I'm building this for a friend and during my testing I noticed that it has a tendency to oscillate. This is most easily seen at the C of Q1 / B of Q2. Noone else seems to have experienced this, so attached is an example from my scope: The yellow trace is C of Q1. I don't remember where I took the purple, probably at E of Q1. The RIAA correction filters high frequencies at the output (blue and green traces) so they cannot be measured at C of Q2 or E of Q3.
47P beween B and C of Q2 cures the problem.
47P beween B and C of Q2 cures the problem.
Attachments
It will probably only be exposed as increased levels of distortion as the RIAA effectively filters it away. It's happening way into the MHz range. Inspired by other, similar 70's style discrete RIAA designs, I initially tried the 47p on Q1 B-C, but interestingly that made matters worse.