US2765117A

US2765117A - Electronic beam controlled multiplication

Info

Publication number: US2765117A
Application number: US226247A
Authority: US
Inventors: George B Loper
Original assignee: Socony Mobil Oil Co Inc
Current assignee: ExxonMobil Oil Corp
Priority date: 1951-05-14
Filing date: 1951-05-14
Publication date: 1956-10-02
Anticipated expiration: 1973-10-02

Description

Oct. 2, 1956 e. B. LOPER ELECTRONIC BEAM CONTROLLED MULTIPLICATION 3 Sheets-Sheet 1 Filed May 14, 1951 650265 5, LOPEQ IN V EN TOR.

BY 163,4; W

AGENT Oct. 2, 1956 LQPER 2,765,117

ELECTRONIC BEAM CONTROLLED MULTIFLICATION Filed May 14, 1951 3 Sheets-Sheet 2 GEOQGE 5 0, 52

IN VEN TOR.

BYAQWW 3 Sheets-$heet 3 Fla. 9

G. B. LOPER ELECTRONIC BEAM CONTROLLED MULTIPLICATION Oct. 2, 1956 Filed May 14, 1951 650/265 51 LOPEQ INVENTOR.

BY 1&6:

AGENT United States atent ELECTRONIC BEAM CONTROLLED MULTIPLICATION George B. Loper, Dallas, Tex, assignor, by mesne assignments, to Socony Mobil Oil Company, line, a corporation of New York Application May 14, 1951, Serial No. 226,247

8 Claims. (Cl. 235-61) This invention relates to the multiplication of two electrical functions and more particularly to control of electron flow by two electrical functions to produce a third function that is proportional to the product of the first two functions.

There are numerous methods, systems and devices known in the art for performing multiplication. Calculations have been performed automatically by using a combination electrical and mechanical system which requires the insertion into the system of some indicia of the quantities of interest whereupon the system produces an output indication proportional to the sum, difierence, product or quotient of the quantities depending upon the operation set up.

Further, it has been proposed to produce a deflection of an electronic beam onto a pattern of electron collecting plates which, when suitably interconnected produce an output voltage proportional to the product of two voltages suitably applied to the beam control system. Such prior art systems in general have disadvantages inherent in them that prohibit their use in many applications.

It is an object of the present invention to provide for multiplication two electrical quantities in which there is produced a third electrical quantity or indication that is proportional to the desired product.

It is a further object of the invention to provide a system utilizing a single vacuum tube to which the quantities of interest may be applied and from which the product indication may be taken.

A further object of the invention is to provide a network Which includes a vacuum tube in which two electrical quantities control the magnitude of an electron beam for the production of a single output indication proportional to the product of the two electrical quantities.

In accordance with the present invention two electrical quantities are multiplied by establishing a stream of electrons normally flowing symmetrically with respect to a boundary, establishing a field of force along the stream to produce non-symmetrical electron flow with respect to the boundary in sense and magnitude dependent upon a first of the quantities, and at a selected location along the electron stream controlling in equal and opposite senses the magnitudes of the portions of the electron stream on opposite sides of the boundary in response to the second of the electrical quantities. The amount of change of the total electron stream as measured at a point in the direction of electron flow beyond the selected location is directly proportional to the product of the two quantities.

For a more complete understanding of the present invention and for a further description thereof reference may now be had to the following description taken in conjunction with the accompanying drawings in which:

Fig. 1 illustrates electronic multiplication in accordance with the present invention;

Fig. 2 is a diagrammatic representation of a suitable form of tube;

ice

Fig. 3 illustrates the anode characteristic for the tube of Figs. 1 and 2;

Fig. 4 is a schematic diagram of a multiplying circuit;

Fig. 5 is a set of anode current curves illustrating the operation of the present invention;

Fig. 6 is a set of curves similar to Fig. 5 illustrating the effect on the beam of the deflection voltage;

Fig. 7 illustrates the transconductance of the tube as a function of the beam deflection voltage;

Fig. 8 is a modified form of electronic tube;

Fig. 9 is a further modification of the electronic tube;

Fig. 10 is an isometric view of a further modification of the tube shown partially cut away; and

Fig. 11 is an elevation view, partially in section, of the' tube of Fig. 10.

