Thirteenth Edition Copyright 2007 L-3 Communications Electron Technologies, Inc.

Transcription

1

2 TWT/ TWTA Handbook

3 The information contained in this handbook/disk is considered to be published information generally accessible or available to the public. It contains basic TWT/TWTA functions and purposes. It also contains general scientific, mathematical or engineering principles commonly taught at colleges and universities. It has been released in the Public Domain through unlimited distribution at conferences, meetings, seminars, trade shows or exhibitions. L-3 ETI would like to extend a special thanks to the engineering staff who contributed a great deal of time and effort in preparing this handbook for publication. Thirteenth Edition Copyright 2007 L-3 Communications Electron Technologies, Inc.

4 TABLE OF CONTENTS TOPIC PAGE Introduction to the TWT...1 The L-3 Communications/TWT Connection...1 General: Microwaves, a High Frequency Range Electromagnetic Radiation...2 Microwave Amplifiers...3 The Components of a TWT...4 How It Works...4 Controlling the Beam...5 Variations on the Slow-Wave Structure...6 Comparison with the Klystron...8 Looking Ahead...10 Specific Applications and TWT Design Trade-Offs...11 Radar TWTs for Airborne and Sea Level Applications...12 ECM TWTs for Airborne and Sea Level Applications...12 The TWT in Space: Communications TWTs for Space Applications...13 Missile TWTs for Active Seekers...13 References...13 Appendix A Glossary of Terms...15 Appendix B Electron Gun and Cathode Design Trade-Offs...37 Appendix C Slow-Wave Circuit and Beam Focusing Trade-Offs...47 Appendix D Collector Design Trade-Offs...59 Appendix E TWT Package Design and Cooling Method Trade-Offs...63 Appendix F Power Supply Interfaces...67 Appendix G TWT Parameters that Affect System Performance...71 Appendix H Factors that Affect Power Combined TWTAs...95 Appendix I Millimeter-Wave TWTs...99 Appendix J Notes on TWTs for Radar Applications Appendix K Notes on TWTs for ECM and Missile Applications Appendix L Notes on TWTs for Space Applications Appendix M Notes on the Care and Handling of TWTs Appendix N Equations and Formulae Commonly Used in TWT Design Index...133

5 Introduction to the TWT The TWT is an electron tube used for amplification at microwave frequencies. Microwave frequencies are generally identified as frequencies above 500 MHz. At microwave frequencies the familiar circuit theory concepts no longer apply and it is necessary to use electromagnetic theory to describe the electric and magnetic fields that exist in electromagnetic waves. The operation of the TWT depends on the interaction of a beam of electrons with an electromagnetic wave. The history of microwave technology is a history of progressive advances in the techniques used to generate, amplify, and process signals at microwave frequencies. Operation at the threshold of the microwave region was provided by triodes that employ special geometries to minimize transit time effects. This was followed by the magnetron and other crossed-field devices, then by the klystron, and today by the traveling-wave tube (TWT). It would be difficult to imagine present-day microwave technology without the TWT and TWT amplifier (TWTA). No other devices can match the TWT s unique combination of bandwidth, power output, and gain. From electronic warfare to space exploration to the relaying of home-video signals, the TWT has expanded the microwave horizon. L-3 Communications Electron Technologies, Inc. (L-3 ETI) has been at the forefront of each of these TWT developments. The purpose of this Handbook is to present an overview of the everchanging TWT technology and the L-3 ETI products which can be used to implement this technology. It is, in a sense, both a history and a prophecy. It will tell you where we are, and point the way toward new TWT innovations and applications at L-3 ETI and in your own development laboratory. The L-3 Communications/TWT Connection The TWT is not a new device. Its remarkable capabilities and some of its potential applications have been known for nearly fifty years. It was invented during the latter part of World War II by an Austrian refugee, Dr. Rudolf Kompfner, while working on microwave tubes for the British Admiralty. The TWT was not utilized during that war and remained an experimental laboratory device until the first practical tube was developed by J. R. Pierce and L. M. Field at the Bell Telephone Laboratories (BTL) in The first details were published in the IRE Transactions for February From 1945 to 1950, most of the development work was done at BTL and Stanford University. By present-day standards, these efforts were relatively low key. BTL, in particular, was interested in the TWT because of its potential application in the communication field. Meanwhile, the military services had other potential applications in mind: radar and electronic countermeasures. The development of radars during World War II had been rapidly followed by the development of countermeasure techniques to deceive and jam them. The evolution of new radars has, therefore, been partially the result of a continuous need to stay ahead of any new countermeasure tactics which might compromise the radar s effectiveness (and vice versa). The trend in search radar, for example, has been toward much higher powers and toward new techniques which would have the effect of increasing visibility even while being jammed. A good anti-jamming radar must 1

6 be able to shift frequency over a wide bandwidth quickly to avoid dwelling at the jammer s source frequency. Similarly, the trend in ECM has been toward wide bandwidth system capabilities where the jammer amplifies wideband noise or may deceptively retransmit the hostile radar pulse to offset the radar s ability to determine the target s position or track. Since wide-frequency bandwidths are essential to the employment of all these tactics, an amplifying device capable of broad operating ranges with sufficient gain, output power and efficiency was needed. The TWT was found to be ideally suited for the task and the military deserves credit for funding many of the early advances in TWT development. Much of this advanced technology work was done at what was then the Hughes Aircraft Company. In the late 1950 s, with the future of the TWT as a key element in a number of application areas assured, a small group of scientists, engineers and skilled technical support people who had been involved in TWT research throughout Hughes were brought into one organization. The organization later became the Electron Dynamics Division (EDD); and with ownership by L-3 Communications, is now L-3 Communications Electron Technologies, Inc. or L-3 ETI. L-3 ETI is the established leader in the development and production of military and commercial TWTs, TWTAs and related subsystems. Some of the earliest successes for L-3 ETI TWTs, in addition to radar applications, were in the area of space applications. L-3 ETI space TWTs and TWTAs have been used in scientific experiments, manned missions, and communication applications by both military and commercial customers. Early programs included Syncom, the ATS series, the Intelsat series, and, more recently, domestic communication satellites both in the U.S. and abroad. To meet the requirements for future space programs, these devices continue to be developed and refined. This work is advancing the state-of-the-art in areas of longer life, lighter weight, higher efficiencies, lower distortion, higher frequencies, and smaller size. L-3 ETI TWTs are also meeting the expanding customer requirements in other application areas, such as radar, electronic counter-measures, ground terminals and microwave instrumentation. In all of these fields, on-going programs for further product refinement and basic research continue to produce devices and subsystems of the most advanced designs. General: Microwaves, a High Frequency Range Electromagnetic Radiation Microwaves exhibit several properties of visible light: 1) they travel in straight lines at the speed of light and are only minimally refracted by the earth s atmosphere; and 2) they can be focused into narrow beams which are subject to complete reflection when they impinge upon a conducting surface. These properties make microwaves especially useful for radar and telecommunications systems. 2

7 Figure 1 The electromagnetic frequency spectrum. Microwave Amplifiers Within the frequency range of interest, two technologies (vacuum tube and solid-state electronics) are used to generate and amplify microwaves. Each offers advantages for specific applications within the performance domain of radio frequency (RF) systems (see Figure 2). Microwave power tubes (the principal product derived from RF vacuum electronics) are preferred for applications requiring both higher frequency and higher power. Electron transport in a vacuum conveys as advantages to microwave power tubes such features as wide band performance, efficiency, high gain, thermal robustness, and radiation hardness. Alternatively, solid-state power amplifiers combine the power from many transistors. The advantages of charge transport in a solid-state media yield compact devices, and competitive efficiency and bandwidth at lower frequency and power. Figure 3 shows the various vacuum devices in the industry within the Microwave Tube general family. This Handbook only delineates the topics related to the Traveling Wave Tube (TWT) family of products. Figure 2 Ranges of applications of microwave TWTs and TWTAs and Solid State Power Amplifiers (SSPA). 3

8 Figure 3 The summary of various types of microwave tube families. The Components of a TWT All TWTs possess four major subassemblies: An electron gun that products a high density electron beam. A microwave slow-wave circuit that supports a traveling wave of electromagnetic energy with which the electron beam can interact. The collector that collects the spent electron beam emerging from the slow-wave circuit. The TWT package, which provides points for attachment to the using system, provides cooling for power dissipated within the TWT, and, in some cases, includes all or part of the beam focusing structure. Other functions may also be included as required. How It Works The basic form of the TWT has changed very little since its invention by R. Kompfner in 1944, although the performance of these devices today is at least an order of magnitude better in all attributes. Amplification in a TWT is attained by causing an electromagnetic RF wave to travel along a propagating structure in close proximity to an electron beam, as indicated in Figure 4. At the left of this simplified diagram is an electron gun assembly. The cathode, when heated, emits a continuous stream of electrons. These electrons are drawn through an aperture in the anode and are then focused into a welldefined cylindrical beam by a magnetic field. The beam is thereby caused totravel inside the slow-wave circuit for the length of the tube. The electrons are finally collected and their kinetic energy is dissipated in the form of heat in the collector. 4

9 Figure 4 Simplified TWT schematic. At the same time that the cylindrical electron beam is moving along the length of the tube axis, the RF signal to be amplified is fed into the slow-wave structure consisting, in this case, of a coiled wire called a helix. The RF energy travels along the helix wire at the velocity of light. However, because of the helical path, the energy progresses along the axial length of the tube at a considerably lower axial velocity, determined primarily by the pitch and diameter of the helix. The phase velocity of the RF wave (the speed at which the phase fronts of the energy appear to move along the length of the tube) is made slightly slower than the velocity of the electron beam. This near-synchronism results in a continuous interaction between the electron beam and the RF signal. Some of the electrons in the beam are slowed by the RF field, while others are accelerated. As the velocity-modulated electrons move down through the helix they form bunches. These bunches, in turn, overtake and interact with the slower helix RF wave, surrendering kinetic energy to the wave on the helix. The result is a cumulative amplification of the RF signal. Single TWTs have been built with power gains of more than 10,000,000 (70 db). Controlling the Beam The electron gun functions somewhat like the lens in a projector. The object is to get as much electron current as possible flowing within a concentrated beam without distortion. Good gun design is extremely important since the gun is the source of electrons for the beam. A wide variety of gun designs have been developed by L-3 ETI in an effort to provide well-behaved electron beams that can be readily adapted to new TWT types. 5

10 Many TWT guns also include control grids or control electrodes that make it possible to turn the electron beam on and off rapidly. This approach to beam modulation uses a much smaller voltage swing than would be required if only the cathode voltage were modulated. The typical grid-controlled gun has six main elements: The gun shell or support structure, which is usually ceramic or a brazed metal-ceramic assembly The heater The cathode or electron emitter A control grid A focus electrode to aid in proper formation of the electron beam An anode which effectively provides the accelerating field for the electrons. Figure 5 shows a typical gun in cross-section. Life and reliability of the end product are largely dependent upon the design and type of cathode material used. Many different types of cathode materials have been used as electron emitters, but two have generally become standard. The first is an oxide type with a nickel base and a barium/ strontium coating. The second is a dispenser type which has a body consisting of porous tungsten material that is impregnated with a mixture of barium, calcium, and aluminum compounds. The impregnants migrate to the emitting surface, dispensing barium and other active materials to replenish material lost through evaporation. A variation of this type of cathode, known as M-type cathode, is coated with a porous layer of osmium to lower the work function and allow a lower operating temperature. Another variation is the scandate cathode, which has a layer of scandium oxide to provide an even lower operating temperature. Another type of dispenser cathode is the coated particle cathode (CPC) which, as the name suggests, is a structure made up of specially coated particles bonded to a nickel support. A choice of a specific cathode type is dependent upon the required beam power (current density) and is a function of individual tube design. Variations on the Slow-Wave Structure Figure 5 A typical grid-controlled electron gun includes a support structure, heater, cathode grid, focus electrode and anode. Although there are many types of slow-wave structures, most are based upon Kompfner s original helix design. The helix is still the widest bandwidth structure available. 6

11 Figure 6 illustrates the principal component parts of a typical metal-ceramic helix TWT. In the illustration, the metal ceramic envelope and PPM focusing structure can be seen in the central portion of the figure. The final assembly, incorporating the balance of the package parts, can be seen in Figure 7. Figure 6 The principal component parts of a typical metal-ceramic helix TWT. Figure 7 Assembled helix TWT. Figure 8 demonstrates the kind of performance characteristics that can be achieved with this type of slow-wave circuit. The extremely broadband performance is ensured by the highly accurate tolerances held during the helixwinding process. This accuracy is essential to the process of interaction between the electron beam and the superimposed RF wave. One reason for limitations on the peak-power capability of helix tubes is that their circuit characteristics are susceptible to backward-wave oscillations when the operating voltage exceeds about 10 kv. In its basic form, the helix is generally restricted to devices having power outputs of less than 3,000 watts. A number of configurations derived from the basic helix structure have been explored at L-3 ETI, extending its properties to provide even higher output powers. Early among these were schemes that involve using two helices wound in opposite directions (the contra-wound helix, as seen in Figure 9a) or in the same direction (the bifilar or folded helix). These devices extend the useful range of operating voltages up to the 20 to 70 kv range and allow larger transverse dimensions at a given frequency range. 7

12 Figure 8 Output power, 658H. Another solution to this problem is the ring-bar tube (Figure 9a), which has distinctly different circuit properties and is not subject to backward-wave oscillations. Ring-bar TWTs are generally designed for voltages in the 12 kv to 30 kv range, with peak-power levels in the order of 10 kw to 20 kw. With sufficiently high voltages, peak-power output levels can be in excess of 100 kw. The coupled-cavity structure, shown in Figure 9b, employs coupled resonant cavities to effectively slow the RF energy. The original coupled-cavity structures provided frequency bandwidths on the order of 10 to 15 percent. Methods have been developed for increasing the bandwidth to 40 percent and more. Tubes utilizing this circuit have been built to produce several hundred kilowatts of peak power at S- through Ku-bands with up to 60 db gain. The inter-digital line is also a version of the coupled-cavity circuit and has found extensive use in low and medium power amplifiers ranging up to one kw peak-power output with gains of about 30 db. Comparison with the Klystron At the input of the TWT circuit, the RF signal level is quite low and the resulting modulation of the electron beam is similar to that of the input cavity of a klystron. In the case of the TWT, however, the circuit is nonresonant in the case of a helix tube and has a low Q resonance in the case of a coupled-cavity structure. The wave actually propagates at a speed close to that of the electrons in the beam. The initial effect on the beam is the introduction of a small amount of velocity modulation that later translates to current modulation. The current modulation then induces an RF current in the circuit. If the proper phase relationships are maintained, this results in an amplification of the signal. The major difference between this mechanism and that of a klystron is that the TWT interaction is continuous over the entire length of the circuit rather than occurring at the gaps of a few resonant cavities. This continuous interaction is the result of a propagating wave, whereas in the klystron the wave does not propagate. In fact, in the klystron there is generally no coupling between any of the cavities except that afforded by the modulation on the electron stream. 8

13 9a Helix-derived circuits. 9b Cross-section of a coupled-cavity TWT circuit and the assembly method. Figure 9 Various alternative traveling-wave circuits are used, each optimized for a particular application. The question can be asked, however, whether any real difference exists between a very narrowband TWT and a broadband klystron, both of which can indeed possess the same bandwidth performance. The answer is that in a true klystron, the wave does not propagate. Each cavity operates independently and in complete isolation from all other cavities. There are exceptions to this rule in the form of hybrid configurations in which the pure klystron concept is 9

14 significantly modified. But these cases do not alter the basic distinctions between the two devices. Obviously, then, the single most powerful attribute of the TWT is bandwidth. Although there are applications for TWTs where the bandwidth requirement is very small, by and large the primary impetus for their continued development has been applications where the bandwidth is 10 percent or more of the center frequency. Another advantage intrinsic in the TWT amplification process is that extremely large gains in the neighborhood of 60 db can be realized with little sacrifice in bandwidth or any of the other desirable properties of the TWT design. Because the gain-bandwidth product is not the result of an unpleasant tradeoff, as is often the case in other microwave amplifiers, there is no reason to be limited by any such figure of merit. Instead, the gain of a TWT is an exponential function of the interaction length. Each incremental increase in length produces the same incremental increase in decibels of gain. The fact that TWT gain in decibels is directly proportional to length gives the TWT a distinct advantage over crossed-field amplifiers where the gain dependence on length is much less favorable. Looking Ahead The TWT is now a mature device, being developed and manufactured by a mature industry, geared to the hard economics of the marketplace. There is, therefore, no reason to look forward to dramatic breakthroughs in the major characteristics of the TWT device. There will be, however, a steady and significant improvement in its performance, reliability, adaptability, and cost. Moreover, it would only be fair to state that the ultimate capabilities of the TWT in terms of bandwidth, power output, efficiency, size, and signal fidelity have not yet been fully exploited by present-day systems. In particular, the efficiency continues to be improved with the latest devices achieving levels of 70 percent and greater (see Figure 10). This situation will undoubtedly change as the systems of the future are pushed toward better performance without corresponding increases in size or complexity. Even though new classes of TWTs are not likely to appear in the foreseeable future, the current effort to improve efficiency, linearity and powerhandling capability will cause a measure of excitement in the industry. The mix of devices will also change as advanced systems replace some of the more obsolete equipment. At the lower power levels, there will be ever increasing competition from solid-state devices. More emphasis will be given to making TWTs more adaptable to the modern airborne environment, which requires compactness and high operating temperatures. 10

15 Figure 10 Space TWT efficiency since Life and reliability will also be given a great deal more attention in the future. The real advances in this area will come as a result of conservative and skillful designs, not as a consequence of legislation and specification writing. Improved cathodes will permit TWTs to be designed for longer life and for higher output power at the higher frequencies. Improved and analytical techniques have resulted in TWTs having higher efficiency and improved linearity (reduced distortion), primarily through the use of improved velocity tapers at the output end of the slow-wave circuit. TWTs having 70 percent efficiency are now in production, and many of these TWTs offer less than 30 degrees of phase shift as the RF drive level is varied from small signal to saturation. The next few years will lead to additional improvements in power output, bandwidth, and efficiency, causing the state-of-the-art curves to move up by a half an order of magnitude. Specific Applications and TWT Design Trade-Offs The design of a TWT originates with the requirements to provide certain amounts of gain and power over a certain frequency band. The final design of the TWT usually results from a trade-off study that includes considerations involving the power supply (sometimes called the EPC, the electronic power conditioner) and interfaces between the TWT and the using system. These considerations lead to trade-offs that affect each of the major subassemblies of the TWT. Those considerations include: The type of slow-wave circuit to be used in meeting the power and bandwidth requirements, including the selection of cathode voltage and current to be used in meeting those requirements. It is important to note that the higher thermal dissipation capability in coupled-cavity TWT circuits can provide two orders of magnitude greater power output capability than available from TWTs having helix circuits, at the penalty of increased size and weight. The method to be employed for focusing the electron beam. 11

