A.C. OPERATED METER FOR SMALL DIRECT CURRENTS A 13ie8is Presented to the Faculty of the Department of Physics University of Southern California In Partial Fulfillment of the Requirements for the Degree Master of Science "by Earle Clinton Enholm June 1940 UMI Number: EP63331 All rights reserved INFORM ATION TO ALL USERS The quality o f this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a com plete m anuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. Dissertation Publishing UMI EP63331 Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author. Microform Edition © ProQuest LLC. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code ProQ uest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 4 8 1 0 6 - 1346 +0 £ V e This thesis, written by EARIj:..CLIIITgiL..ElIH Cm ........................ under the direction of h .% 3 . F ac u lt y Committee, a n d a p p r o v e d by a l l its m e m b e r s , has been presented to and accepted by the Council on Graduate Study and Research in pa r t ia l f u l f i l l m e n t of the r e q u i r e m e n t s f o r the deg re e of MASTER OF SCIE2mtCE Dean Secretary JUHE,1940 F a c u lty C om m ittee C hairm an i. ? • TABLE OF CONTENTS CHAPTER PAGE INTRODUCTION............................ I. THE ELEMENTARY THEORY OF ELECTRONICAMPLIFIERS FOR D.C. CURRENTS ................... i . 1 Sensitivity and its limitations • • ........ 1 Effect of grid resistance ................ 2 Current Amplification • • • .............. 3 Meter sensitivity ........................ 3 Random fluctuations ...................... 4 Choice of tubes ............................ 5 II. DESIGN OF PRESENT AMPLIFIER............ 7 III. ADJUSTMENT AND BALANCING OF AMPLIFIER........ 14 IV. SENSITIVITY.................................. 18 V. DIRECTIONS FOR U S E ........................... 20 Adjustment of Potentiometer ........ Directions for removing chassisfrom cabinet Summary ............. 22 23 23 BIBLIOGRAPHY...................................... 25 APPENDIX 26 ................... List of Parts ........................... 26 LIST OF FIGURES FIGURE PAGE 1 . Simple Electronic Amplifier............ • • • • 1 2* Simplified First Stage Circuit • • • • • • • • 7 3. Circuit to show ^-balancing • • • • • ........ 9 4* Simplified Circuit of Two-stage Amplifier • • • • 9 5* Circuit of Amplifier, complete • • • • • • • • • 11 6 * View showing Interior of Amplifier • • • • • • • 15 7* Wheatsone’s Bridge .................. • • • • • • 1 8 8 * Operating Panel of Amplifier • • • • • • • • • • 20 i IHTRODUCTIOH The problem of measuring very small currents and volt ages originating in high resistance sources has always been & troublesome one. Even when galvanometers of suitable sen sitivity are available, their efficient use is often preclu ded because of the condition that for maximum sensitivity the internal resistance of the galvanometer must be of the same order of magnitude as that of the external circuit.^ If the external resistance is of the order of several megohms, this requirement Is impossible for any galvanometer now available. These small currents may be measured by passing them through a high resistance and observing the potential dif ference developed across the resistor by means of an elec trometer. This method is inconvenient and cumbersome. The most versatile and convenient means for this type of measurement is provided by the thermionic valve or vacuum tube. This device has a very high input resistance, so'that no appreciable power is drawn from the source, and a moder ately low, easily matched output impedance. If the vacuum tube served merely as an impedance matching device between source and galvanometer it would be exceedingly useful, but in addition large amplification may be provided, extending the i*.