More on the changing voltages...as things warm up.
After 5 minutes
R17/Q1: 342mV
R17/D39: -13.5V
R18/Q2: 322mV
R18/D40: -14.6V
After 10 minutes:
R17/Q1: 349mV
R17/D39: -13.08V
R18/Q2: 328mV
R18/D40: -14.45V
After 15 minutes:
R17/Q1: 355mV (near steady, rising maybe 1mV every 20 seconds)
R17/D39: -12.92V (near steady, moving toward 0, maybe 10mV every 20 seconds)
R18/Q2: 332mV (again, near steady, rising slowed considerably)
R18/D40: -14.35V (again, near steady, dropping towards 0 slowed considerably)
After 5 minutes
R17/Q1: 342mV
R17/D39: -13.5V
R18/Q2: 322mV
R18/D40: -14.6V
After 10 minutes:
R17/Q1: 349mV
R17/D39: -13.08V
R18/Q2: 328mV
R18/D40: -14.45V
After 15 minutes:
R17/Q1: 355mV (near steady, rising maybe 1mV every 20 seconds)
R17/D39: -12.92V (near steady, moving toward 0, maybe 10mV every 20 seconds)
R18/Q2: 332mV (again, near steady, rising slowed considerably)
R18/D40: -14.35V (again, near steady, dropping towards 0 slowed considerably)
Essentially yes. Make sure to really drive the probe into the PCB to break through any lacquer etc.
Its because those are the voltages relative to your point of reference (black meter lead on ground).
Concentrate on ONE channel at a time. Lets take Q1. Something is seriously amiss when you are seeing 7 volts here. It should be near zero.
The next test at this point is simply to look at the output on an oscilloscope... can't remember if you have access to one or not
That will tell you whether you are dealing with a dynamic issue such as instability, or just a DC fault.
Hi Mooly. So the 7.3 volts I'm seeing at R21 (and 3.5V at R22) should be near zero. These two resistors connect directly to C9/C10 on the other sides.
This makes me go back to an original question I had about these caps. The service manual and schematic indicate that these caps should be 470uF 25V NP (Non-Polarized), correct? The ones I have, that appear to be original as they still have glue around the bases are polarized. See pic in thread #15.
1. I have NP replacements. Should I pull them, check them on my meter and/or replace them?
2. All that glue around the base overflows on a number of surrounding components. Could that lead to contamination you speak of?
This makes me go back to an original question I had about these caps. The service manual and schematic indicate that these caps should be 470uF 25V NP (Non-Polarized), correct? The ones I have, that appear to be original as they still have glue around the bases are polarized. See pic in thread #15.
1. I have NP replacements. Should I pull them, check them on my meter and/or replace them?
2. All that glue around the base overflows on a number of surrounding components. Could that lead to contamination you speak of?
Yes, there should be zero volts on R21/R22 and also zero volts on the input to the power amp (R2).
On a working amplifier, if you applied a DC voltage to the input (on R2) then the main output of the amplifier should also reach that same DC level, plus or minus, whatever polarity and amplitude the applied voltage happens to be. That is so because the amplifier has a gain of unity or 1 at DC.
So its very important to confirm that there really is no applied voltage present (through leakage or whatever) at R2.
The glue is a known source of problems in a lot of equipment (not just audio). Definitely clean it up and remove it.
On a working amplifier, if you applied a DC voltage to the input (on R2) then the main output of the amplifier should also reach that same DC level, plus or minus, whatever polarity and amplitude the applied voltage happens to be. That is so because the amplifier has a gain of unity or 1 at DC.
So its very important to confirm that there really is no applied voltage present (through leakage or whatever) at R2.
The glue is a known source of problems in a lot of equipment (not just audio). Definitely clean it up and remove it.
Ok. Checked DC Offset and voltages on Q1 & Q2 before starting. Very close to last readings. But must have missed this. I originally stated R22 was positive 3.5V. It was negative. As Mooly indicated, the voltages I see at R21 and R22 are same as DC Offset at speakers. I think I may have also indicate R21 and R22 were V. They are mV. Sorry. I must pay closer attention.
So, I pulled C9 & C10. Cleaned up most of the glue on surrounding components. Tested caps on meter and they are at 471/473. In spec. I also saw positive indicator on board which matches how they were installed, so perhaps schematic is wrong.
Soldered them back on and voltages are about the same
R17: 7.1mV
Left offset: 7.1mV
R18: -4.0mV
Right offset: -4.0mV
So, I pulled C9 & C10. Cleaned up most of the glue on surrounding components. Tested caps on meter and they are at 471/473. In spec. I also saw positive indicator on board which matches how they were installed, so perhaps schematic is wrong.
