Bias cap value change

[LarBal]

I've got an odd thing here with a 70s jtm 50 watter. Perhaps this is just a bit of info I never learned.
It;s got bias circuit fed of high voltage, with a 220K dropper, to a 2 cap, voltage divider with a pot. Pretty standard.
This was a 6550 amp and it had el34s that were biased around 80ma each when it came in! Sounded bad.
Anyways, owner wants 6550s. It's got a ppimv that was installed well, 500K dual gang with 220K safety resistors, so 150K ish value for grid leaks, which I believe is fine for 6550.

I'm monitoring strange behavior in the bias circuit, with or without tubes.
I've replaced both bias caps and tinkered with the 47K to get the 6550s in range, but I was testing different values of bias caps, and noticed that when I took out the cap at the junction of the 47K and 15K, the negative voltage at the grids swung wildly. Further, when I left it as a 1 cap supply, and tacked a 100uf across the lone 10uf, it also swung wildly, as the negative voltage decreased(got more neg). Like -50v to -120 v
Then I put back in the cap at the diode(100uf) making it a 2 cap supply again, the neg voltage also dropped 20v or so.



Is this normal behavior or do I need to look deeper

[Chuck H]

Those capacitors should be 10uf. Not 100uf. There's not much current through the circuit but I think it's possible the charge condition stability and longer charge time with the higher uf value could be problematic.

Also, are you being certain to connect those capacitors in circuit with correct polarity? Which would be + to ground rather than - to ground. If they've ever been connected incorrectly the capacitors may be compromised. Also, how long have you been storing the capacitors you put in there? I don't build often and about five years ago I tossed a bunch of electrolytic caps I'd had around for over twenty years and never used. I could never trust them for a build at that age. IMHO you should always buy electrolytic caps fresh for the task at hand. Unless you're a professional repair tech. Then you do need some inventory on hand. But age does need to be monitored for electrolytic caps.

[Helmholtz]

A simple drawing would tell more than 1000 words.

[LarBal]

My bad, I meant to attach the drawing in the original post.

Yes, all the caps are oriented correctly and are new. Yes, 10uf is the spec, I was curious about 100uf in there.

I think what I was experiencing was charge times for a bigger cap, i.e. 100u, but it took a few minutes, I figured it would be sseconds

[Helmholtz]

If both caps are 100µ it will take more than 1 minute to reach final voltage.
During this time power tubes will overdissipate.
These caps are low value for a reason.

[LarBal]

Great! Thank you

[mhuss]

The three series bias resistors should be 47k (pot end), 15k, and 120k for 6650 tubes