Iet labs 1417 User manual

+

PREC

I

SION

INS

T

RUMENTS

F

OR

TES

T

AND

M

EASUREMENT

t

1417

Four-Terminal

Capacitance Standard

Instruction Manual

1

.

~

ET

~

.

~r.

..

~~~

~

~·

INC.

\!!!!/

GenRad

!B

tl Me.n Slree<,

~\':s

lb<.L")•

,

N'1 11

~

ft0

www.tellabs.

com

TEL {!>1til

3::.4-

Si

~

S

f:t

• !

500i

B

~!!>t

.

B!!

•

!'-

AX \

!>16

}

J

J~

-

~988

:X

:&

l

l ,

l

:&

=I

=t

~

=-!

:j

=I

:j

1417 Four-Terminal

Capacitance Standard

Instruction Manual

Fonn

1417-0100-00

I'

;~

lET

LABS,

INC.

\!.!Y

534 Main Street

Westbury,

NY

11590

www.ietlabs.

com

516-334-5959 Fax: 516-334-5988

The

material in this manual is for informational purposes only and is subject to change, without notice.

QuadTech assumes

no

responsibility for any error

or

for consequential damages thatmay result from the

misinterpretation

of

any procedures

in

this publication.

Contents

Manual Changes

Warranty

Specificatipns

Introduction -Section 1

Operation -Section 2

Use With QuadTech

Bridges-

Section 3

Theory -Section 4

Service and Maintenance -Section 5

WARRANTY

We

warrant that this product is free from defects in material and workmanship and, when properly used, will

perform in accordance with applicable

lET

specifications.

If

within one yearafteroriginal shipment, it is found

notto meetthis standard, it will be repaired

or

,

at

the option

of

I

ET,

replaced

at

no charge when returned to lET.

Changes in this product not approved by

lET

or

application

of

voltages

or

currentsgreater thanthose allowed by

the specifications shall void this warranty. JET shall not be liable for any indirect, special,

or

consequential

damages, even

if

notice has been given to the possibility

of

such damages.

THIS WARRANTY

IS

IN LIEU OF

ALL

OTHER

WARRANTIES, EXPRESSED

OR

IMPLIED, INCLUD-

ING

BUT

NOT

LIMITED TO,

ANY

IMPLIED WARRANTY

OF

MERCHANTIBILITY

OR

FITNESS FOR

ANY

PARTICULAR PURPOSE.

WARNING

OBSERVE

ALL

SAFETY RULES

WHEN WORKING

WTTH

HIGH VOLTAGES

OR

LINE VOLTAGES.

Dangerous voltages may be

present

inside this

instrument.

Do

not

open

the

case

Refer

servicing to qulified personnel

HIGH

VOLTAGES

MAY

BE

PRESENT

AT

THE

TERMINALS

OF

THIS

INSTRUMENT

WHENEVER

HAZARDOUS

VOLTAGES(

> 45 V)

ARE

USED, TAKE

ALL

MEASURES TO

AVOID ACCIDENTAL CONTACT WITH

ANY

LIVE COMPONENTS.

USE MAXIMUM INSULATION

AND

MINIMIZE

THE

USE

OF

BARE

CONDUCTORS WHEN USING THIS INSTRUMENT.

Useextremecaution wh

en

working

with

bare

conductors

or

busbars.

WHEN WORKING WITH HIGH VOLTAGES, POST WARNING SIGNS AND

KEEP UNREQUIRED PERSONNEL SAFELY

AWAY.

CAUTION

DO

NOT

APPLY ANY VOLTAGES

OR

CURRENTS TO THE TERMINALS

OF

THIS

INSTRUMENT IN EXCESS

OF

THE

MAXIMUM LIMITS INDICATED ON

THE

FRONT PANEL

OR

THE

OPERATING

GUIDE

LABEL.

4

=-

~

:a

=a

:a

:j

:t

~

=I

~

=!

4

~

Instruction Manual Changes

These supplementary pages contain information

of

improvements

or

modifications not documented

in

the

main text. All references to GenRad in

the

manual now apply to QuadTech, Inc.

Page

1

1.1

PURPOSE

• Add.

The

direct-reading accuracy

of

the

1417 is not sufficient for checking

the

accuracy

of

many, more

accurate instruments such as the QuadTech Digibridge® impedance meters except at very high values

of

capacitance.

To

check these instruments,

one

must take advantage

of

the

precise ratios

of

the

1417 which

accurately scale

the

1uF value to much higher values up to 1 Farad. Thus

the

1

uF

value should

be

calibrated, at each frequency, to get accurate capacitances

of

higher values. Calibrations

at

higher values

are unnecessary (

see

paragraph 5.3

).

Page

1

1.2

DESCRIPTION

• Add after

the

CAUTION note.

If

the

instrument to

be

tested is able

to

apply

de

at any level use a series

blocking capacitor

to

prevent

de

from flowing

in

the

1417 (see paragraph

2.5

.3, added to this change

secti

on

).

Page

3

2.2.1

Four

Terminal

Connections

• Add to frrst paragraph. Connection for QuadTech Digibridges are given in paragraph 3.1.1 (added to

this change section).

It

should be noted that theoretically

the

CURRENT and POTENTIAL terminals

may be interchanged.

Page

3

2.2.2

Tenninal

Impedances

• Add.

For

most

instruments, including

the

QuadTech Digibridges,

the

potential connections can tolerate

the

higher impedances and thus should

be

connected to

the

1417 POTENTIAL terminals. These series

impedances cause

errors

when making 2-terminal measurements which may

be

reduced

by

paralleling the

CURRENT

and

POTENTIAL

connections (see paragraph 2.8).

Page

3

2.2

.3

Connecting

Leads

•

.AruL.

If

possible,

the

connecting leads should

be

twisted to reduce mutual inductance.

Twist

the two (H

& L) potential leads togeth

er

to reduce susceptibility to magnetic fields and twist the two current leads

together

to

reduce

the

radiation

of

a magnetic field.

Page

3

2.2.4

Mutual Inductance

in

Leads

• Add. Mutual inductances in the connecting leads may

be

greatly reduced

by

twisting them as discussed

in paragraph 2.2.3 added above. Many instruments make a short-circuit calibration correction which can

also greatly reduce mutual inductance errors.

If

such a calibration is made,

try

to keep

the

lead

configuration as unchanged as possible between calibration and measurement.

Page

3

2.3.1

Accuracy

of

the

1417

• Add to fourth paragraph. Fortunately,

the

newer, more precise instruments

for

measuring high

capacitance

can

apply such low signal levels that this non-linear effect causes only small errors.

• Add

to

fifth paragraph. It should be remembered that

the

final application

of

most instruments that

measure high capacitance is to test electrolytic capacitors that generally

have

broad tolerances. Highly

preci

se

instruments

do

not

ha

ve to

be

calibrated to

the

ir

full accuracy

if

this is their only end use.

Instruction Manual Changes (continued)

Page

4 2.3.2 Required Accuracy

•

Add

to

last paragraph.

Note

that

often

full accuracy is

not

required particularly

if

the

end

use

is for

testing low accuracy electrolytic capacitors.

Page 5 2.4.1

The

luF

Calibration

• Re.place complete paragraph.

If

accuracy better than

0.25%

is required, a calibration measurement

of

the

1

uF

value

at

all

three

frequencies is required.

This

can

give

an accuracy equal to

the

ratio accuracy

plus

the

1

uF

calibration accuracy

at

low

signal levels.

At

higher levels,

the

accuracy based

on

the

1

uF

values should

be

within

0.1%

at 100

or

120Hz

if

Ev

is held

under

5V

or

at 1

kHz

if

Ev

is less than

20V

(refer

to

paragraph

2.3.3).

The

1

uF

values

of

the

1417

may

be

calibrated directly

on

a precision

bridge

or

by

comparison with a

calibrated 1

uF

standard,

such

as

the

QuadTech

1409Y

(a stable silvered-mica unit),

that

has been

calibrated

on

a

precision

bridge. Recommended precision bridges

are

the

QuadTech 1615 (1620 system)

or

the

QuadTech 1616 (1621 system). Both

of

these bridges apply

very

low

voltage

at

this capacitance

value

and

have

accuracy

of

0.

01%

or

better.

If

a calibrated 1

uF

standard is used,

the

instrument being

calibrated

can

usually

be

used

to

make

the

comparison because

it

will usually

have

resolution

substantially

better

than

its accuracy.

This

is particularly

true

of

those

instruments (such

as

the

QuadTech 1689, 1692

and

1693)

that

can

read

differences

in

percent.

It

is recommended

that

the

transformer

cores

be

cyclically demagnetized before calibration

to

ensure

that

there

is

no

residual magnetization

due

to

the

previous application

of

de

or

a

large

transient

current

(refer

to

paragraph

5.5.1). Also

refer

to paragraph

5.3

for

more

information

on

calibration and adjustment.

Page 6 2.5.1

Magnetization

of

the

Divider

Core

