ENGINEERING RESEARCH INSTITUTE THE UNIVERSITY OF MICHIGAN ANN ARBOR Scientific Report ES-1 A RADIOACTIVE-IONIZATION-GAGE PRESSURE-MEASUREMENT SYSTEM N. W. Spencer R. Lo Boggess L. R. Brace M. A. E1-Moslimany ERI Projects 2096, 2406, 2597 GEOPHYSICS RESEARCH DIRECTORATE AIR FORCE CAMBRIDGE RESEARCH CENTER, ARDC CONTRACT NOS. AF 19(604)-545, AF 19(604)-1511, AF 19(604)-1948 LAURENCE Go HANSCOM FIELD BEDFORD, MASSACHUSETTS U. S. NATIONAL COMMITTEE FOR IGY NATIONAL SCIENCE FOUNDATION GRANT Y/10o20/249 May 1958

This report is presented in partial fulfillment of the requirements of Contract Nos. AF 19(604)-545, AF 19(604)-1511, and AF 19(604)-1948.

TABLE OF CONTENTS Page LIST OF ILLUSTRATIONS iv ABSTRACT v 1.0 INTRODUCTION 1 2.0 GENERAL ASPECTS OF THE SYSTEM6 3.0 IONIZATION CHAMBER 13 4.0 AMPLIFIER 24 5.0 RANGE-CHANGING CIRCUIT 29 6.0 TRANSISTOR POWER SUPPLY 33 7.0 MISCELLANEOUS 38 7.1 External Control 38 7.2 Battery Pack 40 7.3 Calibration 40 7.4 Choice of Hi-Meg Resistance Values 44 7.5 Constructional Aspects 52 7.6 Performance History 54 8.0 CONCLUSIONS 59 9.0 ACKNOWLEDGMENT 60 10.0 REFERENCES 61 iii

LIST OF ILLUSTRATIONS Table Page 7o61 Disposition of Pressure-Measurement Systems 56 Figure 1l1 Complete system in enclosing tubing. 2 102 Complete system out of enclosing tubing. 3 201 Block diagram of radioactive pressure-measurement system. 7 2.2 Output voltage and hi-meg resistor relationships. 9 203 Typical i-p curve for ionization chamber showing corresponding load resistances. 10 2o4 System schematic diagramo 11 3.1 i-p curve of typical ionization chamber showing low-pressure and high-pressure characteristics. 15 3.2 Assembled ionization chamber with attached low-leakage wafer and hi-meg resistorso 16 353 End plate of ionization chamber showing titanium tritide sourceO 17 3.4 Ionization-chamber elements. 19 3o5 lonization-chamber element assemblyo 20 3o6 Ionization-chamber polarizing electrode. 21 357 Ionization-chamber collector assembly. 22 4. 1 Amplifier block diagramO 25 4,2 Amplifier schematic diagramo 27 4 3 Amplifier subassembly. 28 5.1 Solenoid assemblyo 30 5.2 Range-changing-circuit schematic diagram. 31 6o1 Power-supply schematic diagram. 34 602 Power-supply assemblyo 37 7.1 Schematic diagram of external control circuit. 39 7.2 Exploded view of unit subassemblies. 41 7.3 Vacuum system used for calibrating pressure-measurement system. 42 7.4 Calibration setup at vacuum system. 43 7.5 Calibration curve, range lo 45 7.6 Calibration curve, range 1. 46 707 Calibration curve, range 2. 47 7o8 Calibration curve, range 3~ 48 7~9 Calibration curve, range 5. 49 7o10 Calibration curve, range 5. 50 7.11 Vacuum-fitting arrangement for connection of unit to nosecone wallo 53 7o12 Range position vso time for 3 units on Nike-Cajun AM 6.38. 55 iv

ABSTRACT An air-pressure-measurement system employing a radioactive ionization gage has been developed for use in rockets for the determination of surface pressures which can be interpreted in terms of the ambient pressure, density, and temperature of the atmosphere. The system has proven useful for the measurement of pressures from atmospheric to at least 10-3 mb. v

lo INTRODUCTION The remote measurement of a quantity with a precision of approximately 1 part in 50, when the quantitative variation covers several decades, poses interesting and challenging problemso Such a quantity is the ambient air pressure of the earth's atmosphere, whose magnitude decreases from approximately 1000 millibars (mb) at sea level to 10-3 mb at an altitude of 50 mileso This report is concerned with a description of a pressure-measurement system illustrated in Figso 1ol and lo2 that has evolved through several years of upper-atmosphere investigations carried out by The University of Michigan as part of the upper-atmosphere sounding program of the Geophysics Research Directorate of the Air Force Cambridge Research Centero Very crude measurements were first accomplished through the use of thermionic ionization gages (VG-1A) which were employed on several V-2 flightso The VG-1A, however, was not useful at pressures greater than approximately 10-2 mb, and thus alternative designs were soughto Possible use of an ionization gage employing a radioactive ionizing source was brought to the writer's attention by Professor Wo G. Dow of The University of MichiganVs Electrical Engineering Department, who was at that time directing the research program, and by Dro Jo R. Downing, the Inventor of the Alphatron,* the name of the commercial version of the deviceo Consideration of radioactive ionization gages in the light of known operating limitations imposed by the rigors of a rocket flight demonstrated that the device possesses considerable merit for rocket-borne service. Pertinent factors include: a) undamaged by exposure to atmospheric density; b) physically rugged; c) nearly instantaneous electrical response; d) manufacturing precision unnecessary; and e) current vso pressure range adequate for ground-to-50-mile measurements Certain drawbacks are likewise evident, such as: a) very small output current magnitude; b) radioactive ionizing source constitutes potential health hazard; Copyright by National Research Corporationo 1

::j i:: L I. ', 1 1 1 1,. ',.......... ";; - -. 1 I - -. 1 I I 1 1 I.......1. I I:::::::! iil::::l i i~:; ~....:il::::I:], ~:: i~I.1i'.:I I. 1 -, I.: X.:........ I.::.::::::::.'-:.::::::: 11,., -111 - 1 ~~~~~~:i:.1:.,-.- I I - I:..,:. - -:: 1:..:.::::::l ~ -~~i!~i!ii ~!iiii;i~lii i:::;l::i i::::::]:,.:::Il- I'll... I 11............,, 1,.::.. -,~~~~~~i::::::::...,....-~~i:ill: -..."::::.D~::.. -:,ll-i-x:::l-:- -;;~~j - i-Bii:i-~'' I~i..... -1I: -i:i:: c t..::::::.... I coI::::::..'::::::::::.. II-.. '' 1..-... -~~~~~~~~~~~~~~~~~~~~:. I. I I 11 I -,~~~~~~~~~~~~~: - -i: I -~I-::I r,l '::,:::::~::::::::.::''.::::.::::: '''":':' r-i:ii:: l.::: i:::,:i:::::: i l:::::i::::i ''::::'~i ii..,-i ddI:.. --- li.......... I I.iiiii'.ii i':::': I.... I...:X il '::.,::,:,:.,..:.!::::,,:: F j - 1.....: I I II ':-::::::..:::::::::..::::::::.,:::::,:.,.;::::::::::iii~:~~i~i:;::::::~:~i.-i]..... ~.- I _-: 1: ~~~~~~::,::,. ~~~~~~~~~~~~~~:. ~~~~~~~:,,:..:. I..;........ I..:.,:....: ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:.... -........_ 2..2.,:: illi il:...iI: I~:,::l:::lili,::. -i:: iii_:iir..::,1 i::: -'-::~iii:::~:: -:::i l i.:.:.........1 II:.:I I.. G q I:. 1~::::.::,,:.::,::::: 11- "::x::::::i 4::- I l.l.Ij.I;- - Iiiii -. l il, -~ I...:-:::!j.l:::lii-lii -.l':: ' ':': l:7::.... -iiii:Ii~i,:i I '.-1- 1 111li....::::::, -I I.~~~~~~~~~~~~~~~~~~~:;(: I:::!-ii:::::,:,1 i:I::- - - i:..::::;............. ll -: —:..:.....-:.:::I-:I:,~;1 ~,- I 1- 1 ii::-:::::::i; — -.-. - i~ l i:;::i'i ii i.-. iii...::i: i.-l::- - -l Ii.:: -....... I -. --..,.,.,...i~:. l I.i- I Il-:..:.I:.:I I -.:: I. -::i:..ji ~. - jl:.ii:::::i:::;-::l i_;.i.:'i:..'il~ ~ ~ ~~~~il.-:,~:~~~ ~ I..........;::i::i:,ii::i~::,:,:i:. I.1i..,:i.,i."ii~:-: ~i...... i::*i:l:!x.,,,:'I l'11.... I - 11-1 1 1 —l". -:. 1:1:~~~~~~:::,:,:.: ~~~~~~~:::~~~~~~~~,~~~~~~:~~:~~:~~:~~~~~~~:~~~~~~i~~:i::~~~~:~~:::::~~~~:::~~~:~~:~~ 1... ~ ~ ~ ~ ~ ~ ~......I. ~ ~ ~ ~ ~ ~ I - ' 'I'll I- I.. ''... — i,::;:i::. -ii:.. - ll-::::::.:-.. I --:-.11::.,::. I, -..... I -...... - ].~ ~::i::i:i:::::..,,::.. i::II ll.,.: -;-1- 1 1 I. I 1 11. -:,, -.~~~~~~~~` —.....1.4-1.1I -liiiii I - 1 1,,i'~~:. - I ip,.-i::': I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I-.::. I,::::::,.-OX: ~~ ~~ ~~ ~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~;::::'::::'::::::'':i:ii'i~::~~iii' ~ 1.. I.:.: ~~~::.-::~::::.X - 1. I -. -, —l-I.-::- I ~ s:.::: ': -......::.:.:-.:-::.,,-..'i~ ~i~i~ i i... - 111. 1 1 1-1..'.::!:. ii::~:i: i -,l i i.-.1. I..;: i., I ~.:~ I ':::.......:.::::.::::::::::.:::..::::::::::: j:ijii:... -i.i.iiiilii:::~::~ii.,i:I I: ~::I I. - I iiii iliiiI..., iiii ~:::: '::~:::::':i::::~::::~:'~'~'~":~:~:~:;':i:~:::: ~~~:~:~::::::::l:i.:::::ii i:i i''i'i'i'iil'i':':iliiiii~i:::::::~::::::::::.::.::::~::::::::::::::::: '::::' i::: $..:.....::l.:::jlii~:.:;:i:::::.::.:i: iili::::::::::::::::::':':':: ':~::: ii:i: i:::

