US3278865A - Device using multipactor discharge - Google Patents

Description

Oct. 11, 1966 M. P. FORRER DEVICE USING MULTIPACTOR DISCHARGE 2 Sheets-Sheet 1 Filed May 31. 1965 E m T E L C W W D H s m H INVENTOR.

BY Max P Forrer Attorneys Oct. 11, 1966 M. P. FORRER 3,278,365

DEVICE USING MULTIPACTOR DISCHARGE Filed May 51, 1963 2 Sheets-Sheet 2 Fig.3

k o O n a Q 5 2 u U E 5 LL Q E 2 INVENTOR.

Max P Forrer Attorneys United States Patent 3,278,865 DEVICE USING MULTIPACTOR DISCHARGE Max P. Forrer, Palo Alto, Calif, assignor to Kane Engineering Laboratories, Palo Alto, Calif., a corporation of California Filed May 31, 1963, Ser. No. 284,575 15 Claims. (Cl. 33313) This invention relates to a device using multipactor discharge and more particularly to a device using multipactor discharge for high power nanosecond switching.

The secondary electron resonance phenomenon, commonly known as multipactor, is well known. The term multipactor is derived from the words multiple electron impact. A multipactor discharge consists of electrons in high vacuum which are driven back and forth between two electrodes by an RF electric field. This electron motion is in synchronism with the applied electric field. In the simplest case, each transit occurs in onehalf of the RF period, but higher order modes are also possible where one transit requires any odd number of half RF periods. Upon impact of the electrons with an electrode, secondary electrons are emitted to reconstitute the discharge. It is evident that such a discharge may sustain itself if the secondary emission ratio is larger than one or unity and if the frequency, voltage and gap spacing are such that synchronous motion is possible. The multipactor discharge, being dependent on the voltage, represents a voltage sensitive impedance. This characteristic has caused investigations to be conducted to utilize multipactor discharges for duplexing and switching. Recent investigations into such uses are disclosed in an article entitled Duplexing and Switching With Multipactor Discharges by M. P. Forrer and C. Milazzo in the Proceedings of the 'IRE, volume 50, No. 4, April 1962, pages 442-450; and in another article entitled Microwave High-Power Nanosecond Switch Using Multipactor Discharge by C. Milazzo, published in The Microwave Journal, March 1962 issue. In these articles, there is disclosed a multipactor switch cavity in which one electrode is D.-C. insulated from the remainder of the assembly by a coaxial low pass filter so that a D.-C. switching bias can be applied to this electrode to prevent a multipactor discharge from building up or which quenches a multipactor discharge which may already exist. By the application of the D.-C. bias to the stop electrode, it is possible to electronically control the multipactor discharge to, in effect, provide an electronic switch. The presence of the multipactor discharge in the cavity causes electromagnetic energy which is incident upon a surface forming the cavity to be refiected. In the absence of multipactor discharge, elec tromagnetic energy incident upon a surface forming the cavity is freely transmitted through the device. The device described in the articles named above consists of an input wave guide window and an output wave guide Window. An evacuated reentrant cavity containing a plane parallel multipactor gap and also containing the D.-C. isolation means hereinbefore described is provided. A tuning ring is also provided for adjusting the resonant frequency of the cavity. Although such devices have been made to operate, they have been subject to a number of disadvantages. One is that a very high D.-C. con trol voltage is required which must have a very fast rise time in order to obtain the rapid switching desired. The generation of such a high voltage with rapid rise time presents many problems which are very difficult, if not impossible, to solve. In addition, it has been found that the plane parallel gap provided in such a cavity has limited surface area for secondary emission and that such surfaces must have a large secondary emission ratio to sustain the required discharge current. There is, therefore, a need for a new and improved device utilizing multipactor which can be used for switching operations.

In general, it is an object of the present invention to provide a device utilizing multipactor discharge which is particularly adapted for high speed switching.

Another object of the invention is to provide a device of the above character in which relatively low control voltages can be utilized.

Another object of the invention is to provide a device of the above character which can operate with low current densities.

Another object of the invention is to provide a device of the above character in which small grazing or impact angles are utilized for the primary electrons to increase secondary emission yields.

Another object of the invention is to provide a device of the above character in which a crossed magnetic field is utilized to reduce materially the control or bias voltage required for discharge suppression.

