[Diabolical Artificer]

No, only two valves. I found two "sense" resistors had gone high, from 10 ohm to 400k. This was putting the cathodes at 23v. What I can't get my head round is that 23/400000 = 575 uA cathode current. This obviously isn't the case or they wouldn't red plate. As the OP stage is running on fixed bias, this complicates things. I didn't have chance to take readings to suss out things further as I switched off sharpish.

Earlier in the thread I mentioned ways of protection should any OP valve draw too much current. Putting small wattage sense R's on the OP valve cathodes was muted as a possible protection "device". I've read the same suggestion over on DIY audio. However not all R's go short under high current conditions, as proven by these R's going high. Whether they went high as a result of over heating or are just faulty am not sure. When looking for replacements I found another "new" one, that had also gone high, around the same value, IE 400k.

Anyhoo, replaced said resistors, all well again - ish. However, I still got the "squeak" fault today. It sounds like the squeak you get with dry ice. Adjusting the CCS cures this. It's almost like the pot and some capacitive component, maybe the 0.1u cap across the zener, is acting like a tank circuit. I can only think that the CCS isn't stable, which puts the PS into oscillation, which in turn sends the OP stage into oscillation. The fault usually occurs at switch on, adjusting the CCS current kills the fault.

I've put a 1u and 0.1u cap after the 1235u cap "bank" to shunt HF to ground. Will also put some ferrite beads on the CCS transistor legs, OP anodes, g2's.

When stable and working, it works grand, with 120w OP. Measured ripple voltage is 200mV P-P measured at OP stage anodes.

A.