DISTORTION—Its Cause and Cure

Radio Record, Volume V, Issue 26, 8 January 1932, Page 16

DISTORTION—Its Cause and Cure

(By the Technical Editor) Week by week we are going ‘through the questions asked in the recently held radio serviceman’s examination. Last week we dealt with the aerial-Question 1, and this week we go on to Question 2. ‘ An owner complains that the loudspeaker of his set will not ‘andle reasonable volume without distorting. Explain how you vould go about curing the trouble. (a) In the case of a battery operated set with magnetic speaker. (b) In the case of a mains-operated set with dynamic speaker. a |

ISTORTION in its widest sense is radio itself, for it is inherent in every component, It is sometimes very difficult to trace and. often more difficult to cure. However, let us examine ; the question a little more closely and see exactly what is wanted, for one could write for the whole three hours on the subject. The owner complains that his loudspeaker will not handle sufficient volume without distorting. That is the usual complaint. To the average owner all distortion is in his speaker which, being the last link, is the one to be blamed. However, the radio serviceman must not take it for granted that the distortion is in the speaker, for in nine cases out of ten it will be elsewhere. So the first thing when going on a job such as this is to turn on the set and listen for yourself. Is the man expecting too much of his radio? Has he heard a new one that will handle far more than his older battery set. become dissatisfied, and blamed _ his speaker? We must take it for granted that he is not expecting too much. that there is real distortion. Types of Distortion. Now distortion can be of two kinds -wave form distortion and tone distortion. The former is of the type which shows itself in 4a rattle, blasting, etc.. when the volume is turned up. The latter manifests itself in an absence of certain notes in the musical scale. It may be the bass, the middle notes, or the treble. The question, however, says that he complains that it will not handle sufficient volume, therefore we can take it for granted that the trouble is wave form distortion. The Speaker. HE first duty of the serviceman is naturally to take a cursory glance at the speaker, making any adjustments possible. . If this does not effect a cure, as it will probably not do, then he must look elsewhere. An experienced serviceman will be able to locate

roughly from the sound in the speaker the source of the trouble. However, it must uct be taken for granted that the examiner knows anything about radio and we must detail exactly what to do. As it is rather difficult to describe the varying types of noises and how they represent distortion in different parts of the set, we shall make a systematic search, beginning of course with the batteries. The Batteries. URN the set on. Jeave it on for a few minutes, and then test every

battery commencing with the "A," which is tested both with the yoltmeter and hydrometer. The "B" battery is tested not only from a positive maximum to negative, but to each of the tappings, as it is possible that a high resistance cell may cause all the trouble. Likewise the "©" battery is measured. When we are taking the measurements we must notice particularly if all connections are correct. Trace these out along the battery cable to the set, checking particularly the polarity of the "©" battery. Often this has been replaced and either the batteries are connected together wrongly or the battery as A whole is reversed. Now, having checked the batteries, check the voltages at the valves themselves. If the serviceman is possessed of a high quality tester he need merely slip the plug into each socket in turn and run through the’ various measurements, plate current, plate voltage, filament voltage, filament current, etc., and he can soon see if a valve is operating under its correct

characteristics. This will often reveal trouble in nearby components. It will uso act as a test on the valves. Valves. HILE doing this note particularly that the correct valves are used in their right places. For instance, in a resistance capacity set you will not expect to find valves of a low impedance unless, of course, they were in the last stage. Similarly, in a transformer coupled set you would not expect to find a semi-power valve in the second last stage. Actually quite a large number of sets employ a yaive of this type in that socket, and @#listortion is "nused in a rather inte ing manner. This valve has aS plate current which, when fidwing through the primary of the preceding transformer, lowers its inductance, matching becomes impossible, and distortion results. The first audio valve should have an impedance of about 10,000 ohms. although if it is followed by resistance’ coupling, high frequency or resistance capacity type is necesgary, in which case great care must be taken to ensure correct operating conditions. The second Lf. valye should be of low impedance of approximately 6000 ohms. Where a heayy speaker is employed und more volume is. expected, the impedance should be even lower. Use of a Milliameter. We mentioned a moment ago that we should test each stage by a high quality commercial tester. Many servicemen cannot afford these rather expensive instruments, and must use the ordinary mil!’.eter. Turn the set off and place the milliameter in turn in the B+ lead to each valve, commencing with the detector. If we are to make an exacting test, a meter. with several scales will be required, because whereas we may only need a maximum of 5 mils for the detector, one with 20 may be needed for the power valve. When the meter is in the circuit it should, when the set is turned on and timed, remain fairly steady, or at least flicker not greater than 10 per cent. If, however, there is a kicking and it is occurring downward each time, there is too little grid bias, or too much "B" voltage for the-~pias used, and the remedy here is Increase grid bias or decrease the "B" until the kicking ceases or is failly equal in either direction. If, however, the milliameter kicks upward we hive