Soldered them back on and voltages are about the same
R17: 7.1mV
Left offset: 7.1mV
R18: -4.0mV
Right offset: -4.0mV
So the latest voltages I am seeing are when its stable and plays fine. I'll get new measurements the next time it acts up to provide comparison. What is frustrating (besides my bad numbers) is that it doesn't always act up after it warms up. Last time it acted up was when I first turned it on. And that was on CD not Phono. Previously I saw the issue twice on Phono.
Feedback network caps can be tricky polarity wise. Normally the DC voltage across them is so small (a millivolt or two) that it doesn't matter which way around a standard polarised cap is fitted. Ideally you could measure the voltage and fit the caps accordingly, even if one channel ends up different to the other.
So those offsets are fine at the moment.
So those offsets are fine at the moment.
Yes, last night voltages seemed to be normal on both channels. Next step is to try and reproduce bad voltages, whether that is caused by switching to CD or Phono, overheating or a bad connection/short somewhere remains to be seen.
As I re-read this thread, here is a summary:
1. I tested this amp at seller's home and we got music out of right channel with static. No music from left channel.
2. Got it home and pulled front face and cleaned all contacts.
3. Replaced fried/cracked C39 (and C41).
4. Started getting clean music from both channels on Aux1 playing iPod (for about 10 minutes).
5. Connected Turntable and music stopped in left speaker after about 3 minutes. Music stopped playing in right speaker about 10 seconds after that.
6. Measurements showed high DC offset on left channel. Bad voltages on Voltage Amp Assy at most transistors on right channel.
7. Day later, DC offset on both speakers down to 7.1mV (L) and -4mV (R). Working again.
8. Day later, went to set Phono EQ adjustments which were at 400mV (L) and 1400mV (R) to start. Got them down to less than 100mV, but they were fluctuating, creeping back up. Left heatsink noticeably hot. Right DC offset at -4mV, but left up over 800mV.
9. Two hours later, amp has cooled and DC offsets have returned to 7.2mV and -3.5mV.
10. From this point, right channel has remained stable, no bad voltages.
11. Day later, turned amp on CD (plugged iPod cable into CD this time), and immediately saw left channel DC offset back up near 800 mV.
12. Voltage across dual resistors 180/-180/180mV and climbing
13. Replaced burned out "green" Power on Pilot Lamp. And now have both red light at turn on, and green light when relay clicks on startup working (neither worked before replacing the green light).
14. With focus on Q1/Q2, amp has once again returned to "normal" and voltages at Q1 and Q2 appear to be normal.
15. Pulled C9/C10 to remove factory glue from caps and surrounding components and test the caps. Caps are still within spec (471/473).
16. After re-assembly, amp still showing DC offsets are good at 7.1mV and -4mV.
As I re-read this thread, here is a summary:
1. I tested this amp at seller's home and we got music out of right channel with static. No music from left channel.
2. Got it home and pulled front face and cleaned all contacts.
3. Replaced fried/cracked C39 (and C41).
4. Started getting clean music from both channels on Aux1 playing iPod (for about 10 minutes).
5. Connected Turntable and music stopped in left speaker after about 3 minutes. Music stopped playing in right speaker about 10 seconds after that.
6. Measurements showed high DC offset on left channel. Bad voltages on Voltage Amp Assy at most transistors on right channel.
7. Day later, DC offset on both speakers down to 7.1mV (L) and -4mV (R). Working again.
8. Day later, went to set Phono EQ adjustments which were at 400mV (L) and 1400mV (R) to start. Got them down to less than 100mV, but they were fluctuating, creeping back up. Left heatsink noticeably hot. Right DC offset at -4mV, but left up over 800mV.
9. Two hours later, amp has cooled and DC offsets have returned to 7.2mV and -3.5mV.
10. From this point, right channel has remained stable, no bad voltages.
11. Day later, turned amp on CD (plugged iPod cable into CD this time), and immediately saw left channel DC offset back up near 800 mV.
12. Voltage across dual resistors 180/-180/180mV and climbing
13. Replaced burned out "green" Power on Pilot Lamp. And now have both red light at turn on, and green light when relay clicks on startup working (neither worked before replacing the green light).
14. With focus on Q1/Q2, amp has once again returned to "normal" and voltages at Q1 and Q2 appear to be normal.
15. Pulled C9/C10 to remove factory glue from caps and surrounding components and test the caps. Caps are still within spec (471/473).
