US2122932A - Gaseous discharge tube - Google Patents

Description

July 5, 1938- 0. s. DUFFENDACK ET AL 2,122,932

GASEOUS DISCHARGE TUBE Filed March 23, 1934 2 Sheets-Sheet 1 y 1938- o. s. DUFFENDACK Er AL 2,122,932

GASEOUS DISCHARGE TUBE I Filed March 23, 1934 2 Sheets-Sheet 2 This is a continuation in part of our prior application Serial Numbe Patented July 5, 1938 GASEOUS mscnmcn TUBE Ora S. Duflendack and Ralph A. Wolfe, 'Ann Arbor, Mich.

Application March 23, 1934, Serial No. 716,972 10 Claims. (Cl. 250-275) r 668,753, filed May 1,

erably not greater than on the order of 1.75 mm. From another point of view an essential re- This invention has to do with a gaseous discharge tube distinguished from prior tubes in. that when relatively low voltages are applied to it, it will permit passage of very large currents without undue disintegration ofthe electrodes. Thus when subjected to voltages as low as 100 to 200 volts peak currents of the order of 100 to 125 amperes may find their way through the tube. Our tube is capable of operating in this manner at very short intervals and enduring such operation for a long period of time.

Another characteristic of our improved tube is that the breakdown voltage is critical. Thus until the required voltage is reached, the tube acts as an excellent insulator, but when that voltage is attained, the tube becomes almost instantly an excellent conductor. This is of advantage especially in engine ignition circuits where the amount of energy available to produce the spark is limited and needs to be conserved.

The rapid breakdown of the tube sets up in the tube circuit a high frequency wave. The high frequency of the discharge is of especial use in ignition systems in that it practically removes the possibility of failure of plugs to spark because of leakage paths such as result from carbon and other deposits on the insulator.

Another feature of our preferred form of tube is that with certain designs and types of electrodes it is sensitive to the action of light, especially after it has been in use a sufficient time so that metal volatilized by the arc has formed a coating on the interior of the tube.

Our improved tube in its preferred form is characterized by electrodes of durable material of low work function in an atmosphere consisting of an inert gas or mixtures of inert gases, with or without the addition of mercury vapor to reduce the break-down voltage.

The tube has been especially developed for use.

with the impulse ignition circuits described and claimed in the prior application of Ora S. Duffendack, Donald W. Randolph and Ralph A. Wolfe, Serial No. 668,754, filed May 1, 1933, and in the application of Donald W. Randolph and Hector Rabezzana, Serial No. 727,888, filed May 28, 1934, and for such use we have found it essential that the work function of the electrodes shall be less than on the order of 2.4 electron volts, that the gas pressure shall not be less than on the order of 2.4 cm., and that the distance between electrodes shall be not less than .20 mm., and prefquirement other than low work function is that the gas pressure and distance between the electrodes shall be such that the latter is at least greater than the mean free path of the electrons.

In the use of the tube as a photoelectric tube, the design requirements are necessarily different. Fundamentally the only essentials seem to be that the low work function electrodes immersed in the inert gas be impressed with a voltage slightly less than the breakdown voltage of the tube and that there be provided within the tube, either on the electrodes themselves, or on the wall of the container or in the form of a third member, a surface which gives oif photoelectrons when exposed to light. These photoelectrons then ionize the gas and so reduce the voltage across the gap that an are forms. This mode of operation is distinguished from that commonly employed in photoelectric tubes in that the electrons from the photoelectrically active surface do not serve as the chief conductor for the arc current but function primarily to trigger off" the are between electrodes which are of substantial size and of good conductivity. Another important difference is that in the conventional photoelectric tube the photoelectrons serve as the principal carriers of current while in our tube their principal function is to ionize the gap to permit passage of current in the manner customary in are or glow discharge.

In the drawings we have illustrated in Figure 1 one form of tube.

Figure 2 is a top plan view and Figure 3 is a bottom plan view of one of the electrodes of Figure 1.