Referring now to Fig. 1 there is illustrated a system for multiplication of two electrical quantities such as the voltages, es and eg. In this system there is produced an output voltage e0 that is directly proportional to the product of en and e The system includes an electron source such as an indirectly heated cathode 1G partially enclosed in a suitable electron beam forming structure 11. An anode 12 connected in circuit with the cathode 10 by Way of conductor 13, an anode resistance 14 and B-battery 15 receives the electron beam 16 which normally occupiesthe volume shown stippled. The pair of

deflecting plates

17 and 18 are positioned on opposite sides of the electron stream 16 at a first point in the electron flow pattern for deflecting the beam in its travel between cathode 10 and anode 12.

A control grid structure is also interposed in the electron beam. The control grid structure includes in first element 20 (shown with vertical grids) and a second element 21 (shown with horizontal grids). The first of the input voltages ed is applied to the

plates

17 and 18 by way of

conductors

25 and 26. The

plates

17 and 18 may be biased to form a suitably shaped beam. In Fig. 1, for example, the

plates

17 and 18 are maintained at a nega tive potential by a battery 27 connected to ground and to the

plates

17 and 18 by

resistances

28 and 29 respectively. Depending upon the form of tube structure and the voltages used, it may be found preferable to maintain

plates

17 and 18 at a positive potential.

The second input voltage a is applied to the

grid elements

20 and 21 by way of

conductors

30 and 31. The

grid elements

20 and 21 are biased to a selected operating point by a battery 32 which is connected to ground and to

elements

29 and 21 by way of

resistances

33 and 34 respectively.

When the voltages ed and 8g are both zero, assume that the beam is so formed and directed that it impinges the a anode 12 in an area that is represented by the dotted outline 35. The

grid elements

20 and 21 are positioned in the electron stream 16 so that one-half the electron stream passes through each of the elements. By applying a voltage ea to the

deflecting plates

17 and 18, the electron beam 16 may be deflected to a new position such that all electrons pass through one or the other of the

elements

20 and 21, depending upon the polarity of the voltage ed applied to the deflecting plates. For example, as plate 17 is made more positive and plate 18 is made more negative than the normal bias, the electron beam is deflected to impinge the anode 12 such as in the area indicated by the dotted outline 36. When such is the case substantially all of the electrons pass through the grid element 20 since pattern 36 is to the left side of the line 37 which normally marks the center of the electron beam 16. Similarly if plate 18 is made positive with respect to plate 17 most of the electrons will pass through the grid element 21 in their travel to anode 12.

In multiplying two voltages, the first voltage ea controls the apportionment of the electrons flowing to or through grids-29 and 21 and the second voltage 8g controls the magnitude of each of the two portions of the electron stream that reaches. the anode 12. Electrons flowing through each of'the grid elements failing to reach;

anode 12 fall onto. and are. collected by auxiliary electrodes which, for purpose of clarity, have been omitted from Fig. 1'.

To illustrate operation of the system, assume. first that voltages ea and eg are zero. With the bias battery 32 of suitable voltage; and the deflection bias set for equal division of the beam through the gridsv 2t and 21 there will be a norrnaianode. current flowing. through conductor 13, and resistance 14. The resultant voltage appearing betwcenthe anode 12;and.-.gr.ound may be. suitably nulled in a measuring circuit as by a battery as connected to ground andto the outputindicator 41. When the voltage at point 42' is the same as the potential of the anode 12, the.voltage e appearing on the meter 41 will be zero.

