Thermionic Valve

The other electrodes of a thermionic valve are made of nickel or molybdenum or other conductors having high melting points. The anode is usually cylindrical, surrounding the grid and fila ment, but may consist of two parallel flat plates situated on opposite sides of the filament. The grid is usually a spiral of wire or a mesh wound on a cylinder with the filament as axis. The electrode system is fixed at the centre of the containing glass, silica or metal envelope and is supported by a glass (or silica) "pinch" through which pass the connecting wires between the electrodes and the outside terminals. In a two-electrode valve there are three outside terminals, one for the plate and two for the positive and negative ends of the filament heating battery. In the case of the three-electrode valve there are the same three terminals, together with an extra one for the grid.

The majority of valves are those of the high-vacuum type, so that in the process of manufacture various methods are used to rid the electrodes and glass envelope of occluded gas both before and during the process of pumping the air from the envelope. Such methods are also employed even with a tube which is sub sequently made "soft" by the admission of helium.

Internal Action of a Thermionic Valve.

The performance of a thermionic valve in an electrical circuit is usually interpreted in terms of its static characteristic curves which represent the relations between the currents flowing to the collecting electrodes and the potentials of these electrodes with respect to the filament. Since the data for such curves are obtained using steady electrode potentials certain discrepancies are often found between the behaviour predicted from these curves and the actual performance of a tube when functioning with rapidly alternating electrode potentials and currents. In such cases the discrepancy can usually be traced to the effects of the electrode capacities or to the small but finite time required for the electrode currents to respond to the electrode voltage changes. (a) The Two-Electrode Valve. —in the case of the diode or two electrode valve the most impor tant characteristic is that showing the relation between the anode potential and the anode current (ia), the filament current being maintained constant at the normal operating value. A typical diode characteristic is shown in fig. 2 and may be interpreted in terms of the internal action of the tube. When the anode is at a positive potential it attracts the negative electrons and so an anode current flows. On the other hand when the anode is negatively charged it repels the electrons back into the filament as fast as they are emitted and no anode current flows. From the characteristic it will be seen

that as the positive potential of the anode is increased the anode current also increases until a certain maximum value is reached. A further increase of anode potential brings about no further increase of anode current and the latter is then said to have reached saturation value The saturation current is reached when all the electrons emitted by the filament are attracted and caught by the anode. The necessity for a fairly large anode poten tial to produce saturation current is caused by the mutual elec trical repulsion between the electrons which are situated at any moment in the space between the filament and the anode. This body of electrons is termed the space charge. The electrons which are on their way to the anode repel those just starting from the filament, thus neutralizing to a large extent the pull which the positively charged anode exerts on the latter.

The effect of the electron space charge in opposing the establish ment of saturation conditions has been examined quantitatively by C. D. Child and by I. Langmuir. The essential basis of their calculation is the assumption that the disposition and magnitude of the space charge is so adjusted that an electron at the filament surface is repelled by the electrons in front of it with a force equal and opposite to that due to the positive charge on the anode. For such conditions the space charge must obviously be equal and opposite to the charge on the anode. In the case of a plane emitting surface with a parallel plane anode such assumptions lead to the Child-Langmuir f ormula: Tables of values of 02 for different values of have been given ro by I. Langmuir and K. B. Blodgett from which it is found that when is greater than io, a condition usually met in practice, ro the value of (3 is little different from unity.

Although the Child-Langmuir formulae are sufficiently accurate for most practical purposes, the assumption of an unlimited supply of electrons from the filament makes the formulae satisfactory only for cases in which the anode current is appreciably below saturation value. Moreover the electrons, on emerging from the filament are all assumed to possess uniform velocities, either zero or finite, whereas it is known that such electrons possess velocities with a Maxwellian distribution. The effects of both finite filament emission and of the Maxwellian distribution of electron velocities have been taken into account in more elaborate treatments of the problem by P. S. Epstein and T. C. Fry.