US2928021A - Duplex traveling-wave tube amplifier - Google Patents

Description

March 8, 1960 D. v. GEPPERT 2,928,021

DUPLEX TRAVELING-WAVE TUBE AMPLIFIER Filed Aug. 19, 195'? 2 Sheets-Sheet l irrt/vanne INVENTOR. 00A/a ww I4 Gip/ier March 8, 1960 D. v. GEPPERT 2,928,021

DUPLEX TRAvELING-WAVE TUBEv AMPLIFIER Filed Aug. 19, 1957 2 Sheets-Sheet 2 m'. la v INVENTOR. o/vow/v l( Gif/ier Wwf/E@ 2,928,021 K DUPLEX TRAVELlNG-WAVE TUBE AMPLIFIER Donovan V. Geppert, Santa Clara, Calif., assignor, by mesne assignments, to Sylvania Electric Products Inc., Wilmington, Del., a corporation of Delaware Appuaaon August 19, 1957, vserial No. 673,892 s claims. (ci. sis- 3.6)

This invention relates to electron tubes of the traveling-wave type. It provides-novel electron tubes, especially remain low when operated under conditions appropriate to vbroad-band signal amplification. Reasons for this low eciency need not be explored for an understanding of the present invention. However, it can be noted that a large portion of the long electron beam in a conventional traveling-wave tube interacts with a relatively useful for amplifying electric signals over broad frequency ranges at microwave frequencies, that have higher efficiencies and greater outputpower capabilities than those obtainable with conventional traveling-wave tubes. Relatively small size, freedom from backward-wave interaction and other advantages are also achieved while the basic advantages of traveling-wave operationV are retained.

The -amount of output microwave power that can be obtained from a conventional traveling-wave tube amplifier islimited by several factors that are well known to those skilled in the art. Conventional traveling-wave tubes require a long electron beam of small cross-sec'- tional area. In suchl a tube, sufficient beam current for a large power output would necessitate very high current densities in the electron'beam. Thus, a very high rate of electron `emission from a small cathode area is needed. Even with heavy magnets it is hard to keep a long, high-density electron `beam focused into a cyl- :inder of small cross-sectional area. .The mutual repul- :sion between electrons in a beam fhaving high.space `change densities produces powerful forces tending to ,spread the beam. If electro-magnets are used for beam focusing the considerable power consumed by the magnets further reduces the already low over-all efficiency of the amplifier.

At the higher microwave frequencies the solution of these problems becomes increasingly diicult in conventional traveling-wave tubes because the necessary smallness at such frequencies of the helix (or other slow-wave structure) limits the permissible cross-sectional area of the electron beam to very small values. Backward-wave interaction in the tube further limits the permissible maxi- -mum helix diameter and therefore limits the permissibleY cross-sectional area of the electron beam, even in cases where a hollow cylindrical beamY outside the helix is employed. In conventional traveling-wave tubes designed for operation in the higher microwavefrequency ranges the helix diameter and the cross-sectional area of the electron beam become quite small.

y Not only the power output but also the frequency response ofthe amplifier suffers as a consequence of the` small helix diameters required in conventional travelingwave tubes for operation in the higherfrequency ranges. This is because a helix becomes dispersive as its diameter is made small--that is, the phase velocity of electromagnetic waves transmitted along the helix varies as a function of frequency, so that interaction with the Y electron-beam can be maintained lover only a relatively narrow band of frequencies for any given beam velocity.

Similarly, the use of dispersive filter-type circuits and come other difficulties, narrows the frequency bandwidth.V

Even if all of the'foregoing diliiculties were solved the eiciencyof the conventionaltraveling-wave tube would weak RF field and therefore a large portion of the RF field is not strong enough to extract a considerable percentage of the kinetic energy from the beam. A strong RF field, capable of producing large changes in the electron velocities, exists only in a relatively short section near the output end of the slow-wave structure and even here over-bunching of the beam, saturation',` effects, and other factors may limit the amount of energy that can be transferred to the RF field. The remaining kinetic energy ofthe electron beam is convertedinto heat when the electrons strike the collector electrode and other parts 1 Not only is this energy wasted, with conof thetube. sequent low eciency of the amplifier, but also the heat must be removed from the tube. In typical cases the energy so wasted is several times larger than the energy transferred from the electron beam to the RF field. The use of RF attenuators, necessary in a conventional traveling-wave tube lto prevent undesirable oscillations, further lowers the eiiiciency.

Heretofore, extensive efforts by many workers in the art to solve the foregoing and other problems have resulted in the proposal and evolution of many ingenious amplifier tubes of Various types and of mixed types, too numerous to describe here. In general, tubes known to the prior art for amplifying signals at high microwave frequencies either are ineiiicient and incapable of high power output or are relatively narrow-band amplifiers. Some of the prior art proposals that initially appeared to be promising have failed inpractice to yield any'useful power gain. 4 Briefly stated, in accordance with certain aspects of this invention, a new type of electron tube, for amplifying electric signals at microwave frequencies, has parallel input and output wave-transmission structures each adapted to transmit electromagnetic waves from one t0 the other of its two ends. ln a preferred embodiment the two wave-transmission structures are two electrically conductive helices extending parallel to each other. These two helices are appropriately called the input helix and the output helix.' Input electric signals are supplied to input terminals at one end of the input helix and travel alongthe length` of the input helix with velocities Vdetermined chiey by the helix geometry according `t0 principles that are well-known to those skilled in the art. Electromagnetic waves induced in the output helix, in

a manner hereinafter described, travel in the same di-y words, the two helices constitute parallel input and output slow-wave structures.

An electron stream, provided by suitable electron gun means, flows closely across the input and output helices in sequence, in interacting Arelation to the electromagnetic waves transmitted thereby. The waves transmitted by the input helix velocity-modulate the electron stream, and the modulated stream supplies energy to Waves transmitted by the output helix, so that the waves onthe output helix progressively increase in energy as they travel toward the output terminals of the tube.

A transverse dimension of the electron stream extends for. a considerable distancetmany wavelengths) along the length of each helix so that different transverse portions ofthe electron stream are modulated by waveson different lengthwise segments of the input .helix with .dilferentinstantaneous,phase-angles. Consequently, the

Patented Mar. s, 1960 ever, .since the waves transmitted in a forward direction on both helices travel in the same direction at substantially the same velocity, all of the wave components induced on the output helix arrive at the output terminals n of the tube in additive phase relation.