Another object of the invention is to provide a device of the above character which has increased life properties.

Another object of the invention is to provide a device of the above character in which it is possible to sustain a multipactor discharge at high peak powers.

Another object of the invention it to provide a device of the above character which can also be used for duplexing.

Additional objects and features of the invention will appear from the following description in which the preferred embodiment is set forth in detail in conjunction with the accompanying drawings.

Referring to the drawings:

FIGURE 1 is a front elevational view, partially in cross-section, of a device using multipactor discharge incorporating my invention.

FIGURE 2 is a top plan view of the device shown in FIGURE 1.

FIGURE 3 is an end elevational view looking toward the left-hand end of the device shown in FIGURE 1 With certain parts broken away.

FIGURE 4 is a schematic diagram illustrating the trajectories of the electrons in the coaxial gap configuration utilized in the device shown in FIGURES 1, 2 and 3.

FIGURE 5 is a graph showing secondary emission yield for and 10 angles of impact.

In general, my device using multipactor discharge consists of wave guide means which forms a cavity. A pair of coaxial electrodes are disposed in the cavity. Means is provided for applying a radial electric field to the electrodes. Means is also provided for applying an axial magnetic field to the electrodes whereby the electric and magnetic fields cause the electrons passing from one of the electrodes to the other electrode to follow curved trajectories and to strike said other electrodes at angles substantially less than 90. Means is provided for applying a biasing voltage across the electrodes to prevent the electrons from striking said other electrode.

More in particular as shown in the drawings, my device using multipactor discharge consists of a wave guide body 11 having a rectangular configuration with the narrow side Walls having an inner dimension represented by the letter A, and the Wider side walls having an inner dimension represented by the letter B. A multipactor structure 12 is mounted in the body 11 and extends through the wide side walls of the wave guide body 11, as shown particularly in FIGURES 1 and 2. The multipactor structure 12 includes a coaxial cavity 13 as hereinafter described.

Impedance matching means is provided in the wave guide on each side of the cavity 13 to permit coupling ,able means such as brazing. with a rectangular opening 24 of the same dimensions as the inner dimensions of the wave guide body 11.

has a cylindrical inner surface.

into and out of the cavity without reflection and consists of a step transformer 16 of a type well known to those skilled in the art formed in the body on each side of the cavity 13. A window assembly 17 is mounted on each end of the wave guide body. Each window assembly consists of a circular flange 18 which is provided with a rec tangular opening 19. The flange 18 is mounted on the associated end of the wave guide body 11 and is secured V thereto by suitable means such as brazing. A window 21 ,of suitable material such as a ceramic is brazed within a ring 22 of suitable material such as stainless steel. One side-of the ring 22 is secured to the flange 18 by suitable ,means such as brazing. A rectangular flange 23 having an annular portion 23a is secured to the ring 22 by suit- The flange 23 is provided The multipactor structure 12 consists of an outer annular electrode 26 disposed within the cavity 13 and which A cyindrical inner electrode 27 is also disposed within the cavity 13 and coaxially within the annular outer electrode 26. The diameter of the inner electrode 27 is such that an annular .space 28 is provided between the outer cylindrical surface of the inner electrode 27 and the inner cylindrical surface of the outer annular electrode 26. This space 28 serves as an annular multipactor gap disposed within the cavity 13.

The outer annular electrode 26 is mounted in the lower .side of the wave guide body 11 which is provided with a large rectangular recess 31 and is secured to the'body -11 by suitable means such as brazing. The cylindrical inner .electrode 27 is secured to a pipe or tubular member 32 extending perpendicular to the longitudinal axis of the wave guide body 11.

Means is provided for securing the tubular member 32 and the electrode 27 mounted thereon upon the wave guide body 11 and which, at the same time, insulates the inner electrode 27 and electrically provides a short circuit to microwave energy at the input or output planes of the cavity. This means consists of an upper choke assembly 36 anda lower choke assembly 37. The upper choke consists of a cup-shaped choke section 38 which is secured to the tubular member 32 by suitable means .such as brazing. The lower surface of the choke section ,v 38 is parallel to the inner surface of the upper wide wall of the wave guide body and-has its open end extending upwardly. The choke section 38 also consists of an additional cup-shaped member 39 which is secured to the member 32 by suitable means such as brazing. A cylindrical ceramic member 41 has its lower end brazed to the cup-shaped member 39 and has its upper end brazed to .a metal cylindrical member 42. The metal member 42 is secured to the upper end of an outer case 43 by suitable means such as welding. The lower end of the outer case 43 is mounted in the upper wide wall of the Wave guide body 11 so that it is relatively close to the upper side of the cavity 13.

