THE CONSTRUCTEON AND TEE-HMS OF A 5UPERCONDUC?ENG CHOPPER AMPUFNER T‘hssis 9:» ma Degree c? M. S. MKWGAN STATE UNWERSE‘E’Y Wéfiiiam A. Hum W62 IIIIIIIIIIIIIIIIIIIIIIIIIIIIII 01687 5894 L I B R A R Y Michigan Sta tc Univcmity THE CONSTRUCTION AND TESTING OF A SUPERCONDUCTING CHOPPER AMPLIFIER BY William A. Hunt AN ABS TRACT Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE . Department of-Physics ABSTRACT THE CONSTRUCTION AND TESTING OF A SUPERCONDUCTING CHOPPER AMPLIFIER by William A, Hunt The theory of operation and the experimental results obtained from several devices utilizing superconductivity Which are used to detect and record voltages of the order of 10-—8 to 10-12 volt are reviewed. Details of construction and principles of operation of a superconducting chopper amplifier are discussed in detail. and the advantages are presented of using a super- conducting solenoid in the apparatus instead of a solenoid wound from copper wire, which was employed in a similar device built by De Vroomen and van Baarle. The procedure is given for testing the device both at room temperature and at liquid helium temperature. Even though the sensitivity of the device was limited by instability in the frequency selective amplifier. it was capable of detecting 0.09 microvolt. THE CONSTRUCTION AND TESTING OF A SUPERCONDUCTING CHDPPER.AMPLIFIER BY J William A: Hunt A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Physics 1962 '.. gar.“ ,7 I ///2/n"’ , ‘ . ..' '...’ ‘6'- ACKNOWLEDGMENTS I wish to express my sincere appreciation to Drs. F. J. Blatt and M. Garber for suggesting the problem and to Dr. Blatt for his continued interest in bringing the research to a conclusion. I am grateful to B. LaRoy for his assistance in solving many of the electronics problems. Financial support by the Office of Ordnance Research under contract DA-ZO-OlB-ORD—l6883 is gratefully acknowledged. ********** ii TABLE OF CONTENTS Page I. INTRODUCTION . . . . . . . . . . . . . . . . . 1 II. DESCRIPTION AND THEORY . . . . . . . . . . . . ll Principles of Operation . . . . . . . . . . 11 Low Temperature Parts . . . . . . . . . . . 12 Electronic Components . . . . . . . . . . . 20 III. OPERATION . . . . . . . . . . . . . . . . . . . 26 IV. APPENDICES . . . . . . . . . . . . . . . . . . 33 V. BIBLIOGRAPHY . . . . . . . . . . . . . . . . . 36 iii LIST OF FIGURES Figure l. The superconducting modulator . . . . . . . . 2. Templeton's superconducting reversing switch . 3. Comparison of DeVroomen's and TEmpleton's methods of modulating the superconducting Wire 0 O O O O O 0 O O O O O O O O O O O O 4. Schematic of apparatus as used for the detection of thermoelectric power . . . . 5. Comparison of the critical fields of tantalum. lead. and niobium . . . . . . . 6. Low temperature parts mounted on plastic . . . 7. Circuit for generating the alternating solenoid current showing the current waveforms at various points . . . . . . . 8. Frequency selective amplifier . . . . . . . 9. Phase sensitive detector . . . . . . . . . . . 10. Low temperature circuit used to test the superconducting chopper amplifier . . . . ll. Curve of voltage vs. current used to determine the calibration resistance . . . iv I. INTRODUCTION There have been many advances in the recent past in the theoretical and experimental studies of the trans- port properties of metals at low temperatures. The physical processes causing many of the phenomena are thought to be understood and there is agreement between experimental results and the derived theoretical models. Hewever. there are still areas where there is little or no agreement between the two. The remaining disagreements between the theoretical and experimental results may be due to an inaccurate mathematical model of the physical process under consideration or a model demanding mathematical techniques beyond present day capabilities, or they may be traced to poor or incorrect data. The measurement of thermoelectric data and other sources of small static voltages is one of the important tools by which information on electrons and phonons and their interactions with each other is obtained. These measurements are difficult to obtain for two reasons: crucial measurements frequently must be made at low temperatures (from about 1° K to 4° K) so the lattice vibrations are reduced to a minimum; and the effects measured 8 to 10-12 volt). Therefore, are small (of the order of 10- detection devices are needed which are sensitive enough to obtain measurements of this order. Another requirement of the devices is that they must be small enough to fit in the cryostat used or provide for the elimination of thermal voltages generated in the leads from the cryostat. These thermal voltages are of the order of 10'.8 volt and can swamp the smaller