the reverse, too much grid bias, or too little "B," and here again the remedy is obvious. In recommending a power valve for the last stage, one must take into consideration the current it. will use. If the owner of the battery set does not have an eliminator, and a& means for charging an "A" battery, it would be unwise to try to sell him & power valve which will run his batteries down rapidly. -Far rather do the best one can for his set as it is, and tell him that he cannot handle much more volume without putting himself to a great expense. In Old Sets. "THE few points we have now gone over are the most common causes of distortion, particularly in factoryt receivers. However, in sets of ‘1: design and those made by home structors, there, are many other ports that can cause trouble. Probably transformers and grid leaks are the most common. Very many transformers are too small and, as explained before, have not the primary inductance and cannot match the valve preceding it; consequently distortion takes place. J urthermore, they cause tonal distortion by cutting off, usually, the low notes. The serviceman should make certain that good transformers are used in any set for which good quality is. expected. Look over the transformers to see if there are any loose bolts, as often a loose lamination will give rise to a rattling on very loud passages. In resistance capacity coupled sets, check over the values of the grid leaks and condensers. Here is the rule for values. The resistance in the anode circuit should be from two to three times the resistance of the valve impedance. The grid resistance should be four times the anode resistance and the coupling condenser can now be obtained from the table given herewith.

Defective Connections. HECK over the set for defective connections, particularly open grid circuits. Try reversing the windings of one or the other or of both transformers, if two are used, and see if this makes any difference. As a matter of fact, after resistance coupling the transformer will in very many cases have the secondary connected up in the reverse manner in order to get the best results. Detector Overloading. ISTORTION on the high frequency side of a set is not uncommon, and being mixed up with some form of instability, often gets through to the audio stages and causes distortion. Often distortion is caused in the radio amplifier by too great a signal being imposed on

the detector, causing ceteclor overload: ing. Detector overloading can be detected by an accentuation of sibilants and the upper frequencies. These disappear when the volume is reduced. The voltage on a grid leak detector should be from 20 to 50 volts. It needs to be fairly high in order to handle the vol. ume. Where an anode resistance is used in the detector circuit the voltage must be considerably higher in order to compensate for the loss through the resistance. Distortion due to the detector can

orten de 1aentined by the double hump in the tuning, that is louder signals are obtained on either side of the resonant point on the tuning dial. The cure in this case is either to make an extra aerial tapping so that the aerial can be brought in direct to the detector or use a smaller aerial. Of course one could ehange over to anode bend detection. but this is not always possible. In the case of anode bend detection check very carefully the grid bias. Try increasing it to see if any improvement is made. A choke which has a peak somewhere about on the. wave-length of the station to which is listening will also cause instability and so will battery coupling. Too high a value of grid leak will often cause distortion and a not infrequent cause is a free grid on the detector or screen grid valve. In the case of the detector this may lead to grid choking, or there may be sufiicient leakage in the valve holder, or in the grid condensers to prevent excessive choking, while in the case of the &.g. valve very often almost satisfactory results can be obtained even though the grid is free. Where loud signals are to be handled the value of the grid leak resistance should not be more than 2 megs. It could be even reduced to one for local station work. If h.f. cur-

rent is getting into the audio amplifier. a choke and condenser should be used in the detector circuit. The. choke should have a value of about 200,000 microhenries and low self capacity of three or four microfarads. The by-pass condenser should be of the capacity of .0003 mfds. . L.S, Coupling, APART from h.f. amplification . another fault may possibly appear, due to incorrect coupling between the last valve and the speaker. Tf an output transformer is used, check up the ratio. The ratio of the output transformer should be equal to the square root of twice the valve impedance divided by the speaker impedance or i valve impedanve X 2 R= v¥V speaker impedance Where push-pull valves are used one does not multiply the valve impedance by two. This, however, is not a common cause of distortion, because if put in commercially, they are usually correct. _ Speaker impedance should be that at 250 cycles. Often the inclusion of a filter will improve a set that is prone to distortion. (THE second part ‘of the question relates to distortion in an a.c, set using a dynamic speaker. In general what has been-laid down for the bat-tery-operated set holds good for the a.c. set. However, the procedure is slightly different and there are other complications which may possibly enter in. In the first place the set must be turned on and the type of distortion noted. Tune very carefully to a station and test for overload. It is possible that the owner of the set has not been tuning his set correctly. Particularly is this true for super-hetero-dyne receivers, which distort chronically if all the stages are not in resonunce and these tuned correctly to the incoming wave. By listening to the speaker it is usually easy to tell if the distortion is located either in the speaker itself ool Se

