THE UNIVERSITY OF MICHIGAN ANN ARBOR, MICHIGAN MANUAL FOR PARTICLE TRAJECTORY CALCUL ATOR AND COMPONENT POISSON CELL Electron Physics Laboratory Department of Electrical Engineering By R. J. Martin K. R.'AcCrath J. E. Murray D. R. Terry Approved by: i.[_________ _, J. E. Rowe, Director Electron Physics Laboratory Project 05058 CONTRACT NO. DA- 3 )-09 SC-c)2385 DEPARTMENT OF THE ARMY PLACED BY THE U.S. ARM Y ELECTRONICS RESEARCH AND DEVELOPM TENT LABORATORY FORT MONMOUTH, NEW JERSEY March, 1)64

TABLE OF CONTENTS Page LIST OF ILLUSTRATIONS vi LIST OF DRAWINGS vi i i I. GENERAL INTRODUCTION TO EQUIPMENT I I1 1 ol Application of System I 1 1.2 Introduction to Components of System IA. 1o3 Poisson Cell Complex 1.2 1 4 Computer I c2 1.5 Current Source Console I,3 1 6 Recording x-y Plotter 15 3 II. OPERATION OF SYSTEM I l.1 2.1 Introduc tion II- A1 2~2 Poisson Cell Complex II,3 2o2.1 Introduction 1I 3 2.2,2 Poisson Cell (Graphite Plates)'II,3 2 a2 3 Vacuum Hold-Down Ii ~ 5 2 2 4 Plotter TI 5 2,,3 Computer I.:6 2,4 Current Source Console TI7 7 2 e 41 Introduction II.7 24I-)o2 Current Source Module I s.8 2.4.2a Theory and Operation of Current Source,:IJ 8 2),4,2b Output Impedance II 9 2G4),2c Stability I:I: 1. 2)4c3 Current Source Panel IT.11 2, 9K4 Current Source Console 1 1,11 2.4,4a Console Control Panel II: 12 2,4,4b Console Power Supplies II.14 2 4,5 Addressing System 1 T14 2 5 Recording x-y Plotter II 16 -iii

III. THE USE OF THE SYSTEM IN THE SOLUTIONT OF' TRAJECTORY PROBLEMS I I.1 3.1 Introduction III,1 3 2 The Poisson Cell III j2 3 3 Probe Assembly I 1 8 3.4 Analog Computer III 12 3.5 Solution for Space Charge Distribution II 23 3.6 Solution of a Problem II 33 3,57 Test Problem III356 IV, MAINTENANCE AND SERVICING IV 1 4-.1 Introduction IV.l 4,2 Maintenance'V 1 4, 2,.l Computer and Plotters IV 1 4.2.la Operational Amplif-iers — (Performed Biweekly) 1 V.2.2 1 b Reference Supply, Output Noise Check -\-.( Performed Biweekly) I V: 2 4.-2.lc Power Supplies Check — (Perfiormed Biweekly) -V V3 l. 2,ld Multiplier Checkl- (Performed Biweekly) IV. 4 4.2.le Variplotters —(Performed Biweekly) 4.2.1lf Attenuators - -( Peri ormed Biwee:ly) I.c 4o2,1lg Reference Supplies and Electronic Dini tal Voltme ter — ( Performe d Ml. onthly ) I1 4,2. lh Computer Fan Filters —(Performed Monthly) IVi $.2 li Variplotters - -(Performed Quarterly IlV.$ 4 G2 lj Standard Cells -( Semi annual y) IV? c. 2 2 Poisson Cell Complex Iv.L L 4,2,3 Current Source Console IV (10 l. 3 Servicing.IV ]_( -iv -

Page 4-o301 System as a Whole IV olO 4.3.2 Poisson Cell Complex IV1l 4,33 Computer IV' 1.2 4.304 Current Source Console IVr12 4o 34a Current Source Modules IV. 12 4 3.4b Power Distribution Pa:nel TV 13 4 03.4c Control Panel IV, 14 4o 3 o 4d Filament Supply I v. 14

LIST OF ILLUSTRATIONS Figure Page IIol Impedance Characteristic of Current Source; I.10 IIo2 Plan Diagram of Current Source Elements of Poisson Cell (Top View) II: J15 III.1 Derivation of Laplace's Equation III3 III 2 Simulation of Thin Edge of Wedge-Type Poisson Cell, III o7 III3 Four-Point Probe Geometry II o9 III,4 Five-Point Probe Geometry, III lO III5 Solution of the Ballistics Equations. III 13 IIL6 Generation of Electron Position by Computero IIIl19 IIIi7 Calculation of Space-Charge Density (Laminar Flow) III.27 IIL8 Calculation of Space-Charge Density for Crossing Trajectories III 931 IVli Circuit for Gain and Damping Adjustment of Servo Multipliers Iv, 5 IV 2 Circuit for Alignment of Servo Multipliers. IV 6 IVy3 Probes for Probe Assembly. IV o9 Vol Particle Trajectory Tracer. VeN1l V o2 Flow Diagram of Operation of Trajectory Tracer. V 2 V. 3 Operation of Poisson Cell., V,3 Vo4 Poisson Cell Vacuum Hold-Down. V 4 V e5 Plotter Control Console. VA 5 V 6 Plotter Plugboard o V.6 V~7 Plotter Connector Panel. V. 7 V o.8 View of 205-S Plotter with Control Panel Lid Raisedo V.8 V.9 View of 205-S Plotter with Cover Removed V. 9 -vi -

Figure Page V e10 Interior of the Supporting Stand of the 205-S Plotter,, 10 V.11 Current Source Console, v 1i V,12 Current Source Panel (Front View)~ V.12 V 13 Current Source Panel (Rear View). V 13 Vol4 Voltmeter and Control. Panel of Current Source Console V,14 V ol5 Current Source Console Power Supplies. V li5 Vol6 Unloading Amplifier Attenuators. Vwl6 -vi,

LIST OF DRAWINhGS VI. DRAWINGS OF PARTICLE TRAJECTORY CAL''UJLATOR SYSTEM General VI.1 Schematic of Inter-Unit Wiring~ VI 1 VI. 2 Inter-Unit Wiring - List of Individuual Wiri.ng Connections. VI 02 VIL3 Wiring Diagram of Poisson Cell Probe Switching~ VL.03 VI 4 Wiring Diagram of External Switching of 205-S and 1100-E Variplotters VI,4 VIo5 Wiring Diagram of 1100-E Variplotter Connector Mounted at Base of Vertical Wiring Duct~ VIK5 Poisson Cell Complex VIK6 Schematic Drawing of Current Injection Wiring Between Connector Panel9 Plugboard, and Vacuum Hold-Downo VT o6 VI,7 110 Volt 60 Cycle Power Distribution in 205-S Plotter. VI 7 VIL8 Block Diagram of Component Placemeri.t i.n 205-S Plotter. VEo 8 VI 9 Plotter Control Panel - Schematic Diagram of Coefficient Setting Potentiometer,!.T,9 VI. 10 Plotter Control Panel - Wiri.ng of Mode Control Switching5 V 0O VI.11 Plotter Control Panel - Wiring of Indicator Lights o VI. 1. VI.12 Plotter - Utility Reference Voltage Sour/ce for Poisson Cell. VI 1.2 Computer Console VI.13 Computer Roadmap of Two-Dimensional Ballistics Problems'V' l13 VI 14 Computer Console Wiring of Inter-Unit Cable. rVI 1.4 VIL15 Wiring of Unloading Amnplifier AttenLuators. V II" 15 -viii

Page VI.16 Wiring Changes in Computer Mode Control Panel. VI.16 VI.17 Additions of Components at Back of Computer Patch Bay. VI 17 Current Source Console VI.18 Schematic Diagram of Current Source Console (Front View). VI 18 VI.19 Schematic Diagram of Current Source Console (Rear View). VI,1 VI.20 Schematic of Console Wiring. VIo20 VI.21 Schematic Diagram of' Intra-Console Terminal Board. VI 21 VI.22 Wiring Diagram of Current Source Module. VI.22 VI.23 Wiring Diagram of Control Panel. Vi 2!' VI.24 Schematic of Power Distribution Panel. VI,24 VI.25 Wiring Diagram of Constant Voltage Filament Transformer Supply. VI.25 -ix

MANUAL FOR PARTICLE TRAJECTORY CALCULATOR AND COMPONENT POISSON CELL I. GENERAL INTRODUCTION TO EQUIPMENT 1.1 Application of System The particle trajectory tracer provides a means for'solution of many boundary value problems involving the Poisson equation. Basically, the solutions are obtained by analog techniques. Solutions may be obtained for two-dimensional rectangular and cylindrical geometries. Though useful for many other applicationsl, this manual will cover its use in electron beam studies. This section will serve to give the reader a general introduction to the equipment. Section II concerns the physical operation of the system and Section III will describe the use of7 the system in the solution of trajectory problems. Section IV discusses the maintenance and trouble shooting of the system, while Section V consists of figures illustrating the physical aspects of the system. Finally, Section VI contains the drawings necessary for the servicing of the entire system, 1.2 Introduction to Components of System A general view of the particle trajectory tracer is shown in Fig. V.1. The flow diagram and the interrelationship of the various components is given in Fig. V.2. These major components are the Poisson cell complex, analog computer, current source console, and the recording 1. Martin, R. J., Masnari, N. A. and Rowe, J. E., "Analog Representation of Poisson's Equation in Two Dimensions", IRE Trans. on Electronic Computers, vol. EC-9, No. 4; December, 1960.

-I G2x-y plotter. A general introduction to each of these major oomponen ts is given below. 1,3 Poisson Cell Complex This unit (Figs V.3 through V 10) consists of a Poisson cell mounted on an Electronic Associates 205-S x-y plotter, The cell is the means by which an analog of the problem is set up and upon which. the actual solution of the problem occurs. The plotter serves as a support for the cell as well as providing a means for obtaining the electric fields from the analog. The Poisson cell is mounted on a vacuum hold-down unit which contains 2688 spring-loaded plungers. Currents are injected through these plungers and into the cell and this provides the method of simulating the effect;s of space charge in kthe analog. The electric field measuring probe is mounted on* the servo driven arm of the plotter. In this manner, the probe provides readings of the electric field on the analog that are fed into the computer The -two positional outputs of the computer are fed back to the plotter and drive the probe causing it to move over the cell analogous to the motion of an electron in an actual device. 104 Computer The computer is a modified Electronic Associates 231-R computer (see computer manuals). The computer is used to solve the twodimensional ballistic equations for an electron moving in a give'r field configuration. The inputs to the computer are the voltages that. simulate electric fields from the Poisson cello The output voltages from the computer are the two-dimensional coordinates of the position

of an electron. The two ballistic equations and an error check circuit are "patched" into the computer by means of the computer patch panel; This one wired patch panel board will serve both rectangular and cylindrical geometries, The error check circuit is an energy comparitor which compares the energy equivalent of velocities as determined in the computer with the positional potential of the probe on the Poisson cell. 1,5 Current Source Console Various views of the current source console are shown in Figs. V,11 through V.15, In this unit are mounted sixteen hundred separate high-impedance current generators. Each of these generators, called current sources, are used to inject space-charge simulating currents into discrete regions of the Poisson cell. It is by this means that the effect of space charge can be represented in the solution of electron gun problems. The input to the current source console is the setting of the current sources. These values are determined from calculations, The console output is the space-charge simulating currents which are injected into the Poisson cell. 1,6 Recording xy Plotter The Electronic Associates series 1100-E x-y recorder is utilized to give a permanent record of the trajectories of electrons (Figs. V 1 and V,2), From this record the space-charge distribution is determined that will be injected into the Poisson cell during the solution of a problem. Also, a record is obtained of the completed solution trajectories. The inputs to the recorder are the two-coordinate electron

position outputs from the computers i,e,, the same in.puts t.h.at. go to the electron simulating probe of the Poisson cell complex

-II olII. OPERATI ON OF SYSTEM 2,1 Introduction In general, this system is used to solve two types of problems, The first is the analysis type, i.e., given a desired set of boundary conditions (electrode geometry), find the electron beam configuration. The second is the synthesis type, i e,, given a desired beam configuration, find a practical electrode structure that will produce the desired beam, In either case, one starts with a given or assumed set o:f boundary conditions and wants to find a beam configuration, The solution of the problem is obtained by analog methods, that is, an analog is created whose variables will follow the same laws and satisfy the same boundary conditions as those of the desired problem, The desired solution is then. arrived at by conye-rsion of variables between the actual and analog problem after the analog solution is obtained, In the trajectory tracer, the analog of the problem is set up on the Poisson cell; This is a volume conducting material which when boundary conditions are established with electrodes sets up fields that satisfy the Laplace equation, The Poisson equation is satisfied by adding in space-charge simulating currents through the base of the cell (see Fig. Vi4), The potential and electric field distribution can be determined by probing the surface of the cell. The operation of the system can be seen in the flow diagram, Figl V 2. There are two closed loopse The first is the trajectory tracing loop which ties together the Poisson cell and the computer, and the second is the space-charge injection loops This latter loop is the outer one in Fig, Vb2 and includes all parts of the system as well as