voltages which one desires to measure. The property of superconductivity which many metals exhibit at low temperatures has been employed in the develop- ment of means for detecting small voltages. Circuits with no resistance in them are often very valuable in making these delicate measurements. and in some instances investi— gators have incorporated the phenomenon of superconductivity directly into the measuring apparatus itself. A sensitive but delicate superconducting galvanometer was built by Pippard and Pullan (l) and was used quite successfully by Pullan in the measurement of the absolute thermoelectric power of tin and silver (2). The galvanometer. which was immersed in liquid helium. consisted of two separated loops of superconducting lead wire with a small magnet suspended from a fiber between them. When a current passed through the loops the generated magnetic field de- flected the magnet. twisting the fiber. which in turn rotated a small mirror. The deflection of a beam of light reflected from the mirror was then recorded. In practice the device was used as a null detector by inserting a small resistance of approximately 10-7 ohms in the circuit and passing a small known current through the resistor in opposition to the thermal e.m.f. The galvanometer was capable of detecting a current of 10-5 amp. hence a voltage of 10—12 volt could be measured with the apparatus. However. the very delicate suspension of the magnet and its sensitivity to very small vibrations and slight changes of the magnetic field in the cryostat made the device extremely difficult to assemble and operate. Pullan overcame these difficulties and was able to observe that the thermoelectric power of tin vanishes abruptly at its transition temperature contradicting earlier findings but confirming theoretical predictions. The results of his measurements of the thermoelectric power of silver and of tin in the normal state disagreed with predictions of the free electron theory (2). A device developed by Templeton. the superconducting modulator (3). worked on an entirely different principle. The modulator had a portion of a thermoelectric circuit made of fine tantalum wire. Tantalum at liquid helium temperature is superconducting. but it can be rendered normal by placing it in a magnetic field of about 60 gauss. The tantalum wire in the modulator had a resistance when normal of about 0.1 ohm. The resistance of the voltage source was only about 10-3 ohm. and the remainder of the circuit was made of superconducting lead wire. and so the modulating resistance. R. was large compared to any other resistance in the circuit. R (see Fig. l) was held in a steady magnetic field very close to the critical field of tantalum and was then "swung" in and out of superconductivity by a small magnetic field oscillating at an audio-frequency of 800 cps. Thus during each half cycle the current. which the source of e.m.f. forces through C is effectively interrupted as l. R.becomes normal. The A.C. signal induced in C2 can then be amplified and measured without danger that the measure— ments might be affected by the thermal e.m.fs. in the output leads. According to Templeton the device described here is capable of detecting 5 x 10-.9 volt with the noise level below 1% of half scale. In the superconducting modulator the modulating field has the same frequency as the alternating T """"""""" ’i I MODULAT me I I Rm 5 1'4 uc E I : 1 v svucunamzu 503:“ : C u AMp'uF/EA I E,M.F I 1 I (D I AND I u asuoouuron I I ‘ I I L. ......... - .I LIQUID HELIUM Figure l. The superconducting modulator r ------------- j I I I ; 4) I ' CD I I I 4 g ‘° I MEAamms L _J' INJTRuMmT: ; Homo REL/UM ' Figure 2. Templeton's superconducting reversing switch . voltage which is amplified and measured. Thus there is much danger of voltage pick-up in the thermoelectric circuit and in the external measuring circuits. Bucking voltages were applied at various points. but all of the pick-up could not be eliminated. Templeton also developed another device using superconductivity. a superconducting reversing switch (4). which was much easier to construct than the superconducting modulator and easier to operate. The switch reversed the polarity of a thermal e.m.f. or of some other small voltage source imposed on the output leads from the cryostat. By taking the difference of two readings made with the output in opposite directions one could eliminate the effects of thermal voltages in the leads and obtain the desired thermal e.m.f. of the sample. Referring to Figure 2. resistances a. a'. b. and b' are small coils of tantalum wire. Coils a and a' are between the poles of one small electromagnet. A, and coils b and b' are between the poles of another electromagnet. B. The normal resistance of the