Find the schematic here:
http://www.tech-systems-labs.com/heath_schematic.htmI had one of these in the late 1980's. It seemed to be quite sensitive and very quiet (it does not have much IF gain) but I did not like the audio derived AGC on CW. It hummed a bit too. The tuning control felt cheesy. As you tuned through the signal, the AGC level changed too abruptly. It was OK on SSB. This rig used dual gate mosfets and op amps pretty much throughout. The first IF filter is a discrete bandpass LC type at 8.45 MHz. That will not be a problem as it is plenty wide.
Here are the issues:
The aforementioned Audio derived AGC works only on SSB and CW.
The second IF is a 4 crystal lattice at 3.395 MHz which is too narrow without modification.
No AM Detector.
So to do AM you will need to make 3.395 bandpass filter with enough AM bandwidth. I would make this out of a couple or three top coupled torroid tuned circuits.
I think you could make the product detector into a synchronous detector for AM but you would need to add a PLL that would work at 3.395 MHz to lock the BFO. That would be a cute conversion.
Of course you could do a simple AM diode detector off a FET follower tacked on to the filter output.
But you still need a conventional AGC that produces a voltage that can drive the existing mosfet AGC line which takes a POSITIVE voltage by the way. More positive = less gain. See the solid state handbook by Wes Hayward - he has a couple of these circuits.
Finally you must be able to switch out (off) the SSB stuff and BFO and send AM quality audio to the top of the volume control so this means another switch or a mode switch conversion.
Sound fun?
Mike WU2D