Tube matching & drift

[RupertB]

How close should a "matched" set of tubes be?

I was putting the finishing touches on a California Dreamer build (ax84) last night. When I was setting the bias I found that one of the EH 6V6s (sold as a matched set) was only passing 60-65% of the cathode current that the other one passed (12ma when the other was set at 20ma). I thought I had made a mistake with my wiring (I had, but it wasn't affecting the bias :p ), but the fault followed the tube when I switched them 'round. When I replaced them with the new JJ 6V6s, those biased up to within .2 ma of each other.

The EHs had seen 40-50 hours in my 5E3. I know that tubes can drift in value, but this seems like an extreme case.

Thoughts?

[Old Tele man]

rule of thumb:

20% = normal
10% = good
5% = very good
<3% = excellent
1% = perfect (but for how long?)

...just because they'r "matched" now is no guarantee they'll remain that way after break-in / use...tubes DO drift!

...Most people don't realize that identically "matched" tubes seldom stay that way! Their specs tend to "drift" (sometimes UP, sometimes DOWN) with usage and over time. Which means those "matched" tubes could either: (A) drift TOGETHER (get better); (B) drift SIMILARLY (stay same); (C) drift APART (get worse).

[John Phillips]

Check that it isn't a screen-grid problem. The screen current is included in the cathode current too (which is why I don't like to check bias by cathode current). Some of those Russian 6V6s seem to be prone to screen failure or excess current, especially at higher voltages. If it is that, you need to raise the value of the screen-grid resistors.

If it's not that, they probably aren't matched well enough to get great results - although they might sound fine, it's hard to tell...

[Blue Strat]

...and be sure that the screen grid resistors are of equal value.

The least "drifty" and most durable 6V6s I've ever seen are the Brimar 6V6GTYs. They seldom drift during 24 hour burn in.