coils is large compared to the output impedance of the source. When the coils of magnet B are energized paths 1-3 and 2-4 are superconducting the paths 1—4 and 2-3 have essentially infinite resistance. When magnet B is turned off and magnet A turned on. paths 1—4 and 2-3 become superconducting and pathsl~3 and 2-4 have very large resistance. and the output from the cryostat is reversed. The sensitivity of the D.C. measuring instruments used in conjunction with the switch determines the limit to which measurements can be made. This limit is usually of the order of 0.01 microvolt. Templeton and Pearson used the superconducting reversing switch and the superconducting modulator in making measure- ments of thermoelectric potentials at low temperatures and in the redetermination of the absolute thermoelectric scale. (5). A superconducting chopper amplifier capable of detecting D.C. voltages as small as 10-11 volt was built by De Vroomen and van Baarle (6). The amplifier was designed specifically to measure the very small voltages from a low impedance source. The problem was to tranSfer the electrical energy out of the cryostat into an amplifier with minimum loss and with as little electric disturbance as possible. This transfer of energy was accomplished in a manner similar to that employed by Templeton in the superconducting modulator. but the method of generating the magnetic field to render the tantalum normal was quite different. In the chopper amplifier the fine piece of tantalum wire was placed inside a solenoid. A low frequency non-sinusoidal alternating current was sent through the coils of the solenoid generating at the current peaks a magnetic field great enough to render the tantalum normal. This method of rendering the tantalum alternately superconducting and normal. shown in Figure 3. contrasts with Templeton's method in that a transition to the normal state occurs twice during each cycle of the current in the solenoid instead of once. Consequently. the direct current in the thermoelectric circuit is modulated at a frequency 2f. where f is the frequency of the modulating current. From the waveform of Figure 3 it follows that the magnetic field can induce voltages only of frequency f and odd harmonics thereof. A superconducting chopper amplifier operating in essentially the same way as the one built by De Vroomen and van Baarle will be described here. However. several changes have been incorporated into the design. the most important one being the use of a superconducting solenoid so that larger alternating magnetic fields can be obtained with no power loss. This improvement may make it possible to use lead instead of tantalum as the superconductor in the low ' I J x I a: ' I _ O _ d J! I I I I I I I I I 1 F I” f‘b- - u o - - - - -- q ‘ I I! ' .r t, _ _. a: 4... P I Figure 3. Comparison of DeVroomen's and Templeton's methods of modulating the superconducting wire lO temperature circuit which would greatly increase the temperature range of the device. Details of construction and principles of operation of the superconducting chopper amplifier follow. II. DESCRIPTION AND THEORY Principles of Operation The superconducting chopper amplifier works on the principle that at low temperatures. i.e.. liquid helium temperature and below. the resistance of metals is lowered enough to permit voltages of the order of 10-11 volt to be amplified and detected. The transfer of the electrical energy out of the cryostat with a minimum loss of energy is accomplished by the periodic modulation of a resistance in the circuit containing the low impedance voltage source. The current in the low temperature circuit is. therefore. also modulated. and the wire bearing this current is used as the primary of a small step-up transformer which is incorporated into the low temperature circuit. Consequently. the current induced in the secondary winding of the trans- former. which has a step—up ratio of either 1:50 or 1:100. will be laternating. and thus can be lead out of the cryo- stat with no interference from D.C. thermal voltages generated in the leads. The modulating resistance is a piece of 0.005 inch tantalum wire 8 cm in length which is made alternately superconducting and normal by the periodic application of a magnetic field. 