• Correction to

end

of

first paragraph.

It

can

be

removed

by

cyclic demagnetization,

refer

to paragraph

5.5.1.

Page7 2.5.3 Use

of

Blocking Capacitors

•

Add

paragraph.

DC

current

should

not

be applied to

the

1417 terminals as

it

will affect

the

magnetization

of

the

transformer

divider

cores

changing their input impedance

and

thus

have

an affect

on

the

capacitance values, particularly

at

the

lower

frequencies (refer

to

paragraph

2.3.

1).

If

there

is a

possibility

that

either

the

CURRENT

or

the

POTENTIAL

circuits

of

the

measuring instrument

can

apply

even

a low-valued

de,

blocking capacitors should

be

placed

in

series with these connections.

These

may

be

low-leakage electrolytic capacitors

and

should

have

capacitances

of

200

uF

or

higher.

The

CURRENT

circuit

of

the

QuadTech

Digibridge

impedance meter

may

apply a few millivolts

of

de and

therefore

it

is suggested that a capacitor

be

placed

in

series with

the

H

CURRENT

connection to

the

1417

in

order

to achieve

the

best

measurement accuracy.

Page 7 ·2.6 USE

OF

AN EXTERNAL STANDARD

•

Change

first

sentence

in

paragraph

one

to

read.

An

external standard

may

be

used to

get

referenced

values between

those

decade values

provided,

to

extend

the

range

or

to

reduce

the

effect

of

inductance

error.

•

Change

paragraph

after the formula to read.

The

larger

the

external capacitor,

the

less

error

(percent)

will

occur.

Using

a

10

uF

standard will

give

the

same

decade values

of

capacitance

as

the

internal

standard (except 1

uF)

with less

error

due

to signal level,

but

calibrating

the

10

uF

value is quite

difficult

If

the

external capacitor is much smaller than 1 uF, this

error

becomes

very

critical.

!

..

L.

j

-

j

Instruction Manual Changes (continued)

Page

17 5.3 CALIBRATION PROCEDURE

• Addition

to

beginning

of

the

first paragraph.

The

only calibration required is

the

calibration

of

the

1 uF

setting at each

of

the

test frequencies. These should

be

well inside

the

0.25%

tolerance.

•

,&kL.

Calibration Requirements. The 1417

direct~reading

accuracy can

be

checked

by

measuring the 1 uF

values at each test frequency.

The

direct reading values

of

the

other values should

be

within the

specification (0.25% plus the ratio accuracy)

if

the 1 uF.values

aie

well within specifications and thus no

calibration

of

them is required.

For

higher accuracy,

the

calibrated values

of

the 1uF setting should

be

scaled

by

the appropriate power

of

ten. This mode uses

the

very

precise and stable accuracy

of

the internal transformer dividers. These

are

checked in

the

manufacturer (and may

be

tested later, refer to paragraph 5.4.1),

but

should not

change except

in

the

case

of

gross damage.

It

makes little sense to measure and record

the

higher

capacitance values because they also depend

on

the

internal capacitor which is less stable than

the

ratios

and has a rather high temperature coefficient. Moreover, there is no known instrument that is adequate

for

such a calibration!

The

1417 should

be

considered a precise ratio device requiring only

the

precise

calibration

of

its internal 1uF standard (and careful use) to

get

high accuracy.

It

is preferable to make

these precise 1

uF

calibrations

at

the time

of

use because

of

possible variations in the internal standard

(refer to paragraph 2.4.1)

However, because

the

effective capacitance

of

all settings depends somewhat

on

signal level, particularly

at

the

lower frequencies (refer to paragraphs 2.3.1 and 4.3), it may make sense to make capacitance

calibrations for a specific application that uses relatively high signal levels at

100

or

120Hz

.

Page

18 5.4.1 Divider Ratios.

• Change second paragraph to read.

The

ratios should

be

those for a and

bin

Table 4-1 to the following

limits in ppm

of

input

or

in percent

of

reading.

10

100

103

to4

105

106

16

ppm

5 ppm

1.6

ppm

.75

ppm

.8ppm

.6ppm

%Limit

.005%

.

0075%

.024%

.06%

A high precision divider is required. Note that the

1417

could have application as a precision divider.

Page 18 5.5.1 Demagnetization

of

Divider Cores

• Change last sentence

of

paragraph

one

to read. A QuadTech 1311

set

for full output

on

its

30

V range at

50

Hz

can

give

a suitable signal.

• Change step c

to

read. Set

the

1311

to

50

Hz

and to its

30

V range and increase

the

output for

maximum.

Specifications

Approximate

Capacitance

Rat

io

Accuracy

D

Accuracy

Terminal Impedance

Value

{Internal

100&

Standard

I

120Hz

1kHz

120Hz

1

~tf

--

-

~.001

10

~tf

0.02% 0.04%

~.001

100

~tf

0.02"~

0.04%

:!:.001

1

mF

0.02%

0.06%

:!:.001

10mF

0.03%

0.2"~

:!:.001

100mF

0.1% .

::.003

1 F

0.25%

.

:.01

Capacitance: Internal Standard: 1

~tF

to

1 F in 7 switch-sel

ected

decade

values. External

Standard:

Indicated

cap

a

citance,

multip

lied

by

C

ext/1~tf.

Capacitance

Accuracy

, direc:t·reading:

0.25%

plus

ratio

ac:cura~y

(see

table)

at

100Hz.

1

20Hz,

and

1kHz

.

20

to

25°C.

':""'th

low

apg

lied

vol'!age

(<1/4

E max) using internal

standard

and

a

proper

four-

terminal

measurement.

(May also

be

used

as a

two-termi

nal

standard,

with

a D < 1

and

a

capacitance

change

from

the

four-terminal value

of<

1/2%

up

to

1 mF

at

120Hz

or

tess.)

Capacitance

Ratio

Accuracy: See

table.

Dissipation

factor:

0.01

at

100Hz,

120Hz

and

1 kHz.

ForD

accurac

y,

see

table.

Terminal Impedances: See figure

and

table

(approx

values given).

Temper

at

ur

e

Coefficient

:

Approximately

-140

ppm

rc.

Voltage

Char

acteristic:

Approximately

+0.3%

change

from

Ov

to

E

max

(see

table)

at

100

Hz. Less

at

higher frequencies.

Maeh:anical: DIMENSIONS:

lwxhxd):

8.

5x5.9x5.25

in.

(2

1.5

x14.7x

13.2

em).

WEIGHT: 6

lb

(2.7 kg)

net,

11

lb

(5

kg) shipping.

II

ZA

Z8

1kHz

(n)

~

.001

0.03 0.03

~.001

7.0

15.5

:!:.001

3.1

6.4

:!:.002 1.1

2.2

::.005

0.37

0

.72

.

0.13

0.23

. 0.04

0.05

ZA

H

CURRENT

Ic

L

0

'V\1\

<

10mn

Description

1417

Four-Terminal Capacitance Stand:ard

EMax

lVI

20

6

2

0

.8

0.5

0.25

0.06

•

Not

Specified

POTENTIAL

'V\1'.,.

<1

0mn

L

0

Catalog

Numb~r

141

7-97()0 j

..

!

:1

j

;

-

j

i

_

..

;

.r

i

-

j

--

.

j

Introduction-Section

1

1.1 PURPOSE.

The 1417 Four-Terminal Capacitance Standard

is

a

standard used

to

calibrate bridges and meters

that

measure

high values

of

capacitance with a 4-terminal connection.

It

also has some use

in

the

calibration

of

instruments

that

make

a 2-terminal connection. Most instruments

that

measure high

capacitance

do

so

at

100

Hz,

120

Hz

or

1 kHz;

the

1417

is

compensated

to

have good direct-reading accuracy

at

these

frequencies.

It

can be calibrated

to

be highly precise, making

use

of

its precise capacitance-scaling ability.

1.2

DESCRIPTION.

The

1417 uses 2 precision transformers

to

multiply

the

value

of

a capacitance standard by factors

of

10

up

to

10

6•

It has an internal standard

of

1 /lF

so

that

7 decade values

of 1

JJF

to

1

Fare

switch selected. Other values may be

obtained by use

of

an external standard.

This standard does

not

act

as

a capacitor

at

de

and

the

application

of

de can alter

the

capacitance values.

CAUTION

Do

not

apply de.

1.3

CONTROLS AND CONNECTIONS.

There are 2 panel knobs driving rotary switches. The

CAPACITANCE switch selects the capacitance value as

indicated when

the

intern

al

standard

is

used. When an exter-

nal standard

is

used,

the

value

is

that

indicated

by

the Cdial,

multiplied by Cext/1 /lF (see para 2.6).

The

TEST FREQUENCY switch selects proper padding

capacitors and resistors

to

give good direct-reading capac

i·

tance accuracy and

aD

of .

01

at

100Hz,

120Hz

