Hey folks,
I need some guidance here. I have some experience on class a/b amps, have troubleshooted and restored a lot of pioneers, marantz, realistic, etc...
But this is my first Nikko power amp.
I have replaced all electrolytic caps, and a few faulty components: 4 out of 8 output transistors were shorted, so I replaced all output transistors and a burnt fusible resistor.
2SA1095 -> 2SA1215 (Q733,734,737,738)
2SC2565 -> 2SC2921 (Q731,732,735,736)
2SD882 -> KSD1691 (Q727, 728)
2SB772 -> KSB1151 (Q729, 730)
There is no bias offset adjustment on this amp, just IDLE current adjustment trim pots, but the TP testing points are not showing anything usable.
Power supply seems fine, the rails are providing correct voltage but some of the output transistors are showing incorrect voltages.
On the left channel the negative side of the circuit has displaying some positive voltage on most transistors, Q733 and Q737 bases are 26v, when I presume should be showing a negative value.
On the right channel a similar problem is happening but in the positive side of the circuit...
I simply "copied" the original placements of the transistors (there are no shorts with the heatsink).
I am not leaving it on for long since it can burn something, It is drawing about 1A @ 127V.
I need a peer to review and help me to troubleshoot this issue.
DMM, o-scope, thermal camera, etc are available...
I need some guidance here. I have some experience on class a/b amps, have troubleshooted and restored a lot of pioneers, marantz, realistic, etc...
But this is my first Nikko power amp.
I have replaced all electrolytic caps, and a few faulty components: 4 out of 8 output transistors were shorted, so I replaced all output transistors and a burnt fusible resistor.
2SA1095 -> 2SA1215 (Q733,734,737,738)
2SC2565 -> 2SC2921 (Q731,732,735,736)
2SD882 -> KSD1691 (Q727, 728)
2SB772 -> KSB1151 (Q729, 730)
There is no bias offset adjustment on this amp, just IDLE current adjustment trim pots, but the TP testing points are not showing anything usable.
Power supply seems fine, the rails are providing correct voltage but some of the output transistors are showing incorrect voltages.
On the left channel the negative side of the circuit has displaying some positive voltage on most transistors, Q733 and Q737 bases are 26v, when I presume should be showing a negative value.
On the right channel a similar problem is happening but in the positive side of the circuit...
I simply "copied" the original placements of the transistors (there are no shorts with the heatsink).
I am not leaving it on for long since it can burn something, It is drawing about 1A @ 127V.
I need a peer to review and help me to troubleshoot this issue.
DMM, o-scope, thermal camera, etc are available...
This amp has emitter follower output stages, so all base and emitter voltages of the output transistors should be in the neighborhood of 0V when operating properly. The amp also features "DC servo" nulling, so it doesn't need offset adjustment pots.
Many members advocate Dim Bulb Testers. You've already powered the unit, and found about 125W power draw. That may be excessive, but far less than the 480W, 620VA noted in the specs. Standby power isn't noted. It's possible that any impending damage has already occurred.
What's status of power and protect LEDs?
When and if you're comfortable taking more measurements, I suggest measuring the supply voltages (+/- 58V, and +/- 12.5V for the opamps). Next checks would report output transistor voltages and earlier stages.
Many members advocate Dim Bulb Testers. You've already powered the unit, and found about 125W power draw. That may be excessive, but far less than the 480W, 620VA noted in the specs. Standby power isn't noted. It's possible that any impending damage has already occurred.
What's status of power and protect LEDs?
When and if you're comfortable taking more measurements, I suggest measuring the supply voltages (+/- 58V, and +/- 12.5V for the opamps). Next checks would report output transistor voltages and earlier stages.
Thank you for your reply.
A long waited variac is on the way now.
Here is a run down of the transistors starting with the left channel:
Some of the measurements are not very precise since I am not waiting too long for the voltage to stabilize, but I can repeat where necessary.
True. I do use a DMT in most cases, but I do not have the proper bulb wattage for this amp and the voltages I was getting with the DMT in the circuit were complicating the power supply check, I couldn't tell if they were low because of the DMT or if there was a short bringing them down.
A long waited variac is on the way now.