16 The method to be used for varying the beam current, including the method used for turning the TWT on and off as well as any modulation required during TWT operation. The operating life requirements. The environmental conditions under which the TWT will operate (ambient pressure, ambient temperature, shock and vibration levels, etc.). The type of cooling available. Size and weight limitations. Cost. The final design of the TWT can be affected as much by the priorities assigned to the design considerations as by the considerations themselves. It is not uncommon to find that TWTs designed for similar frequencies and power levels may differ widely in design. For example, consider a TWT designed for use in an aircraft where dielectric cooling fluid is available. The TWT might well use that fluid both for cooling and as a dielectric around high voltage regions. Alternatively, that same design could not be used in an application where the cooling fluid is not available or is not a good dielectric. The major applications for TWTs include: Radar TWTs for airborne and sea level applications. ECM TWTs for airborne and sea level applications. Missile TWTs for terminal seekers. TWTs for space applications, especially for communications satellites but also for radar and space probe applications. Instrumentation TWTs for laboratory and test equipment. The design approaches relevant to each application will be discussed. Radar TWTs for Airborne and Sea Level Applications The features that influence the design of these TWTs include: High peak power with moderate average power. Moderate bandwidth Cooling can usually be conduction, forced air, or circulating liquid. Operation at altitude usually requires that high voltage leads are encapsulated or enclosed in dielectric fluid. Operation in extreme environments requires rugged construction. These requirements are usually met by using TWTs that have either helix or coupled-cavity slow-wave structures and solenoid focusing, although the latest available magnet materials are allowing some of these devices to utilize periodic-permanent-magnet (PPM) focusing structures. ECM TWTs for Airborne and Sea Level Applications ECM TWTs tend to fall into two categories, those that require high peak power and those that require high CW (continuous wave) or average power. In most cases, the bandwidth must be wider than for radar TWTs because the exact frequency of operation is not defined until the threat is identified. Helix or helix-derived circuits are essential for those applications requiring octave or multiple octave bandwidths. Coupled-cavity circuits are required when the power requirement exceeds the capability of helix-type circuits and the reduced 12

17 bandwidth can be tolerated. Permanent magnet focusing is widely used in these devices. The TWT in Space: Communications TWTs for Space Applications Features that influence the design of these TWTs include: Long life and high reliability High efficiency Moderate power output (usually less than 300 W) Moderate bandwidth (1 percent to 5 percent) Low phase and gain distortion (high linearity) Well-controlled manufacturing facility and manufacturing documentation These features usually dictate that the TWTs have helix slow-wave circuits (usually with a velocity taper to improve both linearity and efficiency), high area compression ungridded electron guns, PPM focusing, multiple stage depressed collectors, and conduction cooling (in some applications, the collector is cooled by direct radiation into free space). The use of a linearizer is often dictated by the linearity requirements. Linearized TWTAs (LTWTAs) provided by L-3 ETI represent an integrated design wherein the designs of the TWT, EPC and linearizer are closely coordinated. Missile TWTs for Active Seekers Features that influence the design of these TWTs include: Minimal size and weight Narrow to moderate bandwidth Off to fully operational turn-on times of 1 second or less High efficiency High reliability after long inactive storage periods These TWTs are normally of the periodic-permanent-magnet (PPM) focused helix variety. They normally utilize unique cathode-heater designs to provide the very fast warm-up required. They typically have multiple stage depressed collectors with conduction cooling. References 1. Theory and Design of Electron Beams, J. R. Pierce, D. Van Nostrand Company, Inc., Traveling-Wave Tubes, J. R. Pierce, D. Van Nostrand Company, Inc., Electron Optics, O. Klemperer, Cambridge at the University Press, Fundamentals of Microwave Electronics, M. Chodorow and C. Susskind, McGraw-Hill, Microwave Tubes, A. S. Gilmour, Jr., Artech House, US TWTs from 1 to 100 GHz, Microwave Journal, 1989 State of the Art Reference. 7. Principles of Traveling Wave Tubes, A.S. Gilmour, Jr., Artech House

18 APPENDIX A GLOSSARY OF TERMS

19 AM near-carrier noise AM noise and spurious outputs at frequencies close to the carrier frequency, usually 100 Hz to 500 KHz relative to the carrier. Components not related to the presence of the carrier (signal) are usually caused by ripple on power supply voltages. Interactions in the TWT are usually responsible for components that exist only when a carrier (signal) is present. (See also, Near-carrier noise.) Amplitude pulling factor The amount gain change experienced when the VSWR of the input source or the load is changed. Amplitude pushing factor The amount of gain change experienced as the voltage on a particular tube element is varied, usually expressed in db per Volt. AM/AM conversion An abbreviated form of the phrase amplitude modulation-to-amplitude modulation conversion, used to describe a type of distortion in microwave amplifiers. It is the change in the output RF voltage produced by variations in input signal level, usually expressed in db/db. An ideal amplifier has an AM/AM conversion of 1 db per db. As the amplifier approaches its saturated output power (the power output at the peak of the RF output versus RF input curve) the amplification process becomes nonlinear and gain compression occurs. At saturation, the AM/AM conversion is 0 db per db. (See Figure A-1.) Figure A-1 Typical power output and phase shift as a function of RF input power for a communications-type TWT. 16

20 AM/PM conversion An abbreviated form of the phrase amplitude modulation-to-phase modulation conversion, used to describe a type of distortion in microwave amplifiers. It is the change in phase angle of the output RF voltage produced by variations in input signal level, usually expressed in degrees/db. An ideal amplifier has an AM/PM conversion of 0 degrees per db. (See Figures A-1 and A-2.) Figure A-2 For a single carrier condition, AM/PM conversion rises sharply as drive is increased. AM/PM transfer An abbreviated form of the phrase amplitude modulation-to-phase modulation transfer, used to describe a type of distortion in microwave amplifiers. It is the change in phase angle of the output RF voltage on one signal produced by variations in input signal level on a signal at another frequency, usually expressed in degrees/db. An ideal amplifier has an AM/PM transfer of 0 degrees per db. Anode 1. In a klystron or a TWT, a positively charged electrode; the stream of electrons that leaves the cathode flows toward the anode. 2. In a gridded vacuum tube (such as a triode, tetrode or pentode), it is called a plate. 17

21 3. In a cathode-ray tube, the anodes are connected to a positive potential source. The anodes concentrate and accelerate the electron beam for focusing. Average Power 1. A value of power equal to the time integral of the instantaneous power over one period, divided by the period. The period is the time for one cycle; for pulsed signals, the period is the reciprocal of the pulse repetition frequency. In other words, the average is obtained by finding the area under a plot of power versus time and dividing the result by the time for one cycle. For pulsed signals, average power is equal to the peak power times the duty cycle, where the duty cycle is the ratio of the on time to the period. 2. In circuits containing reactance and resistance, the current and voltage values may have different phase angles and must be represented by vector quantities. The average power is the result obtained when the vector quantity E is multiplied by the in-phase component of the vector quantity I. Backstreaming A condition in which a portion of the electron beam is reflected from the collector and travels backward toward the electron gun. This is an undesirable effect, distorting the primary electron beam and any modulation that may be present. Backward wave oscillator Abbreviated BWO. A wideband voltage tunable oscillator related to a traveling-wave tube in somewhat the same way that a klystron oscillator is related to a klystron amplifier. It uses a broadband circuit similar to a TWT. Regeneration exists because the backward wave mode on the circuit propagates RF energy in a direction opposite to that of the electron stream, oscillation occurs when the loop gain is equal to or greater than unity. Higher power TWTs can experience undesired backward wave oscillations when the backward wave mode has sufficient regeneration. Beam efficiency The RF output power divided by the beam power (cathode current times cathode voltage). Body/helix protection A combination of circuit elements in the TWT power supply, ranging from a simple resistive network to a complex crowbar device. The circuitry is designed to prevent damage to the TWT slow-wave structure as the result of arcing or unusually high intercept current. Brillouin field See Brillouin flow. Brillouin flow A magnetic focusing scheme for electron beams in klystrons and TWTs in which the magnetic field that provides focusing is parallel to the direction of electron flow and the electron gun is shielded from magnetic field. The amount of 18

22 magnetic field to focus a non-thermal beam is called the Brillouin field (a nonthermal beam is a beam that contains none of the random components of velocity that result from emission from a hot cathode). In practice, the presence of thermal velocities requires that the actual field be somewhat greater than the Brillouin value. (See also, Confined flow, PPM.) Cathode A negatively charged electrode which emits electrons. In a TWT this is a thermionic emitter where electron emission results from operation of temperatures in the vicinity of 1000 C. Cathode loading The current density at the emitting surface of the cathode usually expressed in amperes per square centimeter. (See also, Current density.) Charge density The amount of electric charge contained in a unit of volume, usually expressed in Coulombs per cubic meter. Charge density can be calculated from current density divided by the velocity of the electrons. Cold match The input or output match of a TWT obtained when the TWT is not operating (the TWT is electrically cold ). Collector 1. In a TWT or a klystron, the collector is the element that collects the electrons after they have been used to provide microwave amplification or oscillation. The unconverted kinetic energy in the electron beam is converted to thermal energy (heat) in the collector. 2. In a transistor, the element which collects the current that passes through the base region. The collector is the output element in a transistor and performs a function similar to that of the plate in a vacuum tube. Confined flow A type of beam focusing similar to Brillouin flow except that a small amount of magnetic flux is caused to thread the cathode. More field is required than for Brillouin flow, but the beam is less affected by RF bunching (less scalloping is introduced when RF is applied to the tube, body current is affected less by the application of RF drive). (See also, Brillouin flow, PPM.) Contrawound helix A helix slow-wave structure where two helices, wound in opposite directions, are superimposed into a single structure. This circuit offers substantially higher power than a conventional helix with some sacrifice of bandwidth. Control grid An electrode mounted between the cathode and the anode of a tube to control the flow of electrons. An appropriate negative voltage (with respect to the 19

23 cathode) reduces the electron flow (beam current) to zero or cut-off and an appropriate positive (or, in some cases, less negative) voltage allows current to flow. This electrode is usually some sort of mesh structure. Coupled-cavity tube A TWT with a slow-wave structure made up of a number of cavities electrically coupled by means of coupling holes or slots. This circuit is capable of very high-power operation. Crossed-field device A high-vacuum electron tube in which a direct, alternating or pulsed voltage is applied to produce an electric field perpendicular to a static magnetic field. The values of the magnetic field and the electric field are selected to cause the electrons emitted from the cathode to travel in a direction that is nearly perpendicular to both the electric and magnetic fields and parallel to the direction of propagation of energy on a nearby microwave frequency delay line (slowwave circuit). The electron beam interacts synchronously with a slow wave on the delay line. Current density The current per unit area. For electrons in an electron beam, the current density is the beam current divided by the cross-sectional area of the beam. The current density at the cathode surface (cathode loading) is equal to the current divided by the surface area of the cathode and is usually expressed in amperes per square centimeter. Charge density can be calculated from current density divided by the velocity of the electrons. CW An abbreviation of Continuous Wave. A TWT in CW operation is provided DC voltages and the electron beam operates continuously, as opposed to pulsed operation where at least one of the voltages applied to the TWT is a pulsed voltage and the electron beam operates intermittently. In some applications, the TWT operates in the CW mode and the RF signal is pulsed. db See Decibel. dbc The level of a spurious output relative to the main signal (carrier). The dbc is defined as 10 times the log to the base 10 of the ratio of the power in the carrier to the power in the spurious output. (See also, Decibel.) dbm Power expressed on a decibel scale. The dbm is defined as 10 times the log to the base 10 of the power in milliwatts. (See also, Decibel.) dbw Power expressed on a decibel scale. The dbw is defined as 10 times the log to the base 10 of the power in Watts. (See also, Decibel.) 20

24 Decibel The Bel is a logarithmic scale for expressing gain, power, and loss, based on logarithms to the base 10 (common logarithms). The decibel is defined as one tenth of a Bel and is abbreviated as db. Usually based on power ratios, such as: Gain in db = 10 x log (power output/power input) Loss in db = 10 x log (power input/power output). If the input and output impedances of an amplifier are the same, gain can be calculated from the voltages of the input and output signals: Gain in db = 20 x log (output voltage/input voltage). Depressed collector Collector depression is the process of applying a negative potential (with respect to the tube body or ground ) to the collector of a TWT or klystron to reduce the electron beam velocity as the electrons enter the collector. This reduces the kinetic energy in the electron beam, causing less energy to be converted to heat when the beam impinges on the internal surfaces of the collector. The result is that the conversion efficiency of the device is greatly improved (less waste heat is generated). A collector designed to operate in this mode is called a depressed collector. Depressed collectors can have one or more depressed stages. Multiple-stage depressed collectors employ velocitysorting techniques to direct the high velocity electrons to the stages having the greatest depression (the greatest amount of retarding field) and the slow electrons to the stages having the least depression. Four-stage collectors are common today with five-stage units starting to be applied. Dispenser cathode A cathode that has a body made of porous tungsten. The pores contain the active materials that enhance electron emission. When the cathode is heated to its operating temperature, the material in the pores reacts with the tungsten freeing up barium which migrates toward the emitting surface, continuously replenishing the supply of active material (barium) at the emitting surface. Dispersion A term used to describe changes in phase velocity of an RF wave with respect to frequency. A non-dispersive circuit propagates RF energy at a phase velocity that is constant as frequency is varied. Drift tube A section of metal tubing held at a fixed potential to form a drift space where the electron beam is unaffected by external forces. Drive A term to indicate the RF input or RF signal to an electronic device. 21

25 Dual-mode Any device having more than one set of operating parameters, i.e., a TWT operating in both a low-power CW mode and a high-power pulse mode. DVT (Dynamic velocity taper) A technique for changing the pitch of the helix as a function of distance along the helix to improve efficiency and linearity. (See also, Tapered velocity.) Dynamic range for linear operation The range covered by the input drive level that produces 1 db or less of gain compression and the input drive level at which the signal input is equal to the noise input (ktb) in dbm plus the noise figure in db. Usually expressed in db. (See Figure A-3.) Figure A-3 Dynamic characteristics of a typical traveling-wave tube. Earth station A surface-mounted transmitter or receiver designed to communicate to or via a satellite. Mobile earth stations can be vehicle mounted land or sea. ECCM (Electronic counter-counter-measures) That sector of electronic warfare dealing with the neutralization of unfriendly jamming or deception devices. 22

26 ECM (Electronic counter-measures) That sector of electronic warfare dealing with the neutralization or deception of unfriendly detection devices. Efficiency The RF output power divided by the sum of all power provided by the power supply. The RF input power is usually ignored when calculating the efficiency of a TWT because the TWT has high gain. Electron The smallest known negatively-charged stable particle. It has a charge of x coulombs; all electric charges are presumed to be integral multiples of this number. Electrons constitute the extra-nuclear structure of atoms, and hence are present in all matter. High speed electrons emitted during radioactive decay are called beta rays. Electrons released from a negativelycharged electrode by the action of heat, light, ions or intense electrical fields constitute electron emission (sometimes called cathode rays). Electrons released from an electrode because of electron or ion bombardment are called secondary electrons. Electrons scattered from near the surface of an electrode because of the presence of other charged particles are called reflected primary electrons. Equalizers A passive device providing selective loss over an operating band such that the net gain of the equalized amplifier matches a required profile. EPC (Electronic power conditioner) A sophisticated power supply/modulator usually associated with spacequalified TWTAs. In addition to providing power, EPCs commonly include protection circuits, circuits for receiving commands and circuits for providing telemetry data. ESM (Electronic support measures) That division of EW involving actions to search for, intercept, locate and identify radiated electromagnetic energy for the purpose of threat recognition. Focus electrode An element in the electron gun that is used to focus the electrons into a well-defined beam. Sometimes called the beam-forming electrode (BFE). German manufacturers usually call this electrode the Wehnelt. Folded-helix circuit A helix slow-wave structure where two helices, wound in the same direction, are superimposed into a single structure. This circuit is sometimes called a bifilar helix and offers substantially higher power than a conventional helix with some sacrifice of bandwidth. Folded-waveguide circuit Another name for coupled-cavity circuits. 23

27 Frequency designations An officially (FCC, DOD, etc.) approved alphabetic designation for a range of frequencies. Gain The ratio of output voltage, current or power to the input voltage, current or power, respectively, in an amplifier stage, receiver or system. Power gain is usually expressed in decibels (ten times the log of the ratio of output power to input power). Getter A device which, when activated, absorbs gasses within an electron tube. The operation of a getter does not depend upon the continuous application of power. Grid An electrode mounted between the cathode and the anode of a radio or electronic tube to control the flow of electrons from cathode to anode, to serve as an electrostatic shield between the cathode and the anode, or to suppress secondary emission from the anode. When used to control the amount of electron current in a TWT or klystron, the grid is called a control grid. The grid electrode may be a cylindrical-shaped ring, a section of wire screen or mesh, or a spiral of wire through which electrons can readily move. Group delay The distortion that results when the time delay of a signal being processed through a device is not constant as a function of frequency. Areas of interest are linear delay, parabolic delay and the ripple component. The linear component is the difference in nanoseconds of delay between two frequencies within the specified bandwidth. Parabolic delay is determined by fitting an algebraic equation of the form y = ax + bx 2 + c to a plot of delay versus frequency. Ripple is the maximum peak-to-peak variation in nanoseconds of the ripple on the test data, relative to a plot of the smoothed curve. Group velocity The velocity at which energy advances along the axial direction for a signal that is propagated along the RF circuit. For a coaxial transmission line with no dielectric, the group velocity is the speed of light. For waveguide, the group velocity is less than the speed of light. Einstein s theory of relativity requires that the group velocity cannot be greater than the speed of light. (See also, Phase velocity.) Harmonic drive The inclusion of phase-conditioned harmonic power in the input RF signal to reduce harmonic capture and improve efficiency at the low end of the operating band of a TWT. 24

28 Harmonic interaction The effect of the harmonic content of the RF input signal on the beam modulation. This is generally undesirable and usually reduces the fundamental power output. Heat pipe A passive device used to transfer heat from a hot region where the thermal density is high to a cooler region at lower thermal density. A fluid is caused to boil at the hot end of the heat pipe and condensed at the cool end. A wick transports the condensate back to the hot end, no external pump or power source is needed. Heater The heater is used to raise the cathode to its operating temperature. The most common type of heater consists of a coil of tungsten wire adjacent to or embedded within the cathode body. The heater is connected to a DC or AC power supply. The use of AC on the heater sometimes induces a low level of phase modulation on the signal being amplified. Hot match The match at the input or output of a TWT, obtained when the TWT is operating (electrically hot ). HPA (High Power Amplifier) Usually refers to a subsystem used in Satellite Ground Terminal applications. Insertion loss The reduction in signal strength obtained when a passive device is installed (inserted) into an RF transmission line. Insertion loss is caused by power being absorbed in the device and by reflections from discontinuities at the interfaces between the device and the transmission line. Usually expressed in db. (See also, Decibel.) Intercepting grid A control grid that is not mechanically shielded from electrons emitted by the cathode. The grid intercepts some electrons, especially when a positive potential is applied. Inter-digital line A slow-wave structure composed of a comb-like structure with alternate segments being connected together at one end, remaining segments connected together at the opposite end. 25