Figures 4, 5, 6, and 7 are views showing modified forms of electrodes.

Figure 8 is a perspective view of the electrodes shown in Figure 7.

Figure 9 is a section through another design of electrodes while Figure 10 is a perspective view of the electrodes of Figure 9.

Figures 11 and 12 show other forms of electrodes.

Figure 13 shows our improved photoelectric tube.

Inasmuch as this tube was developed primarily for use in the impulse ignition circuits previously referred to we shall first refer briefly to the requirements for successful operation of a'tube in such circuits and shall then describe the construction of the tube which satisfies them. In such circuits a magneto, or battery and interrupter, or the like are connected to the terminals of a and the primary of a step-up transformer, ar-

ranged in series, are connected across the terminals of the condenser. The secondary 'of the transformer is connected through a conventional distributor to the usual spark plugs. When the charge on the condenser has attained the necessary potential at discharge takes place between the tube electrodes producing a spark at the plug to which the distributor directs current. It is required of the tube that it shall not break down until the voltage applied to the condenser reaches a sufliciently high value so that the energy stored in it is ample to provide a spark at the plug capable of igniting the mixture. At the same time the break down voltage of the tube must be sufficiently low to permit the use of magnetos or other charging devices of simple, inexpensive construction. Satisfactory tubes have carried peak currents as high as 125 amperes when subjected to around volts.

We have illustrated one of the tubes in Figures 1 to 3. The tube consists of an envelope, preferably of glass, within which are sealed two electrodes, slightly spaced apart. The envelope contains an inert gas "or mixture of gases, with or without addition of a metallic vapor such as mercury as described later on.

In Figure 4 we have shown one of the first forms of electrodes we have used. The surfaces of the electrodes are closest to each other along the axis and then separate. as the edge of the electrodes is approached. This design has the advantage that the discharge will be confined to the region adjacent to the axis instead of playing upon the electrode edges and at times jumping to the backs of the electrodes. How-- ever it has the disadvantage that the life of the tube is likely to be reduced because the discharge plays upon but a small part of the electrode surface and may soon destroy it, thereby increasing the distance between electrodes so that the tube may fail to operate.

We have also used electrodes having the shape shown in Figure 5. Here the flattened central portions afford a considerable area for the play of the discharge.

We have also successfully employed electrodes of the typeshown in Figure 6 but here again the life of the tube maybe reduced owing to the fact that the discharge is confined to a relatively small portion of the surfaces of the electrodes. i

In Figures 1 to 3 we have shown a design where exposure of the tube to light is not objectionable. This design is much the same as that shown in Figure 5 except that a hole I! is drilled in the center of each electrode. The hole reduces the initial starting voltage, operating upon what is known as the hollow cathode principle. This action is due in part to the greater curvature at the edges of the hole. The effect, however, does not persist on account of the transfer of electrode material from the cathode to the anode. Another characteristic of this design is the hollowing out of the back of the electrodes at l5. This is done to reduce the mass of the electrodes so that less treatment will be required to outgas them.

- from the action of light.

trodes be made of uniform material.

. condenser to intermittently charge it. The tube of the electrodesof Figures 1 to 3,-and'in addition the advantage that the discharge is shielded Such shielding is of value in case the electrodes contain or give off materials which become active when exposed to light and affect the starting voltage of the tube as hereinafter described.

In Figures 9 and 10 there is illustrated a modification of the electrodes of Figures 7 and 8 in that one end of the cylindrical electrode is closed.

Figure 11 shows electrodes having hollowed portions facing each other. The advantage of this is that the electrodes may be set with the edges so close together that the discharge will not strike to the edges but plays on the inner hollowed portions and thus the disintegration of the edges is prevented. Y

After some experience with all of these electrode shapes we have standardized on electrodes in the form of simple disks such as are shown in Figure 12 with faces approximately 18 mm. in diameter and 2 mm. thick. With this shape there is a' tendency for the discharge to strike to the edges of the disks and to shift about in position. Such shapes, of course, have the advantageof simplicity of manufacture.