m id~~~~~~~~~~~~~~~::::~::i:i::;:;:::;::: X 4f: ni — l. i.'i~: 00::0:::":::'' B 0 =TiR E:i f:::::::,:::~::;i:xijir0 ~~a~ it: a.:.::::::-::::., /:::ij,:::x::j i:: /::r:: iiiiilii~i:~i~:ai:-::::::::: ir- r ii::::::::.::::: Fig. 1.2. Complete system out of enclosing tubing.

c) relatively large physical volume (at that time) of existing gages (cylinder 3-1/2-in, diam and 4-in. length); and d) nonlinearity of current-pressure characteristic at both high and low pressures Consideration of these several factors led to the conclusion that the device could best be used by employing a d-c amplifier with unity negative feedback as a means of representing the small ionization current in terms of a voltage suitable for telemetering. Thus, following an original design by Dr. Downing (communicated personally), a completely battery-powered amplifier employing an electrometer input stage and a cathode-follower output stage was developed. Advantage was taken of the multi-decade useful range of the ionization chamber by incorporating a series of several hi-meg* input resistors which enabled the measurement of currents in the range 10-7 to 10-12 ampere, and subsequent interpretation in terms of pressure with adequate resolution considering the experimental technique employedO Since the equipment had to operate unattended in a rocket, automatic insertion of the proper hi-meg resistor consistent with the chamber pressure during rocket ascent was essential. Thus a d-c servo circuit was devised which would accomplish the desired operation. Ionization chambers** obtained from the National Research Corporation were employed, and several units constructed and flown in Aerobee rockets, resulting in measurements of upper-atmosphere pressure, temperature and density.1,2,3 Experience with these units and the constructional techniques employed at that time made evident the desirability of redesigning 'the circuitry to (a) eliminate many of the several batteries employed, (b) consolidate the various circuit portions into a single unit, and (c) bring about a simplification in the then complicated constructional technique. At the same time it was desired to decrease the chamber volume, increase the ionization current, and if possible reduce the hazard associated with the use of a radioactive sourceo Aside from these factors, the desirability of conducting relatively extensive pressure measurements during the IGY rocket program, which would involve nearly 50 pressure-measuring units, made it imperative to develop a unit embodying the improvements mentioned above, which could be easily hand-produced in quantities of 10 to 20 units. Accordingly, the development was undertaken and carried out, resulting in a pressure-measurement unit which is the subject of this report. The work was initiated under two contracts/ and essentially concluded under a third// with *Victoreen RX-l. **National Research Corporation, No. 510, patented. fAF 19(604)-1511, AF 19(604)-545o //AF 19(604)-1948,

additional substantial support from a U. So National Committee-National Science Foundation IGY Project grant. Initial studies which led to the design of an improved ionization chamber have been recorded in a report prepared in partial fulfillment of the obligations under Contract AF 19(604)-1511. 5

2.0 GENERAL ASPECTS OF THE SYSTEM The general instrumentation scheme underlying the pressure-measurement unit is essentially identical to that originally reported in the literature,3 and can be illustrated by a block diagram, as in Fig. 2.1. (See also the complete schematic diagram, Fig. 2.4.) The ionization chamber supplies, as a function of the density, composition, polarizing voltage, and, in some cases, temperature of the gas contained within its walls, a current which is passed through an appropriately valued hi-meg resistor. The voltage drop across this resistor appears as a portion of the voltage in the feedback loop of a d-c amplifier for which P, the feedback factor, is unity. Thus a voltage change effected at the "output" of the amplifier is nearly equivalent in magnitude to a voltage change across the hi-meg resistor (the input signal) but opposite in senseo The degree of equivalence is determined by the magnitude of the system loop gain as indicated by the familiar expression V.G. = A (2.1) 1 - AfB where V.G. is the voltage gain with feedback, A is the loop gain, and 5 is the fraction of output voltage returned to the input, unity in this case. In this system the loop gain is the order of several thousand so the voltage gain is, for the purposes of this instrumentation, unity. Thus one can consider the amplifier system as an impedance matching device or as a current amplifier, whichever is most convenient. The current produced by the ionization chamber lies in the approximate range of 10-12 to 10-8 amp as will be detailed in a later section of the report. Applicational aspects of the units require a resolution necessitating "division" of the total current range into 5 "sub-ranges." The necessity for maintaining the output signal in a nominal 0-5 volt range (determined by telemeter requirements) thus requires that 5 different ionization-chamber load resistances be employed. Since the voltage gain of the system is unity, it is apparent that these resistances must be of sufficient magnitude to produce the desired magnitude of output signal, Equation (2.1) means that the voltage gain will be essentially unity with no phase shift for considerable variation of circuit parameters, tube character 6

ION CHAMBER LOAD RES I STANCE ELECTRO- D-C METER AMPLIFIER if ---- TELEMETER I I I I --- RANGE CHANGING CIRCUIT TRANSISTOR POWER SUPPLY SILVER CELL BATTERY SUPPLY Fig. 2.1. Block diagram of radioactive pressure-measurement system.

istics, etc,, provided the loop gain A remains sufficiently high. This concept is applicable when considering a change in the input signal and a corresponding change in the output signal, but does not require any particular d-c output level corresponding to a particular, constant d-c input levelo Thus an arbitrary operating level may be chosen for reference. Consider the situation for zero ionization-chamber current, consequently zero volts across the hi-meg resistor. The system is so arranged, that is, voltages, circuit constants, etc., are so chosen, that the output operating level will be 600 volts, herein referred to as the "zero pressure voltage,' or "ZPVo" Now, if the pressure in the ionization gage is increased from zero, the output voltage will correspondingly decrease from 6o0 volts. When the drop across the gage load resistor (hi-meg) reaches 6 volts, the output will be zero volts, and the point will nave been reached (limit of telemeter) where the next desired hi-meg resistor in the series of 5 should be inserted. Its value is chosen so that the output voltage will be 5o0 volts when it is insertedo If the ionization-gage current is permitted to increase further, the output will again decrease, allowing, when 0 volt is reached, repetition of the above operational sequence. These considerations are illustrated in Fig. 2.2, which shows the output voltage corresponding to several decades of input current, It is apparent that for proper operation the load resistance should be changed each time the output reaches 0 or 5 voltso To this end, the rangechanging circuit uses the system output voltage as an "error signal" to control operation of a bi-directional rotary solenoid stepping switch. Self-induced switching is accomplished in either the increasing or decreasing current direction, hence either increasing or decreasing pressure direction, and thus the range-changing circuit acts to select the proper resistance for an on-scale output indication throughout the total acceptable pressure-measurement range of the instrument, Figure 2o3 illustrates a typical current-pressure relationship for this equipment. The corresponding subdivision into "sub-ranges" by the ionization-chamber load resistances is shown by superposing a calibration voltageresistance grido It is apparent from Figo 203 that peak currents near 100-mb pressure, and low currents near 10-12 amp lie outside the grid, that is,no provision is made for determining these valueso The corresponding outputs will lie below zero and above 50, volts, respectivelyo This situation is not detrimental to operational use of the instrument as these regions lie outside the useful range of the instrument. Although, normally, range-changing would occur, it will not under these two circumstances as the circuit is disabled to prevent switching below range 1 and above range 5o 8

