I suspect the short across C26 was not present when you took the data in post 43, as I would expect the bases of Q18 and Q6 to be equal and your data doesn't show that. That doesn't change suspicion of Q18--- I still believe it's failed.
Your might reinstall the C26 short and check again. If you end up removing Q18, you might recheck +/- 60V rail voltage with Q18 absent and C26 shorted. I'd be more confident we'd found the defect if we saw the rail voltages rise.
Your might reinstall the C26 short and check again. If you end up removing Q18, you might recheck +/- 60V rail voltage with Q18 absent and C26 shorted. I'd be more confident we'd found the defect if we saw the rail voltages rise.
The charging time constant is lengthy—- around 25 seconds and made worse by bridge rectifier. Slow is expected.
You should repeat this experiment once you’ve replaced the transistor and should get the same behavior. Leave current limit resistors in place.
Once you’ve cleared that hurdle, confirm amp output biases near 0V.
Next will be experiments to confirm bias spreader
You should repeat this experiment once you’ve replaced the transistor and should get the same behavior. Leave current limit resistors in place.
Once you’ve cleared that hurdle, confirm amp output biases near 0V.
Next will be experiments to confirm bias spreader
With any luck, your amp is now repaired aside from reassembly and adjustment. But, being cautious since we don't know how far damage might have propagated, I recommend checking the bias spreader (aka Vbe multiplier) since failure there could damage output devices all over again.
I wish I'd thought to suggest these tests before installing the new Q18--- there'd be a bit less vulnerability to the outputs with it absent. Probably Ok in either case with the current limit resistors.
So, with current limit resistors still in place, explore behavior of the bias adjust circuit. Initially, set the R53 pot for maximum resistance--- i.e. wiper at the Q8 base end. The remove the short across C26 and monitor the voltage across C26 instead. I estimate is should be roughly 3.3V.
If Q18 is installed, measure voltage across R67. There will likely be a a few tenths of a volt present. It should vary smoothly with R53 pot rotation, but be cautious; when R67 voltage approaches 1.2V, Q17 and Q7 should begin conduction and voltage across R68 should appear--- you're near conduction of the output pairs. When satisfied that behavior is correct, set the R53 pot back to minimum voltage across C26. With bias at minimum, you might see rise in 60V rails, but there may be enough current flowing to preclude that.
[If Q18 is absent, you should be able to advance R53 and see smooth increase of C26 voltage; the output devices should be unable to conduct since Q18 is removed. When bias circuit looks good, return pot to minimum applied driver voltage, install Q18 and go through the procedure above.]
At this point, you're near the big moment. I advise confirming supply voltages at U3, and that U3 pin 1 isn't stuck on a supply rail. My prediction is that it's fine.
Reassemble, remove the current limit resistors, steel your nerves, and apply power. Hopefully, all that remains is bias current adjustment.
Good luck!
Ahh... R53 is a 12 turn pot. I thought I had it turned all the way correctly.
I checked R53 and it's not showing any resistance changes... in fact, it shows continuity on all 3 pads. It's junk!
Now... the funny, odd, weird, coincidental, etc bit... I have a new R53 replacement! I bought it because the bias isn't as stable as I would like and thought maybe R53 was out of spec. Ha! What luck?!
The new one is showing 100ohms as delivered. I will adjust it out to 200 and install it. It's there anything else to consider before testing again?
Wow this is interesting. Thanks again for all the help!
I checked R53 and it's not showing any resistance changes... in fact, it shows continuity on all 3 pads. It's junk!
Now... the funny, odd, weird, coincidental, etc bit... I have a new R53 replacement! I bought it because the bias isn't as stable as I would like and thought maybe R53 was out of spec. Ha! What luck?!
The new one is showing 100ohms as delivered. I will adjust it out to 200 and install it. It's there anything else to consider before testing again?
Wow this is interesting. Thanks again for all the help!
I overlooked the clamping behavior of D7 and D25; they conspire to prevent prevent supply current from being delivered into the bias circuit.
That you see voltage increasing across C26 suggests that limitation may clear itself if the 60V rails rise several volts. Check if that is in fact occurring, and if the bias begins to behave. Please let me know either way.
Thanks.
Another thought. Set R53 to max resistance. This minimizes bias voltage and may help 60V rails to rise.
That you see voltage increasing across C26 suggests that limitation may clear itself if the 60V rails rise several volts. Check if that is in fact occurring, and if the bias begins to behave. Please let me know either way.
Thanks.
Another thought. Set R53 to max resistance. This minimizes bias voltage and may help 60V rails to rise.
Last edited: