Matching Tubes with an emissions tester?

[mosfetmaniac]

Greetings,

While on lookout for a couple of tubes I noticed one site that had a pair of "matched tubes", fairly steeply priced but a matched pair did appeal... until I read further.

There was a question from another party asking for the test results, something I was going to ask myself.

The answer was along the lines of " Tested on EICO 625"...

I don't own one but I thought they were only a basic emissions tester and didn't test transconductance or have I missed something?

Regards.
Dave.

[XTC]

Probably a question better posed in the audio section.

However, as far as I know, 'matched valves' are matched for gm (mA/V) and the EICO is a go/no go tester along the lines of the Mullard tester. Presumably the scale amounts to transconductance, but it could be highly non-linear.

AVO valve characteristic meters allow you to measure gm under a range of plausible conditions although I suspect that most matching doesn't involve more than setting the tester up as stated in the data book and checking that the gms are matched to within some specification, whatever that may be.

I recall that the QUADII literature specified valves to be matched 'within 25%' but didn't explain further. 25% sounds like a generous margin.


I've read that Tektronix had their own special test rigs for matched valves.

So, in short, were I looking to buy matched valves, which I'm not, especially were they used or NOS, I'd be looking for a bit more reassurance than tested on the EICO if the matching meant paying more money.

That said, if the EICO said they were both good and the needle was roughly in the same place for both, I'd guess the valves would be OK for all but the most discerning buyers.

Pete.

[G8HQP Dave]

You need to bear in mind that to some internet sellers 'matched pair' simply means that the Philips etch code shows that they were made in the same factory in the same week. I'm not sure if this is due to ignorance or dishonesty.

About the only place where a matched pair is really needed is a push-pull output with shared bias arrangements. Individual cathode bias or adjustable 'fixed' bias should cope unless the pair are wildly mismatched.

25% matching shows that Quad knew what they were doing. A random (i.e. unmatched) pair from the same factory might be up to 50% different in anode current or gm, but typically more like 20-30%. If you buy good NOS valves you are likely to be within Quad's requirements anyway.

[Oldtimer74]

I have done some matching of 6146 valves for parallel use in transmitters. The gm can vary enormously depending on the setting of the Vg1 and Ia. Rarely will two valves zero at exactly the same Vg1 or Ia on a valve tester but you can usually get quite close. I tend to set the Ia at the test value and vary the Vg1 to zero the meter, then see what gm you get. See what difference there is the the Vg1 between valves.Then do the same setting, the Vg1 and varying the Ia. You can get quite close, certainly within 25% I would say. I have managed 10% or better but it may depend on having a number of valves to have a go at.
Pat G3IKR

[XTC]

Interesting.

So Mosfetmaniac is being offered a pair of good NOS valves, allegedly matched on an EICO, but basically any pair of good NOS valves would serve perfectly well in most audio amps and any two single NOS valves would be likely be good enough to pass any reasonable test. That closely matching valves on a valve characteristic meter would yield better results in service, is a dubious proposition.

It's questionable whether QUAD bothered to match their O/P valves, or at any rate to match them other than crudely.

I think the offer should be treated as being offered two NOS valves and no more paid for the matching than being offered two different NOS valves.

A couple of years back I bought a couple of Selectron EL34s on ebay. I tested them on an AVO IV and they both came out at dead on 11mA/V, which I recall is the spec in the AVO data book.

I think it's easy to be conned into paying far over the odds for this matching business.

Pete.

[mosfetmaniac]

Greetings and thanks for the replies.

After reading them I realised that I really don't need perfectly matched tubes, but was considering paying more because they supposedly were, which is certainly helpful to the back pocket!

A gentleman up the road from me whom has been extremely generous with his time advising me how to restore tube radios kept telling me "as long as they work fine and are strong, what's the problem?", OK I thought but most radios dont exactly call for a pair of tubes the same, obviously I need to take that advice

Thanks again.

Dave