"Vof ts rms or dc.

OPERATION 3-21

d. Calculate the bias from a known E B, using the formula given above, except that R'B depends on the settings of
FREQUENCY range and OUTPUT VOLTAGE RANGE on
the 1316 Oscillator. (R' B can be as large as 2.4 kQ.)
Alternatively,
measure Ex' (F igure 3·11 )
"GENERATOR INPUT" CONNECTOR

GR 1238
OETECTOR

c. Turn the oscillator OUTPUT
and then slowly ccw to zero again.

ADJUST

3.15 REVERSED CONFIGURATION.

cw to MAX

Figure 3-11.

3.15.1 Explanation.

etJ
I

OPTIONAL
FLOATING
81AS SUPPLY

To apply higher ac test voltage (up to 500 V) to the
unknown capacitor than that normally permitted), use this
configuration, in which the oscillator is connected to the
bridge DETECTOR OUTPUT connector and the detector is
connected to the bridge GENERATOR INPUT. Use this
reversed configuration when lower sensitivity can be tolerated.

H

In normal operation, the maximum voltage across the
unknown capacitor would be limited by transformer-core
saturation to E M A X
0.16 f (para. 3.8) However, in
reversed operation, the maximum voltage is lirnited by
power dissipation in bridge resistors, or by insulation breakGR !316
OSCILLATOR

GR 1616 8RIDGE

Figure 3-11. Reversed configuration - bridge circuit
with oscillator and detector interchanged, showing optional application of bias to unknown capacitor.

down. The former depends on the CONDUCTANCE
plier setting
refer to EM A X in Table 3·2.
When the configuration

multi-

is reversed, the ac test voltage

across the unknown capacitor is equal to the generator
voltage at balance, or between zero and twice that voltage
for any unbalance

3.15.2 Procedure.
3.14.3 Dc in the Ratio Transformer / Demagnetization.
When the capacitor that is measured with bias (by either
of the preceding configurations) passes leakage current, de
will flow through the bridge ratio transformer. This current
will magnetize the core and affect the accuracy of the

a. Interchange 2 connections at the rear of the bridge
(Figure 1·3) so you have oscillator POWER OUTPUT tied
to bridge DETECTOR OUTPUT and bridge GENERATOR
INPUT tied to detector INPUT SIGNAL.
b. Balance the bridge as usual (para. 3.5) but observe the
limits of Table 3·2 instead of para. 3.8.

ratios, and such use generally is not recommended. It is,
however, possible to operate the bridge with some de in the
transformer,
if you keep th is cu rrent below the I M A X
values shown in Table 3-3 to avoid saturation of the core.

Table 3-3 --------BIAS CURRENT LIMITS FOR
TRANSFORMER SATURATION

----------

Operation within these limits will not damage the bridge,
but demagnetize the transformer after such use to insure
accuracy in normal operation. Demagnetize as follows
a. Connect oscillator to bridge in the normal confiqura-

READOUT MULTIPLIER

tion.
b. Set oscillator FREQUENCY
to 100 Hz, OUTPUT
VOLTAGE RANGE to 50, OUTPUT ADJUST to zero. Set
bridge TERMINAL SELECTOR to CAL.

XlO

3-22 OPERATION

X 100
X 1

200 mA
20

2

• An elastic limit ~ for best resolution these values are too high; for
some measurements,
currents of 5 'MAX
may be tolerable.

Theory-Section
4
4.1
4.2
4.3
4.4
4.5
4.6
4.7

INTRODUCTION
.....
PROPERTIES OF CAPACITORS
BASIC BRIDGE CIRCUITRY
.
CIRCUITRY OF THE 1616 BRIDGE
C·STANDARDS ACCURACY
G·STANDARDS ACCURACY
RATIO ACCURACY

4-1
4·1
4-5
4·7

4·9
.4·10
. 4-11

4.1 INTRODUCTION.

4.2 PROPERTIES OF CAPACITORS.

The 1616 Precision Capacitance Bridge is a standardslaboratory instrument
of exceptionally
high precision
Together with the other components of the 1621 Precision
Capacitance-Measurement System. it is designed for accurate measurements of capacitance, conductance, and therefore the properties of dielectrics, as well as high-resolution
comparisons among capacitance standards
This section deals with the 1616 bridge For theoretical
discussions of the 1316 Oscillator and 1238 Detector,
please refer to their individual instruction manuals

4.2.1 BasicComponents of Capacitance.

Figure 4·1.

Three Terminals. Most physical capacitors can be precisely represented by the three capacitances shown in Figure 4-1. the direct capacitance, C H L' between the terminals
Hand L (capacitance between the plates of the capacitor),
and the two terminal capacitances, C H G and C L G , from the
corresponding terminals and plates to the case, surrounding
objects and ground (to which the case is connected to
either conductively or by its relatively high capacitance to
ground)
A 3-terminal capacitor (Figure 4-2) has connected to the
G terrninal a shield that cornpletely surrounds at least one
of the terminals (H), its connecting wires, and its plates
except for the field that produces the desired direct capacitance to the other terminal (L) Changes in the environment
and the connections can vary the terminal capacitances,
usually
C H G and C L G , but the direct capacitance C H L
referred to simply as the capacitance of the three-terminal
capacitor - is determined only by the internal structure.

Figure 4·1. Schematic diagram of a capacitor, showing
the direct capacitance and the associated terminal
capacitances.

f----f

.. IL

Figure 4·2. Structure of a 3·terminal capacitor with 2
coaxial connectors.

This direct capacitance can be calibrated by 3-terminal
measurement methods, utilizing guard circuits or transformer-ratio-arm bridges, which exclude the terminal capacitances.
The direct capacitance can be made as small as desired,
since the shield between terminals can be complete except
for a suitably small aperture The losses in the direct capacitance can also be made very low because dielectric losses in
the insulating materials can be made a part of the terminal
impedances. When the 3-terminal capacitor is reconnected
as 2-terminal, the 2-terrninal capacitance will exceed the
calibrated 3-terminal value, C H L, by the terminal capacitance C H G .

Two Terminals. In the common 2-terminal connection,
the capacitor has Land G terminals connected together,
ie, L terminal connected to case The terminal capaci-

THEORY

4·1

m
oklJ

tance C L G is thus shorted, and the total capacitance is the
sum of C H Land C HG.
In general, since one component

of the terminal capaci-

tance C H G is the capacitance between the terminal and
surrounding objects, the total capacitance is changed by
changes in the environment of the capacitor and particularly by the introduction
of the wires required to make
connection to the capacitor.

(o)

-x

STRUCTURE,
IT MUST
BE UNGROUNDED

Lb 1 CAPACITANCE
COMPONENTS

NOTE

C"

There is a possible confusion between Hand L.
Usually the 2-terminal capacitor is labeled H for
the more completely shielded connection, L for
the one connected to shield (if any) However,
the termi nals of the 1616 bridge are labeled the
opposite way L for the inner, H for the outer
part of the 2-terminal port Bridge H connects
to capacitor Land
G, but G must not be
grounded.
The uncertainties in the calibrated value of this 2-terminal capacitor can be of the order of tenths of a picofarad, if
the geometry, not only of the capacitor proper but also of
the environment and of the connections, is not carefully
defined and specified. For capacitors of 100 pF and more,
the capacitance can usually be adequately defined for an
accuracy of a few hundredths percent, if the terminals and
method of connection used for cal ibration are specified
For smaller capacitances or for higher accuracy, the noncoaxial 2-terminal capacitor is seldom practical. The 3-terminal arrangement is generally preferred.
Nevertheless, there
are very accurate capacitors (in all but the smallest sizes)
with 2, coax ial, ungrounded terminals.

Two Terminals, Coax Connection (Figure 4-3) In the
coaxial 2-terminal structure, capacitor terminal L is again
connected to the case, but not ground The case is a
complete shield around the field that determines the direct
capacitance C H L. Fields outside the case contribute only to
C LG, which can be excluded from the measurement So
most environmental changes have little effect on accuracv.
If the bridge is designed to measure such a capacitor and
each incorporates a suitable precision coaxial connector. the
uncertainties of measurement can be as low as 100
attofarads. The realization of such precision depends largely
on exacting definition of the boundary between capacitor and
bridge. This boundary is represented in Figure 4-3 by line XB. Physically. the boundary is the plane of the mating face of
the outer conductor of the G900 connector. i.e.. the
reference plane of the connector. For precise measurements.
the electric field must be entirely parallel to that reference
plane so the portion represented by C is unequivocally part
x2
of C. whereas C
and C are not.
x

4-2 THEORY

Fx

0

C"

--IE--~­
8
C,

Co

-1E- -IE-

~
tel

BRIDGE

(e1) CAPACITOR
BRIDGE

II~
ON

Figure 4-3. A coaxial 2-terminal capacitor, its structure, component capacitances, and connection to a
bridge.

The fringing capacitance C F x makes the freestanding
capacitance of the capacitor slightly greater than C, How
ever, C; is the value we always measure, calibrate, and refer
to as the capacitance of the device.
The fringing capacitance C
makes the bridge terminal
capacitance slightly greater ~~en open- circuited than the
value C which obtains when a suitable capacitor is attached.
ADJUST cannot be so accurately set with the
The
UNKNOWN port open as with a standard coaxial 2-terminal
capacitor installed. Limitations on the accuracy with which
GenRad can specify C involve uncertainty as to the position
.
of spring-loaded
partsFB.In the G90 0 connector an d reduce d
resolution because of "noise" picked up by the unshielded
LOW terminal. However the repeatability is much better than
the tolerance given C
in para. 3.4. i.e.. ±8 fF.

zmo

FB

4.2.2 Inductive and LossyComponents.

Figure 4-4.

No physical capacitor is ideal, ie., free of inductance
and dissipation, although some are excellent So also, the
equivalent networks that are used to represent a nonideal
capacitor do so imperfectly
However, they are satisfactory
in many instances, particularly if some of the parameters
are understood to be quasi-constants
(They may vary
somewhat with temperature, humidity, pressure. frequency,
acceleration, aging. illumination, etc.)
Generalized Circuit. Figure 4-4 (a) represents the nonideal direct capacitance (C H L of Figure 4-1) with 5 lumped

constants R represents the metallic resistance in the leads,
supports and plates; L, the series inductance of the leads
and plates; C I , the capacitance between the plates; C K' the
capacitance of the supporting structure Conductance G
represents the dielectric losses in the supporting insulators,
the losses in the air or solid dielectric between capacitor
plates, and the doc leakage conductance. For most purposes,
C, and C K are added as C (Notice that C K is zero if
support-structure capacitance is entirely with in C H G and/or

CL G )

1,

HIGH

r

more lag in IT By itself, G has no effect on apparent
capacitance because the bridge resolves them separately
Equivalent Parallel Circuit. Figure 4-.4 (c) is the equivalent circuit based on the 2 components measured directly
by the bridge Terminal properties (E T, IT, (J) are identical
with diagrams a and b Diagram d shows the vector relationships
The algebraic relationships between these components
and those of diagram a are given below, followed by simplifying approximations that are convenient for studying the
effects of one parameter at a time. It is the capacitance C;
(not CSE ) that we generally mean by the term "apparent
capacitance".
C (1 - w 2 LC)

C =------------2
x

ET

(1

I
G

LOW

G2 L

w LC + RG)2 + (wGL + WRC)2

G (1 + RG) + w 2RC 2
w 2 LC + RG)2 + (wGL + WRC)2

= --------------

x

(1

(e)

There is a resonance between Land C at a frequency f o .
We are concerned only with the condition f
f o (and C; is
positive)
At sufficiently low frequency (f«
f o ), we observe that
w 2 LC «
1, but various losses may still make C; =I=-C,
thus

<

I,

Ie;.

i:-fl'
I

C,

REFERENCE
DIRECTION

L

OF E,.

(c)

(f)

'---"'-'-'1

ERs·RSI,.
I

I

--REFERENCE
DIRECTlQN

I

G "'::

I
I
I

x

I

nu

__

I
I
I
I

x

C - G2 L

C - G2 L
"'::----

(1 + RG)2 + (wGL + WRC)2

OF IT

I

__

C "'::

G(1+RG)+w

(1 + RG)2

2RC 2

(1+RG)2+(wGL+wRC)2

"'::---

G

1 + RG

E.

Figure 4-4. Capacitor lumped-parameter
circu its and vector diagrams.

equivalent

The corresponding vector diagram, Figure 4-4 (b) shows
how the capacitive and conductive current components add
to make IT, and how the capacitive, inductive, and resistive
voltage components add to form ET. IT leads ET by the
phase angle (J, between 0° and 90°, as is characteristic of a
physical capacitor. (We are not interested in conditions at
such high frequency that IT lags ET) The complement of (J
is 8, the dielectric loss angle
Notice that, regardless of the relative magnitudes of the
other vectors, a small inductive component (IELI < IEel)
always makes lET I < IEeI In other words, the presence of
series inductance makes a capacitor apparently increase in
capacitance as frequency (or the inductance) increases.
The effect of series resistance on apparent capacitance is
opposite, because ER tends to make IETI > IEel But the
effect of shunt conductance G on apparent capacitance is
mostly to accentuate the effect of R, if any, by causing

where the fractions at the right are valid at sufficiently
values of Rand L.

low

<

fo
If, however, we relax the frequency restriction to f
but consider the case of negligible series loss (wRC «
1)
and reasonably small shunt loss (G < WC), then C; increaseswith a term proportional to f2 as follows

With this information, the capacitance at, for example, a
frequency of 2 MHz can be con puted with high accuracy
from the calibrated value at 1 kHz. For fifo up to 0.3 or so,
the accuracy may be greater than that of a measurement at
2 MHz because of the difficulties in determining the measurement errors produced by residuals in the connecting
leads outside the capacitor (unless, of course, one uses a
bridge particularly designed for measurement at 2 MHz)
You may use a grid-dip meter for measuring f o, with the
capacitor terminals shorted.

