US2663811A - Vapor-electric device - Google Patents

Description

Dec. 22, 1953 WITNESSES:

Fig.l.

J. L. BOYER VAPOR-ELECTRIC DEVICE Filed Dec. 20, 1950 Breakdown Voltage who 2600 3600 4600 5600 PD(P in microns ofHg,D in cm) Fig.2.

INVENTOR John LBdyer.

ATTORNEY Patented Dec. 22, 1953 VAPOR-ELECTRIC DEVICE John L. Boyer, Pittsburgh, Pa., assignor to Westinghouse Electric Corporation, East Pittsburgh, Pa., a corporation of Pennsylvania Application December 20, 1950, Serial N 0. 201,784

23 Claims.

My invention relates to vapor-electric devices or tubes, and it has particular relation to small low-Voltage high-current low-arc-drop hotcathode arc-discharge devices using a vaporizabie discharge-metal selected from the alkalimetal subgroup consisting of cesium," rubidium and potassium. K 7

it is well known that the breakdown-voltage of an evacuated or low-pressure gap-device is a function of the pd product, where p is the pressure and d the distance between electrodes. At extremely low values of this product, the breakdown-voltage is very high, u into the thousands of volts, but as this product increases, the breakdown-voltage rapidlyfalls until it reaches a minimum breakdown-voltage, after which any further increase in the pdproduct will cause the breakdown-voltage to rise again, slowly at first, and later on more rapidly. The minimum breakdown-voltage is of the order of some 200 to 500 volts, depending upon the gaseous medium in the space'separating the electrodes, and the pressure-distance value pd at which this minimum breakdown-voltage occurs is also variable,

depending upon' the gaseous material. When the gaseous 'medium consists substantially entirely of the vapor of a vaporizable reconstructing cathode-metal or discharge-metal, the pressure :0, indie pd product, will be the vapor-pressure of the metal at the coolest internal temperature of the tube, during the operation of the device, or the condensatioil-temperature of the discharge metal. The curve which plots the vapor-pressure of different metals, as a function of temperature, is a characteristic curve of the metal, which dififers according to the metal.

Before the successful introduction of alkalilnetal rectifiers, all vaporizable-cathode rectilfiers have had such a high arc-drop that their efficiency was quite low at voltages under about 600 volts direct current. Among the many distinguishing characteristics of alkali-metal rectifiers are their extremely low arc-drop, which improves their eificiency, and the unusually high temperatures at which they have a vapor-pressure in the useful rectifier-range, which makes it possible to use ambient-atmospheric cooling, without the necessity for heavy water-jackets.

Heretofore, alkali-metal rectifiers have been designed for operation high up on the low-pressure side of the breakdown-pd curve, in order to operate at the customary vapor-arc rectifiervoltages of the order of 600 volts or higher. This has necessitated the use of a ,pd product which is smaller than the product at which the minimum breakdown-voltage occurs. The permissible rated output-voltage of the rectifier is only about one-half (or less) of the breakdown-voltage under the operating'conditions which prevail in the rectifier. To make the pd product small, either the electrode-separation distance d should be small, or the vapor-pressure 10 should be small, or both. It has been'customary heretofore, to design alkali-metal tubes which must be operated with inverse voltages of 600 to 1000 volts, so as to have the minimum practicable electrode-spacing, which is about A inch, and this has necessitated the use of vapor-pressures of the order-of '70 microns of mercury.

My new low-voltage alkali-metal rectifiertube is designed to operate with inverse voltages which are in the range of 100 to 200 volts, or sometimes 250 volts, or lower, which makes it possible to let the minimum internal temperature or condensation-temperature of the tube run up to temperatures which will give a pal product higher than the value corresponding to the minimum breakdown-voltage of the device.

A characteristic feature of my new low-voltage rectifier-tube is, therefore, the high pd product at which it operates, which has a threefold advantage: it relieves the designer of holding his design to the minimum possible spacing between the anode and the cathode, which removes a serious design-limitation; it permits the use of higher condensation-temperatures,- corresponding to higher vapor-pressures in the tube, thus making it easier to cool the tube by direct radiation to the atmosphere; and it makes it possible to operate the tube with higher currentdensities on the cathode, thus obtaining a higher current-rating for a given size of tube.