— — ... ---------- _ H.B. Brooks: Sensitivity of a galvanometer as a function of its resistance. Bur. Standards Jour. Research 4: 297 (1930) ii possibility oi measurement to ranges far below that provided by the galvanometer alone* The vacum tube when used to amplify small d*c. currents i is subject to many abberations. This has made necessary the development of special tubes and circuits for the compensa tion and elimination of such abberations. By the use of these circuits^ and by observing elaborate precautions cur rents as small as 10" * amperes, or a flow of electrons per second h^ve been measured. La;rge storage batteries and a sensitive galvanometer are required for these ultra sensitive circuits; but if, on the other hand, moderate sensitivities — to 10**12 amperes — are stifficlent, an a.c. operated de vice, using a rugged milliamme,ter, is possible. The con struction of such an instrument is the subject of this thesis. In common with other recent deslgns,^*^ the Instrument to be described has several advantages, namely: it Is port able, readings are easily and quickly made, and ordinary re ceiving £ubes are used, rather than the expensive electro meter tubes. In addition, it Is a.c. operated, and uses a rugged milliammeter Instead of a sensitive galvanometer. £ Penick: Direct Current Amplifier Circuits for use with the Electrometer Tube ( a bibliography is included) Rev. Sci.Inst. 8 , 196-198 (1937) ® Roberts : A feedback Micro-micro Ammeter. R.S.I. 10:181-183 (1939) ^ Gabus and Pool: A portable Photo-tube unit using an RCA 964 tube. R.S.I. 8:196-198 (1937$ • * * 111 The features which are peculiar to this design are as follows: the input resistance ia 20 megohms, thereby avoid! ing the use of the somewhat unreliable resistors with values of thousands of megohms used in other designs; unusually high voltage sensitivity, making possible the use of the instru ment as a high resistance millivoltmeter; and a built-in potentiometer for on-the-spot- calibration -- eliminating the need for calibration curves and the need for recalibrating as the tubes change their characteristics. The instrument was designed for use as a d.c. millivolt meter of 20 megohms input resistance, or as a device to measure small currents down to 10~^1 amperes, originating in high resistance sources. It is useful in measuring photo electric, leakage, or ionization currents. will suggest themselves. Many other uses 1 CHAPTER I THE ELEMENTARY THEORY OF ELECTRONIC AMPLIFIERS FOR D.C. CURRENTS The simplest form of such an amplifier consists essentially of a high resistance, Rg, through which the cur rent to he measured is passed, followed by a vacuum tube connected as shown in Fig. 1, with a suitable meter M in the plate circuit. ^ ^ f Xnfwt| The voltage drop produced by the current i ^ ^ n zizer» ^ passing through the input resistor Rg is impressed upon the grid of the vacuum tube, and the corresponding change i. in plate current noted. If the amplifier has been calibrated, the current in Rg, or the voltage acrossi£t, is known. As the circuit stands, all of the plate current passes through the meter, necessitating the use of one which is relatively insensitive* but by the use of another battery and a simple circuit, the steady plate current can be bucked out, so that only the changes in plate current are indicated by the meter. This permits a more sensitive meter to be used. SENSITIVITY AND ITS LIMITATIONS The current sensitivity of such a circuit depends upon three factors, each of which is, however, dependent upon limited by many other factors. and These are the grid or input resistance Rg* the current amplification within the tube* and the sensitivity of the meter in the plate circuit. Effect of Rg. That the sensitivity is directly proportional to the magnitude of Rg is obvious, for if this resistance be increased tenfold, the current to be measured may be decreased tenfold without changing the IR drop through the resistor. The voltage on the grid of the tube, and therefore the reading on the plate meter are the same as before. In a current measuring device, therefore Rg is made as large as possible. But the maximum value of Rg depends upon the insulation of the tube itself and upon the amount of the grid current. The tube resistance acts as a shunt across the input resistance Rg , so that Rg must ordinarily be small compared to the resistance of the tube itself. Grid current also tends to lower the input resistance of the tube itself. It must be small enough so that, passing through the grid re sistor with the current to be detected, its effects will not mask those of the applied current. Special tubes, such as the General Electric S’.P. - 54 and the Western Electric D-96475 tubes, are available having quartz insulation of the tube elements, and also designed so as to have very small grid current: Rg from 10^ to 1014 ohms. these may use values of But for receiving