The plot thickens on this. I was able to get some more testing done and am more confused than when I started. I removed relay A3K1 (start disconnect relay) on the engine control board. As expected when I turn the master switch to start nothing happens. I insert the relay, start the engine, then pull the relay. The generator is not excited at this point. I checked my charging circuit voltages and they are all good. I also check the coil voltage across A1K1 main panel aux start relay and found zero volts as there should be. To be safe I removed the wire to the starter soloniod in case the A1K1 relay engages for some reason as it had recently. I excite the F+/F- with a 9 volt battery. The generator powers up perfectly and provide good stable current. I notice a buzzing that should not be there and decided to check across A1K1 coil again. To my surprise I am getting 120VAC to the coil and it is engaged. Fortunatly the starter was disconnected. I have checked every connection in the generator and cannot find any reason to get that kind of voltage there, especially when it should only see 24vdc when engaged. Aside from the 6 wires that went to the main AVR I have never removed another wire in the cabinet. Any thoughts?
Well, I think the A3CR8 diode may be bad on the control board. If it failed to a short or leaky condition then that would explain the backfeed to A1K1. Was it possible that you were measuring running field voltage at the A1K1? If A3CR8 was shorted then the field would have to go quite high because there is a resistor to go through as well as the A1K1 coil in addition to getting enough on the field to make the gen A/C.
Additionally, a leaky A3CR8 diode could be bleeding some of the flashing voltage which is preventing the AVR from picking up and taking over at cranking.
Look at the diagram, lower right area, page 4-95 of the 288 page TM. The Army TM number is TM 5-6115-615-12.
You could also swap the A3K1 relay with A3K2 (glow plug preheat) to make sure the A3K1 relay is okay.
I think it would be pretty important to have the f+ and f- wires from the new AVR on the correct terminals as well.
One thing about the 460 or any other substitute regulator is that there could be trouble during the handoff from forced field flash to self regulated operation. The forced field flash ends when the a/c output from the dc charging system builds enough voltage to pass the zener diode thus allowing the A3K1 to stop the flash and lockout the starter. Meanwhile, and all along the AVR is begining to see A/C voltage build up in the generator. At some point it is taking over. Hopefully, all of that occurs without stressing the diode. I did not look further on the diagram to determine if the A/C AVR is held off line until the flash is terminated. maybe a good modification with a substitute regulator. Because the next question, if a failed diode is why did it fail?
I suggest pulling the control board and looking it over carefully.