I received a NAD M3 quite a while ago with issues. 100% down to bad capacitors, that thing was full of bad capacitors, even the main filters were complete garbage. I was stupid in trading away and fairly nice Wyred4Sound amplifier in exchange for it. Oh well.
The NAD is beautiful, it took me a while, but I did a full recap on it, every single electrolytic in it. I got it all back together and adjusted bias on the first channel and then when I went to set it on the other channel, I found I had no adjustment whatsoever. I said it down and walked away as I had plenty of other stuff to do. Now I kind of wanna just finish it up so I can, check it out. I went through the usuals, like checking, transistors and resistors and so far have not found any issues.
I have course checked the two devices mounted to the heat sink as they generally have something to do with bias current. One of the items is a thermal breaker and it has continuity so it is good. The other thing is a “posistor” which this is the first time I’ve ever heard of this device.
I checked the schematic and this is all that I could find in there regarding it, it goes between two pins on a ribbon cable. From what I understand, it increases in resistance as it gets hotter.
I figured it would be similar to a thermistor and would measure similar to a diode. I measured the device on the channel and sure enough it measured as a diode with a voltage drop of .2v
I measured the one on the bad channel and it came up as no device detected
But I’m wondering if that’s OK, because the description of it is that it’s a device that increases resistance as a heats. The good one measures 42 ohms across it. The “bad” one measures 53 ohms. Though I believe this helps answer my question as that 53 ohms is kind of flaky.
I would swap the components between the channels, but it’s an absolute pain to get to the hardware of the devices and to dismantle and assemble the channels/amplifier. So I’m thinking I would like to just purchase a device and throw it in if you all think that it is indeed bad
What would I throw in there? It’s marked with “F1 63”
The manual does not give much information other than the part number being 07-70471-08 and T=80c.
I am having trouble tracking down a PTC that is rated for 80c and 42 ohms, I’m guessing that’s what it is supposed to be since that’s what the good one measures. If I go with those specs, I’m limited to some surface mount components. I am not seen anything for being mountable to a heat sink.
Thank you so much!
Dan
The NAD is beautiful, it took me a while, but I did a full recap on it, every single electrolytic in it. I got it all back together and adjusted bias on the first channel and then when I went to set it on the other channel, I found I had no adjustment whatsoever. I said it down and walked away as I had plenty of other stuff to do. Now I kind of wanna just finish it up so I can, check it out. I went through the usuals, like checking, transistors and resistors and so far have not found any issues.
I have course checked the two devices mounted to the heat sink as they generally have something to do with bias current. One of the items is a thermal breaker and it has continuity so it is good. The other thing is a “posistor” which this is the first time I’ve ever heard of this device.
I checked the schematic and this is all that I could find in there regarding it, it goes between two pins on a ribbon cable. From what I understand, it increases in resistance as it gets hotter.
I figured it would be similar to a thermistor and would measure similar to a diode. I measured the device on the channel and sure enough it measured as a diode with a voltage drop of .2v
I measured the one on the bad channel and it came up as no device detected
But I’m wondering if that’s OK, because the description of it is that it’s a device that increases resistance as a heats. The good one measures 42 ohms across it. The “bad” one measures 53 ohms. Though I believe this helps answer my question as that 53 ohms is kind of flaky.
I would swap the components between the channels, but it’s an absolute pain to get to the hardware of the devices and to dismantle and assemble the channels/amplifier. So I’m thinking I would like to just purchase a device and throw it in if you all think that it is indeed bad
What would I throw in there? It’s marked with “F1 63”
The manual does not give much information other than the part number being 07-70471-08 and T=80c.
I am having trouble tracking down a PTC that is rated for 80c and 42 ohms, I’m guessing that’s what it is supposed to be since that’s what the good one measures. If I go with those specs, I’m limited to some surface mount components. I am not seen anything for being mountable to a heat sink.
Thank you so much!
Dan
PTCs are large tolerance. You could easily have two same part number being 42 and 53, especiallyy if you mesured them at two different times at two different temps. This one is +/-20%: https://www.mouser.be/datasheet/2/281/PRG21AR420MB1RA-e-1485222.pdf
Upshot: the 53 ohm one probably is fine.
Jan
Upshot: the 53 ohm one probably is fine.
