Unstable bias on 12xxx at 90v

[JTM100]

I built a 12xxx with EH 6CA7's. The specs are pretty much the EVH/Friedman specs. If I run the amp at normal 110v, the bias is stable at 38ma idle 450 b+. However, when I lower the voltage to 90v with the variac and then crank the bias up to around 48ma at 370v B+. The bias drifts downward to about 20 ma with the bias trim pot all the way up. The bias circuit is standard 12xxx. The only thing is, in order to get enough range in the bias trim pot on 90v, I had to bring the 47k resistor that connects to the leg of the pot down to 5.6k. I didn't change anything else. I can put the 47k back in and raise the voltage back up to 110v and the bias will be stable. So, it is not liking something at 90v. Am I missing something here?

[Strat78]

What is the value of the trem pot? The b+ probably changed too when you adjusted the bias so you have to keep recalculating. There should be a happy medium resistor/trem pot combo that will be in range with a wider variety of wall voltages. Leave the 47k on the pot leg and adjust the other resistor next to the diode. Oh, are the EH 6ca7's burned in somewhat, I've found that they tend to drift quite a bit the first couple of days.

[JTM100]

The trem pot is 25k. The EH 6CA7's are fairly new but I have been blasting the amp for over week now. They did drift for first couple of days like you said but are stable now on 110v. So, you are saying to experiment with values of the trem pot and the 27k resistor connected to the diode?

[Strat78]

25k is good for the pot, and it's just easier to adjust the bias range from the bias supply dropping resistor position.

[Sparky4444]

...I am exactly as you...amp spec'ed exactly like yours, same tubes, adjusted the resistor like you did..

...and I see the drift as well...

...what I tried doing was to power up the amp at 120V, which will give a really hot bias...let it juice up for a bit...lower the voltage to 90-something, then bring it back up to around 95-100ish...seems to be a bit more stable and the bias settles around 40mA, if I remember correctly..

...but that is interesting you're experiencing the exact same as me...

[JTM100]

Yes, it is strange. Dave Friedman did say that he biased Ed's amp to 85ma and THEN dropped the voltage 90v. I just figured that I would bias to 50ma at 90v. I think I am going to try what Strat78 suggested and leave the 47k resistor and 25k pot in place and adjust the voltage by changing the value of the 27k resistor going from the bias tap into the diode. Earlier today I was reading here http://www.el34world.com/charts/Biascircuits.htm" onclick="window.open(this.href);return false; and although the values are different and this is for a Fender it looks like the 470 ohm resistor( Fender) is the bias range resistor and your sweep range on your bias pot can be adjusted by raising or lowering the value depending on what your needs are. It looks like a I need to raise the 27k(Marshall) to a higher value to get into the range that I want to be in.