Assume now that voltage as iszero and a D. C. voltage e has a finite value so that the grid 21 is more positive thannormaland grid. is more negative. The electron beam 16 is symmetrical with respect an imaginary boundary common to center line 37 and'to the axis of cathode 10, and thus impinges grids 2i) and 21 equally. However, due to the eifect of, the voltage ed, more electrons are permitted to flow through the grid 21 while the half of the beam impinging the grid 20 is reduced by an. equal and. opposite amount. As a result the total number of electrons reaching anode 12 remains constant; thus the voltage e0 on indicator 41 remains Zero.

Assume. now that e is zero and that Ed has a finite value. Since grid elements 20 and 21' are at thesame potential, the total number of electrons reaching plate- 12 is the same regardless of which grid the beam passes through and the output voltage e0 is still zero. Thus when either voltage es and e is zero, theoutput voltage e0 is also zero, satisfying one of the requirements of a multiplyingoperation.

If both 6g and ed have finite valuesthere will be an increase or. a decrease in the number of electrons reaching theanode 12 depending upon the relative polarity of the voltages es and a F or example, if plate 17 is'made more positive and plate 13 more negative than normal, and grid It is made more positive with grid 21 more negative than normal, a larger proportion of the beam 16 will.

pass through grid 29, which grid offers less resistance to electronflow thandoes grid 21. In such cases there will bea substantial increase in the totalnumber of electrons:

reaching plate llper. unit time, which will, result in a decrease in the voltage as as measured by meter-41.

The-foregoing illustrates operation with D. C. voltages.

Similar results will be obtained with alternating current voltages. In the latter case output indicator; 41, is an alternating; current device; utilizing or'displaying'theinbattery 49 or an equivalent.

For accurate and reliable multiplication; in; eitherthe A. C. or D. C. case it isrequired -that.the..variation in theapportionmentof electrons between. the two grids be a linear, lunction of the beam deflecting voltage. If the electron pattern is uniform and rectangular asillustrated in Fig. l, the physical deflection. ofbeam: 16 must be a linear function of the deflecting voltage es.. Further, the control. that the rids 2i andZl have over the magni tudcs of the portions ofthebeam 16 that passes therethrough also must be a linear function. of the control voltage 8g. When the above-mentioned apportionment and the grid control are both linear functions of the re spcctive voltages, the variations" in anode current ac: curately represent the product" of the respectivevoltages.

While in Fig. l the tube has been representedas being comprised of asource 16, beam forming-means 11', de-

fleeting plates

17 and 18,

control grids

20 and 21 and anode 12', further structure may be required to produce the linear responses above described. In Fig. 2 there is illustrated a diagrammatic representation of a tube that is suitable and in which additional elements are shown. Where consistent, like parts'have been given the same reference characters as in Fig. 1.

in this modification, control grids 20 and 21- together form a cylindrical grid enclosure around cathode 10. The grids 2t) and 21 are electrically separate but. have? closely adjacent terminuses along lines, at. the. center. of the beam 16 when the beam is in. its normal position. A screen. grid structure formed by an inner mesh 50 and an outer mesh 51 encases the

grids

20 and 21. A suppressor grid 52 is positioned between the. outer screen 51 and the anode 12. The entire structure is enclosed in an evacuated glass chamber formed by the envelope 53. With. such structure operation similar to that normally found in conventional pentode tubes, is provided.

Fig. 3 illustrates the plate current characteristic for; such pentodes, the family of curves representing different: values of control grid bias voltages. It will benotedithati a separation between the various curves is constant for equal increases in bias voltage ezz. Thus the plate current 1 isa linear function of the control grid voltage;

The deflection of the beam as a function of the voltage; ea from the center line 37, Fig. 1,, will, for smallangleg. be a linear function of the deflecting voltage es. If.rela-. tively wide angle deflection is employed, the beam may be; suitably shaped so that the apportionment of the electrons:v between the

grids

20 and 21 is a linear function of thade. fleeting voltage.

