Presence Control in AB763 with parasitic oscillation

[norstroms]

Hello all,
I've been working on this amp for a long time (on and off for ~15 years?), making improvements to where it now sounds almost perfect. Not "Fender" perfect, but the way I want it perfect. One of the last bits I'm trying to tackle is in the negative feedback loop. I use an 820-ohm resistor as described on many schematics of that era. I like to have the option of making it sound somewhat stock, but I really prefer to have less NFB in that circuit so I installed a 5k-l pot after the resistor. I also thought well why not add a presence control while I'm at it?

So I have experimented with both major versions that I've seen described on circuits of the previous era: a 25k-l pot/4.7k/.1uf circuit and a 5k-l pot/1.5k/.1uf circuit. The 25k pot sounded more like a presence control at all volumes, the 5k pot produced a barely audible result unless I was really cranking the pre-amp. In either configuration I get parasitic oscillating feedback when turning the presence control up past roughly noon, with the reverb also on anywhere above roughly noon. Basically it's impossible to turn any combination of the reverb control, NFB control, and presence control up at the same time.

I read that sometimes reversing the OT primaries will solve something similar to this, so I tried it and got a very bad motorboating result and switched it back immediately.

Any thoughts on where to approach parasitic oscillation in the negative feedback loop? Is it just not such a good idea to include a presence control in a circuit more or less based on the AB763 design? Is this partly why Fender ditched the presence control in this era?

I've included a schematic (below). The NFB pot is noted as 25k-l because I am also experimenting with even less NFB in the circuit. Maybe this is also inadvisable for this design... I'll find out sooner or later. Using my ears, so far it sounds like there isn't much of a difference past about 2.5k or so...
Thanks for any thoughts - N

[HotBluePlates]

You have massively more negative feedback with your arrangement than the stock Fender. If you borrow or copy something from an amp, you have to copy all it's elements for it to work the same.

So what do I mean? The 820 ohm resistor is only half of the negative feedback loop. The other half is the 47 ohm (or 100 ohm, depending on model) resistor to ground under the tail of the inverter.

So your 4.7k resistor is 100 times bigger than it should be. That being the case, the easiest route is to make the 820 ohm resistor 100 times bigger as well. Swap an 82k resistor for that 820 ohm resistor, and see if the oscillation tames down.

Too much negative feedback will make an amp unstable.

Reducing the negative feedback will make the presence control seem to have less effect. If you feel you need more feedback, or more effect from the presence pot, reduce that 82k resistor incrementally. If you start getting instability again, you'll need to raise it. While some instability might be caused by wire routing, there is[/i] a finite amount of feedback at which a given circuit design and OT will be unstable no matter what you do.