.In other words, electromagnetic wave components traveling in the forward direction on the output helix are reinforced and progressively increased energy-wise by additional induced wave components in each lengthwise portion of the output helix. Therefore, a growing wavetravels along the output helix toward the output terminals of the tube. On the other hand, any wave components that may travel in a backward direction (away fromrthe output terminals) along the output helix are somewhat out-of-phase with the wave components induced lin successive lengthwise segments of the output helix, and are progressively attenuated. Consequently, this type of Vbackwardwave interaction is no problem in the Vnew tubes. Other types of backward-wave interaction likewise present .no problem, forreasons hereinafter discussed.

In a preferred embodiment the electron stream has.k

the form of a planar sheet of moving electrons, produced by an electron gun comprising an elongated strip cathode and focusing electrode; or, two such sheets of electrons ample, by a plurality of electron guns having small-area f cathodes.

In the description thus far, interaction between -the electron stream and the traveling electromagnetic waves has Vbeen assumed without specifying the mechanism whereby adequate interaction is accomplished. This is a crucial point at which certain prior art proposals have failed. If the amount of interaction is too small the tube willnot yield a useful microwave power gain.

In tubes constructed according to the present invention, interaction between theelectron stream and the traveling electromagnetic waves takes place in a manner somewhat analogous to the interaction principle of conventional traveling-w'ave tubes. That is, each of the two wave-transmitting structures is a slow-wave structure adapted to transmit electromagnetic waves having at least one wave component with a phase velocity that is small compared tothe velocity of light. A conductive helix is a Well-known example of such a slow-wave structure. lheV direction and velocity `of electron ow in the electron .stream is so chosen that the electron flow has a velocity component in the direction of said phase velocity 'and'at approximately the same speed so that electrons 'f passing through the RF field of the transmitted waves remain in approximately constant phase-positional relation to said wave component over a distance larger than a microwave wavelength, preferably several waveand output helices (but usually not intersecting the 'helix axes) so that the electron flow has a velocity component .in a direction of low-velocity electromagnetic wave propagation on both helices. This velocity component is made approximately equal to and slightly greater Vthan the slowwave velocity in the same direction by adjustments herein described, vto 4establish and maintain the well-known 4VV phase-positional relation required for interactions of the traveling-wave 'amplifier type.

Preferably, the two helices are ilattened in a plane parallel to the electron stream (as viewed from one end, each helix has a race-trac shape) so that the electron stream can pass close to a large portion of the attened cylindrical surface of the helix. Thus, the electron stream travels for a considerable distance (several wavelengths) in close proximity to each helix, and therefore in relatively strong portions of the RF eld associated with each of 'the transmitted electromagnetic waves. Consequently, a fairly strong interaction of the traveling` wave 'amplifier type zis 'achieved Ybetween the electron stream and the electromagnetic waves transmitted by each of the two helices.'

The interaction soobtained differs greatly from the interaction produced Yby directing an electron stream Vacross a single interaction gap in klystron-like structures and simpleslotted waveguides, not only quantitatively but also in basic principles Vand consequences of fundamental importance tothe operation of the amplier as a whole.

-In fact, each of the two wave-transmission structures, Yin association with the electron stream, is in itself, without regard to the other wave-transmission structure, a traveling-wave amplierhaving a low power gain, preferably unity power gain. While such a structure, standing alone, may not produce a useful power gain, the two structures and the electron stream in combination, according to the principles herein disclosed, can yield large power gains and supply a large microwave power output at considerably higher etliciencies than have been realized with prior broad-band amplifiers at high microwave frequencies.

The electron paths in the'new tubes are usually shorter than in conventional travelingfwave'tubes. More irnportant, 'the cross-sectional area of the electron streams in the new tubes is relatively large so that large electron currents can lbe obtained without resort to extremely high current densities. Consequently, the space-charge density in the electron 'streamis relatively low; and this fact, in combination with 'the shortness of the electron paths, makes it easy to solve the beam-'focusing problem. Also, the electron vgun construction is .simplified in certain rev spects since a large-area cathode is feasible and extremely high rates of electron emission from small cathode areas are not required. Furthermore, the elimination of baci;- ward-'wav'e interaction problems makes possible the use of larger helices kfor any selected frequency range 'and thus makes possible a better frequency response'and therefore greater bandwidth, especially Vrat the higher microwave frequencies.

The. foregoing and other aspects of this invention may be better understood from the following description 'of illustrative examples taken in connection with the ac- 'companying drawings. The 'scope of the invention is defined by the appended claims.

In the drawings:

Fig. 1 is a somewhat schematic lengthwise sectionV of an amplifier tube embodying principles of thiswinvention;

Fig. '2 is a transverse section taken generallyalong the line 2-2 of-Fig. l; and

Fig. 3 is a transverse section taken generally along th line 3 3 of Fig. l.

Referring tothe drawings, an evacuated metal envelope Imay consist of a hollow, non-circular cylindrical side por vltion l and two end plates Z and 3, shaped and joined together as shown. An upper vportion of the lenvelope serves asia collector velectrode for the electron streams hereinafter described, and is provided with a plurality of Vcooling tins '4, 5, T6, 7 and .8 for dissipating the heat generated by electrons vstriking the envelope.` Extending lengthwise 'within the evacuated envelope there are two parallel conductive `helices 9 and 't respectively wound lupon and supported by strips .of .insulation-1t .and 12.

The two helices are flattened in a vertical plane lon each side so that when viewed from one end each helix has a narrow race-track shape, as Yshown in Figs. 2 and 3. If desired, for reducing dielectric loading of the helices or for other purposes, the sides of strips 11 and 12 may be recessed, or each strip of insulation may be replaced by two parallel non-conductive rods, to reduce the amount of dielectric material in immediate proximity to the flattened sides of the helices. Also, various other supporting structures may be substituted for strips'll and 12 to support the helices within theY envelope.

