Two examples of the use of the two-electrode valve are its applications as a detector of wireless signals and as a rectifier of alternating current for power purposes. Both of these applications depend on the fact that when an alternating voltage is applied between anode and filament a current flows through the valve only during the half cycle when the anode is positive. The prac tical use of the two-electrode valve as a detector of wireless signals was first made by J. A. Fleming in though the fact that a tube containing rarefied gas with one hot and one cold electrode showed unilateral conductivity had been previously demonstrated by Elster and Geitel in 1889. A simple wireless circuit with diode rectifier is represented in fig. 3. As a result of the high frequency oscillatory potentials produced between the points A. and B. by the action of wireless waves on the aerial, unidirectional currents pass through and actuate the telephone or galvanometer.
For the rectification of alternating current from a power cir cuit, the arrangement shown in fig. 4 is commonly employed.
With such an arrangement the alternating supply is first trans formed to a convenient voltage and then applied to two diodes, the filaments of which are connected together. With this arrange ment both halves of the alternating voltage cycle are used, the diodes passing current alternately. This circuit is used in many high-tension eliminators for use with broadcast receivers.
(b) The Three-Electrode Valve.—The introduction of a third electrode between the filament and anode was first made by Lee 'British Patent No. 2485o, Nov. 16, 1904, U.S.A. Patent No. 803684, April 59, 1905.
de Forest 2 in 1907. De Forest had previously, like Fleming, used a two-electrode tube for detecting wireless signals and had named such an instrument an audion. The same name was, however, retained for the three-electrode tube of de Forest in which a zig zag wire or grid was introduced between the filament and anode. Recently the name audion has been applied to a three-electrode tube.
Because of its closer proximity to the filament, the grid of a triode permits of a more effective control of tube current than does the anode. This control is effected by way of space-charge neutralization and this is determined quantitatively in terms of the positive charges on the grid and anode. For example, if the grid and anode are maintained at the same positive potential with respect to the filament the charge on the grid is greater than that on the anode. Actually it is m times as large where m is the amplification factor of the tube. For normal operating conditions the grid and anode potentials are not the same, but the effect is similar, for, if we increase the grid potential by say one volt the increased charge it acquires is m times as large as that acquired by the anode when its potential is similarly increased.
The electron current leaving the filament which, when the emission is adequate, is controlled by the electric field produced there by the grid and anode charges, passes out towards the grid and anode. In doing so the electrons acquire high speeds and shoot through the holes in the grid and are collected by the anode. Prac tically only the electrons which leave the filament opposite a grid wire are caught by the grid when positively charged. When the grid is at a negative potential it does not collect electrons at all. It is because the grid is efficient in attracting electrons, but in efficient in catching them that the current variations produced by changes of grid potential are registered mainly in the anode current and not in the grid current. If a triode were not evacuated but were filled with a gas at atmospheric pressure the electrons, mov ing comparatively slowly, would follow the lines of electric force and the thermionic current would be divided between the grid and anode in the ratio of their respective electric charges. For the case mentioned above in which grid and anode potentials were identical, the grid current would be m times the anode current.
If the filament, grid and anode are represented as conductors I, 2, 3 respectively the charge on the filament Qf may be written where the C's represent capacity coefficients and V, and the grid and anode potentials with respect to the filament which is assumed to be at zero potential. Since the electric field Ef at the filament is directly proportional to Qf we have The value of m depends only on the geometrical configuration of the electrodes and, for a case in which the filament, grid and anode are parallel planes and the grid consists of equidistant 'U.S.A. Patent No. 841387, Jan. 15, 1907.
that the grid potential is zero. The current flowing to the anode will then be represented by OG. If the grid potential is made increasingly positive the anode current is increased until the saturation value is reached at B. On the other hand when the grid potential is made increasingly negative with respect to the fila ment, more and more electrons are sent back into the filament and ultimately the anode current is reduced to zero, as at A. For these particular conditions the negative charge on the grid is equal and opposite to that on the anode, the numerical ratio of anode and grid potentials being equal to the amplification factor in.