tubes operated at ordinary voltages, 10 to 20 megohms is the limit. current will be discussed more fully later. Grid 3 Current Amplification* The current amplification of a tube depends upon the grid resistance Rg and the mutual conductance (gm) defined by the relation ip « plate current gm = dlp/deg eg = grid voltage. But i(j = Galvanometer current k = Galvanometer constant d Galvanometer deflect ion of meter i * input current The current amplification A is 1q /1 » so that The sensitivity S = d/i * SmRg/k mny^amp The last relation restates the facts mentioned above, that the sensitivity depends upon the input resistance, the am plification provided by the tube, and sensitivity of the gal vanometer* The current which will provide a deflection on one mil limeter is then given by the relation l/S = k/gmRg For an :FP-54 tube, gm s 25 micromhos, and if Rg * 109 ohms, and k = 10"® amp/mm, a current of 4x10"^ amperes will produce a deflection of 1 mm* Meter Sensitivity* Both the meter sensitivity and the amplification within the tube are limited, for if either is in creased beyond a certain limit, it will be found that the in 4 strument becomes subject to random fluctuations which in terfere with the usefulness of the device. If the am plification is small, the fluctuations will be correspond ingly small, and a sensitive meter may be used* but if the amplification of a given circuit is increased, the fluc tuations increase correspondingly, and a less sensitive meter, must be used. Random Fluctuations. Since the occurrence of random fluctuations places a limit on the sensitivity of the am plifier, it is necessary to mention the causes of these fluctuations and to indicate methods for their reduction. These causes are as follows: 1. Lack of adequate shielding. 2. Fluctuation of supply voltages. 3. Microphonics within the tube. 4. Ionization within the tube. 5. Secondary emission. 6. Uneven filament emission. A great deal of work has been done along these lines. Input circuits and tubes are carefully shielded, using iron if there are strong magnetic fields. Fluctuations in volt age supply are minimized by using large new storage batter ies or by devising special circuits which compensate for voltage changes2). Tube microphonics are reduced through proper tube design and by cushioning the tube sockets. Ionization within the tube may be reduced by using a highly evacuated tube at low voltages - below the ionization po tential if possible. The effects of secondary emission, if iroltages high enough to produce it areused, are reduced by means ofa 5)ecial grid ( #3 in a pentode tube.) Uneven fil ament emission is minimized by using low temperatures on the filament, stable current supply, by use of filament material of maximum evenness of emission, and by special cir cuits. Choice of Tubes The general Electric F3? 54 Electro meter tube, which has been mentioned before, was especially designed with these disturbing elements in mind* it is quartz insulated, and since the plate voltage is 12 volts, the grid current ( due to ionization in the tube) is ex tremely low. With a sensitive galvanometer to measure the changes in plate current, currents down to 10 " ^ amperes can be easily measured, and with great precautions and special —18 circuits, measurements down to 10 amperes, may be made. Where only moderate sensitivity isrequired, an a.c. operated device using the less expensive receiving type tubes and a rugged meter is more convenient. Several types of these instruments have been made, and are mentioned in the references. In selecting a receiving type of tube for this purpose, one having the grid connection to the top of the tube rather 6 than to the base is preferable because of the better insu lation of the grid* If the tubes are to be a.c. operated, we have the choice of the following triple grid tubes: 954 , 57, 6C 6 , and 1603. The 954 is of the so-called acorn type. No tests were made on this tube, but its physical and electrical character istics certainly warrant an investigation as to its availa bility for this class of service. The 57 and 6C6 are identical except for the filament, the former operating on 2.5 volts, and the latter on 6.3 volts. They vary greatly among themselves, and should be carefully chosen. The 1603 is identical to the 6C 6 electrically, but is of low noise design for use