-II 2human calculation and adjustment of the space-charge simula4ing current sources o First, it should be noted that if the space-charge injection loop is not utilized one obtains solutions to the problem in which the spacecharge effects have been disregarded, To solve a problem, one first sets up'the boundary conditions, i 0e,.,' the electrode configuration is set up in the Poisson cell. The Laplace equation is now satisfied, To solve for the beam. configuration, the trajectory tracing loop is utilized, The probe (see Fig, V o3) is the analog of an electronl; it reads the electric fields analogous to those which an electron would see, and because the computer is programmed -to move the probe ballistically as an electron, the probe follows a path analogous to an electron trajectory0 This is simply what the trajectory loop does —it traces trajectories of individual elect;rons. o.en one has traced a representative number of electron trajectories consistent wi th the boundary conditions, th;e space-charge-free solution is obtained. In. order to determine the space-charge solu.ti.onr one utilizes the space-charge loop> After the space charge-free solution. is obtained (recorded upon the 100-E Variplotter), the resul.t,ing space.-charge distribution is computed and set into the Poisson cell,o EHo'.wever, this chanLges t:he electric fiel.d distributiion of the problem ard thus the electron trajectories. Therefore, the beam trajectories are again taken and the space charge recomputed. If the loop is repeated ofte:n enough (generally two or *three -times), there will be no change i'n beam configuration or space-charge distribution and thus a self-consistent soLuti on has been obtained0

-II.3In this section, the operation of various parts of the system will'be described. 2.2 Poisson Cell Complex 2.2o1 Introduction. The Poisson Cell is mounted upon a 205-S Variplotter which serves as an adjustable support for the cell. The plotter also contains the mechanism for moving the probe that serves as input to the computer, The cell is a volume-conducting slab which is supported upon a vacuum hold-down. The injection of currents into the slab through spring-loaded plungers in the hold-down allows solutions of the Poisson equation. Various parts of the Poisson cell complex are discussed below., 2.2.2 Poisson Cell (Graphite Plates). The volume-conducting plates are made of graphite and Fydrostone, a plaster binder. The conductivity of these plates is due to the graphite which is mixed homogeneously with the Hydrostone. Various degrees of conductivity can be obt;ained by varying the ratio of Hydrostone to graphite in each mixture.'IThe present percentages being used are 24.5 percent graphite, 75,5 percent Hydrostone. This combination produces a plate with a conductivity of approximately 2.5 x 10-4 mhos/cm, I:n preparing the mixture for molding, a definite procedure is followed. The powdered mixture of graphite and Hydrostone is ball-milled for a 24-hour period previous to adding any liquid to i't, The purpose of this is to produce a homogeneous mixture, to separate any agglomeration, of particles, and generally to improve the mixture's texture. This mixture is then removed from the mill and sifted into a water sol.ution. This solution contains a negative catalyst, sodium citrate, which prolongs the set-up time of the mixture. The actual mixing can be done either mechanically or manually

Next the mixture is put into an air-tight vessel and the pressure in this vessel reduced by 25 feet of water, Thie purpose of ~this i.s to remove air bubbles trapped in the m.i.xt;ure duri.g mixing~ T.e rixture is then poured into a mold, leveled, and allowed to set, When i t has set enough to move, it is put into a humidity-controlled drying room to dry thoroughly O After remaining in the drying room for approximately a week, the pl.ate is tested to see if it has the conductive and linear qialities needed for use in the trajectory tracer system., This tes'ting procedure involves various steps. First, the plate is trimmed and electrodes are silver painted on the edglesof the plate which run parallel to *the proposed electron gun axial direction, Second, the conductivity of the plate is determined by taking a resistance measurement between the painted ends (i.e., a planar or cylindrical diode-,type measurement); it can. then be calculated from the dimensions of the plate, Finally, equipotentials are probed out on. the plate's'top and bottom surfaces, Thi s is done by applying + 100 volts to one of the painted electxrodes and grounding the othero Then the 10, 20, 30, etco, volt equipctentials are probed out and recorded, If the plate has the desired conductivity and th.e poten-tial. distribution does not vary more than +~ pereent f rom the theoretical distribution, the pl.ate i s accep tabl.e for u se I..e acceptable plates are trimmed down'to 21-!i/4 x.18-1/4 inches in p:reparation for silver imprinting cf the base of the plate, The purpose of silver imprinting (see Fig. V4) is'to assuere a uniform area of current injection. into the plate when it is placed on t;he vacuum hold-downo This imprinting is accompl.i.shed by use of a template and a spraying technique o

-II 52o2.3 Vacuum Hold-Down, Use of the vacuum hold-down provides a highly flexible method of injecting space-charge simulating currents into -the graphite plates, To change plates, all that is necessary is'to release the vacuum., remove the plate, insert a new plate and resume the vacuum, When positioning -the new plate care shoul".d be taken to make sure that the p:late is snugly against th.e three positioning posts. The plates should be held in this position until the vacuum takes effect. If the plate draws down too tightly or doesn't draw far enough down, its position can be adjusted by regulating the presstulre via the pressure adjustment valve on the vacuum pump system. The cons'truction of the vacuum hold-down facilitates extreme ease of current injection into the plates. The spring-loaded plungers compensate for any irregularities in. the contact surface of the pla te. In addition, with t;he springs and vacuum there is a nearly uniform distribution of force over both surfaces of the plate. 2.2.4 Plotter, The 205-S Variplotter has bee:n modified'to contain the vacuum ho:l.d-dow:n, various associated space charge simulating equipment, and a, probe assembl.y, As mentioned previously, the vacuum hold-down provides a very flexible means of current injection in-to a graphite platteo The associated equipme:nt, contained in the plotter, such as the connector and plug boards, supplies a compact central locationr from which. currents msy'be directed to the various sources on the plate, and'by which sources may be shorted to establish desired electrode configurations The probe assembly is spring loaded; hence, as it moves along, the probes maintain uniform contact -with the plate, These springs also provide the electrical connectionr between the probes and their wiring.

The controls used on the 205-S Variplotter for operation of the equipment as a trajectory tracer are the power switch, vacuum switch, pen operate switch, parallax potentiometers, scale factors and associated, potentiometers. For detailed information on the func.tioning of these various components reference should be made to pages 14 through l16 of equipment manual on. the Variplotter, Model 205-S and 205-T,which is supplied. by EA.I The probe assembly, using either 4 or 5 probes, permits the measurement of potentials and potential. gradients that are to be continuously taken from the Poisson cell. Generally, however, -the four probe system is used to read the potential gradients from the Poisson cell. For discussion of associated wiring and operation of this system., refer to Section III of this manual~ 2j3 Computer The Electronic Associates 2315R computer has been modified to the extent necessary -to its inclusion in the trajectory t.racer system..It is connected in the system by means of one cable.. Remo.-irlg this cable arnd attaching the shorting cable conrlnector will. a'.l.ow i'ts use'total.ly isolated from the remainder of'the system., As detailed in Section 21. the compu'ter serves,o solve,the ba:..listics equation-. An additional circuit, the energy bala.ncie circuit, aids in checking for error in the trajectory t;racing loop of'the system. Referral to Fig V I.13 and a general. knowledge of the computer are all that are necessary for running trajectory traces, Mounted in the bay above the patchboard on t.he computer is the unloading ampli.fier attenuator panel. These a~ttenuators are connected

in according to the roadmap, Drawing VI.13., The circuit is used to unload the gradient reading probe of the Poisson cell complex, The theory of the operation of the unloading amplifier is in the literature2. To balance these amplifiers, one opens the input;, If the amplifier output voltage rises or falls, the system is un-balanced. One then adjusts the feedback potentiometer (the unloading amplifier attenuators) until the output voltage remains at a fixed potential. The unloading circuit is then properly adjusted 2,4 Current Source Console 2,4.1 Introduction. In the Poisson cell, space charge is simulated by a change in the current flowing through the volume conducting medium., The space charge simulated is i A where p = tube space-charge density, dT = Podsson cell volume element, S K = a constant dependent upon the resistivity of the volumeconducting medium of the cell, i the change in current flowing in the volume element dT~ This is is the current injected into each particular volume element dT through the base of the Poisson cell. 2D Korn, G. A. and Korn, Tv M., Electronic Analog Computers, McGrawHill Book Co, p 317; 1956.

-II.8Each source supplying the current i for each, volume e Lement dTs S S is a separate supply and will. henceforth be called a current source The current sources are constructed in modular form and packaged in panels in units of 32, There are fifty of these panels inr. the current source console. Each panel of sources may'be connected by cable to any 32 unit section called groups of the Poisson cell,'T:I.uus, 1600 current sources can, flexibly be utilized to serve 1600 of the 2688 discrete volume elements (dTs) of the Poisson cello In the following, a single current source module is first discussed. This includes the theory of operation, output impedance, stability, and panel controls. Then, the panel unit is covered followed by a discussion of the source console and its operation. Final1l.y, the system of addressing of space-charge simulating areas of the Poisson cell is discussed, 2'o4 2 Current Source Module 2 o42a Theory and Operation of Currer.nt Source, Figure VI.22 shows the schematic of a current source modules. It consists of a 6661-6EH6 pentode with a large amount of degeneraftive feedback in the cathode circuit -to provide plate current stability. The screen gri.d of the tube is directly connected to its power supply to increase the tube output impedance~ The control grid is maintained at a fixed poten'tial, The current range of approximately 1 to 500 microamperes is adjusted by varying the resistance in the cathode circuit0 To provide the desired sensitivity of adjustment, the current ran-,rge adjustment is accomplished in two stages by a lever switch. The center positio:n of the lever switch opens the cathode circuit~

-I1 9The -tube fil.ament is undervoltaged and is series connected to the heater of a bimetal switch. In this manner, a panel mounted neon light indicates if there is an open tube filament, Operation of the push button connect+'s the par.ticular current source to a digital voltmeter for adjustment of *the space-charge simulating injection current~ The voltmeter is connected across a 10 kilohm precision resistor which is in series with the plate (injecting current) circuit. This push button also contains an. "idiot" circuit which indicates if two or more current sources are simultaneously connected to the voltme'ter The indication is an audible noise and a light on the console power panel. As seen in the picture of the operating controls of a current source, Fig V o12g the lever switch adjusts -the two ranges of the current source with center position off. The push button connects the current source to the voltmeter, The knob adjusts *the injecting current value. The neon. light is nunbered to indicate the current source address and lights either if the push button connects the source to the voltmeter or if there is no cathode heater curreent- 2,4 J2b Out-put Impedance. It is v\ery desirable for this application. that dach current source have an output impedance that is as high. as possible, This minimizes the l.oading effect of one suppl.y upon another. Because one could need, 1600 supplies -that must be adjusted, it can be readily understood that high output impedance is a necessity. As shown in Fig, IIel, the culrrent sources are designed *to have a minimum output impedance of six megohms in the range of 1 to 500 microamperes, This value of impedance is determined on the'basis of a plate voltage variation of 100 vol.ts. This plate voltage variation is

i~~~ 0~~~~~~~~~~~~~~~~~~~~~~ I I I III I I I. bw l l I I__i I I I l l l I I I I1 I Iloc 0 0~~~~~~~~~~~~~~~~~~~~~~ 0~~~~~~~~~~~~~~~~~~~~~~~~ (SYYH093YY ) 3ONV03dY-l _ 0 QZH — H ~~~~~~~~~~~~~~~~~~~~~0 00 ~~~~~~~~~~~~~~($AHOE3A 0 0V~lI

-II o1.greater than any change that should occur due to loading in the trajectory calculator. Thus, the maximum. loading effect upon any current source should be of the order of one percent of the injecting current.Q 2,4,2c Stability, From life tests9 it was determin:ed that stability was better than 2 percent for an operating time of 2000 hours, This was with some select-ion of tubes, In order to obtain good stability it is advisable to age new tubes at maximum pLate dissipation and then select for stability. The failure rate of the tubes utilized in these clurrent sources will be small because the cathode heater is undervoltaged and only small currents are drawn from the tube~ 2~453 Current Source Panel. FiguresV,12 and V,13 show front and back views of a current source panel., A panel cont;ains 32 current source modules num'bered successively from. I to 32. These numbers specify the source numbers of one group (panel) of sources. At the top righbt-harld corner is a numbered amber light, which indicates the group number, This::number may -be changed from 1. to 84'by replacing t.he n.umn.bered iens, Bot:h of the above sets of num'bers (group rum'ber, 1-84 and source:antber:1.1-32) relate to the regio:n of tbhe Poisson cell. to Wbhich,'the current source panel is con:,ectted. The re-plac —:eabl e n umbe r i:n the t1op center of the pane:lel indicates *-he positfiQon of'the panne:. irL the consol.e 2,44 Current Source Console. The current source consol:e (Figs, V.11 l through V.15) consists of seven vertical relay racks connec'ted'toge'ther -to form the co:rnsole The center rack contain.s %the power supplies and controls necessary -to operate the console,, Two current source panels are also included in, this racko The four remai:ning racks

