I was wondering if you'd got the panel sorted on this yet!
Well I've thought about it quite a bit :tongue:
Anyway, moving on, there is progress!
I've built up the replacement for the dodgy part of the amp onto a proper piece of stripboard with nice new components, and sort-of faked it driving the output stage. Stability is excellent. No nasties when hit with sharp edged square waves, nor with slowly ramping triangles. From cold (where the offset has been zeroed) there's no more than 1mV drift after about half an hour. When a sine wave input is run massively into overload (to almost a square wave) the offset shift is less than 2mV.
I'll do a few more tests to see if I can make it hiccup, then if all is OK I'll unhook one of the existing sections of the amp and try this in its place . I'll probably use the sub amp as if it all goes tits-up that's the one I don't currently use anyway.
Once I'm more confident I'll post drawings and pix.