in critical applications. It is pro bable that this is the logical tube for the amplifier to be described, but it was not tested because of considerations of cost. It may be placed in the amplifier at any future time * without making any changes other than normal readjustment. If the decrease in "noise11 should warrant an increase in sensiti vity, this may be easily accomplished by making a few changes which will be described later. The 6C6 was chosen on the basis of cost over the 954 and the 1603, and over the 57 because, having been developed for automobile service, it was considered probable that the cathode would be more sturdy and less subject to microphonics. CHAPTER 11 DESIGN OP PRESENT AMPLIFIER The amplifier whose construction and operation are the subject of this thesis has two stages of amplification. The first stage, which is similar to an amplifier devised by Harnwell and Van Voorhis,^ is stabilized in two ways, which will be considered separately. As will be seen by the simplified diagram of the first stage (Pig* 2) , in which grids #2 and #3, as well as tfee *x balancing arrangement, are omitted for clarity, the circuit is essentially a Wheatstone bridge in which two of the resist ors have been replaced with vacuum tubes. As with the ordin ary Wheatstone bridge the various constants may be adjusted so that the galvanometer M will show no current, and the cir cuit is said to be in balance. If now a voltage is applied to the terminals so that the grid of the upper tube is made, say, more positive, the conduc tivity of the tube will there fore be increased and the current 5 through the upper portion of the network will be increased. Fit L Be cause of the greater current Showing first stage cir carried by R^ the potential of cuit without /Lt-balance. S G. P. Harnwell and S.N. Van Voorhis, Balanced Electro meter Tube and Amplifier circuits for small Direct Currents: Rev. Sci. Inst: 5:244 (1934) 8 of point A will decrease and the circuit will he unbalanced* The amont of unbalance as indicated by the galvanometer is a measure of the applied voltage* The advantage of using such a circuit lies in wellknown fact that the itfheatstone bridge is unaffected by variations of supply voltage, for if the voltage EB is in creased, the voltage drop through and R2 will change in a balanced bridge, so that the potential of point A will still equal that of Point B. The change of supply voltage is therefore not indicated by the galvanometer. If the two tubes were actually identical no further balancing other than that inherent in the bridge circuitw ould be required* But while the tubes were matched as nearly as possible at the time of purchase, small differences in mutual conduct ance and in cathode heating rate prevent the two tubes from behaving in an identical manner when the supply voltages are changed: perfect balance is therefore not achieved, and an additional method of balancing the residual fluctuations is required. The method developed by Turner and called by him \i bal ancing is also incorporated in the first stage* readily understood by reference to Pig. 3 It may be Suppose the plate voltage Ep to increase slightly due to an increase in line voltage. The increased voltage would increase the plate current if no other factors are considered. But the current through ri and T£ would also increase, raising the potential of the cathode. As the grid is now more negative to the cat hode, the current would tend to de crease. And if ijl is the amplification -W W W factor of the tube, and if r^ and T2 are chosen so that there will be no Fig 3 change in plate current with change in Ep. A single stage amplifier embodying these two methods of stabilization is very stable, and may be used with excell ent results using a micro-ammeter or a wall galvanometer. In order that a rugged mi H i ammeter might be used without loss in sensitivity, a second stage was added. This second stage utilizes a type 53 tube, which contains two triodes in one envelope. These are used to make a bridge circuit similar to that of the first stage, the grids being connected to what would be the galvanometer binding posts of (ijl balancing, grids §2 and 3, and volt age supplies omitted.) of the first stage. The millammeter is connected into the 10 second stage in the usual manner. "both stages is shown in Fig. 4. A