29 Intercept point On a plot of intermodulation distortion data as a function of RF input drive, the carrier data at small signal drive levels has a slope of 1:1, third-order intermodulation products have a slope of 3:1 and fifth-order intermodulation products have a slope of 5:1. Extrapolating the slopes for the carriers and the third-order products produces an intercept point. The output power at this point is called the third-order intercept point, as shown in Figure A-4. Similarly, the fifth-order intercept point is the point at which the carrier slope intercepts the fifth-order product slope. (See also, Intermodulation distortion.) Interfering mode A higher order mode which, when excited, detracts from or distorts the signal in a transmission system. Intermodulation distortion 1. Impairment of fidelity resulting from the production of frequencies that are the sum of, and the difference between, frequencies contained in the applied waveform. 2. When a signal containing two or more frequencies is applied to the input of a nonlinear device, the output Figure A-4 Typical third-order intermodulation data for a helix TWT. consists of waves having the original frequencies plus additional new frequencies, as shown in Figure A-4. These new frequencies are the result of intermodulation distortion in the nonlinear device. Intermodulation is undesirable in audio amplifiers and microwave tubes such as klystrons and traveling-wave tubes. The most troublesome intermodulation products in a TWT are the thirdorder intermodulation products at frequencies of 2f1 f2 and 2f2 f1. Typical data are taken on a plot where RF input level is shown on the abscissa (X-axis) and RF output levels are shown on the ordinate (Y-axis). Three curves for output levels are usually plotted: output with one drive signal (one carrier), signal output with two drive signals (two carriers), and intermodulation products when two signals (carriers) are present. In the small signal region (carrier outputs 10 db or more below saturated output), the plots for carrier outputs have a slope of 1 db per db, the plots for third-order intermodulation products have a slope of 3 db per db and plots for fifth-order intermodulation products have a slope of 5 db per db (the fifth-order intermodulation products are at frequencies of 3f1 2f2 and 3f2 2f1). (See also, Intercept point.) 26

30 Intrinsic efficiency Efficiency calculated by dividing the RF output power by the sum of all power provided by the power supply with the exception of the power provided to the heater for the cathode. When permanent-magnet-focused TWT designs are scaled by less than a factor of two in power output level, the intrinsic efficiency is usually constant. Isolation filters A passive device or network which isolates a circuit or device from the effects of connected or surrounding circuits or devices. Isolator A passive device that has nonreciprocal propagation characteristics that result in low insertion loss for energy traveling in one direction and high insertion loss for energy traveling in the other direction. The most common usage is to isolate active devices from power reflected at discontinuities in RF transmission systems. Jammer An active electronic counter-measures (ECM) device designed to deny intelligence to unfriendly detectors or to disrupt communications. Klystron A microwave tube which uses the interaction between an electron beam and the RF energy on microwave cavities to provide signal amplification. The klystron operates on principles of velocity modulation very similar to those in a TWT except that klystron interaction takes place at discrete locations along the electron beam. Common types of klystrons are the reflex klystron (an oscillator having only one cavity), two-cavity klystron amplifiers and oscillators, and multi-cavity klystron amplifiers. Linearizer A device that improves the linearity of the AM/AM conversion curve of an amplifier (causing the slope of power output versus power input to be 1 db per db over a wide range of drive levels). Linearizers are usually designed to also reduce the AM/PM conversion (causing the phase shift to remain nearly constant over a wide range of drive levels). The use of a linearizer reduces the intermodulation distortion. (See also, Intermodulation distortion.) Loss buttons In coupled-cavity TWTs, a patented method for inserting frequency-selective loss in order to inhibit the excitation of higher-order modes. Magnetron A crossed-field microwave oscillator tube containing concentric cylinders; the inner cylinder is the cathode and the outer cylinder is an anode that contains embedded resonant cavities. A strong axial magnetic field causes a cloud of electrons to orbit between the cathode and the anode. The RF voltages across gaps in the resonators modulate the velocities of the electrons. This causes the 27

31 orbiting electrons to form into spokes that rotate around the tube axis. As the spokes of electrons rotate past the resonator gaps, they induce currents that shock excite the cavities. The RF voltages build up to large levels. High power output is obtained at moderately high efficiency. Match A measure of the quality of the impedance match between the device input or output port and the transmission line to which the device is connected. (See also, Percentage reflection, Reflection coefficient, Return loss, and VSWR.) Metal-ceramic A term applied to tube assemblies that employ no glass as a part of their vacuum walls. Insulation between elements operating at different voltages is achieved by using ceramic insulators. The assembly is integrated by using braze joints between ceramic structures and metal structures. Mode interference See Interfering mode. Multi-mode Having the capability of operating with more than a single set of parameters. (See also, Dual mode.) Multi-octave Capable of operating satisfactorily over a frequency range of 2 or more octaves. Multipactor or Multipaction A term to denote an electron-rf field interaction in which the electrons take energy from the RF fields and give up this energy to the surface on which the electrons are collected. The initial electrons are supplied by field emission, electrons that stray from the electron beam, or cosmic radiation, but the number of electrons is increased by secondary emission. In microwave tubes, multipactor is generally considered to be an undesirable effect and can occur across cavity or drift tube gaps, in output waveguides, or can involve the ceramic output window. It is detected by observing the output power as the signal level is increased. The onset of multipactor is indicated by the presence of a hysteresis loop as a function of signal level. Increasing the signal level produces a sudden drop in the RF power output followed by low output even when the drive to the amplifier is decreased below the point of initial onset. On high power tubes, there will be heating of the surface or surfaces involved. Multi-stage collector A collector with several segments, each successive segment being depressed more than the preceding segment. This further enhances the collection efficiency, thus, the overall efficiency of the device. (See also, Depressed collector.) 28

32 Near-carrier noise AM noise, PM noise and spurious outputs at frequencies close to the carrier frequency, usually 100 Hz to 500 KHz relative to the carrier. Components not related to the presence of the carrier (signal) are usually caused by ripple on power supply voltages. Interactions in the TWT are usually responsible for components that exist only when a carrier (signal) is present. Special measurement techniques are usually required to measure near-carrier noise because the resolution bandwidth (IF amplifier bandwidth) and local oscillator stability of standard spectrum analyzers are not adequate to provide accurate test data. Noise measurement test sets (NMTS) are often developed for the specific application. (See also, Noise measurement test sets.) Noise figure The ratio of the signal-to-noise on the input of a device to the signal-to-noise on the output. It is important because it indicates the amount of noise the amplifier contributes to the signal and it is an absolute indicator of the sensitivity of the device. The noise figure is usually expressed in db, and is abbreviated NF or F. Noise measurement test sets (NMTS) Special test sets developed to measure AM noise and PM noise at frequencies close to the carrier frequency, usually 100 Hz to 500 KHz relative to the carrier. Components not related to the presence of the carrier (signal) are usually caused by ripple on power supply voltages. Interactions in the TWT are usually responsible for components that exist only when a carrier (signal) is present. Special measurement techniques are usually required to measure nearcarrier noise because the resolution bandwidth (IF amplifier bandwidth) and local oscillator stability of standard spectrum analyzers are not adequate to provide accurate test data. Also, standard spectrum analyzers usually cannot distinguish between AM (amplitude modulation) noise and PM (phase modulation) noise. Noise measurement test sets (NMTS) are often developed for the specific application. Two major types exist. In one type, the signal is passed through a mixer (down-converter) that has a highly stable local oscillator, feeding into a very narrow band IF amplifier, and detected by either an AM or a PM detector. The video output from the detector is analyzed by a wave analyzer or video spectrum analyzer. The AM detector is a simple crystal detector, the PM detector employs two detectors and a delay line or a resonant cavity to provide a discriminator characteristic. An important component of either type of test set is a calibrator that provides a reference signal and a known level of either type of modulation. Noise power The noise generated by a device or amplifier when measured at the output port when the input port is terminated and no RF drive is applied. Usually symbolized as NPO and measured in milliwatts or dbm. Noise power output in dbm can be estimated from: NPO (in dbm) = NF + G + 10 x log(bw) 29

33 Where NF is the measured noise figure of the device, G is the small signal gain in db (assuming that the noise power does not drive the amplifier near its saturated output level) and BW is the bandwidth in MHz. Noise power ratio (NPR) A measure of intermodulation when an infinite number of carriers is being amplified. The input to the amplifier is a well-defined bandwidth of random noise with a deep notch in the center. Intermodulation in the amplifier produces intermodulation products that partially fill the notch. The ratio of noise power in the passband to the noise power in the notch, measured at the output of the amplifier, is the NPR. Non-intercepting grid A control grid that is mechanically shielded from the cathode by a shadow grid in such a way that no electrons are intercepted by the control grid, even when a positive voltage is applied to the control grid (with respect to cathode). Octave A term borrowed from music to designate a range of frequencies where the highest frequency is twice the lowest frequency. Outgassing A term used to describe the emission of various gasses from internal surfaces during the processing and testing of thermionic devices. Overdrive An input signal level greater than that required for saturation, resulting in decreasing output power and increasing distortion. Peak power 1. The power at the maximum of a pulse of power, excluding spikes. 2. The output power at the maximum amplitude that can occur with any combination of signals to be transmitted. Percentage reflection A measure of the quality of the impedance match between the device input or output port and the transmission line to which the device is connected. Expressed as 100 times the voltage in the reflected wave divided by the voltage in the incident wave. (See also, Match, Reflection coefficient, Return loss, and VSWR.) Perveance 1. A numerical constant for an electron gun design, determined by the size, spacing and geometry of the electrodes in the electron gun. This value is often called the gun design perveance. The measured perveance for the gun depends on the tolerances on the geometry and the condition of the cathode. 2. A quantity based on the measurement of beam current and voltage defined as the current divided by the voltage to the three-halves power. When the voltage is the cathode-to-anode voltage in the electron gun, the perveance is 30

34 called the gun perveance. When the voltage is the cathode-to-circuit voltage, the perveance is called the beam perveance. Phase linearity A term referring to the degree of deviation from a straight line of the phase versus frequency characteristic of a device. Phase pulling factor The amount by which a change in input or output VSWR changes the phase shift in the TWT. Phase pushing factor The amount by which a change in the voltage on an electrode changes the phase shift in the TWT, usually expressed in degrees per Volt. Phase tracking The closeness or similarity of the phase characteristics of a number of devices. This is an important consideration when power combining the outputs from two or more devices. Phase velocity The velocity at which phase fronts advance along the axial direction for a signal that is propagated along the RF circuit. For a coaxial transmission line with no dielectric, the phase velocity is the speed of light. For waveguide, the phase velocity is greater than the speed of light. Slow wave circuits are designed to provide a phase velocity that is much less than the speed of light. (See also, Group velocity.) PM near-carrier noise PM (phase modulation) noise and spurious outputs at frequencies close to the carrier frequency, usually 100 Hz to 500 KHz relative to the carrier. Components not related to the presence of the carrier (signal) are usually caused by ripple on power supply voltages. Interactions in the TWT are usually responsible for components that exist only when a carrier (signal) is present. (See also, Near-carrier noise.) Power combining A scheme whereby the outputs from two or more amplifiers are combined to provide a greater output power than available from a single amplifier. Power curve A plot of output power versus input drive or input frequency. (For power curve versus input drive, see also, AM/AM conversion.) PPM (Periodic-Permanent-Magnet) A method of focusing a TWT where permanent magnets of opposite polarity are placed side by side along the length of the tube. 31

35 Pulse compression A matched filter technique used to discriminate against signals which do not correspond to the transmitted signal. Used in radar systems for improved detection capability. Pulse-up ratio The ratio, usually expressed in db, between the CW power level and the pulse-power level in a dual-mode device. Radar Acronym for radio detecting and ranging. A system where a relatively high frequency radio pulse is used to bounce a signal off a distant object. The direction and time of response give the location of the object. Redundancy, automatic In a communications system, a feature which automatically switches to a standby unit in the event of a failure. Reflection coefficient A measure of the quality of the impedance match between the device input or output port and the transmission line to which the device is connected. Expressed as the voltage in the reflected wave divided by the voltage in the incident wave. (See also, Match, Percentage reflection, Return loss, and VSWR.) 32

36 Resonant cavity A short piece of waveguide or other transmission line terminated at both ends with a metal piston, an iris diaphragm, an open circuit, or some other wavereflecting device. It can be used as a component of a slow wave transmission line (circuit), a filter, a coupler between transmission systems, or as an impedance transformation (matching) network. Klystrons use resonant cavities as input and output couplers to couple microwave energy into and out of the electron beam. Coupled-cavity TWTs use tightly coupled resonant cavities as slow wave circuits to provide continuous coupling between the waves on the circuit and the electron beam. Return loss A measure of the quality of the impedance match between the device input or output port and the transmission line to which the device is connected. Expressed as -20 times the log of the ratio of voltage in the reflected wave divided by the voltage in the incident wave. (See also, Match, Percentage reflection, Reflection coefficient, and VSWR.) Ring-bar tube A TWT with a slow-wave structure composed of ring-like segments connected by straps or bars. This device is capable of higher power levels than a conventional helix tube at a significant reduction in bandwidth. Saturated power output A term used to describe that point on the RF power output versus RF power input characteristic where an increase in input power does not produce an increase in output power. Screen grid A grid structure placed between the control grid and anode to reduce the capacitive coupling between the control grid and the anode. In a multi-mode electron gun the element used to control beam current from the edge of the cathode. Serrodyne An operating mode for a TWT in which the input signal is translated to a new frequency at the output. A sawtooth voltage is superimposed on the DC voltage between the cathode and the slow wave circuit, causing the phase shift in the TWT to change in a linear fashion as a function of time. The amplitude of the sawtooth voltage can be adjusted to provide 360 degrees of phase shift during each cycle, causing the output of the TWT to be offset by the frequency of the sawtooth waveform (the flyback time of the sawtooth must be very short so the TWT thinks that the phase is being offset continuously). Shadow grid A grid structure placed between the cathode and control grid and electrically connected to the cathode. This element shields the control grid from interception. 33

37 Single-stage collector A TWT or klystron collector that has only one element for collecting electrons. The stage may be operated at ground potential or may be depressed. (See also, Depressed collector and Multi-stage collector.) Slow-wave circuit Any structure which slows the effective axial phase velocity of an RF wave in order to establish synchronism between that wave and an electron beam. Slow-wave structure A microwave propagating structure that provides an axial phase velocity that is less than the speed of light. (See also, Slow-wave circuit.) Space charge The electrical charge in electrons that are in a vacuum environment (as opposed to electrons within the crystal structure of a metal.) Space charge limited current The mode of electron gun operation in which the cathode emits a copious quantity of electrons and the flow of current is limited only by the electric field introduced by the charge on the electrons in the space near the cathode emitting surface. (See also, Temperature limited current.) Space charge emission See Space charge limited current. Spherical diode A two element (cathode and anode) structure built in such a manner as to duplicate the spherical geometry of the cathode in a typical electron gun. Used for evaluation and analysis of electron emission, analysis of electron gun designs and for realistic cathode life testing. Tapered termination A gradual increase in the amount of loss applied to a slow-wave structure to control reflections within a TWT. Tapered velocity A change in the pitch of a helix, the height of cavities, or some other axial dimensions of a slow wave circuit to change the phase velocity of the RF wave. This is done because the extraction of kinetic energy from the electron beam causes the beam to slow down near the output of the TWT. The taper helps to maintain synchronism between the RF wave and the electron beam; this greatly improves the efficiency of the TWT. (See also, DVT.) Temperature limited current The mode of electron gun operation in which the cathode does not emit a sufficient quantity of electrons to support space charge limited current and the flow of current is limited by the amount of current emitted from the cathode. Temperature limited current occurs when a cathode is operated well below its 34

38 normal operating temperature or when a cathode is depleted. (See also, Space charge limited current.) Temperature limited emission See Temperature limited current. Tetrode A thermionic device having four elements, usually a cathode, control grid, screen grid, and anode. In multi-mode TWTs, a term describing the electron gun (cathode, shadow grid, control grid and screen grid). Time-out circuitry A combination of circuit elements designed to actuate after the lapse of a certain period of time. Commonly used to apply high voltage to a tube following an appropriate warm-up time after application of heater voltage. Transfer curve 1. The family of curves for various values of plate voltage in which plate current is plotted as a function of control grid voltage. 2. Specifically with regard to microwave tubes, a curve or family of curves in which output power is plotted as a function of input drive power at a fixed beam voltage. Sometimes referred to as a gain-curve. (See also, AM/AM transfer.) Triode A three element thermionic device composed of a cathode, a control grid and an anode. TR limiter (TR switch) A Transmit/Receive switching device which limits the amount of power transmitted. Usually employed as a receiver protect device in a radar system. TWT (Traveling-wave tube) A microwave tube of special design using a broadband circuit in which a beam of electrons interacts continuously with a guided electromagnetic field to amplify microwave frequencies. TWTA (Traveling-wave-tube amplifier) A combination of a power supply, a modulator (for pulsed systems), and a traveling-wave tube, often packaged in a common enclosure. Uplink/downlink Uplink refers to the transmission of intelligence to a satellite while downlink refers to the re-transmission to a ground station. Vacuum envelope Any structure containing or capable of containing a high vacuum environment. Usually refers to the body structure of a thermionic tube. 35

39 Velocity resynchronization Any method for changing the axial velocity of an RF wave or of an electron beam to improve the synchronism between that wave and an electron beam. (See also, Tapered velocity and Velocity jump.) Velocity jump A method of maintaining the proper relationship between beam velocity and circuit phase velocity near the output of a TWT whereby the beam velocity is caused to increase (as opposed to causing the circuit phase velocity to decrease). The increase in beam velocity is obtained by operating the output portion of the slow wave circuit at a voltage that is positive with respect to the remainder of the circuit. Velocity step See Tapered velocity. Velocity taper See Tapered velocity. VSWR An acronym for Voltage Standing Wave Ratio. A measure of the quality of the impedance match between the device input or output port and the transmission line to which the device is connected. Measured by observing the standing wave pattern produced by the interaction of the incident wave with the reflected wave. The VSWR is the ratio of the voltage at a peak divided by the voltage at a minimum point on the standing wave pattern. (See also, Match, Reflection coefficient, and Return loss.) Work function The amount of energy it takes to get an electron out of the cathode s crystal structure and into the vacuum environment in front of the cathode, measured in electron-volts (ev). A cathode having a low work function can operate at a lower temperature than a cathode having a high work function. 36

40 APPENDIX B ELECTRON GUN AND CATHODE DESIGN TRADE-OFFS

41 With increasing performance demands on TWTs, the quality of the electron gun design is a key factor. There is, therefore, a specialized group within L-3 ETI which concentrates in gun design, focusing, and related electron-optical problems. Design tools, such as computer programs for analysis and gun analyzers for experimental evaluation, are continuously upgraded and improved. A file on previously designed guns of all types is maintained for reference and as a basis for new designs. Electron guns used in traveling-wave tubes are generally convergent. This means that the current density at the cathode, i.e., the cathode loading, is significantly lower than the current density in the beam and below a specified maximum value. Cathode loading is related to the cathode life; reduced cathode loading will usually allow the cathode to be operated at a lower temperature and will provide a longer cathode life. The design of high perveance, convergent guns is well established and is based on Pierce s spherical diode concept in which a conical flow of electrons converges into a well-defined beam having uniform current density across the beam cross-sections. The design procedures provide data on cathode and anode radii of curvature and aperture angles, but do not give sufficient details on the shape of the beam-focusing electrode and anode. The design of these electrodes can be determined with a large scale computer program that enables the designer to evaluate a variety of electrode shapes and to establish an optimized configuration. Electron Guns for TWTs Requiring Electron Beam Modulation In some applications it is necessary to depart from this simpler electron gun geometry to provide the capability of modulating with smaller voltage swings. This can be accomplished in any of a number of fashions. The simplest approach is to isolate all or part of the focus electrode from the cathode and then provide a separate connection for that TWT gun element. Another approach involves placing an isolated electrode between the focus electrode and the anode and designing the shapes of the electrodes to provide the necessary electric field profile. This is called aperture grid modulation; for high perveance guns, this usually requires a voltage swing that is a large fraction of the cathode voltage. If smaller voltage swings are required, a radial vaned ( target shaped) grid can be placed in front of the cathode as shown in Figure B-1. The use of a grid will allow the cathode current to be modulated by a swing of a few hundred volts, but some of the current that would normally pass through the anode and enter the beam hole in the slow wave circuit will be intercepted by the grid (in fact, this type of grid is called an intercepting grid). The portion of the cathode current that will be intercepted by the grid is determined by the screening factor of the grid. As a result, single-gridded guns, such as that shown in Figure B-2, can only be used for low-average-power tubes because of the thermal power limitations of the intercepting grid. The modulator must supply the current intercepted by the grid; it is necessary to verify that the thermal load on the grid due to interception is within acceptable limits. 38

