Bias circuits for that type transistor are all over the map. Seems like everybody has their own secret sauce. Something we have found in the past is that an amplifier layout that isn't completely stable will still behave fine in Class C, or seem to. In one case it took between a quarter and a half Watt to make the transistors turn on and amplify at all. The transistors completely ignored any input below that level.
But when bias was added, the amplifier wanted to feed back and oscillate. Adding bias had made it sensitive to input levels of a tenth of a Watt and far less. That amplifier's layout caused it to have roughly the equivalent of 1/20th of a Watt of feedback going on. So long as the radio's input was greater than that, it would swamp out the feedback as if it were not there at all. But adding bias would cause the amplifier to stay keyed after the radio returned to receive mode.
Oops.
Adding bead chokes, more negative feedback and tweaking some capacitor values calmed it down.
Could be that adding bias will go smoothly without drama.
Haven't done it in a while. I'll see if any archival docs survive and post them.
73