11 12 There is. however. danger of pick-up from the relatively large current needed in the solenoid to produce the magnetic field of about 60 gauss required to render the tantalum normal. Two steps are taken to minimize this pick-up. The current leads to the cryostat from the relay (see Fig. 4) are made of commercial eight-weave wire surrounded by magnetic shielding. and the coil itself is placed in a lead can which is superconducting at liquid helium temperature and thus is a perfect magnetic shield localizing the fields. Superconductivity is restored twice during each period of the alternating magnetic field which has a frequency f; consequently. the alternating current induced in the secondary of the transformer has a period 2f. Thus the field current can induce signals only of period f and harmonics of f. and if the even harmonics in the field current can be removed. pick-up from it in the output to the frequency selective amplifier can be reduced to a point where it is negligible. Low Temperature Parts While the device was in the process of development several solenoids similar to the one described in (6) were wound and tested. In an early experimental model of the Figure 4. l3 POWER 7.3 CPS I I AMPLIFIER "‘_“""” OSCILLATOR ' “REC“ ”A PHASE ' V SYNCH I—-—~ SENSITIVE I RECTIFIER dare-g TOR - AND ' ' RELAY ‘ + I COMMUTA TOR ~ 23 CPS Dc CA LIB RATION F R EQUENCY SIGNAL SELECTIVE If V AMPLIFIER I I ‘ r' _ .. -1 Samoa: auann mu ' ' -- - '1; RESISTANCE Ith- ' I III- : I ; I . . Pam/m: I ' '- -J I' ----- :ZI . I I I I ' : I I '.flMN£ ‘ I IT . __J I"L 1' I . : TANTALUM wms : I I I I. .. .. .. .. .. --.. I. .. ....I VAcwM TWORMER J II. AND H. - I'IEATER cons 1; AND 15 - THERMOME'TERS‘ . LIQUID MEL/UM of thermoelectric power Schematic of apparatus as used for the detection l4 device a coil of 1000 turns of #40 enamelled copper wire was wound on a paper core and tested. It was found that the power dissipated in this coil boiled away the helium more quickly than expected. quicker in fact than the time necessary to run an experiment of reasonable length. To remedy this difficulty it was decided to wind a superconducting solenoid which would be able to carry a reasonably large current with no power loss. Niobium wire was chosen fro the solenoid because of its high transition temperature (9.220 K) and high critical field. Figure 5 shows a comparison of the critical fields of tantalum. lead. and niobium. A length of 0.005 inch niobium wire was electro- plated with 0.0001 inch of copper which. when the niobium is superconducting. acts as electrical insulation. Also if. when a large current is flowing through the solenoid. a portion of therfiobium'wire should become normal. the copper can carry the current better than any other type of insulation. thus decreasing the possibility of burning out the coil. The niobium wire was first etched in dilute ihydrofluoric acid. It was then pulled slowly through the <:enter of a copper tube filled with the standard copper (alectroplating solution of CuSO and H $0 The solenoid 4 2 4° of 1000 turns was wound on a paper core with an inner 15 Eownoac can .pmoa .Esamucmu mo moaofiw Hmowufluu mnu mo comwummfioo .m shaman CS vtatumemu Ox G W N o _ Pi) I o Inoom g2 . 2* A3§tfixu orXf 16 diameter of 0.1 inch and a length of 0.8 inch. As a result of the copper electroplate the leads of the solenoid can be soldered directly to the output leads from the relay commutator eliminating the need for spot welding. The major advantage of the superconducting solenoid is that a fairly large magnetic field can be generated in its core without power loss. Consequently. metals other than tantalum could be used as the superconductor in the primary circuit. Tantalum has a transition temperature of 4.380 K and has the advantage that at liquid helium temper- ature of 4.220 K its critical field is only about 60 gauss which can be easily attained using a relatively small current in the solenoid. However. tantalum's critical field rises quite sharply as the temperature is lowered below liquid helium temperature and reaches a value of 860 gauss at 00 K (7). Therefore. tantalum can be used as the superconductor only at temperatures below 4.380 K. and for temperatures much below 40 K quite large magnetic fields must be generated. On the other hand the transition temperature of lead is 7.220 K. Its critical field rises sharply at first but then levels off and is actually less than the critical field of tantalum at temperatures below 20 K (see Fig. 5). Thus if lead were used as the superconductor the magnetic field 17 required at helium temperature would be greater. but at temperatures below 20 K less. than that needed for tantalum. and the upper temperature limit of the device would be 7.220 K instead of 4.380 K. In an early model of the device of 10 mil wire made from an alloy of 1.6 atomic percent indium and 98.4 percent lead was tried as the superconductor. and it was found that a direct current of 1.6 amps in the solenoid was necessary to render the alloy normal at liquid helium temperature. The indium was added to the lead so that its residual resistance would be increased. Hewever. this alloy could not be used in a working model of the device because. when the alter— nating current is sent through the coil. eddy currents are induced in the superconducting lead shield surrounding the coil. which make it necessary to increase by at least a factor of four the current required to render the super- conductor inside the solenoid normal. The electronic equip- ment used could not generate a current of this size. and so only tantalum was used as the superconductor. The core of the low temperature transformer was made from four 0.014 inch mumetal rings of one-fourth inch inner diameter and three—eights inch outer diameter. After machining. the rings were re-annealed and fastened together 18 with varnish. For the secondary winding a total of 200 turns of #34 enameled copper wire were wound on the core with a tap at 100 turns so the transformer could be used with either 100 or 200 turns on the secondary. It was found that the device worked more effectively when 200 turns were used. The primary of the transformer consisted of two turns of the tantalum wire. The calibration resistance in the low temperature circuit is a 10 cm length of #32 commercial copper wire with a measured resistance at liquid helium temperature of 4.45 x 10_4 ohms. This resistance is of sufficient magni— tude to test and calibrate the apparatus. but when making actual measurements with the device a resistance an order of magnitude smaller is desirable. Then. for example. a current of two microamps through a resistance of 5 x 10-5 ohms would produce a potential drop of 10.10 volt. the order of magnitude for which the device is intended. Two methods of connecting the tantalum wire to the copper calibration resistance were attempted. Electroplating the ends of the tantalum wire with copper and then soldering them to the copper wire was tried first. but it was found that the copper electroplate did not adhere well to tantalum. Globs of copper would stick to the tantalum. but they could be easily scraped off indicating that there was no actual l9 fusion of the two metals. and so the method of electro- plating the tantalum wire and soldering it was discarded. The connections were made by spot welding. Several methods of spot welding were tried and the best welds were achieved by welding both the tantalum and copper wire to a small square of nickel foil. Other methods of spot welding caused the tantalum wire or the copper wire to stick to one of the electrodes of the welder. A few drops of alcohol on the electrodes and on the wires at the points where they were to be welded created. during the weld. a reducing atmosphere which elimitaed oxidation and yielded a clean shiny weld.* The transformer and the solenoid were placed in lead containers which act as perfect shields against pick-up when the lead is superconducting. The complete low temper— ature circuit. including the transformer and solenoid. was then mounted on a 3 £3 x 1'; 9 X‘L ? piece of plastic. 4 l6 16 The current and potential leads were attached to the circuit at the copper-tantalum joints. For the current leads to the solenoid a length of commercial eight-weave wire with :magnetic shielding was used. The current and potential * I am grateful to Dr. P. A. Schroeder who suggested this procedure to me. 20 leads to the low temperature circuit and the leads to the transformer were #36 enamelled copper wire braided into an eight-weave to reduce the pick—up. The position of the individual compenents when mounted on the plastic is shown in Figure 6. Electronic Components The electronic components used in conjunction with the low temperature parts of the chopper amplifier can be divided into two groups. The first group generates the alternating current free from even harmonics which flows through the solenoid producing the periodic magnetic field in its center. The frequency selective amplifier. phase sensitive detector. and the voltage recorder make up the other group. The solenoid current originates in the audio oscil- lator (Fig. 7). is amplified by a 20 watt power amplifier. rectified by a silicon diode. after which alternate waves are commutated by a mechanical relay whose coils are energized by a multivibrator. The multivibrator was designed so that its natural frequency was as close as jpossible to 11.5 cps and is synchronized to exactly half the oscillator frequency. The synchronization is so phased that NICKEL FOIL TAN TA LU“ WIRE -'-—— 21 r COPPER anon-mm Rm: TA Ives TRANSFORMER ,4. I I I Souswno- I I I I I I LEAD . CONTAINER ‘ I I I I I I I I I I I I I I I I I I L_. -—--—-I LEAD ON TA INEH CUBA ENT AN D PO TENTML LEADS ( s - wsAvE) PM 5 TIC SHIELDED e-wsavs wmz ' Figure 6. Low temperature parts mounted on plastic 22 23CP$ OSCILLATOR RDWMTT ' Powefl loo K , AMPLIFIER 63V C) + HEATER 0 850V FILAMENT SILICON I' ---------- 2—- B --, DIODE I I I ? I C2) I &wK I I ' I' ------------ - 'I I [OK [OK I I - I I I I I , I, ‘ i , 2k .027... .027” ; :3 I I I I -’-/2AU7 . I@I —' I I I a 3 : | -------- _ - - -I_ -4. 1' - 35/214117 RELAY : IM IM : ® I I _ I I I A * I i I I I look I SOUHWMD L. _____________ J many/BRA TOR @nnnn \JVV QD_£:\ ll”\ [5\ /‘L GD.J”\. /"\ V V WAVEFORMS Figure 7. Circuit for generating the alternating solenoid current showing the current waveforms at various points 23 switching of the relay occurs during the period of zero current. The frequency of the current flowing through the solenoid is therefore 11.5 cps. and its waveform is such that it should be free of even harmonics. The frequency selective amplifier (Fig. 8) was found to be quite unstable and very sensitive to pick-up. Though it was physically separated from the oscillator and current generating circuits. and metal shielding was placed between the amplification stages. this amplifier determined the lower limit to which measurements could be taken. Details of the operational characteristics of the frequency selective amplifier are found in (6). Figure 9 shows the phase sensitive detector. This detector proved satisfactory. its only drawback being that neither of its outputs is grounded. Consequently. only voltage recorders with a high impedance. balanced input could be employed. A model G—22 Varian recorder was found to work quite satisfactorily with the phase detector. 24 OUTPUT To AIM? SMSI'I'IV'E SO , DETECTOR I5'K 3.3K ng> I “39% J 5:03} 4 6 K ' 2'“ 8 I—III “I we». 68k ‘ 2 68K I I ’57" 3 00K; I-5N<_ l-5'M 300K 1 L: I;— 9 55%.?" a 5:: :2 "I‘ I~ I» — I- . ::: F :2: --—-— —o -'_- -_I__ I 2 1’ “tar 3‘ IM “‘ . 1?: 3K 500k .2: ,M I a. so. ' ' ' ’ K5879 * :1: V. 5’: 7'9 /:'/o'oo . v, 5879 Figure 8. Frequency selective amplifier 25 I- I Houumuww m>fiuwmamm mmmnm .m musmwm Lhnakefifimfigu I; .xan H . 2.13.2.1 I; Q oMHu su>M¢ u III. OPERATION As each individual part of the chopper amplifier was completed it was checked thoroughly to see if it would function properly. All the parts were then assembled. and the complete apparatus tested. The device would normally be used as a null detector. However, for the present our concern is to determine the sensitivity of the chopper amplifier. Figure 10. showing the low temperature circuit. will be referred to in the following description of the testing procedure. It should be noted that the potential leads to the joints of the calibration resistance and the tantalum wire are not needed once the resistance is calibrated. One of the three leads coming from the transformer is the tap placed at 100 turns of the secondary: it was found that the transformer performed more satisfactorily using the full 200 turns. and so the tap can be disregarded. To check and calibrate the electronic amplifier and phase detector the device was tested at room temperature. The output of the 23 cps audio oscillator was fed into a series of step-down transformers and connected to terminals A, and the wire C was cut at the midpoint. MQasuring the 26 27 I v~ .--.~-..- “4 Current leads Potential leads Calibration resistance (Cu wire) Tantalum wire Transformer _Superconducting solenoid ® ©. ® ® (9 (9 Figure 10. Low temperature circuit used to test the - superconducting chopper amplifier. I Lead shields 6) 28 individual drop across each transformer and multiplying them together gave a total voltage drop of 7100 at the calibration resistance. The voltage ratio of the small toroidal trans- former in the low temperature circuit was 160:1 (step—down) at room temperature instead of the expected step-up ratio of 100:1. The reduction in voltage instead of an increase is due to the high resistance of the tantalum in the primary as compared to the inductance of the two turns: at liquid helium temperature. when the tantalum is superconducting. all of the impedance will be caused by the inductance. The total voltage drop between the oscillator and the frequency selective amplifier at room temperature was. therefore. slightly more than 106; hence a measured voltage of 0.1 volt at the oscillator corresponds to approximately 0.1 microvolt at the input to the frequency selective amplifier. With this input. the output from the amplifier was clearly detectable on an oscilloscope and in the output of the phase sensitive detector. Hewever. instability of the amplifier. manifested by 31 w but erratic variations of its gain. prevented detection of smaller signals. Prior to immersion in liquid helium. the calibration resistance was soldered together. A fairly large current (10 ma) was sent through the circuit with no current in the 29 solenoid. and no potential could be detected across the circuit. indicating that the tantalum was superconducting and the welded joints were good. With a current flowing through the low temperature