THEORY

4-3

There are secondary effects on the apparent value of
capacitance in a 3-terminal measurement that will not be
described here except to say that they involve resonances
among Land C components in all 3 basic parts of the
network of Figure 4-1 (The series inductor in Figure 4-4
(a) may represent, at least in part, a wire carrying current
not only to CH L but also to CH G or part of it) These
effects are generally negligible if the following approximation is valid within the desired measurement accuracy

Q

tan

e=

wC x

1

5

x

3. Power factor is the ratio of the real power (watts
of dissipation) to the product of rrns voltage and current
(voltarnperes)
P. F = cos

ERs

e

E

=

sino

T

where f' a is the lowest resonant frequency of the network
Estimate f'o as the free ringing frequency of the largest C
component, such as C H G' (given some initial energy) with
the capacitor-lead and cable inductances; assume the bridge
HIGH terminal is virtually grounded and its LOW terminal
is an open circuit.
Equivalent Series Circuit. Figures 4-4 (e) and 4-4 (f)
present another 2-component equivalent circuit which may
be appropriate if Rand C are the only significant components of diagram a However, the 1616 bridge does not
I n terms of the parallel equivalent
measure Rs directly
circu it

Loss Factors. An important characteristic of a dielectric
material, and hence of the capacitor made from it, is the
ratio of energy dissipated to energy stored, per cycle of ae
There are 3 commonly used "factors" to express the same
characteristic.
1. Dissipation factor or loss tangent 0 is the ratio as

Notice that several other convenient expressions can be
taken from Figure 4-4, such as Q or 0 in terms of current
components in diagrams c and d or voltage components in
diagrams e and f Other convenient expressions are given in
para 36, for use of 0 or Q in converting data from
series-equivalent to [)<.rallel-equivalent circuit forms and
vice versa.
NOTE
For low-loss capacitors, the difference between
o and PF is very small For 0 1CfI , 1Cf2,
and 1Cf3 , (0
PF )/0 is 05%,005%,
and
5X 1Cfs%, respectively

4.2.3 Frequency Characteristics.
The lumped-parameter equivalent circuit such as Figure
4-4, a, for a real, physical capacitor is most appropriate
under the conditions in which the parameters (R, L, C, G)
are constants I n the following examples are some situations
in which they are, others in which we have to treat them as
quasi-constants, or develop special "constants".
Capacitance. At high frequencies, the inductive effect
predominates (C; increases with f2) At low frequencies,
C; is essentially constant if Q is very high and the dielectric
is air. But if the capacitor has a solid dielectric, such as
mica, it is responsible for a slight capacitance change with
frequency; C 1 is then a quasi-constant
10

defined above
fZ

w

li
It

1'0

I

Expressed in terms of the components of Figure 4-4 a
+ w 2RC 2
0=--'----·--wC (1 - w 2LCl - wG 2L
G (l+RG)

- ---r--

0.0 I
10Hz

100

1kHz

10

/

J --.

100

/

IMHZ

"----C

10

.---

FREQUENCY

Figure 4-5. Capacitance vs frequency characteristic of

I f frequency is well below resonance and if both series
and shunt losses are small, a good approximation is

G

0:=::::-+
wC

RwC

.

2 Storage factor or quality factor Q is the ratio of
energy stored to energy dissipated per cycle of ac

4-4 THEORY

a mica capacitor.

The change in capacitance C; with frequency, of a
1000-pF capacitor with mica dielectric, is shown in Figure
4-5. The dashed line slanting downward to the right represents the change in the dielectric constant of mica resulting
from interfacial polarization: that slanting upward to the

right shows the change in effective capacitance C x resulting
from series inductance. Though the former phenomenon
may be true at much higher frequency, it can be neglected
above 300 kHz in this example, because there the latter
effect is so predominant. The magnitude of the change at
low frequencies depends upon the dielectric material and is,
for example, much smaller for polystyrene than for mica
Dissipation Factor. At high frequencies, "series" loss
predominates because of skin effect in the leads and plates
(as explained below) At low frequencies, "shunt" loss
predominates, but G may be far from constant. If goodquality, low-loss, solid dielectrics are used, G is nearly
proportional to w; then it may be useful to treat G/wC as
the quasi-constant DZ
The skin effect can be represented by considering R a
variable, R = Rz J(;J, in the second term, RwC, of the
formu la for 0, then

effective conductance at w
capacitor

1. Thus, for the variable air

At very low frequencies, a "shunt" loss conveniently
represented by G may be significant. Leakage conductance
G is usually negligible at frequencies above a few Hz and is
important only when the capacitor is used at dc for charge
storage. The dominant components of 0 at audio frequencies are the dielectric losses in the insulating structure and
in the dielectric material between the plates

4.3 BASIC BRIDGE CIRCUITRY.

*

4.3.1 Elementary Capacitance Bridges.
Measurements of capacitance, particularly those of high
accuracy, are made by a null method that uses some form
of the basic ratio bridge. shown in Figure 4-7 The capaci-

0= Dz + RZCw 2 /3
The dissipation factor as a function of frequency for a
mica capacitor is shown in Figure 4-6 There 0 is the sum
-2

10

I

0::

~

~163

z
o
~164
o,

f M 1N

--

~

- - -- - --

(f)
(f)

--

-- -- - - --

100

I kHz

10

.-

/

-

°10. 5
10 Hz

!

/,
_1 __

,

/

100

/
Figure 4-7. An elementary

-- -- -- --

- - --

I MHz

- --

100

FREQUENCY

Figure 4-6. Dissipation factor vs frequency characteristic of a mica capacitor.

of three principal components a constant dissipation factor
caused by residual polarizations and shown by the horizontal dashed line; a loss produced by interfacial polarizations,
which contributes the 0 shown by the dashed line slanting
downward to the right; and an ohmic loss with sk in effect
in the leads and plates. which results in a 0 proportional to
the 3/2 power of frequency and is shown as the dashed line
slanting upward to the right The total dissipation factor
has a minimum value at a frequency that depends on the
size of the capacitor; f M I N varies inversely with capacitance
and ranges from 1 k Hz to 1 M Hz for capacitance values
from 1 pF to 100 pF.
In an air capacitor. the losses in the air dielectric and on
the plate surfaces are negl igible under cond itions of moderate humidity and temperature
The loss is, therefore,
largely in the insulating supports, and the above formula
applies. However, if the capacitor is variable. and the varying part is lossless, D Z varies inversely with C Then it is
preferable to treat G/wC as G I w/wC,

where G I

is the

capacitance

measuring bridge.

tance of the unknown, Cx,s. balanced by a calibrated,
variable. standard capacitor, C s' or by a fixed standard
capacitor and a variable ratio arm, such as RA . Such bridges
with resistive ratio arms and with calibrated variable capacitors or resistors can be used over a wide range of both
capacitance and frequency and with a direct-reading accuracy which seldom exceeds 0 1%
For higher accuracy. resolution. and stability in capacitance measurements at audio frequencies, a bridge with
inductively-coupled
or transformer ratio arms has many
advantages, and increasing use of transtorrner-ratio-arrn
bridges is being made in the measurement of many types
and sizes of capacitors

4.3.2 Transformer-Ratio Bridges.
The advantages of transformer ratio arms in a bridge are
that accuracies within a few parts per million are not
difficult to obtain over a wide range of integral values, even
for ratios as high as 1000 to 1, and that these ratios are
almost unaffected by age. temperature, or voltage The low
impedance of the transformer ratio arm also makes it easy
to measure direct impedances and to exclude the ground

'Thomas,

Instruments
Cliffs,

H.

E.,

and

Clarke,

and Measurement

N. J. (1967)

C. A .. Handbook
of Electronic
Hall, Englewood
Techniques, Prentice

p 36·57.

THEORY

4-5

impedances in a three-terminal measurement without the
use of guard circuits and auxiliary balances.
To illustrate these characteristics, a simple capacitance
bridge with transformer ratio arms is shown in Figure 4-8.
On the toroidal core, a primary winding, connected to the
generator, serves only to excite the core; the number of
primary turns, N p , determines the load on the generator
but does not influence the bridge network. If all the magnetic flux is confined to the core - as it is to a high degree
in a symmetrically wound toroid with a high-permeability
core - the ratio of the open-circu it voltages induced in the
two secondary windings must be exactly equal to the ratio
of the number of turns The ratio can be changed by the
use of taps along the two secondaries, but, when the ratio is
fixed, the voltage is highly invariant Changes in the core
permeabil ity with time and temperature have only very
small effects on the effective ratio. It depends not only on
the turns ratio (a perfect integer) but also on leakage flux,
wh ich is not confined to the core in a practical transformer.
The ratio is, therefore, both highly accurate and highly
stable.

\ I,

the terminal G connected to the junction of the ratio arms.
The capacitances between Land G shunt the detector, so
that they affect only the bridge sensitivity The capacitances between Hand G are across the transformer windings. To the extent that the transformer can be assumed
ideal, i.e., with no resistance in the secondary windings and
with no flux that does not link equally both secondaries,
the current drawn by the H-G capacitances does not change
the voltages V 5 and V x or the balance cond itions In practice, the transformer resistances and leakage inductances
can be kept so small that quite low impedances or large
capacitances can be connected from H to G before there is
appreciable error in the bridge.
The junction of the ratio arms, G, is therefore a guard
point, or guard potential, in the bridge All capacitances to
G from the H or L corners of the bridge are excluded from
the measurement In the three-terminal capacitors represented by the H, L, G terminals in Figure 4-8, the bridge
measures only the direct capacitance. C x' of the unknown
in terms of the direct capacitance, Cs ' of a standard without additional guard circuits or balances.
One can take advantage of the accurate and stable ratios
of the transformer by designing a bridge with an "unknown" arm that is fixed and a ratio that can be varied to
balance the bridge Far greater measurement accuracy is
feasible with such a design approach (making C, a fixed
capacitor rather than a variable one) For example, consider
the following alternatives.
Figure 4-9 shows three of the possible ways of balancing
a simple transformer-ratio capacitance bridge. For simplicity, the generator and primary are not shown, but it is
assumed that the two secondaries have 100 tu rns each and
are excited so that there is 1 volt per turn. The capacitor in
the unknown arm is assumed to be 72 pF
IOOV

Figure 4-8. An elementary capacitance bridge with
transformer ratio arms.

-100

V

This balance condition is not affected by the capacitances shown from the Hand L terminals of Cs and C x to

4-6 THEORY

-70V

(0)

(b)

(e)

FIXED RATIO

MULTIPLE DIVlDERS
SINGLE FIXED CAPACITOR

SINGLE DIVIDER
MULTIPLE
FIXED
CAPACITORS

VARIABLE CAPACITOR

In Figure 4-8, the two transformer secondary windings
are used as the ratio arms of the capacitance bridge with the
standard capacitor, Cs, and the unknown, Cx' as the other
two arms in a conventional four-arm bridge network. The
condition for balance, or zero detector current, is easily
shown to be that

IOCV

IOOV

Figure 4-9. Circuitry for 3 methods of balancing a
transformer-ratio capacitance bridge.

In Figure 4-9a, the two ratio arms are equal and the
bridge is balanced in the conventional way with a variable
standard capacitor which is adjusted to 72 pF
The detector current can equally well be adjusted by a
variation in the voltage applied to a fixed standard capacitor. In Figure 4-9b, the standard capacitor is fixed at 100
pF This is balanced against the 72pF "unknown", connected to the lOa-Vend of the transformer, by connect ion of

the standard to 72 V of the opposite phase, obtained from
suitable taps on the transformer windings The inductive
divider shown has a winding of 100 turns with taps every
10 turns and, on the same core, another winding of 10
turns tapped at every turn. If, as shown, the second winding
is connected to the 70 V tap on the first winding and the
capacitor to the 2-V tap on the second winding, the
required 72 V is appl ied to the capacitor. Six or more
decades for high precision can be obtained in a similar
fashion with more windings on one core and the use of
additional transformers driven from the first. Such inductive dividers have very accurate and stable ratios, but the
errors increase with the nu mber of decades because of
loading effects
Another method of balance by voltage variation is
shown in Figure 4-9c, where a single decade divider is used
in combination
with multiple fixed capacitors. The
lOO-turn secondary is tapped every 10 turns to provide
lO-V increments If, then, a 100-pF capacitor is connected
to the 70 V tap and a lo-pF capacitor to the 20 V tap, the
resulting detector current balances that of the 72-pF "un
known", connected to 100 volts. This bridge can be given
6-figure resolution, for example, through the use of 6 fixed
capacitors in decade steps from 100 pF to 0001 pF, each
of which can be connected to anyone of the taps on the
transformer.
In any of these bridges, the bridge ratio can also be
altered by use of taps on the "unknown"
side of the
transformer to select the voltage applied to the capacitor
being measured. For example, a 720-pF capacitor would be
balanced in any of the circu its of Figure 4-9 if connected to
a lO-turn (10 V) tap on the upper wind ing The range of
the bridge can thus be extended to measure capacitors
which are much larger than the standards in the bridge
These advantages of transformer ratio arms and dividers
make possible a bridge of wide capacitance range, and high
accuracy, one that is usefu lover a wide frequency range It
is economically reasonable to construct the relatively few,
fixed capacitance standards to have the necessary stability
and accuracy for such a bridge, one that will measure with
.01% accuracy over 6 decades of capacitance and 3 decades
of frequency. At low frequencies, a limit is imposed on
sensitivity by the maximum voltage obtainable from the
transformer, since, for a given core, the voltage at saturation is proportional to frequency. At high frequencies there
is a decrease in accuracy resulting from the decrease in core
permeability with frequency, from the increased loading of
the transformer by its self-capacitance as well as the bridge
capacitances and, of course, from the usual residual capacitances and inductances in the bridge wiring and components.

4.4 CIRCUITRY OF THE 1616 BRIDGE.

Figure 3-3.

In general, this capacitance bridge is based on the 3rd
method, Figure 4-9, C Each of the 12 internal capacitance

standards is calibrated at a cardinal value and each is
switched to an appropriate tap on the ratio transformer as
you bring the bridge to balance. Conductance (or loss in the
capacitor being measured) is balanced by the same method.
Conductance
standards of fixed, cardinal values are
switched to appropriate taps on the same ratio transformer
The circuitry is described in more detail in the following
paragraphs.

4.4.1 Excitation.
The application of sufficient voltage (power is mostly
reactive) from the oscillator to the ratio transformer provides a test signal so the detector will respond to an unbalance The level of excitation is limited, always below the
saturation level of the transformer core. The primary
winding of 200 turns is driven directly by the oscillator
POWER OUTPUT port. Several design features eliminate
stray coupling to other parts of the bridge and particularly
to the detector.
1. The GENERATOR INPUT port is connected to the
transformer by a shielded wire.
2. This shield is floating in the bridge (grounded only at
the oscillator)
3 In the ratio transformer, the primary winding is surrounded by an electric shield that is grounded In the
bridge

4.4.2 Circuit for the Unknown.

Figure 3-1, 3-2.

