|
While admitting that a drastic reduction in the number of valve types might impose hardship on a few wireless users with highly specialised requirements, the author urges that such a move would be to the ultimate good of the majority. Constructive suggestions as to how reductions may best be made are offered.
Marconi V24.
Each autumn, when The Wireless World Valve Data Supplement appears, the veteran reader can hardly fail to contrast it with the valve list of 1921, the eve of broadcasting. It was not a long list. It consisted of one entry the French R valve. Filament volts: about 4. Filament current: about 0.7 Amp. Amplification factor: not specified; but measurement showed it to be between 5 and 10. Mutual conductance: unheard of; but is found to have approximated to 0.3 mA/V.
There were a few other types, left over from the War, but they were not generally available, except, perhaps, the V24, with its curious shape, designed to reduce inter-electrode capacities. In any case, the general type of them all was the same as that of the R viz., battery triode of incredibly low efficiency.
In the intervening years not only has the efficiency of valves been doubled many times, but the variety has multiplied almost without limit. There is now a type of valve for every purpose. That in itself - as well as the vastly increased efficiency - may be welcomed as a product of progress. The comparison is like that between the business prospects of a raw and clumsy individual - a Jack of all trades - and a large staff of trained specialists. The results - the humble village shop and the mammoth department store - reflect the difference.
New Valves and More New Valves
But commerce came to a stage when it realised that mere multiplicity was no sign of efficiency, but rather the reverse; and there was much heart-searching when the emphasis was shifted to rationalisation, with the weeding out of the surplus employees and establishments.
To some extent that condition may be said to have already begun in the valve world. Two years ago the season brought forth such an amazing crop of new valve types that the makers realised that there was such a thing as excessive progress. In Germany they went so far as to proclaim a valve design holiday. Action was less drastic here, but there has been an appreciable slowing down in the issue of new valve types, and an increased withdrawal of old ones.
Our valve manufacturers looked with some envy at their American brethren, who had confined themselves to a minimum of types, retained year after year, even although conservative in characteristics. But, as this policy allowed them to be produced at a fraction of the cost of British types, who worried about having to use a few more of them to get the required results?
Quaintly enough, just as the wisdom of this seems to have been having some influence in our market, the Americans appear to have succumbed to an orgy of new designs.
Some of the advantages of standardising a minimum number of valve types have been hinted at. The number of each to be manufactured is correspondingly a maximum, and the factory can be organised to turn them out more economically and with less likelihood of troubles due to continually having to change round the machines and operatives to new specifications. Sales and service organisations are simplified - dealers are not forced to the alternative of stocking stupendous quantities of valves or of replying 'I can order it for you, Sir'. In brief, valves can be made cheaper, more reliable and easier to get. What more can one want?
Against this, there is the argument that such conservative policy deprives one of the advantages of frequent improvements in characteristics, and perhaps of some developments altogether. The ardent experimenter, keen on setting up new records of amplification per valve, is hit. The general public, who want trouble-free radio, and don't care whether the results are gained with one valve or twenty so long as the price is right, benefit.
Wasteful Multiplicity of Types
There are several ways in which standardisation can be tackled. One is to weed out types, or even whole classes, of valves that can no longer justify their existence. It is very nice for the designer to be able to pick exactly the specification he wants - 'three sleeve lengths to every collar size', as it were - out of a gargantuan list; but it is sheer laziness.
In this day of specialisation is there really any excuse for a multiplicity of general-purpose triodes? Most of the duties that once had perforce to be undertaken by them can now be performed far better by special types. The disadvantage of the triode is that it does not fully possess the chief attribute of the valve the one-way action that is so valuable in radio circuits; some of the output strays back to the input through the anode-to-grid capacity.
The screen grid valve was devised to eliminate this imperfection, or at least to reduce it to an inappreciable amount. The screen grid valve itself in turn was found to suffer from certain minor flaws as circuit design advanced, and was succeeded by the high-impedance pentode. It is still unexcelled, however, for low anode-grid capacity; in fact, the general run of them are better in this one respect than the latest pentodes. This is not necessarily so; it is a matter of a compromise to achieve the best all-round characteristics. In all other normal respects the pentode is as good or better. The pentode is useless as a dynatron, of course, but the very few who are interested in dynatrons might well be content to have to order them specially, if the masses thereby obtained the rationalisation benefits of sweeping the lists clear of screen grid valves.
The surviving HF pentodes could then be further reduced by a half. The number of occasions when a non-variable-μ valve is used and a variable-μ type could not be substituted with negligible discomfort is so small as hardly to justify their retention in the catalogue of standard valves.
