Three Electrode Vacuum Tube Oscillator.— As previously intimated the De Forest three electrode vacuum tube is now used as a source of high frequency sustained oscillations, and many different arrangements of circuits have been designed to utilize this remarkable feature of these tubes, now in practice f re qucntly designated vacuum tubes or VT's. One such arrangement of circuits to obtain sustained oscillations is shown in simple form in Fig. 4, termed a vacuum tube oscillator. It illustrates the tfeed back') or aregenerativet coupling of circuits, perhaps first devised by Armstrong. F is the usual filament, G is the grid, P is the plate of the tube proper. HB is the heating or lighting battery. r, r' are re generative or feed-back timing coils. C is a condenser across the filament grid-circuit. Closing the circuit or jarring the tube will cause an initial impulse or oscillation in the grid-plate circuit, including coil r. This oscillation in turn reacts on the grid-filament and r circuit, which again reacts on the former circuit, thereby producing a feed-back or re generative oscillating current in the tube that may be termed self-perpetuating. After a brief interval a steady sustained oscillation is pro duced. The coils r, r' and condenser C are adjusted to oscillate at a desired frequency in the usual way. By means of this tube oscilla tions from one-half cycle to 20,000,000 per sec ond are now obtainable.
It has been noted by another writer in ex pounding the oscillating feature of these tubes that uin order to set tube circuits into a state of radio frequency oscillation it is necessary that the connections be so made that the grid end of the grid inductance will be alternately negative and positive, as the plate end of the plate inductance is positive and negative. When the grid and plate radio frequency cir cuits are coupled with the proper phase relation any variation in voltage in either the grid or plate circuits will cause minute disturbance in the oscillation circuits, setting them into oscil lation at whatever frequency they may happen to be adjusted to. For example, a slight varia tion of voltage in the plate circuit by any means whatever will cause its resonant circuit to oscillate at radio frequency, and the resultant current will act upon the grid circuit, setting it into action at the same frequency. The result ing radio frequency fluctuations of the grid potential will act upon the plate at the right time to keep the plate resonance circuit in a state of oscillation, and this state of affairs will continue so long as the proper supply of voltage and filament current is maintained, but not otherwise. The tube is able to generate alternating currents because of its amplifying properties. The energy delivered to the gnd circuit in accordance with the actions just out lined will gradually increase in value until a maximum is reached, which is the maximum output the valve is capable of delivering.0
Consult The Wireless Age, June 1919, p. 15, article on "Wireless Telephone Transmitter for Seaplanes." In general De Forest gives the explanation of the vacuum tube oscillator as follows: ((There is only one oscillating circuit. This circuit is such that a sudden change of poten tial impressed on the plate produces in turn a change in the potential impressed on the grid of such a character as to produce in its turn an opposite change of value of potential on the plate. Thus the to and fro action is reciprocal and self-sustaining.* In some respects the operation of the vacuum tube oscillator is perhaps analogous to the oscillator previously described. Another example of a self-perpetuating oscillation cir cuit is that due to placing a telephone receiver before the mouthpiece of a telephone trans mitter when a shrill whistle or howling is often established. This phenomenon is especially no ticeable in the transmitter and receiver of the acousticon type. This effect is seemingly due to a series of reactions somewhat similar to, but much more complex than those that occur in an electric door bell, another device which self-perpetuates its own vibration; when its circuit is closed by the push-button. In the case of the telephone howler let us assume in explanation of its action that a movement of the diaphragm of the receiver toward its tnag net tends to weaken the air' pressure on the diaphragm of the transmitter, and hence weak ens the pressure of the carbon of the trans mitter. This causes a weakening of the cur rent allowing the diaphragm of the receiver to fall away, with the further result that the air column is compressed, this tending to increase ttga the pressure on the carbon a in and also in creasing the current stren in the circuit, whereby the diaphragm of t e receiver is again attracted, which actions are repeated over and over. Investigation of this phenomenon indi cates, as might be expected, that it is dependent upon the fundamental set of vibration of the receiver and transmitter, the length of the air column enclosed between them and also the oscillation period of this circuit. The above references to the attraction of the diaphragm of the receiver and to its falling away are perhaps rather broad terms, when it is con sidered that as near as can be calculated the amplitude of vibration of the said diaphragm in reproducing speech is about the one twenty millionth of an inch. While this phenomenon has no direct bearing on the vacuum tube oscil lator it is interesting as somewhat analogous thereto and hence may tend to simplify the explanation of the device.