(ZPV) 6.0. 5.0 - u 4.0 -_ 0 > 3.0 -. 2.0 Io 0 1.0 0 - -12 -II -10 -9 -8 -7 IONIZATION CURRENT (AMPERES) Fig. 2.2. Output voltage and hi-meg resistor relationships.

l0-r u-8 w 00' 108 La cr w Cl< 1079 z z -II I0 QQ w 1012 10-lo 00 i Iri M - I- I I I- I i i-t- i-i I - iH H 1i~ —f — 4 1++ H iH I -4- 4-l 4 I- I I I, I I I!I4i- i 1 I I -4-! i I '.00 0017 I I I I RI I f~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ R2 I T~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I f I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ R4~~~ woll,~~~~~~~~~~~~~~~~~~~~~~~~~ L.900~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -..oo~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ *~ 10-2 10'PRESSURE IN 10L MILLIBARS 10' 102 Fig. 2.3. Typical i-p curve for ionization chamber showing corresponding load resistances.

urea ip oTseuuiaqos ma-sE * - Z * Tj' ' - ON al>Ma 3JLva | 3nssi *.- __.| N NO Kiev; o 1:1 31 W31SAS lN3W3UnSV3W 38nSS38d 39V9 NOIVZINOI 3A11V01itaV anx. 8S-1~-C 3eVO AS a 3>03Ho L6SZ I Xo3romd I I %o01 S83HIO tV- 3ONvtU3101 %9 ** lINn HOV3 NI a3isnrpav S3mvA IIV * NVEIHIW MuOgUV NNV NV91H31)W 40 AlSa3AINn 31n.lSNI HOYV3S38 9Nia33N19N3 An a3NIS3a S'1130 3A11S 1-UH-V oA T Uv-noNv,,'oo T WNVWI3a,,'M T -IVNOIXOVIA - 3ONV13-0o. V 01o O3H 3a Lsnw a3IdJ13dS 3SIMN3H1-O SS3Nn SNOISN3r la -rv J I I I I *O'0 S10A 8'S 0-~ 801SIW83Hl -H 'ONI 39NVt 0 - 80LSIWN3Hl 0 - N SO10A (i,- O —p QNnouo o-, e1-S _A 2/1 (-A -N r 01 lo X a H 83dlM -_ rD Z5t - 01-S t~s - c ro I O E4^ 01 8 * _ 00 (0 5 I o s I b2 NI 90 LSZNZ,.6 1XI1 U06e I'j I-dI~~ 8: 1-d 1-d 1-10 ftA> AQ4 Q Q Z -d DO00000,Z z-S 2 Z E-1S l | ^,~N 1 I l 16 NI VU~OA n r A Eri- AV-13N ON3Z -L9Z NZ [X1l 6~ NO. 0s biO N0 1-P ~ g - a I I vtl UVJ I Sl10A 01 + 0- -- I ~~ NO NO1-GH o- ----- ndlno o ---iwo 1- IV\ -n X3Z'a 3~ hi: }s A 091 1ior____ A 0G1 01' oI - I -IL+ est i,1 A 091 go0 A (s1 joFT?I r6~ N1 za E6~ N1. I a 71a X I 'Of t8:) I '63 8 1~d'.A I I -eL J.-LV a LJ - I II l - FT I Pl-T 1 '93W I IZ tl Z-3N WZ'l OZ^S <. 3'1 P^UIOOI90; fzo 5 ~Z c ~'9,I cP9 Z- 3N 93W '0-1 ZZd 'AO0 AQ0O__ s* X AM0R _ AAA 081 < 9r3YU; |93mI xZ i I z I=1Z ' )1Z~~~~~~~~~~~~~~~~~~~~~I~ ~~~4 Ml G.l - 8 Z~S~~~1 >iZI )lZC8 1~c^ --- Adr-Z XLZ 8 ' >M0~ 93W 9~d8 ON06~ '0 AA X I * L~S UO~~: *Sd 8 --- VVVGtr022 4 GZ 8 ti It II8'l II XOHL t -- i l l U0~~< 8eu Ax A Z* e9S 93M O'lI X ~Z d 6Z H _. ^ 61 8 81d 20 g A 003 I0 8bj A 1 T I0 g -S 0 819 'a== 9i -, — Ir0 p-8 L a. ol n *i 09S j,**C)I OZ9 -- t 062( 191 Nl otr 9 =.__ 1Od WIHUl I G.- - 1 0013 ''' I PltltPI ___. OLP, 089 M,91 O8 b 8 GZA ', < tB I I 8 { Z119 **M 062 bU;, V Z,18 I0 8 I^o8 1, I S I. -.a112 -H,NoZ8e1 A: 62 SX01S je l l 8 Ilid L. X 06~ 01 8 i i 4ol t_; L '1 ST130. B6H z DOOf^___ EZ119.^ 800V WHI It== a ' yyM Zs.' Ha r DI Wb! N ACS 2200 13 1 m IM L I > -At 110 0 6. 0./LVNOO, <f AV13d I i I VI-S fN30 Ilx A~Ti san )lS1O3Nd %I< < w 8'9 39V9 NOILVZINOI - -

350 IONIZATION CHAMBER A radioactive ionization gage produces a current that is proportional to the density of the gas within its walls, for a large range of densities. Several factors determine the magnitude of the current at a particular density, including: a) volume of the chamber; b) strength of the radioactive source; c) composition of the gas; and d) electrode potentials. Although the i-p characteristic is essentially linear over a large portion of the chamber's useful range, it departs from linearity at low densities and ultimately at high densities. At the low-density end, the i-p curve approaches a particular low value of current, not gas-density-dependent, known as the dark current. As the density decreases and the ionization current correspondingly decreases, the dark current becomes a larger and larger fraction of the total current, finally masking the ionization current completely At the high-density end of the scale, recombination occurs for a larger and larger fraction of the ions as the density increases, thus limiting the number of particles available for collection. In addition, the useful path lengths of radioactive source particles become shortened by greater energy loss as the density, and hence the number of collisions, increaseo These two major factors cause the i-p characteristic to depart from an extension of the linear portion. One aspect of the i-p characteristic warrants further comment. Many chambers exhibit, in the operating region where recombination plays an important role in determining the output current, a hysteresis effecto That is, if the output current is observed for density change in one direction, for example, to lower densities, and then observed for increasing densities, it becomes apparent that the current is greater for a particular density when dp/dt is negative. The magnitude of the difference is observed to vary with the geometry of the chamber, and is at present believed due primarily to adiabatic temperature changes and their effect upon the recombination and attachment coefficients for the gases in the chamber. The particular ionization chamber employed as the basic sensing element in the pressure-measurement system described in this report has been designed on the basis of a laboratory investigation that considered various geometries, *See Ref. 4 for further details of an investigation of this effect. 13

electrode configurations, electrode potentials, and radioactive sources. In general, the objectives of the investigation were to: a) increase the magnitude of the output current for a given density as compared with an earlier model (see above); b) reduce chamber volume; c) discover the causes for and if possible decrease the dark current; d) discover the causes for and if feasible eliminate the hysteresis effect; e) eliminate hazard due to presence of radioactive ionizing source; and f) determine a design that could be easily fabricated by relatively unskilled technicians. To a substantial degree, all the above objectives are realized in the current design illustrated in Fig. 352, which shows a completed chamber with attached low-leakage ceramic wafer and associated hi-meg resistorso The design changes leading to realization of these objectives are largely interdependent, and the single factor which makes the re-design possible is the radioactive source. Previous units employed a sealed (nickel overlay) stabilized radium source of activity up to 3 mC, which emitted, for useful ionization purposes, adequate 5 Mev alpha particleso The sources also emitted significant gamma radiation of an intensity sufficient to cause hazardous working conditions within 2 ft of the source. The gamma was of no value, producing a wholly insignificant contribution to the useful ionization current. Consultation with the source supplier led to consideration and subsequent adoption of tritium as a source material. Tritium is a high-activity beta source emitting 17 kev particles. Its known characteristics make it very suitable as an ionizing source for this applicationo It is characterized by a half-life of 12 years, adequate for calibration constancy for this application. Under normal conditions it exists as a gas and is thus unsuitable directly. However, it readily combines with titanium to become titanium tritideo This form can be bonded to a stainless-steel foil which in turn can be secured to a convenient backing plate for applicational purposeso Figure 3.3 shows the titanium tritide-stainless-steel foil backed source secured** to an ionizationchamber end plate. In preparation of the source, a thin layer of titanium (j 1 i) is rolled on the stainless foil and then tritiated in a vacuum furnace. A thin layer is necessary to limit penetration of tritium and the consequent unnecessary loss of beta emission in the material, which, because of the low energy of the particles, is substantial in the 1-k layer. Quantitative data on the relationship between effective and total activity are not obtainable utilizing existing *T. W. Taylor, Uo S. Radium Corporation. **Araldite Epoxy Resin No. 502, Catalyst HN951, CIBA, Inc. 14

-8 a. -9 ( — -Ir 0 o - -Il to High-pressure collector charocteristic Low-pressure col lector characteristic -4 -3 -2 0 I 2 3 LOGo PRESSURE, MB Fig. 3.1. i-p curve of typical ionization chamber showing low-pressure and high-pressure characteristics.

