US2912613A

US2912613A - Electron beam tubes and circuits therefor

Info

Publication number: US2912613A
Application number: US367850A
Authority: US
Inventors: Jr John S Donal; Cuccia Carmen Louis
Original assignee: RCA Corp
Current assignee: RCA Corp
Priority date: 1953-07-14
Filing date: 1953-07-14
Publication date: 1959-11-10
Anticipated expiration: 1976-11-10

Description

Nov. 10, 1959 J. s. DONAL, JR, ET 2,912,613

ELECTRON BEAM TUBES AND CIRCUITS THEREFOR Filed July 14. 1953 3 Sheets-Sheet 1 Nov. 10, 1959 J. s. DONAL, JR., ETAL 2,912,613

ELECTRON BEAM TUBES AND CIRCUITS THEREFOR Filed July 14. 1953 3 Sheets-Sheet 2 United States ELECTRON BEAM TUBES AND CIRCUITS THEREFOR Application July 14, 1953, Serial No. 367,850

11 Claims. (Cl. 315-516) The present invention relates to' electron beam tubes and circuits therefor, and particularly to beam tubes of the electron coupler type used as phase discriminators, amplitude modulators or balanced modulators.

In an electron coupler tube, as described in C. L. Cuccia Patent No. 2,452,797, assigned to Radio Corporation of America, an electron beam is projected along a path extending in succession through an input region upon which is impressed a radio frequency electric field transverse to the beam path and an output region to which the beam gives up energy by inducing a second transverse radio frequency field. Each of the two regions includes a pair of opposed field-defining electrodes located on opposite sides of the beam path and coupled to a resonant circuit, which may be either an external circuit or some form of cavity resonator forming part of the tube structure itself. The input and output circuits are adapted to be coupled to a radio frequency source and a useful load, respectively. The beam is subjected to a constant axial magnetic field. The magnetic field strength H is adjusted to a value such that the angular cyclotron frequency of an electron in said field is equal to where e and m are the charge and mass, respectively, of an electron, and w is the angular frequency of the radio frequency field set up in the input region. Under these conditions, as the beam passes through the input region the electrons are caused by the crossed electric and magnetic fields to move in spiral paths of increasing radii about the atent Another object is to provide an improved amplitude modulator tube.

Still another object is to provide a tube which will serve as a balanced modulator.

These and other objects are achieved in accordance with the invention by incorporating the basic principles of the Cuccia electron coupler in an electron coupler tube having three separate radio frequency electric field regions, two of which are input regions and the other of which is an output region, with two coupling beams. Two embodiments of the tube are illustrated, in the first of which the three regions are provided by three coaxial resonators arranged in a row with an electron gun located adjacent to each end resonator and a coloriginal beam axis, absorbing energy from the input region and the radio frequency source coupled thereto. Since they have the same angular and axial velocities, all the spirally traveling electrons in the beam lie at any instant on the linear directrix of a cone, and the envelope of the rotating or revolving beam is a cone. Hence, the beam is sometimes termed a cone-directrix or rotating pencil beam. In the output region the rotating beam gives up spiral energy to the region and coupled load by inducing radio frequency voltages on the field defining plates and thereby setting up a transverse radio frequency field in that region. As spiral energy is abstracted from the beam the radii of the electron paths are progressively reduced, and hence, the envelope of the beam in the output region is an inverted cone. The electron coupler tube may be used as a modulator by coupling an unmodulated carrier signal to the input region and modulating the beam current or the transit time by a modulating signal thus producing a modulated carrier signal in the output load, as disclosed in said Cuccia patent. On the other hand, an amplitude modulated signal coupled to the input region will be faithfully reproduced in the output of the device.

The principal object of the present invention is to provide a single electron tube which will perform all of the functions of a phase-discriminator circuit.

lector located between each pair of adjacentresonators in such manner that each beam passes through an end resonator and the middle resonator only. Two input sources are coupled respectively to the end resonators, and the middle resonator is coupled to a useful load. In the other tube embodiment illustrated, the two input resonators and associated electron guns are arranged side-by-side and the output resonator surrounds the extensions of the two beam paths beyond the input resonators. Schematic circuits are also included to illustrate the uses of the tube as a phase discriminator, an amplitude modulator and a balanced modulator.

