[Granitehill]

It sounds like the transistor amp might be a bodge to get round a very low reading S meter?

Mine was anything but - the meter was very optimistic, and I added a shunt resistor across the meter to pull the calibration down to something more sensible. I set it as 50 microvolts for S9, and this was pretty level across the bands. It was actually the only (reasonably) believable S meter on any RX of mine. Definately one of the better features of the radio.

I'd guess the weird mod. points to a lot of loss in the IF stages - the mechanical filter would be prime suspect, I'd say (see my PM). When the set is 100%, there's possibly a bit much overall gain - the RF gain control is certainly needed.