PV Vintage (ss pre/PI) output tube grid drive oddity

[Gtr_tech]

OK, so this "Vintage" came in needing all kinds of work. I did a complete recap of the pwr supply and preamp signal path, redid the screen supply to a configuration that would be easier on current prod. screens. Output looks cleaner and is free of the slight oscillation that was riding on the audio signal. Here's the final issie....it has real weak output. Plate/screen V's are normal albiet on the high side. What I did was pull the output tubes and check the grid drive signal....looks good with plenty of available drive V. But....once the output tubes are in the sockets, grid drive decreases by about 75%. I was seeing ~40vac (halfway up on the master) on a single grid with tubes pulled, and with tubes in it drops to ~10vac. Knee-jerk reaction tells me its a loading issue, but the control grids shouldn't be loading much if any at all. Waveform doesn't get dirty or noisy, just gets puss.

Bias is stable so the .1u coupling caps should be ok, I don't suspect the driver pair since they aren't be'n asked to deliver current so much as voltage.

Output section is identical to the Deuce...
http://www.webphix.com/schematic%20h...vy_deuce-b.pdf

Any ideas before I pull the damn board again and check the driver pair for leakage?

[Steve Conner]

Putting the power tubes in closes the NFB loop, so of course it will decrease the drive to the grids.

I think your real problem is that the preamp isn't putting out enough juice to the PI. You probably dislodged something during the recap or put an electrolytic in backwards.

[Gtr_tech]

Goddammit....I just typed out a reply and the page freaked and I lost it. Got a theory....hang tight....

[Gtr_tech]

Ok....snipped the FB wire and drive signal is back. Scope tells me its stable. Sounds really good....

Now they're not using a ton of NFB via a 47k feed, doesn't jive with me that the FB should all but kill drive signal. Since it sounds so nice without the FB, I'm tempted to leave it that way since the owner is a blues player and will likely favor the more open tone without NFB. But...I'm also "getting the elbow" to check that driver pair for leakage since everything else checks out and it had to operate correctly as designed.

[Steve Conner]

I don't think the PI transistors have anything to do with it. If they're leaky, the DC levels would be wrong, so check you have 24V at the emitters and 180V at the collectors.

I still suspect you're not getting enough level from the preamp. Did you check the preamp's +24V rail voltage is correct and all the transistor bias voltages are reasonable? Also, is the tremolo working? It could be stuck in the "quiet" phase, maybe from a shorted JFET.

Also check the ground connection running from the 1k half of the NFB divider back to the preamp. If that was open, you'd have 47x more NFB than you should.

[Gtr_tech]

All V's around the driver pair are within spec, rails are good as well. Everything around 1K is within spec, good gnd connection. So this doesn't make a whole lotta sense. But...I just got off the phone with Gene at PV and he agreed with me that as long as its stable it should be OK to run it as is. So I think we're gonna run with it. Sounds pretty nice this way.