Dealing With Amplifier Distortion

Radio Record, Volume I, Issue 34, 9 March 1928, Page 12

Dealing With Amplifier Distortion

A Talk About the Causes

By

Megohm

ame, HVERAL, correspondents. i Gia! complain of ‘‘blasting’’ and distortion of the outHi put of their receivers, mei? situatel’ a considerable distance from 2YA. This is usually a matter between the last audio valve and the loudspeaker, and without actually hearing their performance it is difficult to say whether one or both 1s to blame, Every outfit has its limit as to'the amount of undistorted volume that it can deliver, and listéners very frequently like to obtain «the greatest amount of volume possible from their set. With no reserve of volume whatever, it is a sure thing that blasting will occur just as soon as a speaker gets "worked up," or a band or orchestra atrives at a doubleforte "passage. Distortion may be noticed just as much on high notes as ‘on low ones, and however low an impedance valve may be employed in the last stage, if volume is being- taken without any reserve, blasting or distortion is certain on passages above the average intensity, . |

If a cone speaker has a ‘"papery"’ sound, it is not the fault-of:the cone, but. a sure sign that the last valve is overloaded, and. the cone enthusiastically responds to the high notes introduced. by the overloaded valve. These high notes would be suppressed by a horn speaker. The installation of a power-valve will effect a cure. Our ears get very insensitive on the yery high and low notes, so that unless these are much amplified we do not appreciate them. It is the fault of most amplifying systems that they amplify the central portion of the musical scale quite well, but, according to the particular system employed, they fall off more or less badly on the extremes. And we must drag the loudspeaker into this argament too, for it will not always reproduce perfectly good extreme notes that are put out by the last valve. . Weakness of High Notes. VERY high musical notes are inclined to be weak when reproduced, partly because they lave not been am-_ plified to the same extent as the body of the piece, and partly because the speaker passes them on weakened, or even distorted. This action is noticeable on high violin or piano notes, and whistling. If your outfit reproduces these with volume somewhat proportioned to the whole, then that is good. But the insensitiveness of the ear also creeps in here, and although there may be fair amplification of these high notes, unless it is exaggerated we may only hear them weak-

ly, We distinguish one instrument from another by the overtones or harmonics produced with the fundamental or actual note, and when we get high in the scale we are unable to tell whether we can hear these harmonics or not, and that has a deal to do with the weakness of high notes, and the difficulty sometimes experienced of discriminating between high piano notes and those of a piccolo or a xylophone. Results on Low Notes, AY THIEN we come to the low notes conditions are very similar, though in some ways more difficult. As the ear loses its sensitivity on the low notes, it takes a tremendous amount of energy to make the ear appreciate the sound, and this energy causes a heavy tax on the capabilities of the amplifier and speaker, so that in descending we arrive at a point where it is decided that the ‘‘game is not worth the candle,’? or that it is not worth the expense and bulk of supplying an amplifier that will serve to make pro‘nerly audible any notes below this

point. Our amplifier and speaker, a cone, may be amplifying tolerably well down to 100 or even 80 cycles, but below that it will be attempting to amplify notes, but only producing muffled sound with distortion. ‘The broadcast station may settle the matter for us and decide to cut off at, say 60 evcles, so that if we can bring about some improvement in our lower note amplification, we shall get all there is to get, though in practice the transmitter will be likely to be putting out notes of a~ lower audio frequency than a receiver is likely to reproduce. And if any transmitter were changed to send out no frequencies below 100, probably 2 large number of listeners with reproducers weak on the low notes would notice an actual improvement in their reproduction,

Quality of Miisical Harmonics. Bet this cutting off of low frequencies must not be overdone at the transmitter, because although a horn speaker of a certain type may not be able to reproduce such soinds at their actual pitch, it can lead you to believe that it is doing so in this way, Musical sounds are rich i11 harmonics, as already mettioned, and these give the instrumental quality to the sound. ‘These harmonics, several in number, are all higher than the fundamental note. When a low note comes through a small horn speaker, the latter is probably quite unable to reproduce this note, with a frequency of, say, 60 cycles. If harmonies did not accompany the sound, it is quite likely |

that nothing of this note would be heard behind the melody. But the horn is able to mislead us by’ reproducing the harmonics down to the lowest one that it is capable of passing on, and quite likely the lowest one will be somewhat muffled or distorted, Listening to a horn speaker by itself, it is difficult for even a musical ear to decide just what lower notes are being suppressed, owing to the reproduction of their "ghosts" or harmonics, which may easily be considered to be the {undamentals. ‘Ihe only way to realise the difference of reproduction is to sit between an ordinary horn speaker and a cone that emphasises the lower notes. This has been done by the writer, and is a highly interesting study. Two Loudspeakers Used. FOR several years the Zenith Company of Chicago has supplied two loudspeakers, a cone and a horn, in all its cabinet sets, and some time ago, as already mentioned in this column, the writer adopted this idea. With the two speakers connectel in parallel, and with a gvod choke filter in the

