When Replacement is Due
The attitudes of set-owners towards valves differ considerably. Some people suspect them of being at the bottom of any disease the receiver may develop, and allow them to be removed, like a bodily appendix, on the slightest professional encouragement. Others work them till they drop dead in their tracks, and complain bitterly if this occurs in the first three years or so.
Their similarity to electric lamps may be responsible for the prevalent idea that the time to replace is when they fail to 'light'. In my experience that hardly ever happens with AC heated valves, With battery valves it quite often happens, from three chief causes:- dropping on the floor, genuine old age, or accidental contact with the HT battery. However it occurs, or to whatever type, the need for replacement is too obvious for further comment except to call attention to the wisdom of making quite sure that apparent deadness of the valve is not due to failure of the heater current to reach it; in particular, bad socket contacts should be suspected.
There are a few other faults, such as grid touching cathode, that put a valve right out of action and leave no choice about replacement. But most valves deteriorate less drastically, and sometimes the balance between continuing to tolerate the symptoms and discontinuing to possess the price of a new valve is very even. it may not always be realised how far below normal performance valve deterioration has brought the set. Even the more apparent faults, such as rattles, may perhaps not be traced to valves, and the less obvious falling-off in range, volume or quality, may go unsuspected altogether. These represent two classes of trouble that include most non-fatal valve ailments noisiness and loss of emission.
Hums, Groans or Rattles
If a valve becomes noisy at all it is quite likely to do so at an early stage in its life; so much the better if it is within the guarantee period. Sometimes the noise is purely mechanical, produced by movement of one of its parts; if it buzzes continuously the source of the sound is fairly easy to locate, but if it rattles only when shaken up by the vibration from the loud speaker it may be very difficult to trace the trouble to valves at all, let alone to any individual valve. When the loose part controls the characteristics of the valve in any way the mechanical rattle is usually swallowed up in the much louder noise that it causes from the loud speaker. The variety of types of sound that can be produced in this way is remarkable, from shrill tinkles to deep groans or hums, and they may be continuous or intermittent. The latter are sometimes very elusive, especially when the effect is more of the nature of an obscure type of distortion than of a distinguishable noise. Then there are various crackles, rustlings and hissings due to such electrical faults as leakage between electrodes. It is also possible, and quite common, for the normal slight hum from a set to be increased to offensive proportions as a result of valve faults.
Now I am not going to be drawn into a lengthy treatise on fault location, that can be studied elsewhere, for the continuous types of noise are straightforward examples of it. As for the sound, excited noises, once they have been clearly spotted, and often only certain musical notes make them apparent, the procedure of tapping each valve in turn generally shows up the offender. But it should be noted that it may be necessary to make the test with a silent carrier wave tuned in.
Noisiness, even although it has its obscurer and more exasperating forms, is a positive sort of thing; and, as it takes away a lot of the pleasure of listening, the valve responsible should be replaced without delay. But loss of emission usually comes on so gradually that it is difficult to say when the results justify a new valve or valves. The set-owner himself may be the last to realise that there is anything wrong at all, though a quick change-over to the original standard would leave no doubt about it. The output valve is the most likely to deteriorate in this way, and the result is, of course, a reduction in un-distorted output. If it is far gone, there may be serious distortion at all volumes. It is possible for distortion to follow reduced emission in the other stages, but it is more usual for the first results to be reduced sensitivity, and perhaps total interruption over part of the tuning range (when the frequency-changer stops oscillating).
Now (except for the non-oscillation) without laboratory equipment it is very difficult to tie this down to any particular valve, or even to prove that it is a valve at all. Unless there are definite grounds for suspecting something else, however, it is fairly safe to start by checking the valves. The question is, how to check them. Assuming the possession of a multi-range test-set of reliable quality, it resolves itself into systematic use of this instrument (again, I refuse to sit down and write a book on this subject when there is a perfectly good one by Cocking); but as the majority even of The Wireless World readers probably are devoid of such equipment the problem remains.
Dealers Valve Testers
A few years ago it would just have had to keep on remaining, but in these enlightened days every radio dealer worthy of the name owns a valve tester or set analyser, and he has every reason to encourage people to come and check their valves on it. Although the indication of a valve tester may mean anything or nothing, the scope for faking is actually rather limited, for, although it is to the dealers advantage for valves brought in by the potential customer to register an uncompromising BAD when applied to the instrument, it is essential to his reputation that new valves off the shelf should escape this condemnation, or even the intermediate POOR, and come safely within the GOOD zone, Not all valve testers are marked in this outspoken fashion, though non-technical customers are suspicious (and perhaps rightly so) of the sort that reads only in figures and can the more easily be explained away by a resourceful salesman. But to those who know something about radio. (and who does not after reading 'Cathode Ray' weekly?) the figures are much more informative, always assuming that the instrument is a reliable one. Practically all such valve checkers read one of two things anode-current when certain voltages are applied or mutual conductance. The latter reading fails to distinguish between valves of very different types, but does give some idea of the condition of a valve, even without any other information. An anode-current reading is quite useless on its own -what is 'good' for one type of valve may be 'very bad' for another, but when compared with a table showing what the readings ought to be when tested under those particular conditions it is more informative than the mutual conductance test. Preferably both readings should be available; but instruments intended for counter demonstrations are not usually so technical as this. The knowing customer will ask the dealer to produce his service gear.
Having so far treated the dealer with un-concealed suspicion, I must now join him on his side of the counter to point out to the exacting customer that he must not be too strict in demanding that a new valve should give precisely the reading specified as normal. Valves are allowed to pass out from the manufacturer with readings differing as much as 30 per cent. or even 50 per cent from normal; and except for push-pull purposes there is seldom much point in insisting on limits of 5 or 10 per cent. Similarly, a used valve may depart quite largely from the normal for that type without necessarily being due for replacement. But, if the comparison is made with the readings given by that same individual valve when new, a change of as little as 20 per cent is likely to mean that it is no longer capable of doing its job properly. For that reason one is recommended to scratch a note of the readings on all valves when new. I have never actually done this myself, but I am sure it is excellent advice.
Of course, a reasonable amount of common sense must be exercised, so that readings are not taken one time with batteries that are well up and next time with them run down; nor must it be forgotten that when the readings are taken with the valves in position in the set any one reading depends on the other valves working normally.