Here is a run down of the transistors starting with the left channel:
| E | C | B | |
| Q731 | 26 | 56 | 26 |
| Q733 | 25 | -55 | 26 |
| Q735 | 25 | 56 | 26 |
| Q737 | 25 | -56 | 25 |
| Q727 | 27 | 28 | 28.5 |
| Q729 | 26.5 | 28 | 27 |
| Q723 | 27 | 57 | 57 |
| Q725 | 25 | -55.8 | 26 |
| Q719 | 27 | 55.6 | 26.6 |
| Q721 | 26 | 0.13 | 26 |
| Q715 | 27 | 27.5 | 27.5 |
| Q717 | 26 | 56 | 26 |
| Q709 | 60 | 56 | 62 |
| Q713 | -55 | 20 | -54.2 |
| Q707 | 60 | 60 | 62 |
| Q711 | -54.2 | -54.4 | -54.2 |
| IC701 | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
| 22 | 0 | 1.75 | 1.76 | 1.62 | 12.6 |
Some of the measurements are not very precise since I am not waiting too long for the voltage to stabilize, but I can repeat where necessary.
Thanks! That's an impressive amount of data!
I'm not suggesting that IC701 is the only culprit, but many of its voltages look strange. I assume its supply voltages are about +/- 13V? I would expect the pin 5 voltage to be about 1/11 of the Q731 emitter voltage, or about 2.4V; pin 6 should be about 0V. Would you double check pin 1 voltage? It's way beyond the supply rail.
I haven't yet spotted anything in the transistor voltages that jumps out, but I'm still looking. It's interesting that the output voltage is sort of mid-supply. I think I'm going to need to know the voltages on the dual FET, Q701.
I'm not suggesting that IC701 is the only culprit, but many of its voltages look strange. I assume its supply voltages are about +/- 13V? I would expect the pin 5 voltage to be about 1/11 of the Q731 emitter voltage, or about 2.4V; pin 6 should be about 0V. Would you double check pin 1 voltage? It's way beyond the supply rail.
I haven't yet spotted anything in the transistor voltages that jumps out, but I'm still looking. It's interesting that the output voltage is sort of mid-supply. I think I'm going to need to know the voltages on the dual FET, Q701.
I can not help you with this specific amp, but have experience with two integrated Nico amps of this aera.
To tell the truth, these amps have no audiophile sound, even in best "as new" condition. After an A-B test against a cheap, good TDA 3116d2 I decided to remove the amp section and fit some LM3886 modules. The change in sound quality is like night and day. I have experienced that the pre amp stages of that time are often quite neutral (with integrated amps), but the power amps are simply outdated.
Instead of spending a lot of time in repairing and chasing long obsolete parts, it maybe a better option to implant a cheap, good modern amp section. You may be able to use the old transformer and power supply, if voltages match up. Class D is interesting, as you may get a lot more power from them, compared to old class A/B.
As these amps have no real historic value, such a modernized version should be OK, even with purists.. Like fitting a clean, efficent engine into an old car that is used as a daily driver.
Some hate these "retro" amps, but to me it is legitamate, as you do not change the look. Something like green, recycled HIFI, even helping the planet.
To tell the truth, these amps have no audiophile sound, even in best "as new" condition. After an A-B test against a cheap, good TDA 3116d2 I decided to remove the amp section and fit some LM3886 modules. The change in sound quality is like night and day. I have experienced that the pre amp stages of that time are often quite neutral (with integrated amps), but the power amps are simply outdated.
Instead of spending a lot of time in repairing and chasing long obsolete parts, it maybe a better option to implant a cheap, good modern amp section. You may be able to use the old transformer and power supply, if voltages match up. Class D is interesting, as you may get a lot more power from them, compared to old class A/B.
As these amps have no real historic value, such a modernized version should be OK, even with purists.. Like fitting a clean, efficent engine into an old car that is used as a daily driver.
Some hate these "retro" amps, but to me it is legitamate, as you do not change the look. Something like green, recycled HIFI, even helping the planet.
I have some insight.
The right side of Q701 is heavily on, which turns Q705 off, Q703 on. So Q707 should be on, and Q709 off. Yet Q709 collector is 56V--- very suspicious. Q723 base is 57V (probably variation during measurement) but emitter is 27V. BINGO! Q723 is obviously failed, but I hadn't spotted it earlier.
