So I recently retrieved my Argonaut 505 Amateur Radio Transceiver from storage. Turned it on, no hiss from the speaker. That’s… unusual, points to a dead audio amplifier.
Geek line, do not cross. Geek line, do not cross. (TLDR: I solved the problem)
This is the schematic of the audio amplifier of the 505 (the slightly later, slightly upgraded Argonaut 509 uses an integrated circuit audio amplifier). The Operator’s Manual is also good enough to provide voltage levels at the transistor pins.
So I measured, and found around 0.5V on Q4 base instead of the anticipated 6V. Q5 base was a tad lower. The obvious reasons for this would be Q5 base shorted to ground, or Q3 shorted, or Q3 turned on. The base voltage on Q3 didn’t look high enough to turn it on so…
Pulled Q5, tested it on my multimeter, it was good. So was Q3. Now how is that possible?
So I pulled Q4 as well, and that too tested good.
Next theory, something is disturbing the biasing (there’s a DC feedback loop at play here for the bias, and an AC feedback loop for the gain). The base of Q1 is biased to 5.2V by R2/R3/R4, this determines that the emitter of Q1 needs to be 0.6V higher (Vbe). And small variations in this emitter voltage will cause Q1 to turn on or off, varying the collector current and in turn the voltage over R21.
This voltage determines whether Q3 turns on or off, which drives Q4 and Q5, which determine the voltage on the emitter of Q1. See? Feedback.
Anyway, the point is that if this feedback loop is disturbed, Q3 could turn on and drag the base voltages of Q4 and Q5 down. Removed both capacitors, nothing changed. And all the resistors measured the correct value.
Now what?
With all this in-and-out, I eventually noticed that intermittently things would come right. So I started undoing things until I found …
This is Q4. It’s designed to be mounted on a heatsink — it’s just the standard transistor package inside a heat-conducting easily-mounted aliminium block. It had developed an internal short between the base and the case. First time I’ve ever seen this happen. Grounding the case by mounting the transistor on the heatsink and the heatsink to the chassis pulled the base low.
A mechanical problem, in the end. No wonder it didn’t make sense looking at the schematic.