Transformer-Coupled Audio Amplifiers

Radio Record, Volume I, Issue 49, 22 June 1928, Page 14

Transformer-Coupled Audio Amplifiers

POINTS OF VALUE

HERE two or more transformercoupled amplifying stages. are used trouble may be experienced with the set howling at a high-pitched frequency. This is most common when the A.F. amplifer is preceded by a detector valve in which the grid condenser and leak method of rectification is employed. The grid connections of such a detector are very sensitive to stray electrostatic pick-up, and in particular are likely to be affected by the stray field set up by the large A.I’. potentials generated in the last stage of the amplifier! this provides enough reaction to maintain the whole amplifier in a state of A.I. oscillation, This pick-up by the gridleak detector is the seat of a number of ‘troubles for which the succeeding A.F. stages are apt to be unjustly blamed. The remedy is to arrange the detector layout so that the length of connecting wire between the grid condenser and the grid of the detector valve is an absolute minimum. Also arrange things so that nothing sarrying amplified A.I’. currents comes within several inches of the gril connections of the detector yalve. This form of instability is entirely absent from receiyers employing anode bend detection. GETS are sometimes described in which the detector unit is completely screened by 2 metal compartment of its own, This is an excellent plan, as it materially assists both ALF. and RE. stability. Sometimes, of course, the trouble does actually arise in the amplifier itself. The simple expedient of reyversing the leads to one of the intervalve transformer primaries is often resorted to. This is not, however, invariably effective. especially when there is electrostatic feed-back from tes output to the input of the aimplifier. BREAK-DOWN OF WINDINGS. i OISTURE getting fluto the windings of audio transformers is a frequent cause of what is usually classed as a "burn-out." but is really a corrosion of the wire by electrolysis made possible by the dampness, i ANY manriacturers now adopt yarious inethods of impregnating the windings with moisture-proof compositions, including waxes of various sorts, and there is no doubt that in the better-class transformer moisture is offectually excluded. In a new transformer some very special process is used in which the whole transformer, including the stallory laminations and Jeading-out wires all complete, is absolutely sealed in a special prepara- tion which, as the description says, "seals them for ever against the inroads of moisture, rust, and electrolysis." As a test, instruments which were immersed in water for three

months were afterwards just dried om the exterior, and were then found work as well as those which had not been so treated. TESTING TRANSFORMERS FOR CONTINUITY. "TESTING the windings of transformers for continuity is a simple matter, and should be carried out on a new. transformer when purchased, A single dry-cell, which may be one that has done service, is connected to one end of a pair of ’phones. To the other side of the cell a length of wire is attached, Place the ’phones on the head, touch one end of the primary winding with the wire from the cell, and the other end with the free ’phone tip. A good click should be heard in the ’phones, almost equal to that when the connection is made without including the primary winding. If only a very faint jick is heard, the winding is not continuous. The secondary winding ig tested in the same way. High volt ages ‘must not be used for this test PLATE CURRENT FEED RESISTANCES. OM15 thne ago in this column w: mentioned a system of B accumulator plate current to severa valves by adopting the method of plac: ing a resistance in each common plate lend in order to cut down the voltagé to the required value, thus doing awey with battery tappings and evenly exhausting the whole accumulator. This was taken from an American source, but it appears that this method is adyocated by Ferranti's chief radio engineer, Mr. A. Hall, A.R.C.S., who has been making searching investigation into the matter of preventing oscillation in audio circuits. It has been shown that us a result of feed-back when using two stages transformer coupled with the same B battery feeding both stages, that the overall amplification is by no means the product of the two stages when measured separately, and that in fact this ideal condition could only be obtained by supplying each (audio) plate from a separate B battery, or could nearly be obtained with a separate B battery for the detector. AY R. FLALL found that a low-fre-quency choke of high inductance could be placed in each plate lead and would effectively prevent any backcoupling, provided that the battery end of the choke be connected to earth pos tential through a condenser of 4 mfd. capacity in the same way as obtains across separate output resistances of an eliminator. For the benefit of new readers it may here be explained that back-coupling mentioned above is the action or inter-action of one yalye upon another throngh the medium of the 3 battery. (Continued Next Page.)

Construction | (Continued from page 14. )

. It was then found that instead of chokes wire-bound resistances could be used. These are much cheaper than chokes, and fulfil not only the same function of preventing back-coupling and giving results equivalent to a separate battery for each yalve, but also provide a means of eyenly exhausting the whole B accumulator. And it is. claimed that with this system "motorboating" cannot oecur, and the amplification is as great as in the case of feedihg each plate from a separate B battery. | VERY receiver with audio stages: fed from the same B battery is liable to produce a change in wayeform and amplification characteristic due to interaction, of which interesting curves have been published taken by means of the oscillograph, a piece of arvaratus that traces out and records the form of either high or low-fre-quency waves, One of these curves shows the audio wave of high amplitude, and running through the centre is a high-frequency ripple that should not be there, and which would cause 2 shrill whistle in the loud-speaker. OW-FREQUENCY oscillation is 2 fault that is difficult to remedy, especially as it is sometimes evident as an oscillation above audibility which creates an unpleasant form of distortion. The principle: of interposing suitable resistances in each plate Iead is of some importance, and when provperly appreciated is likely to become widely adopted. A variable resistance included in series will allow of fine adjustment, much as is the practice upon B eliminators in which there is not a eontinuous resistance (potential divider) with tappings. extending from positive to negative. R.F. VERSUS AF. (QUITE likely a large number of listeners haye an idea that audio amplification gives a greater increase than radio amplification per stage. But this conception probably arises from the fact that almost every receiver incorporates one or two stages of audio amplification. This, for the reason that audio is so easily added, and its effect assured with minimum trouble. Nevertheless, it is a fact that R.F, amplification gives the greater increase in signals. The power into the loudspeaker is proportional to the square of the yoltage on the grid of the power tube and the output

of the detector is proportional to the | square of the voltage on its grid. Voltage on the grid, it should be explained, means signal voltage only, and is in no way connected with battery yoltage. , HEN the A.F. amplification is multiplied by 10, for example, the power into the loudspeaker is 100 times greater, but when the R.F. amplification is multiplied by 10, the detector output is 100 times greater and the power into the loudspeaker is 10,000 times greater. The following table shows the comparative increase produced by the two methods of amplification :-

audio side of a receiver forms a handy means of building up large output after detection, so that the detector is not called upon to carry heavy volume as would be the case with only one or no audio stages. A single audio stage is sometimes adopted in order to avoid troublesome, unwanted interstage couplings causing loss of quality. But in spite of these figures, the | IN CONCLUSION. T is scarcely safe to mention "distortion" nowadays, for after an article such as the above has appeared, one or two letters are sure to arrive from individuals who erroneously take it as an attempt to lay the blame for all distortion upon every receiver. ‘Such an idea is far from the intention of the writer at all times, and it is his firm resolve to continue with these artides so long as they appear to be helpful to a majority of constructors. One individual classed them as contentious matter that should be excluded from the paper! In constructional writing the question of how the signals arrive at the aerial, and the condition in which they arrive has little bearing on the case. The constructor builds and improves his receiver so that signals will be as efficiently handled as possible, so that the quality of good signals will be preserved, and the bad quility of others’ not made any worse than necessary.

Added Increase In Speaker. Auplification R.F. AF. 2 16 4 5 623 25 10 10,000 100 20 160,000 400 50 6,250,000 2500