TRITIUM SOURCE PLATE CERAMIC WAFER HI-MEG RESISTORS.... /.. I"...' i — H311 Fig. 3.2. Assembled ionization chamber with attached low-leakage wafer and hi-meg resistors.

I INCH Fig. 35.. End plate of ionization chamber showing titanium tritide source. 17

techniques. It is estimated from chamber measurements, however, that an effective activity of approximately 1 curie per square inch is attainedo The source in Fig. 353 has an area of about lo4 sq ino Figures 3.2 through 357 illustrate the physical aspects of the ionization chamber. It is seen to consist of an outer stainless-steel wall, cylindrical in shape, closed at one end by the source-bearing end plate, and at the other by a disc carrying several glass-metal seals, to provide for external electrical connections, and element mountingo Internally, the gage consists of a cylindrical polarizing electrode (Fig. 3.6) and a collector assembly (Figo 357) which supports 2 collectors, one for the high-density end of the range, the other for the low-density end. It is to be noted that the source faces the open end of the polarizing electrode cylinder, directly adjacent to the high-pressure collector, the circularly formed wire shown in Fig. 35~5 This collector is supported on 2 glassmetal seals which are in turn supported by a bridge which spans the open end of the polarizing electrode. The glass-metal seal visible in Figo 355 in the center of the bridge supports the low-pressure collector which is shielded from direct source radiation by the bridge. The purpose of shielding the low-pressure collector is to minimize direct impact of the source electrons and the consequent secondary emission. The area of the collector is minimized to reduce electron emission that will occur due to x-ray radiation from the surface of the polarizing electrode under the influence, again, of the source electronso These precautions are essential to reduce the dark current, an important consideration, for the system cannot distinguish between collecting + ions and emitting electrons. Extensive tests have shown the dark current to be directly dependent upon the area of the collector (consistent with the x-ray concept as reported for thermionic ionization gages by Alpert5) and upon exposure to beta (or alpha) source emission. The relatively low energy of the radiation means that the path length of the betas in air decreases sharply as the density increases. At higher densities, little ionization is produced in the vicinity of the low-pressure collector and an alternative collector near the source must be employed, Thus the chamber is provided witn both a low- and high-pressure collector. They are proportioned so that in their linear region, near 1 mb, the currents are roughly equivalent, eliminating a step when changing from one to the other, Typical i-p characteristics of each are shown in Fige 351, and a composite curve in Fig. 235o The chamber elements are fashioned from stainless steel, spot welding being employed in the assembly of the structureo Two thermistors are usually employed during flight measurements, one in air, visible in Fig. 3~6, the other held in a small depression in the outside of the polarizing electrode end plate, the de pression being visible in Figo 3560 These thermistors enable estimates of the 18

i:::ji: i:.d -P 0) m 01) F-I a) a, a) -p (Ti Cd N I Hrl Cd.r-4 0 H:E E i: ':;ii j: E iiiiiiii:::i:EERE:: El S i:::: i:: i::: g. iiii iEEEi.ESS.:ES::g:EEREEEii i: IE:i:E: iL:::::: i:iil:i:::i::::::!:EES~ES iE::E; i:::iE~:;; i i:::EE::E~;::::::.::::::::RE~SA:EiSEESEE:::~EEiE!iEfiEE~EEEE:E:EE E::fiE:: gE:-::::: I: ji i::: i:j i:i ~:E ii i: iiiii 0ii:;;::yf;;:;;E i:: E iE: iE EE:EE iE i::E iERE:Ei::Ei.:EER:EiS:E:: i i E i: E E:::::: i i::: l:::::::::::::::::::.:::.E.E.:: E iS:::::-::::E!E~::g~iE:0:D:::~EE:D: i; i:E g f i ii::.:i::;;::;;:Ei y: LEC:gi:iiiiSS;;:;:irii CEiirE.:E::: i::::::::::i::E::::::::i:E::ii::i I;:: g:!:::gE~;:0;:~;::i:::E:;:::::;L l:;:::::::;:i:iiE:::::::::iEi R E: E:::::EEECE!:EtSEE:iE;::: l::E:i::ii:!i:gEEE i~~,:ii::::fI:::::::::t i EE iE: ii. i 2i i:::!;1ESEE EE iE ilyE iEg EE ii iR 0 EE;:::i: i::iE::::::::;:::::::::::::!E:I: f:!::gEERE ':':ECSC!E:.::!::EED!E::::EL: i g~ ~g:::::::i::: iEE;i:::::::: i::::i:::::E: Ii-EEE i:ED:i f::g::: iiEEEEE:E::iE!L;:::..: iE::::iE:::;:E:::::E: i!: E C: i: i:: ili ii:iil;:: iiEEHS C:fii:;::ER::E~:::::::: iD 2:::::E:::i::E:::DDi:RRigE::::::: i::::::::::::: iEEE::!E!S::::RE::iEE:::E::::::: i:::: t:::!E:::::: i:0:E::i:: i: i::::::E;:::::::::E:::::: i:: i::: E:;;i:E::i::;;;;~~~~~~~~~~~~~~~~~~~~~~~~i::::: iEi::i I:iitESE:i:l:Eii E;;iiljl: E i:: ~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~:E:j~::::.iiE:i:j::Eljjiiiii:i::i:::E: S::::::::E:EE:EL~::: ~:::::E: D:'::'L:f i::'::E:ER S::E:EiE:: i::i::l:..................::.: i.i i i.i i: i i: Ei i-E E E E E iEE:i:;:! i..i........::::...:... 19

I,,ji 8: M Fig. 3.5. Ionization-chamber element assembly. 20

THERMISTOR.1 // I * r 1 I Fig. 3.6. Ionization-chamber polarizing electrode. 21

HIGH PRESSURE COLLECTOR LOW PRESSURE / COLLECTOR' I ".NCeS I '' " ' '1 I ON. 0.... ooollwwo.00 400" l....W~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.:..St.. O\+>.:.:..:.. 7:&*:.fs 9. Fig. 3.7. Ionization-chamber collector assembly. 22

gas temperature in the chamber. The collector potential +45 volts is established by an external battery. This arrangement is necessary to obtain a constant potential. It is expected that a Zener diode will ultimately be incorporated for this purpose, as constancy is required. 23

4.0 AMPLIFIER The amplifier, Fig. 4.3, employed in this system has been developed to meet several requirements including (a) high input impedance to make negligible any loading effect on the highest value of chamber load resistance (1o9 x 10 ohms), (b) low output impedance so that externally connected devices such as the telemeter, meters, etco, would not appreciably alter the magnitude of the output signal, (c) reasonable independence from supply-battery voltage changes, and (d) a loop gain of several thousand, so that the gain would not be a factor in the system calibrationo To meet these requirements, the circuit of Fig. 4.1 has been devised. Outwardly, it is a straightforward 5-stage direct-coupled amplifier with an electrometer input and a cathode follower output stage (see block diagram, Fig. 4.1). The electrometer input stage satisfies the first requirement for high input impedance, while the loop gain and cathode-follower output enables a low output impedance, which can be shown to be z 1 (4.1) gm A where gm is the transconductance of the cathode-follower tube and A is the loop gain, a real quantity for the usual operating range of the amplifiero The five stages are centered about the electrometer (V1), a voltage amplifier (V2A), a differential stage (V3), another voltage amplifier (V2B), and the cathode follower (V4B). The individual stage gains taken singly are approximately, in the above order, 2, 20, 10, 10, and o08. The product of these gains is not, however, the loop gain, for the internal resistances of the power supply (both B+ and B-) have a profound effect upon operation of the amplifiero The effect of these impedances is to impose a complicated feedback situation, both positive and negativeo For example, a change in current in the cathode follower reflects into the preceding stages a feedback voltage in the form of a B+ changeo The polarity is effectively different for each stage, and the magnitude is sufficient to overwhelm the normal signal. Thus it is not possible to determine the loop gain by simply multiplying the individual stage gains. It may be obtained, however, by determining the ratio of the output voltage swing to the electrometer grid voltage swing for a representative change in voltage in the feedback lineo Measurements of the gains of groups of stages indicates that the gain V2A is much greater than anticipated, indicating that it is probably operating as a starvation amplifier, very close to cut-off. In contrast, the gain of a representative system from the grid of V3 to the output is generally only about 10. 24

ELECT -- Ist MILLER STAG E COMP 2nd B+ STAGE COMP. 3rd STAGE C. F. 4 4 4 4.I CHAMBER LOAD RESISTANCE I1 o E OUTPUT 'I 4 4 4 4 4 Fig. 4.1. Amplifier block diagram.