For cold operation the cathode at least must be made either wholly or partly of materials of low work function. The electrodesmust likewise be durable and capable of resisting the effects of the are over .long periods of time.

In general, we have found it to be essential for the successful operation of the tube in the ignition circuit that the cathode have a work function not greater than 2.4 electron volts.

This work function corresponds approximately been found that they are too unstable to withstand the action of the arc. It has consequently been necessary to use stable alloys or other compounds containing comparatively high. proportions of the low work function material. We have had the best results with alloys of nickel, copper and barium; nickel, copper, chromium and barium; aluminum and barium; copper and barium. The first two give similar characteristics to the tubes and are in most respects equivalent. Aluminum-barium alloys used were somewhat porous and full of gas, the release of which caused somewhat erratic operation of the tube. The specimens of copper-barium alloy tested contained less barium than is necessary for satisfactory operation over a long period of time. p

We have found it very desirable that the elec- In general, it is our conclusion that the sounder, the more homogeneous and the cleaner the alloys, the more satisfactory was the operation of electrodes-made from them.

We have had most success with alloys containing nickel and barium. We have made a number of tests to determine the percentage of barium that is necessary in the electrodes of the neon tubes. Neon filled tubes have been operated successfully for as long as 200 hours-at 6,000 sparks quate.

7 2,122,982 per minute when fitted with electrodes contain- .ing as low as 0.7% barium, and this. amount seems to us to be the lower limit of barium con tent for successful commercial operation. Tubes fitted with alloys containing 6%- barium have been operated for more than 900 hours and it would seem that this amount of bariunnis ade- We have successfully employed in our tubes electrodes made of alloys of nickel, copper and barium and of nickel,.copper, barium and chromium described and claimed in the copending application of D. W. Randolph, S. 740,537,

filed August 18, 1934.

Of all of the compositions so far tested, the following has given the best results:

Percent Nickel 58%, Copper Barium 4 Chromium 2 The electrodes must be thoroughly outgassed in order that the gas in the tube may remain pure and the current and voltage characteristics of the tube remain constant. This outgassing may be accomplished in various ways. One method consists in operating thetube at a low gas pressure and a relatively high voltage, of the order of 2,000 to 10,000 volts alternating current or direct current. A current of 200 milliamperes at 2,000 volts will quickly heat the electrodes to incandescence. The same thing can be accomplished by a less current at higher voltage or by a larger current and lower voltage. Thus itwas found possible to outgas the bariumnickel alloy electrodes by operating the tube on a 220 volt circuit. Either alternating current or direct current may be employed for this purpose, but alternating current is preferred as then both electrodes have the same "treatment,

We have had especially good results by operating the tube fllled'with a charge of neon gas in av 220 volt, 60 cycle alternating current circuit until the electrodes are brightly incandescent. The

' gas used is later pumped out. While this process is somewhat wasteful of gas and the electrodesaresomewhat dulled by the treatment, the results have been very satisfactory.

Another method consists in heating the electrodes with high frequency induced currents in a high'vacuum. Pumping of the tube is continued during the degassing process. The electrodes retain bright surfaces throughout, the treatment. There is some evaporation of the metal and a thin film of metal accumulates on thesurface of the tube but this does not impair its operation. We believe this method is preferable for electrodes of nickel-barium alloy containing 4% or more of barium.

As a modification of the last mentioned method y it may be desirable to give the electrodes preliminary treatment in a separate vacuum chamber before mounting them in the tubes. If a thOI. ough outgassing is accomplished in an auxiliary chamber, it is not necessary to heatthe electrodes to the point where they begin to evaporate rapidly after they are mounted in the tube. By this method the deposition of metal and metallic oxides on the walls of the tube is minimized.