42 High-average-power tubes employ, therefore, nonintercepting gridded guns, which use a precisely aligned pair of grids, with the shadowing grid closest to the cathode electrically connected to the cathode and either on or close to the emitting surface of the cathode. The wires in the shadow grid suppress emission from those portions of the cathode that are in close proximity to the grid wires, eliminating the current that would be intercepted on the control grid. The penalties paid are that the electron gun is more complex Figure B-1 Grid with radial vanes. to construct (since the two grids must be aligned very precisely) and the quality of the beam focusing is degraded compared to the beam quality obtained from an equivalent non-gridded design, since we now have non-uniform emission from the cathode, as shown schematically in Figure B-3. This grid configuration reduces grid interception of the control grid from about 10 percent to a very small fraction of one percent, thus making it possible to substantially increase the average power capability. Figure B-2 Schematic of intercepting gridded gun. 39

43 Figure B-3 Schematic of nonintercepting gridded gun. Figure B-4 shows typical gridded electron gun characteristic curves which determine the specific point of tube operation, as well as the required negative voltage for tube cut-off. Future improvements in the techniques for building gridded electron guns include the use of cathodes in which non-emitting regions perform the function of the shadow grid and the use of grids that are bonded directly to the surface of the cathode. Oxide Cathodes The oxide cathode uses a nickel substrate that is coated with oxides of barium, strontium and (sometimes) other metals. This type of cathode was once the mainstay of the TWT industry, but has now been replaced by the dispenser cathode (see below). Oxide cathodes are still used in certain lower frequency tubes. Although no longer used in ETI TWTs, their description is included here because the reader may still come across them in certain older devices. Oxide cathodes will support current densities of 0.5 Amperes per square centimeter in CW operation and 2 to 10 Amperes per square centimeter in pulsed operation (usually in applications where the pulse width is 50 microseconds or less). The temperature at which the cathode must operate depends on the current density required; the quality of the vacuum inside the TWT, including the partial pressure of each species of gas; and the duty cycle at which current is being drawn from the cathode. For example, a high power CW TWT may have internal parts that liberate gasses because of their high operating temperatures, making it 40

44 Figure B-4 Focusing characteristics of a typical gridded gun. necessary to operate the cathode at a temperature in excess of 750 degrees C. In applications where the RF power level is less than 1000 Watts, the cathode may be operated at temperatures below 680 degrees C (especially when the current density is less than 0.25 Amperes per square centimeter). For most TWTs, the operating temperature is in the range 680 to 750 degrees C. Oxide cathode life expectancy as a function of operating temperature is shown in Figure B-5. The abscissa in the figure is true temperature in degrees C, since the temperature of oxide cathodes is usually calibrated by building special diodes in which a thermocouple is used to determine true temperature as a function of heater power. As seen in Figure B-5, in many applications the life expectancy was greater than 20 years. The model used to generate the data in Figure B-5 is based on the performance of on-orbit space TWTs and life test TWTs having demonstrated lives of 142,600 hours (16.3 years). Dispenser Cathodes Dispenser cathodes consist of a porous tungsten matrix that is filled with compounds of barium, calcium and aluminum. In this simplest form, the cathode is called a B-type or an S-type cathode, depending on the ratio of the compounds in the matrix. These cathodes will support current densities of 5 Amperes per square centimeter and higher, and are less susceptible to gassy environments than the oxide cathode. In pulsed applications they are less dependent upon the duty cycle than the oxide cathode. Operating temperatures for B-type and S-type cathodes are usually in the range 1000 to 1200 degrees C. Operating temperatures can be reduced to the range 900 to 1100 degrees C by coating the emitting surface with a thin layer of osmium and other metals. Such a cathode is designated an M-type cathode. This type of cathode is the current standard cathode for virtually all ETI TWTs. It s combination of long life and resistance to less-than-ideal environmental conditions have it the industry standard for high reliability TWTs. The scandate cathode, which is in development, has a coating of scandium compounds and offers promise of operation at even lower temperatures. 41

45 Figure B-5 Life prediction for oxide cathodes operating under ideal conditions. A life expectancy of greater than 30 years can be obtained when the dispenser cathode is operated at temperatures below 1000 degrees C; this requires that the cathode be operated at a relatively low current density and that the vacuum environment in the TWT is quite good. The life expectancy prediction for dispenser cathodes operating under these conditions is shown in Figure B-6. The abscissa in Figure B-6 is brightness temperature in degrees C, since the temperature usually is calibrated by building special diodes that contain a viewing window. An optical pyrometer is used to determine brightness temperature as a function of heater power (brightness temperature is lower than true temperature by 40 to 60 degrees, depending on the emissivity of the surface being observed). The ordinate is the time at which the cathode current has degraded to 90 percent of its beginning-of-life value (this corresponds to a 0.6 db reduction in saturated output power). In practice, this RF output power drop is avoided in modern TWTA/HPA systems by incorporating an anode or grid voltage control loop that senses cathode current and adjusts the anode or grid voltage as necessary to keep the cathode current at a constant level. With these TWTA control loops, constant RF output power can be maintained for the full duration of the spacecraft lifetime. Additional comments upon TWTA life are discussed in Appendix L. 42

46 Figure B-6 Life prediction for state-of-the-art dispenser cathodes operating under ideal conditions. Gun Design Trade-Offs The voltage-current relationship in a TWT electron gun is defined by the perveance, as follows: gun perveance = cathode current (cathodeto anode voltage) 1.5 or gun microperveance = cathode current (cathodeto anode voltage) 1.5 Typical values of gun perveance range from 0.01 to 1.5 micropervs (the unit for perveance is the perv and the unit for microperveance is the microperv when voltage is in volts and current is in amperes). This is an important electron gun design parameter that affects the design of the entire TWT, not just the electron gun. The efficiency at which electron beam power (cathode voltage times cathode current) is converted to microwave power is called the basic efficiency or the beam efficiency or the circuit efficiency and depends strongly on the perveance of the electron beam. Basic efficiency usually lies in the range 5 percent to 40 percent. The lower values of basic efficiency are usually encountered in designs that have low values of perveance, as would be the case for low power, high frequency TWTs. At a particular power level, the range of 43

47 basic efficiencies will usually fall within a relatively narrow range. For example, the basic efficiency for a 12 GHz 50 Watt TWT may lie in the range 15 percent to 30 percent. This implies that the power in the electron basic must be in the range 167 Watts to 333 Watts. This power level can be achieved by using low, moderate or high values of perveance. The design of the TWT would be different for each set of conditions. Low Voltage, High Perveance, Electron Gun Design Selecting low voltage, high current operation means that the gun perveance must be high, and this presents several practical limitations. High perveance electron guns are more difficult to design and construct than low perveance electron guns; the quality of the electron beam generated is more sensitive to spacing and alignments than for low perveance designs. High perveance operation implies that the current density in the electron beam is high. This results from the fact that the electron beam diameter required for a particular design depends on the frequency of operation and the operating voltage. The combination of high frequency operation and high perveance (low voltage) implies that dimensions are small and current levels are high, resulting in high current density. Reasonable performance in the slow-wave circuit usually requires that the electron beam have a current density in the range 2 Amperes per square centimeter to 250 Amperes per square centimeter. These current densities are often much higher than can be supported by the cathode emitting surface, requiring that the cathode emitting surface be designed as a concave spherical cap. The surface area of the cathode is then much larger than the crosssectional area of the electron beam. The ratio of these two areas is called the area compression or the compression ratio of the electron gun. An electron gun having a large area compression can provide and electron beam having a high current density while maintaining a low current density at the cathode. However, thermal velocity effects and the effect of slight misalignments make it more difficult to focus beams that emanate from an electron gun having high area compression. Area compression ratios of from 10:1 to 50:1 are quite common. Ratios of up to 100:1 or so are used when necessary. However, beyond about 120:1, tolerance control gets extremely difficult, tolerances of the order of to inch are common for these designs. Area compressions greater than this are generally avoided. At the other extreme, low area compression can result in high current density at the cathode surface and require that the cathode be operated at a high temperature. High Voltage, Low Perveance Electron Gun Design Cathode current density and gun perveance considerations can be removed by designing the TWT for high voltage, low current (low perveance) operation. Low perveance operation also offers the advantages that the electrons in the spent electron beam contain a smaller percentage velocity spread than is the case in high perveance designs (making it easier to design a highly efficient collector and obtain high overall efficiency) and the helix in a helix TWT is larger 44

48 (making it easier to fabricate and providing greater thermal capacity). However, low perveance design can introduce a whole new set of problems: High voltage operation requires larger spacings between electrodes within the TWT; geometries must be chosen to keep the value of electric field across vacuum gaps at less than 200 kv/cm. If the surfaces are not smooth, the field is limited to lower values. TWTs operating at higher voltages tend to be less stable than TWTs operating at lower voltages. For example, unless special precautions are implemented, undesirable backward-wave oscillations can occur when a TWT having a helix type slow-wave structure is operated at voltages greater than 8 to 15 kv, depending on the frequency. A Practical Electron Gun Design for a Basic TWT An electron gun for a practical TWT will operate at an intermediate perveance, based on a trade-off study in which the factors shown in Table B-1 are considered. Table B-1 Factors Considered in Selecting the Gun Perveance for a TWT Factors that Favor High Perveance 1. Operating voltage is low, usually avoids high voltage breakdown problems. 2. RF performance is better, especially when broadband operation is needed. 3. Gain per unit length is higher, therefore the structure is shorter. Factors that Favor Low Perveance 1. Less magnetic field required for beam focusing. 2. It is easier to design an electron gun for the area compression that is needed for low cathode loading (for long cathode life). 3. Slow-wave circuit design is simplified because of the larger dimensions, and the design is tolerant to manufacturing variations. The electron gun must be designed for the proper perveance while drawing uniform current density from across the cathode surface and producing an electron beam that has the proper diameter. The profiles of the anode and the focus electrode are selected to provide the electric field patterns that are needed to meet these requirements and to cause the electrons to converge into a welldefined electron beam. TWTs having guns of this geometry are modulated by varying the cathode voltage or the anode voltage, requiring power supplies capable of providing large voltage swings (complete turn-off of the amplification usually dictates that the voltage swing is as great as the operating cathode voltage). Modulating the cathode voltage is called cathode modulation operation and requires switching the entire cathode current as well as the cathode voltage. 45

49 Modulating the anode voltage is called mod-anode operation and requires switching only the anode current, a small fraction of the cathode current. Modanode design requires that the spacings and insulators between the anode and ground must be sufficient to isolate the large negative voltage that must be applied to the anode when the TWT is turned off as well as the spacings to isolate the positive anode from the cathode when the TWT is turned on. An electron gun designed for a given TWT will incorporate only the design features required for its particular application. 46

50 APPENDIX C SLOW-WAVE CIRCUIT AND BEAM FOCUSING TRADE-OFFS

51 Slow-Wave Structures TWT amplification takes place when the velocity of the electron beam is close to (in synchronism with) the velocity of the signal on the slow-wave circuit. Maximum amplification occurs when the electron beam velocity is somewhat greater than the velocity of the signal on the slow-wave circuit. The amplification process causes energy in the electron beam to be converted into microwave energy; this causes the electrons to slow down as they approach the output end of the TWT. The slow-wave circuit must meet several requirements, including: The velocity of propagation on the slow-wave circuit must be nearly synchronous with the velocity of electrons that have been accelerated through the potential between the cathode and the circuit. This condition must be maintained over the entire operating frequency range. For highest efficiency, the velocity of propagation is tapered toward the output end (to account for the slowing of the electrons). The circuit must provide a strong axial RF electric field in close proximity to the electron beam. The circuit must provide good thermal paths to dissipate heat generated by RF losses and by interception of electrons. The most common circuits used to meet these requirements are the helix circuit and the coupled-cavity circuit. These circuit options are shown in Figure C-1a and Figure C-1b. Historically, the simple helix was the first circuit to be investigated by R. Kompfner in Obviously this circuit must possess some outstanding advantages to have persisted so long in such a commanding position. Probably its simplicity appealed to Kompfner and led him to utilize it in his broadband modulation experiments, since at that time the available techniques for fabrication and assembly of vacuum components were crude and primitive compared to those of the present art. The early investigators of this elegant circuit probably did not fully realize that almost all of the theoretical measurement standards would show that the helix was far superior to hundreds of other propagating structures which were to be investigated during the formative years of TWT development. In retrospect it is easy to appreciate the basic physical reasons which have given the helix its unchallenged position of leadership, but in the beginning, many hoped that the very first circuit candidate could be superseded by something much better. Simple first-order considerations of the interaction process between an electron stream and a propagating electromagnetic wave suggest that the following properties determine a figure of merit for the particular circuit being evaluated: Minimum stored energy in the propagating wave Maximum axial electric field in the region of the electron beam Constant phase velocity (no dispersion) over frequency No interfering modes The last property is one which must be qualified since it is highly dependent upon the specific design under consideration, but in general the helix possesses a very manageable mode structure that is indeed superior to almost all other substitutes. 48

52 Figure C-1a Helix slow wave structure and the support rods. Figure C-1b In the L-3 ETI patented coupled-cavity slow-wave circuit, lossy ceramic buttons are used to produce a gradually tapering loss pattern to reduce mismatch between internal terminations and the input and output RF waveguide couplers. The dispersion characteristic of the helix is probably the single most significant property that has caused this circuit to find such acceptance for broadband application. ECM systems and broadband test equipment demand an octave or more of good performance from a TWT to provide maximum coverage in a single amplifier. 49

53 Widest Bandwidth No circuit has ever rivaled the helix in bandwidth capability, and most do not come close in this department. It behaves very much like a single wire above a ground plane, propagating a TEM mode. Such a circuit is, of course, completely nondispersive. Since the helix is much more complicated than a single wire over a ground plane, it does not provide infinite bandwidth and does exhibit mode interference at a very well-defined point where the circumference is exactly one-half the wavelength of the propagating frequency. Since the helix geometry does not involve large opposed metallic surfaces, the stored energy for a given power level is naturally quite low. Almost any conceivable alternative to the helix employs more massive metal surfaces, which provide an equivalent capacitance for the storage of energy and a lowering of this figure of merit. The helix also provides a very convenient electric-field configuration. Inside the structure the field is somewhat constant (it does vary) over the cross section of the pencil beam which is generally utilized as the energy source. It is difficult to imagine another geometry with the same natural uniformity in this regard. Most alternative circuits do not provide, therefore, as strong an interaction between the electric field and the beam. Analytically, the ratio between the axial electric field (squared) in the beam region to the total power flow in the helix circuit is referred to as the circuit interaction impedance. It is referred to as impedance even though it is related to the electric field which is available for interaction with the electron beam and is, therefore, not the same impedance as is generally employed in ordinary microwave circuit investigations. The excellent interaction impedance of a helix has resulted in TWTs with efficiencies that exceed 70%. Mechanical Advantages Aside from purely electrical considerations, the helix is almost ideal from a mechanical viewpoint. It lends itself to simple fabrication techniques which are highly precise and it can be accurately assembled in structures which fit well with the rest of the TWT package. Circuit symmetry is essential if an elegant design is desired at a reasonably low cost. Because of the precise manufacturing capability and the inherent nature of the helix circuit, internal rf wave mismatches are quite small, and this results in low gain and power ripple characteristics across frequency. The first helix TWTs constructed in the late 1940 s were characterized by their fragility and very low thermal capacity. As a consequence, their early development was directed primarily towards low-power applications where the signal power was a few watts or less. These devices were temperamental and short-lived because of the unsophisticated techniques used in their design and construction. Today, by contrast, helix-type TWTs are quite capable of delivering several kilowatts of CW power at S-band and C-band over an octave of frequency coverage. For space communications, the helix TWT has matured to the point where 100 watt TWTs with efficiencies reaching towards 70% are common place at frequencies as high as K-band. Even at V-band, 50 watt space-qualifiable TWTs have been demonstrated since the late-1990 s. The lifetime and reliability of many helix TWTs (satellite applications typically require at least 15 years of 50

54 component life) exceed most other types of active microwave sources. In short, this generic device has advanced in capability by orders of magnitude as a result of sustained development efforts by the major TWT manufacturers. Helix-derived TWTs To create suitable high-power beams for the generation of more than 5 kw of peak RF power, it is almost mandatory to utilize beam voltages in excess of 10 kv when a conventional TWT design approach is to be employed. At beam voltages greater than 8 to 15 kv, however, the pure helix suffers from the nature of its mode structure, especially the backward-wave mode. Historically, therefore, investigators have proceeded in two separate directions, both of which have proved successful in their efforts to develop suitable TWT circuits for use at higher voltages. The first attempts concentrated on a modification of the simple helix circuit by employing another helix coincident with the first, but wound with a reversed pitch. This came to be known as the contra-wound helix, and later versions were designated as the ring-bar circuit and the ring-loop circuit. Figure C-2 illustrates these geometries. Figure C-2 Helix-derived circuits. It can be shown from simple circuit considerations that the phase velocity of these two configurations is going to be much higher than a simple helix, and consequently, synchronous beam voltages far in excess of 10 kv should be realizable. This is, in fact, the case. Many successful TWTs have been developed using this structure. The basic assembly is almost identical to that of a conventional helix tube in that the RF structure is supported on ceramic rods inside a long tubular barrel. Unfortunately, the thermal capacity of such a circuit is not much different from that of the simple helix, so the ultimate average power capability is restricted to roughly the same numbers as for a helix TWT at the same frequency and size. Ring-bar TWTs Ring-bar TWTs can conceptually be considered as structures derived from multiple helix circuits, in particular, the twin crosswound helix (Figure C-2). 51

