NAD C325BEE Will not power.

[paul_g_787]

Hi Everyone. I have here a NAD C325BEE amplifier which will not power and I am a little stuck.

Service Manual: https://elektrotanya.com/nad_c325bee-sm.pdf/download.html

It was dropped off to me with the report being that it was in use, then started distorting and crackling, then it switched off and had an amber light and would not come back on.

So. I firstly found an issue where the main capacitors for the +46/-46 DC rail were bulging (C517 and C 518) and there was +39V and 0V respectively. As reported the amplifier had an amber power light but would not turn on.

I have replaced these now and I am getting +40V / -40V, which is still a bit low...  Not sure here..... But progress...
I also replaced C523 and C524 as their ESR was pretty high. These are for the +37/-37V rails which measure just fine.

It is my belief these capacitors failed / were in the process of failing as they are exactly next to the main heat-sink... a quick search shows this is a common failure in this model.

So that being sorted (I think) I now have the issue where there is no power light at all. Not even amber. And the amplifier won't respond to the power button.

The service manual check sheet (Page 24) says to check the 5V supply circuit on the power board. However there is nothing labelled 5V at all on the power board schematic. I am assuming this would be the pin labelled "TRIG" ??  Confused here... 

Power supply schematic on page 19 of the SM for those playing along.
I have attached a copy of this with my current readings written on it.

All the bits in green I am sure are good. 

All the red bits are where I am confused. 

The TRIG pin is at 0V
The AC DET pin is at 5V AC and -25V DC
The I-V pin is at 0V

What are these ping exactly. I am not sure.?? 

So my questions are, firstly what should TRIG, AC DET, and I-V be measuring?
And also why would the +46 and - 46V rails only measuring +40 and -40?

Can anyone make any suggestions here?

[up8051]

5V is made by IC45 AMS1117 form 8V (CB53 pin 6,7)

IC45 schematic page 17 right-up corner.

[paul_g_787]

5V is made by IC45 AMS1117 form 8V (CB53 pin 6,7)

IC45 schematic page 17 right-up corner.

Aha well spotted sir!! Even though the SM says "on power board"... No wonder I didn't see that!

I will test that part of the circuit next.

[Kim Christensen]

All the red bits are where I am confused. 

The TRIG pin is at 0V
The AC DET pin is at 5V AC and -25V DC
The I-V pin is at 0V

What are these ping exactly. I am not sure.?? 

So my questions are, firstly what should TRIG, AC DET, and I-V be measuring?
And also why would the +46 and - 46V rails only measuring +40 and -40?

It looks to me like the rails marked 46V are actually capable of being switched between 40V and 51V depending on the voltage level on TRIG. (Voltages approximate)
When TRIG is zero, then the 28.5VAC transformer windings are suppling the power via the D51 bridge rectifier.
But when TRIG goes "logic high" then both the SCRs (D531 and D532) are turned on which connects the D55 bridge rectifier outputs (which is fed from the 36.4VAC windings) to the filter caps (C517 & C518) and thus the voltage on "46V" rails (CB54) increases from 40V to 51V.
I would guess this is supposed to happen when the user turns up the volume and the amplifier needs to put out more power.

[feedback.loop]

All the red bits are where I am confused. 

The TRIG pin is at 0V
The AC DET pin is at 5V AC and -25V DC
The I-V pin is at 0V

What are these ping exactly. I am not sure.?? 

So my questions are, firstly what should TRIG, AC DET, and I-V be measuring?
And also why would the +46 and - 46V rails only measuring +40 and -40?

It looks to me like the rails marked 46V are actually capable of being switched between 40V and 51V depending on the voltage level on TRIG. (Voltages approximate)
When TRIG is zero, then the 28.5VAC transformer windings are suppling the power via the D51 bridge rectifier.
But when TRIG goes "logic high" then both the SCRs (D531 and D532) are turned on which connects the D55 bridge rectifier outputs (which is fed from the 36.4VAC windings) to the filter caps (C517 & C518) and thus the voltage on "46V" rails (CB54) increases from 40V to 51V.
I would guess this is supposed to happen when the user turns up the volume and the amplifier needs to put out more power.

I believe this is so-called NAD PowerDrive. It is supposed to switch from the higher voltage rails to the lower voltage rails when more current is needed.

[paul_g_787]

OK so had a look at the 5V rail and I am getting 0V!!

I measured at the capacitor C444 and I am getting 10.37V DC with no obvious AC ripple. 

The strange thing it that the amp came to life after probing the AMS1117-5.0 at pin 2...... I had 4.96V to GND and the amp switched on and worked! Then after turning the amp off and on it is now dead again and I have 0V.

I am wondering maybe a bad joint on the IC45?

I am going to see if I can get my ESR meter across C445 as well.

[paul_g_787]

OK something now.

I noticed some residual crud.. maybe flux. around IC45's pins. I just cleaned this off with IPA and BAM!!! the 5V is back! And then it went back to 0 again.
So I cleaned it again and now it is back and has stayed back. I have tried prodding the IC and shocking the pcb a little but it appears that whatever it was has now been dislodged. Maybe some solder spatter?

Unfortunately I cannot get to C445 to check it without de-soldering the +/-46V wires which go via tubes in the heat-sink to the power transistors. They are double-sided soldered.... Not sure I want to tamper with them.
However C445 does not appear to have bulged or oozed in any way and as the 5V is reading OK now (and there is no ripple I can see) I think it is fine.

But.... When I press the power button... Nothing. The LED is solid Orange. 
So I checked all my rails and these look OK, however the +/-46 and +/- 27 Rails are reading the same? I checked for direct shorts (with probes both ways) and there are none. But seems suspicious...... 

+37V = +39.3V
-37 = -39.1V
+46V = +39.3V
-46V = -39.1V
+5V = +5.0V

However the 18V rails are measuring:
+18V = 0V
-18V = -1.1V

The next thing the Troubleshooting guide says is "Is the standby signal low level?" I assume this is at R401/C41 in the supply circuit? What voltage is this supposed to read here?

[Kim Christensen]

I noticed some residual crud.. maybe flux. around IC45's pins. I just cleaned this off with IPA and BAM!!! the 5V is back! And then it went back to 0 again.

There could still be a cold solder joint that you just can't see. I would reflow IC45's pins even though the 5V seems OK. It may still be intermittent and causing problems on power up even though you see it stable later when you probe it.

[paul_g_787]

I noticed some residual crud.. maybe flux. around IC45's pins. I just cleaned this off with IPA and BAM!!! the 5V is back! And then it went back to 0 again.

There could still be a cold solder joint that you just can't see. I would reflow IC45's pins even though the 5V seems OK. It may still be intermittent and causing problems on power up even though you see it stable later when you probe it.

Well it was tricky but I managed to wick the old solder off the IC and apply new solder.
(please excuse the cotton swab fibres left on the board, I promise I will clean them off even though they are non conductive).

As you can see it is in a tricky location to get to and there is no way to remove the PCB without de-soldering the red, white and black power wires going via the heat-sink tubes.

I put some Kapton tape on the speaker wires just in case I accidentally contacted them with the iron but I did manage it without making contact.

Unfortunately it has made no difference to the operation but at least we know that IC45 now has good connections.