To permit passage of current of suflicientlyi high peak value it is essential that'the tube befilled with gas, preferably inert gas, although this is not essential.

otherwise disappear from the tube. Other gases reducing the starting voltage represented by the saturated The advantage of using inert gases'is that they do not readily clean up or disappear more rapidly, shortening the life of thetube.

The. inert gases suitable for use in the tube are helium, neon, argon, krypton and xenon. Ourexperience indicates the possibility of their use either alone, or in mixtures of'varied proporlions, with or without the addition of metallic vapors such as mercury, caesium, potassium and lithium.

Tubes have been made to operate satisfactorily when filled with neon, helium, argon or krypton. Owing principally to its availability and low cost most of our work has been done with neon gas either alone or as the principal gaseous constituent.

Marked improvement in the operating characteristics of the neon tubes has resulted from the addition of traces of argon, krypton, or mercury vapor. These added materials have lower ionizing potentials than neon, and hence increased ionization results. As a consequence the starting voltage of the tube is reduced. The addition of 0.01 per cent to 1 per cent of argon or 'mercury vapor, for instance, will reduce the duction of sparks of greater energy.

Mercury has the added beneficial effect of acting as a getter, particularly for hydrogen. The action on hydrogen is important because this gas is less easily removed from the rare gases than are other contaminations. It has been observed further that there is no tendency for the electrodes to bridge over and short circuit when the mixed gasesare used, particularly when mercury is present.

The mercury-neon mixture has the disadvantage that the concentration of mercury vapor is not constant but varies with the temperature of the tube. Thus in very cold-weather the amount of mercury vapor is reduced at the start and its effect on the discharge is likewise reduced. Tubes filled with a mixture of neon and argon do not have this defect as argon does not condense at the lowest temperature that would prevail in winter. Therefore, neon-argon mixtures are to be preferred over neon-mercury mixtures. A combination of neon, argon, and mercury has been found to be most satisfactory as the argon of the cold tube sufficiently and the beneficial efiects of the mercury are active shortly after the tube has been put into operation. A mixture containing argon at .01 mm. pressure and the mercury vapor that is vaporat the temperature of the tube has been found to be very satisfactory and tubes so filled have practically no varition in starting voltage with the tempera ture. The neon pressure should be about 65 cm. of mercury.-

We have also introduced into the tube inert gas or mixtures of inert gases together with caesium vapor, potassium vapor or lithium vapors and have obtained substantially the same effects as in the case ofmercury. These vapors not only reduce the starting voltage of the tube and increase the energy of the spark, but also act as getters and alloy with the electrodes to keep the work-rfunction down. In general some or all of theseeffects may be expectedto be attained by tically all the common gases by absorbing them,

whereas not a great quantity of neon is absorbed. In addition, various getters such as magnesium, barium, and lithium, and various combinations of these, were flashed in the tube being filled or in the connecting lines to the gas supply or both. By these methods, much purer gases were introduced into the discharge tubes, and, as a result, the starting voltage of the tube was lowered and remained much more nearly constant. The getter flashed in the tube itself had no effect on the operation of the tube and took upany contaminating gas that might be released from the electrodes or tube walls during the operation of the tube. It seems necessary to flash a getter in each tube not containing mercury in order to insure that it will have a constant starting voltage and this practice has been adopted as standard.

We have varied the gas pressure over a range from .5 cm. of mercury to above atmospheric pressure, 1. e. '76 cm. At the lower pressures there is a tendency for the discharge to take the form of a glow and to limit the current to too I small a value. For this form of discharge, electrodes of larger effective area than have been used are demanded, and the performance will be subject to the disadvantages previously pointed out. The higher the gas pressure, the greater is the peak current of the discharge and hence, the larger the E. M. F. developed in the secondary of the transformer of the ignition system. A high initial pressure is desirable on account of the gradual clean up of the gas. The higher the initial pressure the longer, in general, will be the life of the tube, but there is the accompanying disadvantage that the gap break-down voltage increases with increase in pressure.