-II.l12each contain eight current source panels. The su tot;al. of source panels is 50,. making a total number of 1.600 current sources~ The center rack contains a voltmeter console control. paneL, and, behind the removable door, the three power supplies that operate the current sources~.Behind the blank panel, e there is al.so a small power supply suppl.ying power to the neon paneln l.ights. Tbhe operati on of the control panel will now be discussed, followed by a description of the operation of -the power supplies. 2,4o4a Console Control Panel. The current sources console is separately corltro:.led from the conso:le contro:l pane:L (Fig. V.14), u.,or.nted on the panel are two push-button switches, lettered "FIL" and "DCT", voltmeter input switch, a panel light., and t;hree fuse holders, The operation and use of each con'trol., are detailed'below. "FIL" Switch.. For ini't;ial warm-up of console, Supplies power to, a. filaments of current sources bo filamen.ts of 200 vol.t s'upply c. blowers d. neon lights power supply e. digital. voltmeter This is a push to operat:e, push to release switchT, The swri.tch is lighted if power is supplied to the filament power supply. It is advisable to allow a half hour warm —up period. "DC" Switch., Applies the d-c power t;o the cllrre.nt sources, Supplies power to: a. 200 V-screen grid supply (Lambda) b 1.5 V co.nt:ro:i.-grid supply (Kepco)

-II 13Again, this is a push-push type switch. When shiutting down,9 be sure to release this switch before releasing the "FIL" swi. tc h Voltmeter Input Switch., This eight-uni't switch cornt'roLs't,-he input to the panel mounted Hewe tt-Packard Digital. Voi.tme'ter: The Digit;al Voltmeter is used:for adjustment of the cLurrenri_ sourcl-es ac,.d monitoring of the console power supplies The eiggLt posit- ion:rs otf lt switch are~ 200 - monitors the output of the 200 volt secreen grid supply 15 -monitors the output of'the 15 vot: scree:n grid seppl{Y FIL - monitors the output of filament -transformer supply +100V - monitors the +100V computer reference supply -100V - monitors the -100V computer reference supply I -connects the digital voltmeter to distribution panel for monitoring of currents from current source LINE - monitors -the 110V 60 cycle output to the console OFF - connects voltmeter input to connectors on front panel, of voltmeters. This is for t-he purpose of utilizirig *the voltmeter in trouble shooti ngv N0TE'lThe fronlt pane-l connectors must noit be shor-ted to the frame for correct. measuremen:.ts in other positions (remove shorting bar), Panel Light This is an. incandescent light called an "idiot" light.: Since pushing the push-button switch of a current source c.ion:.;ects the source output to a common buss hlavring more than onsne curren:t source connected to this buss will cause. errors.an the source currents, An. auxi.liary circuit' therefore, has been incliaded to cause -this "idiot" ligh t to:.ight a:nd an audible alarm tzo sound if two or more current

- -L.1.44source push buttons are in the "on.n" position. In order to deactivate thi.s light and alarm., all of the current source push,-but';toYn switches must be in the "off" position. Fuses. These fuses are associated with the app.l.a.oKatlo of power to control functions within the console itself, 2o404b Console Power Sulies The three power supplies necessary for the operation of the current sources are shown in Fig. V 15 These are mounted i.n the center rack behind a removable dooro The bottom supply is a Larmbda 1o5 ampere, 125-325 volt supply0 This supply, set at 200 volts, supplies the screen grids of the 6661-6BH6 pentode in each of the current sources, The next supply supplies heater current to the filaments of'the pentodes0 It is a constant -voltage transformer with a 250 ampere out put. The supply is adjusted to operate the filaments at below rated current0 The third sulpply, a Kepco, supplies the 15 volts to the pentode control grids0 Note that the control grid co:-nection is positive. 2.4~5 Addessing System. In this trajectory'tracer, bookkeeping is a very important factor~ One must definitely kn:.ow whi.ch current, source is connrected to which discrete region (dT') of the Poisson ceL:l. Also, one must know the positi.ons of t-hese discrete regiolns with: respect, to the coordinate axis of the particular problem~ Agai rL.. in calc:ul.a'ting space-charge densities the particular regions of the Poissorr cell. must be kno'wmo Thus, a common addressing system is utilized' throughout trhe trajectory cal.culator and must'be rigidly adhered to in. doi.:ng any mathema-tica1. calculations~ Figure IIo2 shows a p:1.an of thie Poisson

-II.15I 7 8 14 15 21 22 28 29 35 36 42 43 49 50 56 57 63 64 70 71 77 78 81 84 / \\ Y / / I 8 - G 81-8 z 9 I 1 16 17 24 251 1 1 132 GROUP 81 FIG. II.2 PLAN DIAGRAM OF CURRENT SOURCE ELEMENTS OF POISSON CELL (TOP VIEW).

cell and the spring-loaded plungers that supply the space-charge simulating currents. There are 2688 of these plungers, ieo, there are 2688 discrete regions on the cell. These regions are divided into 84 groups of 32 regions~ Each discrete region is addressed as follows: first, the group number (1.-84) is given and then the number of the region. (1-32) in each gro-up as an example G71-6 indicates the sixth source of the seventy-first group as indicated on the plan. All other places where this kind of addressing is important are also numbered accordin4gyo The connector board and plug board (see Figs- V06 and V,7) are connected in a one-to-one correspondence with. the cella Each is addressed as above with the group numbers marked on the boards. The current sources are connected into the Poisson cell through the connector panel. Since there are only 50 groups of 32 current sources, only 50 of the 84 connectors on the connector board can be connected to current sources, thus the reason for the amber light (group number indicator) on each current source panel, When a current source panel is plugged into a desired group of the Poisson ceil at the connector panel, the lens of the amber panel light is changed to correspond to the proper group number. By this method, the 50 groups of current sources can serve the 84 groups of the Poisson cell and give continuity of addressing. 2o5 RecordiE x-y Plotter The input to the llOOE Variplotter is the electron position output from. the computer. This then serves to record the various trajectories that are plotted upon the Poisson cell by the gradient reading probe0 Initially, the record of these trajectories is utilized to determine the amount of

II 1.7space charge to be injected into the Poisson cell. The final set of trajectories serves as a record of the solution of the problemn

-III.lIII, THE USE OF THE SYSTEM IN THE SOLUTION OF TRAJECTORY PROBLEMS 351 Introduction The integrated components in this system can be divided into two main groups: a nonlinear function generator called a Poisson cell which is used to simulate fields analogous to those in the device under consideration, and an analog computer which uses the information from the nonlinear function generator to determine the ballistic equations for electrons in the device, The flow diagram of the tracer is shown in Fig; V,,2 The nonlinear function generator consists of the following three components, 1. The Poisson cell, a conducting media on which potential gradients satisfy Laplace's equation. in two dimensions. 2, The 205-S x-y plotter which has been modified to carry a probe for determining gradients and potentials on the Poisson cell. 3. The current source module from which currents can. be adjusted and directed to any of the 2688 sources on the Poisson cell making possible the solution of Poisson's equation on the cell, As a whole, the system operates as follows, The probe system. which simulates an electron, reads the gradients from. the Poisson cell and passes the information to the computer; the computer, which is programmed to solve the electron ballistic equations, uses these gradients to determine the relative velocity and direction in which an electron would move when faced with these conditions and relays this information to the pen of the modified 205-S plotter; the probe then moves to the new position as directed by the computer and starts the cycle over again by picking up

-III,2the gradients at this new position0 This process is continuous making it possible to solve for and plot the paths of electrons as they move in the actual device, This section will describe in detail the purpose of each of the various components and illustrate how each can. be used to accomplish its purpose. 302 The Poisson Cell In order to generate the electric-field distribution analogous to that in the actual device, a model must be constructed which has some cornveni niet v:, iab..e Sieisfying Laplace's and Poisson's equations As a model, the function generator uses a conducting medium, called a Poisson cell, on which the electrode geometry of the device is constructed; as a variabhli, it uses the potential on the surface of this Poisson cell, To demonstrate how the potentials satisfy Laplace's equatio:n on the Poisson cell, consider an infini.tesimal element of the cel lI with y and z components of current density passing through it (see Fig.!I.ol)o Continuity requires that -the net current leaving'the surfaces of this volume be zero. Thus (iy + dy dz + i + - dz) dy i dz - i dy = 0. (.31) y Z -z y ~ dy dz + -az dz dy = 0 V jv O = (5.2 )

-III.3aiy y t * + dZ -. —I iz + - dY Iiy FIG. III.1. DERIVATION OF LAPLACE'S EQUATION.

-III 4Writing the currents in terms of the voltage gradients, in thre two directions, and substituting into Eqo 3o2 yields1 av 1 av (303) z R y - [ z O 1 vY' + a = 0 (3 4) R = resistivity of the conducting medium (ohms/meter), Thus the voltages on the surface of the Poisson cell satisfy Laplace's equation and can be scaled to represent the voltages of the actual device in areas free of sources or sinks. In order for the potential distribution to satisfy Poisson.'s equation, sources must be introduced -throughout t;he plate Witbh a source of current i conrected to the bottom of the elementl in p Fig0 III ol, the continuity relation, E q, 3:L, becomes: ( + y d zy) d + dy i dz + i y - i dzy - i dz - i dy = 0 3i 6i +-z = i (35) w az p

iIi o.5Substituting Eq. 353 in Eq. 3.5 yields 2V + 2V - Ri (3.6) Therefore, the potential in a region of the cellH over a source {(or si.nk) of current also satisfies Poisson's equation in t;wo dimensions, establishing the Poisson cell as a model to simulate the conditions in an electron device where space charge is a factor. Thus, the existence of a relationship between the fields AL an e:l.ectro:n device and the potentials on a Poisson cell have been established and it remains only to set up the scaled electrode geometry of the device on *the Poisson cell to complete the model. In order -to set up this geometry, it is necessary to establish equipotentials throughout; the thickness of the cell in positions which correspond to the actual electrode positions in the device under study. There are two methods of e.C stablishing these equipotentials on. the Poisson cell. The first, utilizing surface electrodes and shorting on -the underside of the cell, is the more versatile in that geometry c;hanges can be made without changing plates. However, it does not provide a good representation of the real device near these electrodes because it does not establish a definite equipotential plane through the thickness of the cell.. The second method i.;nvolves drilling holes under the electrodes and pai.nting the holes with conducting paint to better establish this equipotential plane. This method also utilizes the surface electrodes and shorting techniques of the previous one and) thus, creates a more

-ii 6 6accurate analog. The disadvantage of this second method is'that %ny increase in dimensions of the geometry necessitates a new plate, For our use, it has been found that, if the cathode is cut out and this edge painted with conducting paint, the surface electrodes provide a good enough representation elsewhere for crossed-field geometries For axially symmetric geometries, the error ir.ltroduced by using only surface electrodes and underside shortings becomes larger as the radial distance of the electrodes increases because the thickness of the plate increases. In some cases, the geometry or the nearness of trajectories lto electrodes may make it necessary to use the second method. The use of the vacuum hold-down makes the second method more feasible by eliminating the time consuming wiring involved in changing plates on, the older system0 In the case of the wedge, an additional problem is introduced by the finite thickness of the edge. Ideally, the plate should taper down'to an infinitely thin section at the axis, and' o keep the alnalogy between the Poisson cell and the corresponding volume in th.e actual device, a resistance network must be constructed to simulate the resistance of the missing thin edge0 A one-dimensional resist.,ance network, which maintains the missing edge resi stane between'the sources of the last row of the plate, is sufficient for this purpose (see Fig. 11I.2). If R is the volume resistivity of the pla'te, A2 is the axial. length between sources, t is the'thickness of the existi ng thin edge, and r is the projected distance from the existing thin edge to the axis.'The resistance to be connected between the sources is

-III.7r /jx/2 7I toi II FIG. III.2 SUTION OF THIN EDGE OF WEDGE-TYPE POISSON CELL. FIG. III.2 SIMULATION OF THIN EDGE OF WEDGE-TYPE POISSON CELL.

Ta! I Ri i o 8-_ x = A = 12 t @7) 2 1 Since the resistor between the edge of the plate and the first source simulates only half as much volume as the others, this resistance is onl:y half as large as shown in Fig, III2, 3.3 Probe Assembly The 205-S plotter has been modified. to carry a probe assembly from which voltages on the surface of the cell can'be read. Th;ese probes are connected directly to the computer and continuously supply it with the electric field information it needs to solve the ballistics equations for the device being tested. The servomechanisms on the plotter, which control the probe position and velocity, receive instructions from the output of the computer, Connected as such, the probe system simulates an actual electron and traces the path that an, electro:n would take if faced by the field configuration on t-he cel.lo IThe plotter can be used with either a fours-probe or a five-probe assembl.y The probe positions for the two cases are illustrat;ed in, Figs0 III 3 and III o4. With' the four-probe system, the y a:nd z gradie:nts are approximated as follows~ V y_ I +Nf -VV( 38) a- AY 2n A C a6 Adz D az Az 21 VC + VD'A IBV (3 9)

-III. 9A..C =B ID FIG. FOUR-POINT D FIG. III.3 FOUR-POINT PROBE GEOMETRY.

-III.10~ A 2 C 1 2 B D FIG. III.4 FIVE-POINT PROBE GEOMETRY.

