US2083107A - Thermionic tube - Google Patents

Description

Patented June 8,V 1937.

THERMIONIC TUBE Felix L. Yerzley, Ithaca, N. Y., assigner to Radio Corporation of America, New York, N. Y., a corporation of Delaware Application August 3, 1934, Serial No. 738,265

2 Claims. (Cl. Z50-57) My invention relates to thermionic tubes and has particular reference to tubes employing heated surfaces for maintaining a current of electroris. Y

shown that together with the electrons there is always some emission of positive ions. The mobility Vof ions and electrons is such that an ion may exist for a long time in the electron space 5 Thermionic tubes, for example those usedras charge before it is neutralized by combination 5 detectors and amplifiers in radio circuits, must with an electron. During the time it is presentin maintain a very uniform rate of emission of space charge region the potential there is someelectrons. Any variations in the rate of emiswhat higher and many more electrons reach the sion become immediately apparent in the output plate than is normal. From theoretical considof the tube. When suiiiciently amplified, they eratons it has been shown that a single positive 10 manifest themselves, for instance, as noise in a ion in a spaceV charge of electrons may release receiving circuit. With uneven electron emishundreds, and under certain conditions thousion the noise amplitude may be great enough Sands, of electrons. Positive ions, then, may be to mask the incoming signals completely. responsible for great variations in the plate cur- An ideal vacuum tube, therefore, must be comrent, and are therefore extremely detrimental to 15 pieteiy free from such fluctuations in order te the proper operation of thermionio tubes. In the maintain maximum possible sensitivity of tube CaSe Of radio tubes the disturbance du@ t0 DOS- circuits. tive ions results in a high noise level.

Extensive experience with the thermionie tubes I discovered. however, that it is possible to elimhas demonstrated, however, that it is extremely lnate almost COmDletely the effect 0f the POSitVe 20 diicult to obtain a perfectly uniform flow of ions on the plate current. My research has shown electrons. Unlike a flowing liquid, the stream of that for-a straight filament the DOStVe i011 emiselectrons consists of nite particles, each having Sion iS limited almost entirely t0 tS 000161' DOI- a discrete electrical charge, and the rate of emistions- I found that by placing ConductOI'S near sion of these particles is not constant. Thus, althese cooler portions of the lament and apply- 25 though the average current is reasonably the same ing to them a negatiVe DOtentiafl it iS POSSble t0 over a period of time, instantaneous values ef the collect the positive ions thereby largely eliminatemission current vary slightly from this mean. ing their eiect on the flow of electrons.

This statistical uctuation represents one of the My invention iS mre fully described in the aomajor problems in the development of thermionic Companying specification and drawing in which- 30 tubes. It is well known as the shot eieet, so Fis. 1 is a diagrammatic view of a thermionic named by W. Schottky who first predicted it in tube employing my DOStl/'e'ion C0lleCt01`S,-Fg 2 1918. is a diagrammatic viewof a tube ywithout my A large part of these fluctuations can be recollectcrs,.Fig. 3 is a curve showing temperature moved by the use of aspace charge over the emitdistribution along the filament of a tube, Fig. 4 35 ting surface of the filament. The space charge is a curve showing the rate of electron and posii i is formed by adjusting relative potentials of the tive ion emission from the filament, and' Figs.

plate and filament to a point where the mutual 5 and 6 are diagrammatic views of another arrepulsion of the electrons begins to take eiect in rangement Of a illafment With my ColleCtOl'S.

40 retarding their movement and causing thereby a The findings of my research are that for a 40 certain accumulation or cloud of free electrons straight filament I, supported by lead wires 2, near the emitting surface of the filament. This the positive ion emission is limited almost encloud or space charge serves as a reservoir abtirely to the end portions of the filament (or a sorbing the statistical uctuations, and its outer similar ion emitting element) where there is a boundary serves as a fairly uniform emitter. steep temperature gradient. This may be ex- 45 'I'he accumulation of electrons forming a space plained by the fact that the cooler portions of the charge alters, however, the potential distribution filament retain most of the impurities originally between the filament and the plate. The high present in the wire, and these impurities, form density of electrons near the emitter is sucient the principal cause of the positive ion emission.

to lower the potential even below that of the Of course, a certain amount of the impurities 50 emitting surface. Such lowering of potential inoriginally present always remain in the central troduces a new and very undesirable effect by portion, and it seems diflicult to reconcile this causing the attraction of positive ions which are fact with the almost total lack of positive ion also emitted by the filament. emission from the central portion. A possible ex- Experience with electron emitting bodies has planation is that in the central region the im- -55 purities leave the surface as neutral atoms or as negative ions.