ov in the receiver. As the speaker is'more complicated than the magnetic cone speaker, more troubles are likely to centre in it, and more attention must be paid to it than in the case of the battery-operated set. Distortion in a speaker may be due to a loose centering device allowing the moving coil to rub against the pole faces when heavy volume is. being handled. Very often by moving the cone of the speaker this rub can be felt and so cured. It may, of course, be necessary to. take the speaker away and make the adjustment on the work bench. If the centring device ap"pears to be quite in order, test the pot for magnetism, which test can best be carried out by holding a_ screwdriver or some metal object against the centre piece, which should, strongly attract the metal. If this is weak, the trouble is in the field supply. This may be of two types. The pot winding may be incorporated in the set as a choke or may be supplied by a separate rectifier. In the latter case (Concluded on page 26.)

Ee3iInnHniLiazhillptEoIIcJpLHIIIILiVIIC3IIIILIIHIEJHIIIIIIE Distortion (Continued from page 17.)

test the rectifier and the connections associated with it. In the case of the former, many factors may enterin. If used as a choke it is possible that there is a breakdown actually in the speaker itself and this can usually be determined by a continuity test. . The service manuals accompanying a.c. sets usually specify the resistance of the dynamic speaker field and this should be checked up with a ohmmeter wl‘ch, by the way, every serviceman should possess. If this is in order and the magnetism is weak, then look to the radio set on the other side of the choke. An open circuit of any description in. the previous portions of the set associated with the power supply would diminish the amount of current flowing through the circuit and so weaken the magnetism in the-speaker. This entails a systematic search of the receiver which can best be done by testing the valves, which is the next step in any ease in our search. Testing the Valves. IRST test the power-valve, not only for operating voltages, but for balance. Power-valves must be balanced to within five or ten per cent. if quality is desired. By balancing is meant that the plate current taken by the two must be the same. From the power-valve go back right through the set testing all valves, not only for weakness in emission, but also for an internal short circuit. The latter, however, more usually causes hum.

UUOLAGUUHASESUAAUNALATASSESHALUITSTA ULES AAHATANNNC SEAHDEUASEOO ESTA FASAALEATER Other Tests. HAVING been assured that all the voltages and valves are correct, test chokes, transformers, ete. for partial short circuit, mainly in the secondaries of transformers. Particularly would a break-down in the secondary of a pushpull transformer cause the trouble, If a gramophone pickup is available, or if the set is one of the combination types, try the audio side, and if this is in order, as it probably is if all the tests have been made and a negative result obtained, the trouble is obviously in the radio section, and in all probability is due to the condensers being out of alignment. Particularly in super heterodynes must the stages be balanced up and generally checked over. An open circuit in one of the grid returns of the hf. valves will also cause distortion, as would a break-down in the bias re sister on any valve. See that the volume control is working properly, for quite often the volume control introduces distortion through one of seyeral channels, depending upon the nature of the trouble. Look over the set for anything that would cause loss of volume, for if the set is not as sensitive as it should be it is quite likely that the man is trying to force his set, causing overloading and so distortion. Lastly, check the output transformer for correct ratio. If the set is a commercially made one it is most unlikely that this will be wrong, unless some defect has taken place in the transformer itself, which is unlikely. Summary. GUMMING up, we would expect to find the trouble in the speaker it-

self, or in the circuit directly associated with it. In the pushpull valves, or in their circuit, particularly in the grid returns and the resistances and condensers associated with them. We have not taken into consideration the inadequacy of design or the faults in design, for it is presumed that a commercial A.C, set with a dynamic speaker is quite satisfactory when it is installed. With the battery set, however, the case is somewhat different, because some of them are quite old, and standards in those days were very much’ different from what they are now, and the radio serviceman would have to make some alterations in design in order to give complete satisfaction to his client.

Grid leak. Condenser. (megs. ) (mfds. ) © 0.5 .02 : 1.0 01 2.0 005 3.0 » 00385 5.0 .002