Text preview for : 147(Model147Nanovolt_Null).pdf part of Keithley 147(Model147Nanovolt Null) Keithley 147 147(Model147Nanovolt_Null).pdf
Back to : 147(Model147Nanovolt_Null | Home
INSTRUCTION MANUAL
MODEL 147
NANOVOLT NULL DETECTOR
OCOPYRIGHT 1972, KEITHLEY INSTRUMENTS, INC.
PRINTED, APRIL 1977, CLEVELAND, OHIO, U. S. A.
CONTENTS MODEL 147
CONTENTS
Section Page
1. GENERAL DESCRIPTZON---------------------------------------------- 1
2. OPERATION-------------------------------------------------------- 5
3. APPLICATIONS----------------------------------------------------- 21
4. C1I
5. MhINTENANCE------------------------------------------------------ 31
6. REPLACEABLE PARTS------------------------------------------------ 51
SCHE~T~CS------------------------------------------------__-----__-- 65
0477 ii
MODEL 147 ILLUSTMTIONS
ILLUSTNATIONS
igure
NO. Title Pag:e
______- i
la. FrontPanel.. ............................. 1
lb. Front Panel with Cable ......................... 3
2. Front Panel Controls .......................... 4
Model 147 Rear Pmel Controls & Connections. .............. 6
a: Model 1481 Low-Thermal Input Cable ................... 10
5. Model 1483 Low-Thermal Connection Kit. ................. 10
6. Bus System for Model 147 ........................ 12
7. Thermal Sink Construction. ....................... 15
a. Normal Wave Form at Demodulator with Input Shorted ........... 17
9. Wave Form at Demodulator Shown with Some Pickup. ............ 17
10. Wave Form at Demodulator when Amplifier is Saturated .......... 17
11. 8-cps Filter Circuit for Recorder Output ................ 18
12. Using Model 147 with 4-Terminal Connections. .............. 18
13. Exploded View for Rack Mounting .................... 19
14. Circuit UsingGuildline9144 with Model 147 Null Detector ........ 21
15. Circuit Using Guildline 4363DL with Model 147 Null Detector. ...... 22
16. Circuit Using Guildline 9120 with Model 147 Null Detector. ....... 22
17. Circuit Using Biddle 605001 with Model 147 Null Detector ........ 23
18. Circuit Using Leeds & Northrup 7556. .................. 23
19. Circuit Using Model 147 Null Detector with ES1 240, 8OOR and KS 925. .. 24
20. Block Diagram of Model 147 Amplifier Circuits. ............. 25
21. Model 1.47 Tnput Circuit. ........................ 26
22. Diagranl of Power Supplies & Battery Charging Circuit .......... 29
23. Model 147 Input Compartment. ...................... 32
24. Correct Wave Form in dc-to-dc Inverter ................. 36
25. Correct Wave Form in Oscillator Circuit. ................ 38
26. Improper Wave Forms in Oscillator Circuit. ............... 38
27. Correct Wave Form at Demodulator Test Jack ............... 41
28. Out-of-Phase Wave Form at Demodulator Test Jack. ............ 41
29. Top View of Model 147 Chassis ..................... 44
30. Bottom View of Model 1~47 Chassis .................... 45
31. Transistor Locations on Printed Circuit 76 ............... 46
32. Capacitor & Diode Locations on Printed Circuit 76. ........... 46
33. Resistor Locations on Printed Circuit 76 , ............... 47
34. Resistor 6 Test Point Locations on Printed Circuit 76. ......... 47
35. Resistor b Test Point Locations on Printed Circuit 74, Bottom Face ... 48
36. Component Locations on Printed Circuit 74, Top Face. .......... 48
37. Resistor & Test Point Location on Printed Circuit 75 .......... 49
38. Capacitor & 'Transistor Locations on Printed Circuit 75 ......... 49
39. Resistor Locations on RANGE Switch (5102). ............... 50
40. Resistor Locations on RANGE Switch (S102). ............... 50
0477 iii
SPECIFICATIONS MODEL 147
SPECIFICATIONS
iv 0477
MODEL 147 NULL DETECTOR GE:NLSAL DESCRIPTIOX
SECTION 1. GENERAL DESCRIPTION
l-1. GENERAL.
a. The Keithley Model 147 is designed specificaLly RS a col,vcnicnt self-contained dc
electronic null detector. Its sensitivity is 0.6 x LOT3 microvolt per mi.llimrter or 0.03
x 10-e microampere per millimeter. Its resolution is better tllnn 3 nanovolts with a LO-oll:il
source resistance and 10 nanovolts with a 300-ohm source resistance. This corresponds Lo
a power sensitivity of 3 x lo-21 watt. Zero shift is less than 15 nanovolts for sourcf
resistance changes from 0 to 300 ohms. Line frequency r~!jcction is better than 5OOO:l 011
its most sensitive range.
b. The Model 147 has 16 ranges from 30 nanovolts full scale to 1.00 millivolts on a
zero-center meter. Meter accuracy is f2% of full scale on ~11 ranges.
c. For reliable and versatile use, the Null Detector is of solid-state design, cxccpi
for the first two input stages. It has high line isolation - lOl(J ohms - and battery
or ac power line operation.
FIGUPJ? La. Front ,>anel,.
0572 I
GENERAL DESCRIPTION MODEL 147
l-2. FEATURFS,
a. Battery operation permits complete isolation from ac power lines, eliminating many
grounding problems. Battery operation also allows flexibility and convenience in use.
The Model 147 automatically recharges the battery if needed when the powercord is connected.
b. As an electronic null detector, the Model 147 is immune to mechanical vibrations.
It will also recover from a Z-volt overload on its inost sensitive range in less than 20
seconds.
c. Besides performing as a null detector, the Model 147 can also be used as a 2% direct-
reading nanovoltmeter.
d. The Null Detector has an output of ;tl volt at up to 1 milliampere for full-scale
meter deflections to drive a recorder or oscilloscope. Output accuracy is ;tl% of full
SCSlS.
e. A zero suppression circuit, furnishing up to 100 microvolts, permits measuring small
changes in a larger dc signal or compensating for thermal emf's.
l-3. APPLICATIONS. (Also see Section 3.)
a. The Model 147 is designed specifically as a null detector. It has sufficient sensi-
tivity to be used in most applications with all commercially available potentiometers, in-
cluding 6-dial models, ratio sets and resistance bridges, including Wenner, Wheatstone and
Kelvin Double Bridges. It can be used to make 4-terminal measurements.
b. Keithley's Model 147 is more sensitive than the finest galvanometer systems. It is
also immune to mechanical vibrations, thus eliminating the need for shock-free mountings.
Additional advantages over galvanometer systems'include the ability to recover from 2-volt
overloads in 20 seconds, much less off-null loading, plus considerably faster speed of
response.
0572
MODEL 147 CKNERAL DESCKIPTION
0572
GENERAL DESCRIPTION MODEL 147
FIGURE 2. Front Panel Controls.
0572
MODEL 147 NULL DlXI:C'l'OR
SECTION 2. OPERATION
2-l. FRONT PANEL CONTROLS. (See Figure 2.)
a. AC CONNECTI?DLamp. l:hc Lamp is lit whcncvcr the unit is connected t0 i.lll xi. pwur
line and the POWERSUPPLY Switch is in tile AC or OFF position,
NOTE
b. BATTERY CHARGING Lamp. When lit, this Lanq indicates the ibattery ins clrnr:ini:. p1,:
charge current determines its brightness. If the lamp is noi l~it, tllc" LllC lbilCLtdl.\ ii
charged.
c. POGJIZR
SUPPLY Switch. Ihe Switch controls tlrc mode of opcratiori for tile powel- suppI!:.
1. AC position: The Null Detector will opcratc from the nc power Lint. 'TIIC
battery will be charged if needed; then, the BATTEQY CIIARGINC Lamp will lil:ht.
2. OFF position: The Model 147 is not operating. Howcvcr , tlic battery wil~l lbc
charged, if nccdcd and if the power cord is connected.
3. BATI'ERY position: The Null Detector is operating from its battery. 'I`liC ni'
power line is internally disconnected, whether or nut tile power cord is coluieclcd;
the AC CONNECTiXD Lamp is off; the battery cannot be charged.
4. BATT. TEST position: When the POWERSUPPLY Switch is held in this posil:ioii,
the Model I.47 shows the state of the battery charge directly on its mctcr. Al I
circuits within the instrument are the same as for battery operation exccpl at Lllc
meter terminals.
POWERSUPPLY Power Cord AC CONNKTED BATTERY CIIARGINC La"{'
Switch Setting I Connected Lamp I (If battery is charging)
AC Yes 0" on
YCS On Oil
OFF
NO Off Battery CllnnOt bc cli;lr::c!d
Yes Off Battfry cannot IbC char;:ed
BATTERY
NO Off Battery cannot hc chnr~cd
Yes Off Battery can,,ot bc cllar~:cd
BATT TEST NO Off LIattcry cannot be charged
TABLE 1. Indicating Lamps and POWERSUPPLY Switch Settings. The Table SllOWS tile
relationship between the front panel lamps, the power cord and the I'Oli'ER SUl'1~'I.Y
Switch setting.
0365
MODEL 147 NULL DETECTOR
OPERATION
d. RANGE Switch. The RANGE Switch selects the full-scale meter sensitivity (either
microvolts or millivolts) for one of eight ranges, from 0.03 to 100.
e. FUNCTION Switch. The FUNCTION Switch selects the function - MICROVOLTS or HILLI-
VOLTS - which is to be measured.
f. ZERO SUPPRESSControls. Two controls determine the amount of zero suppression.
1. The COARSE Control disconnects the suppression circuit (in OFF position) or
selects one of four suppression voltages in decade steps. Refer to Table 3.
2. The FINE Control is a continuously variable adjustment for the suppression
voltage set by the COARSE Control. It adjusts the range between the positive and
negative values of the maximum voltage set by the COARSE Control.
g. INPUT Receptacle. The INPUT Receptacle is of a special low-thermal design. Use
only the Models 14.81, 1482, 1486 and 1488 for mating connectors.
FIGURE 3. Model 147 Rear Panel Controls and Connections. Circuit designations refer
to Replaceable Parts List and schematic diagrams.