and

1kHz

(see para 2.7 for use

at

other

frequencies).

It

also has a

fourth position

that

removes

the

internal

1·JJF

standard"and

cori.nects inste.

ad

the

EXTERNAL STANDARD terminals

(see para 2.6).

There are 4 pairs

of

binding posts

on

standard 19-mm

(3/4-in.) spacing plus 2 ground (case) binding posts.

All

pair-s

have termina

ls

marked H and

L,

and L

term

i

na

ls are

all

common

.

The

pairs

of

terminals

on

the

left are those

that

should

be

...

.:nnected

to

the

measuring instrument. They are labeled

POTENTIAL AND CURRENT,

common

nomenclature for

a 4-terminal standard. Generi!IIY, these are connected

to

bridge terminals with similar markings.

The

VOLTMETER terminals

on

the

right are directly

across

the

standard capacitor used, whether

it

is

the

internal

standard

or

im

external standard. Their primary use is

to

monitor

the

voltage across

the

standard

so

that

corrections

can

be

made when precise measurements are required

{see

para 2.4). These terminals may also

be

used

to

add capa-

citance

to

the internal standard (see para 2.6).

The EXTERNAL STANDARD terminals are used

to

conn-

.

ect

an external capacitor

to

obtain

other

than

decade

values

{see

para 2.6). These terminals are switched in the

circuit

by

the

TEST

FREQUENCY

sw

itch.

Fivure

1·1.

GR

1417

Standard.

1

4.

,.

-.J

j

Operation

-Section

2

.1

GENERAL.

The manner in which the 1417 should be used depends

on

many factors:

the

accuracy required, the

test

frequency,

the

applied voltage,

the

capacitance values required, and

the

instrument with which

it

is

used. The operation

is

very sim·

pie in many applications,

but

can become more complex,

particularly when

the

highest accuracy

is

required. It

is

strongly suggested

that

the

theory and calibration techniques

used

be

thoroughly understood

to

avoid possible errors.

Because different instruments measure under different

conditions and have different accuracy specifications, speci·

fie instructions are given in Section 3 for each

GR

instrument

that

measures high-valued capacitance. These refer

to

the

following paragraphs which describe procedures and pre·

cautions

that

should be used

to

obtain accurate calibrations

for

the

1417, so

that

it may be used

to

make accurate

cal

i·

brations

of

measuring instruments.

2.2 CONNECTIONS TO INSTRUMENTS.

2.2.1 Four-Terminal Connections.

The

connections

to

4-terminal measuring instruments are

made

to

the

Hand

L CURRENT terminals and H

a-td

L

POTENTIAL terminals. Usually

the

terminals

of

the

meas·

uring instrument use a similar notation (although + may be

used instead

of

Hand-

instead

of

L)

and, in genera

l,

these

terminals are connected

to

the

corresponding terminals

of

the

1417,

but

not

always. Section

3,

which describes

the

calibration

of

specific GR instruments, gives

the

proper

connections

for

each instrument.

The 1417 also has terminals

that

tie

directly

to

its case.

It

is

not

usually necessary

to

make any connection

to

this

terminal. However,

if

strong fields are present, grounding the

case may give

better

results. There

is

no direct connection

between

the

case and

the

internal circuitry.

2.2.2

Terminal Impedances.

The

1417

has impedance in series with each

of

its termi-

nals as shown

in

Figure 4-4

and

Table 4-2. These have

no

effect

on

a perfect 4-terminal measurement. However,

no instrument makes such a perfect measurement. so

that

connections should

be

made in such a way

that

the

larger

terminal impedances are connected

to

the bridge

or

meter

terminals

that

are most immune

to

such errors. This

is

why

the

various instruments described

in

Section 3 have different

connections. There

is

a best way

to

connect

to

each unit.

2.2.3

Connecting Leads.

The connecting leads used should

be

short

and use reas-

onably heavy wire

to

avoid increasing these terminal impe·

dances. However, in many cases test lead sets are supplied

for particular instruments. If these lead sets are used also in

the

application

of

the

instrument,

it

is

desirable

to

use

them

for

the

calibration as

wel

l, so

that

the

calibration

is

made

on

the instrument

as

it wi

ll

be used. If special lead sets are

used

it

must be remembered

that

they

can

affect

the

cali-

bration.

2.2.4

Mutual Inductance In Leads

While a gqod four-terminal measurement

can-

be immune

to

.self-inductance in

the

le

ads, mutual inductance between

the

current leads and the potential leads can still cause an

error,

an

error which

is

very

important

at

very high capaci-

tance. The equivale

nt

circuit

of

figure 4-3 gives

the

formula

c

Cmeas = 2

1

±w

CM

The mutual inductance can be positive

or

negative. This

error

is

critical

at

very high capacitance, even

at

100

Hz,

and much more critical

at

1 kHz.

For

example 2.5

nH

g

ives

0.1%

error

when measuring 1

Fat

100Hz

or

when measuring

10

mF

at

1 kHz. The inductance

of

2 coaxial, circular,

single-

turn

coils

both

of

1

-e

m radius, 1 em apart, is about

5 nH;

it

is

obvious

that

great care

must

be

taken

when w2 C

is

large.

Mutual inductance between

the

current

circuit

and the

potential circuit occurs in the 1417, the

connect

i

ng

leads,

and inside

the

measuring instrument. The mutual inductance

inside

the

1417

is

usually negligible compared

to

that

of

the

leads.

At

the

highest capacitance values (100

mF

and 1 F)

it

is

approx 1

nH

.

At

lower values

it

causes negligible error

at

120

Hz

and less

than

.02% at 1 kHz.

The sections following, on the calibration

of

specific in·

struments, note

the

effect,

if

any,

of

the

mutual inductances

of

these instruments and their lead sets.

2.3 ACCURACY.

2.3.1 Accuracy

of

the

1417.

With no calibration,

the

accuracy

of

the

1417

depends

on

many factors:

the

accuracy

of

the

initial adjustmen

t,

drift in

the

standardcapac

it

or,i

ts

temperature coeff

ici

ent,

the

signal level applied. and mutual inductance in i

ts

connec·

tions. Under reasonab

le

conditio

ns

(see para

2.3

.3),

its

un-

calibrated

or

direct-reading accuracy

is

0.25

%.

The ultimate corrected accuracy

is

much

better

. Measure·

ments have been made

to

±.01

%.

This

corrected accuracy

is

possible because

the

inductive voltage dividers used

to

scale

the

capacitance

va

lues (see Section 4) are extremely precise.

This scaling accuracy is

the

RATIO ACCURACY given

in

the

specifications.

Th

is is

the accuracy of

the

ratio bet·

ween

the

capacitance v

al

ue

at

any setting

to

the

value at

1

fJF,

assuming

that

the

si

gnal level

on

the

internal standard

remains constant,

or

that

the

change caused by a changing

3

signal level

is

negligible,

or

can

be precisely determined.

To

utilize the good ratio accuracy, it

is

necessary

to

know

the

1-p

F value precisely.

A changing signal level causes a change in

the

effective

value of

the

standard

because

of

th

e non-linearity

of

the

input

impedance

of

the

voltage dividers

(not

a non-linearity

in

their ratios). The effect

is

largest

at

low frequencies

and

negligible

at

1 kHz. Typical changes with

lev

el are shown

in

figure 2-2. Methods

of

determining this level chan

ge

and

making a suitable

correct

i

on

are discussed

in

para 2.4.2 and

2.4

.3.

At

1 kHz,

the

level effect is negligible (less than .

004

%).

but

(see

para 2.2.4) mutual in

duct

ance effects become

n

ot

iceable.

The

RATIO ACCURACY given in

the

specifica·

tions for 1 kHz include

the

errors caused by mutual induc-

tance insi

de

the

1417

, which are generally negligible

com·

pared

to

that

of

the

leads

or

the

measuring device.

2.3.2 Required Accuracy.

In

calibration,

it

is always

de

sirable

to

have

the

calibra·

tion accuracy

of

the

standard

far

better

than

that

of the

instrument being calibrated. This

is

not

always possible and

is never possible

at

the

highest accuracy

level

for

any

type

of

measurement. As a result,

the

accuracy

of

the

measuring

instrument may depend

on

the

accuracy

of

the

standard

used

to

check

it.

Suppose, for example,

one

has a

standa

rd whose cal

i·

brated value is known (or can be assumed

to

be)

wi

thin

±A%

and a measuring instrument whose accuracy speci-

fica

tion

is

±8

%.

Now we

can

say:

1.

The

instrument

is within calibration if

it

reads

the

calibrated value

of

the

standard

to

±(8-A)%, if B > A, and

2. The instrument

may

be within calibration

if

it reads

the calibrated

va

lue

of

the standard

to

±(A+B)%.

There

is

an area

of

uncertainty

of

±A%

at

both+

and -

tolerance limits of

the

instrument. If A>B, it is impossible

to

say whether

the

instrument

is

within tolerance

or

not.

Fortunately.

most

manufacturers give conservative

tolerances

to

allow for their calibration uncertainties

and

for

aging and environmental effects, so

that

measurements

made in a good environment generally will

be

within

the

B-A tolerance, if

the

instrument

is

working

prope

rly,

but

only

so

long as A

is

not

too

large. If

the

i

nstrument

is

not

within the tolerance,

but

is within the A+B tolerance,

it

can

be readjusted,

or

if this

is

not

practical,

it

should be assigned

a

new

specification of ±(A+8)%.

0

-~

Baran

a

pl

o.~

;:

~

m·

,

"""

s

1

ab

1o~

zed

by

!,

!

~

i ::

tJI'

S.~:J

i

de

•

•1 1

f~

'·

U

"'ne~

f"'ly 1

ns

e•:ed

~

f'or metal t

oo

'

2::===

II

(

/a

b

1

nd

~n

g

post on

ly

"

Spade

term1na1 Clam

ps

,

-.

P

"g

en:ers

.,

Sl

ender

w

1th

• •

:n

roat al• ""eSlles

~

·

.•

""''"'

•~•

"'

' •t · '

"'

"'"

"" t•

"'

dom

• 9

"'

""

>C

Jjass

ur•

ng

cor

tact ev

en

:,

~-

~

1

ts

n•

de

acLtoo

~

uMer

nul .;

..., J : 1 1

~

•

I " •

~ W!

th

O

~,.;

!

c

u1t

1n

g

w

he

~

nut 1S loose , 1

1. 1 11b d ';:

• ..n 0 a In lng ' ,.

,4

posts

.

Figure

2·

1.

Methods

of

connection

to

binding-post

terminals.

0.3

LlC%

0.2

c

0.1

0

0.1 v 0.2 0.5 1 v 2 5

10

v

Ey

·VOLTS

RMS

Figure 2

·2

. Changes in

capacitance

vs

voltage

at

VOLTMETER

terminals.

4

,..

L

-

..o.s

'

-

I

-

!

j

:t

j

AVAILABLE INTERCONNECTION ACCESSORIES

TYPE

NO.

._._.

274-NO

274-NP

._.__

274-NL

274-LLB

274-LLR

In

many

cases,

the

accuracy

of

the

instrument

is

far

bet·

ter

than

required by a specific application,

in

which

case

this

broader

A+B tolerance

is

of

no consequence.

For

example, ifa 0.

1%

instrument

is

used

to

measure

±1

0%

capacitors, errors

of

a few

hundreths

percent in

the

calibra-

tion

of

the

1417

standard

are

of

no

importance. It may

even make sense

to

use

the

simpler calibration procedures

that

give

reduced

accuracy.

2

.3.3

Direct-Reading Measurements.

The

1417

should be accurate

to

±0.25%

without

correc-

tions

if:

1. Proper 4-terminal connections are made (see 2.2.1

and

2.2.2).

2.

The

TEST

FREQUENCY switch

is

set

to

a setting

(100Hz,

120Hz

or

1kHz)

corresponding

to

the

frequency

of

the

test

signal,

3.

The

ambient

temperature

is within

the

range

23°C

± 5

°C

and

4.

At

100Hz

or

120Hz.

the

level

at

the

VOL.TMETER

terminals

is

less

than

5 V

(or

the

voltage

at

the

CURRENT

terminals

is

less

than

0.25%

of

the

value

of

Emax given in

the

specifications)

and

the

dividers used

in

the

1417

have

not

been

left

magnitized (see para 5.3)

or

5. A 1-

kHz

test

frequency

is

used and

the

voltage

on

the

VOLTMETER terminal

is

20

V

or

less (or

the

voltage

at

the

CURRENT

input

terminals

is

less

than

Emax>·

If

these

conditions

are

not

all

met.

the

appropriate

cali-

bration

procedures

given below should

be

used

to

ensure

better

accuracy. It is good practice

to

at

least make

the

1-J.L

F

calibration

checks

to

avoid errors

due

to

long·

term

drift.

2.4

CALIBRATION CORRECTIONS FOR BETTER

ACCURACY.

2..4.1

The

1·JJF Calibration.

If accuracy

better

than 0.25%

is

required,

or

if

the

conditions

of

para

2.3.3

are

not

met,

a calibration measure·

DESCRIPTION CATALOG NO.

Oo

ubl~·plug

patch

cord,

i

n-line

920-mm

(36-in

.)

long

0274-9860

Double-plug

Patch

cord,

ri

ght-angle

920

-

mm

(36-ln.)

0274-9880

long

~hielded

dou1>le-p

lug

patch

cord,

920-mm

(36-in.l

long

0274-9883

Single-Plug

Patch

cofd,

black,

920

-

mm

(36-in

.)

long

0274-9468

Single-plug

patch

cord

,

red

.

920-mm

(36-in.)

long

0274-9492

ment

at

1

J.LF

is

necessary.

Such

a calibration will give

accuracy within 0.1%

for

all capacitance values, if

the

voltage

level

on

the

1417

is

kept

within

the

range

of

items

4

or

5

of

para

2.3.3

. If

at

1

00

Hz

or

120

Hz.

the

level

on

the

voltmeter

terminals,

Ev.

is

held

constant

for

all

measurements,

the

per-

cent

correction

obtained

from

the

1-J.LF

calibrating measure-

ment

can

be

used

for

all

values and

should

give a measure-

ment

accuracy equal

to

the

specified RATIO ACCURACY.

The

level

of

Ey

will

not

affect

the

1-kHz values

of

the

1417

so long as

it

is

below 20 V.

If

at

100

or

120H

z

this

level

(Ey)

is

not

constant,

and

high accuracy

is

required,

see para 2.4.2

and

2.4

.3.

The

1-.uF value

of

the

1417

may

be

calibrated

with

.

either

a precision 2-terminal

(or

3-terminal) bridge.

or

a less

precise bridge

with

good

resolution

and

a precision

1-J.LF

standard.

A

recommended

precision bridge

is

the

GR

1615

(or

1620

system). A 2-terminal

measurement

should

be

made

across

the

VOLTMETER

terminals

with

the

TEST

FRE

-

QUENCY switch

set

to

the

frequency

used.

This

bridge will

apply

a very low voltage

(Ey)

to

the

1417

. Less precise

bridges

may

also

be

used,

such

as

the

GR

1656

or

GR

1608,

if

they

are calibrated

by

measuring a precise

1-J.LF

standard

,

such as

the

GR

1409

-

Y.

Both

these

bridges have .01%

(or

better

) resolution

at

even decade values.

The

standard

and

the

1417

should be measured

and

the

difference between

the

measurements

is

the

difference

between

the

calibrated

value

of

the

standard

and

the

actual value

of

the

1417

.

These bridges normally

apply

a low voltage

to

the

un

known.

Th

is

calibrating meas

urement

may

also be made with

an

instrument

that

one

in

tends

to

calibrate

with

the

1417.

As above.

both

a

1-JJF

standard

and

the

1417

are measured

and

the

difference between

them

(plus

th

e calibration correc-

tion

of

the

standard

) is used as a correction

for

the

1417.

However,

to

be meaningful,

the

measurement

device used

5

must

have a resolution substantially

better

than the accuracy

to

which it

is

to

be calibrated.

2.4.2 Approximate Level Corrections.

If

at

low frequencies

(100Hz

or

120Hz)

the

voltage level

at

the

VOLTMETER terminal,

Ev

. is

not

very low or

is

not

con

stant

at

various 1417 capacitance settings, a level cor-

rection should be made for

best

accuracy. The

method

for

making these corrections

is

given in para 2.4.3. However, for

those GR instruments discussed in Section 3, an approximate

correction, accurate

to

0.1

%,

can be made

without

making

additional calibrating measurements

on

the

1417.

The procedure is as follows:

a. Measure

the

low-level

1-J,IF

value

of

the

1417

at

the

desired

test

frequency as in para 2.4.1

and

note

the

percent

deviation

from

the

nominal value.

b. From

the

tables of Section 3, note the value

of

Ev

for

each specific range

of

the

instrument

used and 1417 set·

ting. Then, from

the

typical calibration curve

of

Figure 2-2.

