VTA board bias slightly unstable

Hello - since my last post I've built and installed the VTA board (in an original ST-70)... and of course, it sounds fantastic.

My only problem right now is that the bias doesn't want to stay stable. It seems to be more unstable on the left channel, and goes to between .35 and up to .48 volts, and the right side seems to fluctuate between about .37 and about .42 volts. Each time I check it it seems to be different. I usually just check and don't change anything, let the amp play for another half hour, then check again, etc and it'll be off even more. Then I'll set the bias back to .4V on all tubes after having it fully warmed up, and it stays for a short while. Then after an hour or so, back off by a bit.

I can generally tell that it needs to be rebiased when it starts to lose its magic superfluous sound... my motto lately is "if it doesn't sound perfect, check the bias" and bam, usually that's the issue. Once I match them up again it really does come back into perfection zone and sounds great.

Any ideas or similar experiences? How stable is it supposed to be? Is this possibly just one bad tube ruining it for the rest? Thanks!

Thanks Bob. These are actually new JJ E34L tubes, probably only about 100 hours on them. I've heard they can be unreliable though. I'll switch sides and see if the issue follows, and go from there. Thanks again.

new production Mullard repros have a similar problem but much worse, they are unstable when used in circuits designed for old style EL34s.
If you are using 270K resistors from R29-R32 change them to 220K, or even lower to 180K or super low 150K.
This is also a definite change needed if you ever use KT88s or similar.
That will definitely require a re-bias but if it's a tube design problem like the repro Mullards have, that should cure the problem.
Also, if one tube has different characteristics than any of the others, the bias pot position will be hugely different than the others.
Standard JJ EL34s always bias exactly at the center position of the bias pot, other brands seem to vary depending on the batch.

Roy,

what modern output tubes do you recommend/like?

I've been running the Russian equivs and like them better than the Shugang EL34Bs.

The Russian tubes sound really nice, especially for the price.

Jason

I just started the st-70 thing a few months ago.

I found some cheap Valve Art EL34's and think they sound very nice. I just don't know about reliability.
My tube timer indicates 460 hours and they still bias perfectly.

I honestly like the sound better than the overly bright Mullards that came with it.YRMV.


Steve

My Shugang EL34B's are Valve Arts, and they are a great value too.

Bob Latino wrote:I made a post about these Shuguang made Valve Art EL34B's mentioned by j4570 and Westy56 on this forum about two years ago. Not only are these very good sounding tubes, they have proven to me to be very durable and reliable tubes. The "B" designation on the end denotes the wider tube glass which flares out a bit at the tube base before the straight sides go up. Link below on the Valve Art EL34B's from March of 2009.

Valve Art EL34B output tubes for ST-70 amps

Bob

The fit was so tight, one of my tubes broke. The VA's are brutal on the 70 sockets.


Steve

Bob-it was your post that got me to buy the Valve Arts. They were very inexpensive and I wasn't going to fire up a rebuilt amp with expensive tubes for sure.

No problems with the replacement sockets I got from Bob.

Jason

Roy, thanks for the advice re: R29-R32. Quick question - what is the effect of that change? Is it safe to do a modest reduction from 270k to say 220k, and what are the reasons behind that in the circuit? Could I still use standard EL34's in the same amp and bias them easily? I'm noticing the bias pots are nearly fully to the right in order to set the JJ E34L's at .4 V.

R29-32 are technically known as "bias feed resistors" not to be confused with the more commonly known grid resistor.
They act similar to more commonly known "pull-up" or pull-down" resistor in a digital circuit, and control the amount of bias current that can go to the screens, which controls the total plate current thru the tube.
Most older circuits and tubes using EL34s in push-pull used 220K-270K resistors, and KT88 amps use 100K-150K.
But for some reason SOME new tubes will go into "bias runaway" with the normal higher values, so those tubes should be run with lower valued "bias feed resistors".
They will affect the bias voltage also, so you may need to adjust the voltage divider that creates the bias voltage, normally changing R39 is the easiest way to change the range of the pot as it effects both channels. To get more range on the bias R39 can be made smaller than 10K, like maybe 2.2K
In regards to sockets and tubes fitting too tight, there are a few tubes where the pins are oversize, but another common problem is there are two types of octal sockets, and one type I do NOT like to use, has an opening that looks like -O- the other type of socket looks like ( ) and fits most tubes better.

Thanks for the thorough explanation, makes sense to me roughly. I might try a 220k instead of a 270 and see where that takes me.