The 2 terminals of C x , HIGH and LOW, must be connected, respectively, to the ratio-transformer secondary (upper
half, in the simplified diagrams) and the summing point
(which drives the detector) Because the midpoint of the
ratio-transformer
secondary winding is grounded, both
terminals of C x must be floating However, the 3rd terminal
or shield of the capacitor under test is grounded at the
3-terminal port
The facility for measuring capacitors larger than the
largest standard, as mentioned above, obtains when the
TERMINAL
SELECTOR switch points to XlO or X100
Then, the HIGH side of C; is connected to the 20- or
2-turn tap, respectively, of the secondary winding When
the selector indicates Xl, the connection is to the 200th
turn from ground. Then the voltage across the unknown
capacitor at balance equals the oscillator voltage
Each of the UNKNOWN terminals is disconnected from
the bridge and grounded whenever the TERMINAL SELECTOR selects CAL. When it selects 3-TERMINAL,
the
2-terminal port is likewise grounded, and vice versa

4.4.3 CapacitanceStandards.

Figure 3-3.

There are 12 capacitance standards, one for each mu It ipie of 10 from 100 nF down to an effective value of 1 aF
The "high" side of each is connected to the front-panel
selected tap on standards side of the transformer secondary

THEORY

4-7

(generally the lower half in the simplified diagram) The
"low" side of each standard is permanently connected to
the same summing point mentioned above, which drives
the detector. However, the 2 largest C standards can be
disconnected from the summing point by dropping the C
MAX lever switch. Thus, stray capacitance shunting the
detector can be reduced while you measure extremely small
values of increments of C,
Each of the first 6 CAPACITANCE
lever switches
indicates values in tenths of the value of its associated
standard. Thus, the lever with a
1, 0, 1,2, ... 9, XpF
readout actually switches the lO-pF standard capacitor
The max imu m current through each of these standard
capacitors (when the corresponding readout is X, i.e.. ten)
obtains when its HI G H side is connected to the transformer
0
secondary at the 200th turn, 180 out of phase from the
unknown-capacitor-X 1 connection. That is the bottom of
the transformer as. shown in simplified diagrams The
intermediate taps are 20 turns apart The zero position is a
ground connection. The
1 position is a connection to a
2G-turn tap on the "unknown"
side of the transformer
secondary.
In order to provide greater stabil ity in the set of capacitance standards it is preferable to use a moderate valued
capacitor connected to fewer turns than an extremely small
valued one connected to the usual number of turns. Therefore a set of taps is brought out from the ratio transformer
at intervals of 2 turns (to a maximum of 20 turns) Each
capacitor that connects to this set of taps is 10 times as
large as it would have to be if it were connected to the
usual (20-turn-per-step) taps The last 6 "C" lever switches
span 2 turns per step. The 6 associated standard capacitors
are each 10 times as large as the effective values shown in
Figure 3-3.
Therefore, each of the last 6 CAPACITANCE
lever
switches indicates values in hundredths of the value of its
associated standard Thus the lever with a - 1, 0, 1,
2,
9, X -aF readout actually switches the 100-aF
standard capacitor For these, of course, the - 1 position is
a connection to a 2-turn tap on the "unknown" side of the
transformer
Physically, the transformer is manufactured with a
separate winding of 22 turns, tapped every 2 turns for the
small C standards. A set of 20 windings of 20 turns each
serves the large C standards and the "unknown", by a series
combination
that acts as a center-grounded 400-turn
secondary Except for the primary (200 turns) all these
windings are very intimately coupled by multifilar con
struction
Each C standard is cal ibrated to the desired value by
means of a trimmer. All 12 are located behind the locked
door in the front panel (Figure 1-2), where they are labeled
with the effective values of the corresponding standards, as
shown in Figure 3-3 Calibration procedures are given in
Section 5.

4-8 THEORY

4.4.4 ConductanceStandards.

Figure 3-3.

There are 5 conductance standards and a multiplier
circuit for reducing their effective values in steps, so that
you have virtually 11 conductance standards, one for each
multiple of 10 from 10 p.S down to 1 fS
Like the capacitance standards, the conductance standards connect as follows high side, via lever switches, to taps
on the standards side of the transformer; low side, via
multiplier circuitry to the summing point that drives the
detector.
Because this bridge is designed for the greatest precision
in C rather than G measurements, it is preferable to have
the main (20-turn-per-tap) secondary winding carry only
C-standards currents Therefore, the G standards are connected to the other (2-turn-per-step) winding described
before. Consequently, each of the 5 conductance standards
is 10 times as large as the effective values shown in Figure
3-3.
It follows that each of the 5 CONDUCTANCE
lever
switches indicates values in hundredths of the value of its
associated standard. Thus the lever with a - 1, 0, 1,
2, ... 9, X . p.S readout actually switches a 100 p.S standard of conductance, i.e.. a precision 10-kSl resistor
When the G multiplier is set to Xl, the connection from
the set of 5 conductance standards to the detector is direct
When that setting is X 10- 1 , a resistor network passesonly
10% of the conductance standards' current to the detector
summing point (The other 90% returns directly to ground)
Similarly, at XlO- 2 , l(Yo passes to the summing point The
"box" (with these G standards and networks) also contains
pots for setting the multiplier
ratios exactly to the
appropriate powers of 10 R 18 for 10- 1 , R 19 for 10-2 etc.
These pots are intended for calibration
purposes as
described in section 5.
Ideally, the current through each conductance standard
should be perfectly in phase with the ratio- transformer
voltage. However, the resistor is bound to have some endto-end capacitance. For example. R 1 has some stray C in
parallel. Without compensation, the current through this
stray C affects the bridge balance condition like increasing
the value of the C standard.
The compensation is represented schematically by C23,
connected between the midpoint of R 1 and ground Refer
to Figure 5-9 Physically, C23 is formed by a threaded,
spring-loaded sleeve surrounding the main part of R 1 and
adjustable by turning with a special wrench. The current
component from the transformer through C23 leads by a
large angle (nearly 90°) causing the voltage at the midpoint
of R 1 to lag slightly with respect to the ratio-transformer
voltage If the compensation is adjusted correctly, the
resulting lag in the current frorn the middle of R 1 to the
detector is just sufficient to cancel the leading current
through the stray C
Not only is each of the 5 G standards compensated in
this way, but also 1 of the 4 series resistors in the G

multiplier network Thus, R9 passesthrough a similar sleeve
with capacitance represented as C22. There is, of course,
more stray capacitance to ground in the G·standard circuit
when the G multiplier is set to some multiple other than
X 1. Consequently, in some of the positions of that switch,
the necessary compensation to achieve zero phase shift is
the addition of capacitance shunting the series resistor. So,
for example, for the io ' multiplier, R6 has the adjustable
C18 in parallel
The desired result is complete independence of C and G
balances, and hence measurements A measure of the
quality of the compensation is Do, the ratio of conductance
to lk Hz parallel susceptance of Cc of the G standard G s
(We refer to the effective conductance, including the multiplier network if it is involved, and the effective susceptance,
including the compensation described above i Refer to
para 3.9.3
Do

Gs

~---­

21Tl03

c,

Because of the compensation described above, Cc is very
small and can be negative Thus, the factory adjustments
assure that the rnaqnitude of Do is very large for any
G·standard and G·mu Iti plier combination you chose
Because it is extremely difficult to obtain external standards of conductance with zero or known shunt capacitance,
we do not recommend any readjustment of the "C"
compensation in the G box, i.e.. do not turn any of the
sleeves, do not replace any resistor passing through a
sprinq-loaded threaded sleeve in the G box. C18 and
C20.
C27 are "factory adjustments"

4.4.5 External Standards.
An external standard, or whatever you connect to that
port, is connected like a capacitance standard (a large one)
The LOW terminal is grounded, when EXT MUL TIPLIER is
set to OFF, and otherwise connected to the detector summing point. The HIGH terminal also connects to ground in
the OFF position, as well as in the
position
HIGH
connects to the ratio- transformer in steps of 20 turns
Unlike the CAPACITANCE
lever switches, the EXT
MU LTI PLIE R indicates directly the fraction you select of

a

the external standard (in farads and Siemens). Thus. the
settings of
.1, 0, 1,2, .9,
1.0 indicate (respectively)
that - 0.1, 0, 01, 0.2,
09, 1 times the external
standard C and G are added to the CAPACITANCE and
CONDUCTANCE indicated by the lever switches.

4.4.6 Zero Adjust.

Figures3-2, 6-3.

This front panel adjustment is primarily intended to
compensate the bridge for a few pF of terminal capacitance
when you chose a 2·terminal position of the TERMINAL
SELECTOR switch The adjustable component is the pot

R24, connected across 2 turns of the ratio transformer (on
the standards side), with its wiper arm capar.itively coupled
through C17 to the detector summing point. Therefore, a
capacitive balance can be made without moving the Clever
switches and a "zero offset" can be chosen so that the
readout is zero for some convenient reference condition in
2·terminal measurements. (See para 3.4.2.)
A second function of the ZERO ADJUST control is to
compensate the bridge for a few aF of variation in effective
zero offset among the CAL, 3·TERMINAL
Xl, and XlO
positions of the TERMINAL
SELECTOR switch. (In the
latter 2 positions, the 3·TERMINAL
UNKNOWN L.OW
connector must be shielded with a Type 874WN OpenCircuit Termination to remove the 125 aF or so of termi nal
capacitance from the bridge) The circuit from R24 through
C17, although it is now disconnected from the detector and
grounded by the TERMI NAL SE LECTOR switch, neverthe
less couples enough signal to the summing point for the
purpose described before (Notice that the range of adjust
ment is now reduced by a factor of 10 6 or so.) A coarser
adjustment, C301, very lightly couples some signal Of the
opposite phase to the detector (via the G box) and is
provided so you can make sure that the ZERO ADJUST
control, sornewhere in its range, reaches the desired zero
condition for each of the first 3 positions of TERMINAL
SELECTOR. (Each position will require a different ZERO
ADJUST setting)
The obvious utility of this second function is to make
the CAPACITANCE readout true, down to the 12th digit,
so you need not add a "zero correction" to each measurement. However, accuracy in that digit is insignificant for C,
above about 100 fF (01 pF) Therefore, only if you must
measure such small capacitors will you need this function
regularly. It does have utility in recalibration of the bridge,
particularly in trimming the smallest internal C standards
with the larger ones as references.

4.5 C-STANDARDS ACCURACY.
4.5.1 Calibration.
The set of 12 internal capacitance standards can be
cal ibrated qu ick Iy and accurately by a series of comparison
balances starting with a single external standard capacitor
of almost any size within the range of the bridge. Since the
8 figure resolution of the bridge permits comparison with a
precision of 1 ppm down to
1 pF, the accuracy of
calibration is usually determined by the accuracy of the
standard.

a

Only one external standard is required, most conven
iently a 3 terminal 100·pF standard, such as the GR Type
1404·B.
With a test
frequencv of
1 kHI,
the
accurate, Internal 0 1 transformer ratio can be used to
ensure accurate decade ratios of the internal capacitance
standards The ~ 1 position of any capacitance balance
switch connects the corresponding internal capacitor to a
20·turn tap on the "unknown"
side of the ratio trans

THEORY

4·9

former. This capacitor can be compared with the next
lower decade capacitor, which is connected to the 200-turn
winding on the standard side when the corresponding lever
is set to the X position. Any adjustments required can then
be made with one of the trimmers accessible beneath a
hinged cover on the front panel Refer to para 54
Such checks or recal ibr ations of the bridge need not be
made often The capacitors are constructed to be so stable
that after cal ibration they may be expected to change less
than 10 ppm per year in normal use. The temperature
coefficients of the internal standards are stated in the
Specifications and the consequent effect of temperature on
accuracy is summarized in para. 3.9.

4.5,2 Sealing.
The "air-dielectric"
standards, except the 4 smallest, are
filled with dry nitrogen and sealed. If they were open to the
atmosphere, their capacitance would change 20 ppm for
each 10% change in humidity and about as much again for
each 300 meters of elevation or 3% change in barometric
pressure because of the weather. Sealing practically eliminates these infl uences.
The 2 largest capacitors, with selected mica dielectric,
are also sealed to keep them dry and stable with respect to
the effects of humidity

cient a is therefore

6C

RM

Example 1 what is the fractional change of the
3rd-decade internal capacitance standard during a 5-minute
measuring process, if the environmental
temperature is
known to be maintained in the range 23° ±l°C) The worst

°

possible case is 0 = 22°C and 60 e = 2°C; From the
specifications, a representative value for a is 3 ppmtC; the
time constant for the lag box in the bridge is 360 minutes
(6 hours) So

6C
C

= (3

ppm) ( 5 min I (2° C) EO
°c
360 min

=

083 ppm

Example 2 What is the effect of a change, 18 hours ago,
from 10° to 23°C)

4.5.3 Thermal Lag.
The 8 largest internal capacitance standards are housed
in a "lag box," a thermally insulated container that comprises most of the "C box," within the bridge For the
purpose of estimating transient temperature effects on precision of measurements, we represent the thermal properties of the lag box simply as follows
1 The mass M inside the box is all at temperature
j
2 Heat is transferred through massless insulation of
thermal resistance R
Then the lag box will respond to a step change 60 e in
to 0) at time zero
environmental temperature (from
0
with a simple, exponential change of internal temperature
0, as follows (A close analogy exists to a series R-C circuit.
capacitance being analogous to M, applied vokaqs to 0, and
the capacitor voltage to OJ For the electrical circuit. the
time constant IS RC, for this thermal model, RM )

°

°

where E IS 2718
and t IS time (in the same units as RMI
Consider rate of change of inside temperature at any

t I me t

d
60 e
dt OJ = RM

4·10 THEORY

During any time interval 6t, that is very small compared to
the time constant, the internal temperature change is that
rate of change multiplied by 6t The corresponding fractional C change for a capaci tor with a temneratu re coeffi-

