Beware of NAD M3 Fire Hazard


My $3k NAD M3 started shooting sparks out the top and burned the shelf that was 8" above. Luckily I was home and not sleeping or the house would have burned down. If anyone has one of these I advise them to unplug it when not in use. I took it to two different repair shops and they said it would be about $800 to just get it running and there may be board issues. They advised not to take the gamble. Anyone have any suggestions on what to do with it?
pwb
Consider sending it to the Hydraulic Press guy channel on You Tube and ask for some $$ when he makes thousands from the subscribers; People send him all kinds of stuff; Very entertaining and it’s for science.

I’ve had some equipment over the years I would have loved to send him;

I can't believe that happened to you; Did you buy it new? How freaking scary!

Or you could throw it from a great height onto a trampoline. Or hey I know, shoot it with a sniper rifle from a mile away! If you do this remember, one camera and mic on the rifle, another high speed on the NAD. 

Or wait! Even better! Hook it up with some sparklers and stuff and recreate the event. First go down to Goodwill, pick up a rack and some stereo gear for like $5, make a video of the whole thing burning down. Not sparkling, burning down. Like you went out to check the mail and left the camera running. Then go down and shoot video of the local fire department. Put it all together so it looks like NAD burned your house down. Email a copy to NAD. Ask about their warranty, and oh by the way would you be interested in commenting before I post this on youtube?
@ pwb

Odd the AC line safety fuse did not blow. The fuse should have blown protecting the unit from becoming a fire hazard. Is the AC Line fuse the original OEM fuse that came with the amp?

If it is the original OEM fuse I suggest you contact NAD service Tech support.

Jim

Have you removed the ground pin?
If not, check that your socket is correctly wired.
Odd the AC line safety fuse did not blow.

@jea48, Jim, I took a look at the service manual for the M3, which can be seen at hifiengine.com if one is registered there. It appears that it has three fuses, all of which are AFTER an EMI filter, consisting of inductors and capacitors, which is connected between the incoming AC and the rest of the amp.

There are two 5 amp fuses (assuming this is a 120 volt model), one for each channel, with the main (rear panel) power switch connected between the filter and these fuses. There is also a 100 ma fuse protecting standby-related circuitry, with a relay connected between the filter and this fuse.

All of the fuses are located inside the unit, and they are not accessible on the rear panel.

All of this would seem to cast suspicion on the EMI filter as the culprit, although secondary damage elsewhere within the unit certainly may have occurred.

Best regards,
-- Al


Did the service manual indicate the correct fuse direction? Might have been switched.
Thanks for the responses. As far as I know the fuses are intact. I stopped the sparks shooting out the top by unplugging it. I took it to the NAD authorized repair facility and they said a faulty power supply caused the round cylinder ( not sure of part name) to explode which probably damaged the board and other parts. I can send pics if anyone wants to see the damage. Thanks.
Safety caps do fail short, not enough though to create a fault condition a fuse or breaker can clear. RIFA caps in old Mark Levinson gear are notorious for this. 

If it's the "round cylinder", it is possible that something conductive shorted the power supply cap or caused a voltage reversal. Who knows,  a discarded component lead was dropped inside the unit during production and wiggled its way to a bad spot during shipping/setup.
Al, that amp has a lot crap in it! How long did it take for you to find the 3 fuses?


pwb said:
I took it to the NAD authorized repair facility and they said a faulty power supply caused the round cylinder ( not sure of part name) to explode

round cylinder in the power supply???

I should have mentioned in my previous post that series arcing will not cause a fuse to blow. In the case of parallel arcing the current may not be high enough to cause a fuse to blow. The arcing can, will, create sparks though.

Jim
.
Al, that amp has a lot crap in it! How long did it take for you to find the 3 fuses?

Not long, Jim. The service manual includes a block diagram, a schematic, and a parts list. Which made it pretty easy to determine what fuses are being used, and where they are located in the circuitry.

