NAD 712 power amp intermittent problem

Rhod

New Member
Good evening, folks. I wonder if there's anyone who can help. I'm looking at my Dad's old NAD receiver for him. The story is: it stopped working for him and he left it for a few weeks. I took a look inside to check the fuses, found they were ok, and then realised on testing the unit was working perfectly. It worked for a little while, then stopped again. Fast forward a few weeks, I bring the unit back to my place for a closer look. Same thing again: on initial testing all worked perfectly, all inputs, all outputs. The next day I tried it again, this time nothing, or more accurately:
Power up: relay clicks, LEDs/display operate as normal, some initial hum audible over headphones which reduces over a short time. No signal audible over headphones or either speaker channel.

Taking a line out of the preamp out confirms that all is well there, and that the problem lies in the power amp section.

Usually I would suspect transistors but not for an intermittent issue!

Do you knowledgeable folk have any thoughts about what I should look at next? I'm armed with a multimeter, a soldering iron, some common sense, some experience with electronic repair but not a lot of electronic theory. I'm willing to learn!
Many thanks for your time. Happy Christmas!
 
It's not an amp I know, but I had a look at the schematics (the service manual is available free on the hifiengine.com website), and can give you a few thoughts that might help:

It's good you've been able to isolated the problem(s) to the main amps, that makes it a bit easier. With both channels affected, it's likely to be something 'common' to both channels, e.g., dirty switches / contacts, power supply problems, or protection / relay circuits, rather than anything wrong 'within' the main amp circuits, which are separate circuits for the 2 channels.

It would be worth properly cleaning / deoxit-ing any speaker related switches & the headphone socket (that I think also incorporates switching), instructions are on the forum: http://audiokarma.org/forums/index.php?threads/the-idiots-guide-to-using-deoxit-revisited.207005/ - You have nothing to lose, and dirty / oxidized contacts are one of THE most common (and often intermittent) problems with older equipment.

On the protection / relay side - the NAD 712 has no speaker protection relay, but uses a similar type of 'thermal breaker' (for 'protection') as the NAD 3020 does, e.g., a thermal breaker in each channel output. Both breakers would be unlikely to be faulty at the same time, so I think you can safely ignore everything protection or relay related. The relay that you can hear activating when you turn the amp on is part of the power supply, not a speaker protection relay.

The main amp power supplies would seem to be the logical next step after that. Initially, just have a good look around, ideally with magnification, for anything which has been running hot / may be heat damaged, as well as any solder joints that look possibly damaged ('dry joints' can often lead to intermittent faults - and NAD are generally 'famous' for solder joint issues), or caps that might be leaking.

Most of the amp runs on the various +/- 5V & +/-12V supplies, which, as the pre-amp seems to be working fine, are presumably all still good (?). The main amps are different though, and use the higher voltage supplies, -28V, +/-33V, and +38V, where I suspect there may be a problem (possibly a heat failed solder joint, maybe... ?). I'd probably start with the visual checks, and confirm that all the supply voltages are present & correct (as per SM pgs 17 & 18), and go from there.

Oh, and welcome to AK !
 
Dear Goldie99,
Thank you for your welcome, and your guidance: really helpful! I will follow your instructions, particularly those about the main amp power supply (having already been reasonably liberal with the contact cleaner!).
I will report back!

It's not an amp I know, but I had a look at the schematics (the service manual is available free on the hifiengine.com website), and can give you a few thoughts that might help:

It's good you've been able to isolated the problem(s) to the main amps, that makes it a bit easier. With both channels affected, it's likely to be something 'common' to both channels, e.g., dirty switches / contacts, power supply problems, or protection / relay circuits, rather than anything wrong 'within' the main amp circuits, which are separate circuits for the 2 channels.

It would be worth properly cleaning / deoxit-ing any speaker related switches & the headphone socket (that I think also incorporates switching), instructions are on the forum: http://audiokarma.org/forums/index.php?threads/the-idiots-guide-to-using-deoxit-revisited.207005/ - You have nothing to lose, and dirty / oxidized contacts are one of THE most common (and often intermittent) problems with older equipment.

