ROTEL RA-820BX3 not working

[sgrossklass]

Originally Posted by sangeet2

Across R697 I am measuring no voltage.
R692 / D651 I am measuring no voltage.

That's why Q639 doesn't do anything.

Is the R692/D651 junction exactly at 0 V or more like -0.2 V? That would make the difference between a Q638 that's dead short (or a shorted C652) and one that merely is doing its job.

What's the voltage at the R694/R696/D652 junction? It should be positive to -0.6 V, a value of about -1.0..1.2 V would indicate that Q638 rightfully turned on. If that should be OK, you can temporarily remove C652.

The DC offset and amplifier "performance" indicate that there must be more amiss, unfortunately.

But since both channels are equally affected, this could be a bad connection of +/-B1 supplies and ground from the power supply to the power amp. Should be easy to probe with the unit off. Could it be that the amp has been dropped or otherwise handled roughly?

[sangeet2]

Originally Posted by sgrossklass

That's why Q639 doesn't do anything.

Is the R692/D651 junction exactly at 0 V or more like -0.2 V? That would make the difference between a Q638 that's dead short (or a shorted C652) and one that merely is doing its job.

What's the voltage at the R694/R696/D652 junction? It should be positive to -0.6 V, a value of about -1.0..1.2 V would indicate that Q638 rightfully turned on. If that should be OK, you can temporarily remove C652.

The DC offset and amplifier "performance" indicate that there must be more amiss, unfortunately.

But since both channels are equally affected, this could be a bad connection of +/-B1 supplies and ground from the power supply to the power amp. Should be easy to probe with the unit off. Could it be that the amp has been dropped or otherwise handled roughly?

I am measuring the following (with relay jumpered):

R692/D651 = 9 mvolt
R694/R696/D652 = 1,2 volt


ground seems ok
+/-B1 also

The sound is ok with my headphone, my testspeaker is bad - sorry for my fault, not checking this better before

[sangeet2]

Originally Posted by jaycee

Hm, so there is a problem. If you have some 10uF capacitors, try swapping out C601/C602. Signal coupling electrolytics have a habit of going bad and causing distortion

Still more then 200 mv DC at the speaker output after exchange of the caps

[sgrossklass]

Originally Posted by sangeet2

I am measuring the following (with relay jumpered):

R692/D651 = 9 mvolt
R694/R696/D652 = 1,2 volt

That's MINUS 1.2 V, I suppose? That would explain why the relay driver circuit does not engage.

What do you measure on the other ends of R694 and R696? I have a hunch that the indicated 36 V may be quite a bit low.

You said the thermistor measured 70k. That should be OK, it's a nominally 50 kOhm part, NTC. As it warms up, its resistance reduces, thereby shifting voltage at R694/R696/D652 into negative territory until eventually (overtemperature condition) Q638 starts conducting and thus turns off Q639.

Originally Posted by sangeet2

Still more then 200 mv DC at the speaker output after exchange of the caps

How much can you influence this by turning the input stage balance adjust pot, VR603?

This design is a little strange. First of all it seems to be an inverting amplifier at a gain of only about 22 dB?! Input resistance is somewhere between about 5k and 20k depending on volume.
Input pair balance is very odd - DC resistances at the bases are deliberately imbalanced (with an offset control provided to get output offset in check), at signal frequencies impedances become approximately balanced. Whoever designed these Rotels had a penchant for very strange design decisions at times (I once helped a guy mod his RA-980BX for much reduced noise, turns out they used a rather noisy AD712 opamp after the volume control).

[sangeet2]

Originally Posted by sgrossklass

That's MINUS 1.2 V, I suppose? That would explain why the relay driver circuit does not engage.

thats not MINUS

Originally Posted by sgrossklass

What do you measure on the other ends of R694 and R696? I have a hunch that the indicated 36 V may be quite a bit low.

around 12 V

Originally Posted by sgrossklass

You said the thermistor measured 70k. That should be OK, it's a nominally 50 kOhm part, NTC. As it warms up, its resistance reduces, thereby shifting voltage at R694/R696/D652 into negative territory until eventually (overtemperature condition) Q638 starts conducting and thus turns off Q639.

I measure the thermistor again, its 56k

Originally Posted by sgrossklass

How much can you influence this by turning the input stage balance adjust pot, VR603?

not enough, but I can influence the DC

Originally Posted by sgrossklass

This design is a little strange. First of all it seems to be an inverting amplifier at a gain of only about 22 dB?! Input resistance is somewhere between about 5k and 20k depending on volume.
Input pair balance is very odd - DC resistances at the bases are deliberately imbalanced (with an offset control provided to get output offset in check), at signal frequencies impedances become approximately balanced. Whoever designed these Rotels had a penchant for very strange design decisions at times (I once helped a guy mod his RA-980BX for much reduced noise, turns out they used a rather noisy AD712 opamp after the volume control).

well, this doesn't mean much to me, as I don't understand it
Can you tell me what disadvantage this DC offset has?

[sangeet2]

Ok, after I found out that my multimeters plus and minus inputs were exchanged I can verify the voltage at R694/R696/D652 = 1,2 volt.

The DC on the speaker terminals is going down from 270 mV to 190 mV on one side.
On the other side its going down from 200 mV to 116 mV.

This has been measured after the amp has been running for about 20 minutes.

But thats still too much. After reading for some time in the web I assume that its possible drivers are bad...

[sgrossklass]

Originally Posted by sangeet2

thats not MINUS

Hmm.

What do you measure across R695?

The transistors are both original 2SA608K?

Originally Posted by sangeet2

around 12 V

Now that doesn't sound right at all. The voltage on the other end of R696 should be +36 V (maybe +33 V here), and R694 goes to the thermistor, and with the thermistor value you determined, its other end should be at -2..-3 V or so.

Verify function of D609 in off state (diode test in both directions), and measure the voltage at its cathode (should be same as one end of R696 - if it isn't, there must be a bad solder joint at work).

If the diode looks fine but the voltage still stays at +12V instead of +33V, try swapping in a new electrolytic for C653 (10µ / 50V or higher).

Originally Posted by sangeet2

not enough, but I can influence the DC

That'll need further investigation then. Looks like loop gain is very low - at least that's the only explanation I can come up with. But before we get to this, let's sort out the relay problem first.

[sangeet2]

Originally Posted by sgrossklass

Hmm

Yes it is Minus, see my post above

What do you measure across R695? 0,6 Volt

The transistors are both original 2SA608K?
No, they are AC698KF8G, and if I look at the other transistors, guess what: they also don't match the ones from the schematic

Now that doesn't sound right at all. The voltage on the other end of R696 should be +36 V (maybe +33 V here), and R694 goes to the thermistor, and with the thermistor value you determined, its other end should be at -2..-3 V or so.
From ground to R694 I measure almost 30 V and MINUS 1,2 V

Verify function of D609 in off state (diode test in both directions), and measure the voltage at its cathode (should be same as one end of R696 - if it isn't, there must be a bad solder joint at work).
Diode is ok, voltage at its cathode is 0 V anode is almost 30 V.

If the diode looks fine but the voltage still stays at +12V instead of +33V, try swapping in a new electrolytic for C653 (10µ / 50V or higher).
I have done this already...

That'll need further investigation then. Looks like loop gain is very low - at least that's the only explanation I can come up with. But before we get to this, let's sort out the relay problem first.

Yes please

[sgrossklass]

Originally Posted by sangeet2

The transistors are both original 2SA608K?
No, they are AC698KF8G, and if I look at the other transistors, guess what: they also don't match the ones from the schematic

They appear to be doing their job in any case.

What other kinds of discrepancies did you notice?

Originally Posted by sangeet2

From ground to R694 I measure almost 30 V and MINUS 1,2 V

I suppose you meant R696? Wouldn't make sense otherwise.

If everything were OK, those -1.2 V should be something like +14 V. The thermistor would have to go down to about 1k for the relay driver to turn off.

Please check the values of R696 and R694. The only explanation I have now is R696 being way out of spec (high).

[sangeet2]

Originally Posted by sgrossklass

Please check the values of R696 and R694. The only explanation I have now is R696 being way out of spec (high).

YEAH!
It turned out that R696 was faulty. After I exchanged it, the relay works
Thank you so much for your guidance and patience!

I still have to get the DC on speaker output down...
Could it be that the power supply electrolytics are bad after 20 years? They are Rubycons 85 degree