RM
E

4.6 G·STANDARDS ACCURACY.
The basic conductance accuracy of 01% depends on
precision resistors and multiplier
networks that can be
calibrated against external standards The very high degree
of independence between G and C balances depends on
factory-set compensation in the G box, measured in terms
of 0o, and the minute losses in the C standards, which are
specified
As indicated in the parts list, the greatest accuracy is
provided in the standards for the first 2 positions of the G
readout, R 1 and R2 The nex t, R3, has a tolerance of
±O 1%; the following, R4 and R5, of ±1% Intercomparisons
can easily be made using the bridge ratios, in a manner
simi lar to the cal ibration of C standards You can measure
the error in R5, for example, (against R 1 or R2) but no
calibrating adjustment is provided
Adjustments for the multiplier network are described in
para 5.4. ,Essentially, one makes the bridge balance with
the correct G readout while a standard external conductance (or network) is connected to the UNKNOWN port
The factory adjustments make 00 at least 103 for any G
readout This means. for example, if the unknown capacitor

has a 0 of 104 , the 1kHI error in C measurement due to
stray capacitance in the G standards will not exceed 0 1
2
ppm However, if the unknown conductor has a 0 of 10 at
1 kHI, the error in C measurement may be as much as 10%
Refer to para 39 and 44 As a rule of thumb, at 1 kHz

DC
C

1
00 0

o
Do

that approximation is invalid for small numbers of turns
CG represents stray ground capacitance that is not associ
ated with the transformer, but appears in parallel with CG T
at any given setting of the switches Typical parameters are
given in Table 4-1
These residual impedances make the voltage V applied to
the capacitor C, or C; differ from the voltage E induced in
the transformer winding For small residuals, the relationship between these voltages is

4.7 RATIO ACCURACY.
To measure accurately with this bridge, you must
properly balance a well defined unknown capacitor against
accurate standards via accurately known ratios Section 3
deals with system operation; 42, definition of the device
being measured, and 4.5 accuracy of internal standards The
following paragraph deals with the accuracy of significant
ratios in the bridge and their effect on your measurement
The transformers in the Type 1616 Precision Caoacitance Bridge are not qu ite the ideal transformers shown, for
example, in Figures 1-1 and 3-3 The resistance, leakage
inductance, and capacitances of the ratio-transformer
windings, which are assumed to be zero in the simplified
bridge theory, have been kept sufficiently small in the
instrument so that errors from these residual impedances
are less than 10 ppm for capacitances up to 0 1 IlF at a
frequency of 1 k HI However, the residual impedances
make the voltages at the transformer terminals differ
slightly from the voltages induced in the windings and
produce bridge errors that increase with frequency and with
the rnagnitude of the measured capacitance
The accuracy of the ratios when the transformer is
lightly loaded is better than 1 ppm for the unity ratio and
is better than 3 ppm for the 0 1 ratio at 1000 HI or lower
frequencies The winding self-capacitances act as a more
significant load as frequency increases, so that the error in a
0.1 ratio increases to about 30 ppm at 10 kHz and to 05%
at 100 kHz The phase errors are, in general, somewhat
larger than the magnitude errors of the ratios At 1000 Hz,
the phase error is probably within ± 10 11radians, but the
error increases in approxirnate proportion to ratio and to
the square of frequency

4.7.1 Residual Impedances.

where CT = C + CG T + CG, and we use add itional subscripts s or x to designate the standard or "unknown" arm
of the bridge
For convenience we postulate "effective" values C' of
the main capacitor C in each arm (Figure 4-8) C', and C'x
are related by the simple balance condition of para. 4.3,
ie, they are inversely proportional to the turns ratio. From
the circuit of Figure 4-10, for small residuals,

and subscripts x or s can be used as before Residuals on the
C, and your readout rather low;
standard side make C's
C x and your
residuals on the "unknown" side make C' x
readout tends to be high. There is Iikely to be some cancellation of errors, particularly when the turns ratio is 1 1

>

>

t
I

E

I

E

~

GE"I[RATOR

Figure 4-10. Simplified bridge schematic diagram
showing residual impedances of the ratio transformer.

Figure 4-10.

The simplified bridge diagram of Figure 4-10 is similar to
that of Figure 3-2 but shows, in the equivalent circuit of
the transformer, winding resistances (r), leakage inductances (Q), and winding capacitances (C G T ) in both the
standard and unknown ratio arms The magnitudes of these
residuals vary on the standard side as the lever switches are
moved, and on the unknown side as the TERMINAL
SELECTOR is reset. The resistance r and inductance Q are
approximately proportional
to turns in the ratio transformer itself However, because we include the series
impedances of intimately associated wiring and switches,

----------

Table 4-1 ----------

APPROXIMATE

MAGNITUDES

OF RESIDUAL

PARAMETERS

Transformer
turns

r

2

06£1

31'H

08

4

10 nF

500

5

1 nF

150

7

10 pF

75

10

100 fF

100

1 fF

35

6
20

15

60

35

200

105
Note

Q

CGX';:; 10 pF

CGT

Standard
capacitor
100 nF

300 pF

CGS
850 pF

THEORY 4-11

The difficulty
in determining the residual parameters for
the many switch combinations makes impractical the detailed calculation of errors for correction of most measurements However, the uncertainty or worst-case errors are
presented in para. 3.9. Also, if the capacitance being measured is invariant with frequency, you can determine the
error due to residuals for any given measurement with a few
ancillary measurements and calculations
(Notice that the
error term is proportional
to frequency squared.l More
convenient formulas are given below.
An effective conductance G' can be defined, analogous
to C'. G' is larger than G by a term that is a hypothetical
conductance adding a component of current to the detector
summing point (under the ideal conditions of zero rand
zero Q) equal to the true current component that flows
through C in phase with E (because of r ) We assume the
"unknown"
is a capacitor: a slightly different analysis
would be appropriate if it were a resistor.

Subscripts x or s can be used, as before
By definition of the primed symbols, they are related by
the turns ratio. Usually, only the most significant standard
is considered:

C' x
C' s

residual errors from
small residuals

C

=------

x

N s C's

~ +w

C error

2

Assuming

NsC"s

N x(l + w2QxCTxl

C"s = C s

The fractional

both arms of the bridge

(Qs C T s

Nx
QXCTxj

is [the increment

differs from C"sl/C"s' for simplicity,
C, in the denominator

by which Cs

we replace C", with

Notice that the coefficient of w 2 can be determined
experimentally, for any given C; and switch settings for
multipliers and most significant digits, by mak ing 2 measurements at different frequencies, assume C"s constant
The effects of residual impedances on accuracy are
summarized in 3 observations
1. The effective value of the internal C standards is
mu Itipl ied by (1 + k f2) where k is constant as long as the
ratio-transformer switches are not changed.
2 In general, k has either sign, and its magnitude is
within the bounds indicated graphically in para 3.9
3 Because of the factor f2, the residuals cause the
predominant error at high frequency, above 10 kHz
Other errors predominate below 2 kHz

However, if more than one standard of similar significance
is involved, the relationship is generalized as

4,7.2 Examplewith 1:1 Ratio.

An alalogous formula can be used for G
However, if we assume for simplicity that only one G
standard is being used, we can express G and the G error,
x
(readout - G) in Siemens, thus:
x

The errors in the bridge readings can be determined by the
measurement of a calibrated capacitor. A convenient standard of
capacitance and loss is a three-terminal air capacitor that can be
connected directlyto the bridge terminalsto add a minimum of series
inductance and resistance. For such a capacitor it can be assumed
with good accuracy that the capacitance and loss have negligible
changeswith frequencyup to 100 Hz and that any changesin bridge
readings with frequency indicate bridge errors. Ahhough the
dissipationfactor is not generallya simple function of frequency, in a
cleanair capacitorthe magnitudeshould be sufficientlylow, i.e., in the
tens of ppm, that it can also be neglected.

6G

Assuming, similarly, only one C standard is being used,
we write the following expression for C, and define a new
effective capacitance standard C", which incorporates the

4·12 THEORY

The tabulated results were obtained when such a 1000-pF
Capacitor was measured on a 1621
CapacitanceMeasurement System; see Table 4-2.
Note that the apparent capacitance change at 100 k Hz is
only 200 ppm, as compared to the estimated standards-arm

----------

Frequency

10 Hz
100
1 kHz

Table 4·2 --------ERROR EXAMPLE. RATIO 1.1

X00500

G readout

Relative error

C readout

0035S

+7 ppm

pF

0.136

+2

XOO.495
XOO.493

0

5
15X 102
10-

(--1)240

10- 5

10

XOO.488

-5

(-1)884

10- 2

50

XOO.385

-108

1-1)608

io"

100

XOO.164

-329

1--117XO

1

error (W 2 £sCTs) of 5650 ppm The smaller measured error
results from a partial cancellation of transformer-impedance
errors in the 2 ratio arms of the bridge. When. as in this
example. the ratio is 1 1. the residual rand £ are equal on
both the standard and unk nown sides of the transformer in
Figure 4-10 If the total load, CT' is also the same on both
sides. the coefficient (£sCT s - £xCTx) is zero and C, = C",
= C, at any frequency
In this example. and in general, the errors in the two
bridge arms do not quite cancel because the total C T on
one side does not equal that on the other The internal and
external stray capacitances to ground. CG 5 and CG x' are
seldom equal, since the bridge shields and wiring make CGs
Cs and the
relatively high. Therefore. commonly. C",
bridge readout is low For example, other things being

>

equal. if CGs - C Gx = 100 pF, the bridge would read low
by 1600 ppm at 100 kHz
This kind of error. being dependent on the internal
ground capacitances. is altered by any rearrangement of
decade switches to select alternative internal standards. In
the following example, the same 1000-pF capacitor is
measured with 2 of the several possible decade settings
1 kc

X00498
9X0498

100 kc

pF

XOO 163 pF
998686

For a bridge ratio of 1 1, therefore. the
err ors In
C and G are typically less than. say, 1000 ppm at 100 Hz
Both errors are proportional to frequency squared I t is not
usually practical to apply corrections for these errors to the
bridge readings, chiefly because the magnitudes of the stray
capacitances to ground and their variations with switch
settings are not easily determined

4.7.3 The 10:1 Ratio.
Bridge errors from transforrner residuals can, in theory,
be reduced or eliminated by symmetry in the bridge rirr.uit
for any transformer ratio, Just as in the example of the 1 1
£sCT s or £x/£s
ratio All that is required is that £xCT x
CT/C T x to make C, = Cs ' hence, the residuals should be
proportional to the ratio
In practice. there are several reasons why the errors.
2
2
W £CT or w rCT Cs' cannot be kept the same on both Sides
of the bridge as the ratio is increased The residual impedances rand £ (Table 4-1) differ slightly from proportionality
to ratio, even when that is small For high ratios, the rand £
on the low Side can never be less than the minimum values
set by the wiring, switch, and terrninal impedances Any
wires used to connect the unk nown capacitor to the bridge
also increase these residual impedances The capacitance
residuals also are seldom proportional
to ratio Although
the bridge ratio is determined essentially by the ratio of the
unknown
and standard capacitances. Cx/C s' the error
depends upon total capacitance in the bridge arm, ego C T
= C + CG T + CG Neither the transformer wind Ing capaci
ranees. CG T' nor the ground capacitances of the capacitors,
CG , are proportional to ratio.
Generally, for a ratio of 10 1. the bridge errors In C may
be as high as 03% at 100 k Hz The errors are dependent
upon decade settings and stray capacitances in the bridge
and also outside, so that corrections are not easiIy made

4.7.4 The 1: 100 Ratio.
At 1 kHz. a setting of 9X was equivalent to XO. as it
should be At 100 k Hz , not only was the measurement in
error because of residual impedances. but the change from
XO to 9X resulted in a new balance requirement so the final
readout was 99 In fact the two 100-kHz measurements
differed from each other by considerably more than the
original error due to frequency The read ing was lower for
the second 100-k Hz measurement than for the first. because the capacitance C G s was larger With the XOO setting.
the internal 1000-pF standard and its ground and wiring
capacitance loaded the transformer predominantly;
with
the 998 setting. both the 100-pF and 10-pF standards were
also connected and the ground capacitance was thereby
increased about 170 pF. Also, several other capacitors were
switched to higher taps in the final balance. This example is
nearly an extreme case.

A typical situation in which you would use a 100 1 ratio
is measuring a capacitor larger than 1 pF The large ratio
generally means a large bridge error resulting from the
residual impedances, but also simplicity in their effect and,
therefore, a chance to correct your measurement by calculation
Capacitance to ground from the UNKNOWN
HIGH
terminal, due to the capacitor attached there, can be
measured (with another bridge) or estimated That and CG x
(Table 41) are usually negligible in determining
the
dominant factor C T x in the error

THEORY

4-13

Notice that this error is positive (readout is too high), that
it is fractional (a dimensionless percentage, not a number of
picofarads), and that it is proportional
to frequency
squared, the value of capacitor measured, and the total
series inductance (including transformer leakage and those
components of \\ in cables outside the bridge). Whether Qx
includes also the series inductance within the structure of
the unknown capacitor depends on definition. If not, then
C, is an effective "terminal" capacitance and is liable to be
frequency dependent. If yes, then Cx is the series capacitance within that structure, and may be essentially independent of frequency.
The G error is more conveniently treated as an additive
term (mhos) rather than a fractional error Because of the
sense of the large ratio and the limitations of resistance in
switches and connections, the term containing r x predominates, thus

described, another source of error at high frequencies is the
voltage induced in the internal bridge wiring connected to
the detector circuit by currents flowing in the bridge
circuits connected to the transformer The bridge does not
have the short, coaxial current paths required for radiofrequency accuracy, and mutual inductances of the order of
1 IlH between bridge arms may be expected Since the
currents drawn from the transformer by the bridge capaci
tors increase with frequency (i = wCE) and the voltages
induced in the detector circuit are proportional to wMi,
these error voltages, e = w 2 MCE, increase with the square
of frequency The errors depend also, in a complicated way,
upon the internal capacitors used, the transformer voltages
applied to them, and therefore the front-panel switch settings. Experiments confirm, however, that these errors
appear not as a percentage of the measured capacitance but
as a more-or-less constant picofarad error or C offset at any
one frequency The order of magnitude of the error is 100
aF at 100 kHz

a

4.7.6 Ground-Circuit Impedance.

Again, this error is positive (readout is high) However, it is
a number of Siemens, not a ratio. It is proportional to frequency squared, C; 2, and to the total series resistance
including those components of r x in cables outside the
bridge. Whether r x includes any lossy component inside the
capacitor under test depends on your definition. At any
one frequency, the conductance can be expressed as an
equivalent series component Rs' as shown in para 3.6.
Usually that is not included in r x However a known series
resistance (often called "ohmic" resistance or d-e series
resistance) may be treated as part of r x and thereby
excluded from your measurement
As a rule of thumb, calculated corrections of the types
given above (for C and G, for 1·100 ratio) are accurate to 5
or 10% Use them to reduce the error of high-frequency
measurement an order of magnitude; but if frequency is 30
k Hz (for example) do not ex pect corrected measurements
to be as accurate as l-kHz measurements

4.7.5 C Offset Due to Induction.
In

addition

4-14 THEORY

to

the

residual

impedances

previously

Another source of error, which may be significant when
you measure a 3-terminal capacitor remotely at high
frequency, is the finite impedance of the path between the
shield of the capacitor and ground in the bridge This path,
usually through one or both shields of the cables connecting the capacitor to the 3-TERMINAL UNKNOWN port,
has series Rand L components ZSg = RSg + jwL Sg (Figure
3-1) Currents through ground capacitances Cis and
return through ZSg The effects on the measurement are
negative errors in both C and G; thus the. correction is
additive

c.,

The error is most commonly noticed when Greadout
is
negative, both because it is often the larger of the two
components and because negative loss in a capacitor is an
obvious indication of the presence of error. An example is
given in para 3.9.