-III. 11.For an approximation of the voltage at the center of the assembly, the four-probe voltages are averaged as follows:=4 A B C DL+V I (3.10) The five-probe assembly utilizes the same technique for determining the gradients. In addition, the fifth probe makes it possible to obtain the center voltage exactly and makes possible an approximation of the second-order partial derivatives, The second-order partial derivatives can then be used in Poisson's equation to obtain an approximation of the space-charge density present at the particular position of the probe assembly. av +a2v p ay2 az2 C0 A _ AVy/- 2 (v +vC E (2E - VB - VD) - y Ay ay2 Q72 1/2 i~2 - rVA + VB + VC + VD - 4V 1 1. A a 2V 1 _______1 2 (C+ VD 2VE ) 2 V - V) z Z - 2 Q/2 L1/2 ~/2 2 l L VA + VB +VC VD - 4VE thus

-,III 12 4e P = A v + VB + VC + V 4D o The advantage of 4the five-probe system is thatl it pr~ovides anr independernt met;hod of calcualating the aiouunt of space charge present, at any one place on the analogo This calcul.ation car; n then be compared. with the value used to calculate the injected currents, giving anr idea of the accuracy of the represe.n tationL and. providing a method by which it can be improvedo The distinct disadvantage is that the introduction of the fifth probe requires an increase in probe spa-cing, making the gradient determination less accurate0 Thus,'the four-probe system is always utilized for the calculation of trajectories. 3,o4 Analog Computer With the nonlinear function generator (io.e. Poisso:n cell) set up to generate the gradients in the device, the next step is to set up the computer to use these gradients in solvling the, bal li.shtic problem0, To derive the equations th at + the computer must so.Lve, Consider first a particle traveling in a crossed-field devii:e (the equation from t;his derivationz can be used for e Lec trostitly,.a l:focused devices if -the magnetic field is made zero) as shown i:?r Fig IIIT. 5 0 From the generalized force equation., F = ma = q(E + v x B) and -the simplification of the geometry, B = iB x E = jE + kE Y z

-III.13x 0 B k Z q,m FIG. III SOUTION OF BAISTICS EQUATIONS FIG. III.3 SOLUTION OF THE BALLISTICS EQUATIONS.

L dt2 dt q (E + B v) + (E -B v ) k Thus d2x dt2 dY - [E + B v ] dt2 m Y x z and d2Z q [E - B v ] dt2 m z x y dt2 Since the field is conservative, the electron will also satisfy the following energy equation and it can thus be used to check t-he accuracy of the outputo 1 2 -mv = -qU+ o where U = the tube potential 2q U dz+ 2 + ~ m dt dt o P C0 is dependent upon the initial conditions of the problem. Defi ning, A A dt aB / -_ c p = d2 m c dt2

-III 15The equations to be programmed for the computer are p2y = - N Ey - wc pZ (3,11) y c pEz = - N E + ( Py - (3.12) z c 2 + 2 2 2NU = py + pz (3 13) Next, these equations must be converted to machine equations for use in the computer*2 r3 To simplify 1lbi, 4J, i.vfs. symbols, definitions., and scale factors will be usedc y, z tube coordinates in meters, Ya' za analog coordinates in meters, t tube time, T machine time, P =-d tube time derivative, d P machine time derivative, dT Y, Z% machine variables proportional to y and z,:, Z machine time derivatives of variab:les propor-tional] to py and pz, u tube potentiali u0 maximum tube potent;ial, V analog- potential., V maximum analog potential,. * It is presumed that the reader is familiar with the operation of analog computers, See References 2 and 3, 3. Johnson, C. L., Analog Computer Techniques, McGraw-Hill Book Co,; 19635

-III,16Y ~Ymax. 1 MU fMU Y Y Ymaxax. I meter max, max, where MU = machine units. Z Zmax., 1 MU MU z z -z z meter max. max, b Y ymaxO 1 MU z MU - sec. s IP max. IY max -Z' 1 -M U MU - sec. z Pz Pz (Umeter max. M _s Zs (Magnification of analog relative y z to actual device.) 9 V Vo 1 MU U U U volt 0 0 A T P machine time > t P tube time Equations 3511, 3.12, and 5313 must be written in terms of machine variables in order to be solved. Substituting computer time for tube time yields p2y = E - P (3014) P2z = - E + (315)

2 U + C = (Py)2 + (Pz)2 (5316) 0 The gradients in the y and z directions are obtained from the probe voltages and probe spacings. (AU) Y AYz (AU) E = - z Az In terms of analog quantities, Ay= A( -a ) Za M A ~L Y ~(,M) M. M where ~ is the analog probe spacing,. Then" E = - (AV) E = (a) With these changes, Eqs, 3,14, 3.115 and 3516 become Py - M (=v) c pz (5 17) (V + v) = (Py)2 ~ (Pz)2. (3:19) a4ndl

-III 18The integrators in the computer respond such that an input of Pf produces an output, -f, i.eo PfdT = dT = df = - f + const, Therefore, to obtain an output of Y = a y or Z = a z, *the input y z must be of the form - a Py Or - a Pz, respectively. Similarly, to y z obtain Y = b P or Z = b Pz as outputs, the inputs must be of the form YY z - PY= - bP y, or - PZ= - b P z, respectively, To simplify the calculation and the resulting equations, let b = b = b. Thus Y and Z y z can be easily generated with the circuits as shown in Fig. III.6. To put the equations in their final form, multiply Eqs. 3,17 and 3518 by b and Eq. 3o19 by b. pP2 bNM c Y= kA2 (V)y A Z Defining a bNM k k A2 Co Ac = A v+ z A _Nb2 C = kA2 gives

DaP-~y I I \-DI'y-II I.19YbP y b2 - PY oy y 1/bp bPz - I bPz -a Pz a z -z b z z P'I I - PZ CZ FIG. III.6 GENERATION OF ELECTRON POSITION B;Y COMPUTER.

-III.20bP52 = C (AV)y-C Z, (3.20) 1 2 bP2y = C (A) z+ C Y, (3.21) 1 z 2 c (v + V ) = y2 + 4 (3.22) 3 0 In addition, define C /- ay- I(~ (3.23) 4 - C b (3.24) 5 b Some of the scale factors can be written in a more useful form to simplify the computation of constants C through C o For instance, the time scale factor, A, can be reformulated as follows: T P y PY PY A t P y P (y /M) p a P Ya (PY)max. M (Py ) - a max. Since the field is conservative, (py) ma the maximum tube max. velocity can be related to the maximum tube potential as follows: )max. 0 and Eq. 3.15 becomes M4 I1BUC

-III.21Also, the maximum probe velocity on the analog, (Py )max, is limited by the system and is specified to be some value that the 205-S plotter can easily achieve. In most cases, (Py )ma = 1/40 m/sec. is satisfactory. The velocity scale factor b can also be restated as follows: -1 M b = = M (3.26) (,M )(PYa )maxll max,4 Using Eqs. 3.25 and 3.26 and recalling that k = 1/Uo, C becomes bNM 1 klA2 (PYa)max. C 21 Similarly, 2Nb2 C = k2 = 1, for all cases. Since a and a can be rewritten as y z 1 M a = y Ymax, a t max. a = z z a max. therefore,

-III 22a M PYa max, l 4 b = I M I a4 b ax. a max. and IPz almax. C = 5a Imax. Thus the five constants can be rewritten: 2 PYa max. 2 A C =,1 almax., a max. Pz= a max. a max. In addition to their use in solving the ballistic and energy equations, several of the computer's components are used in the actual measurement of the probe voltages, in forming the gradients, and in averaging the four-probe voltages. (This last circuit is needed only when using the four-probe assembly,) In order to obtain an accurate measurement of the- probe voltage, it is necessary to introduce unloading amplifiers in the probe circuits. These make it possible for the computer to read the voltages without drawing current from the cell, If the amplifiers were not used, current would be drawn through the probes and the high contact resistance

-III.23between the plate and the probes would introduce a significant drop in the voltage at the computer. The gradients and average voltage are computed according to Eqso 3.8, 3I9, and 3.10. The computer roadmap for the furnished wired computer plugboard is shown in Drawing VI.135 3,5 Solution for Space-Chare Distribution Before the space-charge effects existing in the actual device can be simulated in the nonlinear function generator, an exact relationsh i.p must be derived between the current injected into an element on the Poisson cell and the space-charge density at the corresponding point in the actual device. Lastly, a method must be devised by which the space charge at any point in the device can be computed from measurable quantities taken from the Poisson cell, In this section, these two relationships will be derived and then combined to give the final form of the formula for calculating the injection currents in terms of the Poisson cell quantities In the following two derivations, these scaling constants will be used, Vt = kV voltage scaling. t s dt = Mds distance scaling. (3.27) Also, the subscript "t" will refer hereafter to actual quantities in the device under study, whereas "s" will refer to the simulated quantities of the Poisson cell

-II.24Using these relations, the exact relationship between the injection currents and the space-charge density in the device can be derived as follows: i = Js dA A s whereas A is the area of the volume element into which the current will s be injected, and Js is the current density in that volume, Using the Divergence theorem, i = Y JS dT, (3.28) where T is the volume of the element into which i is injected0 To get the integrand in a more useful form, express the current density in the Poisson cell in terms of the field in the cell and the conductivity of the cell, J = a E s s apply the Laplacian operator to both sides, and convert to actual device quantities V ~ J = V - dE s s S S M V, a(l E

-III 25For this application, the field is constant with respect to time and is conservative; thus, it can be derived from a potential in accordance with E = -W V Therefore, M2 V ~ J = MV (- V V s s - t (- vt Vt) = - [[ V2Vt] Use of Poisson's equation gives V2 V = P C 0 o 2 Pt V * J = - - (5.29) s s k (3.29) 0 Substituting Eq. 3.29 into Eq. 3528 yields i = t ( )k ) d T and assuming the integratd:oonstant over the volume k ~ Pt e AT5, (3.30) where AT is the volume of the element into which i is to be injected, Thus, Eq. 3530 allows a simulation current to be calcuiated from the spacecharge density at the point in question in the actual device.

-III 426The next problem to be tackled is the determination of the spacecharge density at any point in the device in terms of Poisson cell quantities. This can be approached from two different angles, both of which utilize data taken from the analog. The first method is based on the following formula: Jt Jt PtV or Pt - (3531) where Jt = the current density at the point in. question in the actual device, = the space-charge density at that point, and v = the velocity of the electrons at that point. In order to determine the current density at a specific point, the electron flow must be assumed laminar so that a relation can be established between the current density at a point and the current density at the cathode. With this assumption, the region between any two electron trajectories can be considered a flux tube and the following relationship existse J = G c (3J32) where G is the ratio of the cross-sectional area of the flux tube at the cathode -to the area at the point in question., and J is the current density at the cathode (see Fig. III 7)~ Since the field is conservative *the velocity of the beam at a point can be related to the potential at that point using conservation of energy as follows:

-III.27POINT FOR WHICH CURRENT INJECTION r dt IS TO BE CAL-_ CULATED - = X TRAJECTORES TUBE dt wt H c d0w~G= _5 dC e ~~dc Wc WHERE w IS THE THICKNESS OF THE PLATE AT c a t FIG. 111.7 CALCULATION OF SPACE-CHARGE DENSITY (LAMINAR FLOW).

1 mv = qV Thus v = 2Vt = 2kV (33) where r is the charge-to-mass ratio of an electron and Vt is the potential at the point in question, Substituting Eqs. 3532 and 3533 into Eq. 3.31 yields GJ P C (3034) s Thus, from Eq. 3530, the injection current is aM2 G J i = 2C At __C ke o 24V k C Gat (3 G35) 1 giwhere M2 J C k3/2 and C is constant for any one particular flux tube v On the Poisson cel., the trajectories correspond to the boundaries between flux tubes, Thus on a plate with constant thickness (a flat). G reduces to the ratio of the trajectory spacing at the cathode tothe

-III.29trajectory spacing at the point in questions On a wedge the thickness is proportional to the radius, thus the radial distance of the point in question is also required to determine G. The volume element, At, is merely the volume into which the source is injecting current; it is constant for a flat plate, a function of the radial distance for a wedge. The voltage over each current source can be picked up by the probe assembly on the 205-S plotter and monitored on the computer's voltmeter. Correct positioning of the 205-S probe can be accomplished by tying the 1100-E plotter in with the 205-S plotter and positioning the 1100-E pen over the source in question on a scaled drawing of the Poisson cell which shows the relative source positions. Calculating the injection currents with this method then involves individual determinations of the voltage, the distance between trajectories, and, in the case of a wedge, the axial distance for each source, and it can be correctly applied only in the region where the flow remains laminar —from the cathode to the first interaction. The second method of computation is based on the fact that the space charge present in any region at any instant is equal to the sum of the individual charge contributions of electrons in the region. Assuming a small but finite volume, this statement can be written in equation form as follows: n = (3~36) jt -- _6) j=1 where qj is the charge contribution madp by the jth electron as it passes through the volume ATt, and n is the number of electrons passing through the volume Further, since

-III 530q = idt i. dt PI r.d 9 (3,37) t = AT -j= t where i. is the current associated with a single electron trajectory, and it is integrated over the time that the electron spends in the volume element. Since AT is small, i. can be assumed constant in the volume; thus n i At Pt = At (337 j=l where At is the length of time spent in the volume element. On the Poisson cell the electron beam is broken up into a finite number of trajectories, the current associated with any one trajectory being determined by the current density at the cathode, Thus, for this application, ii can be interpreted as the total current associated with the jth trajectory that passes through the volume element under consideration on the Poisson cell, and n becomes the number of such trajectories. In terms of analog quantities then, J. A. i. ~ = I j 9 (3p38) j M22 where A. is the Poisson cell cathode area associated with the jth trajectory, Jj is the current density at the cathode for the area Aji and M is the distance scale factor (see Fig, III 8)> Since the field is conservative, At can be rewritten in terms of the voltage at the point as follows,

-III.31Aj = dj wj WHERE w IS THE THICKNESS OF THE PLATE TRAJECTORY |IG. II.CA- UT OFSPACE-CHE / FEVOLUME ELEMENT FIG. III.8 CALCULATION OF SPACE-CHARGE DENSITY FOR CROSSING TRAJECTORIES.

