For the Record: Transistor Replacement Problems
February 1966 Electronics World

February 1966 Electronics World

February 1966 Electronics World Cover - RF CafeTable of Contents

Wax nostalgic about and learn from the history of early electronics. See articles from Electronics World, published May 1959 - December 1971. All copyrights hereby acknowledged.

By 1966 the dominance of vacuum tubes, particularly in new products, as stated by Electronics World magazine editor William Stocklin, "...is gradually passing behind us." The Transistor Replacement Problems highlighted in this article refer primarily to finding second (or third or more) sources for transistor types already designed into existing products as semiconductor manufacturers move on to new and better (often much better) devices. There was little motivation for companies to dedicate scare resources to making low volumes of legacy devices just to service radios, televisions, phonographs, etc., which, to be honest, they probably preferred owners would abandon after breaking down and replace them with a newer model. Of course that mentality was nothing new since all manufacturers prefer a throw-away mindset since it serves two primary objectives: selling new products and disposing of products already in the field so as to not have to service them. Mr. Stocklin worried that the situation could get even more serious once integrated circuits entered widespread service.

For the Record: Transistor Replacement Problems

For the Record: Transistor Replacement Problems, February 1966 Electronics World - RF CafeWm. A. Stocklin, Editor

The era of the vacuum tube is gradually passing behind us. In its stead we now have solid-state components. The change was inevitable, and just as progress is impossible without change, so is change without problems. Design engineers, technicians, and manufacturers have been confronted with their own problems over the past few years. Now, with most of them solved, solid-state products are flowing off production lines by the thousands. Most efforts had been directed toward military, industrial, and commercial products, but since the beginning of this year we have all seen many more new types of consumer items that are designed around solid-state components. Almost all hi-fi systems now use solid-state components and many transistorized TV portables, CB, and ham equipment are being marketed.

Although theoretically transistors are supposed to last forever, they may be damaged by excessive heat, improper hook-up, or by the failure of some related component. Hence it is safe to assume that the consumer may be faced with the need for replacement some day.

It is obvious, when reviewing the schematic diagrams and parts lists of most consumer products, that many transistors are special devices available only from the manufacturer. This is a serious problem not only in increasing the cost of an item but in causing serious delays in servicing and maintenance.

Rest assured, though, that with few exceptions manufacturers themselves dislike the situation. To supply items for replacement is never a profitable enterprise for equipment manufacturers.

We are all familiar with tube replacement where one simply had to select a tube type from any manufacturer and get practically identical performance. Transistors, however, even with the same type number, may differ widely in performance.

It is common knowledge that semiconductor manufacturers, when producing a specific type of transistor, will obtain as many as half a dozen differently designated transistors with different characteristics from one run. Even after selection, each type has a wide range of limits.

This presents further problems to end users. In one particular case, H. H. Scott for example may order as many as 20,000 2N2926 transistors. They, in turn, test each one and segregate them into five groups; each is then color-coded with a dot signifying a different beta. Almost all transistors designed for high-quality or critical applications go through similar selection processes by the end user. Fisher Radio does the same thing and so do the TV manufacturers such as Zenith, Motorola, Emerson, etc. In view of this, it should be apparent that any direct replacement for such transistors must inevitably come from the manufacturer and not from the customary electronics parts dealer. This is extremely unfortunate for the consumer, the radio-TV service technician, and certainly for the parts dealer.

There are no obvious solutions at this point that seem promising. Companies like Scott, Fisher, Zenith, Admiral, G-E, and others have service centers across the country, each of which is completely stocked with special components that are therefore readily available at reasonable cost.

The independent TV service technician is certainly in trouble. He must, as soon as possible, set up liaison with the manufacturers so that there will be no delay in obtaining special components. He must work out his cost problem so bulk purchases at certain points will prove more economical. He must, under all circumstances, provide service within a reasonable time and at comparable cost. Otherwise, he will gradually give way to the captive service groups. Certainly, simple substitutions for portable transistor radio sets are in order since quality is the least important factor. But when one wants to get the utmost performance from any hi-f system or TV set, haphazard replacement should not be considered. The technician can't make a hit-and-miss substitution, though no doubt many will. A recent test group of transistors were replaced at random and a variation of as much a 20 db in gain was encountered.

Why then don't semiconductor manufacturers produce transistors with tighter limits? The answer is quite simple if one takes a look at any of the present industrial parts dealers' catalogues. We have more different types of solid-state components at the present time, in the short history of the industry, than we ever had over all the years of the vacuum tube. To ask for further selection would seem out of the question.

A solution that might have some merit would be for EIA to standardize on a specific beta test unit which could be used by every service organization in the country. If this were done, each manufacturer could indicate on his schematic diagram or parts list a standard transistor type number along with the specific gain requirements. Service technicians could then select as required.

We should all ask ourselves if the service industry is going in a direction similar to that of the white goods field. It might be. Certainly refrigerators, washing machines, dryers, etc. are made mostly of specialized components. Will it go further in that direction when we reach the integrated-circuit period? At the moment, it seems that these will definitely be extremely specialized components.

 

 

Posted July 21, 2022