On the protection / relay side - the NAD 712 has no speaker protection relay, but uses a similar type of 'thermal breaker' (for 'protection') as the NAD 3020 does, e.g., a thermal breaker in each channel output. Both breakers would be unlikely to be faulty at the same time, so I think you can safely ignore everything protection or relay related. The relay that you can hear activating when you turn the amp on is part of the power supply, not a speaker protection relay.

The main amp power supplies would seem to be the logical next step after that. Initially, just have a good look around, ideally with magnification, for anything which has been running hot / may be heat damaged, as well as any solder joints that look possibly damaged ('dry joints' can often lead to intermittent faults - and NAD are generally 'famous' for solder joint issues), or caps that might be leaking.

Most of the amp runs on the various +/- 5V & +/-12V supplies, which, as the pre-amp seems to be working fine, are presumably all still good (?). The main amps are different though, and use the higher voltage supplies, -28V, +/-33V, and +38V, where I suspect there may be a problem (possibly a heat failed solder joint, maybe... ?). I'd probably start with the visual checks, and confirm that all the supply voltages are present & correct (as per SM pgs 17 & 18), and go from there.

Oh, and welcome to AK !
 
Ok, an update. I started to check the voltages. I wonder if there's an easier way to do this?! I was looking at the values on the circuit diagram (e.g. +38v at C527+; -28v at C526-: I have to admit I was expecting a value of 66v on the meter here, but got 28v) and measuring the voltage between them following the circuit diagram which of course is a mirror image of the back of the pcb: tricky!

Was measuring other voltages when I decided to check and blow me if the blessed thing wasn't working again!

I did notice some heat damage around one of the rectifier transistors which has been previously repaired (pic attached - you can see the 3 heat sink screwheads, and the resoldered traces) the repaired area was looking a little cruddy so could it be this playing up again?

I guess I need to wait for it to stop working again to continue testing? Am I right do you think?
Thank you!
 

Attachments

  • IMG_20181231_000016.jpg
    IMG_20181231_000016.jpg
    81.4 KB · Views: 35
Rhod Welcome to AK. I would check all the solder connections. Any the look iffy would need to be reflowed just in case. It might just take one bad connection in the right place to silence the amp. Intermittent failures are the worst, some times to figure out.
 
Thanks, NAD80. Will do. I might also reseat that transistor against its heatsink with some new thermal paste.
If/when it stops working again I may be back to ask for some help in interpreting voltage checks!
Many thanks, and a Happy New year to all.
 
Rhod - looks like you may have found at least the area of your problem - it sounds like you have one (or more) failed solder joints, somewhere in the PS section, that's 'connecting' sometimes, and sometimes not, hence the intermittent nature of the problem. I suspect when it's not connected you're losing one (or more) of the power supplies, and maybe not just to the main amps.

I think the 'repaired' section is around IC502 (LM7812), which is the main +12V regulator, but IC503 (LM7912) for the -12V supply also looks to have been running quite hot, albeit less so. The +/-33V main amp supplies are also taken from the feed to those regulators. There could be thermal damage to joints or components around either of those areas, and I wouldn't assume it's the repaired section - I'd probably start just by trying to get some better high quality photos of those areas, top and bottom of the pcb, to look in more detail for any obvious issues, e.g., burnt components / leaking caps / broken joints, etc.

Beyond that, you might also be able to localize the problem with a chopstick, e.g., with the amp on, you could try gently prodding the various components in the PS section with a wooden chopstick (or similar, non-conducting, tool). You're trying to identify a failed joint by just moving it very slightly, so it either makes or breaks electrical contact, and causes the problem to appear (or go away). If you're lucky that might help identify exactly where the problem(s) is / are. You could do the same while monitoring the various PS voltages, if you use a set of mini-grabber DMM leads (if you're not using mini-grabbers, I hope you already have your probes insulated right down to the last ca. 1mm at the tips...).

With that amount of heat damage, I would try to minimize any soldering / de-soldering work, until you've (hopefully) been able to localize the problem a bit more - heat damaged traces / pads will be much more prone to lifting and further damage, hence the point to point rewiring that had to be used in the previous repair.

Could you please also advise which part of the world you're in (the schematics vary a little) ?
 
Last edited:
Thanks, Goldie99. I'm in the UK. Have prodded with my trusty bamboo kebab skewer and not been able to stop it working (!). I am currently borrowing my daughter's decent compact camera to take a couple of hi-res photos...the saga continues...