V4A provides compensation* for cathode temperature change and its effect upon the amplifier operating reference level. Apparently it does not play a decisive role in determining the system gain, although this possibility has not been investigated, Several special points warrant comment: a) V3A, although part of the differential amplifier stage necessary for proper phase relationships, acts also to compensate for the effect of B+ change upon the reference level of the amplifier. Adjustment of the grid's operating point, R14, permits quantitative variation of the compensation. b) R6, R9, and Z comprise a regulator (Zener) for stability of the electrometer filament current. Two series connected diodes are employed to attain the desired voltage level c) Relay RY-1, when closed, forces a constant and known voltage in the feedback loop across the hi-meg resistor group and thus simulates a pressure-derived current. The corresponding output voltage becomes the reference level for the system, so that changes from an adopted value can indicate necessary corrections of output indications d) The ZPV is selected by adjusting the d-c battery voltage inserted in the feedback loop. This adjustment is accomplished by the voltage divider associated with battery BA2, Figo 4.2, e) Capacitor C1 prevents high-frequency oscillation within the loop. f) The frequency response of the system is largely a function of the magnitude of the hi-meg resistor in use and the capacitance across the hi-meg resistor. This response is difficult to measure satisfactorily because of the low frequencies involved, and the necessity for inserting the test voltage (in lieu of a pressure change) in the feedback line. Measurements have shown, however, that the response is flat, for practical purposes, to at least 3 to 4 cps for the highest input resistance used. Generally, rates of change encountered during flight (for example, due to roll) do not require a response in excess of 1 cps, g) Stability of the reference level of the amplifier is considered excellent. The reference output, generally 253 volts, is usually stable throughout a typical flight within 0.01 to 0,02 volt, the limit of most telemeters. From cold to operating temperature, the drift of the reference output generally does not exceed 0o03 volt. *Miller Compensator. 26

V R7 7.5.K:R.6 27. SG 22 *Z 470 K R4 4.;:R8 *330Q +6 RELAY Fig. 4.2. Amplifier schematic diagram.

'ZERO" RELAY ELECTROMETERS -';~~~~~~losl ", -1,,....... I.:,x:-... Fig. 4.3. Amplifier subassembly. 28

5.0 RANGE-CHANGING CIRCUIT As noted in Section 2.0, the purpose of the range-changing circuit is to enable automatic insertion of any one of 5 possible ionization-chamber load resistances consistent with the current and fulfilling the requirement that there be maintained an on-scale output indicationo Thus each time the output signal exceeds 5 or becomes less than zero volts a switching action is required. The output signal accordingly offers a natural means to initiate the switching action. Physically, the resistor changes are brought about by bi-directional rotary solenoids actuating a ceramic wafer switcho See Figso 51l and 352 for some mechanical details, Typical rates of pressure change during rocket ascent do not indicate a need for range changes more frequently than every 10-15 sec. For this reason, the average energy requirements placed on the power supply for this mechanical function are relatively modesto However, the impulse required to activate a solenoid is substantial (typically 0O5 w sec) and so an energy storage system is indicated. In this equipment (see Figo 5.2), the energy is stored in a capacitor (C6, 100 mfd) and is expended in the solenoid (L1 for decreasing pressure or L2 for increasing pressure) through the thyratrons V7 and V6, respectively, causing one solenoid action and thus stepping the ceramic wafer one positiono Capacitor C6 is charged through V5 from winding S-5 of the main power transformer. The R-C networks R18 or R19 and C5 hold the charging tube V5 cut-off during discharge, preventing continued conduction through the thyratron once C6 has discharged. As soon as C5 has discharged, V5 will conduct, permitting the storage capacitor C6 to become charged again to its normal level, approximately 145 volts. V7 conducts when the system output reaches 5~0 voltso A voltage related to the output level appears at the emitter of transistor Tu whose base is held at 5.8 v for reference. Roughly speaking, when the emitter exceeds the base potential, emitter-collector current flows, This current, flowing in R29, R30, and R55, swings the control grid potential of V7 positive, firing the tubeo A similar situation exists for V6 when the output signal goes negative. The emitter of transistor TD is referenced to a fraction of the 5~8 volts by R27 and R28. When the output of the system passes through zero, base and hence collector current flows, swinging the grid of V6 positive, causing conductiono There are several additional aspects of the circuit which are necessary for proper operation: 29

:: j: (~:::::~:. -. i:: i: Fig. 5.1. Solenoid assembly. 30

OUTPUT I +5.8 v Fig. 5.2. Range-changing-circuit schematic diagram.

a) The system must continue to carry out the switching function under a circumstance which would maintain the control grid of either thyratron positiveo Such a condition arises, for example, in performing a total head pressure measurement, where a sudden decrease in current is encountered due to a rapid increase in ram pressure during boost phase. The system may call for a 2- or 5-step range change which will hold the V7 control grid positive. The controlling function of the shield grid is utilized under these circumstances. It can be noted (Fig. 5.2) that the 2 shield grids are returned to -45 volts through R36 and to a fraction of the voltage of C6, determined by R20 and R21, through the neon tube NE-2-2. Thus, when conduction of a thyratron ceases, the shield is biased negative, and remains so, preventing re-ignition even though the control grid may be positive, until NE-2-2 conducts again. R20 and R21 are chosen so that the NE-2-2 will not re-ignite until there is sufficient charge in C6 to effect correct switching. Thus the shield grid controls the switching function in this circumstance, and permits repetitive switching when needed. b) The RC combinations R23, C3 and R22, C2 are employed to impose a time delay in the thyratron ignition sequence. During rocket ascent when the angle of attack of the nose cone is less than 1-2 degrees, the cone wall pressure does not exhibit significant roll modulation. However, for larger angles of attack, which are not great enough to render the data useless, the output signal may momentarily exceed the switching limit. Under this circumstance, switching is not desired, for more data will be obtained by maintaining the mean value range in use. The RC delay is introduced to prevent a range change under these circumstanceso The time constant is not great enough to delay "normal" functioning. c) NE-2-1 is employed only during activation or deactivation of the system to prevent inopportune firing of V6 and does not function during normal operation. d) Capacitors C7 and C8 in shunt with the solenoids assist in extinguishing the thyratrons when C6 becomes dischargedo 32

6o0 TRANSISTOR POWER SUPPLY The usefulness of any instrumentation power supply depends on its ability to supply the required load voltage with acceptable voltage regulation, noise, and efficiency, but in upper-atmosphere research instrumentation a vital first consideration is reliabilityo The amplifier and switching circuit described in this report were designed with this thought in mindo Accordingly the use of a negative voltage supply for grid return enabled the elimination of several bias batteries in the d-c amplifier and thus all interstage biasing could be accomplished through the use of a common supply, thereby minimizing the number of energy sources. These considerations, and others relative to available space, and the necessity for physical ruggedness imposed by rocket use, indicated the desirability of employing a transistorized power supplyo Transistor-oscillator power supplies were in the early development stage at that time; considerable effort was applied to the problem at several laboratories. The authors are indebted to Mr. George C. Uchrin of the Uo So Army Signal Corps Engineering Laboratories, Fort Monmouth, New Jersey, for suggesting a fundamental circuit and for contributing his experience to the design of the power supply which is to be discussed. Mr. Uchrin had successfully built several supplies, including one which delivered 10 watts at 1000 voltso The basic circuit employed utilizes two transistors, a transformer having a center-tapped primary, and a center-tapped feedback winding in addition to the load secondary winding. These components correspond to T1, T2, P1, and P2, the two halves of the primary windings, and Fl, F2 the feedback windings, as shown in Fig. 6.1. To minimize switching transients, it is necessary that the leakage reactance of the transformer be as low as possible. The toroidal core which is used for this reason also provides a square hysteresis loop, A shorter switching time results, allowing a power supply to be designed for maximum efficiency. It is possible with such a transformer to switch a high current safely, causing the transistors to exhibit a greater-than-rated power dissipation during the short switching cycle. However, for the sake of reliability, the power supply to be described is not operated under these conditions, that is, at maximum power-transfer capability. The operation of the circuit can be briefly explained as follows: while one transistor is conducting, the voltage induced in the feedback winding is e = -N(d0/dt). The windings are connected so that the conducting transistor 33