In the accompanying drawings:

Fig. l is a longitudinal sectional view, partly schematic, of one embodiment of a phase discriminator electron coupler tube incorporating the invention;

Figs. 2 and 3 are transverse sectional views taken on lines 22 and 3-3, respectively, of Fig. 1;

Fig. 4 is a view similar to Fig. 1 of another embodiment of a phase discriminator tube incorporating the invention;

Fig. 5 is a transverse sectional view taken on line 55 of Fig. 4;

Fig. 6 is a vector diagram showing the combination of radio-frequency voltages in the operation of an embodiment of the invention;

Fig. 7 is a longitudinal sectional view, partly schematic, showing a particular utilization of a phase discriminator tube incorporating the invention;

Fig. 8 is a transverse sectional view, partly schematic, showing another particular utilization of a phase discriminator tube incorporating the invention; and

Fig. 9 is a schematic transverse sectional view showing still another particular utilization of a phase discriminator tube incorporating the invention.

Referring first to Figs. 13, there is shown an electron tube having an elongated metal envelope ll. Within the envelope, and spaced as shown in Fig. 1 on opposite sides of the longitudinal central axis ZZ of the tube, are mounted by suitable means two electron guns consisting of heaters 3 and 5, cathodes 7 and 9, control grids 11 and i3, and one or more apertured accelerating electrodes 15 and 17 shown here for convenience as mechanically internally connected to the envelope and hence electrically at the same potential and at the potential of the envelope. The electron guns are arranged to project two electron beams with their average direction parallel to the axis of the electron tube. A constant magnetic field of strength where m and e are the electronic mass and charge, re spectively, and m is the angular frequency of the R.F. voltage used to excite the electron tube, as explained in greater detail below, is caused to extend along the axis of the device, within the envelope. This magnetic field may be produced by any suitable means, as, for example,

7 now be explained by reference to Figs. 1-3.

a maximum value.

by an electro magnet, or a permanent magnet external to the tube and making use of the pole pieces P-P- of Fig. 1.

Within the envelope of the tube, in theembodiment of Fig. l, are three sets of

pole faces

19, 2 1 and "2 3. In each pair, such as 19, the two pole faces are spaced on opposite sides of the tube axis. Although the pole faces are shown as flat plates, supported by

extensions

24, 25, 27of the envelope, the pole faces may be arcuate in form and may be supported in any other suitable manner, such as from partitions located at the ends thereof, a method of support which is illustrated in Fig. 4 and discussed in connection .with that figure. In Figs. 1-3 each of the electron beams passes through two of the regions of the tube containing the pairs of pole faces. In Fig. 1 the pairs of pole faces are so supported, and the discharge device is so divided into three regions, by the accelerating electrodes 15 and 17 which also serve as partitions and by two

other partitions

29 and 31 perpendicular to theaxis of the tube, that three separate resonant cavities or cavity resonators are formed. This construction is shown as an example only. Alternatively, electrical connection'may be made between the pole faces and external circuits, so that resonant circuits are formed which are situated partly within and partly exterior to the envelope of the tube. In the form illustrated in Fig. 1 the entire resonant circuit is in each case contained within the envelope with coupling means such as

loops

33, 35 and 37 f suitable form connected to

external transmission lines

39, 41 and 43 so thatenergy at radio frequency may be introduced into the resonant cavities or removed from them.

In the operation of the embodiment of the invention shown in Figs. 1-3 the envelope 1 is connected to a source of direct current potential, such as the battery 45, and

the cathodes 7 and 9 are connected to the same source of potential, in a manner such that the envelope is maintained at a potential highly positive with respect to the potentials of the cathodes. The cathodes may beat the same or at somewhat different potentials. The intensities of the two election beams are adjusted by varying the potentials of the control grids, which potentials are preferably negative with respect to the potentials of the corresponding cathodes. The two electron beams are collected at the

anodes

47 and 49 which are suitably cooled and maybe maintained individually at potentials which are the same or different from the potential of the envelope 1.