simplified diagram of As the current am plification of the second stage is small* — about 60, no stablization other than that inherent in the bridge circuit is used. The entire circuit is shown in Figure 5. It will be noted that this includes a built-in potentiometer for use in calibrating the instrument. The current for this potentio meter is supplied by a dry cell, and since the current drain is slightly more than one milliampere, the dry cell provides good r egula,tion and long life. This current is adjusted by means of P-l, and checked for proper value by a shunted 0-1 milliammeter. The voltage in millivolts is read directly on a dial mounted on P-2, a 600 ohm, potentiometer. The other circuit constants for this portion of the circuit are not given, as they must be determined experimentally in each case. This is because the dial does not cover the en tire winding of P-2, and because individual potentiometers willin vary somewhat from their nominal values. It should be mentioned here that the quality of parts used in the amplifier and potentiometer is c ritical. The parts P-2 and P-3 are General type 314-A potentiometers, having multiple contacts. were of the same type. Improvement might result if P-4 All other rheostats except P-6 are of pot : a d j . -6 874 S H I E L D ^ INPUT •VvwwVt\r VT-3 R-3 ■WWvvV S\ /-3 SHIELD INPUT SWITCH SW-2 SENSITIVITY P I L O T LIGHT \ © - R r Z 2 .5 V 6C6 VT-2 603 CHOKE E^6.3V VT-5 © R-9 V T -4 FH5-5 SW.4 TESTING SWITCH ^ GROUND TO CHASSIS ® CLOSED CIRCUIT JACK 12 General Radio make. Resistors should he wired wound. The writer had set this circuit up over a year previously and had abandoned it because of trouble which was later traced to faulty action of smaller radio-type volume controls for rheostat and potentiometer use. The input to the amplifier leads to a three-position input switch (Sw-2) by means of which the grid of the input tube VT-1 may be grounded, connected to the input, or to the calibrating potentiometer. The grid resistors R 5 are 20 megohms, and are shielded by means of heavy brass cylinders placed immediately against the tube shields for VT-1 and VT-2. All the input wiring is shielded, using aviation ignition wire, although the recently introduced Amphenol co-axial cable would provide improvement. The operation of the amplifier is obvious from an ex amination of Rig. 5. The constants of the tubes of the first stage are individually variable through the use of sep screen supplies (from P-5 and P-6 ) while the voltage is held constant across the potentiometers by means of a voltage re gulating tube, UX874. The ji-balance for VT-1 is provided by R-7, P-3, and P-4 as fasr as the slider, corresponding to rg of Rig. r^. The 3, ijl and the remainder of P-4 and R-5 corresponding to balance for VT-2 is provided by the same resistor network except that P-4 is replaced by two fixed resistors. The sensitivity may be reduced by shunting resistors 13 across the plates of the first stage tubes hy means of Sw-3. Jacks are provided hy means of which the plate current of each of the first stage tubes and one of the second stage triodes may be read, as well as a jack to allow the use of the first stage independently of the second. The type 874 tube not only maintains a constant volt age across P-5 and P-6 but provides a low resistance path through thispart of the circuit for the cathode current of they type 53 tube. The otherwise high resistance would cause excessive degeneration or “inverse feedback”. Not shown in the circuit are two resistors of two ohms each in each side of 6.3 volt filament supply, thus reducing the filament voltage to approximately 4.5 volts. This has the effect of reducing the temperature and thereby making filament emission more uniform and also increases the input resistance of the tube. The two Wheatstone bridge circuits are balanced by P-7 and P-8 . These controls are marked “Zero Adjust, Pine" and “Zero Adjust, coarse” respectively on the outside of the case. Other controls needed far routine use are P-l and P-2, and all switches except Sw-4. side. These are controlled from the out All others are located inside the case and are handled except for every frequent readjustment. 