The lower choke assembly 37 is constructed in a similar manner and consists of a cup-shaped member 46 which is secured to a tubular member 32 and which has its upper surface defining the lower side of the cavity 13. There is an additional cup-shaped member 47 facing downwardly and mounted upon the tubular member 32. A cylindrical .ceramic member48 has its upper end bonded to the cup-shaped member 46 and has its lower end brazed or bonded to a metal cylindrical member 49. The member v49 is secured to an outer case 51 by suitable means such as brazing. The member 51 is secured to a ring 52 by suitable means such as brazing and the ring 52 is secured to the wave guide body 11 within the recess 31 by suitable means such as brazing.

The choke joints 36 and 37 hereinbefore described are substantially conventional and provide a short circuit over a very wide band of frequencies as is well known to those skilled in the art and, therefore, their operation will not be described in detail. In general, the choke joints are necessary to provide the necessary D.-C. insulation and RF shorts. It is the break in the cavity wall which permits D.-C. isolation of one of the electrodes. The break is on the diameter of the electrode which makes it the input plane of a coaxial filter. The highly attenuating reactive filter formed by the choke joints 36 and 37 is associated with the cavity design to prevent serious loss of power through the break in the cavity Walls. This filter presents a well defined equivalent short circuit I plane which does not move relative to the filter input drawings, an electromagnetic assembly 56 has been utilized. This electromagnetic assembly consists of a pair of coils 57 symmetrically disposed on opposite sides of the wave guide body 11 parallel to the narrow walls, as shown particularly in FIGURE 2. These coils 57 are wound upon cores 58 of magnetic material. Yokes 59 extend transversely across the top and bottom of the wave guide body 11 and are secured to the magnetic cores 58 by suitable means such as cap screws 61. The yokes 59, which are in the form of flat plates of suitable magnetic material, are mounted upon the outer choke casings 43 and 51 of the choke 36 and 37 so that the choke casings 43 and 51, which have a substantial thickness, together with the yokes 59 form a part of a -magnetic yoke assembly. The casings 43 and 51 serve as pole pieces and introduce an axial magnetic field into the cavity 13. As shown in the drawings, the upper and lower extremities of the choke casings extend to points just outside the cavity 13 so that fringing fields at the ends of the pole pieces enter into the cavity and provide an adequate magnetic field within the cavity 13.

As shown in FIGURE 1, the device is normally used with a suitable high voltage D.-C. source such as a source 69 indicated in block form in FIGURE 1 which has one terminal connected through a switch 70 to the tubular member 32 which is electrically connected to the central electrode 27 but insulated from the body 11. The other side of the high voltage D.-C. source 69 is connected to the member 42 which is electrically connected to the other electrode. Means (not shown) is also provided for energizing the electromagnets 57.

The electrodes 26 and 27 are formed of any suitable material having a secondary emission coefficient above unity and preferably substantially above unity such as silver magnesium.

- Operation and use of my device using multipactor discharge can now be briefly described in conjunction with the graph and diagram shown in FIGURES 4 and 5. Let it be assumed that it is desired to utilize the device asa switch and that the multipactor effect is taking place in the multipactor gap area 28. If it is assumed that no magnetic field is present between the electrodes 26 and 27, the electrons in the multipactor gap 28 will have a trajectory which is perpendicular to the surfaces of the electrodes as represented by the arrow 71 in FIGURE 4. The electrons will be accelerated toward the surface formed by the electrode 26 under the influence of the applied RF electric field. With the proper choice of gap spacing and RF voltage, the primary electron will arrive at a time just at the time when the field is reversing.

by secondary electrons generated each time the cloud strikes one of the surfaces and the electron transits are in synchronism with the applied voltage. For this processto be stable, the electrodes must have a secondary emission coeflicient greater than unity for the impact energy involved.