Before replacing Q723, I suggest a couple more measurements with meter reference on the 60V rail. Would you report base and collector voltage of Q709. And the voltage across R747 should be 0.
The right side of Q701 is heavily on, which turns Q705 off, Q703 on. So Q707 should be on, and Q709 off. Yet Q709 collector is 56V--- very suspicious. Q723 base is 57V (probably variation during measurement) but emitter is 27V. BINGO! Q723 is obviously failed, but I hadn't spotted it earlier.
Before replacing Q723, I suggest a couple more measurements with meter reference on the 60V rail. Would you report base and collector voltage of Q709. And the voltage across R747 should be 0.
Gosh, I have stared at this table for two days without figuring that out.
Q706 (63.1v reference) C 6.06v B 2.7v
R747 = 0v
Q706 (63.1v reference) C 6.06v B 2.7v
R747 = 0v
Q706 measurements look promosing.
I had to mentally signal trace the path noted above before I spotted the defect.
If you're still stuck on the other channel, we can try a similar attack.
Good luck!
I had to mentally signal trace the path noted above before I spotted the defect.
If you're still stuck on the other channel, we can try a similar attack.
Good luck!
Both channels burned usually points to something gone wrong in the power supply.
Another option is high load for a long time with a magazine on top of the vents, but most better amps have a temperature limiter for this foolishness
Another option is high load for a long time with a magazine on top of the vents, but most better amps have a temperature limiter for this foolishness
@BSST
Hello!
Here are my subs:
Q723 Q724 - 2SC2275 -> TTC004B
Q725 Q726 - 2SA985 -> TTA004B
Q719 Q720 - 2SC1815 -> KSC1815YTA
Q721 Q722 - 2SC1015 -> KSA1015YTA
I have tested and replaced all bad transitors and respective pairs on both channels on the driver PCB.
I have also tested the FETs (with the Peak tester) and diodes, they all passed.
There is something still wrong and it seems to be in the left channel.
When I turn the amp on with the DBT, light goes bright, then dims, then goes bright again, then dims, in regular cycles of about 1 second.
Voltage rails vary with the cycle.
The problem goes away if I disconnect the cable between the left channel and the driver pcb. The problem keeps happening if I connect the left and disconnect the right channel.
At this point I am not sure if there is something else wrong in the driver PCB, unless the Peak tester is not picking it up (no pun intended).
If the driver PCB is ok the problem would be either in the output transistors or in the power supply.
I decided to share this before I start chasing ghosts.
There is a chance my output transistors are fake, I have never worked with this supplier before.
Hello!
Here are my subs:
Q723 Q724 - 2SC2275 -> TTC004B
Q725 Q726 - 2SA985 -> TTA004B
Q719 Q720 - 2SC1815 -> KSC1815YTA
Q721 Q722 - 2SC1015 -> KSA1015YTA
I have tested and replaced all bad transitors and respective pairs on both channels on the driver PCB.
I have also tested the FETs (with the Peak tester) and diodes, they all passed.
There is something still wrong and it seems to be in the left channel.
When I turn the amp on with the DBT, light goes bright, then dims, then goes bright again, then dims, in regular cycles of about 1 second.
Voltage rails vary with the cycle.
The problem goes away if I disconnect the cable between the left channel and the driver pcb. The problem keeps happening if I connect the left and disconnect the right channel.
At this point I am not sure if there is something else wrong in the driver PCB, unless the Peak tester is not picking it up (no pun intended).
If the driver PCB is ok the problem would be either in the output transistors or in the power supply.
I decided to share this before I start chasing ghosts.
There is a chance my output transistors are fake, I have never worked with this supplier before.
Earlier, you mentioned a Variac was in route? Do you have an oscilloscope? I always worry about unobservable oscillation that can provoke inordinate current draw.
I suggest setting VR701 to minimum resistance to reduce bias current. That might have some effect upon the DBT behavior.
