sorry, forgot to post some earlier test results...but first a bit of background....I had already changed q90 (as an experiment) and installed an mmbt2222a in that spot when I also changed the predriver and drivers to 2n2222a.? I hadnt made any other changes.? I did not change anything else (I will be re-biasing, but didnt have parts).? Was doing about 5w flat out on 20m.
With that as background, I had been transmitting FT8 when I eventually got fluctuating output power and then dropped (permanently) to very low power.? As the finals were the only thing I hadnt changed out, and I made some measurements that had me thinking the other stages were ok (which I now realize I hadnt posted), I convinced myself I should change them...thinking that one at least might have gone bad.
Measurements I made at the time (while in ptt with no mic audio):
3v across the q90 emitter resistor (r82 - 100 Ohm).? .44v across each of the predriver emitter resistors (r87/r88 - 22 Ohm)
under drive (wsjt-x with 'tune' set on 20M), measured the following with an oscilloscope):?1.6v peak to peak after Q90 (measured at output of RV1); 2.8v p2p and 2.4v p2p at C90 and C91 (after the predriver), 2.8v and 3.2v p2p after the driver transformer, and 13.2v p2p at the antenna output.
so...l had planned all along to make some other changes, and I'm going ahead with those.? As I was taking q90 out of the radio in favor of a better part (mmbth10, which will arrive tomorrow), I noticed that I had done a really bad job soldering in its predecessor, so I suppose that could have caused some issues...? ?As another data point, I just checked the resistors you suggested I look at (r84, r89, r961, rv1) as well as all of the other PA resistors, and they all appear to be in spec.
I'll post an update after I make the first step of the planned changes (q90 replacement with re-biasing for q90).?