-III 532sst Akt = Vt (3539) s M L- 2ikV where As is the length of the trajectory inside the volume element and Vs is the voltage in the element —both measured on the Poisson cell0 Writing the vol.ume element in terms of analog quantities and substituting Eqs. 3.38, 3539, into Eqo 3537 gives the final expression for the spacecharge density. J.A. AS M3 j=l i2 M M 2kV AT As n S X A.nJ J.' (3.40) S ~jU=. s j=l whe re AT AT= (3.41) t M3 and all the quantities to be measured are analog quantities. Substituting Eq. 3540 into Eq. 3530 gives the injection current: cM2 As i = ~ ATT A. J (3.42a) 0 /AT 2kV J J J s j=l

-III 533 - and = M2 JjAs 3/2o j= J Since laminar flow was not assumed in this derivation, this method can be applied in areas where the trajectories cross, In Eq. 3.42 the first term is constant for the entire problem; further, A. J. will be constant for any one trajectory. Th As \s, the trajectory length in the volume element, and V, the voltage at that element, are the only quantities that need to be determined for each current calculation. The voltage can be read using the computer's voltmeter as outlined above. The trajectory segment can be measured using a plot of the trajectories from the U_100-E plotter superimposed on a scale drawing of the Poisson cell and its sources. It is important to note at this point that the quantities from which the space-charge simulating currents are calculated, using either method, are all. dependent on a particular field configuration on the Poisson cell —the field configuration which existed at the time the data for these calculations was taken. However, this field configuration is changed by the injection of these currents, and although the spacecharge effects caused by these currents improve the analogy between the field on the cell and the field in the actual device, the desired analogy cannot be obtained from one set of calculations but can only be approached through a series of such calculations. 3.6 Solution of a Problem There are three main steps to be followed in analysing a device with this system~ The first step more or less programs the system for

the geometry of the device, and the last two steps involive the detfermination of space charge for the particu.Lar operating conditions being a:na.lyzed0 The first step is to scal.e up the geometry of -the device by some convenient number so t;hat it uti lizes the space on t;he Poissonr cell_ to'best advantage. It is important that the scale factors in the y and z directions be the same because of Poissonrl.s equationv Once the scaling factor is decided upon,. the electrode configuration can be calculated and established on the Poisson cell'by o:ne of the methods mentioned in the previous sec-tion. It is altso necessary to maks a sca:le drawing of the Poisson cell showing'the location of the c ur ret,+ sources and the electrodes for the 1100-E plotter. This drawing is usual.ly one half the scale of the Poisson cell geometry. Also included in, this step is the calculation of t+he po'te:ntiorr.eter settings (constants for the machine equations) for the comp-uter s program. A record should'be kept of these se+ttitngs so that'they can'be checked regularly~ The second step is to simulate the space-charge effect s fcr a given field configuration on the Poisson cello 0 iZt can "be accomplished with'the following sequence of steps. First, trajec t.;ori.es are run for thi s field configuration and are recorded on the 1:7.00-E plot0'ter'o Second, using the recorded trajectories and the scale drawin:g of the Poisson cel.l, e:nough data is taken to calculate +ethe injection currerits'by one of the methods mentioned in the last section, Third, the i:njection currents are calculated and injected into t+thei.r respective volume elements i:n the Poisson cel.l., These three steps will suc.ce;sfull' y accomplish the simulation of the space charge for the origi al,:field

III,.35configLur ation, but, as mentioned earlier:i, th.-i's is not the final solution. since the injection of'these currents change the original field, Step three consists of arriving at the final solution, and it has either one or two degrees of freedom depending on the data available about the emission at the cathode, The easiest situation to analyze arises when -the current distribution at -the cathode is klnown. In this case, the known current distribution can be used in the calculations of step number two and only the final space-charge distribution needs to be determined, This can be arrived at by repeating step number two until two consecutive sets of trajectories are exactly the same. A much harder situation to analyze arises when'the emission is space-charge-limited. In this case, the emission current distribution at the cathode is not known, and'this distribution must be arrived at by trial and error, A distribution at the cathode must be assumed and consistent trajectories obtained as above. The correctness of the assumed distribution must then be checked by observing the potential variation near the cathode. Under space -charge-limited operation, the relationship between current density, voltage, and cathode distance is given by Child's Law, -3/2 - J = 233 x 10-6 -- (amps/m2) (3 -435) X2 Thus, Child's Law will give a current density distribution from -the voltages near the cathode and, if this distribution agrees writh the assumed distribution, the problem is solved I-f not, the assumred distribution is incorrect and must be altered, and the process of

-III.36arriving at consistent charge trajectories must be repeated for this altered distribution. This is repeated until Child's Law is satisfied near the cathode. The number of iterations required for a consistent space-charge solution to a specific cathode current distribution depends on the amount of the beam being analyzed and the accuracy required, Thus in the case of the space-charge-limited problem, a few iterations of the current sources near the cathode is usually enough to get an indication of whether or not the assumed distribution is approximately correctt Needless to say, the work associated with the space-charge-limited problem can be greatly reduced by intelligent estimates of the cathode current distribution. 3.7 Test Problem It is necessary to occasionally run a chelck on the probe system, the plotter, and the computer to be certain that they are operating correctly. This can be done quite easily, utilizing the vacuum holddown, by running space-charge-free trajectories on a crossed-field planar diode. Since the check does not involve the space-charge injection system, a thin surface conductor* can be used on which to set up the planar diode geometry, It is important that the underside of the pla-te wi-th this conducting surface be insulated so that any geometry on the vacuum hold-down has no effect on this test problem. Space-charge-free trajectories are then run in the usual manner and the results checked against the analytic solution, The solutions should agree to within about one Surface conductors can be purchased from Electronic Associates Inc, Long Branch, New Jersey.

-III J 37percent; if they do not; either the computer or the probe system is not functioning correctly, The trouble can be isolated to one system or the other by injecting the electric field gradient into the computer directly with a potentiometer and again running the'trajectories, Agreement of this second solution with the analytic solution would indicate trouble in the probe system and disagreement would indicate trouble in the computer. Once the geometry for this check is set up on a conducting surface and the scaling constants and analytic solution worked out, the check can be run easily and quickly, justifying frequent use,

-IV.1IV. MAINTENANCE AND SERVICING 4,l Introduction A very important factor in the use of this equipment is maintaining it in good operating condition. The system is complex enough; the additional factor of having equipment breakdowns can greatly increase the time required for problem solutions. A technician who is well checked out in the maintenance and servicing of the equipment is a definite asset and, also, continuous maintenance of the equipment will offer definite savings. The section on maintenance of this system has been derived from our (The Electron Physics Laboratory) experience with this system, It is suggested that this procedure and/or the maintenance programs outlined by Electronic Associates and the manuals of the other manufacturers be religiously adhered to in order to use this equipment to the best advantage. The last section on servicing will serve as an introduction to the trouble shooting of the trajectory tracer. The servicing of the commercial items of the system will not be covered, The standard factory manuals should be consulted for work with -these items of e quipment. 4,2 Maintenance 4.2,1 Computer and Plotters. Note: Before any operation of the computer, balance all the amplifiers and reference supplies, Also, check voltage source output values.

-IV 24.21la Operational Amplifiers —(Performed Biweekly). For noise, offset, frequency response, and amplifier output checks, refer to maintenance section of d-c amplifier manual. While making the checks, record values. Each amplifier is checked in turn following the suggested procedures. After completing the day's checks, the troubles should be investigated and corrected. Some of the more common noise faults that may occur are: 1. Excessive ambient noise —common causes are generally in the d-c section. Usually it is caused by the 6U8 and sometimes the 6072. (Note: The majority of the ambient noise is 2900 cpsa Approximately the 13th harmonic of the 94 cps chopper frequency.) 2. Sudden bursts of noise —the usual cause is an output tube, a Tung-Sol 7719 Type (Spec.) (particles flaking off of the cathode). 3. Excessive 60 cycle signal in the output —usually this is caused by heater-cathode leakage near the input end. 4. Square wave noise —the chopper may need adjustment; check the 6AW8 tube. 5. 120 cycle spikes —check the 300 V and other regulatsed power supplies o 6. High microphonic noise —almost always caused by 6U8;- replace it. Sometimes caused by the chopper. Common causes of offset trouble. a. input tube; b. chopper. Insufficient frequency response is generally caused by one or more tubes in the d-c section being below standard. 4.2,lb Reference S;, Output Noise Check —(Performed Biweekly). A lead is connected from the +100 volt reference connection

on the patch board to the y-input of a Cathode Ray Oscilloscope (CRO), The ground of the CRO is connected to the ground terminal on the VTVM panel. The CRO is operated at about 041 msec/cm horizontal sweep frequency and at 5 mv/cm vertical sensitivity0 Now the output of the reference supply may be observed on the CRO by setting the computer in-ot operate mode. The output noise observed on the CRO should not exceed 10 mm p-pt To check for microphonic noise, tap the front of the reference supply. The vertically expanded peaks (microphonic noise) should not exceed 20 my p-p. Also, the entire waveform should not change d-c level. Repeat -the procedure for the -100 volt reference. The sources of trouble are the same as in the operational amplifier. 4,-21lc Power Supplies Check —(Performed Biweekly). Patch the CRO to the external voltmeter terminals on the VIIVM panel. With the computer in operate mode the different voltage supply outputs can be scanned through depressing the corresponding push buttons on the VTVM panel. Observe the amount of ripple voltage present and record it o Note any unusual behavior of the supply. Next, observe the supply voltage on the panel meter, Turn the voltage adjust screw and see if the adjustment operates properly, Repeat *the procedure for each power supply in turn, Common causes of trouble: 1. If voltage does not adjust properly check tubes; 6U8, 5651 and 12Ax7, 2. Excessive noise —check same as above, 3. Plate fuse blows repeatedly, check the capacitors, 4. Voltage goes up and will not come down or large noise bursts, check 6336 (gassy)o

-IV 44.2.ld Multiplier Check —(Performed Biweekly) Gain and Damping —Rough Adjustment. Patch in the circuit in Fig. V1oo Operate the CRO at about 0.5 sec/cmo With the computer in operate mode, switch the input between +10 and -10 and observe the trace on the CROo Adjust the gain and damping for one over-shoot, Repeat for each multiplier. Common causes of trouble: (1) if gain and damping do not operate properly check the tubes; (2) if multiplier operates improperly, check the tubes and also the cathode fuse on the 6500 tube. Follow-Up Gain and Damping, Fine Adjustment, Noise and Cup Alignment. Patch the circuit shown in Fig. IV,1 on the following page. With the computer in initial condition mode, switch S(OO) to the left, S(01) to the center, S(02) to the left, and S(03) tho the left Also, switch the pen lift switch to up, Then put the computer in operate mode. The 1100-E plotter will trace the ground line, Switch the pen lift switch to down. Return the computer to I.C (initial condition) mode and switch S(00) to the right and S(01) to the right. Switch the computer to operate mode and switch -the peni lift switch to up. Now, the variplotter will trace out (SM-MF)4 With the scale factor settings on the variplotter as shown in the circuitry of Fig. IV,2 shown on the next page, the vertical scale is 1 v/in. Adjust'the gain and damping controls so that the (SM-MF) trace is as close to the ground line as possible with minimum jitter. After the gain and damping adjustments have been made, put the computer in I.C, mode. Switch S(00) to the left, S(01) to the center, S (02) to the center Put the computer in operate mode to trace the ground line. Switch the

-IV.5+ I00 - 100 GROUND +100 SM - F CT +F MF - A CT +A MA -B CT +B MB MULTIPLIER SEGMENT OF PATCH PANEL FIG. IV.1 CIRCUIT FOR GAIN AND DAMPING ADJUSTMENT OF SERVO MULTIPLIERS.