The intensity of the electron emission is a function of the temperature of the filament. For a hot filament the temperature distribution along the lament may be represented approximately by the curve 3 in Fig. 3, while the electron emission may be represented by a curve 4, Fig. 4. In the latter figure abscissa on this curve represents points on the length of the lament, and the ordinates represent the positive ion emission and the intensity of the electron emission, the

latter decreasing rapidly near the ends of the lament due to the cooling eeot of the supporting wires. Fig. 4 by curves 5 corresponding tov the cooler portions of the filament. Such conditions of the electron and ion emission exist in an ordinary tube, such as a triode illustrated diagrammatically in Fig. 2. Here the filamentV I is surrounded by a grid 6 connected with a grid circuit 1. The grid is surrounded by a plate 8 connected with a plate circuit 'I' through a source of electric current 9. In my tube as shown in Figs. l and 3, I use the same principal elements, but I add shields or electrodes I Il surrounding the cooler portions of the filament. These shields are given a negative potential from a source of current I I. The potential must be so adjusted that the shields will attract and remove the positive ions emitted by the end portions of the lament. The grid and the plate may be correspondingly shortened so as to surround the active portions only of the filament not protected by my shields or ion collectors. Such shortening of the usual elements of the tube will not materially aifect its output, as it removes only comparatively inactive portions corresponding to the zones of low electron emission on the filament. With this arrangement my shields do not in any way interfere with the normal operation of the tube. This follows from the fact that the operation of a thermionic tube depends upon the distribution of potential between the plate and the filament, and not upon the potential distribution in other parts of the tube. There would be no disturbance caused even were it found necessary in a special case to place the positive ion collectors over part of the surface emitting electrons in large numbers, or if there is some overlapping of the usual tube elements and my ion collectors. In general, however, my researches pointed out that the collectors fulfil their purpose satisfactorily when they are placed only over regions where the electron emission is negligible.

The positioning and size of my collectors Vary depending on the emitting properties of the filament, whether it is a pure metallic filament, an oxide coated filament, a thoriated lament, or some other type. It depends also on the temperature gradient and the potential applied, i. e., the

The positive ion emission is represented inA consequent distribution of the electron space charge.

My collectors need not form a part of any of the ordinary tube circuits. They form an essential part of a bleeder circuit provided for removing an undesirable component of the total emission without appreciably altering other conditions of the tube.

IvIy collectors are especially eifective with tubes employing coated filaments and for tubes using heater cathodes. These operate at relatively low temperatures and are especially capable of emitting positive ions in the manner indicated.

For filaments other than a single =straight length of fine Wire, such as hairpin types, or inverted W types, there are more than two short lengths of filament which emit positive ions in large quantities. In all such cases, however, it is possible to provide my collectors over the coolei' portions of the lament, as shown, for instance, in Figs. 5 and 6. In Fig. 5 there is illustrated a straight lament I2 attached to the supports I5 and I 6, the latter or both of which may be lead-in wires. The shields I3 and I4 serve to protect the ilament at its points of support. In Fig. 6 a W lament I 2 is protected by shields I3' and I4' near its points of support I5 and lead wires I3.

Emanuela-A 4.8 mil tungsten wire'lament 7 cm. long was found to require positive ion collectors 5 millimeters long at each end of the iilament. A negative potential of 30 volts was found to be satisfactory relative to the ilament. This arrangement leaves,r therefore, of the length of the filament to be surrounded by the usual grid and plate, utilizing, however, over 99.5% of the total electron emission. v

It should be noted that my positive ion collectors are applicable to all tubes employing heated surfaces for the electron emission and not employing a gas discharge. The addition of my collectors to tubes does not call for any fundamental alterations in diodes, triodes, tetrodes, pentodes, or any other of the multiplicity of tubes now in use or in the process of development.

I claim as my invention:

l. A thermionic system comprising in combination a tube provided with grid and anode electrodes, and an electron emitting member, means for supporting said emitting member, means for heating said emitting member, the portions of which adjacent points of support are cooler than other portions of the member, shields surrounding the cooler portions of the member adjacent its points of support, and means to maintain the shields at a negative potential with respect to the emitting member, said shields being located substantially parallel to said member and nearer thereto than the anode.

2. A thermionic system as dened in claim l, wherein the shields are electrically connected together.

FELIX L. YERZLEY.

Images