Rhod - looks like you may have found at least the area of your problem - it sounds like you have one (or more) failed solder joints, somewhere in the PS section, that's 'connecting' sometimes, and sometimes not, hence the intermittent nature of the problem. I suspect when it's not connected you're losing one (or more) of the power supplies, and maybe not just to the main amps.

I think the 'repaired' section is around IC502 (LM7812), which is the main +12V regulator, but IC503 (LM7912) for the -12V supply also looks to have been running quite hot, albeit less so. The +/-33V main amp supplies are also taken from the feed to those regulators. There could be thermal damage to joints or components around either of those areas, and I wouldn't assume it's the repaired section - I'd probably start just by trying to get some better high quality photos of those areas, top and bottom of the pcb, to look in more detail for any obvious issues, e.g., burnt components / leaking caps / broken joints, etc.

Beyond that, you might also be able to localize the problem with a chopstick, e.g., with the amp on, you could try gently prodding the various components in the PS section with a wooden chopstick (or similar, non-conducting, tool). You're trying to identify a failed joint by just moving it very slightly, so it either makes or breaks electrical contact, and causes the problem to appear (or go away). If you're lucky that might help identify exactly where the problem(s) is / are. You could do the same while monitoring the various PS voltages, if you use a set of mini-grabber DMM leads (if you're not using mini-grabbers, I hope you already have your probes insulated right down to the last ca. 1mm at the tips...).

With that amount of heat damage, I would try to minimize any soldering / de-soldering work, until you've (hopefully) been able to localize the problem a bit more - heat damaged traces / pads will be much more prone to lifting and further damage, hence the point to point rewiring that had to be used in the previous repair.

Could you please also advise which part of the world you're in (the schematics vary a little) ?
 
Thanks, Goldie99. I'm in the UK. Have prodded with my trusty bamboo kebab skewer and not been able to stop it working (!). I am currently borrowing my daughter's decent compact camera to take a couple of hi-res photos...the saga continues...

Bringing you up to date - the machine has been working for a few days now. I have not been able to isolate the problem, but have cleaned up the heatsinks for the three power regulation transistors (IC501-3), reattached the heatsinks more robustly (the fixing screws were previously not tight, possibly due to the continual thermal expansion/contraction), applied new thermal paste (the old was dried and flaky and I'm guessing not doing its job efficiently, especially when mated to wobbly heatsink!) and reflowed solder around the transistor legs where the connections looked dodgy.

It seems to be working reliably now, but IC502 (LM7812) as picked out by Goldie99 does run very hot (heatsink uncomfortable to touch once it's been running for a little while) and much hotter than both IC501 and IC503. This doesn't seem right, as the heatsinks are all the same size.

I'm hoping that someone who is good at interpreting circuit diagrams (hello, Goldie99 ;-) ) may give an opinion as to whether the circuit is designed for IC502 to run this hot, or whether I need to take some more action e.g.:
  • Increase the size of the heat sink
  • Replace one or more components if they are the cause of overheating
  • Anything else?
Many thanks for your help - it's been really good to be doing this project with some additional input from this forum.
 
Not sure without photos, but I assume you have 2X Zener diodes for D515 & D516 ?

If so, (carefully) check the voltage drops across each of them, e.g., what's the voltage where it says +33V on the schematic (cathode end of D515) ? voltage at the junction of the anode of D515 & cathode of D516 ? and the voltage at the anode end of D516 ? you could possibly have one of them shorted, so the regulator overheats & shuts down thermally (maybe ?). All a bit of a guess without voltage data.

The regulators will run hot, otherwise NAD wouldn't have fitted heatsinks, but "uncomfortably hot" does sound possibly a bit too high.
 
Last edited:
Correct about the Zener diodes (photo uploaded).

Voltage measurements:
cathode end of D515: 36.5V
voltage at the junction of the anode of D515 & cathode of D516: 36.1V
voltage at the anode end of D516: 35.4V

I'm looking forward to learning what that tells us!

Many thanks!
 