Each of the two helices is a slow-wave structure capable of transmitted electromagnetic waves with a phase velocity, in the direction of the helix axis, that is small compared to the velocity of light. The size and pitch of the helices are chosen in accordance with principles well-known to those skilled in the art to give optimum transmission characteristics in the frequency range at rwhich the amplifier tube is designed to operate. In general, the helices will be made relatively large when optimum operation at relatively low microwave frequencies 4is desired,and will be rnade relatively small when operaconventional traveling-wave tube designed for operation in the same frequency range. The reason for this is that the avoidance of undesirable backward-wave oscillations imposes a major limitation upon the helix size inconventional traveling-wave tubes', as is well known. In tubes according to the present invention this limitation on the helix size can be disregarded and accordingly the helix can be made larger. This is an important advantage of the new tubes, because very small helices are dispersive, and this fact detracts from the frequency response of conventional traveling-wave tubes at the higher frequencies.

The left end of helix 9 (as viewed .in Fig. l) is connected or otherwise coupled to input terminals for receiving an electric signal that is to be amplified so that an input microwave signal can be introduced onto the left end of helix 9. For example, helix 9 'may be directly connected as shown, or coupled by Vany of numerous microwave coupling arrangements known to those skilled in the art, to the center conductor13 of a conventional coaxial cable connector that comprises, in addition to the center conductor 13, an outer 'conductor or shell 14 attached to and connected to the metal envelope of the tube, anda non-conductive vacuum seal 15. Thus, an inputl microwave signal can be supplied to input terminals 13 and 14 of the tube by a conventional coaxial cable (not shown) attached to the cable connector. Any input signal so introduced travels from left to right along the length of helix 9, with a phase velocity in the direction of the helix axis that depends chiefly upon the helix geometry and is small compared to the velocity of light. Because it transmits the input signal, helix 9 may be called the input helix.

The right end of helix 9 is provided with a low-reflec- `tion RF termination so that microwaves reflected at the right end of helix 9, and thereafter traveling from right to left toward the input terminals, will be substantially t helix 10 are mostly absorbed by attenuator 20, and waves reduced in energy. A low-reflection termination for the helix can be provided easily by means of'a conventional attenuator 1d, which may be formed, for example, simply by spraying a thin coating of lossy conductive material (e.g., aquadag) on the right end of insulating strip 11 reflected at the left end of helix 10 are substantially reduced in energy. A DC connection 21 between theleft end of helix 10`and the metal envelope-'of the tube maintains helix 10 at'substantially the same DC potential as -the tube envelope. Since helix 10 transmits signals to the output terminal it may be called the output helix. It will be noted that the principal microwave components transmitted by both helices travel in the same direction,

vfrom left to right asviewed in Fig. 1.

A conductive metalstrip 22 extends lengthwise between and parallel to the input and output helices, as shown.

Strip 22 is connected to the `metal envelope of the tube and forms an RF shield and grounding strip that minimizesv direct electromagnetic coupling between the two helices. A conductive metal bar 23 connected to the metal envelope and disposed below and parallel to helix 9 also acts to some extent as an RF shield and grounding strip. In addition, bar 23 which is preferably made ofy ferromagnetic material serves as a magnetic ux concentrator for purposes hereinafter explained, and serves as a first anode or accelerating electrode for the electron gun of the tube. g

An elongated cathode 24 is heated by a conventional heater filament 25 or by any other suitable means. A conventional oxide'coating or the like, characterized "by copious thermionic emission of electrons when heated, is coated in two long parallel strips 27 and`28 on the upper surface of the cathode. Disposed above the cathode is a W-shaped focusing electrode 29 having lengthwise slots alined over electron-emissive strips 27 and 28, as shown. The cathode, heater, and focusing electrode are supported at one end by pins embedded in an insulating base 30, and are supported at their other end by leads extending' through an insulating 'vacuum seal 31 in the tube envelope.

Current is supplied through the heater` lead 32, connected toone end of filament 25, and another lead l(not visible in the drawing) connected to the other end of filament 25, in a customary manner for heating the cathode suiciently to produce thermionic emission of electrons by strips 27 and 28. A negative DC potential supplied through cathode lead 33 (by a conventional voltage supply, not shown) maintains the cathode at a negative potential relative to the metal envelope of the* tube and bar 23. Consequently electrons emitted by the cathode are accelerated by the voltage between cathode 24v and bar 23, which acts together with the proximate portions o-f envelope' 1 as a first anode or acceleratingelectrode of an electron gun, and ow to the upper part of the metal envelope, which acts as a collector electrode'. A focusing potential is suppliedthrough lead 26 to focusing electrode 29 by a conventional DC voltage supply, not sh'own. The value of this focusing potential may be adjusted in a well-known manner to` focus the electrons emitted by strips 27 and 28 into two substantially planar and parallel sheets of electron ow that pass closely adjacent to opposite fiat sides of helices 9 and 10. In Fig. 2 the boundaries of one sheet of electron flow are represented by br'oken lines 34 and 35, while the boundaries of the other sheet of electron flow are represented by broken lines 36 and 37. Similarly, in Fig. 3 the bound aries of the two sheets of electron tiow arerepresented by broken lines 34', 35', 36 and 37'. In Fig. l broken lines 38 through 45 represent typical electronV paths or rays in the Sheet 0f decima 110W wanatiafratmip 2%.

long.

In Fig. 1 it will .be noted that the electron-paths Arepresentedby broken lines 38 through 45 bend as they` pass bar 23 so that the electrons Vflow past 'the input and output helices, successively, at an acute angle to the lengthwise axial direction of the helices. This bending of the electron paths `is produced by a magnetic fiel-:l supplied by a permanent magnet 46. Magnet i6 has two elongated poles that vextend along opposite sides of the tube envelope, in approximately the positions shown in Fig. 2, for a distance somewhat greater than the width of the electron streams as viewed in Fig. l. Magnet df: provides a magnetic field that passes transversely through the tube, including that portion of ferromagnetic bar 23 lying between the two sheets of electron flow. Bar 23- is made of ferromagnetic material to concentrate the magnetic flux in a desired regionof the electron streams. The tube envelope is generally not ferromagnetic but, if desired, it may have Vferromagnetic inserts adjacent to the magnet pole pieces to assist in concentrating the magnetic linx.