Fig. 4 illustrates a more complete circuitdiagramihan. that illustrated in Fig. l but in whichthe measuring circuit has been omitted. In thissystem, the inner. and

outer screen grids

50 and 51 are connected directly together: and to the positive terminal of the battery 15. The sup? pressor grid 52 is connected directly to the cathode. The. system also includes a. circuit foradjusting the. initial apportionment of the electron beam between the two signal grids 2i and 21. The biasing battery. 27 is connected-to. the deflecting plate 17 through a balancing network which. includes a potentiometer 279. whose movable tap is connected through resistor 28 to the deflecting plate=172 Similarly a second potentiometer'27b has itsmovable-tap connected through resistor 29 to the deflecting plate 18. The taps, onthe potentiometers27a and 27b are ganged together for'simultaneous adjustment of the voltage to-: move. the. beam either toward or away from a selected. one of the, grids. When properlyadjusted the electron beam impinges equally on the two signal grids: Themeasuring circuit may be coupled to the anode in the manner illustrated in Fig. 1 for D. C; multiplication, or may be. condenser coupled for A. C. measurement.

Fig; 5. illustrates a set of egIp curves for different values of deflecting voltage, the curves being drawn for an as= sumedv value of biasing voltage e32; When the'deflecting voltagees 0, the grid signal voltage'e may bevaried in either direction without altering the magnitude of" the plate-current 1p. However, if a 4 volt signal signaled is now applied, the voltage on plate-"17 will then be 4' voltsmore positive or negative than the-deflectingplate 18-inwhich case a variation in the grid signal voltagee produces a substantialchange in output'current I 5.

Fig. 6v includes a'similar set. of curveswhere the-plate current Ip is plotted as a function of the beam deflecting, voltage ed for various selected valuesof tllescontrol grid bias voltage cs2. The curve 54 illustrates the operatitmzof the system when a signal voltage 8g; of 6-voltsis applied. to the system whenflthebias, voltage e32 hasa value of. -3 volts. The curve, 542. similarly illustratesv the operation when a signal voltage e of. -6 volts. isapplied. Curves 55 and 5521- illustrate theaoperation, for.v signal voltages of 4 volts and 4 volts respectively when: thebias: voltage e32 has a value of -2 volts.

The functions plotted in Fig. 6 are reciprocal in nature with those plotted in Fig. and from them, it may be seen that electronic multiplication may be performed in the system. When either of the input voltages e or Ed is zero, the output will be zero. When the input functions have finite values and so long as operation is limited to the linear portions of the characteristic curves such as illustrated in Figs. 5 and 6, the variation in the output or plate voltage as a result of application of the input functions will be reliably representative of the product of the two in ut functions.

In Fig. 7 the mutual transconductance, defined as de gm is plotted. It will be apparent from Fig. 7 that positive and negative values of mutual transconductance are obtainable by operation of this system. By this means it is possible to produce output indications that are reliable as to sign as well as magnitude.

In Fig. 8 a modified form of electronic tube is illustrated in which the electron beam 16 is controlled by electrostatic rods 60 and 61. The rods 60 and 61 are positioned normally at the center of the beam 16 and between the cathode and the

control grids

20 and 21. When a negative potential is applied to the rods 60 and 61, each portion of the beam 16 is split into two segments since the electrons are repelled by the negatively charged rods. The two segments 16a and 16b of the electron beam impinge equally on split grid structures similar in nature to that illustrated in Fig. 1. However in this modification four grid elements are required since there are essentially four electron beams. The two grid elements 62 and 63 are electrically common. The grid elements 64 and 65 similarly are electrically common. The voltage e may then be applied in a push-pull manner with one terminal of the voltage source connected to grids 62 and 63 and the other terminal connected to grids 64 and 65. Suitable screen and suppressor grids are added to provide conventional pentode characteristics. When using a tube of this type, the magnitude of the biasing voltage on the rods 60 and 61 will be adjusted so that the beams 16a and 16b initially are equally distributed between the respective signal grids.