And a quick look at rear panel photos that are available on the web confirmed that the fuses are not physically located on the panel.

Best regards,
-- Al
pwb --- The tall round can like things or the donut like things??
Take a look at the inside picture here:

https://hometheaterhifi.com/reviews/amplifier/integrated-amplifiers/nad-m3-integrated-stereo-amplifi...

The two round metal covers at the bottom are on top of the transformer.  The next 4 round can things above are capacitors.  Just curious which one.
It was the second capacitor from the right side of the amp that blew. 
Send your pic to NAD.  Link this thread to NAD, as well.  With luck you may get an appropriate response from them.  In any case, don't plug it in again!
Looking at the service manual again that capacitor is a 22,000 uF 80 volt polarized aluminum electrolytic. As can be expected it is connected to the output of a diode-based full wave rectifier. But based on what you (pwb) were told it sounds like something else failed in the power supply which caused that cap to explode. If so, that something else could have been a lot of things, as there are a great many parts in that circuit.

However, although the schematic is a little hard to interpret I’m fairly certain that the DC voltage appearing across that capacitor in normal operation (and across the other three similar capacitors, each channel having two of them, one for a + voltage and one for a - voltage) is between 70 and 74 volts. If so, that seems to me to be a failure waiting to happen, as a cap should be run at a voltage that is much less than its rating, IMO. Preferably at something like half of the rating.

Regards,
-- Al

I agree with edcyn. NAD needs to know about this. 
My old tech friends refer to the dead NAD as a boat anchor.  We use to play golf with am/fm Delcoes in the late 70's, you will need a good sludge hammer.  Very sad to hear of your plight, glad you were home.  I always consider NAD as a really good brand-I have no experience with their equipment.  My good buddy Chrisopher Coffin, superb audio tech, had this to say about fuses.  A 200 picture tube will blow and save the life of a 25 cent fuse(1979).  Not quite his original saying, but you get the drift.

1. Contact NAD
2.Contact who you bought this junk from.
3 Tell Millercaarbon to piss off.
3 Buy some real gear like Simaudio.


Al,

There is absolutely no impact on the design life of modern low voltage electrolytic capacitors (this would be low voltage) by running them up to and at their rated voltage. Little evidence for MTBF impacts either. Most are also rated for 15-20% over voltage so short term voltage transients don’t cause damage (and this provide margin for MTBF under normal usage). This has been the case for some time. High voltage electrolytic capacitors still have some reduction in life but it is not severe.


Unfortunately electrolytic capacitors are prone to much higher random failures than anything else in an amp.


Could be anything, but a shorted diode in the bridge passing AC and cooking the capacitor would be my guess.


W.r.t. junk,
Everybody has failures. It doesn’t mean it is junk unless this happens an unusual number of times.


NAD will want to know in case they have a mfg or design defect.


https://ibb.co/T1VBM5G

I’ve never seen the top vent of an electrolytic cap blown off like the one in the photo.

That’s just plain scary!



heaudio123 Said:
Could be anything, but a shorted diode in the bridge passing AC and cooking the capacitor would be my guess.

Wouldn’t that cause a high current load on the secondary winding of the toroid power transformer causing the primary winding to overload and cause the 5 amp (I assume a slow blow) fuse to blow?

.
No it would not blow the fuse. The capacitor is not a dead short when reverse biased, but a low enough resistance that it will heat quickly and blow up.
ok... you probably got a simple case of defective cap.... most likely there is no more damage to the the unit . you need all 4 caps to be replaced and that is probably it ...  
@heaudio123, thank you for your response to my post. I know you are extremely knowledgeable about such matters, and my relevant design experience dates back a few decades.

I would be interested, though, in your comments about the section entitled "Operating Lifetime Model" on page 13 of the following Cornell Dubilier "Aluminum Electrolytic Capacitor Application Guide":

https://www.cde.com/resources/catalogs/AEappGUIDE.pdf

Based on the equations shown there it appears to me that operating such a capacitor at 50% of rated voltage would double "expected operating lifetime" compared to operating it at 90% of rated voltage, which to a close approximation is what is done in the M3.