circuit, the tendency for speaker overloading is very greatly reduced, and a much more faithful rendering of the studio items is obtained. But in this method there was still the attempt of the horn speaker to reproduce notes below its ability, and to distort them in the process. As the principle of utilising two speakers is for the horn to take care of the high frequencics and the cone the low ones, it was decided that any attempt on the part of the horn to deal with low frequencies must be suppressed, and this was ac-_ complished by placing a fixed condenser of, say, .005 or .0L capacity in series with one of the speaker leads. he result was a considerable improvement in tone, any tendency to "mushiness" on low notes being cut our. "Mhe exact value of condenser must

be found by trial, but if too small it) will cut out at too high a position onthe scale. It is not suggested that every horn speaker used with a cone shoul] be fitted with a series condenser, but the idea is put forward for the experimenter who is continually striving to improve the quality of musical reproduction and speech too, aud who is not prepared to sit back and say that his reproduction is ‘"perfect." It may seem a large outlay to many to run two speakers, but it goes a long way towards settling a vexed question. After all, those who have run one speaker for a time and increased the output of their set can help quality reproduction in a way that is probably impossible with any inter-

nal alteration of components by purchasing a speaker of the opposite type to the one they already possess. An output filter or transformer is of course essential, and after these improvements have been carried out, further refinements may be gradually carried out in the hook-up. A resistance unit is easily tried in place of a transformer. his idea of two speakers has already been dealt with by the writer, but is good enough to be repeated, The Effect of Power Valves. ps the matter of valves, all set- | owners have a free hand, so far as the last audio valve is concerned, with the reservation that where dry B_ batteries are employed the power-valve in the last stage will be limited to a reasonable plate current, in order not to put tov much of a drain on the battery, Where a B+ accumulator or

eliminator is in use the consideration of plate current is practically negligi+ ble. In such case, if the last andio valve is, say, a 112, and is overloaded, then the substitution of a 171 will effect a great improvement, but the appropriate grid-bias must be provided and kept near the correct voltage by occasionally testing. the dry C battery with a voltmeter. ‘I'hen, as a last resort, there is the 210 type of power-valve, which will handle big volume, but takes from 6 to 7} volts and 1.1 to 1.25 amps on the filament, and works best on a plate voltage of 200 to 400, taking up to 22 milliamps. This last-mentioned valve, it will be seen, is essentially one for operation from a B eliminator giving high voltage, and where a transformer winding is available for heating the filament, tlus saving the necessity for extra A battery voltage to run the valve at its best output.

Jt should be noted that no receiver giving loudspeaker volume, and employing a 201A tube as the last audio amplifier, is capable of quality reproduction, as there is sure to be over‘oading and distortion. | Receivers giv- | ing distorted output will usually be improved by changing the last valye for one capable of handling the volume without distortion. For a 201A a 112 may be substituted, for 112 a 171, and for 171 a 210, but the latter only if high plate voltage is available. At 185 volts on the plate the 112 requires the same grid Lias (9 volts) as the 201A, but takes 6 milliamps instead of 2.5 on the plate, ‘This may be economically run from a dry B battery of not too small dimensions, and the writer knows that some listeners are running 20 milliamp valves off dry batteries of ample proportions,

and, needless to say, are getting good results, thongh the batteries avill .not last the same time as when running a set with a total plate current of 8 to 10 milliamps. If you wish to step up from the 112, then the 171 is available. ‘This valve will take a plate voltage up to 180 volts, which requires the high grid bias of 403 volts, and passes 20 milliamps. At 185 volts on the plate the grid bias is 27 volts, and plate current 16 milliamps. ‘The filaments of the above -yalves all run off a 6-yolt accumulator, but the correct voltage at the filament terminals of the valve-holder is 5& yolts. ‘he UX1z0 is only a seini-power valve, and will not carry any great volume. ‘The filament requires a 4-volt battery, or 44 volts, 3 dry cells, for which use it is most suited. Plate voltage runs up to 135, 6.5 milliamps being passed,

Gradual Distortion May Creep In THE most troublesome form of distortion is that which gradually creeps in. When the set is first installed reproduction may he amazingly good, and yet quality will gradually fall off week by week in such'a way that it is not noticed by the listener, until it at last gets obtrusive and calls for attention, It is very easy to forget about amplifier valves, yet they may haye gone off to such an extent that their emission has fallen much he: low the point where they can carry the low audio frequencies without distortion. With an eliminator the deterioration of the rectifying valve may be reducing plate current, or a dry B battery may have parted with its quota of milliamps and have a residual voltage not worth consideration. ‘Then there is the grid-hias battery, so easily neglected, yet so important where quality is sought. It is important to tune in the dials to the exact maximum reading, and reduce volume by other means than detuning with the dials, which may easily introduce distortion. Dimming the R.I’. filaments is one of the best means of volume control, which has the advantage of preventing tiie detector from being overloaded. : MELERTS are many other sources of distortion, but those connected with the last valve and the loudspeaker are the most common, and are at the present time causing anxiety to a number of listeners, which is the reason for this article, which contains much that is not new, but will, it is hoped, be helpful to those suffering distorted reception. :