16. After re-assembly, amp still showing DC offsets are good at 7.1mV and -4mV.
Last edited:
Just looking at the phono stage and I see that is ultimately AC coupled via C23 (C24) to the power amp section. Providing those caps are OK (no leaky glue across them) then no DC fault on the phono section will cause any DC offset at the power amp.
You are saying that as of this moment it is all working OK ?
You are saying that as of this moment it is all working OK ?
That's worth doing as is checking C13 and C14 for the dreaded glue. The DC offset of the phono stage should be quite respectable stability wise although as mentioned, no offset is going to carry through to the power amp as long as C14/15 are OK.
Actually, I would just replace C13 and 14 because they never see any DC voltage across them to 'form' them in normal use. They may well have deteriorated prematurely through 'inaction'.
Actually, I would just replace C13 and 14 because they never see any DC voltage across them to 'form' them in normal use. They may well have deteriorated prematurely through 'inaction'.
Phono Stage update.
1. I pulled C23/C24 and tested them. C23: 23.35uF, C24: 23.43uF. So they still look good (should be 22uF). They were both Elna Audio Grade Non-Polarized and both had been glued, with glue on a couple of surrounding resistors that I cleaned up..
2. I pulled C13/C14 and tested them. C13: 1070uF, C14: 970uF. So they appear to be good (should be 1000uF). They were both green Nichicon Muse caps and both had been glued, with glue on a couple of surrounding resistors that I cleaned up. I noticed while testing these caps that it took about 10 seconds to get the final readings, as they appeared to take that long to charge up? Is that normal on larger caps?
3. After re-installing all 4 caps, I went to adjust Phono EQ (via VR1 & VR2). This time I did not switch the Input Selector on the Amp from Aux to Phono, I left it on Aux and found that I still had current throughout the Phono Stage to make the adjustments. I also grounded the Phono Inputs on the back per the Service Manual. What purpose does this serve?
Once again, I was able to dial in +/- 100 mV on both, at least get them in the ballpark, but they both continue to fluctuate constantly and the VR1/2 pots were very sensitive. A slight turn and voltage would shoot up to 800mV very easily. Stopped here for the night.
4. This morning I went to check Phono EQ voltages on TP1 (left) and TP3 (right) and saw the following (I checked the voltages this time without grounding the Phono Inputs on back):
After amp turn on and protection relay click:
TP1: -900mV and dropping towards 0
TP3: -700mV and dropping towards 0
About 3 minutes later, voltages were:
TP1: +190mV appearing to level off, but fluctuating slightly up and down.
TP2: +165mV appearing to level off, but fluctuating slightly up and down.
5. Question: It appears that it took a few minutes for the Phono stage to warm up and stabilize. Could this be related to the measurements I saw with C13 and C14 taking a while to charge up? Is it an indicator that some of the other electrolytic caps on this board are also old? Or is it more an indicator (with the fluctuating voltages) that transistors may be suspect?
1. I pulled C23/C24 and tested them. C23: 23.35uF, C24: 23.43uF. So they still look good (should be 22uF). They were both Elna Audio Grade Non-Polarized and both had been glued, with glue on a couple of surrounding resistors that I cleaned up..
2. I pulled C13/C14 and tested them. C13: 1070uF, C14: 970uF. So they appear to be good (should be 1000uF). They were both green Nichicon Muse caps and both had been glued, with glue on a couple of surrounding resistors that I cleaned up. I noticed while testing these caps that it took about 10 seconds to get the final readings, as they appeared to take that long to charge up? Is that normal on larger caps?
3. After re-installing all 4 caps, I went to adjust Phono EQ (via VR1 & VR2). This time I did not switch the Input Selector on the Amp from Aux to Phono, I left it on Aux and found that I still had current throughout the Phono Stage to make the adjustments. I also grounded the Phono Inputs on the back per the Service Manual. What purpose does this serve?
Once again, I was able to dial in +/- 100 mV on both, at least get them in the ballpark, but they both continue to fluctuate constantly and the VR1/2 pots were very sensitive. A slight turn and voltage would shoot up to 800mV very easily. Stopped here for the night.
4. This morning I went to check Phono EQ voltages on TP1 (left) and TP3 (right) and saw the following (I checked the voltages this time without grounding the Phono Inputs on back):
After amp turn on and protection relay click:
TP1: -900mV and dropping towards 0
TP3: -700mV and dropping towards 0
About 3 minutes later, voltages were:
TP1: +190mV appearing to level off, but fluctuating slightly up and down.
TP2: +165mV appearing to level off, but fluctuating slightly up and down.
