Hey, I'm trying to repair an amplifier that I destroyed. I was replacing the differential amplifier transistor pairs (they were causing high DC offset), when I saw a spark somewhere on the board (I forgot to discharge the main filter capacitors) and now I have 29VDC offset on one channel (supply voltage is +/-33V). It still outputs sound, but the sound is static only (same sound as an old tv with no signal). The junk speaker I had connected started smelling like burnt voice coil (still works though), which is how I detected the high offset. The other channel works fine. I've tried putting the old transistors back but no difference.
When I unplug the amp, the + supply quickly goes down to about 7V, whereas the - supply goes down to -12V, but at a much slower rate.
I've tried many things, but am totally stuck. I cannot understand what is causing this. Got any ideas for me?
Here's a schematic for anyone interested in having a look: jpg.
When I unplug the amp, the + supply quickly goes down to about 7V, whereas the - supply goes down to -12V, but at a much slower rate.
I've tried many things, but am totally stuck. I cannot understand what is causing this. Got any ideas for me?
Here's a schematic for anyone interested in having a look: jpg.
Yes, only one channel. I use the headphone output for testing, which is in series with a 330R resistor, so that part is covered. 
The bias spreader you say? Would that be Q405/406 on the schematic?
I'll check all the transistors next. Already done a couple, but wasn't sure that would be the issue (seen no problems so far).
The bias spreader you say? Would that be Q405/406 on the schematic?
I'll check all the transistors next. Already done a couple, but wasn't sure that would be the issue (seen no problems so far).
Member
Joined 2009
Paid Member
To get that much dc at the output and to have that rail collapse quickly after power down you must have a lot of current flowing from the +ve rail to ground. By measuring the voltage drop across resistors in the circuit you can discover where this current is flowing - and where it should not be flowing through a blown part. I would suspect the output (and driver) transistor on the +ve side as the first place to look.
Last edited:
You can usually check for shorts in transistors in circuit.
However I have seen weird faults with transistors like hfe =1 where I needed to pull the transistor to check it on a multi meter.
In my experience it is usually an output transistor gone which sometimes takes the driver transistors with it.
That is the one I replaced originally, the producedure that killed the amp (by something shorting), with two KSA992 transistors (wired in correctly). Currently I have put the old 2SA798 back, as a test, but the problem remains the same.
Update on the problem: I have now pulled every singly transistor on the blown side, and none of them are faulty.
And, with all the transistors physically missing - the B+ is still dropping rapidly after power down (whereas the B- goes much slower).
I will try to check some resistors next.
Also note that the phono pre amp is also using the B+ to create its own supply of 29 volts. EDIT: Disconnected. No difference. Tuner's been disconnected as well.
EDIT again: The transistor on the power supply board (Q501) has the intended voltages on the schematic. B = 14.2V C = 28.9V and E = 13.6V. The voltages I get here are: B = 12.1 C = 32.6 E = 11.6. I think this is a bit odd, because B and E are lower than intended, whereas C is higher. The emitter voltage goes to pin 14 on the main amp board (as seen to the left on the schematic). I think it might be important.
Last edited:
Already done that too. They consist of two big, dual diodes. One with + - + polarity, and the other with - + -. They appear fine. 0.6V in all the right directions (and no conductivity in the others). I did physically remove them before testing.
I've already checked all the transistors (out of circuit), and they all work.
Now I've done the following:
Disconnect everything that takes power from the supply, minus the B+ and B- going to the power amp (that means the phono amp, pre amp and tuner are not getting any power).
Move input signal from volume control directly into power amp.
Now the B+ voltage still goes down to 0V a couple of minutes after power down, whereas the B- stays at about full voltage.
DC offset is now 16VDC. A quick five second test reveals: Some hiss through the speaker (DC offset of course) but music plays nice and clear. Robust speaker, that. Started smelling again.
Now I've done the following:
Disconnect everything that takes power from the supply, minus the B+ and B- going to the power amp (that means the phono amp, pre amp and tuner are not getting any power).
Move input signal from volume control directly into power amp.
Now the B+ voltage still goes down to 0V a couple of minutes after power down, whereas the B- stays at about full voltage.
DC offset is now 16VDC. A quick five second test reveals: Some hiss through the speaker (DC offset of course) but music plays nice and clear. Robust speaker, that. Started smelling again.
Last edited:
How embarrassing.
I somehow managed to miss one small TO-92 transistor (Q403 if anyone cares), and of course - it was the culprit. It was open in all the directions.
Preliminary testing: having replaced that garners about 1mV offset (with new diff pair transistors), and music plays fine through headphones. I will stitch everything back to normal now (it's a mess) and see if it still works after.
Thanks for your thoughts, everyone. Greatly appreciated.
I somehow managed to miss one small TO-92 transistor (Q403 if anyone cares), and of course - it was the culprit. It was open in all the directions.
Preliminary testing: having replaced that garners about 1mV offset (with new diff pair transistors), and music plays fine through headphones. I will stitch everything back to normal now (it's a mess) and see if it still works after.
Thanks for your thoughts, everyone. Greatly appreciated.
Just a note from my meager experience, when you see + dc voltage on the output, it usually indicates that something on the - side of the circuit is not functioning. Either a busted transistor of an open resistor. That also accounts for the + rail draining more quickly than the - rail on shutdown.
+++10000 very correct Often people that see + Dc offset in the output run for NPN transistors in the output stage or drivers ...while the real fault is an open resistor on the other side
Well said still4given !
Kind regards
Sakis
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.