ServiceandDiagrams
-Section5
5.1
5.2
5.3
5.4
5.5
5.6

GR FIELD SERVICE
.
MINIMUM PERFORMANCE STANDARDS
DISASSEMBLY
.
RECALIBRATION AND ADJUSTMENTS
TROUBLE ANALYSIS
PARTS LISTS AND DIAGRAMS
Exterior Views (Mech parts)
Federal Manufacturing Code
Figure 5-9. Schematic Diagram
Appendix A
.

5-1
5-1
5-1
5-7

5-8
5-9
. 5-11
.5-13
.5-15

. A-'

11

5.'

WARNING
Dangerous voltages may be present inside this case. These
instructions
are intended for the use of qualified service
personnel only, to avoid electric shock, do not perform any
servicing other than that contained in the operating instructions
unless you are qualified to do so.
GR FIELD SERVICE.
dures. except as noted, set the front-panel controls initially

Our warranty (at the front of this manual) attests the quality of
materialsand workmanship in our products. When difficultiesdo
occur, our service
are available for technical
telephone assistance. If the difficulty cannot be eliminated by
use of the following service Instructions. contact the GenRad
Service Department in Concord. MA giVing full information of
the trouble and of steps taken to remedy it. Describe the
instrument by type serial, and 10 numbers. (Refer to front and
rear panels)

Instrument Return.

When returning an instrument to
GenRad for service. please identify the failure mode as
accurately as possible and include this information with the
instrument. For instruments not covered by the warranty. a
purchase order should be forwarded to avoid unnecessary
delay.
For return shipment, please use packaging that is adequate
to protect the instrument from damage. i.e.. equivalent to the
original packaging. Advice may be obtained from the GenRad
Service Department in Concord, MA

5.2 MINIMUM PERFORMANCE STANDARDS.
5.2.' General.
The equipment, methods, and criteria for verifying the
specified performance of the 1616 Precision Capacitance
Bridge in the 1621 Precision Capacitance-Measurement
System are given below. Recalibration is described in para
5.4. If performance is grossly inadequate, erratic, or cannot
be corrected by the adjustments, refer to trouble analysis,
para. 5.5.
Equipment needed for the measurements and procedures
of this section is listed in Table 5-1. For all these oroce-

to the standard positions, as follows
Oscillator (1316)
POWER - on (after preliminaries, para. 3.1).
FREQUENCY - 101 kHz
OUTPUT VOLTAGE RANGE - 15 V
OUTPUT ADJ UST - 10 on lower scale of meter
Detector (1238)
POWER _ ON (after preliminaries, para. 31)
FREQUENCY _ 101 kHz
TIME CONSTANT - 0.3 S
FIN E ADJUST - each control at midrange
PHASE SHI FT - 180° (Large dial at O~)
SENSITIVITY
maximum (cw)
GAIN
40 dB
COMPRESSION - disabled (button out)
LINE REJECTION
disabled (button out)
Bridge (1616)
C MAX - down (3 windows closed)
CAPACIT ANCE - zero
CONDUCTANCE - zero
Conductance Multipl ier - )( 10-6
EXT MULTIPLIER - OFF
TERMINAL SELECTOR - CAL
ZeRO ADJUST - any position

5.2.2 Zero Setting, Offsets, and Sensitivity.
Verify that the capacitance offset Co can be set to zero
and measure the conductance offset Go' as follows. Sensitivity and precision are also verified.
a Perform the preliminary and functional checks,
tuning, and phase adjustments of para 3.1-3.3.
b. Install an open-circuit shield (Table 5-1) on the

SERVICE & DIAGRAMS

5-'

---------

Table 5-1 -------TEST EQUIPMENT

Item

Requirements
Stability: 20 ppm/year; temperature coefficient:
2 ppmrC; terminals: 3; value: 100 pF ±10 ppm.

Reference standard capacitor,
(GR1404A)

Standard resistor, 100 kn

Accuracy, : ±.01% at 23°C.

Precision metal-film resistors

10 Mn ±0.1%
1 Mn ±0.1 % (4 required)
10.2 kn ±01%.
100 n ±0.1%

DC Resistance

Basic accuracy: ±O.01 %. direct reading: range 1p.S to
1
6- digit resolution.

bridge

rn

(GR1666)

Binding-post adaptor

Fits 2- TERMINAL UNKNOWN port (G900
to
BANANA JACK 3/4" spacing

Patch cords

Doubly shielded coaxial cable with locking G874
connectors: length: 3 ft. (2 required)

Open circuit

Shielded mount for conductance network: G874
connectors: (2 required).

Shielded mount for conductance
connectors; (2 required).

Mount

connector).

network,

G874

Table 5-2
CAPACITANCE RATIO CHECKS
TERMINAL
SELECTOR
3-ter x1
CAL
CAL
CAL
CAL
CAL
CAL
CAL
CAL
CAL
CAL
CAL

5-2 SERVICE & DIAGRAMS

nF

pF

·.

XOO
NXO
'NX
'ON
'00
'00
'00
'00
'00
'00
XOO
000

••

·.
'N
NX

fF
see 5.2.3c
000
000
XOO
NXO
ONX
000
000
000
000
000
000

aF
000
000.0
000.0
000.0
000.0
000.0
XOO.O
NXO.O
ONX.O
OOn.X
000.0
000.0

Tolerance
see 5.2.3d
500aF
50aF
10aF
5aF
5aF
50aF
5aF
5aF
2aF
5fF
50fF

Approx
GAIN

Trimmer
Label

70db
90
110
120
130
130
130
130
130
130
50
30

1nf
100pF
10pF
1pF
100fF
10aF
1aF
100aF
10aF
1aF
10nF
100nF

3-TERMINAL

LOW

connector.

Set

the

TERMINAL

SELECTOR to 3-TERMINAL
X 10.
c. Balance the bridge precisely, with increased GAl Nand
(if necessary) voltage, as described in para. 35, using the
lever switches for G but only the ZE RO ADJUST control
for C. Leave the C readout at zero, except for temporary
changes of a step or two in the last digit to test whether
you

have sufficient

sensitivity

for

9-figure

resolution

Record the G readout as Go' the conductance offset It
should not exceed 02 X 10-6 p.S in magnitude, and may be
either + or d. Sensitivity is normally sufficient so that an oscillator
voltage of 100 V, GAIN setting of 130 dB, and SENSITIVITY

set near midrange result in a deflection

of 25 (5

divisions) for 5 steps of the last C lever switch, near balance
e Repeat step c with the Terminal SELECTOR set to
each of its 6 positions, the final one being 3 TERMINAL
X 1. The bridge is now set for Co = 0 ±O 1 aF, on 3TERMINAL

Xl, and Go = a known offset

Note that the sensitivity

of ZERO ADJUST

is variable

Its setting is significant in 2-terminal measurements, but
3-terminal measurements can be made to specified accuracy
without regard for how this control is set. A rear-panel,
screwdriver adjustment is required if the range of ZERO
ADJUST is inadequate for balance in step e.

5.2.3 Capacitance Accuracy.
Verify that the absolute van.e of one of the prime
internal standard capacitors has been trimmed within specification, as follows
NOTE
If the required external standard is unavailable, it is still
possible to make a calibration verification with high
confidence by making only the Capacitance Ration Checks of
paragraph 5,2.4.
The internal standards are all of high
stability so that checking all their ratios will almost always
detect any excessive errors. The chance of all the standards
changing by the same amount (in %) is very small.
a. Stabilize the temperature of the bridge at 23° ± 1°C
r 24 hours previous to and during the following steps of

ra. 5.2.3 (36 hours is required after the bridge has been
bove 45° or below O°C; refer to para 46)
b. Remove any shields from the 3- TERMINAL UNKNOWN
erminals and connect instead an external reference standard
ith a nominal value of 1000pF known to an accuracy of
Oppm or better. (Note: the following instructions pertain in
etail to the use of a 1000pF standard but a 10 or 100pF
tandard may be used if the calibration principle is
nderstood,

c. Set the readout to the exact value of the 1000pF
standard, starting with an X. For example, if the value is
1000.016pF, set XOO.016pF. If the value is 999.989. set
Set first row of Table 5-2.
XOO.N89 where N is -1.
d. Verify that the final C readout is not more than 10
ppm different from the absolute 3-terminal value of the C
standard I f that is known to ± 12 ppm, for example, the

acceptable

tolerance on this measurement

5.2.4

Capacitance

Ratios.

The following procedure checks each internal standard
against the next higher. or lower. standard starting with the
100pF standard being checked against the Just-calibrated
1000pF standard as indicated in the 2nd row of Table 5-2.
This and the next eight checks work down to the lowest
standard. The last two checks work up from 1000pF to
.1 uF. (If a 100pF or a 10pF standard was used as the
external standard. then this ratio procedure would start further
down on this table and checks should be added that go up
from the value of the standard used to values skipped.) The
procedure for checking all other standards against the internal
1000pF standard is as follows.
a. Set the TERMINAL SELECTOR to CAL. Be sure EXT
MUL TIPLIER is set to OFF. Set C readout to all Zeros with
first three digits masked using the C-MAX switch.
b. Balance the bridge with ZERO ADJUST and the G
levers. Use the lowest G multiplier (10-6).
Get negative G
values using a -1 setting of the smallest digit possible. Leave
the ZERO ADJUST as set You now have Co < 0.1 aF for
bridge configuration set
c. Set up each of the C-readout combinations starting
with the second row of Table 5-2.
For each. balance the
bridge using the smaller C lever switches and the G controls
as required. Verify that each balance is achieved within the
tolerance limits indicated.
Here * indicates a zero digit
covered by a mask as set with the C- MAX lever and N
indicates a setting of -1,
If the detector phase adjustments are made as suggested
by the Condensed Operating Instructions. the IN-PHASE
meter may be zeroed with the C levers and a precise G
balance is not necessary. This meter can also be used to
extrapolate between C lever steps to estimate the error. The
GAIN settings in the table are suggested as initial settings.
more gain may be required.

5.2.5 Conductance Accuracy.
The following

procedure makes use of 2 external stand-

ards to check the 5 basic internal G standards (An equally
valid check could be made by measuring each with a d-e
bridge, a procedure which would require some disassembly
and unplugging of connections

NOTE
The accuracy of a suitable standard depends on
the certified accuracy when calibrated by NBS
or a standards laboratory, its stability and the
time elapsed since calibration,
it temperature
coefficient and the temperature, etc

is ±22 ppm, If

±3, then ±13
e Remove the external standard

to the "G box 'OJ

&
CAUTION
Restrict the oscillator level to 20 V rms or less
when the TERMINAL
SELECTOR is set to
2-TERMINAL
Xl; otherwise safety diodes will
distort the test signal and invalidate the balance
condition.
SERVICE & DIAGRAMS 5-3

At high settings of GAIN, movement of your hands or
other objects near the unshielded 2-TERMINAL
UNKNOWN port will upset the balance. Keep hands away and
nearby objects still.
a. Install an adaptor on the 2-TERMI NAL UNKNOWN
port and plug in a 100-kQ standard resistor (Table 5-1) as
the first G standard.
b. Reset the frequency to 102 Hz, the oscillator level to
10 V, and tune the detector and oscillator to match as
before. However, reset the PHASE SHI FT in the process of
balancing the bridge, as described in para. 3.5 under Final
Balance Phase Shift/Reset.
NOTE
Reset the phase shift, as required, as you approach balance in any step of the procedure:
para. 5.2.5 and 5.2.6.

The need to do so is recognizable when both phasesensitive meters respond to a change of either C or G lever
switches. (The inconvenience of having to reset PHASE
SH IFT is to be expected only for precise measurements of
very lossy capacitors, not for most measurernents.)
c. Set up, in turn, the combination of G standard,
TERMINAL SELECTOR setting, and G readout for each of
the first 5 rows in Table 5-3. Complete steps d through f for
each setup before proceding to the next.
d. Start with a convenient GAl N setting. Balance the
bridge using the C lever switches, increasing the GAIN to
the "dfs" number tabulated. (Reset PHASE SHIFT as required.)
e. Adjust the SENSITIVITY
so that a change of 5
divisions on the QUADRATURE meter corresponds to the
percentage change tabulated as "Tolerance". Notice that a

External
Standard

100 k
100
100

n

10 M
10

100 k
100

CONDUCTANCE
Terminal
Selector

o

1% change is the step from XOO to X01, or XO( -1)
Similarly, a 1-% change is from XO to Xl, or X(-1) A 10-%
change is from X to 9. Change the GAl N if necessary.
f. Verify that the internal G standard is within th
specified accuracy. In other words, the QUADRATUR
meter should read between the points labeled 25 when the
readout is set as tabulated. (The pertinent standards are R 1
through R5, in that order.)

