Osc problem in 4 x EL84

[tubers]

I have had as oscillation problem in one of my 4 x EL84 amps for quite a while. Finally I took the time to try and track it down.

The oscillation is a little tiny thing that tacks itself onto the front end of the wave form when the volume is up loud and a hard not is plucked. Problem is that the power stage turns this into a nasty spike.

I did everything from chopstick, shift low impedance wires everywhere, rewire the board, swap tubes, swap components, shield grid wires, shield tubes, remove NFB, remove presence, move grid and screen resistors to socket. Still the little rascal is there. See the attached scope shot for a picture.

I have narrowed it down to the PI. The signal coming in is pretty clean (although a big, overloading signal), but coming out of the PI there is that little squiggle that is causing me grief.

So I took another, older version of this amp, that was built in a completely different style and swapped the preamp - problem still existed. Then I swapped PI and power stage - problem was still there! So all I can conclude is that the osc is part of the design. Perhaps it is the nature of the LTP PI when faced with an overloading signal(?!?!)

Anyways to my request. I would like to filter this little baby out of the inputs to the power stage. I am using 4xEL84. They have an input capacitance of 35-50pF and I am using 33k input grid resistors, so I slapped in a 470pF to ground on each grid to give me a 9.5khz LPF, but all that did was produce a soft whine. How do I filter this out?

thanks
Dave

[drz400]

Can you see it with a scope feeding amp with Sine wave? Square wave? A squiggle on the sine wave?
I dont think a filter is the answer.
Are you talking about single coil pickup "hack" when you pluck a string hard on a fendery style circuit?

[tubers]

I will try a square wave and see what I get. Problem is, it happens with a guitar so it has to work with a guitar. It happens with any kind of a guitar (Strat, LP, p90).

I agree, a filter is only a band aid. But believe me I have spent dozens of hours ripping this thing apart.

About the only thing I can conclude is that large signals (> 2vpk) into a LTP PI cause something odd to happen...

[tubers]

Here is a 440Hz square wave. Sine or ramp wave produces nothing (not much of a surprise, as it is the hard attack that produces the issue).

[Allynmey]

What are you using for grid load res. on the PI? I had a 4Xel84 amp with 470K res. and I had the same problem. When I backed off to 330K it lessened but what still there. i ended up at 220K and it's gone! Might be a start.

Allynmey

[d2camero]

I am using 470k. I will 220k - anything to help!

[d2camero]

Sigh... I tried 235k (parallel 470k) - no change, other than the input to PI signal dropped 100mv or so.

[klingo]

OK this is a dumb question,

have you the small ceramic cap (~100nF) across the PI output ?

what are your PI values/topologie...schem?

[VacuumVoodoo]

Since you observe the spike already at the PI input (the blue trace says PI input) it is happening somewhere ahead of it. To make sure disconnect the PI and measure at the preamp output. It looks like you have a sharp resonant peak in frequency response at some high frequency. Can you sweep input frequency with sine signal and produce a amplitude (db) vs frequency plot? A full schematic with some pics would be helpful too.

[sebastian]

Post here a full schematics,I think you have done some mistakes in the layout,maybe you have a inductive and/or capacitive coupling in some places....
Try a 100 pf across input of PI and signal ground......

[drz400]

I'm wondering if you are listening to the very common "hack" sound transient distortion when you play single coils too hard through a Fender amp like especially one that has healthy high end like the Reverb side of a Super reverb. In that case I have to ask..... do you really play like that ?
Do you hear this noise under normal playing circumstances? Cause I can make ANY fender style circuit crap out like that with a single coil if I want to, but it isnt pleasing to play that way.