+90 m1 - --- I11i t t - R 29. I K: R 30 I K: DI IN 39 D2, IN 39 I C7 150 v 5/f I C 10 150 v 5 tf IR31 $2.2K _l C8 150 v 5/f I C 11 150 v 5/if R 32 2.2 K I C9 150 v 5/f D4 IN 91 D5. IN 39 c- jD3j S-1 IN 91 SF-IF OO00Q.Q0 A00000.SF-3 4 tS0 I 2 2 S-2 YQ QYQQYQpQpQJQPQpQppQ <Z4 0000000 P-2 TX2 2N257 00000000' CT-I P-l R33 '390Q 4 a F-l F-7 F-2~ D6 IN 34 I -rvI IJ Al 2N257 Fig. 6.1. Power-supply schematic diagram.

will be biased "on" and the nonconducting transistor will be biased "off." However, as the primary current reaches a limit, imposed by the circuit constants, or as the core becomes saturated, the emitter potential will drop to zero and the conducting transistor will cease to conduct. This will result in the collapse of the field and reversal of sign of the feedback voltage, causing the originally conducting transistor to be biased "off" and the nonconducting transistor to conducto This cycle will repeat at a frequency primarily dependent on the magnetic characteristics of the transformer and the battery voltage, the transistors acting essentially as switches. The output of the power supply is a square wave which may carry relatively large switching transients, unless particular care is taken in the choice of circuit parameterso Three basic configurations, common base, common emitter, or common collector, can be used. The operation of any of these is the same in principle; however, in the case of the common collector and common emitter circuit, the feedback windings must be independent. One of our primary considerations is the ability of the power supply to start under loado The power supply is a loaded oscillator and thus overloads may stop oscillation, and further, any load may prevent reliable startingo It is possible, of course, to apply a starting pulse of some type or to energize with reduced load. The first solution involves control complications, while the latter is not dependable and could result in transistor damage. Instead, a simple starting circuit, such as that employing R33 and D6 as seen in Fig. 6.1, has proven very satisfactory. It can be seen from the circuit that the base-to-emitter voltage is that of the battery minus the drop in R33 and the feedback winding Fl and F2. Thus, when the unit is first turned on, both transistors will tend to conduct heavilyo However, the normal unbalance of the circuit components will cause an unbalance of current in P1 and P2. This will introduce different voltages in Fl and F2 which will allow one of the transistors to gain control and cut the other off. The circuit will oscillate just as in the previously explained elementary circuito Once oscillations start, the starting circuit has negligible effect, and since only the base current flows in R33, it is not necessary to remove the starting resistanceo The diode protects the transistor by causing an IR drop in R33 so that full battery voltage is never applied to the transistor. By choosing the number of feedback turns to supply just enough base current for the maximum load and making fine adjustments in R33, it is possible to prevent the transistors from experiencing even transient overloads in either power or currento Thus, this simple starting circuit enables dependable operation and also provides a convenient adjustment of circuit operationo The simplicity of this starting circuit was a primary factor in the choice of the common base circuit, as starting circuits for other configurations were generally more complex. It can be seen by reference to Fig. 6.1 that the primary circuit has the configuration of the common base circuit with the addition of the starting componentso

Two loads must be supplied by the power supply, one steady-state and the other transient and intermittent. The steady-state load requires a B+ of 95 volts at 4 mils and a B- of 45 volts at 0.4 mil. The transient load involves the intermittent recharging of a 100-mfd condenser to 150 volts. Two separate power supplies could have been used but the desired results were obtained in a single supply by utilizing half-wave rectification and supplying each of the two loads on alternate power strokes. Transistor Tl supplies energy for the steady-state requirement and resetting the core following conduction of T2. T2 supplies the energy for the transient load and resets the core following conduction by Tl. This is accomplished by arranging the polarity of the rectifying diodes appropriately. The steady-state load voltage is sufficiently independent of the transient loading. The power supply (see Fig. 6.2) has proven to be a dependable unit. A total of 28 units have been flown and to date no field servicing has been necessary, and no known flight failures have occurred. Its efficiency varies from 20%, when the 100-mfd condenser is charged, to 65% at the start of the recharging cycle. Battery utilization is 85% at the lowest efficiency of the power supply since the amplifier and range-changing circuit's filament load places the largest demand on the battery. 36

, TOROID TRANSFORMER...................:...- '-: i:.::.. Fig. 6.2. Power-supply assembly.

7. 0 MISCELLAEOUS 7T1 EXTERNAL CONTROL External control of the system, including turning on, turning off, range changing, etc., is accomplished through the use of externally applied power and switching. The main power switch, connecting the internal 4-HR1 battery pack (or an external test source) is S2, Fig. 2o4, a cam-actuated microswitch. The bi-directional rotary solenoid (S1) employed for range changing (Section 5,0) carries this cam on its shaft. It can be noted that connections to the positive side of the storage capacitor C6 and the two thyratron plates are available externally, thus enabling external charging of C6 and subsequent discharge through the solenoids to step the rotary solenoid, by externally connecting the appropriate plate lead to groundo In addition to the ceramic hi-meg wafer, the solenoid assembly carries 2 single-throw 12-position wafers, visible in Fig. 51l and shown schematically in Figo 2.4 as SiB and SlC. S1B provides an external indication of the position of the shaft of the rotary solenoids, thus indicating the particular hi-meg resistor in use. Thus, using an external 0-5 voltmeter, 5 volts correspondsto range 5, 4 volts to range 4, etc., and 05 volt to the "off" position, S1C functions to bias the appropriate thyratron grid when the system is in position 1 or 5, preventing possible self-induced switching to the "off" position or past range 5 as discussed at the end of Section 2LO. It is also to be noted that a connection to the -45 volt bus is available, This is employed externally to disable the system during warm-up or turn-off, by applying a sufficiently high negative voltage (approximately -90) to bias every accessible tube grid beyond cut-off. This function is obviously particularly necessary for the thyratrons, and the electrometer which becomes cut off when the cathode potential of V4B goes to zero. Figure 7.1 illustrates the external control circuit, It provides the necessary external power sources either by battery or rectifier supply, and in addition enables monitoring of the pertinent voltages. The terminal indications correspond to those of the system plugo This control system plus the circuit arrangement noted indicate the following procedure for activating the system, assuming the correct setup has been established. 1) Check: a) range position 0,5 volt (M3) d) overbias on (S7) b) reference relay off (S6) e) output negative (Ml) c) battery at 5~8 (or more unloaded) (M2) 58

S3 S4 — 00 5 v I -0 A I ' --— O B ---— O D --— O E --— O F 0 GROUND O SYSTEM OFF 0 SYSTEM ON 0 OUTPUT O COLLECTOR BATTERY 0 REFERENCE RELAY 0 THERMISTOR 0 - 45 VOLTS 1 h1ll + 45 V,, I ',,,.I n11tl 910V m )|,|~ 150 _ Im1 S6 S7 0) O H - J M2 / 0O-IOV D.C_ - - _ "- - '- --- - --- K 0 + 5.8 VOLTS Op-5Vy0- DL 0 RANGE INDICATION S8 - M3 O — -— 0 M 0 + 150 VOLTS 0 N 0 THERMISTOR Fig. 7.1. Schematic diagram of external control circuit.

