Single power tube oscillation (???)

[EFK]

It's been a while since, in my own amateur way, I've had a problem that has really stumped me.

I threw together a 2 X 6L6 head chassis, physical layout much like a JTM. However, I wanted to experiment with class A push pull idling at 100% so this is a JTM45 PT into a big honking choke input (5H, about 75 ohm). Here is my problem: one of the power tubes *I think* is oscillating, as it is pulling a lot of current and if I leave the amp on it will begin to red plate. Plate and screens are about 10V lower than the other tube. I have swapped tubes around (even using extra pairs) - no difference. I changed out that specific socket - no difference. Swapping plate leads - no difference (problem stays with that tube location, the one closest to the rectifier). BTW, I will be using a little NFB but it is not connected as yet. I have systematically pulled ALL the preamp tubes, no difference - it still does this with all preamp tubes pulled. So, the problem seems to be somewhere in the power amp. I have repeatedly gone over wiring layout, chopsticked stuff to death, soldered and resoldered etc. My power supply is 350/0/350 PT, GZ34 rectifier, 5H 75 ohm choke, 50 uf cap to OT & plate leads, 1K resistor, 50 uf cap to screens w/ 470 ohm R on each tube, then power onwards to the PI and preamp. Power tubes are cathode biased with shared cathode resistor.

Are there any glaring things I should be looking for? Any help appreciated!

[g1]

Have you checked the grid voltages to be sure there is no leakage from the coupling caps?
Other than that, your suspicion of oscillation is a good one. Introduce some NFB and see if it helps.

[imaradiostar]

Try splitting the shared cathode resistor into two at twice the value. I've seen the DCR of an output transformer cause a substantial difference in plate current. This could be what you're up against.

jamie

[Gtr_tech]

Or the pair is just so mismatched that one ain't happy with its idle current. Try a different pair of output tubes yet?

[Steve Conner]

Do you have grid stopper resistors on the power tubes? Something like 1.5k. They have to be mounted right on the grid pins of the tube sockets.

[Gtr_tech]

Yea....grid stoppers are always a good idea...as are screen resistors. If you copied a JTM45 directly, it may not have either of those.

http://www.webphix.com/schematic%20h...5_lead_45w.pdf

That shows screen resistors, but no control grid resistors.

[EFK]

Thanks to everyone for sounding off on this. As it turns out, it was a stupid mistake on my part; the inverter is a paraphase, and I accidentally had forgotten to solder one end of the grid leak R on that tube - it's a tight layout and it was easy to miss. I caught it when I left the meter on the plate, watching the lowered voltage, and continued poking around. When I hit that resistor, the voltage was jumping up and down. Silly!

[Chuck H]

Or the pair is just so mismatched that one ain't happy with its idle current. Try a different pair of output tubes yet?

The OP mentioned that the problem follows the socket, not the tube.