As the electrons move through the magnetic field pro-l vided by magnet d6 and concentrated by bar Z3 the electron paths are 'bent in a plane perpendicular to the field, according to well-known principles. The amount of bending, and therefore the angle between the electron paths andthe axial direction of the helices, depends jointly upon the strength of the magnetic field and the electron velocities. The optimum magnetic field strength for a specific tube design vcan be determined from straightforward calculations, which will become obvious to those skilled in the `art from an understanding of the principles disclosed in this specification.

VIn actual tube designs the magnetic field strength required to bend the electron beam as herein disclosed is usually small compared to the magnetic fields required for focusing a high-density electron beam in a conventional traveling-wave tube. Furthermore, in the present instance the flux lines are reasonably short and the reluctance of the magnetic circuit is relatively'low, whereas in a conventional traveling-wave tube the linx lines are relatively llong and the reluctance of the magnetic circuitis high. Consequently, the magnet required with the present tubes is small and inexpensive compared to the' focusingV magnets commonly used with conventional traveling-wave tubes.

There are two principal reasons why the electron paths in the tube under consideration are slanted with respect tothe lengthwise ial direction of the helices. The electromagnetic waves transmitted by the helices have slowwave components with phase velocities, in the lengthwise axial direction-of the helices, that are small compared to the velocity of light. Therefore, these phase velocities are capable of being approximately matched by easily attainable electron velocities. With the slanting electron paths herein described, there is a component of electron velocity in the same direction as the aforesaid phase velocity of the microwaves transmitted on the helices. By

. adjusting the electron velocities, or the angle of the electron paths, or both, the component oi"k electron velocity in the lengthwise axial direction of the helices can be made approximately equal to, `and slightly greater than, the phase velocity in the same direction of a transmitted inicro'wave component. This is the firstl reason for bending the electron paths.

After the 'electrons have'been given a velocity component that approximately matches the aforesaid phase velocity, interactions of the traveling-wave amplier type take 'place between the electron stream and the microwaves transmitted by each helix; provided that the interaction time (the time that each interacting electron, on the average, spends in the microwave fields.) is sufficiently The length of the interaction region between each electron path 'or ray and the RF microwave field associ- 'ated "with 'each helix isa 'function of the angle between the electron `,paths and lengthwise direction of the helix. If lthis angle were made very small, by using a strong fenbughlmagnet to vbend. the electron paths by almost :90 degrees, la vlong length of Veach electron patti would be in a position to interact with `the electromagnetic wave transmitted on one of the helices, the interaction tirne would be relatively large, and a'correspondingly high degree of interaction would be obtained. With larger angles between the electron paths and the lengthwise direction of the helices the distance along each path wherein interaction can occur is shorter, the interaction time is reduced, and the total interaction between the electron stream and the electromagnetic waves is smaller. Thus, within reasonable limits any desired degree of interaction between the electron stream and the microwaves can be achieved by adjusting the direction of the electron paths relative to the helices. This is the second reason Vfor bending the electron paths.

From the foregoing considerations -it is apparentV that each `of the input and output helices, in association with the electron stream, is in some respects like the conventional traveling-wave amplifier. The electromagnetic wave transmitted along each helix interacts continuously f with and may gain some energy from the electron stream,

without regard to effects produced by the presence of the' able power gain can be realized with a single wavetransmission structure.

In tubes constructed according to the present invention each ray of the electron stream passes close to a relatively short lengthwise portion of each helix and interacts with a transmitted electromagnetic wave component over a correspondingly short distance. Other rays Vpass close to and interact with the electromagneticV wave cornponents transmitted on other short lengthwise portions of each helix. Considering each of the two helices alone, and neglecting edects due to presence of the other, each helix in combination with the electron stream can be thought of' as a pluralityof very short traveling-wave amplifiers connected in series. Because of their short lengths, each of these hypothetical ampliiiers has a low power gain. Preferably for broad-band amplification the length ofk each is made such (by adjusting the angle of the electron paths) that each hypothetical amplier has approximately a unity microwave power gainthat is, the energy gainedby the electromagnetic waves from interaction with the electron stream just balances circuit losses.

If each of the hypothetical short amplifiers operates at unity power gain then a plurality or" such amplifiers connected in series must operate at unity power gain. An electromagnetic wave component introduced at any point neither gains nor loses substantial energy as it travels along the helix. Thus, a signal supplied to the input terminals travels along the length of the input helix at a Y substantially constant energy level and equally modulates tion Vis provided at the. .beam `entrance end ol' the wavetransmitting structure. In the embodiment illustrated, non-reflecting terminations are provided by the two attenu'ators 16 and 2'0, 'located -atencls vof ythe helicesoutside y9 of the region where interactions occur between the elec'- tromagnetic microwaves and the electron stream. There is no need for attenuation and attendant power losses in those portions of the helices where interactions with the electron stream occur. This in itself is a significant advantage over conventional traveling-wave tubes.

In actual practice adjusting the angle of the electron paths in the embodiment herein illustrated can be accomplished in a very simple manner. vAll that is necessary is to adjust the magnitude of the DC negative potential supplied tothe cathode, thereby to adjust the electron velocities. Depending upon the strength` and vertical thicknessfof the magnetic field a certain minimum electron velocity, and therefore a certain minimum accelerating voltage, is necessary to carry the electrons through the magnetic field and across the two helices. In other i,

words, the total velocity imparted to the electrons by the accelerating voltage must exceed the horizontal component of velocity imparted to the electrons by their passage through the magnetic field so that there will be a remaining vertical component of velocity sufficient to carry the electrons transversely across the two helices in succession. As the electron velocity is increased beyond Y this minimum value, by making the cathode more negative relative to the tube envelope, the kangle between the electron paths and the lengthwise direction of the helices' increases, and 'the lengthof each path that can provide interactions with the microwave fields, and therefore the interaction time, decreases. 1

The horizontal velocity component ofthe electron flow, in a direction parallel to the longitudinal axes of the helices, is chiefly determined by the magnetic field and is but slightly if at all affected by moderate changes in the total electron velocity. Consequently, once the 'Since the horizontal component of velocity that isf imparted to the electrons by theirpassage through the magnetic field is proportional to the magnetic field strength, the phase-positional relation necessary for v traveling-wave interaction between -the microwave fields vand the electron stream-is achieved simply by adjusting the magnetic field strength. This adjustment preferably is made during manufacture of the-tube by adjusting the strength of permanent magnet 46 but, if desired, means for adjusting the magnetic field strength during use of the tube can be provided either by substituting an electromagnet for the permanent magnet or by providing a magnetic shunt or the like for adjusting the effective strength of the permanent magnet. j

j While each ofthe two vhelices in association with the lengthwise portion .of the output helix, establishing there.`

magnetic energy. Tubes constructed according to the Y present invention are capable of accomplishingthis conversionto a greater degree than has been achieved in conf ventional` traveling-wave tubes and therefore the new -tubes are more efcient power amplifiers.