5.2.6 Conductance Multipliers.
A procedure very similar to the preceding one is used to
check the G multiplier ratios (steps of 10-1). In this paragraph, the maximum accuracy (0.1%) is required for every
step. Unlike the internal G standards, the multipliers can be
trimmed by you, the customer, if you find they fail to
meet the following specifications (para. 5.4)
a. The setup is much the same as before. Initially,
reinstall the lO().kQ standard resistor at the 2-TERMINAL
(see below) to the
port and connect network "A"
3-TERMINAL
UNKNOWN port, using shielded patch
cords.
b. Networks"t>:'and "B" are 3-terminal conductance standards
that you can assemble in a few minutes. Each is a tee of 3
resistors,Figure5-1, enclosedin an electric shield. Use a G874-X
as an enclosurefor each. Measurethe resistanceof each resistor,
after soldering, at room temperature, to an accuracy of ±.03%.
Use a precisionresistancebridge, such as the GRType 1666. For
convenience in making 4-terminal "Kelvin" connections to each
resistorat the test fixture of such a bridge, leavethe resistorpigtails
full length. (After the measurementthey can coiled up for future
use or clipped off.)
c. Calculate the 3-terminal conductance of each network
you have made, thus:

Table 5-3
ACCURACY AND MULTIPLIER CHECKS
CONDUCTANCE
Approx
us
ns
Mul
GAIN
Tolerance

000.0

Trimmer
Resistor

2- TERM X1
2- TERM X10
2-TERM X100

X
0
0

XOO,O
OXO.O

50 dB
70
90

±0.1%
0.1
0.1

none
none
none

n

2- TERM X10
2-TERM X100

0
0

OOX,O
OOO.X

80
90

2
10

none
none

n

X
X
X

000.0
000.0

10-1
10-2

CDE.F

10-3

60 dB
70
80

±0.1%

"A"

2- TERM X10
2- TERM X100
3- TERM X1

R18
R19
R20

"A"
"B"
"B"

3- TERM X10
3- TERM X1
3- TERM X10

X
X
X

CDE.F
GHI.J
GHI.J

10-4
10-5
10-6

The letters C thru J represent
described in paragraph 5.2.6c.

5-4 SERVICE & DIAGRAMS

90

R21
R22
R23

110
120

digits calculated

as

NOTE
To separate the bushing from the knob, if for
any reason they should be combined off of the
shaft, drive a machine tap a turn or two into
the bushing to provide sufficient grip for easy
separation.

Here, C, D

J represent numerical digits, of which F and
Notice that X (ten) is the preferred

J are insignificant

most significant digit in following procedure For example,
if GA IS 99765 n ,express it as X (1 )765 nS
d. Set up, in turn, the combinations of G standard,
TERMINAL SELECTOR setting, and G readout for each of
the last 6 rows in Table 5-3. Complete steps e through g for
eachsetup, before proceding to the next
e. Balance the bridge, as before
f. Adjust the SENSITIVITY as before, for 0 1% change
equals5 divisions
g Verify that the most-significant G standard multiplied
y the selected G multiplier is within ±O 1% of its nominal
alue In other words, the QUADRATURE
meter should
ad between the points labeled 25 when the readout is set
tabulated.
Ra

5.3.2 Cabinet Removal.

~
i
"8"

&

NOTE
Keep POWER OF F during disassembly or reassembly.

Rb

",.::t
NETWORK

To replace a knob
a. Slip bushing on shaft and rotate to correct position as
observed in disassembly of knob.
b. Keep bushing away from panel by at least the thickness of a filing card Pull it out farther if necessary to
prevent tip of shaft from protruding.
c. Tighten the setscrew in the bushing.
d. Place knob on bushing with retention spring opposite
setscrew.
e. Push knob on until it bottoms and pull it lightly, to
check that the retention spring is seated in groove in bushing.

Rf

Figure 5-1. Three-terminal conductance

standards. Ra

= Rb = Rd = Re = 1 Mn. Rc = 10.2 kn. Rf = 100 n.

.3 DISASSEMBLY.

.3.1 Knobs.

&
CAUTION
Do not use a screwdriver or other instrument to
pry off the knob if it is tight.

To remove the knob from a front-panel control, either
replace. one that has been damaged or to replace the
ociated control, proceed as follows
a. Grasp the k nob firmly with dry fingers close to the
nel and pu II the knob straight away from panel
b. Observe the position of setscrew in bushing when the
antral is fully ccw.

a

e. Release the setscrew with an Allen wrench; pu II the
ushing off the shaft.

Figure 2·3.

To remove the bench-model cabinet from the bridge,
first set the instrument in the horizontal position, free of
unnecessary cables, and proceed as follows
a. Remove the 4 dress-panel screws (A) accessible
through holes in the handles
b. With caution not to let the instrument drop out of its
cabinet, turn it face down to rest on the handles.
c. Pull the cabinet up and off. Carefully return the
instrument to a horizontal position.
To remove each instrument from the 1621 system
cabinet or each rack-mounted instrument from its cabinet,
apply step a, above. Then withdraw the instrument forward
carefully as described in para 24 and 25 .

5.3.3 C·Box Removal.

Figure 5·2.

The subassembly containing the capacitance standards
comprises a substantial part of the instruments' weight and
volume Although it is not to be repaired, it can be removed
as follows, either for replacement or for access to miscellaneous wiring.
a With the bridge right-side-up, outside its cabinet, remove the upper 3 screws (of a total of 5) that are about %
in. from the rear edge of the right side panel
b. Remove the corresponding 3 from the left Also on
the left side, remove the 3 top screws, 3 more 6 in. below
those, and finally 2 more for a total of 11 that form a
rectangle.
c. Slide the C box back only '% in, so it has support, and

SERVICE & DIAGRAMS

5·5

disconnect the 3 BNC connectors behind the EXTERNAL
STANDARD port
d. Slide the box back about 3'12 in. and support it by
hand while disconnecting the edge connector from the
etched board and the 2 wires that are separately clipped
near the middle of the bottom edge of that board
e. Lift the C box out toward the rear.
f. In reassembly, be sure to bring the C MA X gears into
mesh as nearly right as possible. That is, preset the gear on
the C box to the ccw position (of 4 detent stops) as seen
from the left; preset the C MAX lever down (all 3 windows
blanked)
g Connect the white wire (left) and black (right) first;

then the etched-board connector; then the 3 BNC connectors in this sequence violet (bottom),
(top)

,

red

h. Reposition the idler gear associated with C MAX, i
necessary, by loosening, moving, and retightening the small
recessed screw in the left side panel. Moving it toward the
rear decreases the backlash but may cause binding, toward
the front makes the action more free.
i Slip the gear engagement (if the attempt in step f was
unsuccessful) as follows Tip the instrument onto its right
side, remove the 11 screws from the left side, and tilt the C
box by hand enough to disengage the gears while you move
the C MAX lever slightly for better alignment

C TRIMMER PANel

J7,,8,,9

gray (middle),

GEAR FOR S3

~,,~

~~
"':

~CBOX

C TRIMMER
ACCESS DOOR

.......J~--Jl0

WHITE (WT5)
RATIO
TRANSFORMER

<,

T1

<; BLACK (WT6)

C301
ACCESS

Figure 5-2. Interior view, upper rear, with the C box removed and shown above.

5-6 SERVICE & DIAGRAMS

•

G-MUlT TRIMMERS: R23, R2l, R20, R19, R1B, R22

~

_f

I

i~

I

FOR

S4

Figure 5-3. Interior bottom view. The G box cover is shown fastened

5.3.4 The G Box.

Figure 5-3.

This "box," containing the conductance standards, is the
most conspicuous subassembly in the bottom part of the
bridge behind the bank of lever switches Although the G
box is not to be repaired, it can be uncovered for access to
the G-multiplier adjustments and it can be removed if
necessary Use this procedure.
a Place the bridge upside down
b. For access to adjustments, remove the 4 cover screws
Move the cover to the position shown and replace 2 of the
screws through the 2 spare holes (nearest the edge) of the
cover. Adjust only with the cover attached
c. For removal of the box, make a note of the positions
of the 10 wires to be disconnected, something like this left
end, front-to-rear
red tracer, green tracer, blue tracer,
brown tracer, black tracer; rear panel, left-to-right
red,
violet, black, grey, brown.
d. Disconnect those 10 wires and the yellow-banded
BNC connector
e. Remove 2 screws in the rear panel (near, but not
holding, the BCD CAPACITANCE OUTPUT connector)
and 4 screws in the right panel.Iarnonq. but not holding, the
slide blocks) while holding the G box so it does not fall.
Lift it free
f. In reassembly, mesh gears properly by presetting the
switch on the G box ccw (as seen from the right) and the G

temporarily

multiplier

in position for trimming

lever "down,"

the G-multiplier

indicating

network.

1()6. If necessary, to

obtain proper mesh, try again
g. Adjust the gears for a reasonable amount of backlash
by moving the G box slightly, while you have the 4 rightside screws temporarily loose
NOTE
Do not reposition or replace any parts or turn
any adjustments in the G box except as instructed in para. 5.4. Otherwise Do will have to be
reset, a factory operation.
5.4 RECALIBRATION

AND ADJUSTMENTS.

Recalibration is principally the trimming of the internal
C standards for absolute accuracy and exact decade ratios
The G multipliers (but not the 5 standards) can be trimmed
also. Adjustments include C301 which affects the ZERO
ADJUST ranges and mechanical adjustments on the lever
switches.
The temperature of the bridge should be maintained well
within the allowable tolerance, just as the accuracy of each
adjustment should be greater than the acceptable accuracy
of the bridge For example, in addition to the temperature
requirements described in para 52, the bridge should be
held at 23° ±05°C for 6 hours previous to and during
recali brat ion

SERVICE & DIAGRAMS

5-7

For special uses, it is possible to recalibrate the bridge at
temperatures other than 23°C. The range of trimming provided for the internal C standards sets the upper and lower
limits at about 21° and 26°C, respectively
If you do
recalibrate at a nonstandard temperature, be sure to tag the
bridge with that information

5.4.1

Adjustment

of Internal Capacitance

Standards.

To recalibrate the one or all of the internal capacitance
standards, proceed as in paragraphs 5_2.3 and 5.2.4 but
make adjustments of the trimmers (listed in Table 5-.2) to
minimize the errors. Additional procedure is as follows.
a. Unlock and open the panel that covers the trimmer
screws (para. 13)
b. Obtain a small. sturdy screwdriver with an insulated
handle and shank (or enclose the shank in a plastic sleeve).

.&.

WARNING
Full oscillator voltage appears on the trimmer
screws.
c. Notice that each screw is labeld in terms of the
effective value obtained when the corresponding lever switch
is set to maximum. Thus, the 3rd screw is labeled 1nF; it
trims C100 which adds 1000pF to the standard arm when
100pF per step lever is set to X. To trim C100 against the
external 1000pF standard (para. 5.2.3) use the nF (or
1000pF) trimmer as indicated in the first line of Table 5-2.
d. To trim a smaller standard against a larger (as in the
next 9 rows of Table 5-2) use the trimmer whose labeled
value is equal to the capacitance setting. See right- hand
column of Table 5- 2.
e. To trim a larger standard against a smaller (as in the last
2 rows of Table 5-2). use the trimmer whose labeled value is
10 times the capacitance setting. See right-hand column of
fable 5-2 .•
f. If you are interested in preserving the accuracy as long
as possible, guard your bridge against any violent motion.
Even the impact of a dangling patch cord swung against the
panel will cause measurable changes in calibration (but
insignificant compared to the specifications)
Close the
access door over the C trimmers with care, otherwise it will
snap in closing.

5.4.2 Conductance Multipliers.
The G-multiplier calibration is similar to that above.
Proceed as in para. 5.2.6 and trim so as to minimize any
errors you find. The additional procedure is as follows.
(Notice that any possible errors in the 5 G standards, para
5.2.5. must be corrected only by GenRad, which has means
for checking and adjusting Do at the same time.
a. Gain access to the set of 6 screwdriver adjustments on
the G box, as in para 5.3.4.
b. In turn, make the connections and settings Indicated in
the last 6 rows of Table 5-3, and as described in paragraph
5.2.6, but instead of determining the error, adjust the
trimming resistor listed in the last column for the best null
possible. Always adjust these trimmers in the order given
(R18 first) because the last three adjustments have some
interdependence.

5-8 SERVICE & DIAGRAMS

5.4.3 C301 / Setting Zero C.
The ZERO ADJUST control varies Co' the capacitance
offset, in any position of the TERMINAL
SELECTOR /
READOUT MULTIPLIER
switch. So does the rear panel
screwdriver adjustment C301 (Figure 5-2) Normally, C301
is set so that the lO-turn range of ZERO ADJUST spans
conveniently the 3 conditions referred to in para 522,
which are conditions of Co = 0 on CAL and both 3
TERMINAL-UNKNOWN
positions
(The 3TERM
LOW
connector must be shielded I The adjustment shou Id be
made at 1 kHz.

5.4.4 Lever-Switch Stiffness.
The front-panel lever switches that control C and G
digits, ie, all but the first and last levers, have detent
mechanisms that can be adjusted as required The factory
setting is approximately
12 07 (350 grams weight), the
force required at the normal finger position to push each
switch from step to step (between 0 and g)
Other settings are possible, to satisfy ind ividual preferences. About 9 07 will provide a light "feel" and enable you
to slide each switch up and down with your thumb on the
front of the knob, but you need more skill to avoid stopping between detents. About 1607 will provide a positive
detent action that may be preferable if the bridge has many
operators, each mak ing relatively few measurements at a
time.
Perhaps more important than the absolute stiffness of
the detent action is consistency If one or more of your
switches responds differently from the majority it will be
an annoyance. Adjust the detent stiffness as follows.
a. Remove the bridge from its cabinet and position the
instrument upside down.
b. Determine which construction you have, by reference
to Figure 5-4. "A" has 3 screws in a row along the bottom
under each detent wheel; the screw farthest from the panel
is an adjustment (not locked) "B" has 2 such screws, both
locked.
c. If you have "A" construction, turn the adjustment
screw, cw to stiffen the detent action, ccw to loosen it
d. If "B," loosen screws Band C just enough to allow
you to move the block that holds the spring Tilt the block
as is required to obtain the desired detent action and
tighten the same screws

5.4.5

Maintenance

Note On Switches

Both rotary and lever switches may become noisy or
erratic from lack of lubrication. For rotary switches such as
the TERMINAL SELECTOR switch and MULTIPLIER switch,
use CRAMOLIN R5 oil. For lever switches, such as C and G
decades, use LUBRIKO H101 grease. Clean contact area of
lever switches with solvent (i.e.. FREON TF) before applying a
thin film of lubricant.