55 The ring-bar circuit has, however, significantly higher interaction impedance than a helix, and is thus capable of more efficient beam power conversion and larger gain per wavelength. Such tubes also exhibit superior RF-stability with respect to backward-wave oscillations (BWO) compared to helix tubes and are, therefore, capable of operating at higher voltages, as well as of producing higher peak-power levels. They are also capable of handling larger average RF power loads and, thus, they frequently use nonintercepting gridded guns rather than intercepting (single) gridded guns characteristic of high-power helix tubes. Unlike most helix tubes, the bandwidth of a ring-bar tube is generally limited to about 10 percent to 20 percent. As a result, the ring-bar design finds its most frequent application in radar systems. Typical is the L-3 ETI 8729H prototype ring-bar TWT, with performance characteristics summarized in Figures C-3 and C-4. Coupled-cavity TWTs The second basic approach to high-power circuits, other than variations on the helix, has been far more popular because of its many distinct advantages and tremendous flexibility. It represents a complete departure from the helix concept both in its electrical behavior and mechanical configurations. The best generic description is a high-power filter circuit with bandpass characteristics, a form of traveling-wave circuitry which was first considered at the very beginnings of the technology in the 1940 s. Any repetitive series of lumped LC elements constitutes a propagating filtertype circuit and the techniques for synthesizing these circuits are well established in the art. Almost any phase characteristic desired can be realized if the proper LC elements are selected. The real test comes when one tries to transform these choices into a practical mechanical structure that can be fabricated and assembled in accordance with accepted vacuum-tube techniques. The early attempts at this task resulted in some very interesting museum pieces, which probably consumed thousands of man-hours of fruitless labor. These were rejected because they lacked simplicity and symmetry, attributes that would be required to make them practical from the viewpoint of cost and flexibility. Probably as a result of these frustrating endeavors, the real objectives were properly identified and the main thrust proceeded in a direction which satisfied the basic requirements of a good universal filter-type circuit. 52

56 Figure C-3 Peak output power of ring-bar tube 8729H. Figure C-4 Saturated gain for ring-bar tube 8729H. All of this early work culminated in the discovery and development of the coupled-cavity circuit, which now constitutes the fundamental building block of an extremely important class of high-power TWTs. Its remarkable acceptance is clear testimony to its inherent superior qualities, which can be summarized as follows: Excellent electrical characteristics in terms of impedance, bandwidth, and mode structure Mechanical simplicity, circular symmetry for easy machining and assembly Form factor ideally suited to PPM focusing Rugged from both a mechanical and thermal viewpoint 53

57 Very versatile; simple procedures for scaling frequency, power, and bandwidth. The versatility of the coupled-cavity circuit is demonstrated by the fact that it is widely used from L-band to millimeter waves and for power levels from 1 to 500 kw. Probably 90 percent of all high-power TWTs employ this basic type of filter structure. The term coupled cavity stems from the striking similarity of the individual unit cells to an ordinary klystron resonant cavity. In the latter case, of course, there is no coupling, so each cavity is completely closed. In the case of the TWT circuit, coupling is provided by a long slot in the wall of each cavity, as illustrated in Figure C-5. Figure C-5 Basic coupled-cavity circuit. This slot strongly couples the magnetic component of the field in adjacent cavities in such a manner that the passband of the circuit is primarily a function of this one variable. For very small slots, or coupling holes, the passband is quite narrow. When the slot angle (0) is somewhat larger than 180 degrees, the passband is close to its practical limits. The drift tube is formed by the re-entrant part of the cavity, just as is the case with a klystron. Its length is determined by beam-interaction considerations, but the optimum design for a given bandwidth is not a critical function of the gap length. In fact, all of the important cavity dimensions can be adjusted over a rather broad range to accommodate trade-offs between thermal requirements and electrical performance without seriously degrading circuit capability. Once the design is made, however, the tolerances of the circuit dimensions must be very closely maintained. Each half-cavity section can be fabricated in laminated form, which is ideal for the assembly and brazing operations. The individual parts are almost self-jigging, which assures very accurate alignment and spacing between cavities. 54

58 Liquid cooling of the circuit can be provided by properly channeling the outer diameter such that the coolant flows around the massive copper walls of the individual sections. In extreme cases, the coolant can be channeled around the drift tubes to absorb beam interception healing directly, at the price of greater fabrication complexity. Beam Focusing Schemes All TWTs require some means of holding the cylindrical electron beam in shape as it travels along the inner diameter of the interaction structure. Without a focusing structure, the beam tends to disperse or spread out as a result of the mutual repulsive electrical forces between electrons (these forces are often referred to as space-charge forces). A magnetic field in varied forms is used for this purpose. Such a field of proper magnitude will confine the electron beam to the pencil-like cylindrical shape it must maintain. The two principal types of magnetic focusing discussed here are illustrated in Figure C-6. Figure C-6 Two principal methods of magnetic focusing. Of these, the solenoid is still regarded as one of the best magnetic focusing structures. Its magnetic lines are parallel to the direction of travel of the electrons and it can be accurately aligned with the beam. It provides excellent beam collimation and will continue to be used in applications where the last bit of average power is required from a tube so long as tube size and weight are not critical factors. Most of the very-high-power TWTs to date have utilized solenoids. In fact, L-3 ETI has pioneered a technique for wrapping the solenoid directly onto the tube barrel. This optimizes the alignment between the tube axis and the magnetic axis. It also brings the solenoid windings as close to the tube axis as possible, providing the required magnetic field with the very minimum size and weight. A secondary benefit is that less DC power is required to provide the magnetic field. In certain structures, however, where the intraction structure is short enough, permanent-magnet focusing is often utilized in lieu of the bulky 55

59 solenoid. Because of the length limitations, this type of focusing is generally restricted to low-gain or low-power tubes. At the lower power levels it is possible to use permanent magnets, even in long TWTs. The most conservative use of magnetic material is encountered in designs that use a periodic-permanent-magnet (PPM) focusing arrangement as shown in Figures C-6 and C-7. This PPM structure provides a nearly sinusoidal magnetic field on the electron beam axis and very low external field. The RMS value of the PPM field is roughly equivalent to the value of field required in a uniform field design (such as a solenoid focused design or a permanentmagnet design using a single, large magnet). A carefully designed PPM focusing structure using samariumcobalt magnets can provide an RMS focusing field in excess of 6000 Gauss. Perhaps the most outstanding advantage of the coupled-cavity circuit from the user s viewpoint is its natural adaptability to lightweight PPM focusing. In many airborne systems, the weight and bulk of a separate focusing solenoid, along with its sizable power supply, are unacceptable. In these situations, a TWT would be rejected if it were not possible to simplify the focusing requirements with a PPM structure. Space satellite communications TWTs virtually require PPM focusing structures due to the weight and power savings. L-3 ETI space TWTs commonly weigh less than 1 kg. Figure C-8 illustrates the manner in which the PPM focusing system and the RF circuit are combined together to bring the magnetic field down to the beam periphery. The individual cavity walls are fabricated from high purity iron, subsequently plated with copper to reduce RF losses. The iron channels the 56 Figure C-7 PPM focusing structure. Figure C-8 Coupled-cavity circuit with integral periodic-permanentmagnet (PPM) focusing. magnetic field in a very efficient way to the beam region where its effectiveness is maximized. If such a geometry were not available for this purpose, it is highly unlikely that the typical high-powered TWT could even be focused with available

60 permanent-magnet materials. Generally these beams are very dense and require powerful magnetic forces to hold them together. On the outside of the vacuum envelope, the iron pieces (extensions of the cavity walls) are made large enough to contain most of the magnetic material utilized in the focusing cells. Such a configuration improves the accuracy of alignment of the magnetic field and also gives good mechanical support to the entire assembly. From Figure C-8, it is apparent that some degree of circular symmetry is lost in the PPM geometry due to the presence of the coupling holes and the cooling channels. With the iron pole pieces, it is generally desirable to provide liquidcooling lines close to the cavity walls to minimize the temperature drop from the internal sections of the tube to the outside environment. The presence of the coupling holes tends to introduce undesirable transverse components of magnetic field. The effect of these transverse fields is diminished by the use of double-period, periodic-permanent-magnet (PPM) focusing. Previous to L-3 ETI s patented developments in this field, the fundamental limitations of the PPM design were thought to be so restrictive that it could only be utilized in low-power TWTs where the beam power density is typically quite low. Yet one of the greatest needs for this lightweight focusing method has been at the high-power levels required for many airborne and space applications where tubes with focusing solenoids have been too large and heavy. PPM focusing has not been successfully utilized in TWTs having high average and peak power. It should be emphasized that iron is not a good thermal conductor when compared to copper and, furthermore, the presence of the permanent magnets creates some difficulty in accommodating simple cooling schemes. For low-tomoderate average-power applications these considerations are not important and less complex geometries are then possible. Figure C-9 shows a cross-section view of the L-3 ETI 18714H, a high-power pulsed helix TWT, complete with the focusing structure and external package. The tube is of metal-ceramic construction having a total weight of only five pounds. The PPM focusing structure is composed of round magnet discs shown in the cross-section. Figure C-9 The L-3 ETI 18714H X-band high-powered helix TWT weighs only five pounds and is rated at 1.25 kw minimum peak power output with a 0.04 duty cycle. 57

61 APPENDIX D COLLECTOR DESIGN TRADE-OFFS

62 The Collector The primary function of the collector is to collect the spent electron beam after it has interacted with the signal on the slow-wave circuit. If operated at the same potential as the body of the TWT, the thermal dissipation in the collector would be extremely high and the overall efficiency of the TWT (where the overall efficiency is defined as the RF power output divided by the DC input power) would be very low, essentially the same as the beam efficiency. Fortunately, the RF interaction removes only a fraction of the kinetic energy from the electrons and most of the remaining energy can be recovered by decelerating the electrons prior to collecting them. This is accomplished by operating the collector at a negative voltage relative to the slow-wave circuit. This mode of operation is called depressed collector operation and is illustrated in Figure D-1. The slow-wave circuit and the beam focusing magnets are to the left of the region being illustrated; the curved lines represent electron trajectories into the collector. Space-charge forces between the electrons cause the beam to spread after it has left the focusing region; this has the advantage of causing the dissipated energy to be spread over a surface that is much larger than the cross-sectional area of the beam. Figure D-1 Schematic of singlestage collector. The electron beam leaving the slow-wave circuit does not possess a single velocity. The RF interaction creates a scrambled electron beam that contains electrons having a wide range of velocities. Because of this velocity spread, there will be some slow electrons that must be collected at relatively small amounts of depressed voltage (high voltages relative to the cathode). Operating the collector with too much depression will cause some electrons to be returned to the circuit region where they can cause heating and RF regeneration to occur. Even with this limitation, the efficiency of a TWT can usually be doubled or even tripled when collector depression is employed. Additional improvements in efficiency (and the corresponding reduction in thermal dissipation) can be achieved by using collector designs in which velocity sorting takes place, causing the slower electrons to be collected on stages that are less depressed and the faster electrons to be collected on stages that are more depressed. Designs having two and three depressed stages are shown in Figures D-2 and D-3. The overall efficiency of a TWT having a collector with four depressed stages is usually in the range 35 percent to 70 percent, the higher efficiencies being found only in narrowband (< 20% bandwidth) applications such as space communications in which efficiency is of foremost priority. In each case, the complexity introduced by the use of multiple depressed stages must be weighed against the performance improvements and any constraints imposed by the using system, especially in view of the fact that the law of diminishing returns sets in very quickly when the number of depressed collector stages is considered. The following numbers are typical for a high efficiency communications TWT: 60

63 Figure D-2 Schematic of two-stage collector. Figure D-3 Schematic of threestage collector. Efficiency with one depressed collector stage: 50 percent Efficiency with two depressed collector stages: 58 percent Efficiency with three depressed collector stages: 64 percent Efficiency with four depressed collector stages: 66 percent Space satellite TWTs commonly use 4 depressed stages. Such high efficiencies also partially result from using special low secondary emission ratio surface treatments or materials (such as carbon or graphite) for the elements that serve as collector stages, in lieu of the copper that is normally used in this application. Dissipating the Beam The collector dissipates the electrons in the form of heat as they emerge from the slow-wave structure. This is usually accomplished by thermal conduction to a colder outside surface where the heat is absorbed by circulated air or liquid. The heat is even radiated away by white metallic fins in certain space satellite applications (the TWT is said to be radiation cooled ). The specific collector design to be employed is determined by the method of cooling used and the amount of energy that must be dissipated. It has long been recognized that significant power savings could be obtained in a TWT by decelerating the spent electron beam and collecting it at a reduced potential. Depressed collector operation will, therefore, improve the tube efficiency, as well as Figure D-4 Multi-stage collectors offer substantial power savings during small-signal operation or without RF drive resulting in a near constant thermal load. reduce the thermal load of the collector and simplify its cooling requirements. Multi-stage collectors also offer the advantage of substantial power savings during small signal operation or without RF drive. This, in turn, offers a near constant thermal load, as shown in Figure D-4, for a 10-watt device. 61

64 APPENDIX E TWT PACKAGE DESIGN AND COOLING METHOD TRADE-OFFS

65 The TWT Package Typical packaged TWTs are illustrated in Figures E-1 and E-2. The TWT in Figure E-1 is a millimeter-wave TWT for radar applications. The package serves as a mechanical support structure for the TWT and the RF input and output connectors, a thermal path for the conduction of waste heat, an EMI shield, and as a protective cover over the high voltage connections and focusing magnets. The TWT in Figure E-2 is a high power coupled-cavity TWT for ground based or airborne applications. Since the structure of a coupled-cavity TWT is more massive than for helix TWTs, the package parts are basically added to the TWT assembly (as opposed to placing the TWT assembly inside a package). The package provides ports for cooling fluid as well as paths for conduction cooling. Figure E-1 L-3 ETI Model 8907H 130 watt, 35 GHz helix TWT is suitable for radar applications. Figure E-2 L-3 ETI Model 18702H 50 kw peak coupled-cavity TWT with PPM focusing provides high average power RF at X-band. It is suitable for surface and airborne applications. 64

66 The Vacuum Envelope One of the major disadvantages of early helix circuit tubes was the fragile glass vacuum envelope that was used to enclose the tube parts. Long ago, however, microwave tube manufacturers switched to all metal-ceramic construction. Tubes fabricated this was can not only withstand higher G loads, but can be vacuum processed at higher temperatures typically 550 to 600 degrees C as opposed to 450 degrees C in the case of glass structures. This ensures considerably more complete bake-out of undesirable gasses entrapped in the tube, providing improved reliability at higher tube operating temperatures. In a practical TWT, attenuators (lossy sections) are placed along the slowwave structure to provide stability by absorbing internal and external mismatch reflections. The attenuators also isolate external system components on the output arm from those on the input arm. A typical high-gain TWT will provide up to 80 db or more isolation or cold insertion loss. Without this loss added to the internal structure, it would be possible for reflected RF power to travel back to the input, causing regeneration. In a high-gain device this would, in turn, cause regeneration or even a self-induced oscillation. The TWT package must be designed to prevent coupling between the TWT input and output. Sources of signals that contribute to this coupling include the signals that are radiated from output RF connections, the collector, and the collector leads. These signals can couple into the input RF connections, the electron gun structure, and the gun leads. Since that portion of the slow-wave structure dedicated to attenuation does not contribute to the gain of the tube, the effect of adding attenuation increases the length of the device. The higher the gain, the more attenuator sections will be required. A rule-ofthumb is about 30 db per section, so a tube with greater than 60 db of gain would probably have three active sections and two attenuator sections. TWT Cooling Methods Depend on the Application Cooling methods employed in TWTs are as diverse as the applications. TWTs used in spacecraft are nearly always conduction cooled; the spacecraft structure then transfers the heat to surfaces that radiate the energy into space. The TWT package is designed to limit the peak thermal flux density to values in the range 5 to 15 Watts per square inch at the interface between the TWT baseplate and the spacecraft panel. This peak value is usually encountered at the collector end of the TWT. In some instances, the TWT is designed so the collector itself radiates thermal energy into space. TWTs for other applications can be cooled by using forced liquid or forced air cooling, usually as a supplement to the conduction cooling that takes place through the TWT baseplate. Liquid cooling systems usually employ flow rates in the range 2 to 20 gallons per minute; this cooling technique is usually used when the power dissipated is in the range 1 kilowatt to 50 kilowatts. Air cooling systems usually employ air flow rates of 20 to 200 cubic feet per minute, at pressures ranging from 0.5 to 3 inches of water, and can handle dissipated powers in the range 100 Watts to 5 kilowatts. 65

67 APPENDIX F POWER SUPPLY INTERFACES

68 Power Supply Interface The TWT and the power supply are the key elements in any power amplifier. Equally important are the interfaces between these elements to insure optimum performance and maximum life of the resultant traveling-wave tube amplifier (TWTA) or high-power amplifier (HPA). There must be a continuing interaction between the TWT and power supply designers since the design of both elements must go far beyond voltage and current requirements. Cooling, protection of both the TWT and the power supply, mechanical and control interfaces require detailed attention. As a result, today s high performance TWTAs and HPAs are not a power supply designed around a TWT, but rather the TWT and power supplies are designed as integrated elements (Figure F-1) to meet performance and reliability criteria. Figure F-1 Schematic of TWT and power supplies. The following are some of the interface considerations used in the design and implementation of L-3 ETI HPAs and TWTAs for commercial, military, space, satellite earth stations, radar, ECM, and instrumentation applications. Power supply regulation need be only as tight as required to meet RF performance characteristics and to prevent defocusing of the TWT electron beam as a function of line, load, and environmental changes. For example, in many cases the TWT collector supply need not be highly regulated and can often be left unregulated to vary with line fluctuations. In this case, a highly regulated power supply may be overly complicated and reduce reliability. Ripple from the power supply and the TWT pushing factors will determine the amount of signal distortion contributed by the TWTA in the areas of amplitude, phase, and frequency modulation. Less sensitive electrodes require less power supply filtering. (See Appendix G.) Conversely, more sensitive electrodes (higher pushing factors) require more power supply filtering to meet sophisticated systems requirements. 68

69 On/off control of the TWT is not only important as an operational consideration, but also from a protection viewpoint. L-3 ETI HPA/TWTA circuitry can rapidly detect high helix or body current, arcs, high reflected power and other abnormal conditions. But unless the TWT beam power can be removed equally rapidly, the value of this protection feature is compromised. Time-out circuitry must be designed to provide a reasonable TWT warm-up time from a cold start and the minimum down time in the event of a momentary power outage. Both direct and proportional type heater timing circuits must assure that the TWT cathode is at the proper operating temperature before the beam power is turned on. Improper timing can cause TWT outgassing and possible failure. Body/helix current overload protection is a critical consideration in the design of any high-power amplifier. Abnormal TWT defocusing can occur as a result of improper power supply voltages, RF overdrive, output high reflected power, and other conditions. The object of a body/helix protection circuit is to limit the amount of time that defocused electrons can intercept the slow-wave structure to prevent TWT failures. This protection circuitry must be fast-acting and tolerant of normal intercept currents due to TWT aging and turn-on/turn-off characteristics. Other interfaces such as mounting, cooling, thermal protection, and fail-safe power supply circuitry must be considered to prevent damage to the power supply and TWT in the event of a failure of one of these elements. RF input and output interfaces must not place unnecessary stress on the TWT RF connection and vacuum windows. In summary, both the TWT and power supply should be designed together from the ground up to insure proper interfaces and a highquality amplifier for a specific application. 69

70 APPENDIX G TWT PARAMETERS THAT AFFECT SYSTEM PERFORMANCE

71 Evaluating TWTs The basic considerations in selecting a TWT for a specific application are center frequency, bandwidth, and power output. A number of other parameters must be considered, however, in the specification process. Power versus Frequency When discussing the power capability of pulsed TWTs, it is important to make a clear distinction between the peak and average power since these two numbers are limited by totally different considerations. The considerations that relate to average power of pulsed TWTs are essentially the same as the considerations relating to CW (continuous wave) operation; the power at a given frequency is almost always limited by thermal considerations relative to the RF propagating circuit. The electron-beam focusing is never perfect, and a sizable fraction of the total beam power is intercepted by the RF circuit. At some point amenable to calculation, the circuit temperature approaches the melting point of the metal used in the slow-wave circuit structure (or the Curie temperature of iron in the case of PPM-focused coupled-cavity TWT designs in which the cavity walls are made of iron). In both cases the tube is close to destruction, and this condition defines the average power or CW capability of that device. Peak RF power capability is closely dependent upon the voltage for which the tube can be designed. The beam current is determined by the gun perveance: 72 where: I beam = KV k-a 3/2 V k-a is the cathode-to-anode voltage K is the electron-gun perveance The beam power is the product of the beam current and the beam voltage: where: Beam power = I beamv beam = KV k-a 3/2 V beam V beam is the beam voltage (the cathode-to-circuit voltage) With the available techniques for the design of solid-beam electron guns with good optics, the perveance is generally limited to a value not much greater than (MKS units) and most existing power tubes utilize a value between 1 and Once the perveance is fixed, the required voltage for a given peak beam power is then uniquely determined. This, in combination with practical efficiency values, fixes the peak power which the design will support. In turn, the voltage uniquely establishes the circuit parameters. For this reason, it can be seen that a TWT can only operate at the design voltage. Theoretically and practically, these limits for coupled-cavity TWTs should be very close to those values which apply to high-power klystrons, since the basic considerations are identical in the two cases. Historically, high-power klystrons came first and, consequently, most of the early multi-megawatt radars were designed with klystrons as the output amplifiers.