14 CHAPTER III ADJUSTMENT AND BALANCING OP AMPLIPIER Note: The adjustment procedure to he described should not be necessary unless the tubes have been changed or the controls inside the cabinet tampered with. Adjustment is tedious, and care should be taken not to remove the tubes nor to change the setting of the internal controls, as this will make adjustment necessary. In adjusting the amplifier, it is important to remember that the grid bias on VT-4 is dependent not only oh the ap plied voltages, but also upon the plate current of VT-1 and VT-2. The grid potential is equal to the applied voltage taken off R-7 diminished by the IR in P -8 and R-8 or R-9. The cathode of VT-4 is connected to the high side of VT-3, and its potential is determined by the setting of P-3. The adjustment is complicated by the fact that not only must \xbalance be obtained for VT-1 and VT-2, but that their plate currents must be adjusted so that the grid of VT-4 will be negative by the proper amount with respect to its cathode, all with the same stet of controls. However, once balance has been obtained, further adjustments will not be necessary until tubes are changed or they change their characteristics. As the tubes are operated under low voltages, they should last almost indefinitely. 15 To make the adjustment, VT-1 and VT-4 are first removed. By means of a cord fitted with phone plugs at both aids, the plate current jack of VT-2 is connected to the jack mounted on the case of the meter M-l, inside the cabinet* taking this connection automatically disconnects M-l from its reg ular circuit.. By means of P - 2 and P-6, VT-2 is adjusted to ^-“balance, the plate current of VT-2 to 0.53 milliamperes* This is done by setting these controls so that when the test ing switch, Sw-4, current. is operated, there is no change in plate VT-1 is then put in place and similarly adjusted using P-4, P-5, and to some extent for final balancing and matching with VT-2, P-8. Placing VT-1 in the circuit will somewhat upset the operation of VT-2, sd that final balancing must be done with both tubes in place.. Pig. 6 16 VT-4 is then placed in the circuit. If the plate current of the preceding tubes is correct, the tube will draw from three to five milliamperes per plate. The jack which is connected with one of the triodes of Vt-4 is fitted with a shunt so that it may be plugged directly into the jack on M-l. Any reading on the upper half of M-l indicates a satisfactory plate current. If the meter reads too low, it is an indication that the first stage tubes are drawing too much current* if the reading is too high, they are not drawing enough. The first stage must then be readjusted. It is often helpful to have VT-4 in place and plugged into a separate 0-1 milliammeter during the preliminary adjust ment, of the first stage. On operating the testing switch, it will be seen that the imperceptible movement of the meter when plugged into the first stage is now considerably magnified. Adjustments are then made to reduce these deflections to zero. Finally, the testing cord is removed entirely, and adjustments con tinued for minimum movement of M-l when the testing switch is changed in position. This procedure eliminates instability due to changes in plate or screen voltages. Changes in filament eurrent are partly compensated by the bridge arrangement and by the fact that the tubes are operated at reduced voltages, but a sharp change in line voltage will produce some drift due to the 17 fact that the heating rate of the first stage tubes are not the same. This drift could be reduced, if thought necessary, either by using storage batteries or a saturated core trans former as a source of filament current. The small residual fluctuations are due to “tube noise,1' and will not interfere with the use of the instrument. It is probable that the use of 1603 type tubes in place of the 6 C6 would result in less of this type of fluctuation, thus increasing the useful sensitivity, or enabling the use of the present sensitivity with "quieter" action. 