2-2. RTtiR PANEL CONTROLSAND CONNECTIONS.
a. Line Voltage Switch. The screwdriver-operated slide switch sets the Model 147 for
117 or 234-volt ac power lines.
b. Fuse.
2. For 117-volt operation, use a 3 AG or MDL Slow-Blow l/@ampere fuse.
2. For 234-volt operation, use only a MDL Slow-Blow l/16-ampere fuse.
c. Power Cord. The 3-wire power cord with the NEMA approved 3-prong plug provides a
ground connection for the cabinet. An adapter for operation from Z-terminal outlets is
provided.
6 0365
MODEL 1~47 NULL DETCC'TOR
NOTE
A note above the power cord shows the ac power line i:rcqucncy for
which the rejection filter is ad.justcd. 'The instrument will work
at any line frequency from 50 t" 1000 cps, but ac rejection is bcsc
at the indicated frequency. Paragraph 2-18 dcscribcs adjustirl~; lilf
filter circuit.
d. DEMODULATORTEST Jack. A phone jack provides access to LllE denl"dLll~ator ior ti.5:
purposes.
f. OUTPUT. 'The OUTPUT Rcceptaclc provides iL volt at one! mill.iampcrc f:or a iilll.-scale
meter deflection on any range.
f. GND and LO Tcrmiasls. 'The ground terminal. (GND) i.s conncctcd to tllc chassis ;~II)~!! 1)~.
third wire of the power cord. The low terminal is connected t" circui.t ground and cl!&:
low side of the input connection.
2-3. MODE 01: OPERATION. The Model 147 operates cithfr irom an ac power line or irom iis
battery. b-or illost "SfS, it functions well from ac. Use battary operation, llo~ifver, i~1~
the ac power line will create ground loop or isolation problems. 1solnfi"n irun low to
ground is complete for battery operation when tllc power cord is disconnccl:cd; i,t is
greater than lOlo ohms wit11 the power cord connected. Al.so use battery operntiiln it>
reduce the 8-cps ripple which may appear at the output with i-11" input shorted in i&c
operation. See paragraph Z-16.
NOTE
Before using the battery operation, tll"r"ugllly read paragraph 2-4.
Improper battery "pcration can damage the battery pack and lead to
inaccurate measurements.
2-4. UXTERY OPERATION.
The Model 1.47 is supplied with a rcchargcablc h-volt, 4 ampere-ilour Iliclr~l-~~l(llnilIr!l
ba:tery pack (Model 1489). Recommended: Do n"t usf the battery more tlliln cigirt COII-
secutive hours without recharging. At this discharge rate, the battery sl10u1d last abOUt
1000 recharge cycles.
NOTE
Permanent damage to the battery pack occurs if it is used for more
than 16 consecutive hours without recharging. At this discharge
rate, the recharge cycles are greatly reduced. Bflorf usi~ng tile
Model 147, check the state of the battery charge.
b. Cheek the battery charge before making a measurement. Hold the POWERSUPl'LY Switch
in the BATT. TEST position. The minimum acceptable charge is a meter indi.cntion oi ,8;
full. charge is shown by the BATTERY CHARGING Lamp not being lit. Recharge il needed.
Otherwise, battery operation is the same as for the ac power line operating mode; rcI:cr
to paragraph 2-5.
0365
OPERATION MODEL 147 NULL DETECTOR
When the battery is used beyond its capacity, two effects aye seen.
There is a shift in zero offset from ac to battery operation. Also,
the power supplies do not regulate and high ripple voltages appear
at the supply outputs. (See paragraph 5-8.)
c. To recharge the battery, connect the power cord to an ac power line. Turn the
POWERSUPPLY Switch to AC or OFF. The BATTERY CHARGING Lamp will light. The battery will
be charged only if needed, and the circuit automatically prevents it from being overcharged.
d. It is suggested that the battery be used during the day and be recharged at night.
Leave the instrument always connected to the ac power line; then turn the POWEKSUPPLY
Switch to OFF at night. After a fully charged battery is used for eight consecutive
hours ) it will recharge within 16 hours. Leaving the power cord connected has little
effect on the isolation: 1010 ohms with the POWERSUPPLY Switch in BATTERY position and
the shorting link between GND and LO Te?.minals disconnected.
2-5. OPERATING PROCEDURES,
a. Set the front panel controls as follows:
POWERSUPPLY Switch OFF
FUNCTION Switch MILLIVOLTS
RANGE Switch 100
ZERO SUPPRESS COARSE Control OFF
NOTE
Make sure the ZERO SUPPRESS COARSE Control if OFF. If it is not, a
suppression voltage is introduced, causing an error in measurements.
b. Connect the voltage swrce or null circuit to the INPUT Receptacle. Refer tO para-
graph 2-6 for suggestions.
c. Check the voltage shown on the rear panel Line Voltage Switch; connect the Model 147
to the ac pcwer line. Make sure the frequency shown above the power cord is the frequency
of the ac power line. At this point, the AC CONNECTEDLamp will light, asp will the
BATTERY CHARGING Lamp if the battery is being charged. If the circuit low is to be at
ground, put the shorting link between the LO and GND posts on the rear panel.
d. Turn the POWERSUPPLY Switch to the desired mode of operation, AC or BATTERY.
e. Increase the Model 147 sensitivity until the meter shows the greatest on-scale de-
flection.