This last-mentioned high-current advantage results from the peculiar nature of the electronemission characteristics of alkali-metal tubes, in which the electron-emission is produced by a. monatomic layer of the discharge-metal which forms on the surface of a heated cathode which is made of another metal (usually nickel) having a work-function which is higher than the ionization potential of the alkali-metal which is being used. This is believed to be the fundamental reason why alkali-metal tubes have phenomenally low arc-drops, ofthe order of. 4 volts, as compared, 'for example with a mercury-vapor rectifier-tube, in which the arc-drop is of the order of 10.4. volts.

However, in order to maintain this monatomic layerof the discharge-metalon the cathode, it is necessary to have an adequate number of discharge-metal atoms which strike the heated nickel cathode and maintain this monatomic surface-layer thereon. In my present low-voltage tube, which operates at a high pd product, Ican use a higher vapor-pressure of the discharge-metal than has heretofore been the practice, and this higher vapor-pressure makes a larger number of discharge-metal atoms available, in. the vapor which bathesthe" surface of the nickel cathode, thus replenishing the monatomic discharge-metal layer on the cathode at a high rate, and making it possible to operatethe tube with a higher current-density onthe cathode, thus increasing the output which is obtainable from any given size of tube.-

I prefer to operate my new tube, therefore, at a condensation-temperature which is high enough to produce a discharge-metal vapor-- pressure of at least 1,000 microns of mercury,

as distinguished from previously used pressures of about 70 microns. At this high vapor-pressure, it is possible tooperate the tube with a high emissivity of the active emitting-portion of. the cathode, at a rate which. is above about 5 amperes per square inch, as determined by the maximum current-output of the device. And this is done without increasing the operating-temperature of the hot cathode beyond the conservative value of. about 750 0., which has prevailed in many of the best alkali-metal tube-designs prior to my present invention.

My new low-voltage high-current vapor-electric device is an improvement over the alkalimetal vapor-electric device which was described and claimed in a copending application of August P. Colaiaco and myself, Serial No. 144,354,

filed February 15, 1950. In this copending application, it was pointed; out that; the three preferred alkali-metals, cesium, rubidium and potassium, which are usable as the: dischargemetal, forms a more or less distinctive class by themselves, which may be described as the alkali metals having four, five and six. shells in their atomic structure, or the stable alkali. metals having more than three shells. The entire group of alkali metals consists of six elements, of which the first two and the last are readily distinguishable from the other three, with which my invention is particularly concerned.

The two lightest alkali metals, lithium (Li) and sodium. (Na) are separated, in some periodic tables, from the heavier light metals. of the alkali-metal group (IA), as being distinctive because of their electron-grouping. The physical and chemical characteristics of these two lightest alkali-metals are also distinctively different from the group comprising potassium, rubidium and cesium. Sodium is too active, chemically. Lithium has a vapor-pressure which is much too low for my purpose, as this low vapor-pressure requires too high a temperature to obtain a practically usable vapor-pressure which is high enough to give a sufiiciently high current-density to be practical for my purpose. 1 The sixth or heaviest alkali metal, No. 87 in the periodic table, was formerly called virginium, but has now been proved to be an element which is called francium (Fr), an unstable atomicpile product which is very radio-active, and which has an extremely short half-life of only a few minutes, so that it is unsuitable for my purpose.

In my present invention, I really prefer cesium or rubidium, for the discharge-metal, although my principal experience has been with cesium;

I believe that potassium is usable, at least in some possible tube-applications, but it requires a rather high condensation-temperature in order to produce a sufliciently high ionization-rate from the surface of the cathode, and this high operating-temperature increases the chemical activity of the discharge-metal, thus making it less desirable from that standpoint.

My invention consists in the apparatus, combinations, systems, structures, parts, and methods of design and operation, hereinafter described and claimed, and illustrated in the ac companyin-g drawing, wherein Figure 1 is a simplified sectional elevation of an illustrative type of vapor-electric device embodying my invention; and

Fig. 2 is a curve of the breakdownvoltage, plotted against the pd product of pressure 1 times distance or electrode-spacing d, for cesium.

The illustrated tube has an evacuated enclosure-means including a metal anode-portion 4,. which is commonly made of iron or steel, and which is shown as being in the shape of a cuplike container, the side walls of which are provided with radially extending cooling-fins 5. The top of the enclosure-means is the top of a cathode-structure B, which is insulated from the anode-portion 4 by an insulator-to-metal sealing-means l. The tube contains a small quantity, which may be a few drops, or less than a spoon ful, of a dischargemetal which is selected from the group comprising cesium, rubidium and potassium, as above explained.

