Hi Don -- here are the results of my test:
1) Unplugged AGC board from main PCB. Jumpered the header on the main PCB. Left all wires connected to the AGC board that was now "hanging in space" above the main PCB. Result: much reduced RFI - still audible but at a tolerable level
2) Same as (1) with jumper removed from header on the main PCB. Result: same as (1) - RFI present but much reduced
3) Same as (2) plus disconnect S-meter lead. Result: Same as (1) and (2)
4) Same as (3) plus short J1 to ground. Result: Same as (1), (2) and (3) - actually no difference with J1 or J4 shorted to ground - I have J5 shorted to ground in all cases.
Similarly, with the AGC board plugged into the header on the main PCB, none of the changes in steps (3) or (4) affect the level of the RFI heard.
Two other things that come to mind:
1) I don't think AGC is working - I don't see any shorts or disconnected components on inspecting the AGC board, but I'm not noticing any difference in received signals. So perhaps a problem with my construction of the board and could this cause the RFI?
2) I used Dupont jumper wires to connect to the headers on the AGC board (and soldered the other end of the Dupont wire to the appropriate control. Wondering if the use of a header and push-on connector could be introducing RFI? It sure has been useful in trying A/B test cases though!