1. Check the source resistance to make sure it is within the maximum value specified
for the range being used. (See Table 2.) If the maximum resistance for the more sen-
sitive ranges is exceeded, the Model 147 may not measure within its specifications.
2. Zero offsets seen when the Zero Suppress Controls are off will vary with the
quality of the circuit's thermal construction, See paragraph Z-14. When a Model 1488
Shorting Plug is connected to the Model 147 INPUT Receptacle, offset should be less
than 0.3 microvolt.
8 046613
OP1:llATION MODEL I,47 NULL I1lYl'ECTOR
1,, Make sure the signal is greater than Johnson noise in the source resistance (par-
agraph 2-12 ).
2. Use materials which generate a low thermal emL (paragraph 2-14).
3. Mini.mize temperature changes and thermal gradients (paragraph Z-14).
4 Reduce magnetically induced signals by proper shielding and minimizing experimen-
tal layout area (paragraph 2-15).
5. Eliminate ground loops through proper grounding and connection to the signal cir-
cuit (paragraph 2-16).
2-6. LOW-THEIMAL INPUT CONNECTIONS.
a Tlie easiest connection to the Model
14~7 input is with the Model 1481 Low-Ther-
mal Input Cable, supplied with the instru-
mcnt Use tile Cable for temporary setups,
for measurements at several points, and
when fast connections are needed. The Mo-
del 148~1 connects directly to the INPUT
1~eceptacl.e.
b. Where more permanent setups are pos-
sible or where very Sow thermal connections FIGUKC 4. Model 1481 Low-Thermal Input
arc needed, use the Model 1482 Low-Thermal Cable. 'The Model 11+82 Low-Thermal Input
Input Cable. It is similar to the Model Cable is similar except it has bare copper
1481 , except it leas bare copper leads in- leads instead of alligator clips.
stead of nl,ligator clips. Clean the bare
wire! width a non-metallic abrasive, such as
Scotch Isrite, before making the connection. Making crimp connections, as possible with
illc> Model 1483 Kit, is best.
('. Si cadmium solder (Model 1503) is used for a connection, make sure the soldering
iron used is clean and that it has not been used with regular solder before. USC only
rosin solder flux. If possible, heat sink
all cadmium-soldered joints together to re-
duce generated thermal emf's. Careful
techniques will keep thermal emf's below
0.1 microvolt.
d. Use crimp connections with copper
wire and lugs for the best low-thermal
joints. Thermal emf's can be reduced to
10 nanovolts or less using the copper wire,
sleeves and Lugs found in the Model 1483
Low-Thermal Connection Kit. The Kit con-
tains a crimp tool, shielded cable, an as-
sortment of copper lugs, copper wire, cad-
mium solder and nylon bolts and nuts. It
is a complete kit for making very low ther-
I'LCUIW 5 . Model 1483 Low-Thermal Connecti mal measuring circuits. The Kit enables
K int . lieTcr to Section 6 for contents. the user 011 the Model 147 to maintain the
0466K
OPERATION MODEL 14,7 NULL DETECTOR
2-8. FMATING OPERATION.
a. The Model 147 can be connected between two potentials, neither of which is at power
line ground. It can be floated up to ?4,00 volts off ground.
CAUTION
The front panel controls are electrically connected to the case. If the
power cord is unplugged, the case may be at a voltage equal to the off-
ground voltage. use necessary
- e precautions.
b. For best results with floating operation, follow the steps below:
1. Remove the shorting link from the LO or GND Post on the rear panel.
2. Connect the input circuit to the Null Detector. Operate as described in paragraph
2-5. The zero suppress controls may also be used. Do not ground any recorders used
wfth this operation, since the low of the Model 14,7 output is no longer grounded.
3. If power line frequency pickup is a problem, battery operation usually provides
better results.
2-9. KECORDEROUTPUT. The Null Detector output for a full-scale meter deflection on any
range is $1 volt at 1 milliampere. Accuracy is ?I% of full scale. Output resistance is
less than 5 ohms within the amplifier pass band.. Output may be used during both ac and
battery operation. If the Model 147 is used for differential measurements, do not ground
the recorder connected to the output.
Model 147 Model 147 Model 147 Model 147
P V v b 4 P 4
LOW
FlGURF 6. Synchronized Buss System for Model 14,7. When two or m"re Null Detectors are
1
used in one system, an oscillator beat may occur; see paragraph 2-10. Synchronize the
instruments by connecting them as shown. See Figure 29 for point II.
12 04,67R
MODEL 1.47 NULL DETECTOR
2-10. USING MOIW THAN ONE MODEL 147 IN A SYSTEM.
a. The Model 1.47 oscillator is adjusted for a nominal freqnency ol 94 cps. ,,Oi.!i~VC~1~
slight variations in frcqucncy do occur between models. Wl,en using two or ,morc Xill I ill
tectors in one system, an oscillator beat may occur.
b. Synchronizing oscillators prevents an objectionable beat. Connect tllc! cwc in:,t iii-
"ents together at the collector of transistor (219 (Figure 29, Ipoint II), using iii: O.'i-
"icro.farad myl.ar capacitor.
c. At times the system is suc11 that the Null Detector lows may not bc conncctc~! di~-~~~t-
ly together. Then, use a 1:l transformer havi~ng a fairly higil impedance between tllc Cvil
instruments. A IlO-volt, low power isolation transformer works wit. USC n 0..5-:nicrfarad isolation capacitor in seri~es with iboth pri~mary and secondary wi~ndin~s o! C!IE i~~:i,i::-
former.
d. For several Null Detectors connected together, l,sc a synct,r"ni~zcil bliss :;ysli'T!, :iii
shown in Figure 6.
2-11. ACCURACY CONSIDERATIONS. For sensitive measurcmcnts - 10 mi~llivolts and bt~1lbi.x -
other considerations beside tile instrument affect accuracy. i:l:Ifcts not Iroticcnblc ,.i'!len
working with higher voltages are very important with mi~crovolt signals. Tl,c, ~lodel vi 7
reads only the signal received at its input; tllcrciorc, it is i~mportant tllat this signal
be properly transmitted from the source. The fol,i~owing paragraphs indi~cate iactor xiii c:.
affect accuracy: thermaI noise , input resistance, thermal emi's, shi~cldilll: iinconnections. 'Table 4 also offers a quick rcfcrence to correct troubles wllich may occ11r.
1. The thermal noi~sc in any ideal resistance can bc dctarmi~ned Srom tiic Joi~nson !~rlis~'
equation:
EZ",, = 4 I< T 11 I; 1:q I
where Erms is the rms noise voltage developed across the voltage source;
T is tile temperature in degrees Kelvin;
R is the source resistance in ohms;
F is the ampl~ificr bandwidth in cps;
k is the Boltzmann constant (1.38 x 10e23 joules / OK).
yor an ideal resistance at room temperature (3OO*K), equation 1 simplilics to
Ii,", = 1.29 x lo-lo (RF)1'2 xi,
2. Peak-to-peak meter indi.cations are of "ore interest than tlrc rms vaiuc. i:xpcri -
mentally, the peak-to-peak Johnson noi.se is about live times the rms value. At r<10111
temperature, equation 2 becomes
= 6.45 x 10-l' (RF)"' I:<,
EPP
where 13 is the peak-to-peak noise voltage developed across tile voltag" source.
PP
0477
Very slow response time
MODEL 147 NULL DETECTOR
2-13. INPUT RESISTANCE. The Model 147 is a feedback amplifier; its input resia:nnc<. is
obtained using high feedback factors. When the source resistance exceeds the amplifier's
physical input resistance - amplifier input resistance without feedback - the fecdbacl;
is partially destroyed. Then the instrument may not operate properly. Normally, do not
exceed the maximum source resistance listed in Table 2. Iligher resistances can be used,
but noise, offsets, slow response and instability may result. On the most sensitive
ranges, the maximum specified source resistance is consistent with Johnson noise considcr-
ations.
2-14. THERMAL EMF'S.
a. Thermal emf's (thermo-electric potentials) are generated by thermal gradients Ihe-
tween any two junctions of dissimilar metals. These can be large compared to the si!;nals
which the Model 147 can measure
b. Thermal emf's can cause the following problems:
1. Meter instability or zero offset much higher than expected. Note, thougll, the
Model 147 can have some offset (paragraph 2-5).
2. Meter is very sensitive to ambient temperature differences. T%is is seen by
touching the circuit, by putting a heat source near the circuit, or by a regular piliter,,
of instability, corresponding to heating and air conditioning systems or changes in sun-
light,
c. To minimize the drift caused by thermal emf's, use the same metal or metals having
the same thermo-electric powers in the input circuit. Gold, silver and low-thermal soldcr
have o thermo-electric power within about f0.25 ,rv/oC of copper. This means a temperature
inbalance of 1oC between these metals would generate a thermal emf of about 0.25 microvolt
At the other extreme, germanium has a thermoelectric power of about 320 ,iv/oC, and silicon
will develop about 420 jrv/% against copper.
Standard physical handbooks contain tables
of thermoelectric powers of materials.
Since the Model 147 input circuit is made
of pure copper, the best junction is copper
to copper. HOWeVer, copper oxide in the
junction will cause thermal emf's on the
order of 100 nanovolts per oC or less.
Also, differences in processing of two
pieces of copper can cause thermal emf's
of up to 0.2 microvolt per OC. The Model
1483 Kit contains all necessary equipment
to make very low thermal copper crimp joints.
See paragraph 2-6.
d. Besides using similar metals, thermal COPPER
"AI"ERI
emf's can be reduced by maintaining constant L Bw.II BOLT
temperatures, Keep all circuits from open
windows, fans, air conditioning vents and FIGURI? 7. Thermal Sink Construction.
similar sources which vary temperature. Connect leads or lugs as close as possible.
