Value drift over time

I can hardly wait! I must have some dating back to before when I was 10 y.o too ;-)

Mike.

I have a few from the 1920s. Don't look a lot like modern Rs. Not sealed either.

NT

Then I shall keep you in suspense no longer.

Here's what I found from a random selection of old components I snipped out. Firstly, pretty much *all* the capacitors were fine. The 350VDC Hunts capacitors could easily have been new. An Erie plate ceramic of

27k became 38.6k

another 27k ----> 29k

10 ----> 10.7

4.1k ----> 5.2k

220k ----> 246k

400k ----> 509k

These were all marked with a silver tolerance band, so clearly Taylor back then at least not *that* bothered about accuracy.

Same type as this:

-- This message may be freely reproduced without limit or charge only via the Usenet protocol. Reproduction in whole or part through other protocols, whether for profit or not, is conditional upon a charge of GBP10.00 per reproduction. Publication in this manner via non-Usenet protocols constitutes acceptance of this condition.

Most Rs in valve kit are far from critical. 5% would have cost them more than 10%. 20% were more common.

NT

I just put it down to post-war lack of availability but your guess is as good as mine. My experience with valves is not that great. I'm really more of the germanium semiconductor era. ;-)

I believe the first computer I got my hands on (just to play with while I was an apprentice in the factory test bay) was made of those...

Mike.

I was in the elctronic surplus bisness from the 60's to the end of the century. Most people thought that Allen-Bradleys were the gold standard. One cusomer complained and we started checking samples. They were all out of tolerance. A-B's speck sheet specified how to measure. For a given resistance range you applied a specified voltage and measured the current.

As an aside: In the 50's I ran across some carbon comp. resistors that had been modified. Apparently the person was short of cash or in a hurry. The person took a CC and a triangular file and raised the CC to the value neede.

CP

We used to do the same sort of thing with xtals in the days when they were expensive and hard to come by.

-- This message may be freely reproduced without limit or charge only via the Usenet protocol. Reproduction in whole or part through other protocols, whether for profit or not, is conditional upon a charge of GBP10.00 per reproduction. Publication in this manner via non-Usenet protocols constitutes acceptance of this condition.

Couldn't do that with my oldest crystal, it's in a valve glass envelope. 5kHz IIRC.

NT

Except you didn't use a file. You used 600 grit silicon carbide paper and a piece of glass for a flat surface.

At the time, the "standard" crystal was the FT-243 which you could take apart to get to the crystal slab inside.

Yup, the principle is the same, though. For the final fine 'adjustment' we'd use Vim, which is a kitchen scouring powder in the UK and many times less aggressive than 600 grit.

Good to know, but the aging of composition resistors doesn't tell us much about carbon film resistors (the common low-spec type nowadays) or metal film (the common high-spec type) and manufacturer coatings and such are likely to be changing from year to year as well.

Probably, because conductive (metallic or semimetallic) items are positive valence, oxidation will raise resistance with time, for almost anything. How much time, is still a mystery (for almost anything we build today, at any rate).

There's too much chemistry involved to make a really good long-life high accuracy projection for most real components. Humidity, ozone, fungus, air pollution... so MANY variables.

You can eliminate all those with glass, vacuum & getter. Then you find one day that the getter is oxidised & the bulb contains hydrogen.

NT

I was under the impression that glass was impermeable even to omnipresent hydrogen. Or is there a path via where the base pins protrude?

Hydrogen atoms are really really small. Trying to keep hydrogen in or out is always problematic.

Also, a kilogram of hydrogen at a given pressure takes more space than any other gas.

Glasslinger did a mass spec analysis to discover that gassy valves contain hydrogen. I don't know whether that permeates through the glass (unlikely s ince most valves stay hard), leaks in through pin sealing defects or is the result of remaining water vapour reacting with the getter. Either way a ge tter that could capture it would be a good thing probably.

NT

In practice you get molecules on the two-fer principle which are much bigger. Helium is bad too because then the atoms come at you singly...

Mike.

No, those resistors have drifted since their original construction. People did have ohm-meters back then and would verify values on resistors particularly if they were colour deficient or colour blind as I did, testing Rs when building kits - red/green deficient vision. I still check almost all resistors I come across with a meter as I don't trust my green, dark red, and brown differentiation.

If they are in power circuits - plate or cathode or voltage dropping then those do drift as they were only compressed carbon and heat/cooling cycles would cause them to change value - usually upwards as you have discovered. I consider your values typical of tube gear that is 40 or more years old.

Like I said earlier you HAVE to verify all the resistors as they wander over time. Modern Rs are much more reliable when operated under their rated wattage.

John :-#)#