The active portion of the cathode-structure 6 comprises a downwardly extending re-entrant cathode-tube H), which. is preferably made of nickel. The active electitonremitting portion of this cathode-tube; I0 is. really the bottom portion thereof, which is. disposed within the anodestructure 4, and which is provided with a large number of fins H, made. of the same cathodematerial (preferably nickel), for increasing the effective surface-area. of. the electron-emitting part of the cathode. A suitable heating-means is provided, in the form. of: a heater l2, which is inserted down into the lower or active portion of the cathode-tube. ID, for heating the active emitting-portion oi the cathode, this heater being energized by means of a terminal-lead l3 which extends through. a suitable insulating seal l4 in the to of the cathode-structure 6.

My present invention resides in the design of the external cooling-means, such as the heatradiating fins 5 or other means for cooling the tube by direct radiation to the ambient atmos phere which is at room-temperature. In the illustrated tube, the minimum internal temperature of the tube is obtained on a part of the anode-portion 4 of the. container. According to one aspect of my present invention, (at whatever point this minimum internal tube-temperature is maintained, during the operation of the tube), the cooling-means which maintains this coolest internal surface of the device must have such a cooling-rate, for the particular tube in question, that the coolest-surface or temperature-range within the tube is at least as high as the temperature at which the discharge-metal has a vapor-pressure of about 1000 microns of mercury.

According to my best present information, this thousand-micron vapor-pressure is obtained at about 278 C. for cesium, 297 C. for rubidium, and 344 C. for potassium. According to this aspect of my invention, therefore, the coolestsurface temperature of the tube should perhaps be at least as high as 270 C. for cesium, 290 C. for rubidium and 330 C. for potassium. The maximum condensation-temperature which is permitted is not so critical, and may range as high as a temperature which would produce a vaporpressure of about 4000 microns of mercury, more or less, which would be perhaps about 350 0., when cesium is the discharge-metal, for example.

The condensation-temperature may sometimes be defined more conveniently in terms of the maximum current which is obtainable from the tube, which is some sort of measure of the rate of emissivity of the active cathode-portion of the tube, and this temperature-definition thus requires no temperature-measuring means for estimating the temperature of the internal condensation-surface of the tube. Theactive electron-emitting surface-area of the cathode is of course known, and the maximum current-output of the tube is easily measured, and the quotient, current divided by area, gives the emissivity of the active emitting-portion of the oathode, which, in accordance with one phase'of my invention, is at a rate of above about 5 amperes per square inch, as determined by this method, when the temperature of the active portion of the cathode-tube I 0 is maintained at about 750 C.

In my new tube-design, as previously intimated, the previous strict limitation as to the small internal electrode-spacing is considerably relieved, and thus, in Fig. l, I show an anode-cathode spacing d, between the cylindrical inside wallsurface of the anode 4 and the outer peripheries of the cathode-fins l I, which is about an inch (2.5 centimeters) or larger, depending upon the balance which is determined between a practically large mechanical spacing at to give a large pressure-distance product pd, and a shorter distance which would give a shorter arc, and hence a somewhat smaller voltage-loss in the device.

According to one important aspect of my invention, it is a characteristic feature of my lowvoltage high-current rectifier-tube that I use a high pressure-distance product pd, which is obtained by multiplying the anode-cathode spacing d by the vapor-pressure p of the dischargemetal which is being used. In accordance with this aspect of my invention, this pressure-distance product pd is materially higher than the pressure-distance product at which minimum breakdown occurs.

Fig. 2 shows the breakdown-pd curve for cesium. It has a minimum breakdown-voltage of about 260 volts as indicated at 16, at a pd product which is equal to about 1300, when the pressure p is measured in microns of mercury, and the distance or spacing at is measured in centimeters. At pd. products lower than this value, or say between fifty and four or five hundred, the breakdown curve is a very steep curve, running up vertically into something like 6,000 volts at a pd of 50. At pd values of the order of 1300, or more, the breakdown-curve is quite flat, with the breakdown-voltage rising to something like 500 volts at a pd value of 4500.