As is well known to those skilled in the art, if the surfaces have a high secondary emission yield, the discharge will be characterized by rapid exponential rise in time. For example, starting with N initial electrons, there will be N electrons after the first impact, where 5 is the secondary emission yield; that is, the number of emitted electrons for each electron striking the surface. The secondary emission yield is dependent upon several factors. One is the type of material utilized and the other is the incident angle of the electrons. In particular, the secondary emission yield is very much higher as the incident electron strikes the surface on a slanted angle as shown in FIGURE 5. It is for this reason that the coaxial arrangement for the electrodes 26 and 27 has been chosen. The application of the axial or crossed static magnetic field supplied by the electromagnets 57 causes the electrons passing from the outer surface of the electrode 27 to follow curved trajectories as indicated by the arrow 72. This magnetic field is called a cross magnetic field because its direction is crosswise to the radial electric field. By causing the electrons to follow a curved trajectory, the electrons strike the circumferential surface of the electrode 26 at a relatively small grazing angle as, for example, to greatly increase the secondary emission yield in comparison to the yield caused by perpendicular impact of the electron. I have found that this has made it possible to increase the secondary emission yield by a factor of 3-5 times. Curves 73 and 74 in FIGURE 5 show the approximate multipactor discharge for incident angles of approximately 90 and 10, respectively.

Because the multipactor gap area is relatively large, a larger secondary emission field is provided to make possible operation at higher discharge current densities with greater isolation and with the further effect to make possible switching at higher peak powers or, in other words, being able to sustain a multipactor discharge at higher peak powers.

As also shown by the curve 74 in FIGURE 5, the utilization of the crossed magnetic field to provide curved trajectories for the electrons has the additional advantage in that a higher 6 (secondary emission yield) i obtained at lower impact energies, and thus a multipactor discharge will start at a lower impact energy level and will not stop until a much higher impact energy level is reached. This is readily apparent from viewing the curve 74 and finding that it crosses the unity line (where multipactoring must cease) at points which are spaced much further apart than the points at which the curve 73 passes the unity line.

The utilization of curved trajectories or paths for the electrons has an additional advantage. It is known that electrodes utilized for such purposes do not have an unlimited life and that the secondary emission yield drops with use. By realizing a higher 8 by utilizing curved trajectories, an effective increase in useful life for the device is obtained because of the greater secondary emission obtained.

After a multipactor charge has commenced, the steady state or saturation current is reached when electron losses resulting from space charge debunching effects balance the gains resulting from high secondary yield. In the device shown in FIGURES l, 2 and 3 of the drawings, the multipactor discharge takes place in the multipactor gap region 28 of the resonant cavity 13 and thus, in effect, it takes place in a high impedance location which can be used for switching purposes. The strong discharge current caused by the multipactor discharge both loads down and detunes the cavity, causing substantial or all of the incident power to be reflected or, in other words,

prevents the transfer of power from one end of the device to the other.

When it is desired to permit power to pass through the device, the switch 70 is closed to apply the high voltage D.-C. to the electrodes 26 and 27 to terminate or quench the multipactor discharge and to permit the device to act as a passive transmission device. Operation with the crossed magnetic field hereinbefore described reatly reduces the control bias voltage required for discharge suppression. This can be readily understood by viewing FIGURE 4 when it is realized that the D.-C. voltage from the source 69 must only be great enough to overcome the perpendicular impact velocity component of the trajectory for the electron. Thus, as shown by the curve 76 in FIG- URE 4, it is only necessary that the biasvoltage be sufficient to cause the electron to miss the opposite electrode. With the arrangement shown, it can be seen that the smaller the grazing angle, the smaller the bias voltage required; and conversely, the greater the grazing angle, the greater the bias voltage required. In practice, it is desirable to adjust the magnet field to provide the desired discharge current density while, at the same time, retaining a relatively low control voltage requirement. In any event, it is readily possible to operate such devices utilizing control voltages ranging from 200 to 1000 volts. The use of such lower control voltages has tremendous ad vantages particularly when it is desired to switch at nanosecond or higher speeds.

From the foregoing, it can be seen that the application of the DC. voltage to the multipactor gap, region 28 serves to quench the multipactor discharge by destroying the synchronous transit-time conditions. For this reason, the presence or absence of the quenching or biasing voltage as determined by the operation of the switch 70 determines the two operating states of the device; that is, as either a passive transmission device or a reflecting device.