In Fig. 9 there is illustrated another type electron tube that may be utilized for electron multiplication. In this system a coil 66 is mounted as to encircle the tube enclosed in the glass envelope 67 with the axis of coil 66 preferably coinciding with that of the tube. Cathode beam 16 is directed to impinge equally the two control or signal grids. In this modification, as in Fig. 8, the control grids are formed by four elements. The grids 2G9. and 20b comprise a structure which functionally corresponds with the grid 29 of Fig. 1 and the

grids

21a and 21b form a structure which functionally corresponds with the grid 21 of Fig. l. The segments of the grids 2t) and 21 are alternately disposed in a circular array.

Segments

20a and 2% are electrically interconnected by means not shown. Similarly

segments

21a and 21b are electrically common. When a magnetic field is present in which the flux lines are parallel to the axis of the tube, the beam 16 will be bent in a circular path to distort the otherwise uniform pattern in the manner illustrated. The direction of the deflection will depend upon the direction of the magnetic field. Thus the deflecting voltage ea, either positive or negative with respect to a given reference point, may be applied to the coil 66 and the grid signal voltage applied to the control grid segments 202., 202), and 21a, 211:. If the radius of the curvature of the electron orbit as controlled by the strength of the magnetic field is relatively small compared to the radius of the anode 12, the defiction of the beam from one signal grid to the other may be a linear function of the magnetic field strength in the tube. The strength of the magnetic field in the tube similarly may be made linearly proportional to the deflecting voltage ea by suitable compensating networks or by supplying coil 66 from a high impedance source. Preferably, regardless of beain area or shape, the anode current will be a linear function of signal applied to coil. Thus either of the modifications of Figs. 2, 8 or 9 may be utilized for multiplication in accordance with the present invention.

In Fig. 9 the screen grid structure is formed by a single element located between the control grids and the suppressor grids. The screen grid structure, whether a single element as in Fig. 9, or multiple elements as in Figs. 2 and 8, serves to collect that portion of the electron beam that is so impeded in its flight, by reason of the potential barrier associated with the control grid, that they may not reach the anode. The desirability of single or multiple elements will, of course, depend upon particular configurations of the other tube elements and the operating parameters chosen, as is well understood by those skilled in the art. Auxiliary electrode systems other than those shown may be employed, the present description being given by way of example only and not by way of limitation.

Fig. 10 illustrates a further modification of tube structure with a section taken along a plane that intersects the axis of the cathode 10. In this figure parts have been given the same reference characters as in Fig. 1 and the screen and suppressor grids have been omitted for the purpose of clarity. The electron beam 16 is established by a structure 11 which surrounds the cathode 10. The beam 16 is in the form of a disc, the electrons being radiated in a substantially uniformly dispersed circular field around the cathode 19. The anode 12 is concentric with the cathode iii and collects the electrons in the beam 16. The

control grids

20 and 21 similarly are concentric with the cathode 1t and form two mesh-like rings of equal diameter with closely spaced but electrically separate adjacent faces. The plane between the two grid rings 20 and 21 is common to a plane lying at the center of the disc-like electron beam whereby the number of electrons passing through grid 29 per unit time is exactly equal to the number of electrons passing through grid 21 under normal operating conditions.

The deflecting means in this tube is formed by a pair of elements in a biconical array having a projected apex at the center of the disc-like beam 16 and at the center of cathode 1G. The element 17 is positioned vertically above the element 18. Application of a voltage between

elements

17 and 18 will deflect or warp the disc-like beam as to form a cone rather than a disc whereby a major portion of the electrons in the beam will pass through one of the grids, i. e. through grid 20 when plate 17 is more positive than plate 18. Concurrently, the number of electrons passing through grid 21 will be substantially reduced. With such structure it Will be apparent that, as long as the foregoing rules of linearity are obeyed, the multiplication steps hereinbefore described may be performed.