We have found the lower practical limit of gas pressure to be on the order of 2.4 cm. of mercury. The preferred pressure is near atmospheric between and cm. Trouble is experienced in sealing the tubes if an attempt is made to use pressures above atmospheric.

A number of tests were made to determine the optimum distance between the electrodes. This distance depends on the pressure of the gas and the starting voltage desired. At low values of this product, however, the law is not valid and the starting voltage again rises with decreasing values of the product. Thus at the lower gas pressures it is possible to have the electrodes so close together that the starting voltage is above the minimum. Over a considerable range of values of this product, near the value for minimum starting voltage, the starting voltage changes-very slowly with changes in the value of this product. It isin this range that it is desirable to have the tube operate.

In general the product of gas pressure and distance between the electrodes should be greater thanthe mean free path of the electrons. From the standpoint of practical manufacture and operation gaps of less than .20 mm. are undesirable because any accumulation of metal on one of the electrodes, due to sputtering or some other process, may cause a bridging of the gap and the short-circuiting' of the tube. Gap distances on the order of 1.75 mm. appear to be the maximum desirable in practice. Best results have been obtained with nickel-barium electrodes of from 4% to 6% barium content, gas pressures of 60 to '15 cm. of mercury, together with gap distances of from .5 mm. to 1.75 mm.

We have discussed the principal factors that determine the performance of this tube, 1. e.; work function of the electrodes, gas pressure, distance between electrodes and the character of gas, with or without metallic vapor, employed in the tube. We have specified that for satisfactory operation in the ignition system herein disclosed the oathode, at least, should have a work function not more than 2.4 electron volts; that the gas pressure should be not less than 2.4 cm. of mercury; that the distance between electrodes should be not less than .020 cm., and, with existing electrode materials, not greater than approximately .175 cm. We have also grouped pressure and distance together under the criterion that the gas pressure times gap distance shall be such that the latter is greater than the mean free path of the electrons. These relations between all of these variables may also be indicated by the following formula:

(Gas densityXgap distance) Work function shall not be less than K, where K is a constant dependent upon the kind of gas or gas and vapor employed and the units used in measuring density, distance and work function. Inserting in the equation the limits of the quantities in the case of neon we have %X.000902 (gr. per cm. .02 (cm.)) 2.4 (electron-volts) ing to these requirements will give satisfactory operation in the ignition circuits of the character previously referred to.

The'interrelation of factors indicated in the formula is, of course, involved in any specific design of tube and may be made clearer by the following discussion. It will be apparent that the lower the work function of the electrode, the higher may the gas pressure be raised and still maintain satisfactory operation characteristics. In generaL'the peak voltage increases with an increase in gas pressure and when the pressure becomes too great the .peak voltage may become higher than the voltage that can be developed across the condenser. But the peak voltage also depends upon the work function and hence a lower work function material will permit increasing the gas pressure. The lower the work function the greater can be the separation of the electrodes for the same reason.

For high speed operation, a low work function cathode is essential. At high speeds the condenser cannot be charged to as high voltage as at lower speeds. Hence, the starting voltage of the tube must not be higher than the voltage that can be attained at the highest operating the emission of electrons.

give satisfactory speed, and yet the peak current passed by the tube must be sufficient to develop an E. M. F. in the secondary of the transformer high enough to cause a spark at the gap. In order to attain the required current, the gas pressure must be adequateand so a low work function material for the cathode is demanded.

Our tube is capable of operation within a wide range of temperatures. We have successfully operated a tube chilled to the temperature of liquid air, i. e. -196 C. We have also employed heated cathodes. In the application of heat there is no limit to the temperatures which may be employed other than such as are imposed by disintegration of the electrodes or of the materials of which the tube walls are made.