With the electrons following perpendicular trajectories, it can be seen that the biasing control voltage required would be at least several times greater than that required for electrons following curved trajectories such as that indicated by the arrow 72.

It is apparent from the foregoing that I have provided a new and improved device using multipactor discharge which is particularly adaptable for switching applications. By the utilization of a crossed magnetic field, it is possible to obtain reasonable discharge current densities while, at the same time, greatly reducing the control bias voltage required for discharge suppression. In addition, the device is constructed in such a manner so that it can be readily cooled by suitable fluid means such as water. Thus, for example, as shown, water can be passed through the tubular member 32 to provide cooling for the central electrode 27. The outer electrode 26 is physically in contact with sufficient mass so that additional cooling means should not be required. The step transformers 16 are preferably formed of a good heat conducting material such as copper so as to readily conduct away the heat generated in the electrode 26.

Although a device using multipactor discharge has been disclosed primarily for use as a switching device utilizing a control voltage, the construction of the device is also such that it can be used as a duplexer. As is well known to those skilled in the art, a duplexer does not require the use of a control voltage. The duplexer is self-actuated by the high power transmitted pulse and becomes full transmitting as soon as the high power pulse disappears.

I claim:

1. In a device using multipactor discharge, wave guide means, a pair of coaxial electrodes mounted in the wave guide means and forming an annular multipactor space therebetween in electrical communication with the wave guide means, each of the electrodes being formed of a material having a secondary emission coefiicient greater than unity, means applying a radial RF alternating electric field in the multipactor space to cause electrons to pass alternately from one electrode and to strike the other electrode to cause a multipactor discharge in the cavity, means applying an axial magnetic field within the multipastor space, said electric and magnetic fields causing the electrons passing between said electrodes to follow curved trajectories and to strike the electrodes at angles substantially less than 90 to thereby greatly increase the secondary emission yield from the electrodes.

2. A device as in claim 1 together with means for applying a biasing voltage to said electrodes to prevent electrons from travelling between the electrodes to quench an existing multipactor discharge or to prevent the formation of a multipactor discharge.

3. A device as in claim 1 wherein the axis of the electrodes is disposed at right angles to the longitudinal axis of the wave guide means.

4. A device as in claim 1 wherein one of the electrodes is in the form of a hollow annular member and wherein the other of the electrodes is in the form of an additional annular member having a diameter less than the diameter of the first annular member.

5. A device as in claim 4 together with means for D.-C. isolating the additional annular member from the wave guide means, said last'named means including choke means preventing serious loss of A.-C. power through the break in the cavity walls, and presenting a well defined equivalent short circuit plane.

6. A device as in claim 4 together with a tubular member extending through said additional annular member and means for supplying a cooling fluid to said tubular member.

7. A device as in claim 1 together with choke means disposed adjacent opposite ends of the multipactor space, said choke means forming equivalent spaced parallel short circuit planes adjacent opposite ends of the multipactor space, means for mounting each of said choke means upon the wave guide means to insulate the choke means from the wave guide means, the mounting and insulating means including a pair of cylindrical members having ends disposed adjacent opposite ends of said multipactor space and wherein the means for applying an axial electric field to the electrodes includes at least one electromagnet and means for connecting magnetic flux lines from the electromagnet to the cylindrical members so that the cylindrical members serve as pole pieces.

8. In a device using multipactor discharge, a rectangular wave guide having parallel wide walls and parallel narrow walls, said wave guide being formed with a hole extending through parallel walls of the wave guide, impedance matching means disposed in each end of the wave guide, a multipactor discharge structure disposed in said hole and mounted upon the wave guide, said multipactor discharge structure including an outer hollow annular electrode mounted in said hole, an inner annular electrode, each of the electrodes being formed of a material having a secondary emission coefficient greater than unity, means for insulating the inner electrode from the outer electrode, said electrodes forming an annular multipactor discharge gap between the same in the wave guide, means applying a radial electric field in the multipactor discharge gap to cause electrons to pass alternately from one of the electrodes and to strike the other of the electrodes to thereby cause a multipactor discharge in the multipactor discharge gap and thereby prevent transfer of power through the wave guide, and means for applying an axial magnetic field in the multipactor discharge gap to cause the electrons passing between the electrodes to follow curved trajectories and to strike the electrodes at angles substantially less than and means for applying a D.-C. control voltage to said electrodes to prevent said electrons from passing between said electrodes to quench any existing multipactor discharge or to prevent the formation of a multipactor discharge in the multipactor gap.