The operation of one embodiment of the invention will A source of radio frequency, designated as input No. 1 is coupled through line 39 and loop 33 to the above-described resonant cavity formed by elements 1, '19, 2.4, 15 and 29. Input No. 1, of angular frequency w =H@/m produces a periodically varying voltage between pole faces 19 ith the electric field directed perpendicularly to these pole faces. As is described in. detail in Patent No. 2,542,797, the electrons in the bearrr moving to the right in Fig. 1 gain energy from the R.F. field arising from input No. 1 and execute circles of increasing diameter as they pass between pole faces 19. One possible conical envelope 51 of their trajectories is shown by dotted lines in Fig. 1. i a directrix of the cone shown so that either dotted line of the cone '51 may be considered to be the positions occupied by the electrons at one instant of time. The dotted line between grid 11 and the left-hand edge of pole faces 19, in Fig. 1, represents, for convenience, an infinitely thin beam, although the beam may be caused, by suitable means, to have a diameter of almost any desired value. 'When the beam is in the plane of the figure, the RF. voltage between the pole faces will have If the force on the electron is directed out of the plane of the figure and toward the ob: server, and is at its maximum value, theindividpal elecr At any instant all electrons lie along q A We trons lie in the plane of the figure and are moving toward the observer. Thus the position of the directrix beam at any instant. is uniquely determined by the phase of the R.F. voltage between the pole faces of the resonant cavity Furthermore the cone directrix beam becomes an element of a cylinder after the electrons leave the pole faces 19. As long as the electrons are in a resi nin item R-F- r e e d the ont n to revolve around the tube axis at the same angular frequency, as they move, individually, in a spiral path down the tube yvith the pitch of thespiral determined only by the axial velocity imparted by the DC. axial field. Thus information as to the phase of the R.F. voltage between the pole faces is retained .by the electrons after they leave the pole faces, for, as an exexample, at any instant when all electrons are in the plane of the figure "and moving toward the observer, the force on the electrons (oppositely directed to the R.F. field) within the pole faces has again assumed a maxivalue and directed perpendicularly to this plane and in a direction toward the observer. I

In the'same manner, the other electron beam, arising at the electron gun formed by elements 5, 9 and 17, but proceeding toward the left in Fig. 1, retains information concerning the phase of the R.F. voltage between the pole faces 23 which, together with

elements

1, 17, 31 and 2 5 form a resonant cavity coupled to input No. .2 by loop 37 and line 43. The direction of rotation. of the trace of each beam on any section, such as Fig. 3, of Fig. 1 is determined by the direction of the constant axial magnetic field. This direction is independent of the direction of axial motion of the spiralling electrons. Therefore, the directions of rotation of the two cone directrix. beams, theenvelopes of which are shown by 51 and '53, are the same, although their instantaneous positions in a plane perpendicular to Fig. 1 are determined only by the phases of the RF. voltages between the pairs of pole faces through which they have separately passed; Each of the 'two cone-directrix beams 51 and 53 is generated independently of the other beam, by the separate electron gun and separate R.F. field means, combined with the common magnetic field. Beam 51 retains information concerning the phase of the.R.F. voltage between plates 19; beam 53 retains information concerning the phase of the R.F. between plates 2$Q These phases may be quite different. It is of course well known that the phases of the RF. voltages b etWee n the pole faces differ from the phases at the sources indicated by input No. 1' and input No. 2, but the phase differences between the inputs and the corresponding pole faces are substantially constant, so that "any change in phase at input No. 1 or input No. 2 is accompanied by substantially the same change in phase at thecorresponding pole faces. The electrons,

' .as high as from one to one times the radio frequency is substantially completely. indicated by the new angular position assumed by the cone-directrix or cylinder element electron beam.

The invention has now been described up'to the point of showing how the electron beams, leaving the resonant cavitiesshewn connected to input No. l and input No. 2 of Fig. 1, have instantaneous positions indicative of the phases, atthesameinstants, of the R.F. voltages between p e faces .9 an .3,- It w l HOW be s wn hatths device can preducean output in a load which output indicates by its magnitudethe relative phases or the R.-F. voltages between poles faces 1-9 and 230. and, after a correction for the constant phase differences mentioned above, the relative phases of inputs No. 1 and No. 2. The central cavity of the device as illustrated in Fig. 1 is coupled by the loop 35 and line 41 to the output load. An additional shunt load 55 is shown in parallel with the output load. The total loading presented by the shunt load and the output load are adjusted to the desired value as described in greater detail below, although the output load alone may alternatively be designed to have the desired characteristics.