72 There are a few notable exceptions today, but, in general, high-power TWTs are available at power levels up to 500 kw and voltages up to 80 kv. If a requirement should develop for a particular system, there is no reason why a multi-megawatt TWT should not be considered. To give an abbreviated picture of some of the more popular current TWT designs, Figures G-1 and G-2 illustrate the difference between peak and average power capability and the difference between PPM- and solenoidfocused coupled cavity TWT designs. The curves follow the general characteristic defined by Power frequency = constant which is different from the popular precept of power varying as the inverse square of the frequency. Which rule is correct? Both are, but one must be careful how they are applied. If a given design is scaled over a limited frequency range and the thermal stress is to be maintained at a constant within the circuit, the linear relationship applies. If one desires to scale a particular device to its ultimate limit in terms of power and frequency such that all of the key parameters are pushed to the state-of-the-art (that is, beam density, cathode loading, magnetic field, voltage, etc.), then the quadratic dependency is more appropriate. It will be noted that if the voltage is increased, the peak and average power capabilities increase considerably. This variation is a direct consequence of the way in which the circuit dimensions and the peak beam power increase with voltage. The larger circuit will accommodate a greater amount of thermal dissipation and the higher beam power will permit more peak RF power. At some point the peak RF power will be limited by waveguide arcing problems and voltage breakdown in the electron-gun region. The curves shown go up to 65 kv, which is certainly not the limit, but encompass the great majority of TWTs in field operation. The CW curve is shown for 20 kv, which is a voltage region representing a good compromise between voltage insulation problems in the power supply and circuit size for reasonable thermal stress levels. The upper boundary of the curve is a conservative design boundary and can easily be exceeded by a factor of two for special applications requiring more average power. Here again, the same rule that the power-frequency product is a constant is maintained for the same reasons previously stated. The curve shown does not indicate bandwidth capability, even though this parameter affects average power capability. In general, for very large percentage bandwidths, the average power capability may have to be reduced as much as 50 percent. Efficiency During the early years of TWT development, when the emphasis was on bandwidth, gain, and noise figure, most TWTs were regarded as lowefficiency 73

73 Figure G-1 Peak-and-average-power capability of typical TWTs in field use. amplifiers compared to the more conventional microwave sources such as klystrons and magnetrons. Industry was initially slow to change this viewpoint, believing that inherent in the energy transfer process between the electron beam and the RF wave were physical constraints causing the low efficiency. Certainly the experimental evidence from a large number of designs indicated a typical efficiency of 10 percent or even less. Klystrons at comparable frequencies and power levels gave more than 30 percent conversion efficiency during the same period in time. The price for large bandwidth capability was thought to be a poor-to-modest efficiency. 74

74 Figure G-2 CW power capability of TWTs operating at nearly 20 kv. Only slowly did researchers discover the key parameters which had to be carefully controlled to significantly improve upon this picture. Improved control of the electron beam trajectories contributed a great deal in changing the situation since the efficiency enhancement achievable with simple depressed collectors was very encouraging. The age of the desktop computer has had a significant impact on TWT efficiency improvement. Computer controlled machining and manufacturing have resulted in TWTs being built with more precision than ever before. Experimental measurements have improved in accuracy. In addition, as available desktop computing power has steadily increased, roughly following Moore s Law, theoretical TWT design models which predict TWT performance 75

75 have been able to incorporate more design detail and fewer assumptions, and the predictive capability of the modeling has steadily improved. The 50 percent efficiency barrier, once thought of as an upper limit dictated by nature, has, like the 4-minute mile, long been broken. The new millenium is seeing space communications customers routinely asking for TWT efficiencies higher than 65 percent, and TWT designers, competing for customer dollars, are edging the bar ever higher. There are two fundamental mechanisms whereby increased efficiency can be realized in a TWT amplifier. The first mechanism is collector depression, using a series of collectors which can be depressed well below the circuit potential so that unused energy can be recovered from the spent electron beam. Such a collector must be carefully designed optically so that a minimum number of electrons are turned around and collected on the circuit. The optics of such a system are quite complicated and depend not only upon the geometry of the collector segments, but also upon the degree of RF modulation, the magnetic field used to focus the beam, the yield of secondary electrons at the collecting surfaces, and the relative potential of the circuit and all segments of the collector. Two-dimensional and three-dimensional computer simulations of the electron beam inside the collector greatly aid the life of a collector designer today, but the design is still quite challenging. The main advantage of working with the collector to enhance the overall efficiency is that such an alteration does not affect the circuit of the TWT and can, therefore, be accomplished independently of the RF design of the tube. The power supply, of course, must be properly designed to take advantage of the energy recovered by the depressed segments of the collector. In a TWT having a well-designed multiple-stage depressed collector, the power dissipated by the TWT is nearly constant as RF input drive is varied. The power consumed by a TWT having a multistage depressed collector, as a function of RF output power level, can be approximated by an expression of the form: where: power consumed by TWT = A + B x RF power output A is the power consumed by the TWT with no RF drive, B is a constant with a value of 0.8 to 0.9, depending on the design of the TWT. The second mechanism of efficiency enhancement is sometimes referred to as velocity resynchronization. It is well known that the electron beam slows down in velocity when it gives up energy to the amplified RF wave on the circuit. As a result, the propagating wave and the electron beam progressively lose synchronism, with the wave moving far ahead of the beam. When this occurs, the electron bunches are no longer favorably phased to give up energy to the moving wave and the amplification process ceases before maximum signal level is achieved. Early models of TWTs almost always gave disappointing efficiencies because of this unfavorable condition within the tube s interaction region. 76

76 First-order corrections for a lack of synchronism between the beam and the wave can be done with a mere increase in the operating voltage. Unfortunately, this step causes the amplification process to suffer in the small-signal region (near the input) of the circuit and it also causes the linearity of the output versus input curve to degrade. It would appear that an easy solution to this problem would be to make the RF wave velocity on the circuit such that synchronism could be maintained everywhere. This simple concept resulted in the tapered velocity -type circuit, which is one of the basic tools that the TWT designer now virtually always employs. In practice, there are an infinite variety of tapers that differ in the degree of velocity change and the variation of velocity as a function of distance along the slowwave structure. Modern desktop computers, by simulating a nonlinear interaction process between a stream of highly bunched electrons and a growing electromagnetic (RF) wave propagating along the TWT circuit, can quickly analyze thousands of tapers and find a velocity taper which optimizes TWT performance. Any velocity taper choice must be a result of a series of compromises which trade one desirable effect for another. For example, in ECM systems, octave bandwidths are quite common. Over such an extreme bandwidth, the important electrical parameters, which are proportional to wavelength, change by a factor of two or more. It is not surprising that optimum conditions within the velocity taper region are impossible to maintain. In spite of this problem, results from efficiency enhancement schemes for ECM power amplifiers are indeed impressive. Figure G-3 illustrates the efficiency of a kilowatt helix-type high-gain ECM TWT. It should be noted that the frequency coverage is 1.5 octaves with an efficiency above 45 percent. Harmonic Injection Another method of efficiency enhancement, unique to broadband helix devices, does not involve any alterations of the internal parts of the TWT. It is generally referred to as harmonic drive because it is associated with special adjustments made to the second-harmonic content of the input RF signal. The solid portion of the curve of Figure G-3 shows the additional enhancement which is afforded with harmonic drive. The phenomenon was discovered quite some time ago Figure G-3 Broadband high-efficiency TWT. 77

77 when it was observed that the wrong type of second-harmonic input would seriously degrade the power output at the fundamental frequency. On the other hand, the correct amount of second harmonic, properly phased, will increase the fundamental power output and suppress the second harmonic at the output of the amplifier. The process is one of cancellation, whereby the injected second-harmonic signal is such that it is 180 degrees out of phase with the second-harmonic signal generated by the nonlinear processes inherent in the interaction mechanism. With careful design of the input circuit, this cancellation can be made reasonably non-critical and quite broadband. The effect is only important at the low end of the tube s amplification band since the second harmonic of these frequencies still lies within the amplification band of the TWT. Above midband, there is no appreciable amplification of the second-harmonic signals and, consequently, the enhancement scheme is not effective above this point. To provide the correct harmonic input signal, a simple circuit consisting of a phase shifter and a microwave diode can be utilized to transform a pure drive signal to one with a significant second-harmonic component. If the drive signal emanates from an overdriven TWT (one operating well into saturation), it is quite likely that the second-harmonic portion is large to begin with and of the wrong phase. It is difficult to compensate for such a drive signal since the adjustments will be generally quite critical and subject to change as the drive level changes. Harmonics Due to the wide bandwidth and high gain of the TWT plus the fact that in saturation, the tube acts as a non-linear device there will be harmonics in the RF output spectrum. Typically, at saturation for a narrowband TWT such as a space communications helix TWT, the second harmonic will be 15 to 25 db below the fundamental signal. However, very broadband devices will have a higher second-harmonic content, and this number can vary widely among TWT designs. Other higher-order harmonics will also be present to a lesser degree. The harmonic magnitude is a function of the fundamental frequency and bandwidth range, with the lower band edge signals having the greater effect. Gamma-a (ga) A commonly computed TWT value called gamma-a, or γa, is a measure of how much the radial profile of the rf field varies, and it is one indicator of how well a TWT will perform. This quantity can be estimated as follows: f (GHz) H.M.D.(in) γa = 4. 25, E K (kv) where γa is dimensionless, f is the TWT operating frequency in GHz, H.M.D. is the helix mean diameter or coupled cavity beam hole diameter in inches, and E K is the cathode voltage in kv. A typical value of γa for an efficient TWT is For higher frequencies, due to mechanical constraints, γa typically increases to values above 1.0. For γa values above 1.5, the TWT performance (both efficiency and linearity) will start to significantly degrade. For values too 78

78 low (less then 0.8), harmonic signal content may be too high and therefore affect the TWT efficiency. Intermodulation Distortion When more than one carrier is introduced at the TWT input, a mixing, or intermodulation (IM) process, takes place via the TWT interaction. This results in intermodulation products which are displaced from the carriers at multiples of the difference frequency. The power levels of these intermodulation products are dependent on the relative power levels of the carriers and the linearity of the TWT. In the case of two balanced carriers, Figure G-4 shows the variation of carrier and IM product power level with total drive power. The single carrier power curve is also plotted for comparison. As in the case with AM/PM conversion (see below), the IM distortion is significantly reduced in the small-signal (linear) region of the RF drive range. In the small-signal region, the third-order IM output power decreases by 3 db for every 1 db decrease in input signal drive power. For this reason, commu-nications TWTs are often operated well below their saturation power level. Transfer Curves The drive characteristics of a typical TWT are shown in Figure G-4. The threshold input signal level for useful operation is determined by the bandwidth and noise figure of the tube. The dynamic range is that region between the threshold input level and the input at which there is departure from small-signal or linear gain. The gain continues to decrease as the input level is increased and is decreased by approximately 5 to 8 db at the point of saturated power output. The overdrive capability of a TWT indicates the range over which the output power will remain in the saturation region as input is increased. When additional input is applied, output Figure G-4 Typical third-order intermodulation data for a helix TWT. power decreases. For certain applications, it is desirable to maintain full or saturated power output over a broad range of input signal 79

79 conditions. Limiter actions to achieve this objective are implemented in the design of the RF structure, by using multiple attenuators, and by cascading two TWTs with additional equalizers and isolation filters. In other applications, the TWT simply must not be damaged even if overdriven by an input signal 20 db higher than the saturating input signal. While this is not an inherent difficulty, it must be considered when desiging and packaging the TWT. AM/PM Conversion Amplitude modulation/phase modulation (AM/PM conversion) is defined as the change in phase angle between input and output signal as the input signal varies. This factor is expressed in degrees per db at a specified value of power output. AM/PM conversion in a TWT is caused by the reduction in beam velocity that occurs as the input signal level is increased and greater amounts of energy are taken from the beam and transferred to the input RF wave. At a level 20 db below the input required for saturation, AM/PM conversion is negligible. Beyond this point, AM/PM conversion increases sharply. Typical power output and relative phase shift characteristics are shown in Figure G-5 for a communications TWT. Here it is seen that phase shift is relatively insensitive to drive in the small signal ( linear ) portion of the RF output power characteristics. As the TWT is driven toward and beyond saturation, the rate of phase change increases. The slope of the phase curve, or AM/PM conversion, is plotted against RF drive in Figure G-6. Figure G-5 Typical power output and phase shift as a function of RF input power for a communications-type TWT. 80

80 Figure G-6 For a single carrier condition, AM/PM conversion rises sharply as drive is increased. The peak AM/PM generally occurs at the saturated drive level, though in certain TWT designs it may occur at a drive level from 3 to 10 db below saturation. The values of AM/PM conversion and phase change generally both increase with frequency over the passband of the TWT. The curves from Figures G5 and G6 show typical performance at the high end of the band, which is the worst case. AM/PM Transfer AM/PM conversion applies to the case of a single carrier. For two or more carriers, transfer takes place, giving PM at the output on one carrier due to AM at the input on the other. The general trend with drive is similar to that for AM/PM conversion, but the specific values are different and are also a complicated function of the relative carrier amplitudes. A common (usually worst case) approximation for the AM/PM transfer value is to use twice the AM/PM conversion value. Phase Sensitivity Any factor which affects the velocity of the electron beam will give rise to phase changes in the RF output signal. If the distributing factor varies with time, then the result will be phase modulation of the input RF signal. The primary factor affecting the velocity of the beam is the cathode voltage. The other voltages, the magnetic fields and cathode temperature have secondary effects. The power supply designer must take into consideration these phasepushing factors when designing the power supplies for a TWT, since the system 81

81 noise requirements will dictate the power supply ripple and stabilities that must be maintained. Typical pushing values for TWTs are: 35 degrees per 1 percent change in cathode voltage 5 degrees per 1 percent change in grid drive 2 degrees per 1 percent change in anode voltage 0.01 degrees per 1 percent change in filament voltage. These values should be considered as order of magnitude since the actual value for any specific tube will be a function of many factors such as type of circuit, gain, perveance, etc. A more exact formula for the phase pushing value of a TWT is listed later in this appendix (under the section on interaction between TWTs and EPCs). GAIN FLATNESS, PHASE LINEARITY, AND GROUP DELAY Broadband Variations TWTAs which are required to operate over a broad range of frequencies are often required to produce a fairly flat RF output signal over that same frequency range. However, imposing a tight specification on the flatness will often force the TWT designer to take measures which will result in a reduction in the efficiency of the TWTA. Meeting a tight specification on RF output flatness usually requires that the cathode voltage be somewhat lower than the voltage which would produce the greatest efficiency. Operating a TWT in this way is often referred to as under-voltaging the TWT. Communications TWTs are also sometimes under-voltaged in order to reduce the AM/PM at the expense of efficiency. In addition to the restriction on the cathode voltage, the TWT designer must select a slow-wave structure design which is consistent with the flatness requirement. For a helix-type TWT, this means that the diameter of the metal wall which surrounds the helix is reduced to load the slow-wave on the helix and thereby flatten its frequency response. For a coupled-cavity TWT, the designer must select a passband characteristic which is sufficiently broad to produce a flat frequency response across the band of interest. In either case, the interaction impedance and, therefore, the efficiency of the TWT is compromised. Combining the effect of the restriction on cathode voltage and the effect of the broadband requirement upon the slow-wave structure design will result in a reduction in TWTA efficiency of 2 to 10 percentage points. The TWTA efficiency achievable depends upon the severity of the bandwidth and RF output flatness requirements which are required of the TWTA. Narrowband Variation Within narrow bandwidths the dominant effect upon RF output flatness is gain ripple. Gain ripple is nearly always caused by some feedback path within or around the TWT. The effect of the signal traversing the feedback path can be regenerative or degenerative, depending upon the closed loop phase shift. Since a TWT is electrically long (a typical TWT has a phase shift of about 10,000 degrees, or about 28 wavelengths), a small change in frequency can cause the closed loop phase shift to change by 360 degrees. In most TWTAs, the frequency periodicity (the frequency range resulting in closed loop phase change of 360 degrees) is 100 to 200 MHz, as seen in Figure G-7. 82

82 Some of the feedback paths which can contribute to TWTA gain ripple are: Feedback outside the TWTA, from RF output port to RF input port. This effect is usually negligible if proper care is taken to insure that RF energy will not leak out of the RF connectors and components (such as isolators) used at the RF ports of the TWTA. Returning electrons within the TWT electron beam. This effect is negligible if proper care is taken in the design of the TWT collector, the electron beam is carefully focused to insure that all electrons are collected at the collector surfaces, and the proper voltages are applied to the various stages of the collector. Reflections at the attenuator within the TWT at the end of the input section of the slow-wave structure. The path for this feedback loop is completed by reflections at the RF input port of the TWTA. This effect is usually kept within limits by careful design of the attenuator within the TWT and by using an isolator at the TWTA RF input port. Some TWTs have two severs. The center section of slow-wave structure has an internal attenuator at each end. Reflections from these attenuators can create a feedback loop. This effect is usually kept within limits by careful design of the attenuators used within the TWT. Figure G-7 Gain ripple versus frequency for a typical helix TWT. Reflections at the RF output port of the TWTA. The path for this feedback loop is completed by reflections from the attenuator at the beginning of the output section of the slow-wave structure. Again, careful design of the internal attenuator will reduce the resulting gain ripple. This feedback path is usually the dominant path because the gain in the output section is 83