Jan
https://www.mouser.com/ProductDetail/Murata-Electronics/PRG21AR420MB1RA?qs=fDvdbkAhQ5aXy9dG4Mew/w==
So this one right here would be perfectly fine then? I would just need to solder a couple leads to either side and then figure out a way of mounting it to the sink.
Well, your comment has me thinking, even though it’s going to be an absolute pain I’m gonna swap the PTCs between the two channels and see if the issue swaps channels.
Dan
Are you sure the positor has anything to do with bias adjustment? It looks more like over temperature sensing to me.
Bias voltage generator looks conventional enough. Measure voltage across Q316 and see what range of voltage adjustment you have:
Bias voltage generator looks conventional enough. Measure voltage across Q316 and see what range of voltage adjustment you have:
It doesn’t, I did not look at the full schematic to see where those pins go. The amplifier is a bit more sophisticated than what I normally work on. Regardless, it was an absolute pain, but I swapped the two devices between the channels and made no difference. The right channel can still be easily biased to 8 mVwhile the left sits at 2.7 and has absolutely zero adjustment from there.
I’m going to pull the left channel back out and take a look in that area. Unfortunately, there is no easy way of measuring across that transistor while in use, so if I don’t find anything wrong, I will solder some flying leads across it so that I can measure it. it sucks, the board and heat sync module has to be fully seated and installed to power up at all.
Out of curiosity, am I able to power this amplifier board outside of the amplifier? Here are the connections on the schematic and on the board itself.
The four at the bottom, P301-P304 are these four connections here. Pretty robust.
The 10 pin connector is the one on the side. Can I connect the connectors in the bottom that are supposed to see +/- 74v to the 10 pin header on the side that sees +/- 72v? So pin 1 to P301 and pin 2 to P302.
I’m not sure what other voltages it is supposed to see to at least power on and let me sense for bias adjustment, 24v at pin 9 and then is there supposed to be something on the AC sense line?
I have a power supply that can give me +/- 75v and another that can give +/- 32v. it would be so much easier if I was able to power this up on my bench and work without the heatsink. It would save me A LOT of time.
All right, I’m going to tear into this and look in that portion of the circuit
Dan
I appreciate you pointing those components out. I have to take a small break and watch of my daughters for a little bit and I’ll jump back onto it in about an hour or two. I did pull Q316 and Q315 and they both measure good out of circuit with appropriate gain. All of the resistors within the red area measured dead on. I also measured the ceramic capacitors, but being in circuit, the information probably isn’t all that useful. as soon as I can get back to it, I am going to place the two transistors. I took out and measured back into the board and attach some leads so that I can measure voltage on all three pins of that transistor. Out of curiosity and other places, I should look before I place the board back in? I did also check R331 and R339 which are fuse resistors.
Edit: I also pulled Q313 and Q314 and both measure okay.
Dan
Last edited:
I just pulled and checked the drivers Q317, Q318, Q337, and Q338 and all looks good. I’m going to continue looking for a bit before putting everything back together and testing.
Dan
Dan
Can you get at the resistors that go to the same points? or the base of the drivers? Those 4.7 ohm can be disregarded and you can measure on either end of those instead of Q316
I wouldn't start removing parts in hope, gather all the evidence first by careful voltage measurement.
I figured out some places to measure from, I was just going through the board to see if there was anything that was obvious, nothing was found though.
My apologies, but I’m not exactly sure what you mean by measure across Q316. Really the only voltage I can measure is if I go from emitter to collector or emitter to base. In the full range of the pot I’m able to adjust for 676mV to 712mV.
On a sidenote, I decided to just re-flow the board as it’s very clear. It’s had a bit of heat in its life and the joints looked fairly dry. The amplifier now powers on and after about 10 seconds it goes into protection, ugh. I removed that channel from the amplifier and it powers up and stays on just fine, so it definitely has to do with the channel I am working on. I will tear apart once again and see if maybe I have any bridges. Either that or I didn’t have dry joints and now the circuit is completed and a faulty component is throwing it into protect.
But last night I want to say, I pretty much checked every single resistor on this board, and I checked every single transistor on this board. At least check them out of circuit with a Peak Atlas DCA75. The only pieces of silicon I have not checked are the diodes out of circuit (I did check them in circuit with the diode function, couldn’t tell me if the Zeners were accurate) and I have not checked either opamp (TL082CP and uPC1237HA).
Dan
Red meter lead on the emitter and black lead on the collector and you should see the voltage change as you alter the bias preset but it sounds like you have other problems now 🙁
If its going into protect the first thing to check is the DC offset to see if that is the problem. The IC's are highly unlikely to be a problem.
Maybe somebody can answer this for me, does this look like it could possibly be powered up out of the amplifier with no heat sink attached? Just at least to get me to a point where I can measure and adjust for bias. It would give me access to the backside of the board. I have a whole bunch of 2SC5200/2SA1943 so I could just solder some to those points in order to complete the circuit.
Here is the portion of the circuit and labels of all of the pins.
Here is how I have the board.
I would feed +72v into P301 (red) and it has a clip to feed it to pin 1 on the board.
I would feed -72v into P302 (black) and it has a clip to feed it to pin 2 on the board.
I am not sure if I would connect all of the grounds together, but I have P303 and P304 clipped together (green) with the clip going over to pin 10. This connects GND, SPGND, and PRGND
I would feed 24v into pin 9 and then not sure about pin 4, it’s and AC sense if I’m reading that correctly. Hopefully I don’t need that hooked up in order get the bias circuit to work.
The 10 pin connector is a small ribbon, so I would just have to solder little clip points on the backside. By the way that little cable is a pain.
Thoughts?
Dan
Here is the portion of the circuit and labels of all of the pins.
Here is how I have the board.
I would feed +72v into P301 (red) and it has a clip to feed it to pin 1 on the board.
I would feed -72v into P302 (black) and it has a clip to feed it to pin 2 on the board.
I am not sure if I would connect all of the grounds together, but I have P303 and P304 clipped together (green) with the clip going over to pin 10. This connects GND, SPGND, and PRGND
I would feed 24v into pin 9 and then not sure about pin 4, it’s and AC sense if I’m reading that correctly. Hopefully I don’t need that hooked up in order get the bias circuit to work.
The 10 pin connector is a small ribbon, so I would just have to solder little clip points on the backside. By the way that little cable is a pain.
Thoughts?
Dan
Yes, I am going through that now. Even though it does go into protect after 10 seconds or so I still have the same bias voltage on the test points, being unchanging. How you described it is exactly how I measured it. So emitter to collector I was seeing 676 mV up to 712 mV within the full range of the trimmer turn.
Appreciate your assistance in this!
Dan
The circuit you show is the protection and tbh I can't see why you might want to power that up independently. The bias should adjust independently of whatever happens on the protection board assuming of course there isn;t a problem stopping the bias adjusting. The protection won't influence the bias in any way all else being equal.
Running boards out of situ is fraught with problems and running something with no heatsink is asking for trouble.
Listen 🙂 ... having no or low bias current on its own will not stop the amp working. It would play and perform just fine with only a tiny increase in distortion to that show the bias was low and that increase would probably not even be detectable on listening.
The voltage across Q316 has to be high enough to turn on both driver transistors (forget they are parallel paired, it makes no difference to this). That means you would need around 1.2 volts to get them into conduction (0.6 volts turn on base/emitter volts for the NPN and PNP drivers).
As mentioned, even zero bias current still gives you a 99% correctly functioning amp.
Well, I did find out why it’s going into protection
70 V on the output would do that. I went through the board very carefully and there are no bridges so it’s something else. Something to do with a negative side of the rail.
The reason I wanted to test it outside of the amplifier is because it’s just a huge pain to take it out put it in, take it out put it in. And it’s not like I can just put it near the amp make connections and go, depends on the back have to slide into the output board, I have to hook up the small 10 pin ribbon cable
Then on the bottom side, I have to put in for Allen screws that are down deep in order to connect the main grounds and the positive and negative rails
And then on top of that, I have to remove the face plate in order to get heat sink in and out
It’s just a lot of work. I recently watched a video from Mend it Mark where he makes repairs to a Yamaha amplifier board. That board only needed +/-, ground, and then a small sense signal, I think he fed it 12v. Unfortunately, these boards have a lot more connections and are a bit more sophisticated. When I mentioned running without the heat sink I meant just having it idle so that I could get to a point where I can adjust bias. if I could get the bias to go anywhere above 5 V then I would immediately turn it off and mount it back to the heatsink.
I get that the amp will still function, but in my head, I don’t like the fact that it isn’t working correctly. If I could get to eight mV, even if as high as I could get is 7 mV I would live with it. But sitting around 2 mV it would be cool to get it a bit higher.
