AB763 Pre Amp Section Issue

[tdvt]

 So my Reverb Champ with the AB763 preamp & reverb modules is working pretty well but it isn't quite sorted.

There is still a high freq oscillation/squeal (3-6K) that occurs when the volume & treble pot are both near the upper end of adjustment, starting around BF “8”.  Tried multiple 12AX7s, some better than others, but same issue with all.

While it is unlikely you would actually play with those control settings, I would like to eliminate the issue if possible. I have pared down the issue somewhat but can't seem to completely cure it.

As there are probably millions of iterations of this circuit without the control interaction/oscillation/squeal, I would have to believe that the physical layout is part of the problem, but I also have to think that it can be addressed in one way or another.


What I have, built on two boards, is the Hoffman 1-channel AB763, with a few component value changes, schematic below. These boards feed into a single cathode-biased 6V6.


Differences from the Hoffman drawing are; 10uF V1 bypass caps (C7, C12), no bright switch or bright cap (C11) & the reverb insertion point resistor (R26) is 2.2M.
The problem only occurs when an instrument is plugged in, the amp is nearly dead-quiet at idle with the input jack grounded.


I have totally bypassed the entire reverb board, including the 3^rd gain stage (V3B), feeding directly to the 6V6 & the problem persisted, so the problem seems contained to the pre-amp board & panel controls (V1 & the tone stack). Removing the V1A bypass cap eliminates the problem, which led me to try a 12AY7 (with the bypass cap in place), which also eliminates the problem, so the issue seems to be gain related...?


Tried Merlin's 10K/100pF grid-stop on V1B, which did work with a 250pF, but bleeds off too much treble. The grid-stopper is still in place with a smaller cap, but there is still squeal.


So the amp is usable & sounds quite nice at lower volume but I am not sure where to look next to eliminate the problem. I am thinking that a 5751 tube might strike a workable balance between gain & not oscillating but don't have one on hand.


Any insights are appreciated.

 

[pdf64]

The layout isn’t good, output near input, pin 2 grid wiring circuit close by pin 6 anode circuit.

[glass54]

tdvt,
Just for interest, try splitting the load on V1 (R17) ie take TC off 47k and 47k tap.
Does this help ?
Just another thought, C8 Silver Mica. A favourite topic at the moment 
Maybe try another type (Ceramic?)
(+1 with pdf64, experimental layout can give trouble)
Regards
Mirek

[tdvt]

Thanks for the replies.

This was indeed an experiment & I realized the mistake I made with the location of output early on, after I fired it up.

It is almost all there, just that last bit missing, so probably better to make a new version than to patch it up.

I will try the split load though, see if that helps.

The yellow cap is actually C8, a ceramic Vishay. The little SM is part of the Merlin grid-stop trial, what was on hand.


Happily, the reverb board seems solid & functions well first try.

[WimWalther]

Have you considered installing neutralization cap(s) on the tube(s) you think are oscillating?

Try starting around 50-100pF from plate to grid, one section at a time, trial & error. I've generally used silver mica parts for this, but you can also try it with ceramics. And keep the cap leads as short as possible!

The goal, of course, is to use as few as possible of the smallest possible  values that achieve the desired effect. =)

[tdvt]

FIXED, just bad layout..

Thanks for the replies & especially to pdf64 for the nudge (intended or not) to address the root issue.

At times I find it too easy to get caught up in not wanting to undo any more than necessary, as opposed to really jumping in, even if it only seems I'm going backwards.


Anyway, while it isn't pretty, I was able to add an eyelet from the top for one extra junction & re-orient the plate nodes. Noise is gone, everything works as it should.


PS I did try bypassing the plate resistors earlier, didn't help in this instance.

Also I discovered that as the tube sockets/shields are not chassis grounded, there is some potential noise there.

I was able to address that while the module was loose with a strip of copper shielding tape on the bottom of the board, tube mounting bolt to board mounting bolt.