I have a stable 5V and the amber standby light is on, but pressing the button does nothing.

So it might be something with the standby circuit? I will try and measure the standby signal voltage and see what it reads.

Although the standby is supposed to be on until the button is pressed.. so something is stopping it coming 'out of standby'

[Kim Christensen]

The next thing the Troubleshooting guide says is "Is the standby signal low level?" I assume this is at R401/C41 in the supply circuit? What voltage is this supposed to read here?

When the unit is NOT in standby, I would say the "standby signal" at R401/C41 would be less than 0.6V or so. Once standby goes above apx 1.4V, current will flow through R104 and Q43 will start to turn on more. If the standby voltage goes high enough, then Q43 will saturate and the 18V supplies will turn off.

[paul_g_787]

OK so I took some measurements at connector CB34 which is conveniently labelled.

SOFT CLIPPING is at 0.001V
MUTE is at 0.004V
STANDBY is at 0.001V
PROTECT is at 0.002V
+5V is at 4.99V

But still pressing the standby button does not power on the unit. The Amber LED is lit solid.

[paul_g_787]

According to the troubleshoot in the SM I should check the protect circuit.

[paul_g_787]

OK so quick look at the protect circuit shows something.

Page 17 of the service manual for schematic.

IC44 Pin readings as follows:
1 = 0.01V
2 = 0.01V
3 = 0.01V
4 = -22.5V
5 = 0.00V
6 = -0.35V
7 = 0.15V
8 = 0.15V

Interestingly though IC44 is actually a PCB and not an IC but if you look at Page 37 of the service manual is details changing the IC44 for a protect module and shows the new schematic and some other component changes.

[Kim Christensen]

Page 17 of the service manual for schematic.
IC44 Pin readings as follows:

Unfortunately, due to the changes in "IC44" and other components, you'll have to ignore most of those protection circuit voltages listed on page 17.

A good example would be pin 4 of "IC44": Because D513 has been reversed, now instead of +2V you'll have a negative voltage on the "AC DET" line which is the new normal. So reading -22.5V on pin 4 is probably not unreasonable.

The one that looks wrong to me is pin 8 of "IC44". There should be around 3.6V or so there. Is R441 OK? Also, with power off, what does an ohmmeter and "diode check" show when measuring between pins 5 & 8 of "IC44"?

[paul_g_787]

The one that looks wrong to me is pin 8 of "IC44". There should be around 3.6V or so there. Is R441 OK? Also, with power off, what does an ohmmeter and "diode check" show when measuring between pins 5 & 8 of "IC44"?

R441 measures 4.6k ohms so looks good!

With the meter + on pin 5 and - on pin 8
0.712 in Diode test mode
3.8k ohm amd slowly increasing

With the meter - on pin 5 and + on pin 8
1.68 and slowly increasing in Diode test mode
8.68k ohm and slowly increasing

Why is the value of these readings going up when I leave the meter connected a while? Charging capacitors maybe?

Just powered the amp on again to check pin 8 of "IC44" and it measures 0.15, however... when I press the power button this voltage on pin 8 goes up to 4.68V which then settles to 4.23V after a few seconds.

So before pressing the power button "IC44" Pin readings as follows:
1 = 0.01V
2 = 0.01V
3 = 0.01V
4 = -22.5V
5 = 0.00V
6 = -0.35V
7 = 0.15V
8 = 0.15V

After pressing the power button "IC44" Pin readings as follows:
1 = 0.01V
2 = -0.02V
3 = 0.01V
4 = -22.2V
5 = 0.01V
6 = 0.01V
7 = 1.43V
8 = 4.23V

[paul_g_787]

Just went to take measurements again to confirm and my 5V rail was dead again!!!!   
10.3V at input and 0V at output of IC45 (the 5V regulator)

Again.. I cleaned IC45's pins with IPA and dried it with cotton swabs. Powered it on and it is working.......... This is very puzzling.... 

Anyway I just took the measurements of IC44 again and got the same values before and after power button press, and my 5V rail is working still.

[Kim Christensen]

Just powered theamp on again to check pin 8 of "IC44" and it measures 0.15, however... when I press the power button this voltage on pin 8 goes up to 4.68V!!

I wonder if D431 (3.6V zener) is actually missing since the voltage rises to 4.68V... This would make sense because there is a 4.7V zener (D13) on the "IC44" PCB which is effectively in parallel with D431.
It is possible that Q412 is being turned on by the standby input via R437 or maybe the overtemp sensor (R438/439 combo). If Q412 is on, this would drop the voltage at pin 8 of "IC44" to near zero.

[Kim Christensen]

Again.. I cleaned IC45's pins with IPA and dried it with cotton swabs. Powered it on and it is working.......... This is very puzzling...

It might be that the cooling effect of evaporating alcohol is making that regulator work properly until it heats up and fails.

[paul_g_787]

Again.. I cleaned IC45's pins with IPA and dried it with cotton swabs. Powered it on and it is working.......... This is very puzzling...

It might be that the cooling effect of evaporating alcohol is making that regulator work properly until it heats up and fails.

Good thought, however I left the amplifier over-night (approx 11 hours) and the next day it still worked fine. Would the alcohol not have evaporated overnight? 

Perhaps whatever the fault elsewhere is, is drawing too much current from the regulator and it is getting hot and switching off?

If this happens again I will power it off at the mains for a while and refrain from cleaning it with IPA and see if it "cools down" and works again to confirm if IC45 is overheating.

[paul_g_787]

Just powered theamp on again to check pin 8 of "IC44" and it measures 0.15, however... when I press the power button this voltage on pin 8 goes up to 4.68V!!

I wonder if D431 (3.6V zener) is actually missing since the voltage rises to 4.68V... This would make sense because there is a 4.7V zener (D13) on the "IC44" PCB which is effectively in parallel with D431.
It is possible that Q412 is being turned on by the standby input via R437 or maybe the overtemp sensor (R438/439 combo). If Q412 is on, this would drop the voltage at pin 8 of "IC44" to near zero.

Interestingly D431 is present on the schematic but not on the PCB layout on page 20 in the service manual....?
I can't see this Diode on the real PCB but not sure where to look.

[Kim Christensen]

Perhaps whatever the fault elsewhere is, is drawing too much current from the regulator and it is getting hot and switching off?

Could be. Though putting your finger on it would probably confirm overcurrent fairly easily. You could put a clampon ammeter (Or break the lead with regular VOM) in series with the 8 VAC wire that feeds D430 (wires coming off CB47) and see what changes when you lose the 5V.

Interestingly D431 is present on the schematic but not on the PCB layout on page 20 in the service manual....?
I can't see this Diode on the real PCB but not sure where to look.

Looks like a "typo" in their documentation. It wouldn't make sense to have both zeners installed.

[paul_g_787]

Perhaps whatever the fault elsewhere is, is drawing too much current from the regulator and it is getting hot and switching off?

Could be. Though putting your finger on it would probably confirm overcurrent fairly easily. You could put a clampon ammeter (Or break the lead with regular VOM) in series with the 8 VAC wire that feeds D430 (wires coming off CB47) and see what changes when you lose the 5V.

Interestingly D431 is present on the schematic but not on the PCB layout on page 20 in the service manual....?
I can't see this Diode on the real PCB but not sure where to look.

