so this has been driving me absolutely nuts for about three weeks now. my IC-7300 started doing this thing where it transmits fine for maybe 20-30 minutes and then the output just... drops. like goes from 100w down to maybe 3-5w and stays there. doesnt matter what band. receive is totally fine the whole time, SWR reads normal on a dummy load, it just stops putting out power.

if i turn it off and let it sit for 10-15 minutes it comes back full power, which is what makes me think its thermal. ive already reseated the finals heatsink compound thinking maybe someone had done a bad job on it at some point but no difference. fan is spinning fine, doesnt feel excessively hot to touch on the outside. havent pulled the board yet to actually probe anything but im getting there.

anybody seen this one before? wondering if its worth just sending it to Icom or if this is something diagnosable at home. i do have a scope and a decent LCR meter but im not super experienced on RF finals specifically.

had almost the exact same symptom on a friends 7300 last year. turned out to be one of the driver transistors running hot and going into thermal shutdown before the finals themselves got warm enough to trigger any protection. the ALC was also doing weird things on the scope when it happened - were you watching the ALC meter when it drops out? if the ALC slams to max and holds there right when you lose power that points pretty strongly at a driver stage issue rather than the finals themselves.

the tricky part is catching it in the act with probes on it because the board layout is pretty tight. what we ended up doing was running it with the lid off pointed at a small desk fan, which delayed the symptom by like 45 minutes, so we could at least rule out ambient airflow being the problem. didnt fix it obviously but helped narrow things down. icom actually has decent service docs for that rig if you can find them, the voltage rails at certain test points should tell you a lot.

thermal issues on these are usually a capacitor going leaky when hot tbh. not always the obvious stuff like finals or drivers. ive seen electrolytics in the PA stage bias circuit cause exactly this - they measure fine cold and then change value enough when warm to shift the bias point way off. easy to miss if youre only looking at the big parts.

also worth checking if there are any known service bulletins for your serial number range. icom has quietly issued a couple for the 7300 over the years and dealers sometimes dont mention them.