For purposes of analysis, assume that the two sheet of electron fiow are broken upA into'separate parallel electron beams and that each of the broken lines 38 through 45 in Fig. 1 represents a separate electron beam. The input signal traveling from left to right on input helix 9 velocity-modulates each of these electronrbeams by substantiallyequal amounts. Furthermore, the modulation of beam 39 is delayed with respect to the modulation of beam 38 by the time interval required for the input wave to travel along the input helix in the direction'of the helix axis a distance equal to the spacing between beams 38 and 39, and the modulation of each following beam is likewise delayed with respect to the preceding beam. A certain amount of electron bunching occurs in each of the velocity-modulated beams as it travels between the input and output helices. As beam 38 passes adjacentto output helix 10 it induces on the -output helix an electromagnetic microwave component that travels from left to right toward the output terminals. The-amplitude of thisinduced signal component maybe quite smally and its RF field correspondingly weak.A Because it is interacting with a weak RF field only a small proportion of the kinetic energy of beam 38 is converted into microwave electromagnetic energy. Therefore, the power ef,- ficiency associated with beam 38 is low.

Beamy 39 induces another microwave component on output helix 10. However, 4since the modulation of beam 39 is delayed with respect to the modulation of beam 38 by the time interval required for a microwave to travel along one of the helices a distance equal to the spacing between the beams, the wave component induced on helix 10 by beam 38 arrives at that portion of the output helix interacting with beam 39 at just the .right time to electron stream is in itself, without the other, a travelingwave amplifier, preferably having a power gainof approximately unity, the two helices and the electron stream" in combination, according to the principles and examples' herein set forth, constitute a microwave power amplifier operable at high gain andlarge microwave output power,

:with efficiencies appreciably `higher than those previously Y attained with broad-band-amplifiers at microwave frequencies. As the inputA signal travels along theinput helix it velocity-modulates successive tranverse portions of the electron stream. The velocity-modulation causes some degree of bunching in the electron stream as it flows towardtheoutput helix. Each transverse Vportion of the'bunched 'streamV transfers energyto -jadiferent be in additive phase relation to the wave component induced on helix 10 by beam 39. Consequently, the two wave components reinforce each other, and produce a correspondingly stronger RF field interacting with beam 39. This stronger RF field extracts more 'energy from beam 39 than was extracted from beam 38, and therefore beam 39 loses a larger proportion of its kinetic energy to the RF field. Thus, the power efficiency associated with beam 39 is higher than the power efiiciency associated with beam 38.

' In like manner each succeeding one 1of the electron beams interacts with a stronger RF field on helix 10 and therefore transfers a greater proportion of its kinetic energy to the electromagnetic wave traveling along the helix. Thus, the efficiency associated with each succeeding electron beam is` somewhat higher than thev efiiciency asso-y ciated with the preceding'beam. l Things continue inthis way until saturation effects set in. EventuallyfthefRF field traveling along helix 10 becomes strong enough at its cyclical peaks to stop the horizontal motion of the electrons. crease in the field strength other factors remaining unchanged, the motion ofat least some electrons would be reversed--that is, these electrons would be accelerated by the RF field and move toward the left of Fig. 1.

action of veach succeeding electron beam would tend to belesseiiicient, instead of more efiicient, than that'of the preedinszbeam- This difficulty .aan .beoovercgme eed.

Upon any further instill higher power output and `higherfefiiciency'achieved, as'will now be explained.` Y

Assume that vsaturation conditions, as above described, are first encountered atelectron beam 4Z. To bemore precise, assume that beam t2 encounters an RF field of such strength that maximum efficiency is associated with beam 42, and any greater interaction between an electron ybeam and the RF eld would lower the efficiency.` Since the electromagnetic'Y wave is increasing in energy as it travels from left to right on helix if) the RF eld on that portion of helix ifi adjacent to beam 43 is stronger than the RF field 'adjacent to beam 452.V However, the RF field strength falls ofi rapidly with increasing distance away from the helix. Therefore, if beam 43 is tilted outward away from helix l@ slightly so that-it does not pass asrclose to the helix as beam 42, beam d3 can be made to pass through an RF field of substantially the same strength as the RF field that beam 42 encounters, and therefore both of theheams fifiand 43 will transfer maximum energy to the electromagnetic wave.

Similarly, beam dii-.will be tilted outward away from helix iii slightly more than beam 43, and beam d5 will be .tilted outward still more. By this means each beam to the right of beam il can be made to transfer maximum energy to the microwave field and the over-all ef fciency of the tube asymptotically approaches the theoretical maximum as the length of the tube is increased. Furthermore, this is accomplished without tapering the electrical characteristics'of the helices, so that uniformly wound helices can be employed, which is a substantial advantage in tube manufacture.

Where theelectron flow takes the form of continuous sheets rather than discrete beams as it does in the ernbodiment illustrated, each transverse portion of the sheet follows the same path that it would follow if it Vwere a separatebearn.Y in other words, the respective planes of the two sheets of electron flow in theembodirnent illustrated are slightly warped so that the two sheets of electrons are parallel and pass as .close as possible to both helices (as illustrated in Fig. 2) toward the left end of the .tube as viewed in Fig. l and are slightly tilted apart (as illustrated in Fig. 3) toward the right end of the tube. This tilting of the electron stream is easily accomplished by a slight bending of the focusing electrode 2.9, as is illustrated in Fig. 3 of the drawings.

A considerable gain in efficiency is accomplished by the means discussed above. If the length of the tube were terminated at the po-int where maximum beam efficiency is first obtained, or if the-tube were continued beyond this length without some form of compensation for the saturation effects, only one of many parallel electron beams, or a corresponding portion of an electron sheet, could deliver maximum energy to the electromagnetic waves on the output helix. Under such Vcircumstances, the over-all efficiency would be comparatively low because of the relatively large number of beams or equivalent portions of a sheet operating at low efficiencies. By extending the length of the tube and compensating for saturation effects in the manner herein disclosed, the number of high-efiiciency beams or equivalent sections of an electron sheet can be increased without increasing the number of low-eftic'iency beams and therefore the overall efficiency can beraised considerably. High efficiencies in the order ofthose obtainable'with narrow-band amplifiers, such as klystrons, ,are thus obtainablein an amplifier capable of exceptionally Vbroad-band operation.

Numerous variations in tube structure can be `made while utilizing some or all of the inventive principles herein disclosed. For example, in the preceding descripv tron guns'rnaybe orientedin the Adesired direction of eleotron flow, at an acute angle tothe lengthwise Vaxial direction of the -two helices.Y VThe magnet for bending the electron beam then becomes unnecessary. Such a modificationmay 'be-quite attractive lfor moderate output power applications. For achieving high output power as well as achieving certain other advantages the illustrated ernbodiment appears preferable at present.

YIn another modification, an elongated strip cathode like that illustrated in Fig. 1 may be employed, but magnet 4o omitted, so that substantially parallel 'sheets of electrons fiow past opposite sides of the two helices perpendicular to the lengthwise direction of vthe-tube. 'In this embodiment the helices are skewedthat is,Y each helix turn extends down one side of the insulating strip with a relatively smalllslope in a left-to-right direction and back up the other side of the insulating strip with a slightly greater slope in a right-to-left direction.V By this means the wave fronts of the transmitted microwaves are slanted relative to the electron fiow, instead of slant'ing the electron iiow relative to the wave fronts. A difiicultywith this arrangement is that the interaction time is likely to be low unless a very large helix is used. But a very large helix transmits microwaves with a low phase velocity and therefore requires a low-velocity electron stream, which limits the output power that can be obtained. Y

According to a still another modication, each helix may be physically divided into a plurality of short helical sections electrically connected in series and each section v of the subdivided helix is then turned by 9() degrees from the orientation of the helices illustrated in the drawing. In other words, the horizontal input helix 9 (illustrated in Fig. 1) would be subdivided into a plurality of vertically disposed short helical sections arranged in a horizontal row and electrically connected in series. Similarly, output helix it? would be divided into a plurality of vvertically disposed helical sections arrangedin ahorizontal row and each vertically disposed section of helix llt) would beV vertically alined with a vertically disposed section of helix 9.

An electron beam is then providedin interacting relation with each pair of vertically alined helix sections. That is, one electron beam would pass through a first short section of helix 9 and a first short section of helix i@ in sequence, another electron beam would pass through a second short section of helix 9 and a second short section of helix liti in sequence, etc. Preferably, each helix section in association `with an electron beamwould be of an appropriate length to yield a traveling-wave power gain of approximately unity, so that input signals would travel from one section to the next of the kinput structurev at a substantially constant power level and any signal component induced in the outputstructure would travel to output terminals at a substantially constant power level. Each electron beam and associated pair of short helix sectionsforrns a low-gain amplifier section but the signals produced in the output circuit of each such amplifier section combine in additive phase relation with signals produced in the output circuits of the other amplifier sections to provide a'substantial power gain in the lover-all structure. Y

Y Numerous other changes and modifications will occur to those skilled in the art. In some applications other slow-wave structures may he found to be equally suitable or even preferabletohelices. In general, however, Vit appears at present that helices are the preferred slowwave structures for broad-band amplification.

The tilting of the electron sheets to compensate for saturation effects, as herein described, is -not essential to` Vother aspects of ythis invention and may beomitted -where the utmost improvement in etiiciency is not required. In some cases other VVcompensating means maybe fernpl'oyed, such as winding helix i0 non-uniformly to @provide :at-decreasing helix impedance from left-to rright-but infgneral thisis'less advantageous than the preferred 'i meanshe'rein described because of-.manufacturing diiculties and other reasons. #The tubes herein described are easily distinguishable from prior well-known microwave [amplifiers as well as from certain unsuccessful proposals made in the past. The new tubes are believed to be a new type of microwave-amplifier in the traveling-wave family.

Traveling-wave tubes using a sheet-like electron flow across a skewed helix have been proposed previously. The present tubes are fundamentally different from such prior tubes since in the present tubes, when operated as broad-band amplifiers in the preferred manner herein disclosed, no attempt is madeto securev much gain from a transverse beam interacting with a single helix. In the present tube broad-band amplification is kachieved through successiveY interactions with two separate slowwave structures. I

Tubes have previously been proposed and sometimes called traveling-wave-tubes wherein a sheet of electrons ows through elongated'interaction gaps in input and output waveguides successively. However,raccord ing to such prior proposals the interaction between the electromagnetic waves and the electron stream occurs in klystron-like fashion across a single interaction gap in each waveguide. The present tubes are fundamentally different from such prior proposals in that the interaction between the electron stream and each electromagnetic wave occurs over an interaction distance of several wavelengths, with a corresponding long interaction time, in accordance with traveling-wave amplifier principles. Consequently, the interactions that occur in the present tubes' are of a different degree of magnitude from those occurring acrossthe single gap of the aforesaid klystron-like prior art proposals and a fundamentally different typeofY operation is obtained with totally different end results'.

Furthermore, the present tubes are not distributed am-v plifiers which utilize control grids and space-charge effects as the primary means for modulating the electron tiow. In the present tubes control grids may be incorporated in the 'electron guns for auxiliary'purposes as they are in conventional Vtraveling-wave tubes but such control grids are not the primary, nor even a necessary, means for modulating the electron stream at microwave frequencies. Modulation of the electron stream and the extraction of energy therefrom both proceed according to the travelingwave amplifier principle as herein explained.