Looks like a "typo" in their documentation. It wouldn't make sense to have both zeners installed.

Yes OK if the 5V fails again I will check if the regulator IC45 is hot, then let it cool off, then see if it works again after.

And yes I am suspecting that the Zener is not actually there either.

So what is the next step? I can't get to the IC45 to remove it at all without causing a lot of work!

Perhaps there is a short circuit somewhere from the 5V rail which is causing both my faults. - overheating (possibly) IC45 and also the protect kicking in? maybe?

Ouch my brain on this one hehe

[Kim Christensen]

So what is the next step? I can't get to the IC45 to remove it at all without causing a lot of work!
Perhaps there is a short circuit somewhere from the 5V rail which is causing both my faults. - overheating (possibly) IC45 and also the protect kicking in? maybe?

It looks like IC45 is the only thing fed from the 8VAC winding on the transformer. (via D420 and C444)... The good news is, there is a fuse (F55) on the 8 VAC side of D420 on the power board. If you have a 2nd meter, you could flip this fuse up and insert the meter as an AC milliammeter and note what the current is during normal operation. Then wait and see how the current changes when you lose the 5VDC.
If the current dramatically increases, then you've most likely got a short on the 5V output of IC45. (Tiny chance IC45 is bad. More likely some capacitor on the 5V line somewhere else)
But if current drops almost to zero, then the regulator is going open and is faulty.
If you don't have a 2nd meter, measure the AC voltage drop across the fuse (will be in AC millivolts) during normal operation (write it down) and compare to when the fault shows up. ie: Use the fuse as a pseudo "current" shunt.

I had a look at the AMS1117-5.0 datasheet and it is rated for 1A. F55 is rated at 0.5A so you'd think that if the output gets shorted that the fuse would blow. Though IC45 has thermal limiting so current may be less than 1A if it gets very hot.

[paul_g_787]

So what is the next step? I can't get to the IC45 to remove it at all without causing a lot of work!
Perhaps there is a short circuit somewhere from the 5V rail which is causing both my faults. - overheating (possibly) IC45 and also the protect kicking in? maybe?

It looks like IC45 is the only thing fed from the 8VAC winding on the transformer. (via D420 and C444)... The good news is, there is a fuse (F55) on the 8 VAC side of D420 on the power board. If you have a 2nd meter, you could flip this fuse up and insert the meter as an AC milliammeter and note what the current is during normal operation. Then wait and see how the current changes when you lose the 5VDC.
If the current dramatically increases, then you've most likely got a short on the 5V output of IC45. (Tiny chance IC45 is bad. More likely some capacitor on the 5V line somewhere else)
But if current drops almost to zero, then the regulator is going open and is faulty.
If you don't have a 2nd meter, measure the AC voltage drop across the fuse (will be in AC millivolts) during normal operation (write it down) and compare to when the fault shows up. ie: Use the fuse as a pseudo "current" shunt.

I had a look at the AMS1117-5.0 datasheet and it is rated for 1A. F55 is rated at 0.5A so you'd think that if the output gets shorted that the fuse would blow. Though IC45 has thermal limiting so current may be less than 1A if it gets very hot.

I did exactly that. Removed F55 and connected my ammeter to the Fuse holder.

On power up from mains it draws 44mA of power.
After pressing the power button this goes up to 46mA.

Left it running for 10 mins with no issue. Still at 46mA. Haven't lost the 5V again yet.

[paul_g_787]

OK, well I haven't been able to lose the 5V rail again yet so don't know what's happening with that. it has been on for a few hours today with no problem.

Still not sure what to do next with "IC44" and the protect circuit.

What would trigger the protect circuit in this amplifier?

[Kim Christensen]

What would trigger the protect circuit in this amplifier?

Overheating (R439)
Standby being active (via R437)
Output overcurrent. DC bias not being zero or failing to get to zero soon enough after power on. (Q12/1, Q11, Q21, etc)

[paul_g_787]

What would trigger the protect circuit in this amplifier?

Overheating (R439)
Standby being active (via R437)
Output overcurrent. DC bias not being zero or failing to get to zero soon enough after power on. (Q12/1, Q11, Q21, etc)

OK thanks I will take some measurements at these components and see what I find.

[paul_g_787]

What would trigger the protect circuit in this amplifier?

Overheating (R439)
Standby being active (via R437)
Output overcurrent. DC bias not being zero or failing to get to zero soon enough after power on. (Q12/1, Q11, Q21, etc)

Thanks for pointing me in this direction. I took some measurements at the input of the protect circuit I measured the Temperature and Standby Signals at the anodes of D42 and D43.

D42 (Temperature) Is at 0.001V and does not changed after the standby / power button is pressed.
D43 (standby is at 1.2V and changes to 0.000V after the standby / power switch is pressed.

The amplifier however still has an amber LED and does not power on.

So I believe that this proves the standby circuit is working fine and that it is not overheating. 

I also checked the power amp output before the relay (RL12/1 pins 5 and 3) and these are both at -1.4V DC 

I think you may be thinking along the right lines of the DC bias but my knowledge in this area is limited. 
I am a little confused how can I check this on this amplifier?

My +12V, +18V and -18V power supply rails are all still at 0V. Would I be right to assume that these are turned on by the standby circuit but the protect is keeping them turned off?

Edit: I left an important note out. The output relay switch is not switching on.  I notice the protect circuit taps from after the relay switch via R446 and R445, so this must be where the protect circuit "checks" for clipping and DC offset right?

[paul_g_787]

OK so I found something!!



Read the comments, the chap here found that his R45 and R46 were open circuit.

So I checked mine and R46 is fine, 32.8Ω but R45 is open circuit! The pads underneath are a little blackened too.
Edit: I de-soldered R45 and the pads are obliterated! They have de-laminated and disintegrated! It must have got very very hot!
I will have to run some nice neat jumper wires over to these spots but for now I have just tacked the replacement on with some wires hanging out the side of the amp for "testing purposes".

On the schematic it is labelled 33Ω 0.5W FS. What does the FS mean? Is it a fusible resistor I assume?
The parts list specifies 33R, 1/2W, ±5%, FS

[paul_g_787]

OK so I just replaced R45 with a new 33Ω 0.5W Metal Film resistor. Now the problem has changed.

The amplifier turns on after pressing the power button but the power LED is Red and soon after the output relay turns off!  The user manual says this is the protect circuit.

I just had a check of D42 and D43 again (Temperature and Standby Signal). These both stay at 0V still so it is not those.

So I measured the output on both channels before the protect relay and the Left channel has -12.4V DC on it! This must be the fault.

Edit: Was just taking the measurements again and the new R45 fusible resistor blew again! R46 is still fine. So there might be a short somewhere on the left channel causing the -12.4V. But how does R45 blow from this?
Edit 2: Consulting the Troubleshooting guide says if the output DC level is not normal to "check the servo circuit". There is no other word "servo" in the service manual. What is a servo circuit? I haven't come across this term with amplifiers before. (You can tell I am new to this lol).
Edit 3: I just measured the idling current on both channels between TP1/TP2 (Left Channel) and TP3/TP4 (Right Channel). The service manual says it should be 4mV, however the left channel has 14.8mV and the right channel has 9.2mV. Related perhaps?