5.5 TROUBLE ANALYSIS.
If the bridge requires service beyond the realm of recalibration and adjustment described above, or the categories

of repair described below, please return the instrument to
GenRad. Refer to para. 5.1.

5.5.2 BCD circuits.
If the BCD code availableat the rear panel does not indicatethe
measurementas described in para. 2.8, trace the circuit with the
help of the schematic diagram, Figure 5- 7. If the trouble is in the
wiring, repair it. If in the switching, return the instrument to
GenRad.
NOTE
The lever switches are not replaceable, except
by GR.

r:
. -

,~,,~"""

-

~
.

FRONT PANEL

5.5.3 Non-Repairable Subassemblies.

._c-,

SCREWSB,C

....-. BRACKET

I

I

I

I

~

DETENT WHEEL

/

I

The following components are not to be disassembled for
repair. the ratio transformer T1, the coaxial choke T2, the G box.
and the C box. In the unusual event that one of them apparently
needs replacement. please consult with GenRad or return the
instrument as described before.

x

5.5.4 Typical Parameters.
The following parameters are typical They are listed for
reference in trouble shooting. Unless otherwise noted, set
the controls as follows
TERMINAL SELECTOR - CAL
EXT MULTIPLIER C and G readout - zeros
C MAX - shutters closed
G multiplier - 10-6

a

(B)

Figure 5-4. Detent mechanism of a typical lever
switch, A and B constructions. The instrument is upside down Ishowing the bottom edge of the front
panel). The type of wrench shown in part 10.25 inch,
80°, open end) is recommended for instruments
having the B construction.

5.5.1 Mechanical Damage.
Parts, such as connectors at front or rear, that are
damaged or worn, can be replaced A wire broken loose can
be resoldered or replaced. However, the replacement of any
lever switches is not recommended. Only the simplest repairs to the mechanisms of the C-MAX or the G-multiplier
levers is recommended.

At the GENERATOR
INPUT Jack 3.7 H at 100 Hz,
dc, ungrounded Demagnetize after dc measurement;

n

para 314. Measured series inductance increases with test
voltage; may be 5 H at 1 V.
At the EXTERNAL STANDARD port, with EXT MULTIPLI ER at 10, 37 H at 100 Hz, 1 n dc, qrounded
Demagnetize as above
From DETECTOR OUTPUT to EXTERNAL STANDARD LOW, with the EXTERNAL MULTIPLIER set to 10,
14 mH (below l-V test voltage) at 1 kHz; R < 1 n dc
5.6 PARTS LISTS AND DIAGRAMS.
The following pages contain mechanical and electrical
parts Iists. schematic diagrams, and supplementary information.

SERVICE & DIAGRAMS

5-9

MECHANICAL PARTS LIST
Fig
Ref

Qnt

FRONT
1
I.
2
2.
1
3.
I
4.
2

5.

2

6.
7.
8.
9.

2

10.

19

II.
12.

Description

GR Part No.

Fed
Mfg Code

Mfg Part No.

Lock with 2 keys.
Handle.
Cabinet gasket.
Bench cabinet assembly;
detailed breakdown listed separately. *
Panel locking connector asm.,
J I,EXTERNAL STANDARD HIGH;
J4,3-TERMINAL UNKNOWN HIGH.
Panel locking connector asm.,
J2, EXTERNAL STANDARD LOW;
B,3-TERMINAL UNKNOWN LOW.
Banana jack, J5, ground connection.
Precision connector, J6, 2-TERMINAL
UNKNOWN.
Knob asm., ZERO ADJUST, including:
retainer.
Knob asm., TERMINAL SELECTOR;
EXT MULTIPLIER, including:
retainer.
Knob for CAPACIT ANCE or CONDUCTANCE lever switch.
Door assembly.

5605-0100
5360-2032
5331-2220
4172-4016

24655
24655
24655
24655

5605-0100
5360-2032
5331-2220
4172-4016

0874-4006

24655

0874-4006

0874-4005

24655

0874-4005

4150-0900
0900-4410

24655
24655

4150-0900
0900-4410

5520-5333
5220-5402
5500-5321

24655
24655
24655

5520-5333
5220-5402
5500-5321

5220-5402
5500-5120

24655
24655

5220-5402
5500-5120

1616-1041

24655

1616-1041

Multiple socket, J 12, BCD CONDUCTANCE OUTPUT.
Multiple socket, J13, BCD CAPACIT ANCE OUTPUT.
BNC connector, JII, GENERATOR
OUTPUT; J14, DETECTOR OUTPUT
Foot (resilient strip).

4230-4036

93916

57-40360

4230-4049

93916

57-40500

4230-1200

91836

KC-19-161

4171-7010

24655

4171-7010

Fed Stock No.

REAR
I.

2.
2

3.

2

4.

NOTE
Electricel parts information in this section is presented
in such a WIY that all tha dlta for I pert-numberld sub·
assembly Ira visible in I single opening of tha mlnult.
Thus. thl perts list Ippeerl on Ilft·hend plges. whila the
part-location diagram Ion tha Ipron) Ind the schemltic
dillGram hip out] Ira on right-hand Plges.
'Cabinet

kits and hardward

are listed on page 5-12.

REFERENCE DESIGNATOR ABBREVIATONS
B
BT
C
CR
OS
F
J
K
KL
KS
L
M
MK

5-10 SERVICE & DIAGRAMS

=

•
•
•
•
•
•
•
•

Motor
Battery
Capacitor
Diode
Lamp
Fuse
Jack
Reley
Reley Coil
Relay Switch
Inductor
Meter
Microphona

P

Plug
T"nsistor

Q

R
S
T

U
VR
X
Y
Z

..
•
•
•
•
•
•
•

Resistor

Switch
T,andormer
Integreted Circuit
Diode. Zenar
Socket for Plug-In
Crystal
Network

ReferellC8l
ASA Y32.16end MIL-STD-16C

5935-062-1776

2

9

10

11
Figure 5-5. Front view; mechanical replaceable parts identified.

3

2

4
Figure 5-6. Rear view; mechanical

replac;eable parts

!T

404

i;~~ZI~~~~~SI~~~~~~A~;, ~o~, 16';; ,l~: 10-1 I

VOL TAGES EXPLAINED fN INSTRUCTION BOQI'( SERVICE NOTES
c:==::J'
PANEL CONTROL
REAR CONTROL
0" SCREWDRIVER CONTROL WT ~ WIRE TIE TP ~ TEST POfNT
COMPLETE REFERENCE DESfGNA TlON INCLUDES SUBASSEMBl Y
LE TTER, C·RI, B·R1, ETC

(:::::::1

Ie

F

ROTARY

CONNECTfONS

SWITCH NUMBERING

L,W.,,, S"",ACE

'RON',

-..........+

REAR

-

DUTPUTLEAVE~SUElASSEMBLY

~ INPUT FROM DIFFERENT SUBASSEMB

-' --CONTACTS
FIRSTCONTACTCW
FROM STRUT SCREW ABOVE I'(EY IS 01
SECTION SECTlON NEAREST PANEL IS '1

----t>

OUTPUT REMAfNS ON SUBASSEMBLY
fNPUT F ROM SAME SUBASSEMBLY

[)----

ROTORS SHOWN CCW

MAX-!
53

'-----t--+-----------,

I

WT51

+WT5'3

WT5e-

J 10

3

i

55

i

5<0

x

57

0

9
8
7
5
43
2
I
0
-I

'"

WT2.

WT4-

YE-VT-BK

~J
Ii,:no

0

YE-GY-BU YEGY-BK YE-GY-GN

24-

WTN

23

1

WT4

WH-C,NI

J13"47

~

-000-

WH-BU

j'"'
,
WH-BR

'1-c;,

22-

I,

I~

f-~---i WHoO
WTI

,WH-SK

t

t
4-8

~:

0

WTB

Y
JI3""4:'l

I~
,

c

-------.---'-....

CAPACITANCE

=0

THRU

CONDUCTANCE

SIG

517

THRU

52.1

o

WT5

C
,

J
g:;1

~I

,

J : 2.. ¥'- \4

'"

i")

oJ

f\)

m

~

J

r-'-_

,

tw r
'70

'------i---------.

VV- 8

-

\iI

::;

(\)

1'1
r<'>

'"
'-'! t
-s

I>

J

WT 170 - WT 210

VI I
OWTI

R

,

RD-BK

l:l

J 12."'18

J I 2.

-l~NC?.U~Tf\ N C~-.!?U TPU~
: co~
GY-BK,

I

GY-BR
GY-BU
G"!-GN

17

'1(3

VT-BK
VT'BR
VT-BU
VT-GN

14

OR-13K
OR-BR
OR-BU
OP,-(;N

IS

4

8

31

4

i'.8
29

1(;

2

3
4

s
I
2
4

e

, --

I
2

2G
27

VT-GY-8K

VT-GY-BR
VT-GY'BU
VT-C,Y'GN

,J 12 tt
-."
I

2

i'.

4

19

S

5
G
30

I

WT tv"

I

15 "
4
B

20
3
q

21

22

OR-GY,BK
OR-GY-BR
OR-GY-BU
OR'GY-GN

13

d

YL -f\D-BK
YE'RD -SR
YE-P,D-BU
YE-RD'G N

2.0

2-

8

4

25
2<;,

2.\

4

27

B

28

I

5

2.

IQ
23
2'1-

8
I

B
I

2

7

9

10

WIRE COLO""
YE-RD'BK
YE'RD'BR
WH-RD-BK,
WH'RD'BR

YE'BU
,'lWT24

~

-z:

CHA:,SIS

JIZ'*3G,

OR-GY'BK
OR-GY-BR
WH-BK
WH-BR
VT-GY-BK
VT-GY- BR
YT-GY-GN
YE-GY-BK
YT'RD-BK
VT-RD-BR
YE-GY-BR
YE-RD-BU
GY-RD-BK
GY-f\D-BR
YE-RD-0N
OR-GY-BU

Figure 5-7. Wiring diagram of BCD output circuits.

SERVICE & DIAGRAMS 5·11

ELECTRICAL PARTS LIST

CHASSIS MOuNTED PARTS
DESCRI PTlON

REFDES
C

C
C
C
C
C
C
C

C
C
C
C
C
C
C

C
C
C

C
C
C
C
C
C
C
C

C
C
C
C
C

C

C

**
,.,.

J
J

**
**

J
J
J
J
J

,.,.

J
J
J
J
J
J
J
J

* R
R
R
R

,.,.,.

R
R

R
R
R
R
R

S
S
S
S
S
S
S

S
S
S
S

*

**
***

1
2
3
4
5
6

7
8
9
10
11
12
13
14
15
10
17
18
20
21
22
23

24
25
20
27
28
100
101
102
103
300
301
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

CAPACITOR .0<;993 UF
CAPACITOR 9990 PF
FACTCRY SELECT
CAPACITOR AIR 3-32PF
FACTCRY SELECT
CAPACITCR AIR 1.7-8.7PF
TRIMMER CAPACITOR ASM
TRIMMER CAPACITOR ASM
TRIMMER CAPACITOR ASM
TRIMMER CAPACITOR ASM
TRIMMER CAPACITOR ASM
TRIMMER CAPACITOR ASM
TRIMMER CAPACITOR ASM
TRIMMER CAPACITOR ASM
TRIMMER CAPACITOR ASH
TRIMMER CAPACITOR ASM
CAP PCLYPRCPYL 500PF
10PCT 200V
TRIMMER CAPACI TOR ASM
TRIMMER CAPACITOR ASM
TRIMMER CAPACITOR ASM
TRIMMER
TRIMMER
TRI MMER
TRIMMER
TRIMMER
TRI MMER
PF
5PCT 500V
CAP MICA 02
CAPACITOR ASM
CAPACITOR ASM
CAPACITOR ASM
CAPACITOR ASM
FACTORY ADJUST
TRIMMER CAPACITOR ASM
PANEL CONNECTOR
PANEL C(~~ECTOR
PANEL CC~~ECTOR
PANEL CONNECTOR
BUSHING ThREADED BANANATURRET
PANEL CCMEC TOR
RECPT BNC
RECPT BNC
RECPT BNC
CONNECTCR PC 15 POS SR .150 SP
RECPT BNC BULKHEAD.200 CABLE
RECPT MICRO RIB 30 CONT
RECPT MICRC RIB 50 CONl
RECPT BNC BULKHEAD.206 CABLE
RECPT BNC

ORES
1 RES
2 RES
3 RES
4 RES
5 RES
6 RES
7 RES
8 RES
9 RES
24 POT
1
2
3
4

5
6

7
8
9

10
11

CCMP 100 K 5PCT 1/4W
FLM 10K
51l0l>PCTl5PPM1/4W
FLM lOOK
5/100PCT15PPM1/4W
1/10PC T 15PPM1I4W
FLM 1M
FLM 10 M IPCT 100PPM 1/4W
FLM 100 M IPCT 100PPM 11410
FLM 80K
1/10PCT 15PPM1/4W
FLM 800K
1/10PCT 15PPMl/4W
FLM
8 M IPCT 100PPM 1/4W
FLM 80 M IPCT 100PPM 1/4W
loW Kf\CB lK OHM 5 PCT lOT

SWITCH
SWITCH
SWITCH
SWITCH
SWITCH
SWITCH
SwiTCH
SWITCH
SWITCH
SWITCH
SWITCH

RCTARY ASM
RCTARY ASM
SHIELD
RCTARY ASM
AS"
ASM
ASM
ASM
ASH
ASM
ASM

PART OF SWITCH ASM 1616-2500
SUPPLY AS LOOSE PARTS
ORIENTATION OF PART IS CRITICAL

5·12 SERVICE & DIAGRAMS

PIN

1016-3000

NO.