As explained in detail in Cuccia Patent No. 2,542,797, when the beam 51, for example, enters the central resonant cavity, a voltage is induced in the pole faces 21 in a phase exactly opposite to the phase of the voltage between pole faces 19. The radius of revolution of the electrons is progressively reduced as they pass between pole faces 21, so that the cone directrix beam is a directrix of a cone converging as shown by the envelope 57. The radius of revolution of the electrons is smaller when they leave the pole faces 21 than upon their entrance. Since the total spiral energy per electron varies as the square of the radius of revolution, energy has been lost. This spiral energy has been transformed to the resonant circuit and, neglecting circuit losses, to the output load. The degree of convergence of the envelope between the pole faces depends upon the resistive portion of the total output load, here the shunt load 55 plus the output load as illustrated. When the resistance R of the output load, expressed as the transformed resistance presented across the pole faces 21, has the specific value defined by:

V d 2 hir) (2) where V is beam accelerating voltage in volts, I is the electron beam current in amperes, L is the length of the pole faces and d is the distance between them, both in centimeters, the beam will converge so that the radius of revolution of the electrons is Zero as they leave the region between the poles faces 21. Equation 2 corresponds to Equation 15 of Cuccia Patent No. 2,542,797. If R has a value greater than R the beam will converge before it has passed completely between the pole faces, beyond which point of convergence the electrons will rotate with a radius of revolution increasing again so that the beam is the directrix of a cone with its apex pointed to the left in Fig. 1. If R has a value less than R the beam will not be completely converged when it leaves the region between the pole faces.

In accordance with the invention both beams 51 and 52 absorb energy as they pass between pole faces 19 and 23, respectively, and both beams induce voltages in the output pole faces 21. Only one voltage can exist between pole faces 21, which voltage is the vectorial sum of the voltages induced by the two beams.

As an example, it is assumed that the two beam currents are adjusted to equality and that the amplitude of the radio frequency voltage between poles faces 19 is equal to the amplitude of that between pole faces 23, although neither condition is required for the effective operation of the invention. It is further assumed as an example that R equals R given by Equation 2 where I is the sum of the currents in the two beams. When the radio frequency voltages between the pairs of pole faces 19 and 23 are exactly in phase, the beams entering the pole faces 21 are rotating in the same direction and they are in the same spatial phase, i.e., if at one instant beam 51 is in the plane of the section of Fig. 1 and moving toward the observer, the other beam is also in the plane of Fig. 1 and moving toward the observer. The voltages induced separately by beams 51 and 53 in pole faces 21 are in exactly the same phase so that their vectorial sum is in the same phase as either voltage considered separately. Since Equation 2 is satisfied, both beams converge to zero radius of revolution just as they leave the pole faces 21. As will be made clear in greater detail below, the preferred value of R should be equal to or less than R given by Equation 2 where I is the sum of the beam currents. For R somewhat less than R given by Equation 2, the beams will converge equally, but not quite to zero radius of revolution, as shown in Fig. 1.

The example given above, in which the two beams were in the same spatial phase upon entering the pole faces 21, resulted in a total induced voltage, between pole faces 21, equal to the sum of the separately induced voltages. Each electron beam gave up an amount of power proportional to the square of the difference between the electron radius of revolution upon entering and upon leaving the pole faces 21. The voltage developed across the output load was equal to the square root of the quantity equal to the sum of the powers lost by both beams divided by the resistance of 55 and the output load in parallel.

Let it be assumed that the beam currents remain unchanged but that the phase of the radio frequency voltage across pole faces 19, for example, is altered by 180 electrical degrees. The beams now enter the output pole faces 180 degrees out of phase spatially. Since the beam currents are equal, as assumed, the two beams tend to induce equal voltages in pole faces 21 but these voltages are 180 out of phase so that the net induced voltage is zero, and the beams are not collapsed at all. No power is given up to the resonant cavity and no voltage appears across the output load.

As a third example, it is assumed that the radio frequency voltages across the pole faces 19 and 23 are electrically out of phase. The two beams enter the pole faces 21 with their instantaneous radii of revolution at 90 in space. The voltages induced in pole faces 21 are 90 out of phase in time so that the vectors representing the two voltages must be drawn at 90 to each other The sum vector lies, in magnitude, between zero and the maximum value reaches when the radio frequency voltages were in phase between the pole faces 19 and 23.

Thus it is seen that the device shown in Fig. 1 produces an RF. voltage across the load which is indicative of the relative phase of the RF. voltages between pole faces 19 and 23, and hence, indicative of the relative phase of the radio frequency voltages of input No. 1 and input No. 2, although due to the line-lengths between the inputs and the resonant cavities, and between the central resonant cavity and the output, the voltage across the output load may not be a maximum when the two inputs are in phase. This last is not of fundamental importance, since account can be taken of the fixed line lengths. For example, the two inputs can each be situated electrically an integral number of wavelengths at angular frequency (0 from the electrical planes of pole faces 19 and 23, respectively, and the output load can be electrically situated an integral number of wavelengths at angular frequency w from the electrical plane of pole faces 21. The radio frequency "oltage across the output load is then a maximum when the radio frequency voltages of inputs No. 1 and No. 2 are in phase, and decreases as these radio frequency voltages become out of phase.