determine

the

approximate

percent

capacitance change.

c. Apply

the

corrections

of

a

and

b above

to

the

meas-

urements made

on

the

higher

1417

settings.

Better accuracy can be

obtained

if a capacitance

vs

level

curve

is

made for

the

specific

unit

used.

2.4.3 The Capacitance

vs

Level Corrections.

To

obtain

the

best

accuracy

at

low frequencies, it

is

necessary

to

make a precise capacitance-vs-level correction

chart

for

the

specific

1417

used,

to

measure

the

voltage

at

the VOLTMETER terminals

at

each setting

of

the 1417

du

ring use,

and

to

apply the previously determined correc-

tion. The curves should be similar

to

the typical correction

curves

of

Fioure 2-2. (This is

not

necessary

at

1kHz).

It

is

preferable

t~

demagnetize

the

dividers

of

the

1417

before

these capacitance-vs-level corrections are determined

(se

e para

5.3

).

A precision bridge is

not

necessary

to

make

this

correc-

ti

on

curve,

but

the

bridge should have good resolution and

must be able

to

apply

20

V

to

the

1-1-!F

value

or

at

least

as

much voltage

as

will be required in use. The bridge may be 2·

terminal because

the

connection of

the

1417 is

to

the

VOLTMETER terminals.

The GR

1615

Precision Capacitance

Br

idge

is

not

suitable

for these calibration measurements

at

low frequencies because

in its normal connection it applies very little voltage

to

a ·.

1-J,IF

capacitor

and

in its reverse connection the

detector

mismatch results in poor sensitivitY.

The GR

1656

Impedance Bridge

may

be used,

and,

with

a

GR

1311 oscillator, allows measurements

to

20

Vat

100

and

120Hz.

To

obtain

precision

at

low-levels,

t~e

recom-

mended connection

of

Figure 2-3 should be used, the opera·

tor,

bench, and all nearby

equipment

should be grounded,

and the

operator

should

not

touch

the

metallic surfaces

of

the

1656

during balance. While this connection leaves

the

case

of

the

1656

ungrounded,

the

1656

is battery

powered,

so

that

hum from external sources only can affect

the

measurement. This connection grounds the

1656

HIGH UNKNOWN terminal

and

the L VOLTMETER

terminal

of

the

1417,

so

that

a voltmeter with a grounded

low terminal can be used. The voltmeter should

be

dis-

connected from

the

circuit for

the

final balance,

but

the

voltage should be measured when

the

1656

is

fairly well

balanced, because the voltage

on

the

1417

depends

on

bridge adjustments. With this

method,

resolution

to

.01%

down

to

0.2

V,

is

possible if care

is

used.

The

1656

should be

set

to

the

100

nF

MULTIPLIER

setting, so

that

readings will be made

at

full scale

to

obtain

.01%

readout

resolution.

All

readings should be made using

a 9 as

the

first (left most) digit

of

the

readout. This allows

balance

up

to

9XXX =

10110,

which

is

more

than

adequate

and

minimizes

the

effect

of

errors in

the

main bridge

adjustment.

Since

the

1656

is

only 0.1% accurate and because

the

connection may introduce a small,

but

constant

error,

the

resulting calibration curve is only precise

for

relative

values.

The

precise value should be obtained

at

a specific

level (preferably a low level)

by

calibrating the

1656

in

th

is

connection with a precision standard

or

by a measurement

·on a precision bridge,

as

described

in

para 2.4.1.

2.5

PRECAUTIONS.

2.5.1 Magnetization

of

the Divider Core.

CAUTION

Do

not

apply de

to

the

1417.

De

applied

to

the

1417

can make a severe permanent

change

in

the

inductance

of

the

transformer dividers,

which can affect

the

low·frequency calibrated value. A

CAUTION

6

A

500-J..LF

capacitor must always be

connected

between

the

H terminal of the 141 7

and

the

current terminal of any bridge

connected

to

it

This protects the 141 7 from possible damage.

j

-

...,

~

1311

GR 1560-P95

PATCH CORD

1656

141 7

L H

UNKNOWN

VOLTMETER

T

ERMIN

ALS

/

B

IAS

L.-.----------------1

TERMINALS

Figure

2-3.

Connection

diagram

for

1417

calibration

checks.

noticeable change can.be made

by

as

little as 0.1 ampere·

turn

and, since there are

4000

turns

on

one

divider. it could

be affected by as

li

ttle

as

25

p.A. The result

is

a reduction

in

the

inductance of

the

cores and a reduction in

the

effect·

ive

value

of

the

1417. This effect

is

a result

of

residual magneti·

zat

i

on

of

the

cores. It can

be

removed

by

cyclic demagneti·

zation, see para 5.3.

A residual magnetization can also result from switching

large ac signals

on

and

off

.

If

the

signal level,

Ev.

is

above

10

V

at

1

00

or

120

Hz, it

is

preferable

to

reduce

the

drive

signal slowly before connecting

or

disconnecting

the

1417.

2.5

.2

Maximum

Input

Voltage.

The maximum input voltage

is

given in the specifications.

This limit

is

based

on

a maximum power

of

2

W,

a

current

of

2 A and a voltage

of

20

V.

whichever

is

limiting,

at

100Hz

.

These limits are adequate for testing all GR instruments.

Aside from possible overheating,

the

reason

for

these

limits

is

the

two

1.5-W Zener diodes (in series opposing)

that

shunt

the

VOLTMETER terminals. These clamp

the

output

to

a safe value

to

protect

the operator and

the

instrument.

If

the

voltage

at

these terminals exceeds

about

27

V rms,

these diodes will c

li

p and cause measurement errors. A maxi-

mum

of

20

V rms

is

recommended. If

the

27-V limit

is

exceeded and

the

source can deliver high power,

the

pro-

tecting Zener diodes may be destroyed, either shorting,

making

the

1417 unusable,

or

opening, and removing

the

protection feature. Usually

they

will

short

unless

extreme

power

is

applied.

2.6

USE

OF

AN EXTERNAL STANDARD.

An external standard may be used

to

get referenced values

between those decade values provided

or

to

extend

the

range.

This can be accomplished in

two

ways: by using

the

external

standard

by

itself

or

by adding it

to

the

internal

1-p.F

standard.

When

an

external stan

dard

is

connected across

the

EXT·

ERNAL STANDARD terminals and the TEST FREQUENCY

switch

is

set

to

that

position,

the

4-terminal capacitance

is

ideally

that

indicated

on

the

CAPACITANCE switch multi·

plied by Cextf

1p.F

. This value is subject

to

the

errors dis·

cussed in para 4.3.

The low frequency

error

of

the effective value

of

the EXT CAP

is

The larger

the

external capacitance, the less (percent)

error

will occur. This error can be quite large

at

low fre·

quencies if Cext

is

substantially

less

than 1 p.F. This error

can be corrected for by use

of

the procedures

of

para 2.4

If Cext

is

low, these corrections will be critical.

When using an external capacitor connected this way,

the

0 value will be

the

D of t

he

external capacitor plus a small

quantity. The D should

be

determined by a measurement

at

the

1-p.F

position and it may be set

to

a desired value

by

adding resistance

in

series with

the

external standard.

When an external standard

is

connected

to

the

VOLT-

METER terminals,

it

is

added (in parallel) wi

th

the internal

standard and thus increases all values. Ideally,

the

measured

capacitance

is

the

value indicated by

the

CAPACITANCE

switch multiplied by

(1

+ Cextl1

JJ.F).

This connection has

one

advantage in

that

the

internal padding capaci

tors

used

to

compensate for

the

low-frequency inductance

error

(para 4.3) are

st

i

ll

effect

iv

e (if

the

proper TEST FRE-

QUENCY position

is

us

ed).

The

effecti

ve

value

is

C' = 1

p.F

+ Cext + Cp - 1

/w

2 L

7

where Cp

is

the

padding capacitor

that

cancels

the

induc-

tance

term.

Moreover,

the

larger Cext is,

the

less

the

(per·

cent) error caused

by

this

term.

However, for precise meas-

urements this

error

should

be

determined

as

before

para

2.4.