Regards,
-- Al
I'd recommend, seriously, this time, not attempting to put resources into repairing it; When the electrolytic off-gassed and spewed its internal magic smoke, conductive fluid (electrolyte) went all over the other parts (IC's in particular) making new instantaneous conductive pathways and likely altering the nominal set points; It's any ones guess about the future reliability.

From a practical point of view, how relaxed would any of us be if the amp IS reworked? I sure as heck would not be able to enjoy music with this refurbished amp. I would be on edge...wondering when the next magic smoke makes an appearance;

It would be wonderful if NAD would pick up the tab on a proper failure analysis and next steps; 

BTW, I had an M3 in 2006 and really liked it; Very smooth and neutral; Beautiful to look at too; Impressive chassis design.

 
As an authorized NAD servicer for over 40 years this is the first time I’ve heard of this.
I can’t see how this would occur UNLESS the user keeps their electronics on all of the time.
The user did not mention they were listening the time it failed, so I guess they do keep it on at all times. In standby, the capacitor that failed is not in the circuit.
If powered to a fully on condition, the capacitor is live and before any regulation. If there are issues with incoming AC power they will translate thru to these 4 capacitors.
If the user does not keep the unit on at all times this can still occur, but it’s less of an issue.
Capacitors can and do fail due to many reasons but there is nothing endemic in this model.
If the incoming AC power is above the nominal voltage (typ. 114-126) for long periods of time, this can put a stress on filter caps in pre-regulated circuits.
Here is a good suggestion for all electronics... do not place anything combustible above the equipment and the "burning" component will not have anything to ignite.
I do agree with the poster directly above this as the highly corrosive electrolyte is likely spewed onto many parts and the PC board, too.  It will cause metal component leads as well as. PCB traces to corrode and fail. Also, the unit uses double sided PC boards and many small "plate-throughs" (micro rivets that connect one side of the board to the other) that are easily damaged from the electrolyte.

I knew you would bring up the Cornell Dubilier app note for which their equation has not been updated in decades nor any differentiation made on capacitor voltage and size. I doubt they have not changed their electrolyte in 30 years so going to go with oversite or laziness on their part to not differentiate 

From Nichicon, "The degree that applied voltage effects the life of the
capacitor when used below the rated voltage is small,compared to the degree that ambient temperature and
ripple current effects life. Therefore, when estimating the life of a capacitor, the voltage coefficient to the applied voltage (Fu) is calculated as 1. An example of the test results is shown in Fig.2-16."


UCC: "*Kv : Derating voltage factor (a snap-in type capacitor with the
rated voltage of less than 160Vdc and a screw terminal type capacitor with the rated voltage of less than 350Vdc :1)" .... Ie there is no defeating for < 160 volt, and if more robust screw style, 350V


Illinois Capacitor: No voltage derating factor for V<=160. (see their lifetime calculator notes)
Thank you everyone for the useful information. I guess I will just have to take the loss and move on.  I will never purchase an NAD product again. Anyone have a recommendation for a good  integrated under 2k. It’s for a second system driving Merlin tsm-mmm speakers. Thanks.
heaudio123 Said:
Could be anything, but a shorted diode in the bridge passing AC and cooking the capacitor would be my guess.

jea48 response:
Wouldn’t that cause a high current load on the secondary winding of the toroid power transformer causing the primary winding to overload and cause the 5 amp (I assume a slow blow) fuse to blow?


heaudio123 Response:

heaudio123100 posts

03-15-2020
3:19am

No it would not blow the fuse. The capacitor is not a dead short when reverse biased, but a low enough resistance that it will heat quickly and blow up.


heaudio123,
Thank you for your response.
I think I understand what you are saying in your above statement. The cap had a lot of stored energy and it was suddenly fed with a reversed polarity by the shorting diode. Kind of like connecting two 12Vdc car batteries in parallel out of polarity with one another.