Nor are the present tubesl magnetrons ywhich have crossed electric and magnetic 4fields inthe interaction regions. The present tubesV may use magnetic elds for beam-focusing purposes orrfor bending the beam trajectories in the'manner herein explained but interaction between. the electron stream and the electromagnetic waves does not voccur in crossed"static electric and mag-l netic fields of the magnetron type. Y

Amplifier tubes embodying principles of this invention and in particular the preferred embodiment illustrated Yin the accompanying drawings may be connected in a l yariety of circuits for various purposes.` Connections` for operating the tube illustrated is a .broad-band microwaveamplifier have already been described in this specification; p Many other connections and modes of operationl will be apparent to those skilled in the art. For example, a lmicrowave oscillator can be formed by providing a conventional regenerative feedback circuit between the output and input terminals of the tube. Also, various ineansrmay be employed to l'modulate the microwave signalswhich means may or'may not'include modifications or additions to the internal structure of the tube, such la's'the provision of one or more additional electrodes in the electron gun for varying the electroncurrent in the electron stream. n -fThe illustrated embodiment wherein a magnetic field -is used to vbend the electron paths has certain unusual characteristics that .areof particular interest with respect' can be-provided. Input signals 'are supplied to the left Y to 'the use of the tube as a modulator. As :hereinbefore noted. the-'phase-positional relation between 1 the microwave traveling on the twohelices and the moving electrons in the electron stream is established by the strength andvertical thickness of the magnetic field and is substantially independent, within reasonable limits, of the total velocity imparted to the electrons by the accelerating For example, assume that an alternating'modulating voltage is applied to cathode 24 in addition to theA normal DC accelerating potential. The modulating `voltagevwi1l` have little if any leffect on the horizontal component of .velocity of the electrons in thev bent electron paths and thus will have little if any efect upon the phase-positional relation between the traveling microwaves andthe moving electrons. Thus, the modulating voltage does not appreciably affect the frequency response of the tube, whether it is being used as an amplifier or as an oscillator, or for other purposes. The alternating modulating voltage does produce a periodic variation in the angle between the bent electron paths and the longitudinal axial direction ofthe helices and this produces a periodic variation in thel microwave power gain of the tube which amplitudemodulates the microwave` power delivered at the output f terminals. I

For .broad-band amplificationthe-magnitude of the accelerating voltage preferably is adjusted to give approximately unity power gain in the traveling-wave'amplifier action associated withfeach helix independently of the other, as hereinbeforeexplained. In narrow-band and signal-generator applications this is not necessarily the case. For example, assume that the-accelerating voltage (between cathode Y24 and the tube envelope) is adjusted to a valuejust slightly greater than thev minimum value required to carry the electrons through the magnetic field so that the electron paths cross the two helices at a small acute angle to the longitudinal axial direction of the helices. from the traveling-wave amplifier action associated with each helix, independently of the other, may become sufivcierntly greater .than unity that sustained backward-wave "oscillations are generated on input helix 9. Thus7 microwave oscillations can be generated without resort to ex- 'ternalfeedback circuits and microwave energy can be withdrawn either from terminals 13 and 14 at the left end lolfjlielix 9,v or fromterminals suitably coupled to helix 10. When the output energy is withdrawn directly from helix ward-wave oscillators can be tuned by adjusting the magneticliield strength, and can be amplitude-modulated by adding an alternating modulating voltage to the DC acceler'ating voltage.

If the accelerating voltage is adjusted to a value slightly above the highest value that produces sustained oscillations a narrow-bandamplifier of exceptionally high gain end of helix 9 and output signals can be withdrawn from the right end of either helix. Such an amplifier may 'also be useful as a microwave negative resistance. v

It-should be understood that this Vinvention in its broader aspects is not limited yto specific examples herein illustrated and described, and that theY following claims are intended to cover all changes and modifications withinthe truespir'it and scope of theAv invention. 1 l' -1 Under these conditions the power gain arising` MWhat is claimed is:

-bination'z means enclosing kan-evacuated space; two substantially parallel coextensive slow-wave structures-elec- -tromagnetically isolated from each other and each adapted 1totransmit an electromagnetic wave component with the phase velocity that is small compared to the 'velocity of light; and means for providing within said'evacuated space -an'electron stream iiowing closely across said two slowlwave structures in sequence, said stream having a velocity `component inthe same direction as and approximately -equal to saidsphase velocity, whereby low-gain interactions 4of the traveling-wave .ampliiier type occur between said stream and electromagnetic waves on each of said slowwave Istructures while high-gain interactions occur be- Atween waves on'respective ones of saidstructures'through modulation of the electron stream. f

V2. A `traveling-wave amplifier tube comprising the following combination: means enclosing -an evacuated space; irst .and second slow-wave structures within said space velectromagnetically isolated from each-other and Aeach consisting essentially of a plurality of. conductive helical .sections coupled together in series, each Vof said helical ysections being adapted to transmit an electromagnetic Vwave component with a phase velocity smaller than the velocity of light; means for providing within said Vevacuated space an-electron stream vlowing closely by said iirst and second slow-wave structures in sequence, said stream including a plurality of portions respectively 'following lparallel electron paths each passing close to substantially only one of said helical sections of each slow-wave structure; coupling Vmeans for supplying ,microwave signals to :said 'iirst slow-wavestructure, said signals traveling along said iirst structure from one helical section to'an'other -for modulating the several portions of said stream successively, the so-modulated portions of said stream inducing microwave components in respective helical sections of said second slow-wave structure, the so-induced components traveling along saidseco'nd 'structure from one helical'section to another and combining with one another in additive phase relation, whereby an amplied micro- -wave signal Vis produced on said second slow-wave structure; and coupling means forextracting said amplified signal from said second slow-wave structure.