18 CHAPTER IV SENSITIVITY Because of small fluctuations which cannot be balanced out the amplifier sensitivity was purposely limited. The voltage gain of the first stage is limited by the fact that R-8 and R-9 are set at 100,000 ohms. If this value were raised to 250,000 ohms the sensitivity would be increased 3.3 times. Nor are R-13 and R-14 the optimum values to de liver maximum power from VT-4 to M-2. For maximum power into the meter, the following re lation should exist:® 1 Fig 7 Rgi = _ 1_ + 1 RI+R3 RS + R4 Rq2 is fixed, and is approximately 33 ohms. R]_ and R3 are the tube resistances, and are 22,000 ohms each. Rg and R4 are the resistances in the plate circuit, and are equal. Solving for Rg and R^ we find 17 ohms the optimum value for these resistances. The voltage sensitivity of the amplifier as built is 2/3 millivolt per division of M-l. The current sensitivity is, since 20 megohm input resistors are used: 2/3 x 10-3 x 1/2 x 10“7 = .33 x 10“10 or 3.3 x 10”11. amperes/division. The voltage range of the instrument is limited to a ® Page and Adams: Principles of Electricity: p 180 19 maximum of 100 millivolts. This is due to the fact that an input voltage ts much greater than this, after amplifi cation by the first stage, is sufficient to cause the tube elements of ttee VT-4 to block. The current sensitivity is therefore limited to the range from 5 x 10“^ amperes as the maximum to a minimum detectable of lO*1^ amperes. 20 CHAPTER V DIRECTIONS EOB USE CAUTION: DO NOT TAMPER WITH ANY CONTROLS INSIDE THE CABINET. M - L L> Pig 8 1* Turn the sensitivity control to the left. 2* Turn on the power (extreme lower left switch). Due to the fact that the tube heaters warm up at different rates, the amplifier meter M-l (on the left) will fluctuate widely. Keep the needle on scale as nearly as possible by use of the two adjustment controls. The "Pine” adjustment will have the most effect in this case. The large fluctuations of the meter will soon stop, leaving a fairly steady drift which will stop within a few minutes, when temperature equilibrium is established. The 21 residual small fluctuations are due to tube HnoiseM and will not interfere with the use of the instrument. 3. Connect the source of voltage or current to he measured to the terminals on the right side of the case. The leads should he well shielded, and of very high resistance between the grounded shield and the conductor. Amphenol coaxial cable is recommended for uhis purpose. Connections should be made so that the positive lead ^oiee to the porcelain bushing. ed. The binding post is ground With the connections made, but with the E.M.P. on, and adjust the sensitivity so that the meter will give as large a deflection as possible, but remain on scale. Chang ing the sensitivity will require a resetting of the zero point. 4. With the sensitivity properly set, and the zero set with input circuit in place, turn on the unknown E.M.P., and see the reading of the amplifier meter. Turn on the E.M.P. and note whether the meter again reads zero. If not, readjust and take another reading. 5. Turn on the potentiometer (Sw-l) which has been previous ly adjusted as described below. Turn the input switch to "Potentiometer" and adjust the amplifier to zero, being sure that the potentiometer is set for zero. Then turn the pot entiometer dial until the amplifier meter gives the same reading as that produced by the unknown E. M. P. The dial then 22 given the same reading as that produced by the unknown E. M.P. measured across the 20 megohm input resistance in millivolts. The current may he calculated by Ohm’s law. Adjustment of Potentiometer 1. Note zero reading of the potentiometer meter M-2 (on right). This reading is not the zero of the meter, but the first small division past the zero mark. This may be adjusted with a smaL1 acrewdriver if necessary. 2. Turn on the potentiometer. As the winding on the potent iometer is not quite uniform on the lower end, it is nec essary to have slightly different currents passing through the windings for different regions on the dial. These are as follows15 Dial Reading, Mv Pot. Meter Reading 0-5 Special Mark off Scale 6-29 .99 ma 30-100 .98 ma The readings from 6-100 mv are always within two per cent, and in most cases well under one percent. accuracy is less. Prom 0-5 mv., the The settings are as follows: Dial Reading 1 2 3 4 5 Actual millivolts .82 2.8 1.7 4.13 4.8 Ajust the current through the potentiometer according to the 23 table above as soon as the approximate range is determined, using the HPot. Adj.) control, which is P-l. Directions for removing the chassis from the cabinet* In case service work should be found necessary, the chassis should be removed according to the following directions 3 1* Remove right hand screw from bakelite jack cover on rear of chassis, and loosen left hand screw. 