You mention disconnecting a cable between left driver board and the output stages. Does this connector pass the 9 signals that are grouped within graphic brackets in the schematic? Do I infer correctly that a separate cable passes +/- 63.1V and that it remains powered, and that the right channel remains fully connected? If so, this is an interesting configuration. What "output" voltage do you observe at the junction of R719 and R751, and what is voltage at IC701 pin 7?
I suggest doing some testing of the right channel. If all is well with the right channel, that lends confidence to returning to left channel.
I suggest setting VR701 to minimum resistance to reduce bias current. That might have some effect upon the DBT behavior.
You mention disconnecting a cable between left driver board and the output stages. Does this connector pass the 9 signals that are grouped within graphic brackets in the schematic? Do I infer correctly that a separate cable passes +/- 63.1V and that it remains powered, and that the right channel remains fully connected? If so, this is an interesting configuration. What "output" voltage do you observe at the junction of R719 and R751, and what is voltage at IC701 pin 7?
I suggest doing some testing of the right channel. If all is well with the right channel, that lends confidence to returning to left channel.
Variac hasn't arrived yet. new ETA is next week acording to the seller. Yes I do have an o-scope.
I tried that earlier and I was under the assumption that the IDLE current was too high or too low, but no notable change in the behavior.
Correct. There are three sockets, one for each channel and the middle for power supply (pins 20-23).
With all cables connected.
IC701 PIN 7 = 0.06v and stable
R719 = 0~1v RMS : 10vpp! (picture attached) period is about 1.5s
R751 = 0.2~0.5 RMS : 0.4 vpp (picture attached) same 1.5s period
Thank you for all your help so far!
Thanks! Interesting pics.
I probably didn't explain clearly what I was suggesting. If I understand correctly, if you unplug the left channel output stages, the DBT stops its limit-cycle behavior and becomes stable? Is it then safe to remove the DBT and test performance of the right channel? If all is proper with the right channel, there's reason to think transistor substitutes ought to serve in the left channel as well.
Also of interest is how the left channel driver board behaves even though the output stage is disconnected. That is, with output stage disconnected, what is R719 voltage and the IC701 pin 7 voltage? If they are well behaved despite absence of the output transistors, that would suggest DBT cycles are due to trouble in the output devices and driver circuit is innocent.
I probably didn't explain clearly what I was suggesting. If I understand correctly, if you unplug the left channel output stages, the DBT stops its limit-cycle behavior and becomes stable? Is it then safe to remove the DBT and test performance of the right channel? If all is proper with the right channel, there's reason to think transistor substitutes ought to serve in the left channel as well.
Also of interest is how the left channel driver board behaves even though the output stage is disconnected. That is, with output stage disconnected, what is R719 voltage and the IC701 pin 7 voltage? If they are well behaved despite absence of the output transistors, that would suggest DBT cycles are due to trouble in the output devices and driver circuit is innocent.
I probably jumped right in without really thinking.
Correct, it even comes out of protection.
That is assuming they are all equal. They are showing very similar hFE.
But, would they draw more current than the originals? which would be chocking the power supply?
I know... maybe ghosts.
With the left channel disconnected (and out of DBT):
R719 starts at 1.1v when switched on and then stabilizes at 2mV
R720 is always around 1~2mV
R751 is at 0.85v
R752 is at 0.45v
IC701 pin 7 starts -0.6v then stabilizes at +60mV
Last edited:
I agree that assumption of equal parts is a leap.
Re your latest data, I want to make sure I'm not ignoring a detail. I mentioned the junction of R719 and R751 (i.e. same node as emitters of Q719 and Q721), so I would expect the same settled voltage on those two resistors. So you might double check. But I'm very pleased about resulting DBT behavior when output stages are disconnected.
I'm still ruminating about next test. I believe we're in an improved position because we can attack outputs and driver board more independently.
Nope you are not, I was ignoring the word "junction" and measuring across the resistors.
Q719 E = 2mv
Q721 E = 2mv
If I'm interpreting the schematic correctly, with cable between drive board disconnected, there is still +/- 55VDC applied to the collectors of the output transistors, Q731, Q735, and their PNP complements? (Ref Des are unclear in drawing.). Would you confirm presence of B+/B- ? And you report that in this state, DBT and protection are OK. If so, I think this suggests output transistors are OK and points back to Driver and suspicion of Q723 and or Q725.