Within the scope of the invention, the radio frequency voltage across the output load may be used directly for control purposes, for example, as a radio frequency voltage of an amplitude indicating the relative phase of the radio frequency voltages of inputs No. 1 and No. 2. Alternatively, the voltage across the output load may be rectified and filtered so that a DC. voltage is obtained with a magnitude indicative of the relative phase of the radio frequency voltages across inputs No. 1 and No. 2. In this case the DC. voltage will show whether the radio frequency voltages of the inputs are in phase, out of phase, or at some intermediate phase. In the latter case, however, the DC. voltage will not show which input voltage is ahead in time phase. In many cases this information is entirely unnecessary, as will be made clear by a later example.

It was stated above that in the preferred form of the invention the value of the total load resistance is that givenby Equation 2, or is less than this value. Assume, as anexample, that the total load resistance R is greater han that given by 2. When the two beams entering the pole faces 21 are in phase, spatially, each will be completely converged and will begin to diverge again before it leaves the pole faces, Power will be given up, however, and a voltage will appear across the output load. Now suppose that the input voltages alter their phase so that the beams'are somewhat out of phase spatially as they enter the region between pole faces 21. The total radio frequency voltage induced in pole faces 21 will be reduced as the two vectors assume an angle to each other. The convergence of the beams will be decreased, so that for a particular phase they will be completely converged just as they leave the pole faces 21. The power remaining in the beams is reduced, so that the power going to the output load is increased. The voltage across the output load rises, rather than falls as in the above example, as the input radio frequency voltages go out of phase. Since the voltage across the output load falls again for still greater phase differences between the input.

voltages, there will be a single voltage across the output load for two values of phase difference between the input voltages. While in many cases this will not degrade the performance of the invention, the possible existence of such a behavior should be taken into account.

Figure 4 is a longitudinal sectional view of another embodiment of the invention and Fig. 5 is a transverse sectional view taken on line 55 of Fig. 4. In this case two input resonant cavities, 5t; and 59, are placed side by side within the envelope 1, and two electron guns at and 63,.also placed'side by side, cause electron beams to pass along parallel paths in the same direction. Another of the possible forms of construction of the resonant cavities, equally applicable to Fig. 1, has been shown in Figs. 4

a and 5. Arcuate pole faces 65 and 67 are supported at "patent application of C. L. Cuccia, Serial No. 216,320,

filed March 19, 1951, now Patent No. 2,806,172, the beam at any instant is not the directrix of a cone, but all electrons lie at any instant in a plane parallel to the axis of the resonant cavity and passing through this axis. The

path of the beam is not'a straight line, but a curve as shown at 75.

Except as described, and the mode of operation of the embodiment of Figs. 4 and 5 is similar to that of Figs. 1-3. The parts serving the same functions as Fig. 1 have been given the same numbers. The radio frequency voltage across the output load is again indicative of the relative phase of the radio frequency voltages of the two inputs. i

An example will now be given of the use of the invention'for the control of radio frequency oscillators. Let it be assumed that a magnetron oscillator is to be'synchronized with a crystal controlled oscillator by the proce dure known as phase locking, so that under equilibrium conditions the frequencies of the two oscillators are identical and thereis a fixed phase relation between the two oscillators unless an effort is made to perturb the frequency of one of the oscillators, as by plate modulating the magnetron. The embodiments of either Fig. 1 or Fig. 4 would serve for this purpose. Fig. 1 will be chosen forpurposes of illustration.

i Let input No. 1 represent the magnetron and input No. 2 represent the crystal controlled reference oscillator. It