Attachments

  • DSC04485.JPG
    DSC04485.JPG
    78 KB · Views: 21
Certainly not the values I was expecting - the schematic lists D515 & D516 as 3.3V Zeners, so I would have expected something around 6.6V drop across the pair - instead you have only 1.4V, so the input voltage to the IC502 regulator is close to, or even above, the maximum for an LM7812. I suspect that's why it's been running so hot.

That's my best take on it anyway - I'd put new 3.3V Zeners in - but you'll need to be very careful with all the trace damage you have (maybe fit & solder them to the existing leads above the pcb ?).

It would be interesting to see what voltages you measure for the corresponding positions around D517 & D518.
 
Thank you, Goldie99.

Voltage measurements from the same physical positions on these diodes (polarity of diodes is reversed compared to D515/516)

Anode end of D517: -37.2V
Junction of D517 and D518: -33.2V
Cathode end of D518: -29.2V

Voltage difference seems to correspond more with what you suggested, am I right?

I have a bunch of zener diodes that I acquired with a big box of various tools and components. Sadly there are no part no.s and I haven't got a variable supply to work out the ratings! I will order some...

Thanks for your help again...
 
The D517 / D518 voltages are much closer to what I'd expect, and since the IC503 regulator is less (current) loaded than IC502, are probably OK.

The RS Zeners are 5W rated, so should be fine, if they fit physically - keep them mounted off the pcb for cooling. If they're too big, Cricklewood Electronics still have 3.3V, 1.3W rated.

Good luck.
 
Ok, diodes arrived: ended up ordering from Cricklewood as I hadn't appreciated how much wider the legs are on the 5w ones. Good efficient service from them. Fitted them: decided to take your advice, Goldie99, and fit to the existing leads, given the amount of point to point repair underneath.

Tested, and works ok so far. Measurements across the diodes are now 37.3v, 33.7v and 30.4v, which looks a bit more healthy!

I will test it for a longer session and see how warm the transistors are running, but so far, so good! Thank you for your help!

Now, the next little project is to replace the blown display lamps with a couple of LEDs, I think. I suspect there will be a thread or two about that!
 
Ok, run the amp for an hour or so. Actually sounding pretty good with a couple of old Goodmans Maxim 2s that I've got knocking around - good low budget stuff!

Anyway, IC502 is still running pretty hot: ICs 501 and 503 are warm, but 502 still running very hot!

Anything else I should check do you think?!
Thank you...
 
Those voltages sound a bit better, at least the new Zeners are working, the others weren't.

The supply voltages are still quite high, e.g., the +37.3V on the +ve supply, and the -37.2V you measured earlier for the -ve supply. The schematic indicated values (+/-33V) seem to be for a 230VAC mains transformer, so they're probably just higher than 'expected' because the UK still runs at 240VAC, and sometimes higher. In any case, with them being high, the IC502 & 503 regulators are having to work harder to maintain the +/-12V regulated supplies, with the 'work' appearing as extra heat.

IC502 is worse / hotter than IC503 simply because there is a higher current demand on the +12V supply, than on the -12V supply. I wonder if it's not just a rather 'weak' point of the design, especially for use in UK (?).

Without going down the route of modifying the design, I'm not sure there's much more you can do except keep an eye on it, as you sort any other problem(s). The LM7812 regulator does incorporate internal thermal 'protection', which should shut down the regulator if it gets too hot - I suspect that's possibly why the amp was intermittently cutting out before, but it's harder to prove. Anyway, now you have the new Zeners in, the supply voltage to the LM7812 is already 5V lower than it was (now 30.4V, was +35.4V), hopefully that'll be enough.
 
Really helpful, thank you. I wondered if the hot running was built into the design as the original repair was done in 1998! I didn't think about the implications of the mains voltage.
I'm not sure that the thermal cutout was the only cause of the fault unless it can take several days, powered off, to reset. Maybe it can!

I think I will extend the heatsink.... and keep it on long term test....!
Thank you!
 
I shouldn't worry too much about the mains supply voltage. Any equipment for the U.K. (and E.U.) market must be suitable for a voltage range of 220 to 240 volts ac. The U.K. stated mains voltage is 230vac nominal, but quite a few areas are still at 240 volts. This is quite permissible under the Electricity Supply and Continuity Regulations.
n.b. I work in the U.K. power industry.
 
Back
Top Bottom