-I00 -I(00 I)00 ~GROUND.2.02 _100 +100 1 |-1 lI MA ll l SM S(02) MB | S(03) |(01), aSM_ SE*O SMD L A R L A R L A R I oARM SCALE - I.4C. 75 FACTOR 10 PEN SCALE.75 A A h hFACTOR 10

computer to I.C0 mode. Switch S(00) and S(01) to the right and again put the computer into operate mode. The variplotter will trace (MA-MF) Repeat this for each cup. The following positions of S(02) and S(03) will give: Input to S_(02 S(03 ) Amplifier 2 Left Left SM.Center Left MA Right Left MB Right Center MC Right Right MD Remove lead from MD to Right ME S(03)R and replace it with lead from ME to S(04)R. Common faults are: 1, Bursts of noise on follow-up trace —decrease gain slightly, If this does not help remove the multiplier and look at *the wiper and follow-up cup. The trouble may be due to either or both of these being dirty. Clean with carbon tetrachloride and apply a light film of oil. to the wiper assemnbly, If noise persists in sufficient magnitude, the multiplier should be sent back to the factory for repair. 2. Cups not aligned to follow-up cup (trace going too far above and/or below the ground trace). The cups are difficult to realign, A bad case should be sent back to the factory, 3. (SM-MF) trace too far above ground trace, Check the tubes; one or more are probably low in gain,

4.2.le Variplotters —(Performed Biweekly). Clean the ways on the 205-S and 1100-E Variplotters with methanol. Apply a thin film of light machine oil (spindle oil) to the ways after they are clean. 4.2.1f Attenuators —(Performed Biweekly). Check the pot fuse (continuity) by applying a voltage to it with its associated switch. See if there is any output roltage that can be read using the digital v-oltmeter. 4.2.1g Reference Supplies and Electronic Digital Voltmeter — (Performed Monthly) Check all the tubes in the +100 and -100 volt supplies. Adjust the electronic digital voltmeter. 4.2.lh Computer Fan Filters —(ferformed Monthly). Clean the filters which are located under the computer by removing and vacuuming them. Replace any badly soiled filter, 4.2,li Variplotters —(Performed Quarterly). Check all the tubes in the 1100-E Variplotter. Log these values. Periodically flush the vacuum pump on the 1100-E Variplotter. Add oil to intakes of variplotters. 4.2.1j Standard Cells —(Semiannually)o Change the standard mercury cells in the 205-S and the 1100-E Variplotters and the computer +100 volt reference cells. 4.2.2 Poisson Cell Complex. The probe pick-up is very touchy. Dirt particles will hinder the free sliding motion of the metal pins. This system should be washed with soap and water if the probe does not function properly. Wear of the probe tips will necessitate replacement of these probes, It will be necessary to hand fit these probes. Figure IV.3 contains the information required for manufacturing these probes as well as the probe springs.

-IV.9SCALE: 4X FIG. IV.3 PROBES FOR PROBE ASSEMBLY.

-IV.10Very little other maintenance is required on the Poisson cell complex. As listed previously the mechanical system of the plotters should be maintained as well as the vacuum pump and replacement of the mercury cells. The only other problem spot is the plungers in the vacuum hold-down. These should be protected from dust and dirt at all times. 4.2.3 Current Source Console. Very little is required to maintain this unit, Cleaning of the permanent intake filters and checking of tubes in the Lambda supply should be done periodically. The HewlettPackard digital voltmeter should be maintained according to the factory manual. The current sources should be continuously monitored for maximum injection current output. This will allow for replacement of tubes for deterioration in emission current. 4.3 Servicing 4.3o1 System as a Whole. The inter-unit wiring of the trajectory tracer is shown in Figure IVo.l The inter-unit control cable is the only connection between the computer and the remainder of the system. The material on the connections in this cable is in Figure IV,2. All of the wiring to the Poisson cell complex is in the metal overhead duct. The duct contains a continuation of the inter-unit cable and the fifty 32 wire cables from the current sources. The power input to the current sources console also powers the Poisson cell complex. The 1100-E Variplotter may be incorporated into the system through either of two places. One connection is beneath the floorplate of the computer console. The other connector is at the base of

-IV 11the vertical section of the overhead duct. Except for the means of obtaining 110 volt 60 cycle power, these connectors are in parallel. Facilities have been incorporated for the operation of the computer from the 205-S plotter console. The wiring for this convenience are in Drawings VI.3, VI,4, VIK9, VI.10, VI.11, and VI.16. 4L3.2 Poisson Cell Complex. The circuitry in the Poisson cell complex can be divided into two sections: 1. The space-charge simulating current injection wiring, 2. The power and control wiring. The space-charge current injection wiring is shown schematically in I'igure IV,6. The Poisson cell vacuum hold-down probes, the plug board, and the connector panel are connected together in a one-to-one correspondence, i.e.,, one point on each of the three units is common. The interconnecting wiring is by means of 32 wire cables, This corresponds with the groups of 32 utilized in the common addressing system of the trajectory tracer, All power and control wiring enters the plotter complex at the terminal strip T. B. 16 (see Fig. V.8), The only exception is the 110 volt 60 cycle input which goes to T. B. 5. The standard input connection to the sevo systems of the plotter, as shown in the manufacturer's manual, therefore, comes from T. Bo 16. The additional wiring involves the fact that a system control panel has been incorporated into the plotter console. This control panel includes operation of the analog computer, control of the operation of the plotters, control of the probe pickup voltages, control of reference voltages, and setting of the initial conditions of the trajectories, The wiring diagrams of

-IV.12these circuits are in Drawings VI. 7T through VI.12. Drawings VI13 and VI.4 contain the circuits for the probe switching and plotter control. 43.3. Computer. The computer is a standard Electronic Associates 231-R. The only modifications are those necessary to its incorporation in the trajectory tracer system, Except for the listed changes, one should refer to the factory manuals for servicing of the unit, First, wiring to the inter-unit cable had to be incorporated into the computer. This is listed in Drawing VI.14. Behind the upper-right door above the computer patch-bay a panel unit has been added, The panel contains the feedback potentiometers used in the unloading amplifiers (Drawing VI.15). The remaining additions are components that have been connected into the back of the patch bay within the oven (Drawing VI.17), 4.3.4 Current Source Console 413.4a Current Source Modules. The current source console consists of 1600 individual source modules and the circuitry and power supplies necessary for their operation. The circuit diagram for the module is shown in Drawing VI.22, As shown, the source is a pentode amplifier with degenerative feedback in the cathode circuit., Variation of this resistance controls the plate current, The plate lead is connected into the Poisson cell. The precision resistor is used to measure the injection current by means of switching into the voltmeter circuit. The additional contacts of the switch activate the module neon light and also connect into the "idiot" circuit. The pentode filament is in series with a bimetallic switch which is utilized to indicate a burned-out filament, Individual current sources may be removed from the panel in the following manner:

-IV.131. Remove the potentiometer knob. 2. Remove the potentiometer nut. 3. Remove the lever switch nut. 4. Disconnect the module cable by unplugging the kliptite connectors at the rear of the current source panel. Thirty two of these current sources are mounted together on a reliay rack panel. Each panel is mounted upon slides for accessability for servicing. Complete removal of the panel requires disconnecting the black.nt power cables and the grey current source cable, Except for the current source lead of each source all leads from the 32 modules are connected i.r. parallel. This is done with kliptite connectors at the rear of each. current source panel. Each current source lead connects to the con:nec ori at the rear of the panel and thence to the Poisson cell by means of the grey 32-wire cable. The power and current source cables from each panel then feed into the console ducting (Drawings VI.20 and VIo21). The current source cables then go into the overhead duct to the Poisson cell complex. The black power leads go to the distribution panel and to the filament power supply in the center rack. The black power leads are interrupted at the intra-console terminal boards; This is to allow for ease of transportation by separation of the console into three parts. 4-3.4b Power Distribution Panel. The power distribution panel (see Drawing VI.24) contains the common busses for all power except the filament supply. It also contains the d-c supply for powering the neon lights of the current source modules. The fuse for power into the Poisson cell complex is also located here.

4.3A4c Control Panel Depressing PBS-1 (fil,, on-off) switch (see Drawing VI,23) energizes relay CR19, the time delay relay TDR1 and the filament supply contactor, Energizi.ng the CR1 relay applies power to the console fans, the digital voltmeter) and supplies filament power to the HNV. supply (200v screen supply) and all current sources.. Relay TDR1 is a three minute time delay relay which assures filament warm-up before the plate, screen and bias supplies may be turned on., After the three minute warm-up, PBS2 (d-c on-off) may be depressed to turn on the d-c voltages. Transformer T1 supplies 6.3 VAC to the indicator lamps in the digital voltmeter selector switch, PBS1 and PBS2 indicator lamp voltages are supplied by the main filament supply. Resistors R1 and R2 are voltage dropping resistors because the filament supply voltage is 9 VAC, Also located on the control panel chassis is a simple transistorized audio oscillator. When more than one current source push button at a time is depressed an erroneous digital voltmeter measuremenrt will result; the audio alarm and warning lamp L1 serve to indicate this condition to -the operator. Depressing two or more current source push buttons at one time will ener-gize relay CR4 which energizes relay CR3, Energizing relay CR3 turns on the audio oscillator, warnni.-r lamp, opens the irput connections to -the VrVM, a:nd opens t',h.e +t00v reference supply from the computer, 4,3,4d Filament _SuppL y The current source filament supply, Drawing VI25, is located in the lower center section of the console~ It consists of a line and load regulating transformer with

-IV.15 - fused outputs to the current source panels. Each fused output supplies filament voltage to two current source panels.

-V.1 - FIGURE V.1 Particle Trajectory Tracer In the background left is an Electronic Associates 231-R analog computer. To the right is the current source console. In the foreground left is the Electronic Associates 205-S x-y plotter containing the Poisson cell complex and at the right is the 1100-E recorder. All cabling to the plotter is by way of the overhead duct shown at the top of the figure.

is(y,Z) HUMAN p(yz) I, L Pry z 205 S VARIPLOTTER ANALOG CURRENT p(yz) POISSON SURRENT CELL E COMPUTER SOURCE CONSOLE PROBE POISSON CELL COMPLEX FIG. V.2 FLOW DIAGRAM OF OPERATION OF TRAJECTORY TRACER.

FIGURE V. 3 Operation of Poisson Cell A view of the Poisson cell installed on vacuum hold-down and ready for tracing of trajectories. Placed upon the surface of the cell are electrodes that determine the space-charge-free field configuration; a gradient reading probe is connected to the plotter arm.

-.4FIGURE V. 4 Poisson Cell Vacuum Hold-down A view of the top of the 205-S plotter showing the vacuum hold-down with 2688 spring-loaded plungers; each plunger is an injection point for space-charge simulating currents. In the top background is a Poisson cell showing the bottom side that lays on the plungers.

f.00000g,:iX'l;;00000.-:i.4................. 0-.,.'-.,',. -gg _ _ |............|:::s: pae cotinsiiE: all: contro no *crr or th runn of gZf the trajectories. At left: are six coefficient ettig potentiometers. sitce; te to ee ui~~tch u ppliesreernc votba,'e to the Posso oil',..''' the""'-"-.''.;''"'"" c r s itch controls the p e volt s ad t lo..... s w i.............................. t.a..it p a....................: -'igEiE::E~ C~:::0S~S:':.g:000.00.:.-;" —S.... -.'. "-. -.'............;.'.........:1.- i:::- i::-:E:: -:::: i-:'::.:.:: i.i'-:.-E i::R.':-' -. i:':: E.' il -:E::':E-..........:.:-...:..........E. i - i~::E E:R:-i i I^:: t::;::: fti~~~i;0 00ti0;: 0 Ur;> - 000-i::::-i-::::-: 00:: -;t: — 0000t:::;;:::S tt::E;:t............ -......... " s /j%~~~~~~~~~~~~~~~~~~~...M~~E i;;!: —!:::.i~E. XE.:;.:t-0-00:i~:''.'"'-.::-::.".'SEi......:-...... w I ^ w~~~~~~~~~~~~~~~~~db ^ 4fX? (X/.a X~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.......... O f ~ ~ ~ ~ ~ ~ ~ ~ ~~~~.....tf;0000:0'i~::fVt.......'......!EV:0000000;'S.........-...........,00S00.0000:000i-l0:0000-~;0::00':'00: 111~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...................... _E~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...............................~~~~~~~~~~~ 11 ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~............ == = = 1~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..... - 1=1 I 11~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.................. V.............

-V.6plIotter. FIGURE V. 6 Plotter Pluaboard The plugboard is used for application of external wiring to vacuum hold-down current injection plungers. A resistance networks is at the bottom of the board. The plugboard is located in the base of the 205-S p Lotter.

-V.7"(tiWi,...... FIGURE V.7 Plotter Connector Panel A view of the side of the plotter with doors open showing connector panel. At this panel, cables from the current source console are connected into various areas of the Poisson cell.

-V.8FIGURE v.8 FIGURE V. 8 View of 205-S Plotter With Control Panel Lid Raised This view shows T.B.-16 (the two rows of terminal boards). The majority of the incoming power and control wiring is terminated here.

-v.9FIGURE V.9 View of 205-S Plotter with Cover Removed In this view, the Poisson cell, vacuum hold-down, and its supporting structure are removed. In the center are the cabling connectors that connect to the vacuum hold-down for injection of space-charge simulating currents into the Poisson cell.

-V.10-: rl jj7: FIGURE V.10 Interior of the Supporting Stand of the 205-S Plotter The connector panel and plugboard are at the left and right, respectively.

-V.11 - -v ii-rJ........ FIGURE V. 11 Current Source Console The console contains 50 panels of current sources. In the center racik is shown the voltmeter, console control panel, a sliding shelf, and the door which covers the console power supplies.

-V.12 - FIGURE V.12 Current Source Panel (Front View) A front view of one of the current source panels when pulled out on its slides. There are 32 sources. The panel controls of a source consist of a lever switch, numbered neon light, push button switch, and a potentiometer knob. Note the extra panel light in the upper right-hand corner. This light is numbered and is used in the address system of the trajectory tracer.

-V.13 -............-. I ~!,~...:::'~,.~.. ~ FIGURE V.13 Current Source Panel (Rear View) A rear view of a current source panel installed in the console. The lead in the center foreground is the 52 conductor cable carrying the injection currents to the Poisson cell. The leads entering at the right are the power leads. In the background is shown the individual current source chassis. The foreground shows the kliptite connectors where the individual current sources are connected into the system.