is assumed that the magnetron is equipped with an electronic mechanism for the control of its frequency, such as the frequency control electron guns described in Dona-1 et al. Patent No. 2,602,156, assigned to Radio Corporation of America. The control grids of these frequency control guns serve as the'output load, with the'shu'nt load serving for theadjustrnent of the total load resistance to the desired value as described earlier. When the radio frequency voltages from the magnetron and the control oscillator are in phase at the pole faces 19 and 23 respectively, the radio frequency voltage across the output load, after rectification and filtering, gives a maximum voltage at the grids of the magnetron control guns. If the phase of the magnetron alters, the voltage at the grids of the control gun decreases. If the voltages from the magnetron and from the crystal controlled oscillator are 180 out of phase at the respective pole faces, the voltage at the grids of the control guns will be zero. It is assumed that by Well-known means an independent DC. bias is applied to the grids of the control guns and that when the magnetron phase rnoves either ahead or behind the phase ofthe. crystal controlled oscillator by 90 (the phase of the voltage at the pole faces is meanthere and in the following), producing in either case the same reduced voltage at the grids of the control guns, the D.C. bias is so adjusted that the frequency of the uncontrolled magnetron is equal to that of the crysal-controlled oscillator. It is further assumed, as an example only, that the addition .of more signal from the phase discriminator of Fig. 1, to this bias, causes the beam current of the frequency control guns to increase and causes the frequency of the magnetron to tend to decrease. Such a tendency to decrease would retard the phase of the magnetron with respect to the crystal-controlled oscillator.

- The condition for stable phase difference between the two oscillators will now be examined. Suppose the magnetron is initially 90 in phase ahead of the crystal-controlled oscillator. The vector diagram is shown in Fig. 6.

The magnetron voltage, '77, leads the crystal-controlled oscillator voltage, 79, by 90 at the pole faces 19 and 23, respectively. The resultant output voltage across the load has the magnitude shown at 81. This voltage, rectified, addsto the bias of the frequency-control guns. Suppose the magnetron voltage 77 is caused, by plate modulation, for example, to move ahead in phase to a newposition 83. The resultant voltage at the control grids from the phase discriminator is reduced as shown at 85. The total voltage at'the control grids is reduced and the decreased beam current of the frequency-control guns causes the magnetron frequency to tend to increase. This advances the phase of the magnetron still further and no phase stability has been found. However, the

' magnetron phase advances until it is 90 behind the phase of the crystal-controlled oscillaton'as shown at 87. Here the two radio frequency voltages add to produce a resultant rectified voltage 89' at the grids of the frequency control guns. If now the magnetron tries to advance in phase,

the voltage at the grids is increased, the frequency-control gun beam current is increased, and the magnetron frequency tends to decrease. Any phase advance of the magnetron beyond 87 results in a correcting signal which retards the phase again. Correspondingly, any retardation in phase 87, gives a signal tending to advance the phase. When the magnetron is 90 behind the crystalcontrolle d oscillator, stable phase locking is obtained. Of course, if increasing positive signal on the grids of the frequency-control guns were made to raise the magnetron frequency, rather than to' lower it, stable phase locking would be found with the magnetron ahead of the crystal-controlledoscillator.

From the above it is seen that either of the embodiments of the invention so far described can be used to phase lock two oscillators together; If one of these oscillators is assumed to have .a stable frequency, as in the above case, the second oscillator is made to have a stable frequency.

From the descriptions taken in connection with Figs. 1-6, it is evident that the invention can be used, for example, to control very high power oscillators. Two of the devices shown in Fig. 1 or Fig. 4 could be used, as another example, to control the frequency and phase of two magnetrons with reference to a single crystal-controlled oscillator. By modulating the grid voltages of the frequency-control guns in the magnetrons, the phase of one tube can be caused to advance and the phase of the other to retard. If the outputs of the two magnetrons are properly combined in a load, such as an antenna, an amplitude modulation system results. In combining the powers in an antenna, however, it is difficult to prevent the reaction of one oscillator upon the other. Various means,'such as radio frequency bridges, have been used for this purpose. With this device, when the two voltages are in phase at the load, the power not desired in the antenna is dissipated in a dummy load.

As a further embodiment of the invention, the combination of the power from two phase-controlled oscillators is accomplished by an electron tube similar in principle to that of Figs. 1-5. The two phase-controlled magnetrons, for example, serve as inputs No. 1 and No. 2 of Fig. 1. Their power is absorbed by the beams 51 and 53 and delivered to the output load. When the phases of the two inputs are the same, at pole faces 19 and 23, respectively, all of the power output of both oscillators reaches the output load, provided that the resistance R of the output load (the shunt load is eliminated in this embodiment) satisfies Equation 2 where I is the sum of the two beam currents. For best operation, the two beam currents of Fig. 1 should be equal. When the controlled phase of input No. 1, for example, is advanced, and that of input No. 2 is retarded, the power in the output load is decreased. If the phases of inputs No. 1 and No. 2 are made 180 different, at the pole faces, the power in the output load falls to Zero. The power not reaching the output load is dissipated on the water-cooled

collector electrodes

47 and 49. Due to suitable electromagnetic decoupling between the central output cavity and the two input cavities, or circuits, one oscillator cannot react upon the other. If one oscillator is advanced while the other is retarded by the same amount, and if the beam currents are the same, as proposed, amplitude modulation without phase modulation is produced by this embodiment of the invention.