Fig. 11 is a sectional View of the tube of Fig. 10. The screen and suppressor grids are illustrated in this view in addition to the

control grids

20 and 21. The inner and outer meshes of the screen grid are illustrated as interconnected by the conductor 70. A connection to the upper grid 20 may be made by way of a cap on the tube (not shown) or by circuit means extending outside the anode 12 to a base plug. By utilizing such structure the tube may be mounted on a conventional socket 71, thus conforming to standard techniques in fabrication of auxiliary components.

The tube of Figs. 10 and 11 has greater power capacity than the modifications of Figs. 2, 8 and 9 because a greater beam factor is employed. Further, the grid structures are more readily fabricated than those of the other figures.

From the foregoing description of the several embodiments of the invention and the manner of utilizing the system for multiplication of two electrical quantities it will be seen that in each case the electron stream is normally directed so that it is equally apportioned between the two signal grids and that unequal apportionment is produced by the beam deflecting voltage whereupon a variation in the signal grid voltage will increase or reduce the total number of electrons reaching the anode. The boundary between the signal grids is of course defined or controlled by the configuration of the adjacent edges of the grids. In all of the figures this boundary is a straight line. Thus in Fig. 1 the electrons normally are equally distributed on opposite sides of the boundary which is in effect a plane common to the line 37 and axis of the cathode iii, the plane passing centrally between the grids and 21 and the

plates

17 and 18. it will be apparent that the boundary is not necessarily limited to the above described planar state since the grids 2t and 21 may possibly have serrated edges. However for the requirements of the present invention, it is necessary only that there be normally equal electron distribution with respect to the boundary whatever its shape and that the distribution be unbalanced in proportion to the deflecting voltage. In the device illustrated in Figs. 10 and 11 the boundary of course is a plane normal to the axis of the cathode it at the center of and parallel to the electron beam 16.

In Fig. l, the battery 32 when connected to the signal grids 29 and Z1 establishes a potential barrier in the electron beam which the electrons must overcome it they are to reach the anode. The application of the grid signal voltage 8g varies the magnitude of the potential barrier on opposite sides of the boundary, i. e. on opposite sides of a plane common to line 37, Fig. 1, so that on one side electrons having lower kinetic energy may reach the anode and on the other side electrons with a higher kinetic energy will be prevented from reaching the anode.

While the invention has been illustrated by several modifications, it is to be understood that such illustration has been given not by way of limitation but that further modifications will be apparent to those skilled in the art. It is intended to cover such modifications as fall within the scope of the appended claims.

What is claimed is:

l. A system for multiplying two electrical voltages which comprises means for establishing a stream of electrons normally having an equal distribution on opposite sides of a boundary, field producing means at a first point along said stream for deflecting the electron beam for an unequal apportionment with respect to said boundary proportional in magnitude andcorresponding in sense to the first of said voltages, a grid system beyond said first pointin the direction of electron flow for establishing a potential barrier in the path of said stream which normally impedes electron flow equally on opposite sides of said boundary, means for varying in equal and opposite senses the magnitude of said barrier on opposite sides of said boundary in dependence upon the magnitude of the second of said voltages, and circuit means for measuring the magnitude of electron flow at a point in said stream.

beyond said barrier.

.2. A system for multiplying two electrical voltages which comprises means for establishing a stream of electrons normally having an equal distribution on opposite sides of a boundary, electrostatic means at a first point along said stream for deflecting the electron beam for unequal distribution with respect to said boundary proportional in magnitude and corresponding in sense to the first of said voltages, a grid system beyond said first point in the direction of electron flow for establishing a potential barrier which normally impedes electron tlow equally on opposite sides of said boundary, means for varying in equal and opposite senses the magnitude of said barrier on opposite sides of said boundary in dependence upon the magnitude of the second of said voltages, and circuit means for measuring the magnitude of electron flow at a point in said. stream beyond said barrier.