Minimize thermal gradients by placing all Separate only with insulation of high heat
junctions physically close on a large heat conductivity.
sink. Thoroughly clean all copper leads be-
fore making a connection. Crimp together
1S
OPERATION MODEL 147 NULL DETECTOR
the ends of each copper wire; bolt the lugs for each connection point together; mount all
stacks of lugs on a thick metal plate having high thermal conductivity. Thermal conduc-
tivity between the junctions and the heat sink can be kept at a high level by using mica
washers or high conductivity ceramics for electrical insulation.
e. Several other techniques will reduce the effects of thermal emf's. Use the zero
suppression circuit to buckout constant voltages. If connections must be soldered, use
only cadmium-tin low-thermal solder, such as supplied in the Model 1483 Kit. Unlike
metals - including regular solder - may be used and low thermal emf's obtained if a
well-controlled oil bath or a good heat sink is used. Thermal voltages may be calculated
from the thermoelectric power of the materials in the junction and the possible tempera-
ture difference between the junctions.
z-15. MAGNETIC SHIELDING.
a. In the Low resistance circuitry used with the Model 147, magnetic Lines cutting a
conductor can produce large signals compared to the instrument's sensitivity. The amount
of signal developed is proportional to the area enclosed by the circuit and the rate of
change of magnetic flux. For example, motion of a 3-inch diameter Loop in the earth's
magnetic field will induce a signal of several tenths of a microvolt. Increasing the
size of the Loop or moving it more rapidly will increase the signal. Magnetic fields
from ac power lines will cause even more difficulty.
b. To reduce the effect of magnetic fields, use magnetic shielding. Where high ac
magnetic fields are present, it may be necessary to magnetically shield the measuring
circuit , the unknown emf circuit or auxiliary equipment in the circuit. Magnetic shield-
ing is available from several companies in the form of plates, foil or cable.
c. Twist input leads to minimize the area enclosed by the circuit loop, Planning the
experimental Layout for minimum enclosed area is also of particular value.
Z-16. AC SHIELDING.
a. Due to its narrow bandwidth, the Model 147 is somewhat insensitive to ac voltages
superimposed upon a dc signal at the input terminals. However, ac voltages which are
large compared with the dc signal may drive the Model 147 ac amplifier into saturation,
erroneously producing a dc output at the demodulator. Usually it is sufficient to connect
the cases of all apparatus in the measurement circuit together and ground at one point.
This provides a "tree" configuration, which minimizes ground Loops. The common point at
which all shields are connected should be as near as possible to the circuit ground of the
Null Detector at its input.
b. Improper shielding can cause the Model 147 to react in one or more of the following
ways :
1. Needle jitter or instability, from 10% to 20% of full scale.
2. High offset (dc bias). Changing the power cord polarity or the connection between
the LO and GND Posts may affect the amount of offset.
3. Slow response time, sluggish action and/or inconsistent readings between ranges.
4. Amplifier saturation. Observe the wave form with an oscilloscope connected to the
DEMODULATOR TEST Jack (Figure 3). With the input shorted, it should approximate the
16 0466R
MODEL 147 NULL DETiXTOR o?'I:I?A'TION
----~
FIGURE 8. Normal Wave Form at Demodulator
with Input Shorted. Scale is 0.1 v/cm
vertical and 10 msec/cm horizontal.
wave form shown in Figure 8. If excessive pickup occurs, the wave Lorm will reseml,le
that of Figure 9. The circuit will operate reasonably well as long as the wave form is
not clipped, as shown in Figure 10. At this point a dc offset is introduced.
c. To minimize pickup, keep the circuit away from ac sources. Shield as carcfullp as
possible. Connect all shields together at the low side of the input or at the LO Terminal,
The voltage induced due to a magnetic flux is proportional to the arca of the loop. 'ThCSk-
fore, minimize loop arcas in the shield connections as well as the input circuitry. CO,,-
nect the shield at only one point. Run all wires in the circuit along the same path, so
the loop area is only the small difference in position of two adjacent wires.
d. Strong third harmonic magnetic fields - 180 cps for 60-cps units - may create an
8-cps beat at the Null Detector output and meter. To reduce this effect, turn off all
possible nearby sources, such as heavy-duty transformers. Remove the Node1 147 and the
measuring circuit as far as possible from the magnetic field. If removal does not greatly
reduce the beat, magnetic as well as electrostatic shielding around the circuit ~nlny be
necessary, For information concerning your particular shielding problem, contact Pcrl:cc-
tion Mica Corp., 1322 North Elston Street, Chicago, Illinois.
e. The 8-cps beat will be more apparent at the output terminals, since the meter is
filtered. To minimize the beat, USC the filter circuit shown in Figure 11. This divides
the Null Detector's l-volt output at Eull scale to 10 millivolts. The 8-cps beat is re-
duced by a factor of 1O:l. If the 330 microfarads is objectionably large, increase rhc
resistor sizes by 10 times and use a 33-microfarad capacitor. Since the recorder outpu;
0466R Ii
'OPERATION MODEL 147 NULL DETECTOR
is now only 10 millivolts, a non-polar capacitor is not necessary.
f. The Model 147 line frequency rejec-
tion refers to the total ac voltage appear-
ing at the input terminals, Therefore, the
Null Detector is affected by the sum of the
ripple in the working standard and the un- Model
known source. Because of this, working
147
standards having high ac ripple components
will significantly reduce the amount of ac
voltage which may be tolerated in the un-
known.
g. Shielding is preferable to input fil- FIGURE 11. 8-cps Filter Circuit for Re-
ters. Resistive-capacitive filters add corder Output. If~tha 8-cps beat disturbs
noise (equation 1), and the resistance value the measurements, this circuit will reduce
must be subtracted from the maximum source the beat 1O:l.
resistance in Table 2. Inductive-capacitive
filters rend to create loop instabilities
within the Null Detector. Capacity alone across the input, however, is less Likely to
cause Loop instabilities, and it may be used to filter ac components in some cases.
