So the +12 volts would be on the end of R606 that connects to the OP-AMP, right?
If so the voltage there is +11.46Vdc which is from the output of the OP-AMP, but on Q601 base it is -1.822Vdc. The speaker terminals read -1.79Vdc.
I don't see a clear path to the - voltage rail though so I am not sure how the - voltage is there unless Q601 and Q605 are bad, but shorting Q601 base to ground seems to show both transistors may be ok..
If so the voltage there is +11.46Vdc which is from the output of the OP-AMP, but on Q601 base it is -1.822Vdc. The speaker terminals read -1.79Vdc.
I don't see a clear path to the - voltage rail though so I am not sure how the - voltage is there unless Q601 and Q605 are bad, but shorting Q601 base to ground seems to show both transistors may be ok..
Yes, that's correct.
A couple of peculiar things here. Given the output voltage is -1.79V, the Q601 base voltage is much hight than I would predict. Could it be -182mV on the base?
Given that you find -1.822V on Q601, do you find the same voltage on the lead of R606 next to the body of the 1.5M resistor?
I tried feeding a 400Hz signal to the bad channel and it does produce a clean sinewave on the speaker jacks with no load connected. I then increased the signal until the amp started to distort.
So what would that eliminate if anything as a potential problem?
Also is it plausible that the same part in the other channel could be causing the issue with that channel?
So what would that eliminate if anything as a potential problem?
Also is it plausible that the same part in the other channel could be causing the issue with that channel?
The 400Hz test seems very encouraging. Did you drive all the way to clipping? If you apply a dummy load, is there any change in behavior/distortion or change in DC bias at the output? I'm hopeful that the only issue is excessive DC offset at the output.
I'm posting a very simplified schematic of the left channel to make discussion a bit easier. I'll touch on a few details and we can proceed from there.
I've depicted the power amp as an opamp; its AC gain from its + input to output is 1+R613/R611, = +23.2 V/V or about 27.3dB. You noted that R609 appears in the path to ground; I believe its purpose is to the control the path of ground currents back to the power supply. I believe an ohmmeter test will show 100 Ohms from negative speaker terminal to the shell of the input jack, and that a check of DCV across R609 will show virtually 0mV. If this is the case, it's behaving correctly.
An experiment, "thought" or actually tried: assume the servo opamp is absent or that R605 is opened. Q601 base bias current will flow through R607 and cause a few hundred negative mV to appear at the PA + input. The amp's gain of 23.2 will amplify the bias voltage and is probably the cause of the -1.79V you observed at the + speaker terminal. Then in a subsequent experiment, you shorted Q601 base to ground (making bias resistance 0 rather than 100K), and saw the output rise to +70.2mV. This all seems reasonable.
Next, we should connect R605 to +12V. The expectation is that the PA + input rises and the amp output should rise to positive several volts. What voltages do appear at + input and at output? This feature has to behave as described for the servo to work when it's connected.
Apologies for image--- I'm inept with image creation.
I'm posting a very simplified schematic of the left channel to make discussion a bit easier. I'll touch on a few details and we can proceed from there.
I've depicted the power amp as an opamp; its AC gain from its + input to output is 1+R613/R611, = +23.2 V/V or about 27.3dB. You noted that R609 appears in the path to ground; I believe its purpose is to the control the path of ground currents back to the power supply. I believe an ohmmeter test will show 100 Ohms from negative speaker terminal to the shell of the input jack, and that a check of DCV across R609 will show virtually 0mV. If this is the case, it's behaving correctly.
An experiment, "thought" or actually tried: assume the servo opamp is absent or that R605 is opened. Q601 base bias current will flow through R607 and cause a few hundred negative mV to appear at the PA + input. The amp's gain of 23.2 will amplify the bias voltage and is probably the cause of the -1.79V you observed at the + speaker terminal. Then in a subsequent experiment, you shorted Q601 base to ground (making bias resistance 0 rather than 100K), and saw the output rise to +70.2mV. This all seems reasonable.
Next, we should connect R605 to +12V. The expectation is that the PA + input rises and the amp output should rise to positive several volts. What voltages do appear at + input and at output? This feature has to behave as described for the servo to work when it's connected.
Apologies for image--- I'm inept with image creation.
In a previous reply I said R606 which should have been R605, however R606 is the same resistor just in the other channel I believe.
So given the output of the OP-AMP is +11.46Vdc would that be the same as disconnecting the resistor and applying +12Vdc to that resistor?
There is 100 ohms from - speaker jack to - input jack shell.
That said I probably need to be careful when connecting test equipment so that I don't short that resistor by using a scope and signal generator where the shell of the BNC jacks are grounded to the power cord ground.
I'll look at your previous reply in more detail when I get to work.