In accordance with my invention, I prefer to operate my tube at a pd value which is materially higher than the value corresponding to the minimum breakdown-voltage. When cesium is the discharge-metal which is being used, my pressuredistance product pd should preferably be at least as high as, say, 2000, which is to say that the cesium is operated at a condensation-temperaiii) 'ture corresponding to a'vapor-pressure p, in microns of mercury, which is at least as high as the anode-cathode spacing at in centimeters.

The characteristic voltage-pd curves for rubidium and potassium differ somewhat from the curve which is shown in Fig. 2. In the case of rubidium, my best information is that the vertical part of the curve lies very slightly to the left of the vertical part of the curve for cesium, while the minimum breakdown-voltage for rubidium is slightly higher than for cesium, and probably occurs at a slightly lower value of the pressuredistance product pd. When the discharge-metal is rubidium, I would again prefer to keep the pd product at least as high as 2000, when the pressure p and the distance d are measured in the previously described units.

In the case of potassium, my best information is that the vertical part of the breakdown-voltage curve falls a greater distance to the right of the vertical part of the cesium-curve, and is not as vertical as this part of the cesium-curve. The minimum breakdown-voltage for potassium apparently falls at a much higher pd value than in the case of cesium, and when potassium is the discharge-metal, I prefer to use a pd product which is at least as high as 10,000 when measured in the previously described units.

It should be recognized that the measurements on which these breakdown-pd curves are obtained involve considerable experimental difficulties and are subject to considerable change in accordance with the precise experimental con,- ditions under which the measurements are taken. Such curves should be interpreted, therefore, with the realization that there may be unusually large experimental errors in their precise determination, which is a reason for allowing a considerable margin of safety in designing the apparatus.

Since my new tube operates at a pressuredistance product pd which is larger than the value corresponding to the minimum breakdownvoltage of the tube, it follows that the tube operates at a final steady-state condensation-temperature which is higher than the temperature corresponding to the minimum breakdown-voltage point of the characteristic curve. The direct-current output-voltage of the tube is usually something like one-half of the voltage-stress which is imposed on the tube tending to produce breakdown in the non-conducting periods of the tube, and hence the maximum permissible direct current voltage-rating is something like one-half of the minimum breakdown-voltage, more or less, depending upon the transformer-connections and the wave-form. In this respect, rubidium has probably an advantage over cesium, in having a higher minimum breakdown-voltage, if the tube is .to be rated at the highest possible direct current output-voltage.

It will be observed, from the foregoing, that I have produced a new alkali-metal rectifier-tube, which is useful in the low-voltage range of the order of volts or lower, or with inverse voltages of'the order of 100 to 200 volts. My tube is characterized by having an operating-point corresponding to a pd product which is higher than the valueof the product which would give the minimum breakdown-voltage, this condition being produced by using an external coolest-point cooling-means which has the proper heat-transfer rate to permit'the tube to run relatively hot, so as to produce a vapor-pressure p high enough to bring up this pd product to the desired value.

ture, and hence the high vapor-pressure of my tube, I very materially increase the currentrating of the tube, or the emissivity of the cathode in terms of current per unit area, resulting in something like a fivefold increase in the currentrating of the tube, as compared to tubes operating at lower condensation-temperatures which produce a pd product in the initial steep-slope range of the breakdown-pd curve. As the condensation-temperature (and hence the vaporpressure) is increased so as to obtain increased current-ratings, the arc-drop within the tube increases, but at a much slower rate than the increase in the current-rating, and thus the increased current-rating is obtained at the expense of a slight loss in efiiciency. In general, the current-rating should not be increased beyond the value corresponding to a vapor-pressure of 4000 microns of mercury, and a slightly more efficient tube would be obtained if this vaporpressure were held to about 1000 microns of mercury, as previously indicated.

Since my new high-pressure high-current tube operates at a temperature which is higher than has heretofore been used, my new tube is much more easily cooled, without elaborate external cooling-means, and hence my tube is well adapted for light-weight installations in which cooling by conduction to the ambient atmosphere is relied upon entirely.

By reason of the use of a low-arc-drop discharge-metal, selected from the group consisting of cesium, rubidium and potassium, I produce a tube which still has a high efliciency, even when used at the relatively low voltages for which my new tube is adapted, thus making my tubes competitive with mechanical rectifiers, in the lowvoltage field.

While I have described my invention in a single suggestive or illustrative form of embodiment, in a drawing which has been much simplified for clarity in the illustration of the essential novel features of my invention, I wish it to be understood that I am not at all limited to the particular type of tube-construction which is shown, as various changes may be made, within the knowledge or skill of the persons skilled in this art, and various refinements may be used, in the way of shields, temperature-responsive devices, and many other features which are known in tube-design. I desire, therefore, that the appended claims shall be accorded the broadest construction consistent with their-language.

I claim as my invention:

1. A low-voltage high-currentvapor-electric device having an evacuated enclosure-means including an anode-portion of the enclosure-means, a cathode-portion of the enclosure-means, and an insulator-to-metal sealing-means between said anode and cathode portions, a quantit of a vaporizable discharge-carrying stable alk li metal having more than three shells in'its atomic structure, a heating-means for heating the active emitting-portion of the cathode, and a coolestsurface cooling-means for cooling some internal portion of the device to a predetermined temp rature-range which is cooler than any other internal portion of the device during the operation of the device, said predetermined. temperature range being at least as'high as thetemperature at which the discharge-carrying metal has a vapor-pressure of about 1000 microns of mercury.

2. A low-voltage high-current vapor-electric device having an evacuatedenclosure-means ineluding an anode-portion of the enclosure-means, a cathode-portion of the enclosure-means, and an insulator-to-metal sealing-means between said anode and cathode portions, a quantity of a discharge-metal selected from the group comprising cesium, rubidium and potassium, a heating-means for heating the active emitting-portion of the cathode, and a coolest-surface cooling-means for cooling some internal portion of the devic to a predetermined temperature-range which is cooler than any other internal portion of the device during the operation of the device, said predetermined temperature-range being at least as high as the temperature at which the discharge-metal has a vapor-pressure of about 1000 microns of mercury.

3. The invention as defined in claim 1, characterized by said discharge-metal being cesium, and said coolest-surface temperature being at least as high as 270 C.

4. The invention as defined in claim 1, characterized by said discharge-metal being cesium, and said coolest-surface temperature being at least as high as 278 C.

5. The invention as defined in claim 1, characterized by said discharge-metal being rubidium, and said coolest-surface temperature being at least as high as 290 C.

6. The invention as defined in claim 1, characterized by said discharge metal being rubidium, and said coolest-surface temperature being at least as high as 297 C.

7. The invention as defined in claim 1, characterized by said discharge-metal being potassium, and said coolest-surface temperature being at least as high as 330 C.

8. The invention as defined in claim 1, characterized by said discharge-metal being potassium, and said coolest-surface temperature being at least as high as 344 C.

9. A low-voltage high-current vapor-electric device having an evacuated enclosure-means including an anode-portion of the enclosure-means, a cathode-portion of the enclosure-means, and an insulator-to-metal sealing-means between said anode and cathode portions, the active emittingportion of the cathode being nickel, a quantity of a vaporizable discharge-carrying stable alkali metal having more than three shells in its atomic structure, a heating-means for heating the active emitting-portion of the cathode, and a coolestsurface cooling-means for cooling some internal portion of the device to a predetermined temperature-range which is cooler than any other internal portion of the device during the operation of the device, said predetermined temperature-range being at least as high as the temperature at which the emissivity of the active emitting-portion of the cathode, at a temperature of about 750 0., is at a rate of above about 5 amperes per square inch, as determined by the maximum currentoutput of the device.

10. A low-voltage high-current vapor-electric device having an evacuated enclosure-means including an anode-portion of the enclosure-means, a cathode-portion of the enclosure-means, and n insulator-to-metal sealing-mean between said anode and cathode portions, the active emittingportion of the cathode being nickel, a quantity of a discharge-metal selected from the group comprising cesium, rubidium and potassium, a heating-means for heating the active emitting-portion of the cathode, and a coolest-surface coolingmeans for cooling some internal portio of the device to a predetermined temperature-range which is cooler than any other internal portion of the device during the operation of the device, said predetermined temperature-range being at least as high as the temperature at which the emissivity of the active emitting-portion of the cathode, at a temperature of about 750 C., is at a rate of above about 5 amperes per square inch, as determined by the maximum current-output of the device.

11. The invention as defined in claim 10, characterized by the discharge-metal being cesium.

12. The invention as defined in claim 10, characterized by the discharge-metal being rubidium.

13. The invention as defined in claim 10, characterized by the discharge-metal being potassium. i

14. A low-voltage high-current vapor-electric device having an evacuated enclosure-means including an anode-portion of the enclosure-means, a cathode-portion of the enclosure-means, and an insulator-to-metal sealing-means between said anode and cathode portions, a quantity of a vaporizable discharge-carrying stable alkali metal having more than three shells in its atomic structure, a heating-means for heating the active emitting-portion of the cathode, and a coolest-- surface cooling-means for cooling some internal portion of the device to a predetermined temperature-range which is cooler than any other internal portion of the device during the operation of the device, said predetermined temperature-range being at least as high as the temperature at which the pressure-distance product of the vapor-pressure of the discharge-metal, multiplied by the anode-cathode spacing, i materially higher than the pressure-distance product at which minimum breakdown occurs.

15. The invention as defined in claim 14, characterized by the anode-cathode spacing within the device being above about 2.5 centimeters.

16. A low-voltage high-current vapor-electric device having an evacuated enclosure-means including an anode-portion of the enclosure-means, a cathode-portion of the enclosure-means, and an insulator-to-metal sealing-means between said anode and cathode portions, a quantity of a discharge-metal selected from the group comprising cesium, rubidium and potassium, a, heating-means for heating the active emitting-portion of the cathode, and a coolest-surface cooling-means for cooling some internal portion of the device to a predetermined temperature-range which is cooler than any other internal portion of the device during the operation of the device, said predetermined temperature-range being at least as high as the temperature at which the pressure-distance product of the vapor-pressure of the discharge-metal, multiplied by the anodecathode spacing, is materially higher than the pressure-distance product at which minimum breakdown occurs.

17. The invention as defined in claim 16, characterized by the anode-cathode spacing within the device being above about 2.5 centimeters.

18. A low-voltage high-current vapor-electric device having an evacuated enclosure-means in cluding an anode-portion of the enclosure-means, a cathode-portion of the enclosure-means, and an insulator-to-metal sealing-means between said anode and cathode portions, 9. quantity of cesium to serve as the discharge-metal, a heating-means for heating the active emitting-portion of the cathode, and a coolest-surface cooling-means for cooling some internal portion of the device to a, predetermined temperature-range which is cooler than any other internal portion of the device during the operation of the device, said predetermined temperature-range being at least as high as the temperature at which the discharge-metal has a vapor-pressure, in microns of mercury, which is at least as high as the quotient which is obtained by dividing 2000 by the anode-cathode spacing in centimeters.

19. The invention as defined in claim 18, characterized by the anode-cathode spacing within the device being above about 2.5 centimeters.

20. A low-voltage high-current vapor-electric device having an evacuated enclosure-means including an anode-portion of the enclosure-means, a cathode-portion of the enclosure-means, and an insulator-to-metal sealing-means between said anode and cathode portions, a quantity of rubidium to serve as the discharge-metal, a heating-means for heating the active emitting-portion of the cathode, and a coolest-surface cooling-means for cooling some internal portion of the device to a predetermined temperature-range which is cooler than any other internal portion of the device during the operation of the device,

said predetermined temperature-range being at least as high as the temperature at which the discharge-metal has a vapor-pressure, in microns of mercury, which is at least as high as the quotient which is obtained by dividing 2000 by the anode-cathode spacing in centimeters.

21. The invention as defined in claim 20, characterized by the anode-cathode spacing within the device being above about 2.5 centimeters.

22. A low-voltage high-current vapor-electric device having an evacuated enclosure-means including an anode-portion of the enclosure-means, a cathode-portion of the enclosure-means, and an insulator-to-metal sealing-means between said anode and cathode portions, a quantity of potassium to serve as the discharge-metal, a heating-means for heating the active emittingpcrtion of the cathode, and a coolest-surface cooling-means for cooling some internal portion of the device to a predetermined temperaturerange which is cooler than any other internal portion of the device during the operation of the device, said predetermined temperature-range being at least as high as the temperature at which the discharge-metal has a vapor-pressure, -in microns of mercury, which is at least as high as the quotient which is obtained by dividing 10,000 by the anode-cathode spacing in centimeters.

23. The invention as defined in claim 22, characterized by the anode-cathode spacing within the device being above about 2.5 centimeters.

JOHN L. BOYER.

References Cited in the file of this patent UNITED STATES PATENTS Number Name Date 2,489,891 Hull Nov. 29, 1949

Images