3. A system for multiplying two electrical voltages which comprises means for establishing a stream of electrons normally having an equal distribution on opposite sides of a boundary parallel to the direction of electron flow, electromagnetic means at a first point along said stream for deflecting the electron beam for unequal distribution with respect to said boundary proportional in magnitude and corresponding in sense to the first of said voltages, a grid system beyond said first point in the direction of electron flow for establishing a potential barrier which normally impedes electron flow equally on opposite sides of said boundary, means for varying in equal and opposite senses the magnitude of said barrier on opposite sides of said boundary in dependence upon the magnitude of the second of said voltages, and circuit means for measuring the magnitude of electron flow at a point in said stream beyond said barrier.

4. In a vacuum tube the combination of a cylindrical anode, a source of electrons inside said cylindrical anode, electron control means for producing axially confined radial electron flow between said source and said cylindrical anode, a pair of grid structures between said source and said anode disposed above and below a radial plane extending midway of said axially confined electron flow, the electron flow normally being equally apportioned between said grid structures, and means between said source and said grid structures for deflecting said electron flow from one of said grid structures to the other.

5. In a vacuum tube the combination of a cylindrical anode, a cathode structure at the axis of said anode, electron control means for producing radially of said anode disc-like flow of electrons between said cathode and said anode, a pair of electrically isolated cylindrical grid structures spaced one from the other axially of said anode and positioned between said cathode and said anode, the electron flow between said cathode and said anode normally being equally apportioned between said grid structures, and a pair of axially displaced electron deflecting rings concentrically disposed around said cathode between said cathode and said grid structures for deflecting electrons from one of said grid structures to the other.

6. In a vacuum tube the combination of a cylindrical anode, a cathode structure at the axis of said anode, electron control means for producing disc-like flow of electrons radially of and between said cathode and said anode, a pair of electrically isolated cylindrical grid structures axially spaced one from the other and positioned between said cathode and said anode, and a pair of axially displaced .biconical deflecting rings having their centers at said axis and concentrically disposed around said cathode between said cathode and said signal grids for deflecting said disc-like flow of electrons to vary the division of said electron flow between said grid structures.

7. A system for electronically multiplying two voltages which comprises a vacuum tube having a cylindrical anode, a source of electrons in said cylindrical anode, electron control means for producing axially confined radial electron flow between said source and said anode, a pair of grid structures axially spaced one from the other and between said source and said anode, deflecting means between said source and said grid structures for deflecting said electron flow axially of said anode to vary the division of said electron flow between said grid structures, circuit means for applying the first of said voltages in equal and opposite senses to said grid structures for control of the magnitude of electron flow through them,

means for applying the second of said voltages to said deflecting means for establishing a division of said elec tron flow between said grid structures related to the magnitude of said second of said voltages, and means for measuring the total number of electrons reaching said anode as a quantity representing the product of said two voltages.

8. A system for combining two electrical quantities, comprising means for establishing a flow path for a stream of electrons normally having an equal distribution on opposite sides of a boundary, a grid system in said path forming a flow-impeding barrier for normally impeding electron flow on opposite sides of said boundary by equal amounts, means for varying by equal amounts and in opposite senses the magnitude of said barrier on opposite sides of said boundary in dependence upon the magnitude of a first of said quantities and upon the polarity thereof, field-producing means located upstream of said path from said grid system for deflecting the fiow path of said stream in one direction or in the opposite direction in accordance with the polarity of said second of said quantities and by an amount dependent upon its magnitude to change the total electron flow by an amount proportional to the product of said quantities, whether they be of the same or of a diflerent sign, and circuit means downstream from said grid system for measuring said change in said electron flow.

References Cited in the file of this patent UNITED STATES PATENTS 2,171,490 Dalpayrat Aug. 29, 1939 2,172,859 Toulon Sept. 12, 1939 2,197,041 Gray Apr. 16, 1940 2,251,951 Pray Aug. 12, 1941 2,431,396 Hansell Nov. 25, 1947 2,515,456 Loper July 18, 1950 2,551,024 Levy May 1, 1951