2-17. CIRCUIT CONNECTIONS.
a. When measuring in the microvolt and nanovolt regions, consider the effect the physi-
cal connections will have on the potential being measured. IR drops, which in most cir-
cuits are insignificant, now become important. For example, No. 20 AWG copper wire has
a resistance of approximately 10 milliohms per foot. A l-milliampere current through a
6-inch length of this wire will produce five microvolts. To reduce this drop to 0.5 nano-
volt would mean using a wire 0.0006 inch long.
b. Four-terminal connections can often
be used to eliminate this error. Refer
to Figure 12.
c. If an unwanted IR drop is constant,
the zero suppress may be used to nullify
3 Current
Leads
Model 147
ii
the voltage. Leads
t
d. If the currents or resistances in
the measuring system fluctuate, they will
develop fluctuating voltages which will FIGURE 12. Using Model 147 with 4-Terminal
appear as noise or drift in the system. Connections.
2-18. OPERATING FROM SOURCEOTHER THAN 117 VOLT, 60 CPS.
a. If the ac power source is 234 volts, use a screwdriver to change the Line Voltage
Switch on the back panel (Figure 3). Change the fuse from l/8 ampere to l/16 ampere.
Use only 250~volt MDL fuses. No other adjustment is necessary.
b. For 50-cps ac pave-s sources, change the sideband filter capacitors, CL03 and C104.
The Model 147 can operate satisfactorily from 60 or 50-cps sources, but the best ac re-
jection is achieved when the filter is set for the line frequency. Use Keithley part
MODEL 1.47 NULL DI:TL':CTOR
ClOS-.109M (C103) and C45-.0155M (C104) for 50 cps. Refer t" Figure 32 for componcation.
2-19. RACK MOUNTING. (See Figure 13.)
a. The Model 147 is shipped for bench use with four feet and a tilt-bail. 'i'hC :~i"dcl
4002 Rack Mounting Kit c"nverts the instrument t" rack mounting t" the standard i:I,\ (RL19-inch width.
b. To convert the Model 147, rem"vf the four screws at the bottom ot each sidii "1 t!~instrument case. Lift off the top cover assembly with the handles; save c11e four SCi-fi"S.
To rem"ve the feet and tilt bail from the bottom cover assembly, turn the two scr(ii.:s Il<'ilT
Item Keithlcy
(See Fig. 13) Description Part No. Qunntit!
1 cover Assembly 17162C I
2 Cover Assembly, Bottom (Supplied with
Model 147) 176950 1.
3 Angle, Rack 1462413 7
4 Screw, Slotted, lo-32 UNC-2x1/4 (Supplied ---
with Model 147) 4
5 Front Panel (Supplied with Model 147) --- 1
TABLE 5. Parts List for Model 4002 Rack Mounting Kit.
@SCREW
/
i(j)FR PANEL
IGURE 13. Exploded View for Rack Mounting, Using Model 4002 Rack Mounting Kit.
0466R 19
Ur'Diu~LUN MODEL 147 NULL DETECTOR
the back. The two pawl-type fasteners will release the cover and allow it to drop off.
Remove the feet and the tilt bail and replace the cover (2).
c. Attach the pair of rack angles (3) to the cabinet with the four screws (4) previously
removed. Insert the top cover assembly (1) in place and fasten to the chassis with the
two pawl-type fasteners at the rear. Store the top cover with handles, feet and tilt-
bail for future use.
20 0466R
MODEL 147 NULL DETECTOR APPLICATIONS
SECTION 3. APPLICATIONS
3-l. GENERAL. ThFs section contains dfagrams using the Model 147 with various potentio-
meters in null circuits, These are just samples of the circuits available and they do
not exhaust all the possible circuits. The setups which follow demonstrate how the Elodel
147 may be used.
3-2. WORKING SOURCES. The Model 147 permits resolution compatable with the smallest vol-
tage increment available on 6-dial potentiometers. When working at this resolutton, use
a stable battery working source with the potentiometer, Line-operated working sources
are generally limited by several inherent problems - instability, short-term noise of
several microwIts, pocr line regulation, and several millivolts of ripple, This hig!I
ripple may produce dc voltage due to a slight rectifying action at connection points and
switch contacts. Also, working sources having high ac ripple components will significant-
ly reduce the amount of ac voltage which may be tolerated in the unknown. This is be-
cause the sc rejection of the Model 147 refers to the total ac voltage appearing at its
input terminals, and, therefore, is affected by the sum of the ripple voltages in the work-
ing BOUWB and the unknown source.
NOTE
Follow the operating instructions in Section 2. Pay particular attention to the
p~oints brought up in paragraph 2-11 and following,
KElTHLEY .- MODEL 147
GUILDLINE 9144
POTENTIOMETER
FIGURE 14. Circuit Using Guildline 9144 with Keithley Model 147 Null Detector.