2) Close S8 for a few seconds to charge "+150" (capacitor C6) 3) Close S4 for a few seconds, stepping the solenoid, and permit warm-up for 60 seconds, Range should indicate 1 volto 4) Remove overbias (S7)o Output should indicate an on-scale value for atmospheric chamber pressure, 5) Check reference level by closing S6, thus energizing RY-1o Output should indicate approximately 2,3 voltso 6) Switching action may be checked by forcing (momentary closure of S4) to range 2 (output goes negative) and observing self-induced return to range Io 7) To turn-off, reapply overbias and then step rotary solenoid to off position by closing S3 momentarilyo 8) Open S5 to prevent meter (Ml) drain on internal feedback line battery, 7,2 BATTERY PACK As noted from Fig. 2o4, the power for normal functioning of the system is obtained from 4 HR1 silver cells. This pack (Fig. 7~2) has sufficient capacity to operate the system for approximately 1 hr after having been fully charged at 0,l amp to a terminal voltage of 2o0 volts per cello Nonmilitary grade cells are employed and generally provide Ool to Oo2 volt higher terminal voltage than military procurement, The higher voltage is particularly desirable for the thyratron heaters. A fully charged set of batteries is normally operated for at least 15 min prior to flight to enable the terminal voltage to drop to its long-term value. For this application, which requires approximately 1o75 amp, the initial voltage is generally about 605 volts and the final value very nearly 508 volts, Experience indicates excellent constancy at this value for 30-40 min. 7.3 CALIBRATION Units prepared in this laboratory are calibrated over the range 10-3 mb to atmospheric pressure using the laboratory vacuum system, The setup utilized is illustrated in Figs~ 7,3 and 7.4o For pressures from 10"3 mb to approximately 5 mb, McLoed gages are employed. For pressures greater than 1 mg, Wallace and Tiernan aneroid gages are used, The vacuum system employs mechanical fore pumps and an oil diffusion pump and is capable of pressures as low as!07 mbo

::::::::::::-,:::::I-i:~:::-li::""':::::::::::::-::::::::::::::::":::::::::~::::':-:-::i8i:::::::::.r:i:j: iii:i:i-:i 44,,11"....................... Fig. 7.2. Exploded view of unit subassemblies. P.,

a:,I i:: i -i f:.:: z -ii-:-:::::'::::: ::::i t..:i 4, -P CQ I a) Ul Cd P2 bD -r-l *4, Cd a, U 0 rd 0 0 0 bD -H CQ 42

Fig 7.4. Calibration setup at vacuum system. 43

The system is equipped with typical accessories for rough pressure indication, protection in the event of glass breakage, air dryer inlets, liquid nitrogen traps to prevent contamination by mercury, and an arrangement visible in Fig. 7,3 to permit ready connection of 1 to 3 ionization-chamber systemso The calibration procedure involves a preliminary pump down of the chambers to effect adequate outgassing. Generally, the units are pumped for 1-2 days prior to calibration although a clean chamber will reach about 10-5 mb in less than 2 hr, A point-by-point calibration is carried out, starting at pressures somewhat less than 10-3 mb, gradually admitting dry air to the system, Five to ten points are taken for each range and plotted point by point as taken so that possible errors can be detected during the process, The systems are operated from a regulated power supply which provides the 5.8 volts. The output is measured to the nearest hundredth of a volt with a digital voltmeter. The values obtained, the reference value, and thermistor resistances are recorded, A complete set of calibration curves of a typical system is shown in Figs. 705 through 7.10e 7.4 CHOICE OF III-MEG RESISTANCE VALUES Proper values of the hi-meg resistances are determined by (a) the desired pressure resolution, (b) the range of currents to be measured, (c) the ZPV, and (d) the switching limits, The relationships between these factors can readily be shown to be: ZPV - ELL Rn = = k (7o41) ZPV - EUL Rn-1 imax = k (742) min ZPV = kEU- ELL (7.43) k - 1 where imax = maximum current to be recorded (eo = 0 v), imin = minimum current to be recorded (eo = 5 v), n = number of resistors (determined by required resolution), EUL = upper switching limit, ELL = lower switching limit, and Rn = n th resistance,

w Q. 400 _ IIe 300 LIX 200 100........ 0 -I 0 2 3 4 5 6 OUTPUT, VOLTS Fig. 7.5. Calibration curve, range 1. 45

100 90 80 70 60 50 w 30 \ 20 (0 I0 0 I 0 Fig. 7.6. 2 3 4 OUTPUT, VOLTS Calibration curve, range 1. 5 6 46

to, (o bD bQ O k o o 0 0 0 ' -r-4 Li_ O l (<D tO troc N 8W '3anSS3Id

I 9 7 6 W (t) Q. C 5 4 3 2 Ix10-" 0 -I 0 12 3 4 5 OUTPUT, VOLTS Fig. 7.8. Calibration curve, range 3. 6

2x I0'- _ 1.9 1,8 -1.6 '.5 1.4 '.3 1.2 X IxlO1', —0-1 () LU__ C, 9 8 7 6 5 4 3 2 I xlO- - -I 0 2 3 OUTPUT, VOLTS Fig. 7.9. Calibration curve, range 5. 49

9 8 7 6 U* D 5 n) u) w cr a. 4 3 2 I xl10o ' —,0 --— l I- I —... — I —. —. --- -1 0 I 2 3 4 OUTPUT, VOLTS Fig. 7.10. Calibration curve, range 5. 5.6

As an example, the resistances employed in units at present are determined as follows: Let EuL ELL n = 5.0 v =0.0 v, =5, = 4 x 10-8 amp = 5 x 10-12 amp imax 1mi min and thus from Eq. and from Eq. (7.42) imax imin k = (7.43), ZPV = 0.8 x 104 = k5 6.o0 0 6 5 x6 5 6. v. To determine the lowest resistance value R1, max R R1 = ZPV - ELL = x0- = 1.5 x 108 ohms. 4 x 10-8 (7.44) Further use of Eq. R1 R2 R3 R4 R5 (7.41) enables evaluation of all values; thus = 1.5 x 108 = 9.0 x 108 = 5.4 x 10s = 5.24 x 1010 = 1.944 x 10 ohms. 51

7.5 CONSTRUCTIONAL ASPECTS The general constructional technique employed is illustrated in Figs. 1.2 and 7.2. "Lamacoid" decks with appropriately located eyelets support and are in turn spaced' by the various components. The subassemblies are joined by an interconnecting cable and supported by 3 struts, slotted to receive the lamacoid disks, which are secured with a bonding agent* at final assembly. The ionization chamber at one end, and an aluminum plate at the other, are similarly secured to the struts. An aluminum can in 2 pieces axially covers the entire assembly, and through the use of O-rings forms an air-tight cover and shield. Foam** is employed at various points for mechanical support, insulation, etc. The aluminum tubing is "keyed" to the ionization-gage body to prevent rotational movement between these components. The presence of the can also prevents torsional distortion of the main assembly which is necessary to maintain proper clearances for, and orientation of, the shaft linking the solenoid assembly and the ceramic wafer. Electrically, two particular precautions have been taken: (a) to separate the transistor-oscillator power supply and the amplifier, the solenoid assembly and range-changing circuit have been interposed, and (b) the plate resistor and plate lead of the cathode-follower output tube have been placed reasonably far from the electrometer grid region, to minimize capacitive coupling and the resultant "motor-boating" which might otherwise occur when the input load resistance exceeds approximately 1011 ohmso Several minor points warrant mentiono a) The main battery pack has been located at one end of the assembly and provided with plug-in terminals to facilitate replacement. This arrangement permits battery-pack replacement when a unit is installed in a nose cone. b) Vacuum-tight connection between the ionization-chamber outer wall and rocket nose cone is accomplished through use of a right- and left-hand threaded brass coupling and soft copper sealing ring, illustrated in Fig. 7.11. A hexagonal passage (1/4 in.) through the threaded coupling permits adequate gas flow and enables assembly by use of an "Allen" wrench from outside the nose cone. c) The units are 16 in. in length, 2-1/4 in. in diameter, and weigh 3-1/4 lb. *Carl H. Biggs Co., Bonding Agent R315, Los Angeles, Calif. E-P-Fome, Hardener H270, Electronic Plastics Corp., Jamaica, N. Yo 52

i::!. '::i':!iE:.:i::i:.:I::I:: ~ R:' i '.. ':..... or — %:!i...........!~ ii;~.;~:.~% ~. i~: ~ ~j 0:: i~z.~.".:.:':i: z:z.":::, '':i':! c) 0..........a ) iiiiiiii::::;::::i~j:i:E::: j:::E:E iEE::., k -I *::~ 3::::: ff;: >: E 5 i:ig gg..: E;;i-..0 i::g i:Ei: _i ':: i~ER::i"E l~:': E iE~i::::EE-EE::ESER:.::EE. E:. ~N.: iii.i. i;:i:::: i:i- Ei.......... i g i -iEi:E i-i-:;.jmS -.: i. ~r.-EEiE i:-..:lEE:E:g:;E,::.....i2E.i_ ' & &: g ai L ili: ili ili:i i 0ii il~i ii20 t i ~ i;: i:0: _.. ~~~~~~~~~~~~~~ff: ~:::-.i5i;:ii|:20..iE~:i::;~~~E..........:E:Ei *;. E- i g 0 i 0 0 i g -i i E E i ii l _* f _ ' if0Ri -i E;i i: i2::::: e + f g;:::: i tgE gii:::0 i::;-:: ig t i::; iE:; i t fi;&.:;.:::: i:;.......40i:.lE500gi:;: -- 2;~i;:i-: iii0 l:00g::i~i~i_....................................:C:. 0: rgl.~:; -.:l~.:.:; ig5i.. lg::::.. - -;f:.':. 'l:..'.E:;gfi-g; ti0;f l 5:i. -iE..E.;........li.iii. 0!:;;iff;...................... f ~l i::Ei gEii:iii -7 -f:iiE~00 f: a | b i Cii ig~~~~~i ii i igii - ~~~i- ii;gir: tr:'~li~~~~~~~~~~~~~~........................................................lijij