83 higher than the gain in the input section, since high gain in the output section is needed to produce high efficiency. Using an isolator at the TWTA RF output port will help to control the amount of signal reflected from components beyond the TWTA RF output port. Reflections between imperfections along the slow-wave structure. This effect is usually controlled by exercising proper care in the selection of the components which make up the slow-wave structure. Periodically repeating imperfections (i.e., a pitch variation every 10th turn of a helix) are particularly detrimental in this respect. Analysis of gain ripple periodicity is often an important diagnostic tool in localizing the cause of gain ripple. For example, a frequency periodicity which is consistent with the length of slow-wave structure between the TWT attenuator and the TWTA RF output port would localize the problem. Improvement would result if the reflection coefficients at the attenuator and at the RF output port were reduced. Round-trip signal time is directly equal to the inverse of the ripple period. If the group velocity of the TWT, which can be approximated from the operating voltage, is also known, then the distance between mismatches (corresponding to half the round-trip time) can be estimated: v E (kv) group (in/ µ s) K Mismatchdistance (in) = 369, 2 f (MHz) f (MHz) where v group is the TWT group velocity in inches per micro-second, f is the ripple period in MHz, and E k is the cathode voltage in kv. The primary trade-off against gain ripple is the cost of implementing the necessary controls and screening over the parts used in the slow-wave structure. If these controls are implemented and the using system presents a low reflection coefficient to the TWT RF output port, it is possible to achieve a small signal gain ripple having a peak-to-peak amplitude of less than 0.5 decibel. The amplitude of ripple on the saturated gain or saturated power output would be even less of the limiting effects which take place at saturation. Lowering the gain of the individual TWT sections also reduces gain ripple. For each 6 db reduction of gain in the primary feedback loop, the gain ripple amplitude is cut in half. Phase Non-linearity or Time Delay Distortion The same mechanisms which can cause gain variations as a function of frequency will also cause phase non-linearities. Time (or group) delay is the derivative of the curve of phase versus frequency. Since time delay and phase are so closely related, the mechanisms which cause phase non-linearity are responsible for time delay distortion. Phase non-linearity and time delay distortion effects are usually resolved into broadband and narrowband effects, just as was the case for gain variations. The curves shown in Figure G-8 show both the phase versus frequency and time delay versus frequency for a hypothetical TWT. The frequency band of interest is also shown in the figure to illustrate the fact that, for many applications, the narrowband effects dominate over the broadband effects. This is true when the TWT is specified for operation over only a small portion of the pass-band of the 84

84 slow-wave structure, causing band-edge effects to be far removed from the frequency range of interest. Figure G-8 Phase non-linearity and time delay distortion versus frequency. Ripple in the phase and time delay curves, as illustrated in Figure G-8, results from the same feedback paths that cause gain ripple. A simple theory can be used to predict ripple on the phase and time delay curves from measurements of TWT small signal gain ripple. This is illustrated in Figure G-9, which demonstrates the relationship between small signal gain ripple amplitude and small signal phase ripple amplitude for a small interfering signal (less than 2 db ripple amplitude). A more exact relationship between small signal phase ripple amplitude and small signal gain ripple amplitude is as follows: 85

85 G (db)/20 1 ϕ (deg) = 6.60 G (db), π 10 G (db)/ where ϕ is the small signal phase ripple amplitude (or peak-to-peak) in degrees, and G is the small signal gain ripple amplitude (or peak-to-peak) in db. The relationship to group delay is then: t (ns) = π ϕ (deg) 0.18 f (MHz) 115 G (db), f (MHz) Figure G-9 Relationship between gain ripple and phase ripple. where t is the group delay ripple amplitude (or peak-to-peak) in ns and f is the ripple period in MHz. If the ripples are assumed to be purely sinusoidal (a good assumption), then the worst-case slopes for gain ripple, phase ripple, and time delay can all be computed from the gain ripple amplitude (for small interfering signals): 86

86 π G (db) pk - pk gain ripple slope (db/mhz), f (MHz) phaseripple slope (deg/mhz) 20.7 G (db) pk - pk, and f (MHz) 361 G timedelay slope (ns/mhz) pk - pk [ f (MHz)] 2 (db), where here, G pk-pk must be the peak-to-peak gain ripple (twice the gain ripple amplitude). Finally, when the bandwidth of interest is less than half of the gain ripple period: G (db) ϕ (deg) = sin( π B/ f ), and π G (db) G (db) sin( π B/ f ) t (ns) = 2000 G (db) f (MHz) where B is the bandwidth (same units as f). Phase Tracking In many systems, a requirement exists to operate TWTs in parallel. In this type of operation it is important that the phase variations of those tubes operated in parallel are as close as possible to being identical. It is not so important that the total phase shift through the tubes be the same, since a phase shifter can be employed to adjust the phase delay through the tubes at any particular frequency. It is important, however, that the variation over the frequency band of interest be similar. Helix TWTs can be designed to have tube-to-tube phase tracking of five degrees or less, provided care is taken to minimize reflection effects and fine grain variations. It is more difficult to maintain close phase tracking in a couplecavity tube and variations of between 10 and 15 degrees are to be expected. Noise Figure The noise figure (F) of an amplifier is a measure of the degradation in signal-to-noise (S/N) ratio with passage of the signal through the amplifier and can be expressed as follows: 87

87 S / N N + G a a F (db) = 10 log i i = 10 log S /N G N o o a i where S i and N i are the input signal and noise levels, respectively, S o and N o are the output signal and noise levels, respectively, N a is the noise added by the amplifier, and G a is the gain of the amplifier. The IRE (forerunner of the IEEE) adopted 290 K as the standard temperature for determining noise figure, and since the input noise level is usually thermal noise, it is often referred to by N i = kt ob, where k is Boltzmann s constant, T o is 290 K, and B is the bandwidth (Hz). The value of kt o at 290 K is 174 dbm/hz. The primary source of noise in a TWT is related to the density and electron velocity variations within the electron beam. The level of the noise power is related to the number of electrodes in the gun, the size of the electron gun, and its beam optics. Unfortunately, noise figure is one of the most difficult TWT parameters to accurately predict, even with modern numerical simulations. For medium power tubes, the noise figure is typically 30 to 35 db and increases with power and frequency. A simple formula from Pierce s Traveling Wave Tubes predicts that noise figure (in absolute units, not in db) scales roughly as E k (2/3) and I k (1/3), where E k and I k are cathode voltage and current, respectively. Noise Power Output and Carrier-to-Noise Ratio For some applications, the noise power output (NPO) is of prime importance and may be measured by terminating the input and measuring the NPO at the output of the TWT. The NPO can be estimated from the following equation with all parameters stated in db: NPO (in dbm) = log (BW) + (G SS) + (F), where -114 dbm/mhz is the thermal noise caused by a room temperature termination at the input, BW is the bandwidth in MHz over which the NPO is to be estimated, G SS is the average small signal gain in db across the bandwidth, and F is the average noise figure in db across the bandwidth Since the gain and noise figure are not constant across a wide bandwidth, a more accurate estimate could be obtained by integrating the noise figure-gain product across the total bandwidth. Unless a filter is used to provide a welldefined output bandwidth, the noise figure-gain product should be integrated across the entire range of frequencies where the TWT has appreciable gain. The carrier-to-noise ratio is also a TWT parameter of common interest. This value is exactly what it says: the ratio of the TWT output (carrier) power at some defined operating point (commonly saturation) to the surrounding TWT noise power density: N i [ 174 dbm/hz + G ( db) F ( db) ] C/N (db Hz) = P (dbm) + out noise,, 88

88 where C/N is the carrier-to-noise ratio in db-hz, P out is the single carrier output power in dbm, F is the TWT noise figure in db, and G noise is the gain of the noise in the TWT in the presence of a strong single carrier. Typically, in the presence of a single carrier, the gain of the noise is suppressed. It can be as much as 10 db below small signal gain (depending on the carrier strength), but it is safest to use G noise = G ss as a worst case condition. Dynamic Range for Linear Operation The linear region is sometimes defined as the range for linear operation up to the point where increasing the RF input signal results in a gain compression of 1 db. (See Figure G-10.) This point is called the 1 db compression point. Therefore, the dynamic range for linear operation may be defined as the ratio of this maximum input signal to the reference noise level of the TWT. Figure G-10 One db gain compression point. Spurious Outputs and Stability Spurious outputs are minimized and stability is assured through proper designs of the electron beam optics and magnetic focusing. In addition, the processing and fabrication of the TWTs must be carefully controlled. Spurious oscillation is eliminated or minimized by oscillation suppression techniques, such as special attenuation patterns on the support rods of helix TWTs and loss buttons (L-3 ETI patented technique) in coupled-cavity TWTs. These techniques effectively suppress the tendency to oscillate at the circuit mode edge and render a given TWT completely stable over a wide range of operating voltages, source impedances and load impedances. 89

89 Relative to the stability of a complete amplifier (TWTA), the beam (cathode) power supply is of great importance. This supply determines the velocity of the electron beam, which affects the stability of the TWT. Reliability/life Traveling-wave tubes have, over the years, have gained a reputation for high reliability and long life. There are many factors that affect these parameters, such as the basic design, the interface, protective measures, handling, installation/operation and storage. Basic Design The most important factor relative to life is the type of cathode and the cathode loading factor. In addition, special attention must be paid to electrode size, shape and spacing. The method of packaging must meet the vibration and shock requirements for the application. And to meet wide temperature ranges and high altitude (space) requirements, consideration must be given to the cooling technique; i.e., conduction, radiation, liquid, air, heat pipes, etc. Encapsulation, conformal coating, and, in some cases, the use of dielectric liquid will ensure an arc-free and nearly corona-free device. Interfaces The operation of TWTs must be confined to the limitations of the operating and environmental parameters for which the tube was intended. Some of these parameters can be eliminated and the interface effort minimized by taking advantage of the black box concept; i.e., TWT and power supply are built by L-3 ETI and supplied as an integrated unit. This approach limits the interface to the drive signal, RF input and output loads, and input voltage. Protective Measures Steps must be taken to provide protective measures so that the TWT is not exposed to abnormal extremes, such as voltage surges, temperatures, load mismatches, system arcs, and loss of cooling. All TWTs and TWTAs are supplied with operating instructions and test performance data. Special attention should be given to the recommended precautions and operating instructions. INTERACTIONS BETWEEN TWTs AND EPCs THAT CAN AFFECT SYSTEM PERFORMANCE Several important interactions can occur between the TWT and the EPC, many of which are dependent upon the nature of the RF signal being amplified by the TWT. 90

90 Interactions Which Occur When the RF Signal is Pulsed or Contains Amplitude Variations As mentioned previously, the division of electron current among the various stages of a multi-stage collector is dependent upon the level of the RF signal being amplified, as is the electron current (Iw) intercepted by the slow-wave structure. These changes can cause the voltages at the TWT elements to fluctuate. Fluctuations in the cathode voltage can induce spurious phase modulation on the RF output signal. The fluctuations in the current loading on the various EPC voltages will be reflected back to the primary bus which delivers power to the EPC. These signalinduced variations are especially important if the TWTA is handling signals with large amplitude variations and if other TWTAs are connected to the same primary bus. Modulation on the signal being handled by one TWTA can feed into the EPC voltages of the other TWTAs and cause the RF signals on those other TWTAs to experience spurious modulation. EPCs used in applications where large amounts of amplitude variation are to be accommodated must be designed with the proper amount of regulation and filtering on the primary bus voltage and on the cathode voltage supply. Effect of AC Components on EPC Voltages The use of AC voltage on the heater can introduce a small amount of spurious modulation on the RF signal being amplified by the TWTA. If this spurious modulation is critical, the TWT heater can be operated from a DC voltage. The small amount of AC ripple that would be present on a DC voltage provided to the heater usually produces very small amounts of spurious modulation. Ripple and spurious AC components on the DC voltages provided to the collector (or collectors) of the TWT usually produce very little spurious modulation on the RF signal. As a result, the filtering for the collector voltages is accomplished by inductors and capacitors having small values. Active filtering is not required. Ripple and AC components on the voltage of the anode which controls the beam current (the only anode in TWTs having a single anode; the anode nearest the cathode on multiple anode TWTs) can introduce spurious modulation. Fortunately, the anode draws very little current from the EPC. This makes it possible to achieve adequate filtering using passive components (inductors and capacitors having moderate values). However, the cathode voltage must be well filtered and that filtering usually requires the use of active filter circuits. Ripple and AC components can cause both amplitude modulation (AM) and phase modulation (PM) on the RF signal being amplified by the TWTA. It nearly always turns out that the spurious phase modulation produces the greater effect and we shall concentrate our discussion upon that effect. Most TWT EPCs employ DC-to-DC inverters in which the DC input bus voltage or a DC voltage obtained from an AC input bus voltage is chopped or modulated in a converter operating at a moderately high frequency (typically, 5 khz to 30 khz). The output of this converter is stepped up in voltage by transformers, then rectified and filtered to provide the high DC voltages required 91

91 by the TWT. This scheme results in high efficiency and produces ripple voltages which are easily filtered by small value inductors and capacitors. The AC components on the high voltages result from at least two effects. First, the filtering is never perfect and the DC voltages will contain a small AC frequency component at the fundamental and harmonics of the converter frequency. Second, any AC ripple or transients on the input bus voltage will be stepped up in amplitude by the transformers within the converter circuits and will be impressed upon the TWT voltages unless rejected by the input filter or the EPC regulator circuits. To be thorough, we should understand that there really are other effects, such as noise within the regulator circuitry, stray pick-up from nearby equipment, etc. However, these other effects are usually small compared to the two effects which have been described. Fortunately, it is relatively simple to calculate the phase modulation produced by AC components on the TWT cathode voltage. The TWT produces several thousand degrees of phase shift between the RF input signal and the RF output signal, the approximate value of which is given by the following equation: L (in) f (GHz) ϕ (degrees) = 488 E (kv) K where ϕ is the total TWT phase shift in degrees, L is the pin-to-pin length of the TWT in inches, f is the frequency of operation in GHz, and E k is the cathode voltage in kv. The pushing factor of a TWT is a measure of the sensitivity of the phase shift of the TWT to changes in electrode voltages. This pushing factor is approximated from the following equation: df L (in) f (GHz) (deg/volt) = de E (kv) 3/2 K K where dϕ/de k is the TWT phase pushing factor in degrees per volt and the rest of the variables are as in the previous equation. Note that taking the derivative of the total phase shift equation above to get the phase pushing factor is not quite correct due to the approximations involved in that equation. Using the TWT pushing factor and the measured (or predicted) amplitude of AC voltage on the TWT cathode, it is possible to predict the spurious phase modulation (PM) from: Phase modulation = pushing factor peak AC voltage 92

92 If several frequency components are evident on the AC voltages on the TWT cathode voltage, it becomes necessary to calculate the spurious PM generated by each frequency component individually. Switching regulators use waveforms that are rich in harmonic content, and significant frequency components often extend to 100 MHz. Each frequency component will generate a pair of PM sidebands around the RF signal (carrier) frequency. In the strictest sense, each frequency component generates an infinite set of PM sidebands separated by the same frequency spacing as the frequency of the AC component being considered. Fortunately, in a well-designed EPC, the amplitude of the AC voltage is sufficiently small that the higher order sidebands can be ignored and we need to concern ourselves only with the first sideband pair, one of which is above the carrier by a frequency spacing equal to the frequency of the AC component and the other is below the carrier by the same amount. For example, if the AC component is at a frequency of 5 khz, one sideband will be 5 khz above the carrier frequency and the other sideband will be 5 khz below the carrier frequency. The strength of these PM sidebands, relative to the carrier level, can be calculated using modulation theory and is given by: dbc = 20 log J 1(peak phase shift in radians) J 0 (peak phase shift in radians) where dbc is the ratio of the level of each sideband relative to the carrier level expressed in decibels (db), the log is the logarithm to base 10, J 1 is the Bessel Function of the first kind to the first order, and J 0 is the Bessel Function of the first kind to the zero order. Since the peak phase shift values are (if the EPC is designed properly), the Bessel Functions can be approximated by using the first term in the series expansions of the Bessel Functions. The resulting equation is: dbc = 20 log (peak phase shift in radians/2) where the symbols have the same definition as before. Typical TWTAs produce PM sidebands which are 30 to 100 decibels below the carrier (the values for dbc are 30 to 100 decibels). AM sidebands are nearly always at least 10 decibels lower than the PM sidebands. AM sideband levels are calculated from: dbc = 20 log 10peak variation in RF output / 20 1, 2 93

93 where the peak variation in RF output is the variation in RF output (measured in db) caused by AC on the TWT voltages. If the TWT is operated at saturation, reasonably small changes (< 10-20%) in power output can be approximated from changes in cathode current by using the following: P 4 I out (Watt/Watt) = K ( Amp / Amp), P 3 I out K where P out is the change in output power in Watts, P out is the nominal output power in Watts, I k is the change in cathode current in Amps, and I k is the nominal cathode current in Amps. If the TWT is operated in the small signal region (at an RF output power level that is 10 db or more below the saturated level), the change in gain will also relate to changes in cathode current: G 1 I SS (db/db) = K ( Amp / Amp), G 3 I E K where G E is the electronic gain, given by G ( db) = G ( db) + launchingloss (db) sever loss (db), E SS + where G SS is the small signal gain in db. The launching loss is a result of the fact that the input signal produces three waves on the slow-wave circuit: a fast space-charge wave that is not amplified, a slow space-charge wave that is amplified, and a slow wave that is not amplified. The launching loss is usually about 9 db. The sever loss occurs at the attenuator because the wave on the slow-wave circuit is dissipated in the attenuator (amplification continues beyond the attenuator because modulation on the electron beam re-establishes the wave on the slow-wave circuit). The sever loss is about 6 db for each sever. Most TWTs have one sever (i.e. they an input slow-wave circuit, a sever, and an output slow-wave circuit) and the electronic gain is about 15 db greater than the small signal gain. In some cases, high gain is required and the TWT has three active sections (two severs). For TWTs having two severs, the electronic gain is about 21 db greater than the small signal gain. 94

94 APPENDIX H FACTORS THAT AFFECT POWER COMBINED TWTAs

95 by: The output power from a power combiner having no resistive losses is given P = [ P + P + 2(P P ) cos(a) ] where: P is the output from the combiner in Watts P 1 is the output from TWTA #1 in Watts P 2 is the output from TWTA #2 in Watts a is the phase difference between the outputs from the two TWTAs The output power is further reduced by the resistive losses in the combiner and the associated transmission line components. For Ku-band applications, the total combining loss could be less than 0.2 db. The effects of phase and amplitude imbalance are illustrated in the attached figures. The graph in Figure H-1 shows the effect of phase imbalance between output P 1 and output P 2. The graph in Figure H-2 shows the effect of power imbalance between output P 1 and output P 2.. The data in these two graphs show how the combining loss is affected by phase and power imbalances at initial setup. As environmental conditions change and the TWTAs age, the phase and power balance will change. The graph in Figure H-1 may be used to determine the additional combining loss as the phase balance changes. If the power outputs of the two TWTAs (output P 1 and output P 2) change by the same number of db, the combined output will change by that same amount. If the power output from only one of the TWTAs should degrade, the degradation at the combined output may be determined from Figure H-3. Note that the degradation in the combined output is less than half the degradation of the one TWT. A dual EPC may be used for the paralleled TWT approach. TWTs can be selected to operate with common cathode and collector voltages. The dual EPC design approach is based on the following basic features: Common cathode and collector voltages. The EPC will operate and regulate while being loaded by the power requirement of either or both of the paralleled TWTs. A common regulator/converter and common portions of the logic circuits for reduced size and weight relative to two independent EPCs. Independent heater supplies. Each TWT is provided with a heater voltage that is optimum for that TWT. 96