I noticed the vent cap on the other cap did not look like it was affected by the event. It doesn’t look like there is any bulging in the top vent cover. Wouldn’t the shorting diode have affected that cap as well?


heaudio123 Said:
Could be anything, but a shorted diode in the bridge passing AC and cooking the capacitor would be my guess.
heaudio123,

Would you please explain the event in time of the shorted diode. How long of an event time did the short of the diode last before, I assume, the diode blew apart breaking the short circuit on the secondary winding of the toroid power transformer?

Jim
.
I noticed the vent cap on the other cap did not look like it was affected by the event. It doesn't look like there is any bulging in the top vent cover. Wouldn't the shorting diode have affected that cap as well?
@jea48 Jim, good question, but the answer is that the schematic shows that there is a separate diode bridge (and other circuitry) associated with each of the two capacitors that are in each channel. One set of diodes/capacitors/and other circuitry provides approximately +72 VDC, while the other set provides approximately -72 VDC.

Best regards,
-- Al
 

I haven’t seen capacitor application notes in decades, but I believe heat still applies, right? That is, the capacitor can be aged while on, not necessarily because of the voltage applied but because of the heat in the amp. Further, higher temperature caps also have longer life spans at a given temperature than lower temp caps, or am I mistaken?

I agree, unfortunately, that the failure here seems not worth fixing. That’s a lot of cleaning to do, with no guarantees.

If it were MY personal unit, I would attempt to clean it, and see if the remaining power supply works, and if the logic/preamp circuits appear in tact, then and only then would I attempt to rebuild the burned out side. Still, lots of work with no guaranteed happy ending.
almarg9,445 posts

03-15-2020
11:20am


I noticed the vent cap on the other cap did not look like it was affected by the event. It doesn’t look like there is any bulging in the top vent cover. Wouldn’t the shorting diode have affected that cap as well?

@jea48
Jim, good question, but the answer is that the schematic shows that there is a separate diode bridge (and other circuitry) associated with each of the two capacitors that are in each channel. One set of diodes/capacitors/and other circuitry provides approximately +72 VDC, while the other set provides approximately -72 VDC.

Best regards,
-- Al

Al, thanks for the quick response. You provided the reason why.....



Any thoughts on my comment on this part of my previous post.

heaudio123,

Would you please explain the event in time of the shorted diode. How long of an event time did the short of the diode last before, I assume, the diode blew apart breaking the short circuit on the secondary winding of the toroid power transformer?

From the OP:
pwb OP9 posts  

03-14-2020 
 3:03pm  

Thanks for the responses. As far as I know the fuses are intact. I stopped the sparks shooting out the top by unplugging it.
AC Mains was still supplying energy until the OP unplugged the amp from the wall.

Jim
I will never purchase an NAD product again.
EVERY product suffers one-off failures.

NAD has made millions of units with very high reliability and excellent sonics for very reasonable prices.

NAD are well engineered. Electronics are made of component sourced from dozens of manufacturers. A dozen or more maybe involved in the power supply alone. All are potential points of failure for which NAD unjustly suffers the blame.

It's possible that a line surge coupled with a weakish part at precisely the wrong moment caused the only catastrophic failure on the planet.

The probability than another NAD unit would suffer a similar fate is small.

The unit is an audio computer and computers are replaced quite frequently.  The last firmware update was a decade ago, so the unit could be nearing EOL.

Some questions:
- did you buy it new?
- was the unit left on all the time?
- what is the local line voltage? 
pwb OP9 posts  

03-15-2020  
 10:01am  

Thank you everyone for the useful information. I guess I will just have to take the loss and move on. I will never purchase an NAD product again. Anyone have a recommendation for a good integrated under 2k. It’s for a second system driving Merlin tsm-mmm speakers. Thanks.

pwb,

I wish you would take the time and contact NAD tech support here in the US. Send them the photos you provided on this thread. The photos speak volumes to the damage that was caused by the cap in the amp. Make sure you tell them the sparks were still flying when you walked into the room. Let them know if you had not been home there is no telling what the damage may have been to your home.
Be polite! .....