2. Remove crackle finish screws froirv top and sides. Do not loosen nickle plated screws along the bottom. 3. Unsolder ground and input connections to case at in switch. 4. Tip case on left end. Slide right end of chassis forward slightly, so that the chassis will clear the flange on the case. 5. Watch out for jack strip and cover. Set case flat on table. With a large screwdriver the case may be sprung sufficiently to draw out the chassis. Summary. In use, then, the amplifier is allowed to warm up for a few minutes, after which the well-shielded Input con nections are made, and the m plifier zero is adjusted. The current or EMF to be measured is turned on and the meter read ing noted. The input switch is then switched to "Potentio meter" and the dial turned until the meter gives the same reading as before. The dial then indicates directly the mill ivolts across the input resistance. The current is obtained by dividing this value 2 x 10 7, the value of the input 24 resistance. For very small currents, the value is 3 x lO**^1 amp per div. of M-l. Readings are quickly and easily made, and, because of the potentiometer feature, calibration charts and curves are unnecessary. The instrument is reliable and, as long as the internal controls are not tampered with, as fo&l-proof as an instrument of its sensitivity could be. The current range is from 3 x lCT^ amperes to 3 x 10”*^ amperes, although smaller currents can be detected. voltage range is from 0-100 millivolts. The While normally volt ages will be indicated directly on the potentiometer, it is well to note that a half division on M-l indicates 1/3 milli volt. BIBLIOGRAPHY 25 SUMMARY OF REFERENCES 1* Brooks, H.B., “Sensitivity of a galvanometer as a function of its resistance'*. Bur. Standards Jour. Research 4 :297 2 Penick: “Direct Current Amplifier Circuits for use with the Electrometer Tube. Rev. Sci. Inst. 8:196-198 (1937) This reference provides a summary of various circuits devised for use with the electrometer tube. An extensive bibliography is provided. 3 Roberts, Shepard, "A Feedback Micromicroammeter" Rev. Sci. Inst. 10: 181-183 (1939) An a.c. operated device of excellent stability, providing a sensitivity of 10-12 amperes. Requires large grid resistors. 4. Gabus, G.H., and Poole: "A Portable Phototube Unit using an RCA 954 Tube” Rev. Sci. Inst.8 : 196-198 (1937) A battery operated outfit using a grid resistor of 1011 ohms and providing a sensitivity of l(P-3 amperes. Quickly and easily set up, especially suited for photocell work. 5* G.P* Harnwell and S.N. Van Voorhis, “Balanced Electro meter Tube and Amplifier Circuits for small Direct Currei ts” Rev. Sci. Inst. 5:244 (1934) 6 . Page, Leigh and Norman I. Adams, Principles of Electricity, p 180. Van Nostrand Company, 1931. APPENDIX 26 LIST OP PARTS P-1,3 4, General Radio potentiometer type 214-A 100 ohms P-2 type 314-A 600 ohms P-3 type 314-A 2000 ohms p-5,: ?-6 , type 301-A 20,000 ohms P-7 type 214-A P -8 International Resistance Co potentiometer type CS 200 ohms 10, 000 ohms R-5 Wirewound resistor, 50 ohms, 1 watt. R-6 Carbon resistor, 20 megohms, two required. R-7 Wirewound adjustable resistor, 10,000 ohms, 50 watt. R-8 , R-9, Carbon resistors, 0.1 megohm, 1 watt (wirewound resistors are to be preferred) R-10 Carbon resistor, 0*25 megohm, 1/2 watt, for sensitivity control. R-ll it R-13 R-14, R-15 Wirewound resistor, 200 ohms, 10 watts. ii 0 .1 0 M M M n H Wirewound resistor, 1000 ohm, 1 watt. Two wirewound resistorsm 50 ohms, 1 watt, as shown in Pig. 5. Two wirewound resistors, 2 ohms, 2 watts for insertion in filament supply of VT-1 and VT-2. See text. Sw. Cutler Hammer rotary line switch. Sw-1 Cutler Hammer rotary switch, s.p.s.t. Sw-2 Sw-3. Centralab Isolantite insulated Transmitter Switch, single deck, #2542 Sw-4 Yaxley four pole double throw rotary switch, contacts wired in parallel for s.p.d.t. C-l, !-2. Electrolytic condensers, 8 mfd, 450 volts. 27 Parts list, cont* Choke, Filter, 12 henry, 60 m.a. Transformer, Powere, with 2.5 and 6.3 filament secondaries. Dry Cell, Burgess Little Six, 4FH Closed circuit jacks, as indicated on Fig. 5. Two circuit jack, as shown in Fig. 5 Miscellaneous hardware, wire, and cabinet. Tube complement: 2 type 1603 or 6C6--VT-1 and VT-2 1 type 874 1 type 53 VT-4 1 type 80 VT-5 --VT-3 Three required.