Possible Troubles in Audio Amplifiers

Radio Record, Volume I, Issue 45, 25 May 1928, Page 13

Possible Troubles in Audio Amplifiers

HOW SOME OF THEM MAY BE AVOIDED

T igs a comparatively easy matter for a constructor to build an audio amplifier from specifications and produce a piece of apparatus that will function weil from the start and give good quality reproduction. Yet there are unavoidable circumstances that cause components to be used other than those specified, and certain license may be taken by the constructor in various ways, entailing slight modifications of various kinds. In such cases there is always the chance of introducing instability, which is invariably due to some form of ‘"feedback" or stray coupling, and the greater the amount of amplification, the greater is the danger of audio howl being produced. One of the most frequent causes of oscillation in audio amplifiers is the high resistance of old or small dry cells, or in some cases an eliminator, supplying B current. This high resistance has the effect of forcing, or "feeding-back," an additional positive potential upon the plate of each valve preceding the last. This is assuming that all plates are fed from the same B supply. This effect, in turn, has an opposing action upon the grid voltages caused by the shifting of phases, with the result that there is a tendency to reduce volume of signals. In some circumstances there is the opposite effect of the signals being strengthened by the feed-back impulses, sufficiently strong, in some cases, as to cause oscillation. If each erid is in phase with the preceding plate and out of phase with the suc-

ceeding one, then in an, odd number of stages the signal will be weakened, and in an even number of stages the signal will be strengthened, introducing the tendency to oscillation. In some cases where R.F. amplification is high, and is coupled to the audio side through an inefficient coupling such as a small capacity econdenser, then it may be necessary to include an R.I. stage in the number of stages, as mentioned above. Transformer Coupling. HE above remarks apply more particularly to resistance and choke coupling. With transformer coupling it is possible to make the grid opposite in phase to the preceding plate by reyersing the connections to one winding, but this further complicates matters because the capacity between the windings then acts so as to reduce amplification. Mention was made above of phases. These concern the grid voltages only, which are alternately negative and positive, and are quite independent of the direct B battery current. Certain conditions can cause a "lag" in the phases, so that when this is fed back to meet the original impulses, if the lag results in a total reversal, then the positive phase of the fed-back voltage meets the nvegative phase of the original, and vice-versa. This is a brief idea of the complicated questions inyolved in dealing with phase combinations.

.Merits of "Push-pull" Amplifier. . HERE it is desired to use a common plate supply and yet be free from any -tendency to audio oscillation or instability, the push-pull type of amplifier should be used, as in this method of amplification the signal current does not traverse the battery (or eliminator) at all. From the above remarks the conclusion may be drawn that it is best to use 2 low resistance B supply in order to avoid any possible trouble from interaction between stages, or different portions of the circuit, such as would cause audio howling or oscillation. As dry batteries grow old their internal resistance increases and their voltage drops, so that after a certain drop in yoltage it does not pay to keep them in use if quality reception is desired, partly on account of the noise much resembling static, which accompanies reception when a dry battery is becoming "bone dry." "Motor Boating." "4 OTHING said above has any connection with another audio trouble known as "motor boating." Howling is not motor boating. The latter is more. often experienced in resistance-coupled amplifiers, and is easily recognised by the comparatively slow put-put-put sound. It is a limiting or blocking action, caused by the coupling condensers storing up electrons on the grid side faster than the grid leak can carry them off to the negative side of the filament. The grid of the succeeding tube gradually assumes a more negative bias until it finally becomes so negative that no plate current flows in the output circuit of that valve, which is then said to be "blocked." Then the grid leak passes the accumtulated electrons, and the current is restored, only to continue to repeat the same action. One cure of this trouble is to use an impedance in place of the grid leak, as owing to its comparatively low resistance the electrons will flow away as fast as they arrive instead of accumulating. This makes for quality also, as any tendency | to "motor boating’ is accompanied by distortion, due to the grid charge being at orjie of the extremes, or off that portion of the curve which allows of even amplification. Another improvement, if not incorporated already, is to place condensers of 1 mfd. each across each amplifier B voltage, whether from battery or elim-

inator, and these condensers should be placed within the amplifier. If such condensers are already installed, an increase of their capacity is recommended. Another method of overcoming the trouble is the staggering of resistance values and also the staggering of plate voltage values, Replacing the grid leak in the amplifier by one of lower

value will tend to stabilise the circuit, . but this is only done at the expense of { volume. F Care must be taken in placing the transformers in an audio amplifier, ag ! interaction between these when placed ; 4 too "lose together may cause howling; | especially if the transformers are of, | an unshielded pattern.