I would hope NAD America would want the amp to determine exactly, if possible, what caused the catastrophic event that caused the electrolytic cap to blow its’ top the way it did. I would hope NAD America would pay for the shipping.

Jim
.
I bought the unit used 3 year ago on AudiogonThe unit was never left on when not in useThe local line voltage is 110
 A cd had finished playing 20 minutes earlier and I hadn't turned off the components yet. I was in another room when I heard the noise and ran in and saw the sparks shooting out the top. The room was filled with smoke. Another minute and the house would have been on fire. Very alarming and scary.  I will try to contact NAD.
Iaeles,

You likely nailed the reason. A surge taking out the bridge (shorting a diode) leading to AC into the capacitor causing it to overheat and burst. 


Jea48, the stored energy on the cap would not have had much effects. The issue is that the capacitor is unipolar. If you put reverse voltage on it which can happen with a shorted diode in a bridge, then it will conduct through the capacitor and rapidly heat up (seconds to minutes).


Erik, heat is the bigger killed with life reducing in 1/2 every 10C. Ripple current is also causes faster failure due to additional heating, though you can get rapid failure if you exceed the ripple current rating.


Would I attempt to fix it personally? Depends on evidence of secondary failure. Other lower value caps throughout the amp that are on the same voltage rail are also likely toast. Would also depend on disassembly capability. The electrolyte is probably water soluble. Distilled water can be used for cleaning of most components but avoid displays, pots, etc.
Diodes usually fail open circuit, unless the heat is so bad that the entire junction completely melts.  The only way that could happen is if the AC supply surged tremendously.  But then again, if that happened, you would have noticed other stuff breaking connected to the AC.

Was there anything on top of the amplifier which could have fell in (even a strand of hair) and shorted something ?

It could simply be a bad capacitor, from the factory. Although very rare, these may still happen.

WOW, I thought I had a bad one. Man oh man, it looks like the cap next to it was starting to swell also, I can't see the tops very well. I had a Wyred 4 Sound, 1000 (no R) bought used, blew a cap on one MB. 
The guy was running them on a single, 15 amp outlet. I suspect that caused the problem.  BIG BASS HEAD...I find out after the fact...

That would be a nightmare to clean propper, much less fix and not worry about it.. Sorry your traumatized buddy. That amp will leave you with nightmares. I just had valve amps kick a 20 amp breaker and fry. Must be an electrical virus. LOL it's not funny, I'm walking around swearing, with you for SURE.. Mine was self inflicted, yours..Geez ma neez..
Time for a change.. But NAD!! They have a good product. 

Respectfully and with regard
You can see that one or two of the other 80V caps are ready to pop.

Run too close to the voltage limit. they were getting gassy.

A manufacturer of those caps was absolutely full of it, or a parts supplier for that capacitor company was full of it, or..NAD was penny pinching and purchased dicey product, or the capacitor manufacturer flat out lied their face off about their quality standard.

Or some combination thereof. All I can say is that some of the decisions made, from/during that time period, have really hounded the company. eg, the 370 and associated power amp design...I’ve seen far too many of them with the same failure. Additionally, with shorted/dead caps, from the same company/source. Four in one year. I expect to see more.

It is many years down the road, though, so not really all that predictable. But there is a time to be careful about expenditure and a time to throw the penny pinching accountants out the nearest window.

My Best quick guess is that someone decided to try and save a nickel on capacitor costs, and took a shot, they took a swing, they took a flyer... and it worked.

For a while. for about a decade. Now that accounting decision has come back to bite them a bit...