If it is convenient to heat the electrodes, in spite of the cost, inconvenience, and complication, then a greater choice of electrode materials is afforded for the application of heat increases We have successfully employed in gas filled tubes such as herein disclosed, electrodes made of a base material, such as nickel, coated with compounds of electron emitting materials as in vacuum tube manuiacture, such as the oxides or carbonates of the alkaline earth metals, especially barium and strontium. Such electrodes are preferably acti vated in known manner to increase their emissivity as by heating in vacuum for a certain period of time. Where heating is permissible, a number of elements, such as pure tungsten, will emission, as will also alloys such as described herein with too little material of low work function for cold operation. Heating of the cathode may be accomplished by radiation from an adjacent filament supplied with current from'a battery or other source. Gas filled tubes such as herein disclosed provided with heated cathodes made of a wide range of materials will give satisfactory results, but the necessity of at least preliminary heating makes them less desirable for many uses. Heating of the cathode permits some reduction of the minimum gas pressure but if too low pressure is used, for example on the order of a fraction of a millimeter of mercury, the secondary voltage at the spark is reduced and becomes too low to produce constant sparking of the plugs.

We have also discovered that certain of our tubes possess marked photo-electric properties. The tubes possessing these properties are made with the nickel barium alloys, and the best of them have coatings of metal or metal oxides or both onthe walls of the envelope. The most light sensitive of the tubes are those having electrodes containing nickel, copper and barium, and the coatings on the envelopes are believed to consist of, copper oxide, nickel oxide, and barium, nickel and copper metal volatilized by the discharge. Of these materials barium and copper oxide are known to possess marked photoelectric properties and it is believed they account for the performance. We have also noted this action in tubes containing chromium in addition to the materials just mentioned and in such case it is likely that the coating contains a proportion of chromium metal.

In Figure 13 there is illustrated one form of such tube. It is of the construction previously described consisting of spaced electrodes 36 in an atmosphere of inert gas. 40 indicates a light sensitive coating onthe interior walls of the envelope. The electrodes may be supplied with power from any suitable source.

The tube is so designed that the voltage applied to it is slightly below that necessary to break down the gap in the tube. When light strikes the coating 40 the photoelectrons given off by the action of light ionize the gas in the gap causing it to break down under the action 01' the applied voltage.

It will be understood that our photoelectric tube may also be employed to close any circuits desired, for example, lighting circuits, and if desired no provision need be made for subsequent opening of the circuit other than the usual handoperated switches. In general, the tube will be of use in any of the relations in which photoelectric tubes with their usual relay circuits are now employed. For example the motor illustrated in Figure 13 may be employed to open a door and the circuit breaker or breakers could be operated by the door operating mechanism so that when the movement was completed the circuit would be broken and reset for the next cycle of operations which, by the employment of suitable reversing motor and reversing switches, might, if

desired, accomplish subsequent closing of the door.

It will be understood that when our tube is used as a photoelectric tube to control the flow of a considerable current through the action of light, it is necessary to apply to the electrodes a voltage almost but not quite equal to the breakdown voltage of the gap. The photoelectrons given oif by the action of light on the photosensitive surfaces break down the gap by ionization by impact, so that the phenomenon may properly be described as trigger action.

It is quite clear that the photo-sensitive surface need not be on the walls of the tube but may be on one or both of the electrodes or may be in theform of a third member. It is clear, too, that the light sensitive member need not be made of material volatilized from the electrodes but may be formed separately of any suitable light sensitive materials and mounted in the tube. Or if desired it may be formed as a film on the walls of the envelope by volatilizing metals therein by the same methods now in use in introducing getters inside the tube. It will be apparent that should the tube be used to carry a steady current there is no way in which the flow when once initiated may be interrupted by light action. It will of course be necessary to provide for opening of contacts in the circuit by the device, for example, a servo-motor, that may be actuated by the flow of current through the tube.