FIGURE V.14 Voltmeter and Control Panel of Current Source Console A view of the center rack of the console showing the digital voltmeter and the control panel. The voltmeter is used in adjusting currents of sources; the control panel allows operation of the current source console separately from the computer control system.

FIGURE V.15 Current Source Console Power Supplies At the bottom is the Lambda 200V supply that serves the screen grids of the current sources; above it the filament constant voltage transformer. Above the transformer is the Kepco supply that supplies the grids of the current sources. Just visible above the sliding shelf is the distribution panel. This distribution panel is mounted behind a blank panel between the console control panel and the sliding shelf.

-V.16FIGURE V.16 Unloading Amplifier Attenuators sA view of the top right section of thoe Electronic Associates analog stability of the probe unloading amplifiers-.

-VI.1~\ CURRENT SOURCE CONSOLE F —u AQ_ OVERHEAD DUCT co /\(CONTAINING CURRENT C I f /\ \ / SOURCE LEADS AND INTER-UNIT CABLE) 01 VACUUM I uj HOLDl I IDOWNI C 205 S VARIPLOTTER 4 0 I p 1100 E PLOTTER CABLE J U (either one of two) T 1100E VARIB E PLOTTER _... Ru'A' 110 V AC 60~OINPUT FOR CURRENT SOURCE CONSOLE AND 205S VARIPLOTTER'B' 110 V AC 60 INPUT FOR COMPUTER ONLY DRAWING VI.1 SCKEMATIC OF INTER-UNIT WIRING.

-VI.2a - WIRE DESTINATION TERMINAL COLOR (CGMPUTER) o Black H.Q. GND Buss v Red +100 Volt Buss u Blue -100 Volt Buss REAR VIEW OF t White B+ Switch on Computer INTER-UNIT WIRING CONNECTOR NO.1 s Wh/Yel B+ Switch on Computer r Wh/Brn B+ Switch on Computer (Back of Computer Console) p Wh/Brnl B+ Switch on Computer p Wh/Blu. B+ Switch on Computer b Blue RP2-J5-20 a White RP2-J5-21 g White MC-J-l P 0 ~ f Purple RP2-J1-37 k White TB2-2(DJ1-52) r j White RP2J5-23 E K C e Gray RP2-J5-22 z Wh/Red TB2-5(DJ1-55) | A F s d h L ~ y Brown TB2-3 (DJ1-53) L y d Pink TB2-14 G t e h Yellow PS6-TP1-21 C M Z k e Red PS6-TP1-8 C ~ M X Wh/Grn PS6-TP1-7 \ H u f ~ D ~ 9g W Black TB2-1(DJ1-51) N a I Red MC-J1-51 I b N Purple RP2-J1-42 \ V D Gray MC-J1-50 C Green MC-J1-5 2 H Gray MC-Jl-15 M Purple RP2-J5 -59 G PurplW e RP2-J5-39 NOTE: When Inter-Unit Cable is DisG White RP2-JS-24 B White RP2-J5-25 connected from Connector No.1 L Orange TB2-6(DJ1-56) the Shorting Plug Must Be Attached to the Connector in F Purple RP2-Jl-j7)5 Order for the Computer to A Brown MC-J1-3 Function Properly. K Wh/Blk TB2-4(DJl-54) E Purple PS6-P Buss J Green TB2-18 DRAWING VI. 2 INTER-UNIT WIRING - LISIT OF INDIVIDUAL WIRING CONNECTIONS.

-VI.2bWIRE DESTINATION TERMINHAL TYPE FUNCTION (COMc6PUTER) 1 Coax Probe A Lead Trunk 13 2 Coax Probe B Lead Trunk 14 3 Coax Probe C Lead Trunk 15 4 Coax Probe D Lead Trunk 16 5 Coax Probe E Lead Trunk 17 6 Coax X-Input to 205-S X-Input on Patch Variplotter Panel 7 Coax Y-Input to 205-S Y-Input on Patch REAR VIEW OF INTER-UNI Variplotter Panel 8 Coax X-Input to llOO-E X-Input on Patch WIRING CONNECTOR NO. 2 Variplotter Panel 9 Coax Y-Input to llO0-E Y-Input on Patch Variplotter Panel 10 Coax Hi Side of Pot. P30 Hi Side of Pot. P30 on Patch Panel O O O O 6 11 16 11 Coax Wiper of Pot. P30 Wiper of Pot. P30 on Patch Panel 0 0 0 0 12 Coax Hi Side of Pot. P31 Hi Side of Pot. P31 2 7 12 17 on Patch Panel 13 Coax Wiper of Pot. P31 Wiper of Pot. P31 O O O O on Patch Panel 3 8 13 18 14 Coax Hi Side of Pot. P52 Hi Side of Pot. P32 on Patch Panel 0 0 0 0 15 Coax Wiper of Pot. P32 Wiper of Pot. P52 4 9 14 19 on Patch Panel 0 0 0 0 16 Coax Hi Side of Pot. P33 Hi Side of Pot. P33 5 10 15 2C on Patch Panel 17 Coax Wiper of Pot. P33 Wiper of Pot. P33 0 0 0 on Patch Panel 18 Coax Hi Side of Pot. P34 Hi Side of Pot. P34 on Patch Panel 19 Coax Wiper of Pot. P34 Wiper of Pot. P34 To + GND Busi on Patch Panel 20 Coax Hi Side of Pot. P35 Hi Side of Pot. P35 on Patch Panel 21 Coax Wiper of Pot. P35 Wiper of Pot. P35 on Patch Panel DRAWING VI.2 (cont'd)

PROBE SWITCH TB 16 PROBE SWITCH ON 205 S ON COMPUTER TERMINAL PLOTTER CONSOLE STRIP ON TB 5 PROBE PROBE PA E CARRIAGE 26 1TO TRUNK1 ASSE~MBLY RED A BLACK 74 23 TO TRL.Q GN RED 75 ~~~~DAIA C. 2 2_4 W ITO TRUNK I RED 76 -- EDE D 0 ~ ncK ~ — TO TRL 25 ~~~~TO TR k 1 TO H.Q. GND NOTE' ALL WIRES ARE SHIELDED DRAWING VI.3 WIRING DIAGRAM OF POISSON CELL PROBE SWITCHING.

-vi.4B+ SWITCH ON COMPUTER TB 16 DATA PANEL 3 CONNECTOR J-11 llOO E )'SVARPLOTTE VARIP LOTTER (DUCT CONNECTOR) B+ SWITCH ON 205S VARIPLOTTER DRAWING VI.4 WIRING DIAGRAM OF EXTERNAL SWITCHING OF 205-S AND 1100-E VARIPLOT~~S.

-VI.5TB 16 41 B+ SWITCH ON COMPUTER X-INPUT ON COMPUTER PATCH PANEL 43 B+ SWITCH ON 205S VARIPLOTTER 19 Y-INPUT ON COMPUTER PATCH PANEL A ~TO:IGND H G To5 B 0 F J 4t TO I IOV AC LINE DD 4,8 1 TO llOV AC. LINE TO H.Q. GND BU "$ DRAWMG VI,5 WI~NG DIAGRAH (~ lJ.~-E ~AIPmR COIEECTOR MOUI~J~DAT BASE (F VECAL DUCT,

-vI.6INTERCONNECTING CABLE FROM VACUUM H'OLD —DOWN TO CONNECTOR BO ARDAND PATCH BOARD. SCHEMATIC REPRESENTATION OF ONE 32 WIRE CABLE FOR CURRENT INJECTION TO POISSON CELL Z/Z7//ji7/// _ mL7~7~7~~// ~ 7~L5L~~77~77/// 1 Z7o~zTY77Z7Z7YofYII ~7~7~7~7~7L~7~7~7~ II ~~7~7~~7~~7~7~~~:/JI DR WI G I. CH MA IC D AW NG OF C RR NT IN E TI N IR N B TW EN CO N CT R AN L P UG OA D A D AC U H LDDN

FLOURESCENT LIGHTS TB 5 PLOTTER DOOR ACTUATED FAN TB 2 100 V 60 " INPUT FROM CURRENT SOURCES 48 CONSOLE 66 Vacuum TO N Vacuum P —-l I1/68 PRIMARY ofROM Cl 69 Power Do TO B+SWITCH ON COMPUTER TI PRIMARY ~o I, IFROM B+ i DRAWING VI.7 110 VOLT 60 CYCLE~ POWER DISTRI~BUTION IN 205-S PLOTTER

ITB 17 IIOQE VARIPLOTTER CONNECTOR I -1 I VACUUM PUMP I I I I I I I NOTE: ITEMS SHOWN WITH 0i I a I Li DASHED LINES ARE LOCATED INSIDE THE'n 101 205S VARIPLOTTER 1 S I ~BASE. 0 Li Li TI L J ITB 14I I 1 TB 16 140 4; 411, TB 164 I TB7 11 j411 TB5 J60 61l TB5 180 1 0jI I TB5 120 21[ TB5 140 K7] CONTROL PANEL END DRAWING VI.8 BLOCK DIAGRAM OF COMPONhT PLACEMENT IN 205-S PLOTTER.

-VI. 9iTB 16 P6 -TPI -21'(YELLOW) T13: - 18 14 (GREEN ) PS6- TPI - 7 (WHITE / GREEN)4 TB2 -1 BLACK) TB22 9 ( WHITE ) T132- 3 8 ~~(BROWN)312 H.Q. GND 44 (BLACK ) +100 V (UNSWITCHED ) 61 (RED)._=_ P31O HI SIDE 2 P31 WIPER 3 P32 HI SIDE 31_ P32 WIPER 32 PS 6-P BUSS 0 (ILT _ TB32 -4 7 ( WHITE /BLK.) T132- 5 6 ( WHITE / RED) TB2 -6 5 ( ORANGE ) _ PS6 -TPI- 8 4 (RED) J —-' P34 HI SIDE 35 0 —___~~~~~~ I' I~ 0~~~~~~~~~~~~~~~~~~~~~~~~~~P35 W~~~~~~~~~~~~~IPE 38 DRAWING~~~~~~~~~~~~~~~~~~~~~~~~ VI9 PLTTERCNUTROLRAE SWICHEAI DIARAMOFCEIINTSTINGALT

D~~IH~ITImS %0~IM00 aIOBI dO D9kIHIM - rIFc rIOUMOO M'.ZMOrIcl O I IIA -DIMVct og - ir -ovq l Zo ig - i r - ovi i 2 09-ir- o 1o O9. L2 -Gr -Z d8 01 L9 i - ir -ov4 o.L 62-Gr- Zd 8 Ol 6 0 XS' (m~3indwoo3) c - ip- ow o~~~~3db~~~~~~~~~~~~ia d~~~~~~~~ ~~~~~~1 8%~~~~~~~~~~~~~~~~~ ~~O-P~ ~ ~ ~e.- _

-VI. 11TB 16 (PLOTTER) PLATE I, 6. 8 Q 50 TO RP2-J5-20 ( COMPUTER ) I ZW 1 -21 49 PRW I2 48 -22 (I -23 47 REF 13 (46 -24 ~zz4 ~-25 45 DRAWING VI.11 PLOTTER CONTROL PANEL - WIRING OF INDICATOR LIGHTS.

TB 17 REFERENCE VOLTAGE TB 16 SWITCH ON 205S VARIPLOTTER CONTROL PANEL I +100 VOLTS UNSWTCHED + 0 61 ______ ______VOLT BUSS - 100 VOLTS 62 UNSWITCHED -100 _________ 62 VOLT BUSS 3 H.Q. GND TO H.Q. GND BUSS DRAWING VI.12 PLOTTER - UTILITY REFERENCE VOLTAGE SOURCE FOR POISSON CELL.

