HIGH FREQUENCY AMPLIFIERS

Evening Post, Volume CVII, Issue 146, 21 June 1924, Page 22

HIGH FREQUENCY AMPLIFIERS

STABILISING THE CIRCUITS.

The campaign against radiating receiving sets, now in full swing in the United States, is producing a heavy crop of designs of circuits which will not radiate, and some of these are very interesting. .The prime- cause of radiation by a receiving set is, of course, self-oscillation of one of the circuits by the valve associated with that circuit. It only becomes a nuisance, however, when the aerial circuit is oscillated, because the amount of energy discharged from the other circuits is too small to be noticeable at any distance. Non-radiating sets (using the term to mean those which, cannot cause interference, however they are handled) may be divided into two classes —those in which no self-oscillation can be set up, and those in which it may occur, but without affecting the aerial circuit.

First in the non-self-oscillating group, of course, comes the crystal receiver, either with or without a Valve to amplify the audio-frequency output. This cannot in any circumstances radiate any locally generated energy, and for individual use by means of head telephones is admirable when strong local broadcasting is beins; received, while two stages of amplification will enable a loudspeaker to be used. It is, however, not suitable for long-distance reception. Next comes the valve detector without reaction. This gives rather better results than the crystal so far as volume of sound is concerned, but like the crystal, requires amplification for loud reproduction. The ordinary method of amplification is to use regeneration, which gives all the effect desired, but at once takes the receiver out of the non-radiating class. is, it converts the set into a possible generator of energy which may be, and very often is, discharged through the antenna and creates interference. The valve detector, without regeneration is more sensitive to weak signals than, most crystals.

I _ A most important group of non-radiat-incr valve circuits comprises those which use high-irormency amplification, though some of these may also be radiators. It is quite easy to use a valve, as an amplifier of the received waves before they reach the detector, and to arrange the circuit so that it will not oscillate. If, then, a valve is used as detector, this j may be made to oscillate for the purpose of finding distant stations, and its energy, if the circuit is well designed, will not pass into the aerial circuit. There is a large variety, of high-frequency amplifying circuits designed to prevent selfj oscillation, ranging from quite simple ar- | rangements to elaborate neutrodyne, super-heterodyne, and other complexities. I It was pointed out in these notes recently that there must be oscillating currents in wireless circuits, and that the trouble of interference is due, not to these, but to "self-oscillation." Circuits in which this occurs are liable to be overloaded \vith the oscillating currents, so that some of the energy is thrown off. If the aerial circuit is in such a condition, interference may be caused • but in certain circumstances a circuit may be oscillating violently, and yet be so situated or conditioned that the aerial circuit is not affected. This ide,i is important in connection with high • frequency (radio-frequency), ampfication, for the designer has the choice of two principles. He may either make his set non-self-oscillatory, or he may make it generate oscillations and take precautions to prevent the energy being fed into the aerial. AMPLIFIER A RELAY. Every radio receiver must have a. detector of some sort. If there is no high-frequency amplification, the detector is connected, directly or indirectly, to -the aerial circuit, in which oscillations are induced by the received waves. If a high-frequency amplifier is used, its function is to reproduce these oscillations in another circuit which is directly or indirectly connected to the detector. This means that the amplifying valve circuit is acting, as it were, as a proxy for the real transmitting station. It is producing a magnified copy of the incoming waves. Clearly it must not generate any waves independently, or j these will inevitably interfere with the received signals. But it is highly important that, if •possible, the grid circuit which feeds the detector shall be as free as possible to oscillate under the influence of the waves passed to it by the amplifier. This circuit (corresponding exactly to the grid circuit of a single-valve receiver in which, the detector is fed from the aerial circuit) should be of low resistance, and should be tuned to the incoming signals if the best response is desired. The result of 'keeping the resistance low is to improve the selectivity of the set. The effect of accurate tuning is to increase the voltage variations applied to the grid of the detector valve, and this effect is increased if the usual procedure of using high inductance and low capacity is followed.

A successful and very popular method of" applying high-frequency amplification embodying the oscillating detector gridcircuit is what is known as "tuned anode" or "tuned plate" coupling. The plate of the amplifier valve is connected to one end of an inductance coil, the other end of which is taken to the B battery. The plate end of the coil is connected to the grid of the detector valve through a small condenser; and the coil is shunted by a variable condenser which enables the circuit to be tuned. This gives a circuit which oscillates readily and applies the maximum voltage change to the detector grid. But it has a serious drawback. The high potentials set up by the oscillation of the .plate circuit are applied not only to the detector-grid but to the plate of the amplifier valve. Here they set up, by attraction and repulsion effects, corresponding changes of potential in the filament and grid of the amplifier valve, with the result that self-oscillation is very liable to occur. A common remedy for this defect is to introduce a potentiometer in the grid circuit of the first valve. The effect is to check the oscillations of that and so make it more easily controlled; but this is attained at the expense of the amplification and of the selectivity of the set. A NON-OSCILLATOR. A very simple arrangement will set over the difficulty without any drawbacks. It is only the grid-circuit of the valves which needs actually to oscillate to give maximum response. Hence, if the plate circuit of each valve can be made nonoscillating and the grid circuit oscillatory, the required condition is fulfilled. The detector grid circuit is therefore made of a coil and a condenser connected to filament and grid of the valve. The plate; of the amplifier valve, however, is connected to the B battery by a coil of four or five turns only, wound round the detector grid coil. This small coil carries the fluctuating plate current from the amplifier and passes the fluctuations to the detector grid circuit, which is ojjcilliiied correspondingly, the

magnitude of the oscillations being governed by the changing values of the amplified waves and by the exactness oi the tuning of the detector grid circuit. The detector plate circuit may be ma.de to react upon its grid circuit so as to make this circuit regenerative to any desired extent; and yet, if the amplifier and detector circuits are properly ai-ranged and spaced, the amplified valve will not oscillate, because its plate circuit is not in resonance with the waves, and no-high potential variations occur in the plate. There is no need to use a potentiometer on the grid circuit of the amplifier valve, which can be worked under the best conditions. The condition in the circuit described, in which the detector grid inductance has ten or more times as many turns of wire as the amplifier plate coil, is quite different from that of the tuned transformer, such as is used, for instance, in neutrodyne receivers, as well as other amplifiers. These transformers have primary coils with perhaps half as many turns as the secondary (grid) coil. The result is that oscillations in the grid coil are able to force the closelycoupled grid coil into oscillation, or, at all events, to induce considerable changes of potential in it, so that the plate circuit behaves as if it also is tuned, and self-oscillation occurs unless, as in the neutrodyne, special arrangements are made to prevent it.