I am trying to repair a Nakamichi 620 amplifier, but it is acting strange, and I am running out of ideas.
The problem is; DC at the output. The board is supplied with +-53V and I have usually about -52V at the output.
The strange thing is, that it can sometimes drift all the way to +52V, and in rare occasions, it even centers around 0V and it actually works.

I have tried to separate the input and output stage, by 'cutting' the connection where the collectors of Q007 and Q008 are connected to the output stage, and connecting the feedback resistor R016 to GND. But I still measure -52V at the collectors of Q007 and Q008.

If I connect the input of the output stage via. a 6K8 resistor to GND, and insert a signal via a 10uF capacitor, then the signal arrives at the the output , with a DC offset of a little less than 1V.

I believe that the main problem is at the input stage. Why do I usually measure around -52V at the output of that?
It should be possible to get it to balance, but I am out of ideas, so input are welcome.

Thanks
Tom

 

There is lot's of gain between the diff pair inputs and the collector of Q007. Open loop, any imbalance in the input will push the output to one rail or the other.

Keep the loop closed and measure:
voltage at the output
voltage at the base of Q001
voltage at the base of Q004

Let's start with those.

 

Thanks for helping with this.

I have closed the loop again.
This time I measure:
Vout = +52,48V
VbT1 = -13 mV
VbT4 = -50 mV

A point to note. I got this amp as defect. Someone has tried to repair it before.
He probably had problems sourcing the original transistors, so 2SC1400 has been replaced by BC547, and 2SA750 with BC557 in both channels, only Q006 is the original.
This works in the other channel, but this one is drifting.

 

It sounds like something is open or gone high resistance to me. Have a good look at all the solder joints.
But it could be crook capacitors too. Check all the electrolytic caps as they are prone to give trouble.
Of course, it could be anything else too.

 

Yeah, check for leakage in C008, the resistance of R016 (10K) and the solder connections in that path. Under static conditions, you should be seeing about the same voltage at the output and the base of Q004. Could also be a defective Q004.

 

A quick test with C008 removed should result in a unity gain non-inverting amplifier, idling only a few mV from ground. Easy to check with a 1-2 V RMS 1kHz input and almost any scope on input and output.

 

A quick test with C008 removed should result in a unity gain non-inverting amplifier, idling only a few mV from ground. Easy to check with a 1-2 V RMS 1kHz input and almost any scope on input and output.

I wouldn't recommend doing that.
The amplifier can turn into an oscillator!
Suspicion: resistor R001 (10k), resistor R016 (10k). The presence of these two resistors should set the output voltage to zero.

 

I would like to thank everyone for their inputs.

It turned out that there was an invisible crack in the copper track leading up to one of the pads at R016.
So I soldred a wire from the output to the resistor, and the problem was solved

 

Great!