circuit. the current in the solenoid was turned on and slowly increased until the magnetic field generated in the core of the solenoid was large enough to render the tantalum normal. When the current in the solenoid reached 80 ma the tantalum became normal. and a potential drop appeared across the parallel combination of tantalum and copper wire. The normal resistance of the tantalum at 4.50 K is 0.0099 its resistance at O0 C (8). Since the room temperature resistance of the 8 cm of0tantalum inside the solenoid was determined to be 1 ohm. its resistance at liquid helium temperature will be approximately 10.2 ohms and so can be neglected when measuring the calibration resistance which had an estimated resistance of 10—4 ohms before measurement at liquid helium temperature. Figure 11 shows a plot of the current through the resistance against the voltage drop across it with the solenoid current set at 100 ma. The corresponding calculated resistance is 4.45 x 10- ohms. Next the alternating current from the relay (see Fig. 7) was sent through the solenoid. As mentioned before. 30 V (mscrowm) 5 a—~ I (millimefi) Figure ll. Curve of voltage vs. current used to determine' the calibration resistance.’ 31 due to eddy currents induced in the superconducting lead shield surrounding the solenoid. the peaks of the alternating current had to be increased to about four times the direct current necessary to render the tantalum wire normal. Accordingly. with solenoid current peaks of 0.5 amp. the direct current in the transformer primary was modulated in the manner described above. There was some pidk-up from the solenoid current. but this piCk-up was constant and could be eliminated in the output of the phase detector by use of the zero adjust. The sensitivity was limited by the gain instability of the frequency selective amplifier. The lowest current through the circuit that gave a detestable output from the amplifier and phase detector was 0.2 ma corresponding to a voltage across the circuit of (2 x 10'4) x (4.45 x 10'4) = 9 x 10'8 v = .09 u v. The data taken with other values of the current in the circuit (see Appendix B). with all the gains settings fixed. indicate a linear relationship between the current in the low temperature circuit and the output of the phase sensitive detector. Every component of the system worked as expected; it is certain that once the amplifier instability is eliminated the sensitivity of the device will be increased 13y several orders of magnitude. 32 The device is easily arranged as a null detector. The tantalum wire is opened. and the ends are spot welded to small pieces of nickel foil to eliminate the difficulty of making tantalum connections to the source of e.m.f. Lead wires can then be soldered between the pieces of nickel foil and the voltage source. The lead wires will be superconducting and so will introduce no resistance into the circuit. The polarity of the controlled current through the calibration resistance is arranged so as to be in opposition to the voltage source. The unknown voltage is then obtained by adjusting and recording the current through the known calibration resistance until there is zero deflection on the null detector. APPENDICES APPENDIX A Data for the determination of the calibration resistance. I Current through the resistance V Voltage across the resistance £49111 1.1121 0.37 0.13 0.77 0.21 1.48 0.48 1.95 0.95 3.45 1.35 4.05 1.80 6.46 2.90 9.30 4.20 34 APPENDIX B Output of the phase sensitive detector for various values of the current in the low temperature circuit. Recorder Current Synch. Amplifier Amplifier Output (Milliamps) Gain Gain 1 Gain 2 (Volts) 0.2 6 8 5 .25 0.3 5 5 ?' ~02.i .01 0.8 " ’ ” .06 i .01 1.1 V " " .1 2.3 ’ " " .2 5.0 " " F .4 10.0 ” " " .92 35 10. 11. V. BIBLIOGRAPHY Pippard. A. B. and Pullan. G. T.. Proc. Cambr. Phil. Soc. 48 (1952) 188. Pullan. G. T.. Proc. Roy. Soc.. London A_;LZ (1953) 280. Templeton. I. M.. J. Sci.Instru. 32 (1955) 172. Templeton. I. M.. J. Sci. Instru. 32 (1955) 314. Pearson. W. B. and Templeton. I. M.. Proc. Roy. Soc.. London A 231 (1955) 534. De Vroomen. A. R. and Van Baarle. C.. Physica g; (1957) 785. Scott. R. B. Cryogenic Engineering. D. Van Nostrand Co.. Inc.. Princeton. N.J. (1959) 342. Smithells. C. J.. Metals Reference Book. Interscience Publishers Inc.. New York (1949) 480. MacDonald. D. K. C.. Handbuch Der Physik XIV (1956) 163. Kittel. C.. Introduction to Solid State Physigg. 2nd ed. John Wiley & Sons. Inc.. New York (1956). Furukawa. G. T. and Douglas. T. B.. A compilation in American Institute of Physics HandbOOk. McGraw-Hill Bock Co.. Inc.. New York (1957). Sec. 4. 49. 36 ‘3 "TS-DJ. 7“ L, MICHIGAN STQTE UNIV. LIBRQRIES IIIII IIIII III I||I III III IIII II IIII III III II IIII IIIII III 31293016875894