I’m gonna dig around and see what is up with the negative rail on this channel, don’t know why I all of a sudden have high offset, I didn’t before. Do you have any thoughts as to why the voltage is so low at the bias point?
Dan
Running without a heatsink risks pretty much certain destruction. There would be no thermal feedback to the bias generator and a tiny increase in current would see the output transistors fry. Even a current as low as 50ma generates 3.5 watts heat per output transistor (0.05*70) and that would sizzle in a couple of seconds, the current would run away and bang.
You need to look at why you have this 70 volt offset. If it was essentially all working before apart from the low bias (as in post #1) then something untoward has happened.
The negative rail may well be OK (but check of course) and the high positive offset is caused by a problem within the amp such as the upper part of the output stage being turned fully on.
I get that its a complex amp and not very service friendly but you need to get as much as evidence from voltage checks first.
You need to look at why you have this 70 volt offset. If it was essentially all working before apart from the low bias (as in post #1) then something untoward has happened.
The negative rail may well be OK (but check of course) and the high positive offset is caused by a problem within the amp such as the upper part of the output stage being turned fully on.
I get that its a complex amp and not very service friendly but you need to get as much as evidence from voltage checks first.
It indeed not very service friendly. I get we’re you’re coming from on the heatsink, I’ll make sure it’s attached whenever powered up. No sense in risking it. I can’t get the output transistors off of the heat sink, it’s almost like they are glued. I’m sure I can knock them off with a hammer in the punch, but I really don’t want to do any damage so I figure I’ll just keep soldering and desoldering. I can already sense some metal fatigue in the legs though.
When I said the negative rail, I met the negative side of the circuit. I have negative voltage going into the board just fine, but I was thinking that if there’s full positive voltage on the output, the issue would be on the negative side of the circuit. What a pain. I’m thinking it has something to do with the transistors. I removed all of them from the board to measure them, but I’m 100% positive the went right back into the same spot, same orientation.
Either that or there was always high voltage on the output and the protection circuit wasn’t kicking in Doubtful though.
Dan
Well, I don’t know what I did. I removed all of the transistors mounted to the small heat sinks, like the drivers measured them to make sure none of them had gone and put them back in. Now it powers up and does not go into protection, I did check the offset and it still pretty high, but not rail voltage high. 380-410 mV, far higher than I’d want to see it.
The bias is being kind of weird, I will power it up and it will be at 1.8 next time I power it up It’ll be at 2.7. It’s weird how it won’t even budge 0.1 mV when adjusting the trimmer, but can move nearly a full mil of volt between power cycles.
What can be done about the higher than desired offset? This doesn’t look like it’s a standard differential pair input. I don’t know why I didn’t think of about doing it earlier, but next time I go down stairs, I will grab the offset from the other channel. I am going to have to figure out some thing to do about being able to work on this amplifier and not having to take off the outputs each and every time I want to touch the board. The last time I soldered the outputs to the board I had many of the pins snap from fatigue. Once I get everything all settled and everything looks like it’s going appropriately. I will probably just put brand new outputs in it since I have them.
Dan
The bias is being kind of weird, I will power it up and it will be at 1.8 next time I power it up It’ll be at 2.7. It’s weird how it won’t even budge 0.1 mV when adjusting the trimmer, but can move nearly a full mil of volt between power cycles.
What can be done about the higher than desired offset? This doesn’t look like it’s a standard differential pair input. I don’t know why I didn’t think of about doing it earlier, but next time I go down stairs, I will grab the offset from the other channel. I am going to have to figure out some thing to do about being able to work on this amplifier and not having to take off the outputs each and every time I want to touch the board. The last time I soldered the outputs to the board I had many of the pins snap from fatigue. Once I get everything all settled and everything looks like it’s going appropriately. I will probably just put brand new outputs in it since I have them.
Dan
Attachments
I was having a problem with the volume control, which is one of those digital rotary switches, so got that cleaned up and was able to get the other side checked. The right channel looks outstanding, just a couple of mV offset. The left of course has a few hundred. is there any chance that the high offset and inability to adjust bias have anything to do with each other?
Dan
Dan
They should not interact. The offset should be essentially zero at all times because the amp has a DC servo (the TL082 opamp) which constantly forces this zero offset. Any offset at the output will force the output of the opamp to change and bring the offset back to zero. Make sure the -/+18v supplies to the opamp are OK.
The fact the symptoms keep changing as you work on it make it very difficult to second guess what is going on.