Thanks Oz, I believe you are correct.? I replaced the mechanical filter with a "null" 10dB 2 kohm impedance attenuator in the SSB filter location (with suitable pin coding as per the actual SSB filter).
The ripple went completely, leaving a low-pass roll-off that tallies almost exactly with a simple LTspice model of the active audio filter that follows the product detector.? That is already down by 3.3dB at 2.5kHz.
All interesting and educational - but now to try and track down why the rig is not presenting the SSB filter with the correct impedance.
Simon
M0AZR