5. Question: It appears that it took a few minutes for the Phono stage to warm up and stabilize. Could this be related to the measurements I saw with C13 and C14 taking a while to charge up? Is it an indicator that some of the other electrolytic caps on this board are also old? Or is it more an indicator (with the fluctuating voltages) that transistors may be suspect?
1 and 2:
Its probably normal for your cap meter to take some time settling on a reading for large value electrolytic. As an ex service tech I would never take the value of a cap as being any guide to its 'goodness' in just the same way that measuring battery voltage on a DVM gives no clue as to how healthy the battery would be under load.
3:
Grounding (using shorting plugs) does two things. It kills any stray noise pickup which can influence readings and secondly, on input stages with bjt devices (normal transistors so not this one), it helps minimise offset errors due to leaving the input floating and only ground referenced via a high value resistor. The high value resistor (which is R1 here) would cause significant errors. That's only for bjt's though, not FET's which draw essentially zero gate current at DC.
4 and 5:
Honest answer is that I can't say for sure. The fact both channels behave the same suggests its probably not an issue.
Also, because the phono stage is necessarily of high to very high gain means that slight causes of offset appear magnified. Yes, its quite possible the large value of those caps (1000uf) is the cause and you are simply seeing the time needed for the stage to reach equilibrium. You can prove that be replacing them with a 100uf cap and seeing if the time is correspondingly cut by around a factor of 10.
(as an experiment try reading a new cap on your meter. Now add a series resistor to the cap of say 10 or 100 ohms. Does the meter still read the same uF value ?)
Its probably normal for your cap meter to take some time settling on a reading for large value electrolytic. As an ex service tech I would never take the value of a cap as being any guide to its 'goodness' in just the same way that measuring battery voltage on a DVM gives no clue as to how healthy the battery would be under load.
3:
Grounding (using shorting plugs) does two things. It kills any stray noise pickup which can influence readings and secondly, on input stages with bjt devices (normal transistors so not this one), it helps minimise offset errors due to leaving the input floating and only ground referenced via a high value resistor. The high value resistor (which is R1 here) would cause significant errors. That's only for bjt's though, not FET's which draw essentially zero gate current at DC.
4 and 5:
Honest answer is that I can't say for sure. The fact both channels behave the same suggests its probably not an issue.
Also, because the phono stage is necessarily of high to very high gain means that slight causes of offset appear magnified. Yes, its quite possible the large value of those caps (1000uf) is the cause and you are simply seeing the time needed for the stage to reach equilibrium. You can prove that be replacing them with a 100uf cap and seeing if the time is correspondingly cut by around a factor of 10.
(as an experiment try reading a new cap on your meter. Now add a series resistor to the cap of say 10 or 100 ohms. Does the meter still read the same uF value ?)
Thanks for the detailed reply Mooly. Looks like I need to leave unit on for 5-10 minutes and then look to adjust Phono/EQ trimmers after things have settled. I'll give that I try next.
Next step would then be to test out the Phono stage with a TT and see if it affects the outputs again.
(I'll try the cap test with a resistor in series to see what my DMM says)
Next step would then be to test out the Phono stage with a TT and see if it affects the outputs again.
(I'll try the cap test with a resistor in series to see what my DMM says)
No doubt what so ever on that
I've seen and been in that situation countless times. ---------------------------------------------------------------------------------
No conceivable way that it could do unless it was oscillating at some high frequency. The DC blocking caps (C23/24) see to that.
Just hooked up turntable, turned things on, played iPod on Aux for 6 or so songs, all good. Switched to phono and played entire side of an LP. Again, all good. I enabled subsonic filter while playing turntable. I will need to play for extended period to see if there is a heat issue. Maybe from moving Voltage Amp board and EQ/Phono board while working on them, I resolved a short issue, but other than that, I haven't done anything (other than replace C39/C41 on left channel).
So my new DMM has a temperature feature. So I tried it out and checked how hot the bottoms of the heatsinks are getting next to the output transistors and saw 114 degrees F on both sinks. And that was at 1/4 volume. Hot stuff! I was tempted to check temp of the FETs but wonder if that is advisable given their sensitivity to ESR.
So my new DMM has a temperature feature. So I tried it out and checked how hot the bottoms of the heatsinks are getting next to the output transistors and saw 114 degrees F on both sinks. And that was at 1/4 volume. Hot stuff! I was tempted to check temp of the FETs but wonder if that is advisable given their sensitivity to ESR.
then. All you can do is give the unit an extended soak test and see how it behaves. Faults that seem to 'heal up' and where you can't find a definite reason are very frustrating.- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.