Still another embodiment of the invention is shown partly schematically in Fig. 7. As in Figures 1 and 4, two

electron guns

91 and 93, within an envelope 95, project beams 07 and 99 parallel to an axial magnetic field H and between pairs of pole faces 101 and 103. These pole faces may form portions of resonant cavities or may be connected to external resonant circuits. The beams enter the space between a pair of pole faces 105 connected by suitable means to the output load. The envelope 95 and the

collector electrodes

107 and 109 are held at potentials, which may be the same, high with respect to the potentials, which may be the same, high with respect to the potentials of the cathodes, by means of a source of potential such as the battery 111. The control grids of the electron guns are suitably biased by means of sources of potential such as the

batteries

113 and 115.

The operation of the embodiment of the invention shown in Fig. 7 will be understood from the following description. A radio frequency generator is suitably con-' nected to the pole faces 101 in such a manner that a radio-frequency field is produced between said pole faces. The electrons of the beam 97 gain spiral energy from this field so that a portion of the power output of the radio frequency generator is absorbed by this electron beam in the manner described in connection with Fig. 1. The same radio frequency generator is connected to a phase shifter which is in turn suitably connected to the second pair of pole faces 103 so that power from the radio-frequency generator is absorbed by the electron beam 99. The circuit and the beam currents are so adjusted that substantially equal portions of the power from the radio-frequency generator are absorbed by the beams. The phase shifter is a device known in the art, which might consist of variable reactances suitably connected to lengths of transmission lines, the whole having the property of shifting the phase of the radio-frequency voltage, at the output of the device, with respect to the phase of the radio-frequency voltage at the input of the device. The degree of phase shift is controlled by a modulator. By means of the modulated phase shifter the phase of the radio frequency voltage appearing between pole faces 103 may be shifted with respect to the phase of the radiofrequency voltage appearing between pole faces 101. As

explained by the earlier description taken in connection.

with Fig. 1, the revolving cone-directrix beam formed between the pairs of input pole faces will enter the region between the output pole faces 105 with relative spatial phases of revolution which differ from each other by the electrical phase dilference of the radio frequency voltages impressed between pole faces 101 and 103. When the spatial phases of revolution of the beams are equal, maximum power will be transferred to the output load, which should have a magnitude given by Equation 2 when I is the sum of the two beam currents. When the two beams are in spatial phase opposition, no power is transferred to the output load, but all of the power in the beams is dissipated in the suitably cooled

collectors

107 and 109. For intermediate phases, a portion of the power from the radio frequency generator reaches the output load, the remainder going to the collecting electrodes.

It will now be seen that the invention accomplishes amplitude modulation in response to a signal arising from the modulator. Alternatively two phase shifters might be connected between the radio-frequency generator and the respective pairs of pole faces, so that a single modulator acting upon the two phase shifters would advance the spatial phase of one electron beam while retarding the spatial phase of the second beam. Amplitude modulation would be accomplished, with the difference that the symmetrical operation would maintain constant the phase of the radio frequency voltage across the output load.

Still another embodiment of the invention is shown in Fig. 8. This embodiment performs the useful function of a balanced modulator.

A balanced modulator is an electronic system widely used for applications where carrier suppression in a modulation system is desired. In many systems of radio-telephone transmission, for example, the carrier is suppressed at the transmitter by using a balanced modulator; only the side bands are transmitted which, upon arriving at the receiver, are recombined with a locally produced carrier and then demodulated in customary fashion. Such a system is useful for insuring privacy during the transmission.

Also in many types of frequency-modulated transmitters, the frequency-modulated wave is produced by beginning with an amplitude modulated wave which is then applied to a balanced modulator which removes the carrier. The resulting side bands are then shifted to a proper phase and recombined with a carrier so that the output is a phase or a frequency-modulated wave.

The embodiment shown in Fig. 8 is also an improvement over conventional balanced modulator circuits since in addition to having the capability of functioning as a balanced modulator at frequencies much higher than those at which standard circuits will perform efliciently, it is also very versatile in that it can be used to suppress the carrier in either F.M. or A.M. transmisssion.

