US2012339A - Rectifier - Google Patents

Description

Aug. 27, 1935. v A s ET AL 2,012,339

RECTIFIER Filed Jan. 7, 1931 INVENTORS fiwza/dkZ'du a/ds fi arliffinz'ifi BY ATTORNEYS ensures Aug. 21, less I 2,012,339

UNITED STATES PATENT OFFICE "RECTIFIER Donald v. Edwards, New York, 'N. Y., and Earl K. Smith, East Orange, N. L, assignors to Electrons, Inc. of Delaware, a corporation Dela-. ware Application January 7, 1931, Serial No. 507,052

. 8 Claims. (01. 250-215) This invention relates, to controlled rectiflers as in practice the various parts of the circuit of the gaseous type. in which the tube is used 'will assume various The object of the invention is to provide means shapes, proportions and arrangements. for preventing loss of-grid control. The loss of control above referred to, in our vHeretofore in tubes of this type, after the tube opinion is caused by cold emission of electrons 5 has been in .operation for a short time it has from the grid when barium evaporated fromthg occurred that the tube started despite the fact filament is deposited on the grid; that the posithat the grid voltage was more negative than the tive electrostatic field from the plate pulls these trigger voltage. 1. Y electrons from the grid; and that these electrons 10 The invention will be described with reference on their way to the plate ionize the gas and the 10 to the accompanying drawing, in which tube then starts in spite of a negative charge on Fig. 1 shows a conventional rectifier embodying the grid.

the invention; and Various expedients have been proposed with a Fig. 2 an enlarged portion of a part of the grid view to minimizing the chances of barium so 1.3 structure embodying the invention. depositing. One of such proposals was to reduce In app1icants Patent No. 1,905,692 issued April the current density through the tube, thereby 25, 1933, the structure of an electron tube of the to reduce the possible barium vaporization. We type shown in Fig. l of the present application is propose to provide means for preventing emisdescribed and claimed. sion from the grid even though barium be dei represents the glass tube; 2 the plate; 4 the posited thereon. 20 cathode, and 8 the grid. In the form shown the To this end we employ a poisonous grid coatcathode 4 is in cylindrical form having a pluing, thereby to permit increased density of the rality of filaments 3 connected to the filament current through the tube and the use of a smaller circuit 5. Preferably the filaments 3 are oxide grid, and thereby decreasing the relative heating 2., coated but the interior of the cylinder or. both of the grid, even though the tube be operated at cylinder and filaments may be oxide coated. As increased current density. We further provide a coating we prefer the resultant product of an means for maintaining the grid in the poisoned alkaline earth oxide and an oxide of a metal capacondition for a prolonged time. ble of .forming amphoteric compounds less acid In one embodiment of the invention the grid is than titanic acid, such, for example, as aluminum coated with the oxide of a metal of high work 30 oxide, zirconium oxide or the like, heated in function, for example the oxides of tantalum, vacuum. molybdenum and tungsten. The body of the As gaseous conducting medium we prefer grid may be of other material but preferably is argon, although mercury or other gas or mixtures a metal of high work function such as timgsten.

3.? thereof may be employed. InFig. 2 the oxide coating is conventionally The grid structure shown comprises a disk t shown at 25 on the grid wires l0. To avoid peelhaving an opening therein across which are a ing ofi of the oxide, if tungsten oxide is used, we number of bars Ill, the spaces 9 between the bars perier to use the oxide in amorphous form, inbeing suitably adjusted. Plate 2 and cathode d stead of the crystal formation customarily used.

are connected by circuit 6' receiving power from One way of converting the crystalline formation 40 an alternating current source 7 through transinto the amorphous form is to treat it with amformer ll. Preferably the potential on the grid monia at a, lower temperature than minus 21 is controlled by a potentiometer l5 connected centigrade. across the terminals of a battery I8 through a The efiect of heat radiated from the filament,

limiting resistance H, and a variable pointer [6 heat radiated from the discharge and ionic bom- 45 whereby the battery voltage upon the grid may be bardment should be kept below the temperature adjusted. The battery I8 is connected to the of chemical breakdown of the oxide. Thus the cathode through the secondary 20 of a transcurrent density through the tube may be largely former connected across the plate-cathode circuit increased beyond present practice. This permits and thence through wire E9 to the cathode. the use of asmaller grid than normally used, and 50 Thereby an alternating potential which may be in turn reduces the quantity of heat received by made out of phase by the insertion of a conthe grid and reduces the distance the heat has denser H, or an inductance 22 or both, is also to flow to reach the heat radiating part of the applied to the grid. grid outside of the discharge. In practice we I5 The tube herein described is typical, inasmuch have found it possible to build tubes having more than ten times the current density through grid than in prior tubes. This has made possible a grid having one-tenth of the former grid surface and consequent reduction in grid heat. A tube of medium size, following the former practice, would require, say, ten or more square inches of grid surface, whereas with our method this has been cut to one. square inch. The temperature of the discharge is extremely high regardless of the current density and therefore the heat received by the smaller grid will be about one-tenth that of the other. Furthermore, this much smaller quantity of heat has only to flow through at the most one-half inch of grid wire to reach the heat radiating part of the grid outside of the discharge, whereas with the large grid it would have to flow through at least one and one-half inches of metal. The result of this is that the small high current density grid runs only a few degrees above the heat radiating portion of the grid. When the grid is large the heat gradient must be high, and therefore the center becomes very hot.

