EL84 Plate Voltage Oddity

[pgussler]

I am troubleshooting a 36 Watt type amp with 4 EL84's fed by a standard LTP PI. When I got the amp the tone was very compressed and thin on the top end. Everything looked good inside the amp, voltages, connections, bias etc with the exception of the plate voltages on one pair of the EL84's. When I put the VM on those plates it goes nuts. Like the display shows a diode check, ohms check and voltage. The other pair behaves normally. I have found that removing the PI tube corrects the issue. I have removed the NFB switch from the circuit to simplify the troubleshooting. The top of the PI's R/C network is now sent straight to ground. I have also found that when the .01 cap is removed from the PI's R/C network the problem is corrected and with it in the problem comes back. I have done all of the obvious such checking and/or replacing tubes, caps, resistors, connections, etc.

Has anyone seen this before?

[Steve Conner]

Sounds like you have a problem with parasitic oscillations. The two tubes on that side are oscillating and generating a high RF voltage that drives your meter crazy.

We'd need to know more about the specifics of the amp to help you cure that. What make/model is it, can we see a schematic? Do the power tubes have grid stopper resistors?

[pgussler]

It's a 36W clone similar to the one below. I had considered the parasitic problem and then totally forgot about it. I'll try a cap across the PI plates and see if that does the trick.

[Enzo]

Um... schematics are a lot more helpful than a layout.

I'm with Steve, my immediate reaction was that you are oscillating. This is where a scope is valuable, high level oscillation would be apparent on the screen. I would expect weak, thin, compressed sounding, and maybe a little more hum than you expect.

You report the amp was this way when you got it. Has the amp EVER worked right? Or was it built and the problem discovered?

If you disconnected the NFB and grounded the bottom of the PI tail, then the polarity of the OT primary should not be an issue. My first suspects are those looooong wires from the EL84 grids back to the 470k resistors on the far edge of the part board. Try relocating them away from the PI parts and wires that are now next to them.