How to damp Sziklai oscillation?

[piano3]

I always load my cartridges using the Miller capacitance. The most extreme example was my mono lp/78 preamp where I was using a Shure SC35 in series connection. I found that the best solution, subjectively, was with a much higher than usual C and a much lower than usual R. Something like the DL110, which the amp in this thread is designed for, doesn’t seem to care that much.

 

[piano3]

Also, if Sziklai or Darlington is not used for the input diff pair stage, how do you get away with not using overly large emitter degeneration resistors? Does that not increase noise?

 

[flex2]

Wow! Thank you very much for this, flex2. Also thanks to everyone else: I have learned so much about Sziklai pairs, which I am still trying to absorb. There is a dearth of information in the usual textbooks which I have found very frustrating.
Actually flex2, I had not assumed that the miller capacitance was quite so high (again, I have not found a way to calculate it, so I always try to make a reasonable guess). Why do you always try to limit the bandwidth to as low as 5-6 kHz in your phono input stages?
I am now rebuilding the first stage completely having made the decision to reduce the currents by a factor of 4 using bc559/60 and applying compensation as well. I expect the Miller capacitance will be lower, considerably lower.

Yes, BC550/560 have very low Miller capacitances.
Doing the same test as before the 2nd pole is high enough to not get any gain so I can se no tendency that gain should get out of control, so probably less chance for oscillation.
The first pole shifted from 5,8 kHz to around 28 kHz, so higher linearity and less capacitance.

I didn't reduced currents in this simulation, only exchanged the transistors.