Anyway, no one is perfect, and other than this blip, NAD has done really well. Sooner or later all manufacturers get bit by this sort of thing. It is just a matter of time, but is also tied to design and parts decisions. and it is not always the engineer or designer at fault. It can be accounting that is tied to parts costs, which can make iffy decisions. To do the right hing and satisfy customers who know little about the insides.. but.. a lot about prices they pay. It's a difficult balance at best.
I had an Innersound amp completely melt down.  Luckily it was in a pretty heavy chassis so the fire stayed inside and didn't burn my house down.  Innersound was out of business so no recourse but on something like that NAD should just get you set up with a new amp, doesn't matter if it's in warranty or if you bought it used.  That should not happen and if it does they should cover it, they are a big company.  Let us know what they say. 
The local line voltage is 110
The unit is rated 120v. Nominal is ±5% - 114v to 126v.

Running on 110v increases cap ripple if load high current delivery is required. Ripple increases heat. By chance did the unit smell hot?

Diodes usually fail open circuit
ESD often causes devices to fail as a short. A puddle of molten silicon makes an excellent conductor.

The only caps I've ever seen 'blow' have been due to reversed polarity. 

Failure analysis is often pure speculation... unless someone 'fesses up <:-0


I won't derail the thread, but .... this is the perfect example of why you should never remove the ground pin on an AC plug.

Failures are rare but they do happen.

This case wasn't probably caused by a bad ground pin, but this kind of failure can happen and you want the chassis to be grounded by a big beefy conductor when it does.
Ejlif


Given the age, I don't think they are under any obligation to replace (or fix) unless this has happened to any statistically significant number of units with the same parts. It would be a good PR move. Lots of things "shouldn't", happen, but failures do, and taking into account every potential failure mode would make things expensive and compromise performance.


Not being the original owner, we can't be 100% sure there were never any previous repairs, surge events, etc.  
Sorry for your loss.  The picture was truly worth a thousand words.  NAD should give you some help.  They might not replace it, but should offer you a steep discount on a current production model.  Keep us informed.
Dump this unit asap as its pure garbage.Good luck
Hello, 
I agree with contacting NAD as long as they cover the shipping. It looks like a swap of all the boards is the only real fix. If I owned NAD, I would give you credit for a new one and figure out what failed on the old one. Hopefully, the OP would come back to this thread and note they stood behind their product. This would make it Ok for me to buy NAD. If they don’t I would post it here for the world to know. Anything can happen to a piece of electronic. What is most important is do they stand behind their products and was this a one time occurrence. Sorry for the loss. This is why I prefer separates, but I do get this was not your primary system. 
How old is the unit ?
why not call Nad and ask to speak tothe sales and marketing rep . If he cares about the companies reputation 
then he will want to help making things right . This approach makes sense.
I agree... you should have contacted NAD as soon as this happened. If *I* was NAD, I would want to know about this right away so that I could try to make it right. 

I had something similar happen to my Porsche 10 years ago. I had just purchased a 2003 911 Turbo in 2010, it had about 32k miles on it, and was bone-stock; the car was now 7 years old, and well out of warranty. After 2 months of ownership, while accelerating (normally) from a stop light, the engine died. Come to find out, a timing chain had snapped, and the whole engine had to come apart. I made the mistake of reaching out to several different people and shops, including Porsche. Porsche of America got back to me, said they had only ever heard of this happening maybe 4-5 times, and offered to cover the tear-down, diagnosis, and  rebuild of the engine!!! However, I already had it at a independent shop, and they had already torn the engine down (prior to my approval to do so, actually...), so I was out any sort of charity repair work from the manufacturer.

Moral of the story, always talk to the manufacturer FIRST. Worst case scenario, they apologize and send you off to invest in another brand.

Please keep us posted though, I'm very curious about what they will say...

- justin
"If they don’t I would post it here for the world to know."

Shaking down a supplier for a 10+ year old+ electronic product, with a history you don't know (not the first owner), that could have experienced surge events, or other events outside the suppliers control, is not cool.