So given the output of the OP-AMP is +11.46Vdc would that be the same as disconnecting the resistor and applying +12Vdc to that resistor?
There is 100 ohms from - speaker jack to - input jack shell.
That said I probably need to be careful when connecting test equipment so that I don't short that resistor by using a scope and signal generator where the shell of the BNC jacks are grounded to the power cord ground.
I'll look at your previous reply in more detail when I get to work.
Last edited:
If you keep scope ground on speaker negative terminal and drive input jack in normal manner, all should be OK. There should be drop across those 100 resistors.
The voltage at the power amp + input is -1.822Vdc.
Now the voltage is closer to what is on the + speaker jack which should not normally be possible given R609 goes to ground and would divide the voltage down considerably as the voltage also passes through R613 which is 22.1k.
What I need to do is see what DC voltage is on the shell of the RCA jack to ground.
So combining the resistance of R613 and R611 I get 23.1k and when I plug in the -1.79Vdc across the speaker terminals into a voltage divider calculator I get -8mVdc which should be across R609 and is what I should see from the shell of the RCA jack to ground.
I did verify that the shell of the RCA jack has 100 ohms resistance to ground and I think I made that measurement to the - speaker terminal as it also is ground, but can verify when I get home from work.
To be honest I cannot see how I'd get that much of a negative voltage at Q601 B unless R609 is open or Q605 and Q601 are defective, but shorting Q601B to ground and the amp passing a clean sinewave up until the amp is at full output and just does start to clip seems to prove those transistors are ok.
Now to make the schematic easier to follow I extracted the two schematic pages from the service manual PDF and combined them into one image.
Once I do find the problem, would it be ok to leave the 2N5416 for Q609 or should I use the recommended replacement? The main difference is the 2N5416 has a much lower Ft than the recommended replacement and the original.
Now the voltage is closer to what is on the + speaker jack which should not normally be possible given R609 goes to ground and would divide the voltage down considerably as the voltage also passes through R613 which is 22.1k.
What I need to do is see what DC voltage is on the shell of the RCA jack to ground.
So combining the resistance of R613 and R611 I get 23.1k and when I plug in the -1.79Vdc across the speaker terminals into a voltage divider calculator I get -8mVdc which should be across R609 and is what I should see from the shell of the RCA jack to ground.
I did verify that the shell of the RCA jack has 100 ohms resistance to ground and I think I made that measurement to the - speaker terminal as it also is ground, but can verify when I get home from work.
To be honest I cannot see how I'd get that much of a negative voltage at Q601 B unless R609 is open or Q605 and Q601 are defective, but shorting Q601B to ground and the amp passing a clean sinewave up until the amp is at full output and just does start to clip seems to prove those transistors are ok.
Now to make the schematic easier to follow I extracted the two schematic pages from the service manual PDF and combined them into one image.
Once I do find the problem, would it be ok to leave the 2N5416 for Q609 or should I use the recommended replacement? The main difference is the 2N5416 has a much lower Ft than the recommended replacement and the original.
Typo. That should have been "There should be no drop across those 100 resistors." I'm sorry I missed that. My sketch is also incorrect, as the opamp negative supply comes from E3, not E1. So the 100 ohm resistor carries only bias currents to R605, R607, Q601, and the opamp + input; these should total only a few uA.
So I have no idea why the bias voltage at Q601base is so large. Could R607 be open? Could Beta of Q601 be degraded? And there's similar symptoms in both channels that might be related. Do both channels present the correct AC gain from input to PA out?
I believe Q609 should be insensitive to ft. It seems to supply only a bias current to D607 et al and doesn't provide any signal path gain.
So I have no idea why the bias voltage at Q601base is so large. Could R607 be open? Could Beta of Q601 be degraded? And there's similar symptoms in both channels that might be related. Do both channels present the correct AC gain from input to PA out?
I believe Q609 should be insensitive to ft. It seems to supply only a bias current to D607 et al and doesn't provide any signal path gain.
I think I advocate lifting one end of R605. Then the only path for Q601 base current should be through R607. That should tell something about actual base current. With power off, resistance of R607 should be apparent.
My discussions of current flow in the 100 ohm have been neglecting current flow through R611, so voltage drops across R611 and R609 deserve scrutiny--- my error/oversight. Presence of R609 does seem odd. C605 (1uF) will carry AC currents, but it might be interesting to short R609 as an experiment.
P.S. I don't believe R609 has any relevance to servo/offset error issues, though.
My discussions of current flow in the 100 ohm have been neglecting current flow through R611, so voltage drops across R611 and R609 deserve scrutiny--- my error/oversight. Presence of R609 does seem odd. C605 (1uF) will carry AC currents, but it might be interesting to short R609 as an experiment.
P.S. I don't believe R609 has any relevance to servo/offset error issues, though.
Last edited: