SWR Wrkgn man's 4004 overheating

[Mars Amp Repair]

hey folks,
Working on a Workingman 4004 SWR amp an need to know if anyone has had experience with how hot the heatsink normally gets.

This amp seems to get extremely hot on the heatsink...more so that you would think is healthy. The bias adjusts fine by crossover method suggested by SWR and the resultant idle bias current is well within spec. Yet, when you push the amp to about 60-100 watts, the heatsink is so hot you cannot touch it.

This is a 'newer' FMIC variety. We have tightened the bias transistor mounted on the heatsink as they have a tendency to get loose & no longer give that bias heat feedback. When you remove signal, the bias voltage/current settles right back down to the idle value, so it seems like it's tracking.

any ideas...I'm beginning to think the amp is functioning normally. thanx, glen

[Gtr_tech]

Can't say fer sher, but I wouldn't rule out poor thermal management. I declined repair on a pwr'd mixer recently because the heatsink type/area was clearly not up to the task of dissipating the heat generated by several to-3 transistors. So it blew up for a reason and it'll just happen again.

[J M Fahey]

Maybe you can mount some little fan there, to move air around a bit.
Any air movement will improve things a lot.
Anyway, before that, you should check that there is not something else heating those transistors.
Some instability might cause, perhaps not full RF oscillation, but some of it on certain parts of the audio wave.
Scope it , hooked to a real world speaker, not a load resistor, and drive it with different frequencies, maybe at certain ones you can see some "grass" or "fuzzies" superimposed on the clean sinewave.
Be sure to sweep from , say, 30 to 150Hz, because near resonance speakers become very complex loads, and also , say, from 1 to 5 Khz.
I think these have piezo tweeters or regular ones with some crossover; a short after a capacitor puts an "ugly" load to the power amp.
Or the Zobel network might simply be open.
EDIT: I forgot: maybe it's not *that* hot, simply it seems so.
Try to apply some kind of thermometer, hands are not reliable.
Heat sinks get "X" degrees above ambient temperature, whatever it is.
Now I remember you are in full Summer, and in Colorado to boot; here in Buenos Aires we are freezing and even getting a little snow, *very* unusual here.
The very same amp will feel unbearably hot there and just warm here.
A heatsink at 70ºC will start to feel "untouchable", yet be within spec.

[Enzo]

Or try biasing it by the main current method. or just back it off to a colder setting to see if it make a difference.

[zero cool]

Check the +/-15vdc supply. there is a resistor/zener regulator and poor mounting of the large resistors can cause them to break or come loose. On the SM900's there are two large white 5 ohm resistors in parallel. there is 7 watts going through those resistors in that model and its enough to cause the solder joint between the resistors to melt the solder right out of. this will cause one side of the supply to drop to 7.5V and the output stage to go into full oscillation!