It will be apparent that our improved tube is capable of many uses. It, in effect, constitutes an electric valve which breaks down on the application of relatively low voltages and permits passage of currents of high peak value and is further characterized by long life and constant performance characteristics. It is to be expected that many uses other than those herein mentioned will develop in the future. As a possible example, the tube may find use as a lightning arrester or protector for transmission lines.

It has been previously pointed out that if the tube is photoelectrically active and voltage is applied to it that is slightly below the breakdown voltage of the gap, a beam of light may be used to ionize the gap and cause the tube to discharge. Obviously under the same conditions, the gap may be broken down by a momentary application of higher voltage instead of by the use of light.

If the circuit containing the tube is supplied with alternating current, the discharge in the tube, once initiated, will be either continuous or discontinuous depending on the frequency. If

the frequency is sufllciently high so that there is not suilicient time between alternations to permit the gap to de-ionize, the discharge will be con-' tinuous in the sense that any flow of alternating current is continuous. If the frequency is so low that there is suflicient time for de-ionization between alternations there will be but one discharge through the tube. Where the circuit is supplied with direct current there will be but one surge of current followedby continuous discharge. In such case, of course, but one impulse will be received in the coupled circuit.

Obviously the flow of energy released in either a main circuit or coupled circuit as a result of discharge through the tube maybe used in any way in which electrical energy is used, as to produce, heat, light, radiation or mechanical movement.

We claim: 1. An arc discharge tube comprising a plurality of spaced, durable, electrodes, one of said electrodes comprising material of low work function,-

.7% barium, said tube containing an atmosphere consisting predominantly of neon with small proportions of argon and mercury, the pressure of said atmosphere being not less than 2.4 cm., the

distancebetween electrodes being greater than on the order of .2 mm, but not greater than on the order of 1.75 mm.

4. An arc discharge tube comprising a plurality of spaced, durableelectrodes one of which consists of a homogeneous alloy containing from .7% to 6% barium, said tube containing an atmosphere consisting predominantly of neon with small proportions of another material. of lower ionizing potential at a pressure not less than on the order of 2.4 cm., said electrodes being separated from each other by not less than approximately .2 mm.

5. An arc discharge tube comprising a plurality of spaced, durable electrodes one of which consists of a homogeneous alloy containing from .'I-% to 6% barium, said tube containing an atmosphere consisting predominantly of neon with small proportions of another inert gas and a metallic vapor, said last-named gas and vapor being of lower ionizing potential than neon, the pressure of said atmosphere being not less than on the order of 2.4 cm.

6. An arc discharge tube comprising a plurality of spaced, durable electrodes one of which consists oi. a homogeneous alloy containing nickel,

copper and barium, the barium content being not vless than 3%, said tube containing an atmosphere of inert gas at a pressure not less than 2.4 cm., the distance between electrodes being not less than on the order of .2 mm.

7. An arc discharge tube comprising a plurality of spaced, durable electrodes, one of which consists of a homogeneous alloy containing nickel, copper and barium, the barium content being at least on the order of 31%, said tube containing an atmosphere of neon together with small proportions of argon and mercury at a pressure not less than 2.4 cm., the distance between electrodes being not less than on the order of .2,mm., and not greater than on the order of 1.75 mm.

8. An arc discharge tube comprising a plurality of spaced, durable electrodes one of which contains from 4 to 6% barium, said tube containing an atmosphere consisting predominantly of neon with small proportions of argon and mercury, the pressure of said atmosphere being on the order of i from to centimeters of mercury, said electrodes being separated from each other by not less than approximately .5 millimeter and not more than approximately 1.75 millimeters.

. 9. An arc discharge tube comprising a plurality of spaced, durable electrodes, twb of which comprise material of low work function and having opposed active surfaces of substantial area, said tube containing a gaseous atmosphere inert to the electrode material and the tube wall, the product of gas pressure times gap distance divided by work function being not less than .000000237 times a constant characteristic of the inert gas employed.

10. An arc discharge tube comprising a pair of spaced, durable electrodes each consisting of a

Images