[Kim Christensen]

the Left channel has -12.4V DC on it! This must be the fault.

Yup. That should trigger the protect circuit!

Edit: Was just taking the measurements again and the new R45 fusible resistor blew again! R46 is still fine. So there might be a short somewhere on the left channel causing the -12.4V. But how does R45 blow from this?

R45 is blowing because of over current on the +18V rail or a fault with the +18V regulator circuitry itself. (Q41, Q43, D49, etc)
With the power off, try ohm metering the +18V rail (across C49 or other convenient points). Hard to say exactly what you should see, but a short would be an obvious fault. Or a reading of around 10 ohms would point toward a shorted C13 or C23. Losing the +18V rail would definitely unbalance the amplifier and cause these types of issues. The +18V goes many places. Even the pre-amp.
If you end up replacing R45 again, put a voltmeter on the +18V rail before powering it on. When you power it on, see if the +18V comes up immediately. If not, power it off again right away. (Or if there is +18V but R45 is getting hot)

Edit 2: Consulting the Troubleshooting guide says if the output DC level is not normal to "check the servo circuit". There is no other word "servo" in the service manual. What is a servo circuit? I haven't come across this term with amplifiers before. (You can tell I am new to this lol).

I would guess that they mean the L&R power amplifiers feedback paths / circuits.

[paul_g_787]

Alright I was thinking it would be a short on the +18V rail. Glad I am thinking along the right lines and thank you, your explanation has made it clearer what is going on here and how the DC is being generated at the output. I am still learning and this is super helpful!

I only had one 33 ohm 0.5W metal film resistor, so I have ordered some more of those.

In the meantime I will perform the checks you suggested that I can do now (and the rest of them when my resistors arrive).

[paul_g_787]

R45 is blowing because of over current on the +18V rail or a fault with the +18V regulator circuitry itself. (Q41, Q43, D49, etc)
With the power off, try ohm metering the +18V rail (across C49 or other convenient points). Hard to say exactly what you should see, but a short would be an obvious fault. Or a reading of around 10 ohms would point toward a shorted C13 or C23. Losing the +18V rail would definitely unbalance the amplifier and cause these types of issues. The +18V goes many places. Even the pre-amp.
If you end up replacing R45 again, put a voltmeter on the +18V rail before powering it on. When you power it on, see if the +18V comes up immediately. If not, power it off again right away. (Or if there is +18V but R45 is getting hot)

Just took a measurement across C49 with my ohm meter and it is 349Ω. Is this particularly low?

I also just measured across C410 (-18V rail) as a comparison and this is 4.48kΩ

I might add that this is while the amplifier is powered off of course.

[Kim Christensen]

Just took a measurement across C49 with my ohm meter and it is 349Ω. Is this particularly low?

Unfortunately that isn't very conclusive... But, since we know that R45 is a 33ohm 0.5W resistor we can conclude that if we pass more than 123mA through it we'll exceed it's power rating. Therefore we know that the current supplied by the +18v regulator will be quite a bit less than 123mA.
So if you have a small bench power supply with current limiting you could feed the +18V regulator with that and set it's current limiting to 60mA or so. See diagram attached.
Then you'd power on your amplifier and then power on the bench power supply and slowly increase it's voltage while keeping an eye on the current draw (And the +18V rail voltage) as you raise the voltage towards +37V. This way you can limit the current and have time to find the fault without things burning up. There are a few resistors that you can measure the voltage across to figure out how much current is going through them: R254, R154, R616, R615

[paul_g_787]

But, since we know that R45 is a 33ohm 0.5W resistor we can conclude that if we pass more than 123mA through it we'll exceed it's power rating.

Oh yes I see: √(0.5W÷33Ω) = 0.1231A Or 123mA

So if you have a small bench power supply with current limiting...

Unfortunately I do not have a power supply with current limiting (it is on my list to get one). Only an old ATX PC power supply I converted into a very basic variable voltage power supply.

I just received my new 33Ω 0.5W Metal Film Resistors today! 

I don't mind sacrificing one or two resistors, they cost pennies. Hopefully nothing else blows up in the meantime though.


So I replaced R45 (and also R46 because I sacrificed R46 in R45's place earlier and it is in resistor heaven) with brand new parts.

I measured the voltage across the resistors in standby, protection mode (for a few seconds at power on) and also at power on state.

As you can R45 (on the +18V rail) is indeed drawing more than it's power rating by quite a bit, but only after the protection turns off.
Although it does appear to be drawing a bit too much current for my liking even in protection mode!

R45 = 32.97Ω
STANDBY  0.00V (0mA / 0mW)
PROTECT  3.76V (114mA / 432mW)
ON             5.23V (158mA / 826mW)


Although R46 (on the -18V rail) is looking good to me. 

R46 = 32.78 Ω
STANDBY  0.10V (3mA / 0.3mW)
PROTECT  2.90V (89mA / 258mW)
ON             2.91V (89mA / 259mW)

The good news is that with the two resistors in circuit I have no DC on the speaker outputs and the protect mode does indeed switch off when it is supposed to.

There are a few resistors that you can measure the voltage across to figure out how much current is going through them: R254, R154, R616, R615

I will check these next and see what the current is.

[paul_g_787]

Just had a measure of R615 and R616 on both CB32 (In Module) and CB33 (Out Module)

These resistors are specified as 33Ω 1/6W (0.167W)
So I calculate these would max out at 71mA (a max 2.34V across the resistor)

All four of these resistors measures 0.49V across them which I calculate as approx 15mA or 7.3mW.
So looking good here.

So I next measured R154 on CB11. This is measuring 10Ω.
Then I measured R254 on CB21. This is measuring 67Ω!!! (mwah mwah mwaaahhh) And it is a little blackened too. Finally a clue!

So the short is likely to be in this module CB21??

I just measured Q22 and Q23, these do not appear to be shorted.

[paul_g_787]

OK so my 5V rail disappeared again while I was checking the voltage across R254. It happened about 3 seconds after I had removed my probes from R254 and was writing the voltage down.

I tried feeling IC45 immediately and it is not even warm. I probed it with the multi-meter temperature probe and it is at 17.1°C and the ambient room temperature is 16.9°C.

So I just cleaned very carefully the pins of IC45 again and made sure the alcohol was dry thoroughly. Powered it on and 5V is back!! This is completely confusing me......


Anyway, I measured 17.87V across R254, which is still measuring 67Ω which is 4.76W!! That is a lot.
This would be 31.93W across the un-damaged rating of 10Ω!!

I tested Q201 and If I put the + multimeter probe on the Collector and the - probe on the Base I get a reading of 0.6V. I repeated this test on Q101 and I get the same measurement.
If I swap the probes so the + is on the Base and the - on the collector I get a reading of 0.2V This seems wrong to me... I repeated this test on Q101 and I get an infinite measurement here.

[paul_g_787]

I managed to remove the whole module. Here is a close up photo of it.

A couple of spots look like they got pretty hot. R219 doesn't look too good (under capacitor top right), neither does R205 (under capacitor top left) and of course R254 (middle right).

[paul_g_787]

Just tested the components on this module.

Green components appear to be fine.
Red are not testing as they should be.
Orange components are ones I was getting inconclusive measurements on in circuit.

[paul_g_787]

Just found this thread. Interestingly the last poster is also having issues with the same module as mine.
https://www.diyaudio.com/community/threads/nad-325bee-will-not-turn-on.245053/

[Per Hansson]

Maybe it's not of great help but I repaired this exact model of unit a few months back.
C110, C111,C210,C211 where completely open, they sit right behind the module you have removed between some large power resistors that run around 100°C
Also the SMD caps on those modules where completely shot too, so they where replaced also.
The actual main fault of my unit was that Q213 KSA1220 and Q214 KSA1943 where shorted.

One thought that occurred to me when you bypassed that fuse and the unit worked fine:
Have you ruled out that the fuse F55 is not actually intermittent?
Maybe the wire inside does not make contact when it gets hot and that is why your 5v fails?

[paul_g_787]

Maybe it's not of great help but I repaired this exact model of unit a few months back.
C110, C111,C210,C211 where completely open, they sit right behind the module you have removed between some large power resistors that run around 100°C

Just removed these four capacitors and tested them on my capacitance and ESR meter and they are perfectly OK on my unit.

Also the SMD caps on those modules where completely shot too, so they where replaced also.

The 22µF ones on both modules measure just fine in circuit.
The 10µF ones I can't get a good reading in circuit but all four are behaving the same so these are possibly OK.

The actual main fault of my unit was that Q213 KSA1220 and Q214 KSA1943 where shorted.

I just tested the continuity on these transistors and they appear alright on my unit. No direct shorts between the pins.

One thought that occurred to me when you bypassed that fuse and the unit worked fine:
Have you ruled out that the fuse F55 is not actually intermittent?
Maybe the wire inside does not make contact when it gets hot and that is why your 5v fails?

This is a good point and I could order up some to test, however it would not explain the common solution is to clean the IC45 pins every time and this makes it work fine.
I am wondering if somehow a short is occurring between the pads but I cannot see anything likely in my magnifier.

I am genuinely really stuck on what to try with this amplifier next.

[Kim Christensen]

Just tested the components on this module.
Green components appear to be fine.

When you are checking the transistors, I assume you are using the diode check function on your multimeter. If the BE and BC junctions test OK, make sure you also check between CE (Collector and Emitter) for a short.(Many times I've seen BE & BC check OK but CE is shorted) Also, the right resistor in parallel with a junction can make it look like it is good, but actually open. Other times, a low value resistor in parallel will make it look bad. I always treat a "pass" diode check on a transistor junction as a "maybe it is ok".

You've obviously got multiple faults on this PCB. But double check C23 as well because that seems to be a path that would burn out R254. Especially since Q24, Q201, and R215 visually seem to have escaped being overheated.

[paul_g_787]

When you are checking the transistors, I assume you are using the diode check function on your multimeter. If the BE and BC junctions test OK, make sure you also check between CE (Collector and Emitter) for a short.(Many times I've seen BE & BC check OK but CE is shorted) Also, the right resistor in parallel with a junction can make it look like it is good, but actually open. Other times, a low value resistor in parallel will make it look bad. I always treat a "pass" diode check on a transistor junction as a "maybe it is ok".

Yes I am doing BE, BC. I am getting 0.6ish volts across the junctions. The probes the other way I get OL open circuit. CE I get OL, open circuit in both directions.

You've obviously got multiple faults on this PCB. But double check C23 as well because that seems to be a path that would burn out R254. Especially since Q24, Q201, and R215 visually seem to have escaped being overheated.

Indeed! This is a tricky one for me as I have never had to look this in-depth before at a power amp. This is quite a complicated looking design to me too.

I will see if I can remove C23 and check it with my meters out of circuit.

Do you think I should replace R254? I have some through hole resistors, I could temporarily install one for now and order some SMD ones later.

[paul_g_787]

C23 is measuring 9.99uF with 19.6 ohms ESR. This ESR seems a bit high to me.

C24 is 9.96uF with 9.8 ohm ESR.

[paul_g_787]

OK so I replaced R254 with a new 10Ω resistor. The old one was measuring 72Ω out of circuit and is clearly labelled "100" 10Ω.

I also removed R223 which was measuring 39K in circuit, however once removed it was measuring bang on 47K!! So I have re-installed it. (I really hate SMD resistors, they are so fiddly).

Is this a clue as to the location of the fault maybe?

[Kim Christensen]

I tested Q201 and If I put the + multimeter probe on the Collector and the - probe on the Base I get a reading of 0.6V. I repeated this test on Q101 and I get the same measurement.
If I swap the probes so the + is on the Base and the - on the collector I get a reading of 0.2V This seems wrong to me... I repeated this test on Q101 and I get an infinite measurement here.

Yes, I think you should revisit this. Q201 could be bad. I can't see why the BC junction would measure 0.2V when forward biased.

Yes I am doing BE, BC. I am getting 0.6ish volts across the junctions. The probes the other way I get OL open circuit. CE I get OL, open circuit in both directions.

Yup, sounds good to me.

Check: R230 on the mainboard because it feeds the +46V to the module you took out. (R219 is darkened)
To better understand this amplifier, I've highlighted some of the signal paths. I guess they drew it in this weird manner because the input stage of the amp is on a daughter board. Maybe it'll help others too:

[paul_g_787]

Thank you for that, it is really helpful.

I think the weird layout it throwing me off.

You are right on the money, R230 is looking a bit crusty and it is measuring 18Meg. Attached is a picture.

R130 on the other channel is bang on 82 ohms.

[Per Hansson]

C23 is measuring 9.99uF with 19.6 ohms ESR. This ESR seems a bit high to me.

C24 is 9.96uF with 9.8 ohm ESR.

Yea that is bad, they are the same type of caps so their ESR should be identical.

You are right on the money, R230 is looking a bit crusty and it is measuring 18Meg. Attached is a picture.

R130 on the other channel is bang on 82 ohms.

That is really bizarre how C211 could be ok if that nearby resistor got so hot that it overheated and failed.
I noticed those caps are black on your unit: in mine all original caps where blue and brown.
One cap had been replaced previously in my unit: C47, and it was black too, perhaps someone worked on your unit before also?

[paul_g_787]

Yes I thought that strange too. Maybe if the ESR has gone up due to getting hot?

The owner of the amp said they have had it from new and it has been in his home studio the whole time and had no faults before.

Then it suddenly went all distorted and protection came on.

Then he brought it to me.

[paul_g_787]

So I have replaced C23 and C24 now as I was not happy with their ESR being so high and so different from one another.
And I replaced R254 the other day.
And I also had previously replaced R45.
So these are all brand new parts now.

I have ordered some resistors to replace R219 and R230.

I also previously replaced C517, C518, C523, C524 and C444 as they all had very high ESR and C517, C518 had bulged and were reading waaay low (13.2mF and 12.7mF).


I have probed over and over again all the parts on the SMD-L module which I have removed and I cannot find any shorts on any of the transistors or the diodes.
All the other resistors on the module seem to measure just fine, as do the two electrolytic caps C203 and C204.

I am struggling to find why R254, R219 and R230 and R45 had all burned out. Has anyone got any other suggestions?
Maybe to do with the power supply capacitors I previously replaced? I am seriously guessing here. 

Something must have drawn too much current as mentioned, but what?

Or is it possible the resistors all just failed? It doesn't seem likely to me. 

With the SMD-L module removed, I just took some measurements in the 'power on' state.

I just checked all my voltage rails and they all appear OK:
+46 = 50V
-46 = -50V
+37 = 39.4V
-37 = -39.3V
+18 = +18.3V
-18 = -18.3V
+5 = 4.99V


I just measured the voltage across R45 and R46 again.
I get 2.75V (84mA / 230mW)  across R46 consistently.

However with R45 I get 3.54V (379mW), then the voltage quickly jumps to 4.0V (485mW) then the protect kicks in right away (probably because the SMD-L module is missing) and the voltage returns to 3.54V
Does this current seem high? or is this good? It is under the 500mW limit. 

[SilverSolder]

Is it possible to disconnect power to one of the two channels, and see if one or the other side will come up on its own?

[Kim Christensen]

I just measured the voltage across R45 and R46 again.
I get 2.75V (84mA / 230mW)  across R46 consistently.
However with R45 I get 3.54V (379mW), then the voltage quickly jumps to 4.0V (485mW) then the protect kicks in right away (probably because the SMD-L module is missing) and the voltage returns to 3.54V
Does this current seem high? or is this good? It is under the 500mW limit.

I would say that it's OK. It's only drawing 22mA more than the -18v rail (3.54V vs 2.75v across 33R) and there a few more circuits powered from the +18V vs the -18V rail. The protect relay being one which draws 43mA when the speakers are enabled and protect is off. This would explain the temporary bump in current after power up.

Looking at the circuit, I would estimate that the current draw on the +18V line via R254 would normally be very low (1-2mA) since it only provides base current to Q201 and some current via R29 which is 39K.
What voltage does R154 (Right channel) have across it?
ie: The only high current path to burn out R254 is C23 (which you checked and changed due to ESR) and Q201. You mentioned before that Q201 had some questionable readings on diode check. (0.2V +Base -Collector which seems faulty to me)
This is where a small bench supply would be handy. Especially dual supply one with tracking, even if it only went up to +- 35v or so. Then with a pair of +- regs on a breadboard for the +-18V, you could run the module up and check currents, and general operation without letting any more magic smoke out. Otherwise, it's a bit hit and miss with the DMM diode checks without removing components.

Or is it possible the resistors all just failed? It doesn't seem likely to me.

More likely some transistor or capacitor died and created a cascade of failures.

[paul_g_787]

I think it is safe to say that whatever took out R254 also likely took out R45 too.
And also whatever took out R219 likely also took out R230.

Would you agree with this?

Q203 is also a  bit discoloured but it is right next to R219 so it may have just been that which caused it.

Q201 reads as follows:
B+ to E- = 0.639
B+ to C- = 0.632
E+ to B- = OL
E+ to C- = 2.3V then rises to 3.0V then OL
C+ to B-= OL
C+ to E- = 2.3V then rises to 3.0V then OL

[paul_g_787]

What voltage does R154 (Right channel) have across it?

0.002V in standby and 0.008V in power on state.
so 0.2 to 0.8mA?? (over 10Ω)

This is with no input signal and the volume turned down to zero (as were all my previous measurements).

This is where a small bench supply would be handy. Especially dual supply one with tracking, even if it only went up to +- 35v or so. Then with a pair of +- regs on a breadboard for the +-18V, you could run the module up and check currents, and general operation without letting any more magic smoke out. Otherwise, it's a bit hit and miss with the DMM diode checks without removing components.

Yes it would be handy! If I manage to make any money on this amplifier I will definitely be putting it towards a bench power supply. I have looked at a few on CPC and eBay but the reviews say they don't do good current limiting below 200mA. Which is a bit naff.

More likely some transistor or capacitor died and created a cascade of failures.

I hope it is something this simple but it is proving very tricky to find.

[paul_g_787]

Just measured R146 and R246.
R246 is 1.77K (1.8K) however R146 is only 316R??
Confirmed taking the measurements with probes both ways round but the measurements are taken in circuit.

Edit: Just popped them both out of circuit and they are measuring fine on their own.

[paul_g_787]

Just took a step back and coloured all the components I have tested so far. (Don't think I have missed any).
See attached image.

I popped out R230 and it is indeed a complete open circuit!

[Kim Christensen]

Just measured R146 and R246.
R246 is 1.77K (1.8K) however R146 is only 316R??
Confirmed taking the measurements with probes both ways round but the measurements are taken in circuit.

The in-circuit measurement of R146 makes sense because of R453, R454 (headphone area) and R149 being in series/parallel with it.
ie: R453, R454 and R149 are effectively in series (220 + 100 + 68 = 388 ohms) so when you measure R146 it is like there is a 388 ohm resistor in parallel with the 1.8K (319 ohms) (R454 and R149 are connected together via ground path)

So a reading of 1.77K on R246 in-circuit points to a problem with either R452, R455, R249, or even connector/wire CB45/CZ45 or L21.

[Kim Christensen]

What voltage does R154 (Right channel) have across it?

0.002V in standby and 0.008V in power on state.
so 0.2 to 0.8mA?? (over 10Ω)
This is with no input signal and the volume turned down to zero (as were all my previous measurements).

Yea, I expected it to be quite low. The main power source for the daughter board is the +-46V rail with the +-18V just being bias and power for the very 1st stage.

I think it is safe to say that whatever took out R254 also likely took out R45 too.
And also whatever took out R219 likely also took out R230.
Would you agree with this?

That appears to be a correct line of thought. 

Q201 reads as follows:
B+ to E- = 0.639
B+ to C- = 0.632
E+ to B- = OL
E+ to C- = 2.3V then rises to 3.0V then OL
C+ to B-= OL
C+ to E- = 2.3V then rises to 3.0V then OL

That looks OK. It looks like it survived the high current though it's base that burned up R254. Hard to say 100% if it's really OK with just a diode check, but it'll have to do for now.

One other thing. Since you are powering it up with the SMD-L module removed, how does the rest of the left side amplifier behave? What is the DC voltage on L21? (The output)
I had a thought, that if the rest of the left channel circuitry is OK, then you should be able to make the left output settle at near zero volts by putting a resistor (Try 10K) between pins 3 and 6 of CZ21. This would make the protect circuit happy and kind of verify the output circuitry. Essentially, the voltage L21 should follow the voltage at CZ21-3.
note: CZ21-6 is ground and CZ21-3 is connected to the bases of Q210 & Q211.

[paul_g_787]

The in-circuit measurement of R146 makes sense because of R453, R454 (headphone area) and R149 being in series/parallel with it.
ie: R453, R454 and R149 are effectively in series (220 + 100 + 68 = 388 ohms) so when you measure R146 it is like there is a 388 ohm resistor in parallel with the 1.8K (319 ohms) (R454 and R149 are connected together via ground path)

So a reading of 1.77K on R246 in-circuit points to a problem with either R452, R455, R249, or even connector/wire CB45/CZ45 or L21.

Just checked those resistors and they are all looking good.
R452 = 217.86Ω 
R455 = 98.19Ω
R249 = 68.3Ω

Also checked the same on the Right channel which are measuring a little low.
R453 = 195.58Ω
R454 = 94.0Ω
R149 = 66.3Ω
Edit: I disconnected the headphone board (CB45/CZ45) and these then measured the same as the Left channel. I think this is due to the fact the Left channel module is removed from the main board.

And while I was there I checked:
R150 = 0.03Ω
R250 = 0.04Ω
These are measuring low I believe due to the inductors L11 and L21 being in parallel with them but they are close. I can always remove them and test out of circuit if necessary but they are not discoloured in anyway that suggests they got too hot.

R151 = 9.78Ω
R251 = 9.95Ω
These seem good to me!

[paul_g_787]

I think it is safe to say that whatever took out R254 also likely took out R45 too.
And also whatever took out R219 likely also took out R230.
Would you agree with this?

That appears to be a correct line of thought. 

OK that is good! I am learning (albeit slowly)! 

Q201 reads as follows:
B+ to E- = 0.639
B+ to C- = 0.632
E+ to B- = OL
E+ to C- = 2.3V then rises to 3.0V then OL
C+ to B-= OL
C+ to E- = 2.3V then rises to 3.0V then OL

That looks OK. It looks like it survived the high current though it's base that burned up R254. Hard to say 100% if it's really OK with just a diode check, but it'll have to do for now.

It is the best testing I can do at the moment but there is no obvious direct short circuit on Q201 which is a good sign!!

One other thing. Since you are powering it up with the SMD-L module removed, how does the rest of the left side amplifier behave? What is the DC voltage on L21? (The output)
I had a thought, that if the rest of the left channel circuitry is OK, then you should be able to make the left output settle at near zero volts by putting a resistor (Try 10K) between pins 3 and 6 of CZ21. This would make the protect circuit happy and kind of verify the output circuitry. Essentially, the voltage L21 should follow the voltage at CZ21-3.
note: CZ21-6 is ground and CZ21-3 is connected to the bases of Q210 & Q211.

That is a good idea to check! I assume I would need to wait for my replacement for R230 to be installed first before testing these things? R230 appears to feed the base of Q216 via R269.
I have 10K 0.25W and 1W resistors in my parts bin. Would 0.25W be sufficient?

[Kim Christensen]

I disconnected the headphone board (CB45/CZ45) and these then measured the same as the Left channel. I think this is due to the fact the Left channel module is removed from the main board.

Ah yes of course. Brain fart on my part. R249 wouldn't be in circuit with the Left channel module removed. 

I have 10K 0.25W and 1W resistors in my parts bin. Would 0.25W be sufficient?

0.25W will be good enough. Just kind of tack it in there with full lead length so you can easily tack another lower value in parallel if needed. I'm not sure exactly how much base current Q210/Q211 will require, but starting at 10K ensures nothing really bad can happen with this experiment. Q210/Q211's base current cancel each other out, so the resistor will only needs to pass the mismatch between the two.

[paul_g_787]

So my 82Ω 1W resistors arrived! I have installed a replacement for R230 now.

I had a thought, that if the rest of the left channel circuitry is OK, then you should be able to make the left output settle at near zero volts by putting a resistor (Try 10K) between pins 3 and 6 of CZ21. This would make the protect circuit happy and kind of verify the output circuitry. Essentially, the voltage L21 should follow the voltage at CZ21-3.
note: CZ21-6 is ground and CZ21-3 is connected to the bases of Q210 & Q211.

It did indeed. I installed a 10k resistor as you instructed and the amplifier does indeed power on normally! The protect turns off when it should and I get a green LED!
I measured the voltage across this 10k resistor and it is 0.16V.

One other thing. Since you are powering it up with the SMD-L module removed, how does the rest of the left side amplifier behave? What is the DC voltage on L21? (The output)

Now the amp is switching on normally I measured the DC voltage on both channels output. (With the 10k resistor in place of the left module).
These voltages were stable and not flickering about (after about 1 min power on time).
Left Channel = -0.1678V
Right Channel = -0.0042V

(Yes this is a negative voltage with the + probe connected to + and the - probe connected to GND)

Interesting the left reading is higher than the right but this may be due to the module being removed perhaps?

[Kim Christensen]

Interesting the left reading is higher than the right but this may be due to the module being removed perhaps?

Yes. The negative feedback path, which zeros the DC output voltage and reduces distortion, is part of the module that was removed. So it's not surprising that there is a bit of offset voltage on the output of the left channel. It's pretty good, so the output stage must be fairly well balanced.

Looking back at your picture of the SMD-L module, what caused the goo-stain that appears between Q28 and Q29? I wonder if it was electrolyte, which is conductive, and it might be from C204. Have you checked those two caps for ESR (C204 & C203) or was it just for shorts?

When you do reinstall the repaired module, and before power on, setup your meter(s) to measure the voltage drop across R230 and R254 so you can quickly turn it off again if needed.

EDIT: changed link. The photo Hansson posted was the one I was looking at originally.

[Per Hansson]

Looking back at your picture of the SMD-L module, what caused the goo-stain that appears between Q28 and Q29? I wonder if it was electrolyte, which is conductive, and it might be from C204. Have you checked those two caps for ESR (C204 & C203) or was it just for shorts?

I agree, I think it is even more visible in this picture: https://www.eevblog.com/forum/repair/nad-c325bee-will-not-power/msg3957044/#msg3957044
SMD caps exposed to this much heat and of this vintage are notorious for failing, I would just replace them now that you have the module out Paul since it is not a job for someone without proper tools (that might come after you I mean).

[paul_g_787]

Interesting the left reading is higher than the right but this may be due to the module being removed perhaps?

Yes. The negative feedback path, which zeros the DC output voltage and reduces distortion, is part of the module that was removed. So it's not surprising that there is a bit of offset voltage on the output of the left channel. It's pretty good, so the output stage must be fairly well balanced.

Oh I see! Yes that makes sense to me now. 

Looking back at your picture of the SMD-L module, what caused the goo-stain that appears between Q28 and Q29? I wonder if it was electrolyte, which is conductive, and it might be from C204. Have you checked those two caps for ESR (C204 & C203) or was it just for shorts?

I only measured them in circuit with my capacitance LCR meter and both of them were 22µF exactly.
C203 appears to have some goo too looking closer now you mention it. Perhaps this is what took out R219 and R230!?

I will remove them and check their ESR then report back.

When you do reinstall the repaired module, and before power on, setup your meter(s) to measure the voltage drop across R230 and R254 so you can quickly turn it off again if needed.

Good thinking!

[Kim Christensen]

C203 appears to have some goo too looking closer now you mention it. Perhaps this is what took out R219 and R230!?

Could have. Especially if it flowed under one of the transistors like Q29, bridging CB, and causing it to turn on more than it should. Then the circuit, by the very nature of it's design, would attempt to compensate by driving complementary pairs (Q26 & Q28) harder which would cause excessive current to flow through R219, etc.

[paul_g_787]

OK removed C204 and C204. I think they are pretty toast.

C203 measures 21.3µF 63.2Ω ESR
C204 Measures 22.2µF 45.3Ω ESR

I find it strange that they still have their capacitance despite the ESR being so high.

Now, I originally though the goo on the PCB was just flux, as the whole amplifier has crusty flux all over where they never cleaned it off during manufacturing. 

So perhaps the goo is indeed the electrolyte from these capacitors. It would make sense as the owner said that he was playing music and then it started to sound all distorted.
He described it as if the distortion slowly faded in over about 15 seconds then the red LED came on and he heard a click (the protect circuit). So maybe he was hearing the transistors driving harder as you described due to the short.

This is all starting to add up very slowly!

As for replacement capacitors, I have two options in stock:
22µF 35V NP (Non-polarised) but these are quite hefty.
or 47µF 35V polarised ones which are pretty tiny.

Would either of these types suffice for C203 and C204?
What exactly is the function of C203 and C204 in this case?

Edit: Now that C203 and C204 are removed I was able to get a good reading on R202 and R204, which I am please to say are reading just fine!

[paul_g_787]

Just removed R219 in preparation for a replacement. The bottom pad is absolutely obliterated and there is a large black smoky crater where it used to be!! Poor R219! Must have got pretty toasty before R30 blew.

I have checked for shorts from the damaged trace but it appears to be good. I should be able to join a through hole resistor between D23 and R201 around the edge of the board to make the connection.

Interestingly, NAD replied to me (after a bit of a wait) with a copy of the service manual! How helpful! Top marks on their part, only Marshall and Behringer have ever been this helpful in my experience!!
Their engineer also said that these modules are always failing on this model amplifier and they used to just swap them for a new module, unfortunately the modules are now discontinued.

[SilverSolder]

Can the components that fail be beefed up a little?

[paul_g_787]

Can the components that fail be beefed up a little?

Yes I believe so!

The two SMD resistors I have/intend to replace with 0.25W through hole ones. (Because I have these already). My theory is the larger size of these compared to the SMD ones will allow for better heat dissipation, especially as they are mounted off the edge of the board away from the other parts. This should avoid future pad destruction.

The SMD capacitors have been/will be replaced with through hole types with the legs bent in an upside down T shape to attach to the pads. This will allow them to stand away from the board and away from the direct heat of the other components.

The R45 fusible resistor and R46 also were mounted flat on the PCB, this resulted in the destruction of R45's pads. I have mounted the replacements up off the board with 1cm gap to allow the heat to dissipate better.

The fusible resistors of course have been replaced with the exact same power rating.

I am hoping this will all make this module more reliable in future.

[Kim Christensen]

Would either of these types suffice for C203 and C204?

While it's a bit of a waste of a NP capacitor, I'd go with the 22uF ones just to keep things as close to the original design. The 47uF ones will double the time that it takes on power up for the bias currents through Q26-Q29 to stabilize.

What exactly is the function of C203 and C204 in this case?

They decouple the bias circuits for the output stages on the module. ie: Q203, R201, R202, R203, & C203 set the DC bias voltage for Q26. There's flipped version of that for Q27.

Edit: Now that C203 and C204 are removed I was able to get a good reading on R202 and R204, which I am please to say are reading just fine!

Hmm... Sounds almost like C203 and C204 were leaky in more ways than one. Do you have a capacitor leakage tester? If not, you can use a small power supply or 9V battery to try and charge up C203/C204 via 100K or 1M resistor. You can then calc the leakage current by measuring the voltage remaining across the resistor after a few seconds. If you were curious.
Because a very leaky C203/C204 would increase the bias current and burn up R219, R230, etc.

[paul_g_787]

While it's a bit of a waste of a NP capacitor, I'd go with the 22uF ones just to keep things as close to the original design. The 47uF ones will double the time that it takes on power up for the bias currents through Q26-Q29 to stabilize.

I mean, I have 48 of them in stock as it was cheaper to buy 50 than two for another repair last year. I never can work this pricing out sometimes. I may as well use these rather than paying out for new parts and costing more money.

Hmm... Sounds almost like C203 and C204 were leaky in more ways than one. Do you have a capacitor leakage tester? If not, you can use a small power supply or 9V battery to try and charge up C203/C204 via 100K or 1M resistor. You can then calc the leakage current by measuring the voltage remaining across the resistor after a few seconds. If you were curious.
Because a very leaky C203/C204 would increase the bias current and burn up R219, R230, etc.

I do not have one but that is a good idea! I will certainly try that later on just to see.

In the meantime I have attached a new R219 and the two new capacitors C203 and C204.

[paul_g_787]

View from the other end.

[paul_g_787]

OK, I re-installed the module and the voltage readings across the resistors were as follows:
R254 10.56Ω 0.005V
R230 81.70Ω 0.751V

So all being good here I connected some speakers and it is working great!!!!!!

Kim Christensen thank you for all your helpful advice!!! I certainly would not have managed to repair this if it wasn't for your excellent help! 
I have certainly learned a fair bit on this project.

Please see attached an image showing the two SMD modules. On the left is, well... the left channel. You can work out the rest.

I have given the right channel SMD module a very good clean with IPA to make sure it is spotless, just in case of future right module oozing.

I followed the calibration procedures in the service manual. Both channels were way off! But this is all measuring good now.

[Kim Christensen]

Kim Christensen thank you for all your helpful advice!!! I certainly would not have managed to repair this if it wasn't for your excellent help! 

You're welcome. It was a fun exercise for me as well. Good to hear that all is working now. 

[paul_g_787]

Indeed! I do love a challenge and it is good to have some expert help as well!

I learned a lot more than I knew before about how amplifiers actually work from this one. It was good fun! 😁

[Per Hansson]

I would replace the SMD caps on the other module too, if you re-read my post #40 you will see they where bad on both.

[paul_g_787]

I would replace the SMD caps on the other module too, if you re-read my post #40 you will see they where bad on both.

I checked the ESR on them both in circuit and they were 0.3Ω and 0.7Ω (I forget which was which). Which I didn't actually do on the other module.
I didn't want to tempt fate and remove the module if I didn't need to just in case I made it worse! I had a very hard time soldering those SMD parts as it is!

[jhjove]

New to the forum (typically on diyaudio and audiokarma) and 2+ months late to this topic. Quite a nice dialogue and repair.

FWIW, I went through an almost exact repair sequence in February 2021. I acquired a C325BEE (a for-parts-only special) from ebay.
Although in addition to the +/-18V power supply issues, the main power board was fried on my amp. In the end I have a very nice working amp.

I wanted to add that the failed SMD caps on the CB21 + CB22 riser boards I replaced with similar SMD form-factor parts, but solid-polymer type electrolytics.

Again, I enjoyed the thread, very good info.

[FIXITNOW2003]

not seen a 325 but have seen some 320 versions these do suffer with poor elec caps  in the protection circuit. and output stages i would check the to see if esr is poor

[Mathijs]

Now 2 years later (february 2024) I am experiencing a similar problem with my now 18 year old C325BEE NAD amplifier. Thanks to paul_g_787, Kim Christensen and other readers I was able to repair it. My C325 switched off again after a few seconds after being switched on, so from amber to green and to amber light again. The ultimate cause was the SMD caps in the pre-Amps that may have leaked (C203/204/C103/104). I replaced all SMD caps with new ones, after which the amplifier switches on again without errors. Considering the age, I immediately recapped the entire power supply. It was a bit of a job, but for now the C325BGEE can last for a while. Thanks everyone !