0466R 21
AH'LICATIONS MODEL 147 NULL DETECTOR
KEITHLEY MODEL 147
GUILDLINE 436301.
POTENTlOMETER
FIGURE 15. Circuit Using Guildline 4363DL with Keithley Model 147 Null Detector.
KEITHLEY MODEL 147
"NXHOYH
GUlL DL /NE 9120 EYF
POTENT/OMETER
FIGURE 16. Circuit Using Guildline 9120 with Keithley Model 147 Null Detector.
22 0466
MODEL 147 NULL DI:TIICTOR Al?PI.ICATLoNS
FIGURE 17. Circuit Using Riddle 605001 with Keithley Model 147 Null Detector.
KEITHLEY MODEL /47
I ?
STD.
CELL em
000 isl ,
r;;ir,
LttN 7556
POTENTlOMETER
FIGURE 18. Circuit Using Leeds & Northrup 7556 with Keithlcy Model 147 Null Ljctcctor.
1466 23
APPLICATIONS MODEL 147 NULL DETECTOR
KEITHLEY MODEL 147
r 1
ITl I
D.C. GENERATOR `8 6 6
999 DETECTOR BOOR
SD P
IO
1 KELVlN RATIO BRIDGE p 0 1
ES./. 240 80
f0
I PPP, *ID 0
II
q' ' RESISTANCE
STANDARD RS 925
24 0466
MODEL 147 NULL DETKTOR CIRCUIT DESCRIPTION
SECTION 4. CIRCUIT DESCRIPTION
4-1. GENERAL.
a. The Model 147 consists of a chopper, ac amplifier and demodulator systcim followed 1,::
a dc amplifier. Feedback is applied to the whole loop. (See Figure 20.)
b. A mechanical chopper converts the dc input signal to a 94-cps signal. 'The 312 signal
is amplified, demodulated, dc amplified and applied to the meter. A feedback network
samples the signal at the output and compares it to the input. The dc input signal and
the feedback signal are compared in the input transformer primary. 'The transformer in-
creases the voltage-difference signal between the two. The ac amplifier amplifies the
difference signal; line-frequency sidebands of the 94-cps signal arc filtered out. Thac signal is then demodulated by a saturated transistor switch and enters a dc amplifier,
which has a feedback capacitor to filter out the demodulator ripple. The dc amplifier
output is connected to the meter, the output terminals and the feedback network. Ihe
feedback resistors determine full-scale range. The zero suppress signal is conneclcd to
the feedback point in the input circuit.
FIGURE 20. Block Diagram of Model 147 Amplifier Circuits.
c. The power source for the Model 147 is either line voltage or the rechargeable
battery. Voltage from either source is applied to a dc-to-dc inverter and then to tnrec
highly regulated supplies. The three supplies furnish power to the oscillator and the
amplifier circuits. There is also a battery charging circuit to charge the battery when
it is necessary and when the line voltage is connected.
NOTE
Refer to Schematic Diagrams 18512F, 17352D and 173531) for circuit designations.
4-2. INPUT CIRCUIT.
a. The dc input signal is connected through the high terminal of the INPUT Receptacle,
JlOl, to the center contact of the chopper, GlOl. (See Figure 21.) The feedback signal
0466~ 25
CIRCUIT DESCRIPTION MODEL 147 NULL DETECTOR
r
.
* .
FIGURE 21. Model 147 Input Circuit. The dc input signal, Vin, is applied to the chopper.
The feedback signal, Vf (the dc amplifier output voltage, Vo, times the feedback ratio,
B )l is applied to the transformer primary. The signal, Vd, stepped up by the transformer
is the difference between the two, Vd = Vin - Vf. When the dc input signal is initially
applied to the Model 147, Vf is zero and the voltage across the primary is entirely Vin.
As the output voltage rises, Vf fncreases and Vd decreases to a small value, then Vf =
"in, Or p V, = Vin. Only p, which depends upon the RANGE and FUNCTION Switch settings,
determines the amplifier gain.
is applied to the center tap of the input transformer, TlOl. 'The chopper alternately
applies a positive and a negative square-wave signal across each half of the primary.
The magnitude of the square wave is proportional to the difference between the dc input
and the feedback signals. TlOl steps up this signal and applies it to the grid of tube Vl.
b . The input compartment is designed to insure high thermal stability and to minimize
internal ac pickup.
1.. Thermal stability is obtained in part by using only copper wire in the input
circuitry. The input transformer primary and the chopper leads are pure copper. The
input receptacle is 99.5% copper; the impurities add to the mechanical strength without
creating large thermal emf's. The low voltage portion of the FUNCTION Switch uses pure
copper pins and rotor. All connections to components are made with pure copper crimp
lugs. Connections between components are made by bolting the lugs together - not
soldering - to reduce thermal emf's.
2. The input compartment is doubly shielded against magnetic and electrostatic
pickup on all sides. The wires are physically placed to maintain minimum loop area,
further minimizing pickup.
c. The feedback network is formed from the output of the dc amplifier back to the
center tap of the primary of transformer TlOl. The RANGE Switch, SlO2, selects the
feedback ratio used for each range.
4-3. AC AMP