The

D values will

be

approximately

(.01)

11-LF

/

(1J..LF

+Cext

l-

For precise D calibrations, this value should also be de-

termined

by

a

measurement

at

the

1

-1-~F

setting.

2.7

APPLICATION

OF

OTHER

FREQUENCIES.

The

value

of

the

GR

1417

depends

somewhat

on

fre-

quency

(see Figure 2-4). It

is

padded

to

gi

ve

good direct·

reading accuracy

at

100

Hz,

120

Hz

and

1 kHz,

but

may

be

used

at

other

frequencies.

For

frequencies between

100

Hz

and

1 kHz,

the

full accuracy

of

the

1417

may be realized by

measuring

the

1·/-IF

value

of

the

1417

at

the

desired frequency

in

the

1

·kHz

position

of

the

TEST

FREQUENCY switch

and

either

noting

the

error

for

future

use

or

increasing

the

value

to

1

/-IF,

by adding

padding

capacitors

at

the

VOLT-

METER

terminals

.

Below

100

Hz,

the

low-frequency

error

(see

para

4.3)

increases

and,

while corrections

or

padding should be used,

level effects will be increased

so

that

full

corrected

accuracy

becomes

more

difficult

to

mainta

in.

:.c

.,.

c.

-0.8

-06

- 0 4

-0

.2

100Hz

1kHz

F

REQU

ENCY

Figure 2-4. Typical capacitance change

with

frequency.

Note

corrections

are made

at

100Hz

and

120Hz.

8

Above 1 kHz,

the

low-frequency

error

is

"egligible

but

the

effect

of

mutual inductance (para 4.3)

is

increased

and

eventually

the

divider ratios deteriorate.

The

increased

error

(both ratio

and

direct

reading) should

be

less than

±(.02%) (fkHz)2, if

the

mutual inductance

external

to

the

1417

is

negligible.

The

1-kHz

TEST

FREQUENCY posi-

tion

should be used.

At

frequencies

other

than

100,

120

and

1000Hz

the

D

readings will

not

be

.01.

They

will be

approximately

(.01)

f/fs, where f

is

the

test

frequency

and

fs

the

setting

of

the

TEST

FREQUENCY switch. For accurate D values,

an

external

series

capacitor

and

resistor should

be

used which,

on

the

1

J..LF

position, give a calibrated value

of

1

1-1F

and a

D

of

.01.

2.8

TWO-TERMINAL MEASUREMENTS.

The

lower values

of

the

1417

may

be used as a 2-termi-

nal

standard

at

low frequencies.

The

D

is

increased because

of

the

terminal resistances, and

the

C value

is

changed

slightly because

of

terminal inductance·

and

the

variation in

the

a

and

13

ratios (effective terminal capacitance see para

4.4).

All

those

effects are reduced if

the

CURRENT

and

POTENTIAL

terminals

are

connected

in parallel. Table

2-1

shows

approximate

D values

and

the

change in

capacitance

from

its 4-terminal value

for

both

this parallel

connection

and

for

connections

to

the

current

terminals

only

at

120Hz

.

1417

Setting

1 llF

10

llF

100~o~F

1

mF

10mF

Table

2·1

TWO-TERMINAL

MEASUREMENTS•

Current

&

Potent.

Tet"minals Term. In Parallel

L1C(%1

D

L1CI%1

0

.01

0

+0.3 .063 +0.2

+0.8 0.24 +0.4

-1.4

0.85

-o.t5

+2.5 2.85 +1.5

"TypiClll

C

and

0

values.

D

.01

.045

0.17

0.56

1.92

-

~

-

...

•

!

•

!

•

'!

•

1

-

j

Use

with

GR

Bridges-Section

-3

3.1 USE WITH 1617 BRIDGE.

The

1617

Capacitance Bridge

is

a

1%

manual

unit

whose

accuracy depends

on

its dial calibrati

on

(an internal adjust-

ment). itsstandard <.apacitor, and its range resistors.

The

dial calibration and standard can be checked with a precision

capacitance decade

box

(see 1617 Instruction Manual),

but

a standard

of

high capacitance, such as

the

1417,

is

required

to

check

the

higher range resistors. The 1617 dial

is

logari-

thmic and equally accurate

at

readings

of

1

or

10. Therefore,

the ranges can be checked

at

either

point

, or

both.

Because the 1617 accuracy

is

only 1

%,

and

the

applied

power

is

tow (if

the

recommended

test

voltages are used),

the

0.25% direct-reading accuracy

of

the

1417

~s

usually

adequate,

but

a single calibration

ofthe

1417

at

1-pF (see

para 2.4.1l

is

nevertheless suggested

as

a precaution. The

voltage,

Ev

, reaches 3 V

on

two

measurements, which would

mean a capacitancechange

of

+0.1

%,

but

this should be neg-

ligible compared

to

the

ability

to

read

the

1617 dial.

The

1617

makes a 4-terminal measurement,

but

errors

occur if

the

lead impedances

or

term

inal impedances

of

the

1417 are sufficiently high.

To

keep these errors

at

a mini-

mum, different connections are used

on

different ranges

(see Tables

3-1

and 3-2). The remaining lead errors affect

D

on

some ranges, giving a repeatable error

on

two

lower

ranges

and

a less precise

error

in three higher ranges.

To

check

the D accuracy

on

the

higher ranges, measurements should

be made

at

10

on

the

dial.

Table 3-1 also lists

the

recommended 1617

test

voltages

(GEN LEVEL switch). When measuring high capacitance,

the

1617

is

subject

to

error caused by hum pickup. This

error can and should

be

removed by maki

ng

two

measur

e-

ments with

the

METER swi

tch

set

to

GEN

NO

RM

and

GEN REV and taking

the

average readings.

The 4-

term

inal

le

ad

set

(P/N 1617-2210) should be used

(with

the

shorting links

on

the

1617 unknown terminals

disconnected.)

At

very high values,

the

mutual inductance

~tween

the

leads

is

important. The potential leads (the

two

inner leads) should be tightly twisted

to

reduc

e this

error (see para 2.2).

If

th

is precauti

on

is

not

taken,

the

error can be more

than

10

%

at

1 F.

Table

3-2

TEST

CONNECTIONS

1417

Connections

1817Terminal

-UN KNOWN

(ouuide

)

-UNKNOWN

(inside)

+UNKNOWN

(inside)

+ UNKNOWN

(outside

)

H

POTENTIAL

H

CURRENT

L

CURRENT

L

POTENTIAL

3.2

USE WITH

1657

Oigibridge.

L

CURRENT

L

POTENTIAL

H

POTENTIAL

H

CURREN

T

The accuracy

of

the 1657 depends mainly

on

three pre-

cision resistors and a crystal oscillator. These

can

most easily

be checked by resistance measurements

of

10 n. 1

kQ

and

100

ks-2,

and one precise capacitance measurement. How-

.ever, if

one

wants assurance

that

it

measures high capacitance

values accurately also,

one

may use

the

1417.

Table 3-1

CALIBRATION WITH

1417

1617

Gen

Nom

c 0 0

Ev

Setting

Multiplier

Connection

Level!VI C

Read

.

To

l.

Nominal

To

I. (

120

Hzl

1

pF

100

nF

A

orB

2.0

10

zi%

.01 ±.001

0.24

1

pF

1

pF

A

orB

2.0

1 :!:1% .01 ±.001

1.45

10

pF

1

pF

A

2.0

10

±1%

.008

±.001 0.77

10

JJ.F

10

JJ.F

A 0.5

±1%

.008

±.001

1.2

100

JJ.F

10

JJ.F

A 0.5

10

:!:1%

.009

±.001

0.64

100

pF

lOOJ.LF

A 0.2 1

±1

%

.009

±.001

1.4

1

mF

100

JJ.F

8

0.2

10

±

1%

.0

1

~.001

0.75

1

mF

.1

mF

B 0.2 1

=I%

3.0

10

mF

1

mF

B

0.2

10

=1% .01

±.0011

1.55

10

mF

10

mF

B 0.2

±1%

3.

1

100

mF

10

mF

B

0.2

10

±1%

.01 ±.002

1.50

100

mF

100

mF

8 0.2

±2%

1.55

1 F

100

mF

B

0.5

10

±2%

.0

1 ±.011

1.0

NOTES

(I)

Use

1417

frequency

setting

cor

respo

nd

ing

to

to:st

frequency.

(2)

Make

two

measurements

with

1617

input

direct

and

reversed

and

take

average.

(3)

Twist

lead

s

at

C values

>1

mF

.

9

The

1657

is

0.2% accurate

at

the

lower values, so

that

the

1417

should

be calibrated before it

is

used. However,

because

the

1657

applies low power (see Table 3-3) a single,

low-level calibration (see para 2.4.1) for each frequency

is

all

that

is

required.

Connections between

the

1657

and the

1417

should

be

made

as

given in Table 3·4.

Th

e

1657

I+

connection

may

have a slight de offsetvoltage. A large capacitor (500

pF)

should be

put

in

series with the connection.

A

test

lead

set

should be used. The mutual inductance

of

the

1657

gives

neg!

igible error compared

to

its specifica-

tions,

but

care should be used

in

the

lead configuration when

measuring 1

mF

at

1 kHz.

Note

that

the

maximum reading

is

999.99

mF

at

1 kHz

and

99.999

mF

at

120Hz.

It

would

be

preferable

to

use

an