96 Figure H-1 Loss caused by phase imbalance. Figure H-2 Loss caused by initial power imbalance. 97

97 Figure H-3 Degradation as one TWT degrades. 98

98 APPENDIX I MILLIMETER-WAVE TWTs

99 Millimeter-wave Tubes The work which has been done in the development of very high-power millimeter-wave tubes has been largely influenced by the needs of actual systems operating in this frequency band. In general, millimeter-wave tubes utilize very low perveance electron guns, which create some unusual electron-beam focusing problems associated with the proper containment of the thermal electrons. Proper focusing of these thermal electron beams is the primary challenge of millimeter-wave TWT fabrication. Thus improvements in these tubes is often related to improvements in high energy product magnetic materials since these TWTs are invariably PPM focused. Aside from the focusing structure, the major challenge in the manufacture of millimeter-wave tubes is the precision and tight tolerances required for the extremely small circuit parts. To illustrate this problem, Figure I-1 is a photograph of an OFHC copper circuit part in relationship to the head of a regular straight pin. From this picture, one can appreciate the small size and the assembly difficulties involved in the fabrication of circuits utilizing these parts. Figure I-1 Photograph of copper circuit part for a millimeter-wave tube. 100

100 APPENDIX J NOTES ON TWTs FOR RADAR APPLICATIONS

101 Radar Applications Although a German scientist, C. Hulsmeyer, patented the first primitive radar as early as 1904, it wasn t until the mid-1930 s that practical systems evolved. Early systems employed smooth-bore and, at a later date, cavity magnetrons. Today, a wide variety of microwave tubes are employed, including magnetrons, crossed-field devices, klystrons, twystrons and TWTs. Increasing the Range As radar systems have evolved and grown more sophisticated, the need has developed for higher average power devices to detect targets at greater ranges even in the presence of ECM interference. And because tracking capabilities have often been equally important, or duty cycles have had to be limited, peak-power requirements have also increased. The coupled-cavity TWT has proven to be an ideal device for many highpower radar systems. Such tubes are used in mobile, naval and airborne radars. The coupled-activity tube can provide peak powers in the order of hundreds of kilowatts at Ku-band and megawatts at S-band. With solenoid focusing, average powers of as much as 10 to 20 kws are achievable at Ku-band and as much as 60 to 80 kw at S-band. Bandwidths of 10 percent are common for coupled-cavity tubes (bandwidths up to 30 percent are achievable, employing special designs). This bandwidth is sufficient for most systems employing frequency-hopping or frequency scanning modulation techniques. Pulse Compression Pulse compression systems require that the phase linearity of the transmitter be extremely good. Although good phase linearity can generally be achieved in helix TWTs, a coupled-cavity tube must be carefully designed to achieve flat phase performance. To accomplish this, L-3 ETI employs a patented technique to introduce in-band loss. Deviations from phase linearity of only a few tenths of a degree in bandwidths of 40 to 50 megahertz have been achieved. Phase linearity, if held to reasonable limits, will also enable the tube manufacturer to offer TWTs which closely track in phase between tubes. This is of primary importance when tubes are to be operated in parallel. Many systems are being configured using two to four tubes both helix and coupled-cavity in parallel. L-3 ETI has an excellent understanding of the tube design constraints which must be employed to achieve good phase tracking in both types of tubes, and for this reason the company is in a position to work closely with the radar designer to achieve the system requirements. For airborne radar applications, L-3 ETI has employed a solenoid wrapped directly on the tube body. This technique minimizes the size and weight of the TWT and also reduces the solenoid power. The technique, although first employed for airborne tubes, is now standard practice at L-3 ETI for all new solenoid-focused TWTs. Higher Average Power In a system operating with a high pulse repetition frequency, the total average power can be limited by grid heating. Early TWTs were equipped with a 102

102 single grid, and the intercepted current could cause the grid temperature to rise to the point at which the grid would start to emit or ultimately fail. L-3 ETI pioneered the use of the shadow grid for applications requiring a high average power. There are actually two grids in these tubes. The grid closest to the cathode is very carefully placed directly in alignment with the second or control grid and is held at cathode potential. Electrons are not attracted to this shadow grid, but its presence in front of the control grid reduces the current intercepted on the control grid by an order of magnitude. This provides the capability of operating at much higher average power levels without excessive grid heating. Shadow grids are employed in virtually all gridded tubes at L-3 ETI, and even more effective techniques are currently in the development phase. The new techniques promise even further improved tube life and performance. An Optimum Design TWTs have many peculiarities which must be understood to be certain that system performance will not be compromised. It is important, therefore, that the radar system designer work closely with the tube engineer so that the radar performance is optimized. Among the potential problems: Because the tube turns on in a nonlinear manner, some DC pulse compression occurs. This could create a range error. The RF saturation characteristics of a TWT are not the same across the frequency band of the tube. This could also create a range error. An inductance in the grid lead can result in a triode-type oscillation, which might take months to resolve. Long pulses can result in ion-oscillations. Extra tube processing could alleviate this effect. Other problem areas could be listed. The important point, however, is that the L-3 ETI technical staff, which has designed and built more TWTs for radar than any other group in the world, can help to anticipate difficulties before they occur, shortening the design cycle and increasing the chances for ultimate success. Data on a typical helix TWT for radar applications are shown in Figure J

103 Figure J-1 L-3 ETI Model 18703H TWT output power versus frequency intercepting grid. 104

104 APPENDIX K NOTES ON TWTs FOR ECM AND MISSILE APPLICATIONS

105 Electronic Countermeasures Electronic warfare has been defined as a military action to take advantage of the enemy s use of the electromagnetic spectrum or deny its use to him. It is usually categorized into: ESM (Electronic Support Measures) Actions taken to search for, intercept, locate and identify enemy emitters. ECM (Electronic Counter Measures) Deliberate jamming or deception of an enemy emitter or receiver. ECCM (Electronic Counter Counter Measures) Action to ensure effective use of our equipment despite enemy jamming. The earliest use of RF jamming was in World War I by the German Navy. However, these primitive tactics were not really developed until the military began using radar just prior to World War II. The development of radar was rapidly followed by the introduction of ECM techniques to deceive and jam them. In turn, the evolution of new radars has been partially the result of a continual need to stay ahead of any new countermeasure tactics which might compromise the radar s effectiveness. The trend in search radar, for example, has historically been toward much higher powers and techniques that will increase target visibility, even while being jammed. A good anti-jamming radar necessarily must be able to shift frequency over a wide bandwidth quickly to avoid the jammer s source frequency. ECM trends have also been toward wide bandwidth system capabilities. The jammer on the target may be designed to amplify wideband noise, or to deceptively retransmit the hostile radar pulse to offset the radar s ability to determine the target s position. Because wide frequency bandwidths are essential to the employment of such ECM tactics, an amplifying device capable of broad operating ranges with sufficient output power and efficiency has been needed. The TWT has proven to be ideally suited for this task. Unfortunately, however, airborne ECM is usually not designed as an integral part of the airframe, but rather adapted internally or mounted in pods externally, depending on the military service tactics. As a consequence, TWTs are not standardized with respect to interface parameters and require close liaison between the TWT supplier and the ECM system engineer. Pulsed TWTs L-3 Communications Electron Technologies, Inc. (ETI) has been a leading supplier of broadband, kilowatt-level, helix TWTs for many ECM systems. Proven production capability has been established with tubes in all the major frequency bands. Currently, pulsed kilowatt TWTs cover the ECM spectrum up to 18 GHz. The tubes feature a rugged metal-ceramic construction suitable for airborne or missile environments, and typically use a single-gridded electron gun, PPMfocusing and coaxial couplers. 106

106 L-3 ETI broadband helix TWTs have proven to be reliable building blocks for many ECM applications. Some typical examples of helix TWTs can be seen in Figures K-1 and K-2. Figure K-1 L-3 ETI Model 8790H helix TWT typical performance. Figure K-2 L-3 ETI Model 18714H pulsed helix TWT typical performance. Parallel Pulsed TWTs Figure K-3 is a schematic of a technique to combine several conventional one kilowatt pulsed TWTs to attain higher pulsed power output over a broad bandwidth. In such a combiner system, accurate phase tracking for all components over the frequency range is a critical requirement for satisfactory performance. The input dividers are 3 db hybrids while the output combiners (hybrid or magic T ) have to be capable of handling higher peak and average power levels over large bandwidths. 107

107 Figure K-3 When combining several kilowatt-pulsed TWTs to attain higher output, accurate phase tracking is critical. Phase compensation is required for each tube pair to ensure the correct phase relationship in each combiner. Each of the tubes also requires an amplitude and phase equalizer so phase tracking can be kept within +20 degrees over an octave bandwidth, provided that their grid and cathode voltages are also individually optimized. With all these provisions, the combiner losses are still in the order of 1.5 db over the band. Four tubes at the 2.5 kw level would, therefore, provide a combined peak output power of 7.0 kw. In some applications, coupled-cavity TWTs are needed to provide the RF performance and several L-3 ETI types have been utilized. One example is rated at 10 kilowatts peak, 2 percent duty over the band GHz, and another provides over 1500 watts CW over the same bandwidth, as shown in Figure K-4. Continuous-wave TWTs In addition to these pulsed types, high-power CW TWTs are used in broadband airborne jammer applications (see Figure K-5). An integral solenoid provides a compact, rugged device for tactical environments. A Shift to CW Radar Recent changes in airborne tactics have shifted interest to the advantages of CW radar over the earlier pushed techniques. Pulsed radars transmit short bursts of energy and then turn on the receiver between bursts. CW radar, however, uses antenna isolation and frequency resolution to detect the return signal. 108

108 Figure K-4 Power output with constant drive, coupled-cavity TWT. Figure K-5 Output power with constant RF drive, coupled-cavity TWT. 109

109 The countermeasures to these radars are: Deception jamming which uses transmitted signals to confuse the radar s data processing system. Noise jamming which uses high-power density RF to obscure the radar return. Each approach has strong advantages in varied tactical situations, but they are most effective when used together. One way to achieve this objective is to use two TWTs, operating in parallel from a single power supply. One is a pulsed TWT for deceptive schemes; the other, a high power CW TWT used for barrage jamming modes. This approach is based on the available single-mode TWTs previously described, and recent improvements have been made in both pulsed and CW single-mode TWTs in regard to higher power output, extended bandwidth and improved duty cycle. However, new system concepts utilize a single multi-mode TWT to provide both the pulsed or CW output. This approach has obvious simplicity as well as inherent savings in size and weight. Multi-Mode TWTs Key design features of these new multi-mode devices include a shadow gridded tetrode electron gun to provide the varied beam operating parameters. In addition, nondispersive circuit techniques are used to achieve wider bandwidths. Unique attenuators and velocity step tapering are also used to inhibit backward-wave oscillations and enhance tube stability. Integral barrel PPM-focusing provides excellent RF performances as well as a reliable rugged device which meets the stringent requirements of modern airborne environments. Figure K-6 shows typical performance for a multi-mode TWT. Flexibility in providing for intermediate modes allows the designer to adapt the system to a wide range of output power requirements. Specific applications for such multimode performance should be discussed with L-3 ETI at the time of system design to benefit from the latest developments in this area. Missile TWTs embody the same features as radar TWTs except that they require high efficiency and usually require rapid warm-up time. The high efficiency is achieved by using multi-stage depressed collectors and fast warmup is achieved by using a special cathode and heater design having low thermal mass. 110

110 Figure K-6 Multi-mode performance of a L-3 ETI multi-mode TWT offers the systems designer flexibility in output power requirements. 111

111 APPENDIX L NOTES ON TWTs FOR SPACE APPLICATIONS

112 The TWT in Space Just how long can a TWT/TWTA be expected to perform in a space application? Both ground tests and in-orbit data demonstrate that lifetimes in excess of 15 years are conservative estimates for L-3 ETI space hardware. Shown in Figure L-1 is a typical TWTA for space communications. These highly sophisticated microwave amplifiers provide high efficiency RF power with the high gain that is required for downlink transmission. Figure L-1 C-band space TWTA. L-3 ETI space TWT experience began in 1963 with the launch of the Syncom Satellite series. These Boeing/NASA satellites were the first attempt at placing satellites in synchronous orbit. Since the successful flight of Syncom, a continuing line of L-3 ETI space TWTs has achieved a total in-orbit operating time approaching 250,000,000 hours for M-cathode TWTs alone. This is in addition to over 150,000,000 in-orbit operating hours from earlier TWTs with oxide cathodes and ground life test experience of over 30,000,000 hours. The in-orbit TWT experience has been accumulated on hundreds of major space programs, including Apollo, Mariner, Surveyor, Pioneer, Intelsat series, Skylab, Westar, TDRSS, DSCS-II, DSCS-III, Anik series, Palapa, ATS, Aussat, Galaxy series, Telsat, JCS, JCSAT series Telecom, Arabsat, Spacenet, SBS, Astra series, Eutelsat, Marisat, Moreles, Marcopolo, G-Star, Superbird series, Skynet series, Panamsat series, INSAT series, Apstar series, KoreaSat series, Solidaridad, Morelos series, ICO, and Sirius Satellite Radio. In the case of TWTAs, L-3 ETI experience began in the mid-60 s with the Lunar Orbiter program. Since that time, L-3 ETI TWTAs have been developed and delivered for military, NASA, commercial and international space applications. The RF-power capabilities of these units range from 200 mw to 200 W of CW operation. 114

113 The Space TWT The L-3 ETI design philosophy for space TWTs and EPCs relies on conservative designs with strong heritage, and extensive analysis and testing to provide hardware that is rugged, reliable, lightweight, and with a long service life. L-3 ETI has developed and implemented the most up-to-date state-of-the-art technology in the areas of computer modeling and simulation, metal-ceramics assembly and processing and mechanical and electrical design. This combination of philosophy and technology has produced hardware that has consistently proven itself in on-orbit operation. L-3 ETI has on-going research and development programs to improve RF performance, efficiency, linearity and reliability and to reduce size and mass. Keys to a Long-Life TWT The life-determining design feature of a TWTA is the cathode of the tube. As discussed in the appendix B on electron guns and cathodes, the M-type dispenser cathode is the L-3 ETI standard for high reliability space tubes. The physics of dispenser cathode aging have been exhaustively studied at L-3 ETI and a very well-accepted cathode life model has been developed, validated and published. TWT life predictions are made using this model. There are basically three factors which determine the potential life of an M-type dispenser cathode: The temperature at which the cathode is operated. The initial thickness of the M-coating on the cathode surface. The cathode current loading for which the electron gun is designed. These factors along with other data are incorporated into the L-3 ETI cathode life model and the predicted life of the TWT determined. Life expectancy for space TWTs with dispenser cathodes is shown in Figure L-3. The reader is also referred to the discussion of cathode and electron guns in the earlier Appendix B. Electron gun optics are selected for conservative space charge, limited emission density, and low cathode temperature at a perveance, voltage, and beam size appropriate to the tube design. The electron gun designs in the L-3 ETI space TWTs use a control voltage on an isolated anode to achieve a constant cathode current throughout the operating life. This feature maintains the optimum beam focusing and RF interaction that were established at beginning of life. Other factors which must be taken into account to insure long life are the heater which heats the cathode, the metal-ceramic seals which must maintain a near-perfect vacuum, and the focusing structure which must assure maximum beam transmission. 115

114 Figure L-2 Life prediction for state-ofthe-art oxide cathodes operating in space TWTs. Figure L-3 Life prediction for state-ofthe-art dispenser cathodes operating in space TWTs. Cathode-Knee Temperature An idealized curve of cathode current versus cathode temperature is shown in Figure L-4. The region to the left of the knee is known as the temperature-limited-emission region. The region to the right, the area where tubes are normally operated, is the space-charge-limited-emission region. Typically, in long-life TWTs, some finite margin should exist between operating temperature and knee temperature. Another characteristic of long-life TWTs is that the cathode-knee temperature is relatively low and is Figure L-4 The region to the left of the knee is the temperature-limitedemission region, to the right is the space-charge-limited emission region. stable as a function of time after initial processing. Two methods are used for measuring this parameter, cathode activity test and the time-to-knee test. L-3 ETI TWTs with dispenser cathodes are tested by plotting cathode current as a function of heater power and relating heater power to cathode temperature. Long-life Heaters The heater, which is the hottest element in the TWT, must provide the necessary energy to maintain the correct cathode temperature. For this reason the selection of reliable high-temperature materials and the limiting of the maximum heater temperature through optimum thermal design are necessary factors in obtaining a reliable, long-life heater design. 116

115 It should also be noted that during turn-on and turn-off cycles, the heater must go through a change in mechanical dimensions. The design must provide for this expansion and contraction without over-stressing the heater wire or the insulation between the heater and the cathode. The Vacuum Envelope The required vacuum environment for the cathode can be degraded in several ways, all of which tend to increase the pressure and impair cathode operation. Among these are a leak in the vacuum envelope, internal outgassing due to overheating, or arcing with attendant poisoning and/or ion bombardment of the cathode. The incidence of vacuum envelope leaks is kept extremely low by the use of good design, proven reliable metal-ceramic joining techniques, very high quality materials and monitoring of the vacuum throughout the assembly and testing of the TWT. Careful analysis, together with thorough thermal design and testing, leads to conservatively low operating temperatures within the TWT. Under these circumstances, the TWT bakeout temperature is never approached (except at the cathode) in normal specified tube operation. Hence, the cleaning and outgassing function of bakeout is not sacrificed. A getter provides internal pumping capacity over extended life. Screening and storage tests are performed during the manufacturing cycle to eliminate any possible leaks. Beam Focusing Space TWTs make exclusive use of PPM focusing that is described in Appendix C. The magnets are almost exclusively fabricated from high temperature stabilized samarium-cobalt materials. These materials have been chosen so that the desired magnetic field is achieved with the minimum mass. During TWT processing, the magnetic field is generally adjusted for best focusing by placing small magnetic shunts on the outside diameter of the magnetic stack. Critical Design Parameters The primary design parameters which differ according to the application requirements are frequency, power level, gain, bandwidth and life. Secondary considerations which must be taken into account to achieve the best design for a specific application are efficiency, linearity, size and weight. TWT size is determined by the physical laws which determine the frequency response, linearity, and efficiency. The weight depends on the size, materials used, and the structural techniques employed for the necessary strength to survive the thermal stresses, pyrotechnic shock and vibrations encountered during launch and operations. The Space EPC L-3 ETI has developed and manufactured EPCs for TWTs capable of 200 mw to 400 W of CW RF power. These units are designed to supply cathode voltages from 1.2 to 14 kv and will operate from a regulated or unregulated bus. 117

116 In addition, EPCs have been developed for pulsed applications with cathode voltages in excess of 18 kv. The EPC converts the regulated or unregulated spacecraft bus voltage to dc voltages at the proper levels and with the necessary regulation to operate a given TWT. A simplified block diagram of a typical EPC is shown in Figure L-5. Figure L-5 The heart of the EPC is the new approach to voltage conversion and regulation knows as the Regulating Converter. Also shown is the optional power supply that is used when the TWT is linearized. The heart of the EPC is the system used for voltage conversion and regulation, L-3 ETI has been granted patents on circuit configurations that perform these functions with high efficiency and low parts counts. The converter utilizes a circuit configuration that achieves such desirable features as higher efficiency, a single circuit for regulation and conversion, minimized output filter requirements, and simplified control system applications. The size and weight of the EPC is dependent on the thermal interface, RFI and telemetry requirement, spacecraft power bus, allowable ripple current that the TWTA can inject on the power bus, residual AM and PM noise, and shock and vibration levels to be experience in the launch environment. L-3 ETI has on-going programs exploring EPC efficiency improvements, longer life and reliability electronics, improved packaging techniques, the effects of radiation, multi-stage collector operation, and system interface as it relates to both thermal factors and the power bus of the spacecraft. 118