Regarding materials for poisoning the barium, we have tried oxides of tantalum, molybdenum and tungsten and have found them satisfactory. We have reason to believe, however, that the oxides of all high work function but nonradio active materials will work satisfactorily. In fact, some measurements show that some of the oxides of metals whose oxides are anhydrides of amphoteric compounds, particularly the more highly negative ones, would be satisfactory. We also find that carbides, nitrides and halogen compounds of. high work function nonradio active metals will work satisfactorily.

Since there is a tendency for the filament to continue sending barium atoms to the grid, we have found it advantageous to provide a means for feeding poisoning material to. the grid during the operation of the tube. To this end the plate is made of carbon, which, when the tube is in operation, will occasionally release, bydistintegration, carbon ions. These carbon ions will be attracted to the grid by the electrostatic field in the same manner as the barium atoms are attracted from the cathode to the grid and will deposit on top of such barium atoms as may have reached the grid, and will poison any emission of electrons from the grid. if the tube is so arranged as to have a continuous disintegration of carbon it is possible to poison the barium atoms by freshly deposited carbon atoms about as fast as barium deposits on the grid.

Carbon alone is not poisonous any more than any equally electro-negative metal, but due to its peculiar absorbing powers it always contains considerable quantities of absorbed gases and impurities which are never completely removed on the pump and which serve to poison the barium deposited on it. After the tube has been in operation some time, however, these impurities are gradually cleaned from the surface and pure carbon is exposed, after which the tube loses control. But when the carbon ions are released from the plate by disintegration as above described'the carbon is deposited'on top of the barium and effectively poisons emission from the barium.

the

We offer the following as a theoretical explanation of the effectiveness of the oxides of high work function elements, and the depositing of carbon ions on the oxides, although it is to be understood that it is theory only. We think that barium is attracted to a very electro-negative metal and has a tendency to be held perhaps chemically just as in a vacuum tube hydrogen is very strongly attracted by platinum; that is, a very low work function material is attracted to a high work function material. This first coating of barium is poisoned and does not emit. However, when a fresh coat arrives to cover the first coating of barium it is not poisoned because the effect of the tungsten has been neutralized by combining with the first coat of barium. If, instead of pure metal,

a compound of themetal and another element of relatively high work function is used, the barium has less tendency to be attracted and retained chemically and if it is retained, it would probably displace one of the materials which would remain on the surface and poison the barium or any other barium which may arrive.

We claim:

1. A grid structure coated with poisoningmaterial comprising amorphous tungsten oxide.

2. A gaseous. discharge tube comprising a cathode, anode and grid and means comprising amorphous tungsten oxide adhering to the grid for poisoning evaporated emitter material from the cathode.

3. A grid structure coated with the amorphous oxide of a high work function metal.

4. A discharge tube comprising a gaseous conducting medium, a cathode, anode and a grid comprising means for concentrating the discharge from the cathode to the anode, and a coatingof poisoning material on the grid.

5. The method of preventing grid emission in a gaseous discharge tube which consists in concentrating the discharge from cathode to anode and feeding poisoning material to the grid.

6. A controlled grid rectifier tube comprising a gaseous conductive medium, an electron emissive cathode, an anode, a grid having a perforate portion adapted to concentrate the current therethrough and having aheat radiating portion surrounding said perforate portion, and a compound of a high work function metal on said perforate portion.

7.. A grid coating for a gas-filled, grid controlled rectifier comprising a compound of a highly electro-negative metal and another element of relatively high work function, the said compound being poisonous to the emission of highly electropositive substances.

8. A discharge tube comprising a gaseous conducting medium, an oxide coated cathode, an anode, a grid having a discharge opening, said grid opening being small in size relative to the discharge current to pass through the grid whereby the discharge is concentrated through said opening, and means for disintegrating and feeding poisoning material from the anode to the grid during operation of the tube.

' EARL K. SMITH.

DONALD V. EDWARDS.

Images