Standardised Output Valve
The high-slope pentode (with AC and 'universal' heaters) has now emerged as the most useful type of output valve, and could well be standardised in this department. So far we are doing rather well; we have kept only two types of valve, for mains drive at least. The battery position is not so clearly defined; battery power being so costly it might be necessary to have two sizes of output pentode, one where economy is paramount and one where a greater output can be paid for. In addition, there are the special systems - QPP and Class B. Both of these, in varying subdivisions, have their advocates. It does look as if we are still waiting for the valve designers to settle the matter by producing a quiescent valve that is undeniably the best.
And that reminds us that we have not yet fully examined the case for and against the triode. There is still a function for it in a stage of amplification between the diode detector and output valve. The latter having already been settled as a high-slope pentode, requiring only 2 or 3 Volts grid swing to drive it fully, it is questionable whether there is any real need for LF amplification. The diode works best when handling a fair number of Volts. And this is needed for AVC anyway. Still, it might be a hardship to deprive designers of a source of moderate LF 'mag', particularly in short-wave or all-wave receivers, and for amplified AVC. And the same type of triode would do for Colebrook HF amplifiers. But there is a great difference between this solitary survivor (or, at most, two) and the host of miscellaneous triodes still listed.
In the battery range such a type is still more necessary; in addition to the foregoing uses it. is wanted as a driver. There is little technical justification for the triode as a detector in straight sets. The better the characteristics the more heavily it damps the preceding tuned circuit. The screened pentode cuts out this effect, and offers more amplification.
The triode is still unsurpassed as a high-quality output stage when one can afford to throw away most of the power supplied, particularly when it is used in push-pull. But this almost brings us into the realm of special sets; it seems quite probable that not more than one broadcast receiver in a hundred depends essentially on a triode output.
How about frequency changers? There are heptodes, octodes, triode-pentodes and triode-hexodes. Obviously, this department is not ripe for standardisation. But at least let us have an AVC grid base that matches that of the 'standard' HF pentode.
Rectifiers furnish an example of what has actually been done towards standardisation. There used to be an extremely miscellaneous lot. For several years now there have been three standard specifications. And instead of each manufacturer coming out with a different three, they have all agreed to keep to the same. It remains to be shown why this range is sometimes duplicated by offering each valve in directly and indirectly heated types. It is, of course, necessary to duplicate one of them to provide a 'universal' rectifier, with low current heater.
But, you say, why have separate AC and 'universal' types at all? If the universal lives up to its name, why not use it exclusively and abolish the separate AC types? This has been seriously considered, but the feeling is against it. For a number of reasons the running of heaters in series, for which the universal types are designed, is slightly less reliable and satisfactory than parallel connection. It is not very much less, but when only a very small proportion, of people are confined to DC mains - and that number is in process of disappearing altogether - it would be rather foolish to impose even a slight handicap on the vast majority. It would look particularly foolish to our descendants, when DC is no more. One might run universal valves in parallel in AC models; but should a thoroughly well established (and technically sound) standard be thrown over for the sake of a dwindling minority? The universal valves, then, are a temporary expedient. That is fortunate, for their universality was doomed from the start by the makers adopting diverse current ratings.
Scrap the Multiple Valves?
Electrode assembly of a multiple valve (triple diode triode).
No mention has been made of the numerous combination types - double-diode-triodes, double-diode-pentodes (HF and output), triple-diode-triodes, driver - Class B valves, and others. That is because it would be quite a good idea to make a clean sweep of the lot. A small separate diode, or multiple diode, is wanted in any case for certain purposes. What need, then, to add numerous further types of valve that do no more than those already available? They only complicate design and production and mean that if one part fails the whole must be renewed.
If the valve business were being started all over again, no doubt the socket arrangements would be made rather more consistent. Considering the many stages in the evolution of the valve, we might actually have done far worse. We have got over the side-terminal affair quite nicely. There is still a little trouble with valves that were originally fitted with five pins and then went on to seven. And it is generally held that the Americans were right when they made the top terminal the grid. This has been done with all the more recent types, so now it is only the HF pentodes (and their SG predecessors) that have their anodes brought out here. The grid is the most sensitive electrode, and the one to be kept apart from the others. Even some output pentodes have their grids at the top now. It helps in avoiding hum. And it may be very useful for television.
Lastly, the identification of valves: if the subject can be touched on without risk of apoplexy. If readers select any particular class of valve common to most manufacturers-say the AC triode with an amplification factor of about 30 - and run through the lists, making a note of the names under which it appears, further comment is needless. Some of the names have a hint of rational purpose in them; others have not. No intelligent being would suspect that they all denoted substantially the same article, with only the trifling difference of birthplace. To remember all the Q4s and AC/PG465-SS5 is like trying to memorise the London Telephone Directory. Would it not be delightful if, first, all but a handful of approved valve types were abolished or consigned to replacement on special lists; and, secondly, if each type were given a fixed name or number chosen intelligently to indicate its purpose, such name to be beyond the power of any individual valve manufacturer to 'improve'?
In October 1935 RCA introduced the metal valve Types that became the standard receiver types. See side menu for the link. - Ed.
| 