external

standard

with the

1417

,

of

say

0.9

pF

to

check

these range extremes.

Table

J.J

1657

ACCURACY

1417

120Hz

1kHz

Setting

LIC%

0 Range

LI

C(%)

1

~F

0.2 .

001

2 0.2

10~F

0.2

.001

2 0.2

100

IJF

0.2 .

001

0.2

1

mF

0.2 .

001

1.

10mF .005

1

00

mF

10

..

.05

•Goes

to

999

.

99

~F

.

••Goes

to

99

.

999

mF

1657

P-

(rear

leftl

P+

(rear

right)

1-

(front

left)

1+

(front

right)

Table

J..4

INTERCONNECTIOtJ

3.3

USE

WITH 1683BRIDGE

D Range

.001

2

.001

.

01

1417

L

POTENTIAL

H

POTENTIAL

I..

CURRENT

H

CURRENT

Ev

(120-Hz)

0.15

.07

1.25

0.64

Low

The GR

1683

Automatic

RLC Bridge measures

at

both

120Hz

and

1 kHz

to

a base accuracy

of

0.1% (see Table 3-5)

so

that

calibration

to

full accuracy requires careful calibra-

tion

and use

of

the

1417

.

CAUTION

GR

1683

applies

2.2

V

at

1 A; danger

of

residual

magnetization (see para 2.5.1.)

At

120

Hz, a voltage-vs-level curve should be used

to

obtain full accuracy (see para 2.

4.2).

At

1

mF

full

output

of

the

1683

can give

25.5

Von

t

he

1417

VOLTMETER

terminals, which

is

close

to

the

Zener-diode clipping

point.

It is suggested

that

the

GR

1683

OSC LEVEL be

set

to

1 V

for

this measurement. Because

of

the

high level supplied

by

this bridge, it

is

recommended

that

the

1417

be

con-

nected

and

disconnected

and

all range changes

made

with

the

1683

OSC LEVEL

set

to

0.

At

1 kHz

there

is

negligible level

error

but

the

above pre-

caution

is

recommended

to

avoid changes in

the

1417

cali-

bration

due

to

transients.

The measurement connections between

the

1683

and

1417

are given in Table 3-7 . The mutual inductance

error

10

caused by

the

1683

is

negligible compared

to

its specifica-

tions, if

the

current

leads

are

twisted

together

and

the

po-

tential leads are twisted together.

Table

3-5

1683

ACCURACY•

1417 120Hz 1kHz

Setting

!lC(%)

0

LIC(%1

D

1

p.F

0.11

.0011

0.11

.0011

10

IJF

0.11

.0011

0.11

.

0011

100

iJF

0.11

.0011

0.11

.0011

1

mF

0.11

.0011

1.1

.01

10mF

1.1

.01

6 0.1

100mF 6

0.1

•1683

C

reading

of

10000

and

0 • .01

Table

3-6

TYPICAL

VOLTAGES

AT

VOLTMETER

TERMINALS

1683

OSCILLATOR

SETTING

(120Hz)

1417

Setting

Full

1V

1

~F

2.1 1

10

p.F

5.3 2.5

100~F

14

.7 6.9

1

mF

25.5

12

10mF

17

,8 8.4

100mF 7.6 3.6

Table

3-7

BRIDGE..STANDARO

CONNECTIONS

1683

+CURRENT

-CURRENT

+POTENTIAL

-POTENTIAL

.3.4 WITH 1685/1686 INSTRUMENTS.

1417

CURRENT

H

CURRENT

L

POTENTIAL

H

POTENTIAL

L

The

1685

Digital Impedance Meter

and

the

1686-A

and

1686-9000 Digital Capacitance Meters are very similar in

operation

and

specifications. The

1685

measures resistance

and inductance as well as capacitance

and

its

ratio

and

accuracy can be

most

easily checked

by

measurements

of

pre-

cision resistors. However, a

1417

is useful

for

checking its

operation for high-capacitance measurements. The 1686

only

measures capacitance,

so

a standard such as

the

1417

is

necessary

to

check

its high-capacitance ratio errors.

The

1685

and

1686

have

the

same

s~cifications

for

capacitance,

but

the

1686

applies slightly more power

on

some

ranges (see Tables 3-8

and

3-9). The

1686-A

has

slightly improved specifications compared

to

the

1686-9000

and

has an additional range up

to

200

mF.

Each has rear-

panel level control

that

should

be

set

full

O"n

(cw).

All

these

instruments will

make

a measurement

at

120

Hz up

to

5

times

the

full-scale value

on

the

highest range.

The

error

"E"

will appear in

the

readouts

but

the

measurement is

valid, if

not

too

accurate (see Tables

3-8

and

3·9)

.

The connections are

the

same for all

three

instruments

and are given

in

Table 3-10.

The

test

lead sets

1685-9600

for 1685, P/N

1686-9602

for

1686 should be used. These

instruments

with

their

lead sets have approx.

0.2

~H

mutual

inductance between

the

current

and

potential pairs which

j

1,

1

,

j

~

j

causes an

error

on

the

highest ranges (see Table

3-11

and

para 4.3). This error increases

the

measured value and

should be subtracted from all measurements

to

obtain

the

best accuracy.

Table 3-8

1685/1686-9000

ACCURACY•

120Hz

1686

1685

1417

120Hz

1kHz

ev ev

Setting

AC(%) D

AC(%1

D

Approx

1

JJF

:!:0.35

:!:

.0019 :!:0.12

:!:

.

0019

1 0.95

10

JJF

:!:0.35 :!:.0019 :!:0.12

:!:

.

0019

3.5 3.3

100JJF :!:0.35 :!:.00

19

:!:0.36 :!:.0057 7.7 7.7

1

mF

:!:0

.7 -:.0038

:!:1.2

••

:!:.019

t.7

1.7

10mF

±1.4 :!:.0176

0.9

100mF

6t

0.4

tOverrenge

.

•

At

half

full

scale

and

0 • .01.

• •

Range

not

recommended.

Table 3-9

1686-A

ACCURACVt

1417

120Hz

1kHz

Ev

Setting

AC(

%)

D .6CI%) D (120-Hz)

1

JJF

±0.35

!.0

019

±0.12

10

JJF

±0.35 ±.0019 :t0.12

100

llF

:!:0

.

35

:!:.0019

±0.36

1

mF

:0

.7 ±.0038 :t1.2

10mF

:!:1.05

:.0176

100mF

±3.4

±0

.

11

1 F

28•

•Qverr

ange

tAt

half

full

scale

and

0

~

.01

.

1685-1686 Cable

I+

REO

Table 3-10

CONNECTIONS

P+

REO,

WHITE

SLEEVE

P-

BLACK.

WHITE

SLEEVE

1-

BLACK

±.00

19

!.0019

±.0057

:t.019

1417

H

CURRENT

H

POTENTIAL

L

POTENTIAL

L

CURRENT

1

3.5

7.7

1.7

0.9

0.4

0.1

Table 3-11

MUTUALINOUCTANCE

EAROR-(APPROX.)

1417

Settin~

120

H

1

j.l

0

10

llF

0

1001lF

0

-

1

mF

+.01%

10mF

+0

.11

%

100

mF

+1

.1%

1 F +

11

%

3.5

WITH

2230

SYSTEMS.

1kHz

0

+.01 %

+.08%

+.8

%

+8%

The bridge

of

the

2230

Component

Network System

actually measures C parallel and G parallel,

but

the

computer

in

the

system converts

the

measurement

data

tagive

a·dis-

pl~y

of

Cseries and

D.

It measures capacitors

at

120Hz

and 1 kHz

to

good accuracy (see Table 3-12)

so

that

the

correction procedures of para

2.4.2

or

2.4.3

should be used.

The mutual inductance

of

the

leads

of

this bridge does

not

affect

the

accuracy because

of

its limited capacitance range.

The connections are given in Table 3-13.

One precaution should be considered.

The

I+

connect

ron

is

not

decoupled from

the

signal source. This source should

have a very low offset voltage,

but

if

not

adjusted properly

it can supply direct

current

to

the

1417. If

the

measurements

are in error, a large capacitor should

be

placed in series wi

th

this connection

to

determine if this

is

cause

of

the

inaccuracy.

Cap

.

Value

1tJF

10tJF

100tJF

1mF

Table

3-12

2230

AUTOMATIC

TEST

SYSTEM,

CGAL

BRIDGE

Range

5

6

7

8

120Hz

1kHz

AtXu

ra

ey

c D

Ev

Rang•

:.125% :

.001

1 6

:.

12

% : .

001

3.1 7

:.12%

:.001 8.4 8

·

.1

4%

:.

002 16

TABLE

3-

13

2230

I+

P+

P-

I-

I

NTERCONN

ECTIONS

1417

H

CURRENT

H

POTENTIAL

L

POTENTIAL

L

CURRENT

Accuracy

c

:.12%

:.

12

%

:

14

%

D

:.

001

:.

00

1

:.00

2

11

-

------

----

-----------

--

-

3.6

USE

WITH

QuadTech

Digibridges

3.6.1 General

The

1417 Four-Terminal CapacitanceStandard is

useful for checking the accuracy

of

the QuadTech

low

frequency Digibridges at

high

capacitances. These

instruments

are

calibrated using precision resistors,

but

the 1417 can give assurance that they measure high

capacitance accurately. These instruments have

different specified accuracy

and

different capacitance

ranges so that detailed information

is

given for each

in

the paragraphs that follow. All

of

themhave accuracy

better than

the

0.25% direct-reading accuracy

of

the

1417 and therefore 1uF calibrations

of

the 1417 are

required and some care necessary

to

realize its high

ratio accuracy (refer

to

paragraphs

2.4

.2

and

5.3)_

In

some cases the specified

Digibridge

accuracy

is

as

good

as

the 1417 ratio accuracy so that a true accuracy

verification

is

not

possible but nevertheless, tests with a

1417 gives added confidence.

The

specified accuracy given for all instruments are