Consider first one method of operation of the embodiment in Fig. 8 for the purpose of suppressing the carrier during amplitude-modulation transmission, utilizing one set of phase conditions which yield the desired result. In Fig. 8 the same numerals as in Fig. 7 are used to indicate "ll elements'performing the same function as in the cmbodi-'-' ment of Fig. 7. i 3

Here let the adjustment of a phaseeinverter, between a radio-frequency signal generator and the pole faces 1%, be such that a phase-shifted signal -E cos w t appears across pole faces 103 while a signal E cos w t appears across pole faces 101, where is the angular frequency of the signal and is also the cyclotron angular frequency. Then let a modulating signal E sin ta l from a modulator be applied to the coupling-beam gun 91 and let the. phase inverter in line to coupling beam gun 93 be adjusted so that the signal E M cos w t be applied to this gun, where a is the angular frequency of the modulating signal.

Between the output pole faces 1%, the transverse electric field will be 180 out of phase with respect to the signal. E cos w t being to applied pole faces 1M and 1&3; therefore the current which will appear in the load circuit from output pole faces ltlS will be o=iol+i where 1' is the electron current due to the coupling beam 97 and 1' is that due to the beam 99.

The currents i and i will be of the forms, where M is the modulation factor:

The removal of the carrier or any side frequency from a frequency modulated wave can be accomplished in an equally simple fashion by the embodiment shown schebe applied to the pole faces 103 Where Aw is the maximum frequency deviation and m and w are the angular frequenciesof the carrier wave and of the modulating wave, respectively. Then let a constant amplitude signal (assuming that the carrier is to be suppressed) be impressed across the pole faces 101. The amplitudes V of the beams are maintained constant.

it is evident that a signal of frequency to will not appear in the output load, thereby resulting in carrier suppression.

' Note that should the signal frequency, and the other source is a source of constant p 12 be impressed across pole faces 101, the signal at frequency o +nw will be suppressed inthe output.

' What is claimed is:'

1. An electron'beam tube comprising: first and second means for independently generating two radio frequency beams of electrons spiralling about two parallel axes at ference between said sources.

2. An electron beam tube as in claim'l, wherein said first-named means comprises two electron guns arranged to project two electron beams along said two parallel axes, separate means responsive to two different signals for establishing a radio frequency electric field transverse to each of said two beams in a region traversed by'that beam only, the phase of each electric field being determined by the phase of the respective signal producing said field, and means for establishing a constant magnetic field along both ofsaid beams parallel to said axes.

3'. The combination of the electron beam tube of claim 9 with two radio frequency sources having the same.

angular frequency substantially equal to where H the strength of said magneticfield, and e and m are the charge and mass, respectively, of an elec tron, but having any phases, coupled respectively to said separate means.

4. The combination of claim 3, wherein said two sources are independent radio frequency generators.

5. The combination of claim 3, wherein one of said electric field establishing means is coupled directly to a radio frequency generator of said frequency, and the other electric field establishing means is coupled through a phase-shifting means to the same generator.

6. The combination of claim 5, further including a modulator coupled to said phase shifting means.

7. The combination of claim.5, wherein said phaseshifting means is a phasednverter, and further including means for applying a modulating signal to said two beams in opposite phase. a

8. The combination of claim 3, wherein one of said sources is a voltage source of constant amplitude and index frequency-modulated voltage.

9. An electron beam tube including first, second and third electrode structures, each adapted to be energized to establish a radio frequency electric field transverse to a. predetermined axis of the tube, means for projecting a first beam of electrons parallel to said axis and through the fields of said first and second electrode structures only, in that order, means for projecting a second beam of electrons parallel to said axis and through the fields of said third and second electrode structures only, in that order, and means for establishing a constant magnetic field through all of said structures parallel to said axis, each of said first and third electrode structures having radio frequency input coupling means adapted to be coupled to a separate radio frequency source, said second electrode structure having radio frequency output coupling means adapted to be coupled to an output load. 7 10. An electron beam tube as in claim 9, wherein said first, second and third electrode structures are aligned with each other along said axis in the order named.

11. An electron'beam tube as in claim 9, wherein said first and third electrode structures are juxtaposed on oppo- Sides at sa d xi an a smn a t des tae 14 Donal Aug. 21, 1951 Donal et a1 July 1, 1952 Mesner Sept. 30, 1952 Cuccia May 12, 1953 Cuccia Sept. 10, 1957 OTHER REFERENCES Proceedings of the Institute of Electrical Engineers, vol. 100, No. IV, 5, October 1953, pages 16 to 24.