Classic Warranties

[Alan Stepney]

Also, a lot of TV's were on rental, and there must have been a temptation for the engineer, when asked by a customer what the fault was, to reply" just a valve". Thus saving a long discourse on electronics.

With industrial equipment, reliability was an issue, but that was only "reliable" compared to domestic items.
If you bought a new piece of equipment and it lasted for 10 years, or more, that was all that was required as it would often be replaced by a newer better(?) model by then.

[Maarten]

I just read in a component news bulletin from the 1980's that Philips actually considered their tantalum capacitor range relatively unreliable for professional applications and that they solved this problem by developing several series of solid aluminium capacitors.

[emeritus]

The valve manufacurers themselves may be partly to blame for perceptions about valve (un)reliability, judging from the attached Mullard leaflet that exhorts the reader to get their valves tested every year. I think it dates from the early 1950's.

[Herald1360]

It's that fine line between needing to sell more valves and getting a reputation for unreliability. Perhaps that leaflet (see p4) was produced to help shift a few slow moving rimlock parts

The whole tone is a bit of an appeal to the vanity of the set owner I think.

[mark pirate]

Yes, I agree that it appears to be the case, but remember that many dealers would test the valves from any receiver in for repair, so as to make more profit from the customer, therefore adding to the perception that valves were unreliable!

The reality is that although there were valves that had short lives, or were just a bad design generally, most are very reliable and are still testing good many decades later.

Mark

[wireful3]

Thinking generally about warranties, it is good PR to offer them. It might also be inexpensive if the period is much shorter than the life the manufacturer expects. There is also the expectation of the customer to consider. Most would only think a radio had failed if it did not come on when switched on. Light bulbs are a good example, the life used to be specified by the time taken for output to fall to a specified level. I am sure most householders would only think it had failed if it had a broken filament and gave no light at all.

With valves it is similar, when I worked on analogue computers using quite a few valves, we reckoned that if valve survived a few tens of hours it would last for many more. At the time I think 1000 hrs was given by the makers. A 90 day warranty seems easy to meet at little cost. I wonder if any of our radio servicing members have recollection of how often they needed to use this warranty?

[Welsh Anorak]

Hi
I often wonder if the fact that valves plug in led to everyone thinking they were unreliable? After all, everyone and his uncle could change valves, and it really didn't matter all that much where they went because they fitted, didn't they? How many televisions and wirelesses do we find with 'back missing' after everyone has had a go. Luckily, sockets for ICs and transistors didn't really catch on in domestic equipment, and often caused more harm than good (think of the G11 frame chip). Nowadays it's unusual to find someone has already been at their flat-screen TV - thanks goodness.
I wonder if valves had needed a tool to remove them whenther they would have been deemed so unreliable.
Of course you can understand the valve changing - if a set came in with, say, reduced height you'd pop a valve in then, on finding it wasn't much better, change the resistor or capacitor that was to blame. Then do you leave the valve in or put the old one back? Most repairers would leave the new one in and know the job was less likely to bounce, or often simply forget - remember all those NOS valve boxes that actually had swaps inside them?
Glyn

[G6Tanuki]

Originally Posted by Welsh Anorak

Luckily, sockets for ICs and transistors didn't really catch on in domestic equipment, and often caused more harm than good (think of the G11 frame chip).

Quite a few bits of equipment up until the early-1960s used transistors that plugged into sockets! Whether this was because of unreliability issues or a conservative carried-over-from-valves design attitude I'm not sure.

It's interesting to look at the historic ways transistors were fitted: I've got some gear here where the [metal-sleeved, Mullard] transistors have their leads soldered to the PCB but the body of the transistor is physically attached to the circuit board using a very small metal 'Terry"-type spring-clip. They're switching transistors so it's not for heatsinking.

Other gear firmly tied the transistor [OC201-type] body to the PCB using a length of lacing-cord threaded through suitably-drilled holes, and I've also got some Redifon RF PCBs where each transistor is fitted into a rubber grommet in a hole drilled in the PCB.

All this makes me think there was an enduring belief that transistors were in need of mechanical support, if not always easy-replaceability.

[Radio Wrangler]

As another side effect of valves fitting into sockets, we have the fine wine brigade auditioning every possible variant and insisting only certain vintages are any good, and that it matters whether their getters were fired on the South side of the valley, or the North. Had the bottles been soldered in, I think we might not be lumbered with this.

David