FMC

11505-4029
0505- ..02 8

24055
24655

PART

.*••-••••

.*.*-.*
••

MFGR

PART

NUM8ER

il505-4029
0505-4028

4380-3725

74970

160-130

438v-361>u
1615-2120
1015-2120
1615-2110
1615-2110
1615-2110
1615-2110
1615-2110
1015-2130
1015-213"
1615-2180
4803-1569
1615-22VO
1615-2120
1615-2110
1616-6600
1610-0600
1610-6600
1616-660J
1616-6600
1010-0600
0+700-0364
1010-4810
1616-4810
1616-4810
1610-4810
.f- •• -++f-.
1615-2130

80583
24655
24655
24655
24655
24655
24655
2.. 655
24655
24655
24655
19390
24055
24655
24655
24655
24655
24055
24655
24655
24655
81349
24655
24055
24655
24655

MAC-Iv
1615-2120
1615-2120
1615-211i)
1615-2110
1615- 211 0
1015-21h.l
1615-2110
1015- 2130
1015-2130
1615-2180
PP481 56ilPF
1015-220v
1615-2120
1615-2110
1616-060U
1616-6600
1016-0600
1 bi6-6600
1016-06vO
1616-0600
CMu5ED620JN
1616-4810
1616-4810
1616-4810
1616-4810

24655

1615-2130

0874-4006
0874-4005
J814-4005
1>874-4006
4150-0900
0900-4410
4230-2300
4230-2300
4230-2300
4230-2715
423 .. -120"
4230-4036
4230-4049
4230-1200
4230-2300

24655
H655
24655
24655
24655
24055
24655
24655
24055
26601
91836
02000
02660
91836
24655

v874-4006
0874-4005
0874-4005
0874-4000
4150-0900
0900-4410
4230-2300
4230-2300
4230-23vu
143-015-08
KC-19- 161
57-40360-9
57-40500-4
KC-19-l01
4230-2300

oU99-4105
6019-3450
6019-3455
0619-3400
0188- 5100
6188-6100
661 <;-3465
0619-3470
0188-4800
6188-5800
0000-0142

81349
24655
24655
24655
24655
24055
2.. 655
24655
24055
24655
80294

RCR07GI04J
6619-3450
6619-3455
6619-3460
0188-5100
6188-6100
6619-3465
6619-3470
6188-4800
618B-5800
3540S-1-102

7890-5338
7890-5339
789D-825U
7890- 533 7
1616-2500
1010-2500
1616-2500
1610-2500
1610-2500
1616-2500
1616-250v

24655
24655
24055
24655
24655
24655
24655
24655
24655
24655
24655

7890-5338
7890-5339
7890-8250
7890-5337
1616- 2 500
1616-2500
1616-2500
1616- 2 500
1010-2500
l610-2500
1616-2500

10PCl

ELECTRICAL PARTS LIST (cont)
CHASSIS

SWITCH
SWITCH
SWITCH
SWlT CH
SWI TCH
SWITCH
SWITCH
SWITCH
S\oIIT CH
S\oIITCI-

ASM
ASI"
A SI"
ASM
ASM
ASM
ASM
ASM
ASM
AS'"

S
S
S
S
S
S
S
S
S
S

14
15
16
17
18
19
20
21

T
T

1
2

TOROID TRANSfORMER
CHOKE A 51"

VR
VR

1
2

ZENER lMZ-68 A
ZENER lMZ-68A

13

AS,"I

68V
6av

5PCT
5POT

PC BOARD

R
R
R
R
R
R
R
R
R
R
R
R
R

IJ

11
12

13
14
15
16
17
18
19
20
21
22

1.5\01
1.5\01

PIN

RES
RES
RES
RES
RES
RES
RES
RES
POT
POT
POT
POT
POT

1.02K
1110 PCT 50PPM1I2"
10J OHMlII0PCT
50PPM 1/8\01
1 PCT 1I4W
fL'" t49K
t4.9K
1 PCT 1I4W
flM
64.9K
1 PCT 1/4101
flM
1 PCT 1/4"
FlM t4.9K
1/8W
1 PCT
FlM 6.04K
FLM 715 CHM 1 PCT
1/81>1
CERM TRM 200K OHM 10 PCT 15 T
CERM TRM
20K OHM I.> PCT 15T
CERM TRM 20K OHM 10 PCT 15T
CERM TRM 20K OHM 10 PC T 1ST
CERM TRM
2K OHM 10 PCT 15T

flM
flM

1616-31)0U

NO.

fMC

MfGR

PART

1616-2500
1616-2500
1616-2500
1616-2500
Ib16-25011
1616-25.:10
1616-2500
161b-250J
1616-250J
1616-2500

24655
24655
24655
24655
24655
24655
24655
24655
24655
24655

1616-2500
1616-'25UO
1616-2500
1616-2500
1616-2500
1616-2500
161b-2500
1616- 2 501i
1616-2500
1616-2500

1616-2310
1616-245v

24655
24655

1616-2310
1616-2450

6083-1064
6083-1064

24444
24444

lMZ-68A
lMZ-68A

1.5R68B
1.5R68B

NO.

6193-100J
6619-1600
6350-3649
6350-2649
6350-2649
0350-2649
6250-1604
6250-0715
6049-0193
604

a:

fF

_

iJOf,F

_~ If!

100a~~"F

t:::t:::
r::,J;c:"
,
1'0:
I
__I=t---=-1- _:~~
-1:,:::t~ "
i

0J

0J

~5pF.-.05pF

I

.x
••

I

xX

III

1

-:;----=1
z'

I

.XXpF

I

-

~>

10J

005~F

--l;;;-

""

>-

i~.

't>

I

••

ro

10

,0J

0J

'I;,'_,f

[ZM!V]

,lo.F

co

oL

.o
len

a:

loc

0

1

'°
1

I

L_-=r.

I

!

112.

0

III

110
109
NOTE

/

A.UXllJAI:?V

CONTACT

SI/OWN

~i f-fj,,/1~r~~fJ!:f£~J£j~()N5
THIS'

OF

CIRCUIT IS lFPICAL
S 5 - S.2 / .

lOB

107
10..

v--t-~. ---t--

o

----+ __c-__ -t----.,--

+ .. ('----+---,~)!

v--+-----c'---+---o--._--_.+----<~-+-. "'r-

105 c --.._-t-----c~_+--o---t----~-t-----c--t--o----+--··- .. !3.>!
104 c· - -·\·-c~--+--o---+---·()·_--+--~,.~-t~·-o---t--'.2')

103
102

10/
IYTI2.

WTI3

W!

5

112III

iC,lCD PRECI SIGN
CAPACITANCE
BRIDGE
SCHEMATIC
DIAGRAM

110
10'3
108
107
lOb

105
104

103
102.
101

0,11

0,12

313

IC~PI\C-I

Cli

i)l>(lntatlons of R5, R9, and e17 are significant; properly oriented by factory.

514

TFlNC::E]

SI5

51C,

WTIG

r-{
Ci'lBLE

BR

/_0

.;1
.8
7

-,

TERMINI\L
SELECTOR

RE"'OoUT

U'F;/e

.",

EXT

o

MULTIPLIER
~IO

X I
C"'L
XIOO
)\10
~I

-,--I

~~J .i i o

12.

:0

co

a:

o

"

~,1'Li'PI....,E~

•

5

f-----

------ -+-+----QwT310

WTIE>

Ci'l8LE

Wl-i-VT~
JII

Figure 5-9. Schematic diagram of the 1616 Precision Capacitance Bridge.

SERVICE & DIAGRAMS

5-13

AppendixA

G900 CONNECTORS
~

a
3-TERMINAL
UNKNOWN
PORT

/
\

LOW

GR 1616 BRIDGE

=

CA

IF]
Cx

SPECIAL CABLES
Cs

-

\.

TEST FIXTURE
Figure A-l. Test setup for 2-terminal measurerrent,with fringing eliminated.

1616·38

Appendix A
An Interpretation of Millea's Method*
to Eliminate Fringing from
Two-Terminal Measurement.
If a suitable 2-terminal standard is not available, and the
uncertainty of ±.008 pF in the fringing of the G900

c. Provide a terminal for connecting a wire (274- LM) to
the outer conductor or each G900 connector (points a. b).

connector is unacceptable for your measurements,
method is recommended. (Refer to para. 3.9)

Working Capacitor. Use a 2-terminal coaxial capacitor
with a G900 connector. This is C it may be your unknown
x:
capacitor or one you wish to calibrate as a standard.

this

Test Fixture. The assembly of adaptor, tee, cables, and
wires is shown in Figure I The special cables (2 required)
are not commercially available, but may be constructed as
follows.
a. Remove the G874 connector from one end of a Type
874- R22 LA Patch Cord.
b. Install a G900 cable connector instead, Type 900C58. but do not connect its outer conductor to the cable
shield. Leave a gap as shown, by cutting the shield braid
back as far as the cable Jacket. Touch the cut ends of the
braid back under the Jacket enough to assure insulation. The
two should not be cut so short as to expose the inner
conductor to external fields. The rubber sleeve should provide
some strain relief and the cable retainer (by overlapping the
shielded portion of the cable) sufficient shielding.
It is
imperative that the inner conductor of the connector be rigidly
supported so that CA and C are constants.
B

A-l

Procedure.
a. With the connections as shown, i.e.. Cx connected to
CA , a to h, and b to g, measure. C 1 = C A + CX'
b. With the work ing capacitor transferred to the other
arm of the fixture, i.e.. Cx connected to C B ' b to h, and a
to g, measure C2 = C B + C x.
c. With the fixture closed on itself, i.e., C A connected to
C B ' a or b to h (neither to g), measure C3 = C A + C B
d. Calculate the capacitance without
fringing
(C 1+C 2-C 3)/2

• Millea,
Aurel, "Connector
urement
of Two-Terminal
Research,
3, of the National
4, July-Dec.,
1970.

Cx =

Pair Techniques
for the Accurate
MeasLow-Value
Capacitances,"
Journal
of
Bureau of Standards,
Vol 74C, Nos 3 &

_GenRad

GR1621 Capacitance Measurement System

Supplement to GR1616 Precision
Capacitance Bridge Manual

1

Introduction

This supplement to the GR 1616 Precision Capacitance Bridge instruction manual (form number
1616--0100-(0) provides additional information that applies to the recalibration procedure. Keep this
supplement with the basic manual and make an appropriate note in the margin on page 5-8, just after
step e of paragraph 5.4.1.

2

Supplementary Procedure

If in paragraph 5.4.1 step e, you find that the lo-nF and/or 100-nF capacitance trimmers lack
sufficientrange to obtaina balance,use switches S22 and/or S23 as follows. SeeFigure 5-5.

To Extendthe Range of the 1G-nF Trimmer

2.1
•

Remove the instrument from its cabinet, as described in paragraph2.4 or 2.5.

•

Perform a trial balance, using only the 1-pF, lo-pF, and 100-pF lever switches not the
trimmers,andlimiting the 100-pF switch to just two positions: 0 and -1.

•

If the error at balance is positive (lOO-pF lever switch set to 0), tum S22 one step counterclockwise; but, if it is negative (lOO-pF switch set to -1), tum S22 one step clockwise.

•

Set the 1-pF, lo-pF, and lOO-pF lever switches to their zero positions.

•

Returning to paragraph 5.4.1, step e, if the lo-nF trimmer now has sufficient range, reinstall
the instrument in its cabinet andrepeat the calibration procedure of paragraph 5.4.1. But,
if the lo-nF trimmer still has insufficientrange, repeat the above supplementaryprocedure.

Part Number 1616-0150-00

Page 1

1616 PrecisionCapacitanceBridge,Supplement

2.2

To Extend the Range of the 10G-nF Trimmer
•
•

•
•
•

Remove the instrument from its cabinet, as described in paragraph2.4 or 2.5.
Perform a trial balance, using only the 1-pF, lO--pF,and 100-pF lever switches not the
trimmers,and limiting the 100-pF switch to just two positions: 0 and -1.
H the error at balance is positive (l00-pF lever switch set to 0), tum 823 one step counterclockwise; but, if it is negative (lOO-pF switch set to -1), tum 823 one step clockwise.
Set the 1-pF, 1Q-pF, and lOO-pF lever switches to their zero positions.
Returning to paragraph5.4.1, step e, if the 1QO-nFtrimmer now has sufficient range, reinstall
the instrument in its cabinet and repeat the calibration procedure of paragraph 5.4.1. But,
if the 1QO-nFtrimmer still has insufficientrange, repeat the above supplementaryprocedure.

CBox

822

FrontPanel

61000.0

Figure 5-5. Top View, Cover RemOVed,
Showing locations of Switches S22 and S23.

_GenRad
>""\

300 BakerAvenue,Concord, MA01742-2174

Page 2

Printedin U.S.A.

Part Number 1616-0150-00

\2~~:;::;'/

For more information

mil

or the location of your local GenRad sales office,

contact:

GenRad
U.S.A.

Germany

Asia Pacific

1-800-4-GENRAD

TEL. 089/431990
TLX. 529917 GEND

~kn~a~~~~78-4400

Canada

Italy

TEL. 416890-0160
TLX. 06-986766 GENRADCO
MSGA

TEL. (02) 502951
TLX. 320 373 GENRAD I

United Kingdom

Switzerland

European Headquarters
TEL. 062839181
TLX. 848321 GENRADG

TEL. 01/552420
TLX. 816828 GENRA CH

France

TEL. (040) 450605
TLX. 50535 GRBEN NL

TEL. (1) 47970739
TLX: GENRA 220991 F

Netherlands

TLX. RS37808 GRASIA

All Other Countries
International Department
300 Baker Avenue
Concord, Massachusetts
U.S.A.
TEL. 508 369-4400
TLX. 92-3354

01742

mtIIIIIIGenRad
300 Baker Avenu e , Concor d , Massac h usett s 01,

Printed

in U.S.A.



Source Exif Data:
File Type                       : PDF
File Type Extension             : pdf
MIME Type                       : application/pdf
PDF Version                     : 1.6
Linearized                      : No
XMP Toolkit                     : Adobe XMP Core 4.0-c316 44.253921, Sun Oct 01 2006 17:08:23
Modify Date                     : 2007:09:10 11:52:10-04:00
Create Date                     : 2006:06:04 20:44:43-04:00
Metadata Date                   : 2007:09:10 11:52:10-04:00
Creator Tool                    : Adobe Acrobat 7.0
Format                          : application/pdf
Document ID                     : uuid:c84bc80a-12e2-4b9b-875f-88fa833bea1b
Instance ID                     : uuid:7de31133-bf70-46ee-ae2e-9c887192af8c
Producer                        : Adobe Acrobat 8.0 Paper Capture Plug-in
Page Count                      : 84
Creator                         : Adobe Acrobat 7.0
EXIF Metadata provided by EXIF.tools

Navigation menu