7.6 PERFORMANCE HISTORY At the time of writing this report, 33 units, both prototype and operational, have been constructed. Of these, 28 nave been expended in 9 Nike-Cajun and Aerobee rocket flightso Data from several of these flights are being reduced at present and indicate generally satisfactory performance, Final judgment must of course await completely reduced datao The major difficulty encountered in securing proper operation has been associated with the range-changing circuit and in particular with making certain that the thyratrons would be extinguished following each switching function. This problem has been solved, however, and no further difficulty has been encountered, Figure 7012 illustrates normal range-changing operation as experienced during the flight of Nike-Cajun AM 6,38, Other difficulties, relatively minor, have been corrected. For example, the transistors Tu and TD (Fig, 2~4) which switch the thyratrons have enabled adequate stability of switching points, and the "stabistors" which fix the electrometer filament voltage have eliminated a battery which was at least in 1 case responsible for a failure (serial 6-11). Table 7o61 summarizes the disposition and operational results for the 33 units. 54

z 0 0 a. LIJ 03: \J1 2 5 4 3 2.1 5 4 3 2 1 Ram unit Conewoll 2 aI 1 1 1 I mpct 11H3I 5i! O 20 40 60 80 10 10 40 601820 2024026 20 0032I0I34 36 38 --- ATI I<XIIII...... i:-E:=:z==z==zzr..................= 5 4 3 2 I 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 TIME, SECONDS Fig. 7.12. Range position vs. time for 3 units on Nike-Cajun AM 6.38.

TABLE 7.61. DISPOSITION OF PRESSURE-MEASUREMENT SYSTEMS Disposition On hand Expended in prototype test flight at WSPCGo August, 1956o IGY rocket AM 6030 O Expended in first Fto Churchill rocket, NikeCajun AM 6o31 Expended in first Ft. Churchill Aerobee, AM 2.21 October, 1956 Serial No.* 1 2 3 4 1L' 5-5 9-4 10 6-11 7-10 8-9 11-6 12-12 radioactive source Remarks Prototype unit, used on sled test. Amplifier and range-changing circuit failure during boost phaseo Improper output, Range-changing apparently operated satisfactorily, Operation appears satisfactory to 50-km altitude when amplifier apparently failedo Range-changing circuit continued to function (output above 5 volts) and ultimately switched unit off. Satisfactory operation, Satisfactory operation. Satisfactory operation.I Electrometer filament battery failure at launch. Unit turned itself off during flighto Satisfactory operation. Satisfactory operation. Satisfactory operation, Satisfactory operation. *Second of double number indicates serial numbero

TABLE 7.61 (Continued) Disposition Serial No.* Remarks On hand On hand Expended in AM 6.32 July, 1957 13 14 These two units were constructed like serial numbers 2-12 but were reworked as prototypes to incorporate many changes, including battery pack relocation, plugin gage and hi-megs, elimination of zero adjust motor, addition of trimpots for switching limit adjustment, physical rearrangement of some components, replacement of chamber with new H3 source chamber, modified rangechanging circuit, etc [ 17 Satisfactory operation of all units for duration of limited flight. Second stage ignition did not occur. No significant data. 19 \J1 -W3 20 16-211 18-214 21-210 22 Satisfactory operation of No. 16-211 only. Rangechanging circuits of other four units failed during flight. Common difficulty was continued firing of thyratrons after first or second range change. Reason not established but believed due to excessive heat or physical distortion of units due to nose cone expansion. Expended in Aerobee AM 4.01 September, 1957 23-212 Flown in AM 6.34 December, 1957 F 15-203 24-215 Rocket broke up near end of boost, All units were recovered at impact in moderately undamaged condition. Operation satisfactory for brief flight. Repaired and reflown in AM 6037. 25-207 *Second of double number indicates radioactive source serial number.

TABLE 7.61. (Concluded) Disposition Serial Noo* Remarks Expended in Nike-Cajun AM 6.36 January, 1958 L Expended in Nike-Cajun AM 6.37 February, 1958 27 28 29 15 24-207R 25-215 26-13483 32-13481 33-13479 Satisfactory operation. Satisfactory operationo Satisfactory operation. Satisfactory operation of all three units. (These units recovered from impact of AM 6.34 for re-use.) Satisfactory operationo Satisfactory operation. Nearly satisfactory operation. Upper switching point moved to higher value causing loss of data between 0-2 volts only. Reason unknown. Expended in Nike-Cajun AM 6.38 March, 1958 On hand. On hando To be flown. To be flown. 30-13482 31 Second of double number indiLcates rad.ioactive s(ource serial numbero

8.0 CONCLUSIONS This report describes an instrument which in the opinion of the authors offers a desirable and reliable means of measuring rocket-surface pressures. Many units have been successfully employed in several rocket flights. Use has, however, suggested the possible desirability of additional modifications which would simplify construction and check-out to some degree, and which would at the same time contribute to the reliability. It would be desirable, for example, to simplify the complex feedback environment under which the amplifier operates at present, however satisfactory. To this end, regulation of the power supply (using transistors) would likely be helpful and might make possible a net saving in the number of components and volume the unit occupies, by eliminating at least part of the present need for filtering. Secondly, the addition of another winding of a few turns, with rectifier and Zener diode regulator, might make possible the elimination of the batteries in the feedback loopo Thirdly, it is now felt that the constancy of battery-pack voltage and hence heater voltage make the Miller compensator unnecessary. This change combined with the first might make the elimination of at least 1 tube of the amplifier possible without sacrifice of loop gain and hence stability. Further use will certainly make evident the desirability of additional modifications. However, the system in its present form is more than adequate for the measurements like those performed by this research group. 59

9.0 ACKNOWLEDGMENT The development of this system has taken place gradually with use over a period of many years, as noted in the introduction to this report. Naturally, many individuals have contributed to the effort, in varying degrees. Although the present equipment bears only a modest resemblance to the model first em.ployed in a rocket by this research group (1949), the general outline of the system remains unchanged. H. Fo Schulte carried out much of the early development following adaptation of J. R. Downing's system, and contributed substantially to the early flight successes of the original equipments. H, S. Sicinski first noted and carried out some exploratory measurements of the ionization-chamber hysteresis effect. Several others, primarily student technicians, assisted in development of the early modelso Realization of the present configuration, has likewise involved several individuals. Major contributors to circuit development, physical layout, etc., have been Wo G. Kartlick, G. Burdette, J. A. Cornell, T. Mo Muller, and Jo Horvatho The authors expressly acknowledge the contributions of these individuals, It is obvious but should be stated that the system reported represents the ideas and efforts of allo 6o

10.0 REFERENCES 1. Sicinski, H. So, Spencer, N. W., and Dow, W. G., "Rocket Measurements of Upper Atmosphere Temperature and Pressure in the 30 to 75 Kilometer Region," Jo Appl. Phys., 25 (February, 1954). 2. The Rocket Panel, "Pressures, Densities, and Temperatures in the Upper Atmosphere," Phys. Revo, 88 (December, 1952)o 3. Spencer, No W., Schulte, Ho F., and Sicinski, H. S., "Rocket Instrumentation for Reliable Upper Atmosphere Temperature Determination," Proc IoRoEo, 42, (July, 1954). 4. El-Moslimany, M. A., and Spencer, N. W., Some Studies on a Radioactive-Ionization-Gage Pressure-Measuring System., Univ. of Micho Eng. Res. Insto Report 2406-5-F, Ann Arbor, December, 1957. 5. Alpert, D., and Buritz, Ro S., "Ultra-high Vacuum II. Limiting Factors on the Attainment of Very Low Pressures," J. Appl. Phys., 25 (February, 1954). 6. Uchrin, G. C., and Taylor, W. 0., "A new self-excited square-wave transistor power oscillator," (Letter to Editor), Proc. I.R.E,, 43 (January, 1955)o 61

'