3. A traveling-wave amplier tube comprising the following combination: means enclosing an evacuated space; 'rst and second conductive helices electromagnetically lisolated from each other and extending substantially vparallel to each other within said space, each of said helices being adapted to transmit microwave electromagnetic eld components with moving wavefronts traveling Vslower than ,the velocity of light; means 'for providing 'within said evacuated space an electron streamiiowing successively through the electromagnetic fields transmitted by said first helix and the electromagnetic fields transmitted by said second helix, said stream having a velocity .component perpendicular to said wavefronts and approximately equal to the velocity of said wavefronts in the .saine direction so that electrons in said stream pass through each of said fields in approximately constant phase-positional relation thereto; input coupling means .for supplying input microwave signals to one end of said iirst helix, said input signals traveling along said first -heiix and successively modulating different portions of said stream, the soiodulated electron stream inducing output microwave signals on said second helix; and out- 16 tromagnetic -waves -Witha phase velocity -in the longitudinal axial direction of the helix that is small compared to the velocity of light; means for providingfwithin said space an electron stream iiowing closely across said lirst and second helices successively, the direction of velectron ilow said stream being at an acute angle to the longitudinal axial direction of said helices, and the velocity of said ilow being such `that said stream has a velocity component in the direction of and approximately equal to said phase velocity, whereby interactions occur between said stream and said electromagnetic waves on each helix such that the wavesV on said iirst helix modulate said stream and the so-modulated stream induces amplified waves on said second helix.

5. A traveling-wave.tube-comprising the following combination: means enclosing an evacuated space; an elongated slow-wave structure adapted to transmit electromagnetic wave components in a longitudinal `direction within said evacuated space at a phase velocity-that is small compared to the velocity of-light; an electron-emissive-cathode disposed within said evacuated space; electrode means for accelerating electrons emitted by said cathode into a stream of electrons iiowing closely across said slow-Wave structure, the initial direction of said iiow being substantiaily perpendicular to said longitudinal direction, vand magnet means for producing a transverse magnetic iield extending perpendicularly through said stream of electrons between said cathode and said slow-wave structure, the strength of said iield having a predetermined value so as to cause said electrons to follow bent paths and to iiow across said slow-wave structure substantially outside of said magnetic field along straight-line paths disposed'at an acute angle to said longitudinal direction, the strength of said magnetic iield further being such that saidelectrons as they pass said slow-wave structure Yhave a velocity Vcomponent .in the same direction as and-Substantially equal to said phase velocity, whereby interac- Y tions of the traveling-wave amplifier type occur between put coupling means for extracting said output. signals said electrons and said electromagnetic wave components.

6. A traveling-wave ampliiier tube comprising the following combination: means enclosing an evacuated space; iirst and second elongated conductive helices 4disposed in .laterally spaced relation within said space with the `longitudinal axes of said helices substantially parallel, said helices having sides flattened in a common plane, each of said helices Vbeing a separate slow-wave structure adapted to transmit traveling electromagnetic waves with a yphase velocity in the longitudinal axial direction -of the helices that is small compared to the velocity of light, an electronernissive strip cathode disposed within said evacuated space substantially parallel to the longitudinal axes of said helices;` electrode means for accelerating electrons emitted by said cathode into a sheet of electrons flowing closely across a Viiattened side of said .iirst helix and a flattened side of said second helix in. succession; magnet means having elongated pole pieces of opposite polarity disposed on opposite sides of said sheet of electrons and extending substantially parallel to the longitudinal -axes of said helices, said magnet means providing a magnetic eld that extends perpendicularly through substantially the entire width of said sheet of electrons in va region betweensaid cathode and said first helix, whereby said electrons follow bent paths and flow across said two helices outside of said magnetic lield along substantially straight paths dicsosed at an acute angle to the longitudinal axial direction of the helices, the strength of said magnetic field being such that said electrons have a velocity component in the same direction as and substantially equal lto said phase velocity, whereby low-gain .interactions of the traveling-wave amplifier type-occur between'said electrons and each of the electromagnetic waves transmitted by respective ones of said helices, while high-gaininteractions occur between the waves transmitted lby respective ones of said helices through modulation of the electron ow; in-

put 'lonnections .for supplying electromagnetic waves .to

one end of said first helix; attenuator means providing a low-reection microwave termination lat the other-.end of 7. A traveling-wave amplifierr tube comprising the'ifollowing combination: means enclosing an evacuated space; two substantially parallel elongated slow-waveA structures each adaptedvto transmit an electromagnetic wave com-..

portent in the lengthwise direction of said structures with a phase velocityzthat is small compared to the velocity of light; means for providing within said evacuated space an electron stream iowing closely across said two slowwave structures in sequence at an acute angle to the lengthwise direction of said structures, said stream having a velocity component in the same direction as and approximately equal to said phase velocity, whereby lowgain interactions of the traveling-wave amplifier type occur between said stream and electromagnetic waves on each of said low-wave structures considered independently of the other, said acute angle being so chosen that said low-gain interaction produces approximately unity power gain; input connections for supplying microwave signals to one end of the one of said two slow-wave structures that said electron stream flows across first, said f microwave signals traveling along the length of that slowwave structure and velocity-modulating different portions a phase velocity that is small compared to the velocity of light, means for providing within said evacuated space an electron stream ilowing closely across said two slow-wave structures in sequence, said stream having a velocity component in the same direction as and approximately equal to said phase velocity, whereby low-gain interactions of the traveling-.wave amplifier type occur between said stream and electromagnetic waves on each of said slowwave structures; input connections for supplying input microwave signals to one end of the one of said two slowwave structures that said electron stream ows across iirst; and output connections for withdrawing output electric signals from the other end of the other of said two slow-wave structures; said input signals traveling along the length of said one slow-wave structurefand velocitymodulating successive portions of said electron stream successively, the so-modulated stream inducing microwave signal components in successive portions of said other slow-wave structure, the so-induced components traveling along said other structure toward said output connections and combining with each other in additive phase relation to produce a growing wave Vthat increases in energy as it travels along said other structure, said electron stream clearing said other structure by an amount that increases progressively toward said output 'connection to prevent over-saturated interaction between the growing Wave and the electron stream.

References Cited in the tile of this patent UNITED STATES l PATENTS 2,694,783 Charles Nov. 16, 1954 2,717,327 'lfouraton etal. Sept. 6, 1955 2,802,136 Lindenmad r Aug. 6,; 1957 2,804,511 lKompfner 1 Aug. 27, 1957 l2,809,320 Adleroct. 8, 1957 '2,811,664 Kazan oct. 29,1957 2,849,643 Mourier Aug. 26, 1958 Y FOREIGN PATENTS 1,003,491 France Nov. 21, 1951 709,842v Great Britain June 2, 1954 ponent in the lengthwise direction of said structures with t

Images