-vI.13BALLISTICS EQUATION UNLOADING AMPLIFIERS 22M Cl/20 Q f 2avz 10.10 (110 ARM) VE T15 vA y 25'O O I +Y IM~~~~~~~~/0 + -i~~~~~~~~~~~~~~~~~~~~~~~~~~ ( I100 ~ 20 PEN) 22 M (0o) Z(O \) ~C 0OS -VArs IM -10 0 -10 0 1 16~ ~ ~ ~ ~~~~~~~~~~~~~~~~~~~~~1 0 -i~~+ 22 M ~~~~~Is 01/ 0 ry I3 1 q~lo o:R )ET et ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~-av, i6 M iO-lO p+ PO ISE 4~~~~~~~~~~:N" +V.V T~~~~~~~o~ C/01 p~~~f Q~ _~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - 2AA N -(VA~I +V. 1,.C4 +VD)~ AR -VE (FIVE POINT PROBE) vc -100 - 100 16 Is~~~~~"~ —— ~~-+' -. 21'~~~~~~~~~~~~~~~V I + SPACE-CHARGE CIRUI ( OIT RBEASEML) A 2 N~~~~~~~~~~~~~~~~~~ - ~~~~~~~~~~~~~~~~~~RDET8VcaAE CIC [~~~,~.v~evcv4-i+' ev -(A+V+V g._~ I -VA 11'0 o~~~~~~o)-~ ~ ~ +~sv I~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~V zo2 2~v [V+C-(V g (FOUR~~~~~~~~~~~~~~~ PEIN) iOE -V FIEPIN RBE V Z(O) ~ ~ ~ V 12'VA-V8 II ~ M. V~V-(CV) S P C - C A R E C I ~ U I 5 O M T P~ O E A S E BQ)2'( VA + VB + VC Z O) 4 Q2'/4 (VA+Vs+VC+V D~~~~4VE) Q P.[~~IM( _ 1YCV-4E ~~~~~DAI V. CMUER RORDA 221W-DMENINLBLITC R

UCMAL- YU NCTIONO DESTIMTTION COLORR TB16-1 205-S Variplotter External B+ Switch on ComPuter'Wh/Blue Switching TB16-2 1205-S Va~riplotter External B+. Switch on Commputer Wh/:Br Swi tchi ng TB16-3 205-S Vaxriplotter External B+ Switch on 205-S Variplotter'Wh/Gry Swi tchi ng -TB16-4c Pot. Buss Output to-D.V.M. XSf6-TP1-8 Red TB16-5 Units Address TB2-6(WlJ-56) Or TB16-6 Units Address TB2-5 (Wl-55) Wh/R TB16-7 Units Address TB2-4(wU1-54) W/l TB16-8 Unitsa Address TB2-3 WX1-53) Br TB16-9 Units Address TB2-2 Wr1-52) Wh mB6-1b P and Q Hundreds Print Code PS6-P Buss Vi TB16-il Pot. Bass to D.V.M. Ps6 -TP1 -7 Wh/Grn TB16-12- Units Address TB2-1 (Wl1-51) Blk TB16 213 Tens Print TB2-14e Pink TB16 -14 Print Signal (Tens Units) TB2-18 Grn TB16 -15 Print Control.(Hundreds) P6-TP1-2l Yel TB16-16 -100 V Reference -100 V Buss B1 TB16-17 +100 V Beference +100 V Buss Red TB16-18 X-Input to 1100-E Variplotter X-Input Location on Patch Panel Coax TB16-19 Y-Inrput to 1100-E Va~riplotter Y-Inrput Location on Patch Panel Coax TB16-20 Y-Input to 205-S Variplotter Y-Input Location on Patch Panel Coax TB16-21 X-Input to:~05-S Variplotter X-Input L~ocationi on Patch Panel Coax TB16-22 Leads for Probe D Trunk 17 W&8 Probe Switch on Computer Coax TB16-23 L~ea~d for Probe B T~runk 14 Via Probe Switch on Computer Coax TB16-24 lead for Probe E Trunk 15 Via Probe Switch on Computer Coax MfB6-25 Lead for Probe C, Trunk 16 Via Probe Switch on Comnputer Coax TB16-'-'6 Lead for Probe A Trunk 13 Via Probe Switch on Computer Coax TB16-27 Hi Side of Pot. P30 Pot. P30 Location on Patch Panel Coax TB16-28 Arm of Pot. P30 P(5t, P30 Location on: Patch Panel Coax TB16-29 Hi Side. of Pot. P31 Pot. P3~1 Location on Patch Panel Coax M6I-3o Arm of Pot. P31 Pot. P31 Location on Patch Panel Coax TB16-31 Hi Side of Pot. P32 Pot. P32 Locationn on Pa~tch Panel Coax TB16-32 AXrp of Pot - P32 Pot. P32 Location on Patch Panrel Coax TB16-33 Hi Side of Pot- P33 Pot - P33 Location on Patch Panel Coax TB16-34 Arm of Pot-P33 Pot. P33 Location on Patch Panel Coax TBI16i35 Hi Side of Pot. P34 Pot. P34 Lroc~ation on Pattch Panel Coax TB16-36 Arm of Pot. ]P34 Pot. P34 L~ocattion on Patch Panel Coax M61-37 Hi Side of Pot. P35 Pot. P35 Location on Phtch Pancel Coax TB16-38 Arm of Pot. P35 Pot. P35 Location on; Patch Panel Coax TB16-39 Arm of Pot. P35 Pot. P35 Loca-ation on Patch Panel Coax TB16-4o ~ GIM TB16-41 1100-E Variplotter External B+ Switch on Computer'Wh/Yel Swi tchi.ng TB16-42 1100-E Variplotter External B+ Switch on Computer Wh Swittching; TB16-43 1L100-E Variplotter External B+ Switch on 205-8 Variplotter Wh Switching TB16 -44 11.. GND H.Q. GND Buss Blk TB16-45 Mode Panrel Re~ference Light RP245-25 Wh

-VI.15I -__________ HI I%..... POT A WIPER F(OO) I *FG(OO) I- I E(02) POT {'B F(02) *-,1- *G(02) _ I E(04) POT C { F(04) I7 1 1- *G(04) - I I ____, _ L _ _ _E(IO) POT D{ F(10) G(IO) I I )I I -. E( 12) IT - -t -r —- I Im (12) I * ( 2| G(12) UNLOADING AMPLIFIER PATCH BAY ATTENUATOR PANEL BEHIND DOORS ABOVE PATCH BAY DRAWING VI.15 WIRING OF UNLOADING AMPLIFIER AS'INUATORS.

TO RP2 - JI-24 (GRAY) REF I3 ( TO RP2-J5-24 s( WHITE) TO RP2-JI-25 (WHITE) T4,TO RP2-J5-25 (WHITE) TO RP2-JI-37 (BLUE) IC TO.RP2 -J5-37 (VIOLET) TO RP2-JI-39 (BLUE) PS TO RP2-J5-39 (VIOLET) The No. 2 terminals on the PS and IC lights were open circuited from MC-Jl-50 on the mode control panel so as to allow a series connection through connector No. 1 (located in rear of computer) to IC and PS lights on 205-S Variplotter. From the No. 2 terminals on these lights the gray wire running from TB16-52 back to the computer completes the circuit to MC-J1-50. Similarly, by connecting the No. 2 terminal of the reference light I3 to the wire from RP2-J5-24 and shorting the removed wire from RP2-J1-25 to the wire from RP2-J5-25, this reference light is in series with the I3 reference light on the 205-S Variplotter, DRAWING VI.16 WIRING CHANGES IN CO1PUTYER MODE CONTROL PANEL.

POT. A POT. B POT E POT.C POT.D E OHI Q OHIQ OHI | E OHI OHIO O O IM 22M o.M |~ 2 M F 0 wo ow 22M F o cm o o W IM r-m2 Z~ w C w'L~22M G OL OO OLO OL O O | G OLO 06 OLD O' O O H 0 0 O O O O O O,0 J O:;g H d O' O O O O O 0 I (00) (01) (02)(03) (04) (05) (10) (11) (12) (13) NOTE THE ABOVE AREAS SHOWNR ARE: PIT WEIE AND CTPE AT TASSIGCED" O CO PATCH PANEL. THE RESISTORS AND CONNECTIONS WERE ADD G MODICATI=ON OF CCUTER. DRAWIG VIE17 ADDITIONS OF COM lbS AT BACK OF COMRJTER PATCH BAY.

S I, I I~~~~~~~~~i s S I Cs I, Is~~ AA BB Icl I F DRAWING vi-18 SCHEMATIC DIAGRAM OF CURRENT SOURCE CONSOLE (FRONT VIEW).

'A' FANS a FILTERS'B' CONNECTORS FOR INTERCONNECTING'l~~~~~-B' CABLE FROM THE C L I COMPUTER'C' FILAMENT CONTACTOR'AI 1:1 IAI | 1 AI (INSIDE ENCLOSURE) _'' X II A D'A' ~~'B''1~D'1~ ~'D' A.C. POWER INPUT

(FRONT VIEW _ B [] B _ B AB B B B 6 B B A ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Al LIII~BB B B LIZ~B B B B INTRA-CONSOLE WIRING TERMINAL BOARD ~ PLASTIC WIRE DUCT DRAWING VI.20 SCHEMATIC OF CONSOLE WIRING.

D FILAMENT CABLES (2 CONDUCTOR SHIELDED) C E SH IELDED FILASHIELDED FILD -TO FILAMENT SUPPLY MENT GROU NDED AT THIS POINT1 (TWO CONDUCTOR SHIELDED CABLE 3 PER TERMINAL BOARD) TO DISTRIBUTION PANEL (8 CONDUCTOR CABLE I PER TERMINAL BOARD) 0T __ 0 TERMINAL CURRENT SOURCE BOARD PANEL NUMBER NO. C D E I 1,2 9,10 17,18 C E 2 3,4 11,12 19,20 3 5,6 13,14 21, 22 CURRENT SOURCE 4 7,8 15,16 23,24 5 25,26 33,34 41,42 D POWER CABLES DPCOWERUCABLES 6 27,28 35,36 43,44 (8 CONDUCTOR) 7 29,30 37,38 45,46 8 31,32 39,40 47, 48 PANELS 49,50 FED DIRECTLY FROM DISTRIBUTION PANEL DRAWING VI.21 SCHEMATIC DIAGRAM OF INTRA-CONSOLE TERMINAL BOARD.

-VI.22-;*110 V DC white/black tracer R 3 39KQ __ _V Vrn + blue Vmwhite/red tracer VI Is 6661 IO KS1 I% 2. w white/brown tracer 7.5Ma a,4 1,<8RELAY brown 7 r ae +200 V DC orng +15 V DC green 4 1 J r, [ FIL red L TS I LOW ~ HIGH RI R2 II 500KQ: 24K 5 2 4 FIL block 17.5 Mo a 800 KQ1 DRAWING VI.22W DIAGRM OF URE-200 V DC NEI white -110 V DC yellow DRAWING VI.22 WIRING DIAGRAM OF CURRENT SOURCE MODULE.

-VI.R23H.V. AU DI A LAMT SECONDARY COTCO IV..FLMN OTEE UPY (YELLOW)(GREEN) 9V.A.C. IIO VA.CA 5A.C A BTt FIL~~~~~~~~~~~~~~~~~~~~~~~ on~~~~~~~~. - off' Rli BLACK ~i _ ~~~~~~~~~~~~~~~~~~~~~DIGITAL VOLTMETER WHITE C PB, SStIF lt~e GREEN D C ~ 5 20, ___L_ ~~~~~~~~~~~on - off PBS 2 R2: TDR I' c5i T cR Tl R~~~~~~~~~~ ~ ~~~~~~~~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 2 6 9l C2 - (-IOV, R 3 1-5K 5w C3-c4 4mFD. 450V ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~, R 4 39K 2v C5 aaD. 6oov PBS1 & PBS2-CAPITOL #PLS-IA~~'5 R 2 1 MG lw 02 PNFT.50. R 9 120K 5w C3-C 11 ACCI. D.P0V T 25A.~~~~~~~~~ ~ ~ ~ ~ ~ ~ ~~ ~~ ~~~~~~~ CONTLINE) R10 1MEG1/2 PO. POTRRWEDT 1 A i.v -ASC I/ — - - __ ___ -11 11... I- JIMA -.'2RIOI ~I- D --

-VI.24I10 V. DC OUT + V.T.V.M. ~ -V.T.V.. 0 0 00 0 0 00 0 0 0INDICATOR LIGHTS -V.T. VM.! o o o o, o'o o o o. o o I /IAO LIGHTS LGT +1:5 V. (110V.DC) ~ I10 V.A.C. INPUT -I00 V.REFHi Q GND. -15 V. + I00 V. FROM~ COMPUTER - 200 V +100 V. TO PLOTTER -200 V1 0 0 V. TO -'10 V TOPLOTTER AUXILIARY AC +110 D.C — I o o o o o o o o o o o I PLOTTER PLOTTER FUSE -I10 D.C. o o o o o o o o o o o RELAY I o o o o o o o o o o LOCATED IN CENTER SECTION OF CURRENT SOURCE CONSOLE DRAWING VI.24 SCHEMATIC OF POWER DISTRIBUTION PANEL.

TO CONTROL PANEL CONTACTOR RELAY CURRENT ALL SOURCE FILAMENT FUSES PANEL CABLE 6A SLO BLO NUMBER NUMBER r- 1-1 l 9 T I -.,T I 2 3 j BLK. 110 VA.C. CRI 5 6 9 INPUT iI 78 12'''O ——'. I' 0 t''-BKO 9 10 2 L____ — ~~~J 111 1 t, I I 2 12 5 RELAY LOCATED IN RED I - 1 8 LOWER HALF, LEFT 3 13 14 8 o2~~1 1516 II PIE SECTION OF 417 8 1 CURRENT SOURCE 519 20 4 CONSOLE. ( SEE DRAW- 21 22 7 ING Z[-19) 23 24 10 RED | | +_o 25' 26 13 H OUTPUTS 27 28 16 I E3 7.0 VAC I 29 30 19 I 4 8.0 VAC 31 32 22 I 5 9.0 VAC - o 33, 34 14 #2 CENTER TAP 35 36 17 250 AMP MAX. 37 38 20 39 ~ 40 23 PARTS LIST 41 42 15 43 44 18 TI FILAMENT TRANSFORMER, RAYTHEON RVO 2000 M2 45 46 21 CI 2 x 10 MFD 1100 VAC 47 48 24 CRI MAGNETIC CONTACTOR, SQUARE D # DO-I 49 50 49 50 TO CONTROL PANEL C.T. (TO CATHODE OF CURRENT kUURCES) COMMON DRAWING VI.25 WIRING DIAGRAM OF CONSTANT VOLTAGE-FILAMENT TRANSFORMER SUPPLY.

UNIVERSITY OF MICHIGAN 3 901 5 03095 01 36111 3 9015 03095 0136