for their SLOW measurement rates

if

there

is

a choice

of

test rates. Greaterrepeatability can be achieved by

averaging several measurements either manually

or

automatically

if

this feature is available.

3.6.2 Connections

Four-terminal connections should

be

made to the

1417

as

shown

in

Table 3-14 below using the 1657-

9600 Extender Cable (1657, 1658, 1659, 1689, 1692)

or

the 1689-9602 Extender Cable (1689M, 1693).

The

Digibridge current terminals may havea slight de

voltage so that a large capacitor,

>200

uP,

should

be

connected

in

series

with

the H

CURRENT

terminal to

keep de from flowing

in

the precision transformers (this

may

be

an

electrolytic capacitor).

Table3-14

Digibridge Test Connections

Digibridge Connection

1417

Terminal Cable

Plug

Color

Code

1657 1659, 1689 Test Fixture

Connection

1658 1692, 1693

HPOTENTIAL

LPOTENTIAL

HCURRENT"'

LCURRENT

CASE

*Add series capacitor

Red-white

Black-White

Red-Red

Black-Black

Green

P+

P-

I+

1-

Ground

3.6.3

Zeroing

and

Minimizing

Mutual

Inductance

Mutual inductance between the

test

leads can cause

an

error, particularly

at

I

kHz

(refer to paragraph

2.2.4),

even though the position

of

the terminals

on

the

1417 minimizes this effect. It is also suggested that

the

free ends

of

the

testcable be twisted

to

avoid coupling

between the currentand potential circuits. Twist the

two

POTENTIAL

leads together

and

twist the two

CURRENT

leads together.

Many

instruments have

short-circuit calibration capability that should

be

used

to

remove

this

effect.

The

suggested procedure

is

as

follows: · ·

a.

Set

the Digibridge to the test frequency

to

be

used.

b. Connect

the

1417 as

in

Table

3-14 except plug

the

high potential

plug

(red-white) into the low potential

plug (black-white) which is plugged

in

the L

POTENTIAL

terminal

of

the 1417.

c. Set the 1417

CAPACITANCE

switch

to

IF.

d.

Make

the

SHORT

calibration

of

the test instrument.

e. Reconnect the high potential

plug

into the H

POTENTIAL

terminal being careful to move the leads

as little as possible.

PH

PL

ill

IL

Ground

right rear

left rear

right front

left front

threaded holes

f. Reset

to

desired capacitance value.

3.6.4

Use withthe

QuadTecb

1657

or

1658

Digibridges.

These two instruments both have

three

ranges which

are usually checked

by

three calibrated resistors (10

ohms, 1 kohms,

and

100 kohms).

Their

capacitance

ranges extend to 99999

uF

at

100

or

120Hz

and

999.99

uF

at

1 kHz.

This

means that 100

mF

at

120Hz

and

1000

uF

at 1kHz

may

only

be

measured

if

they are

lower

than the nominal value. A chart

of

the C and D

accuracy

of

the 1657

vs

capacitance

is

given

in

Table

3-

15. Divide the accuracy values

by

two

for the 1658

using its SLOW measurement rate.

Both

of

these instruments can apply enough current

to

give

a value for

Ev

of

over

1 V

at

100

uF

(refer

to

Figure

2-2)

but

both instruments should still measure

well within their accuracy (relative

to

the 1

uF

calibrated value at each frequency).

Neither

of

these instruments makes a SHORT

calibration so that

care

should

be

used

to

keep the

mutual inductance

at

a minimum (refer

to

paragraph

3.1.2).

j

j1

j]

j

Table3-15

1657 Specified Accuracy

1417 100

or

120Hz

1kHz

Setting c D Range c D Range

1

uF

0.2%

.001 2 0.2% .001 2

10uF

0.2%

.001 2 0.2% .001 1

1

00uF

0.2%

.

001

1 0.

2%

.001 1

1

mF

0.2%

.001 1 1%* .01 1

10mF

1% .005 1

lOOmF

10%*

.05

1

*At

range limit: 999.99 uF

at

1kHz,

99999

uP

at

120Hz

3.6

.5 Use

with

the

QuadTech

1659

and

1692

Digibridges.

The

1659

and

1692 Digibridges

have

direct reading

ranges well beyond the

range

of

the

1417 standard

but

their specified accuracy gets very

high

above

1 F so that

there

is

no

point

in

checking accuracy

above

that value.

The

actual accuracy will bewell below the values given

in

Table

3-16

if

a

SHORT

calibration

is

made

at

the

test

frequency and many measurements are

averaged.

The

1659 applies

0.3

V

max

but

the 1692

has a choice

of0

.3 V

or

1 V (

open

circuit).

The

0.3

V

level

of

the

1692 should

be

used

up

to

1 mF

at

120

Hz

and

up

to 100

uF

at

1kHz.

This

0.3

V level will cause

an

noticeable increase

in

the

100

uF

value at 100

or

120

Hz,

about0.03%

,

which

is less than the 1692

specification,

but

which should either

be

allowed for

or

reduced by reducing the signal level by putting about

30

ohms

in

series

with

the H CURRENT connection.

Table 3-16

1659

and

1692 Specified Accuracy

-------------1659-

-

------

- -

1417 100 &

120Hz

1kHz

Setting c D c

1

uF

.1%

.0005

.1%

lOuF

.1%

.0005

.105%

lOOuF

.105%

.0005

.15%

1 mF

.15%

.00075 .6%

10mF

.6%

.0030

5.1%

lOOmF

5.1%

.0255

1 F 50% .25

3.6.6

Use

with

the

QuadTech 1689, 1689M

and

1693.

D

.0005

.00052

.00075

.0030

.0255

These instruments

have

direct reading capacitance

ranges

to

99999

uF

(just

under

100

mF)

but this may

be

extended

by

using the

RATIO

mode.

To

do

this:

a.

Go

to

range

4

by

measuring a

high

C

value

or

se

tting

[4]

[=)[SHIFT]

[SPECIAL] [1)

b.

Enter

1000

as

the

nominal value

[1000]

[=][NOM

VAL]

c. Set

RA

110

mode

[2]

[=)[SHIFf]

[SPECIAL] [6]

The

reading is

now

in

mF

(although there

is

no

unit

displayed)

and

the

range

is

extended

to

99999 mF

or

99.999

Farads.

To return

to

the

normal mode enter

[0] [

=]

[SHIFI'] [SPECIAL] [6]

---

- -- -1692

--------

---------

100 &

120Hz

1

kHz

c D c D

.05% .0003 .05% .0003

.05% .0003 .055% .00033

.055% .00033 .1 % .0006

.10% .0006 .55%

.0033

.55% .0033 5.05% .0303

5.05%

.0303

50% .3

These

three instruments are

very

si

milar

in

behavior

but

are specified

diff

ere

ntly. They all

have

measurement acc

ura

cy

at

1kHz

at

1

uF

and

10

uF

tha

t

is comparable

to

the

1417 ratio acc

ura

cy so that a

tru

e

accuracy verification is diffic

ult

under

these conditions

(refer

to

paragraph 2

.3.2),

but these measurements

can

increase confidence.

The

1417 ratio a

cc

uracy is

adequate under most o

ther

conditions.

At

very high

values the instrument specs

are

very

broad and easily

met.

How

e

ver

,

if

a zero

is

made

and

many

measurements averag

ed

,

the

actual Digibridge accura

cy

can

be

quite good. 1% accuracy at 1

Fat

100Hz

is

possible.

These instruments have

su

ch tight tolerances that the

changes in

the

1417 values

with

signal level

are

important.

This

leads to a somewhat paradoxical

situation

in

that lowering

the

signal level will give

!!!!!!!

--

-

bettervalues

of

the 1417 standard but will result in

reduced specified accuracy

of

the instruments.

Fortunately this occurs only

at

100

or

120

Hz

where the

spec

is

broader. Moreover, the accuracy at reduced

levels

is

much better than specified, particularity

if

Table3-17

several measurements are averaged, so that the specified

accuracy

at

1 V should

be

met and thus Table3-17

gives these values even though

it

lists recommended

signal levels that are substantially lower at 120 (or 100)

Hz.

1689M and 1693 Specified Accuracy

--------1689

&

1689M------

-----------1693-------------

1417 100 &

120Hz

1kHz

100 &

120Hz

Setting c D c D c D

1uF

.04% .0004 .02% .0002 .04% .0004

10

uF

.04% .0004 .02% .0002 .041% .00041

100uF

.04% .0004 .05% .0005 .047% .00047

1

mF

.144% .00144 .41% .0041 .112% .00112

lOmF

1.44% .0144 4% .

04

.76% .0076

100

mF"'"'

14.4% .144 7.24% .0726

1 F"'* 144%

1.4

72% .

72

* Recommended programmed voltage

at

100

or

120

Hz,

see text

**Use

RATIO mode to extend capacitance range

1kHz

c D

.02% .0002

.022% .00022

.04% .0004

.22% .0022

2.02% .0202

TestY

120Hz"'

.25 v

.

25

v

.1 v

.sv

IV

lV

IET Labs 1417 User manual

Popular Touch Terminal manuals by other brands