The Signalhound seems to need a d.c. path on the input. My amplifier design has an output attenuator and has worked fine; I was testing a Miteq amplifier when I obtained the strange result. I put a 3dB pad in the way and it was fine.
Reply To: Long settling time on SA44