9. A device as in claim 8 wherein the means for insulating the inner cylindrical electrode from the outer cylindrical electrode includes a pair of choke joints disposed on opposite sides of the multipactor discharge gap.

10. A device as in claim 9 wherein the choke joints include cylindrical members formed of magnetic material and having their extremities adjacent the gap and wherein the means for applying the axial magnetic field includes means for applying magnetic flux lines to the cylindrical members so that a fringing magnetic field is formed in the gap.

11. A device as in claim 8 wherein the outer cylindrical electrode is mounted in wall of the rectangular wave guide and wherein the multipactor discharge structure includes a tubular member extending through said inner electrode, and means for applying cooling fluid to said tubular member to cool the inner electrode.

12. In a device using multipactor discharge, a pair of coaxial electrodes defining an annular multipactor space therebetween, each of the electrodes having a secondary emission coefiicient greater than unity, wave guide means for supplying the power to the multipactor space and for removing power from the multipactor space, means supplying a radial RF alternating electric field in the multipactor space so that the electrons are alternately attracted by one electrode and then the other electrode, and means supplying an axial magnetic field in the multipactor gap, the electric and magnetic fields in combination causing the electrons passing between the electrodes to follow curved trajectories, the spacing between the electrodes being such so that the electrons arrive at the electrodes at the time that the electric field is reversing, the electrons following the curved trajectories striking the electrode at an angle substantially less than 90 to increase the secondary emission yield from the electrode.

13. A device as in claim 12 together with means for applying a high D.-C. biasing voltage to said electrodes to prevent the electrons from travelling from one electrode to another to thereby prevent the formation of a multipactor discharge or to quench an existing multipacto'r discharge.

14. A device as in claim 12 wherein said wave guide means is in the form of a rectangular wave guide having parallel wide and parallel narrow walls and wherein the axis of the coaxial electrodes is perpendicular to the longitudinal axis of the wave guide means.

15. A device as in claim 14 wherein said coaxial electrodes are in the form of a hollow cylindrical member mounted in parallel walls of the wave guide means and a circular member mounted coaxially Within the cylindrical member, means mounted on the wave guide means for supporting the circular member within the cylindrical member and also serving to insulate the circular member from the cylindrical member, and means for supplying a cooling fluid to the circular electrode.

References Cited by the Examiner UNITED STATES PATENTS 2,856,518 10/1958 Lerbs 333l3 2,925,528 2/1960 Beavis 315-39 DAVID J. GALVIN, Primary Examiner.

Claims (1)

1. IN A DEVICE USING MULTIPACTOR DISCHARGE, WAVE GUIDE MEANS, A PAIR OF COAXIAL ELECTRODES MOUNTED IN THE WAVE GUIDE MEANS AND FORMING AN ANNULAR MULTIPACATOR SPACE THEREBETWEEN IN ELECTRICAL COMMUNICATION WITH THE WAVE GUIDE MEANS, EACH OF THE ELECTRODES BEING FORMED OF A MATERIAL HAVING A SECONDARY EMISSION COEFFICIENT GREATER THAN UNITY, MEANS APPLYING A RADIAL RF ALTERNATING ELECTRIC FIELD IN THE MULTIPACTOR SPACE TO CAUSE ELECTRONS TO PASS ALTERNATELY FROM ONE ELECTRODE AND TO STRIKE THE OTHER ELECTRODE TO CAUSE A MULTIPACTOR DISCHARGE IN THE CAVITY, MEANS APPLYING AN AXIAL MAGNETIC FIELD WITHIN THE MULTIPACTOR SPACE, SAID ELECTRIC AND MAGNETIC FIELDS CAUSING THE ELECTRONS PASSING BETWEEN SAID ELECTRODES TO FOLLOW CURVED TRAJECTORIES AND TO STRIKE THE ELECTRODES AT ANGLES SUBSTANTIALLY LESS THAN 90* TO THEREBY GREATLY INCREASE THE SECONDARY EMMISSION YIELD FROM THE ELECTRODES.

Images