Is this slowly cooking the board?

[tino]

Hello everyone. Glad to be here with people who know much more than I. I came here as a result of something I found in two of my three ART SLA1 amplifiers.

So I've got two of them which are 5-6 years old and they are running 24/7 and most of the time idling and at best, putting out a few watts per channel.  I'm really happy with the sound quality of these things so I've picked up a brand new one.  I've noticed for a while that these amps (all) have a very low level buzzing from the speaker which is only audible if you put your ear right up to the speaker so I went looking for reasons and I found the signal cable was routed very close to the transformer on all units so I re-routed it a bit further on all three amps and that nearly eliminated the issue.

However that's when I discovered the two older amps have a darkening of the circuit board right by one of the larger (I assume 1W) resistors. The new amp of course does not have this. I suspect over time, the heat from the resistor is doing this. How big an issue is this? Many thanks in advance.

PS. Here is the pic of the old amp with the discolored board and the new amp with no discoloration.

[Rog520]

I've seen solder joints fail over time on excessively hot components. The joint won't (shouldn't!) melt, but may become brittle where the lead penetrates it, especially if the component is large or there's vibration or movement.

[SeanB]

Simplest is to buy new fusible resistors and some ceramic standoffs to get the resistor to stand off the board by around 15mm so it can keep cool. If you just get the standoffs just solder a stiff copper wire to the existing resistors.

Something like these on the leads to hold them off the board, just the first image on search.

http://www.hxpceramics.com/index.php/product/Steatite-ceramic-bead.html

[picandmix]

If those resistors were overheating then I would expect to see the pcb darkening under them, whereas your picture seems to show the darkening at the joints, which to me means the joints have gone dry, so whip the board up and clean off all the old solder and clean the resistors wire with a fine file or similar and then resolder.

It also looks like some darkening under that cap and diode below the resistors, that might be down to movement so use some silicone or hot melt glue to secure the cap after resoldering.

Also while in there check the tops of all the electrolytic caps are flat and not blowing outward, indicating they are failing.
Purists might say replace them all with new good quality ones.

Out of interest whos amp is it and what wattage does it do ?

[tino]

thanks everyone for thenawesome suggestions. Im going to take the bottom off the amp and have a look. so weird that both older units show the exact same darkened areas.

oh btw picandmix, the amps are Art sla1 and are rated 100wpc.

[AndyC_772]

Wtf? Solder joints don't "go dry" over time. They may fracture through heat cycling or vibration, but that's not the same thing as a dry joint. A dry joint is one in which the solder doesn't properly wet and adhere to the component or PCB in the first place, usually because of a lack of heat or contamination.

Chances are that resistor runs hot, and the PCB will naturally darken a bit as a result. It really isn't necessarily a problem.

It's also possible that the darkening is just dust which has been drawn into that area by local air currents, and baked on to the surface of the board. In that case, it'll clean off with a stiff brush and some IPA, but I really wouldn't bother.

[wraper]

This could fry/kill two electrolytic capacitors which are nearby. And considering the age and 24x7 operation, probably, they are already dead.

[SeanB]

They are Tocon caps, so very likely the ESR is now in the ohm range. If i was going to pull the board out to do the resistors on longer leads I would at the same time just replace them at the same time. Use 105c devices, as they then will last at least another decade or two even if they are Crapxon devices. Those look like 85C devices.

[tino]

Thank you SeanB, wrapper and AndyC_772.   I went back and examined the board. I don't think the darkness is any kind of dust baked on. Looks like its something embedded into the PCB or underneath (as far as color) so it might be that the PCB is just darkening with heat.  I also did look at the caps and it appears all the electrolytic caps are rated for 105C.  None appear bulged but maybe they are slowly failing if indeed that area is getting very hot.

There is a 40mm fan in the back of the unit which is drawing air through the interior. The power transistors for the channels dissipate through the heatsinks on the side but to ensure nothing gets too hot, there is the little fan. Maybe I can just run the units with the top covers off. There wont be anything mounted above each unit anyway.

[picandmix]

Sounds like you are saying you have only reinspected the component side again rather than look at the track side ?

Can you lift  the board  without having having to unscrew all the power transistors ?

If you can easily get under there it might be well worth remaking those joints even if they appear sound, plus check out all the other joints in case any more seem suspect.

If it is difficult to lift the pcb, think I would run the amp with its covers on for a few hours, then with the power off, quickly remove the covers and carefully feel or use a temp sensor on the pcb by those resistors to see if it is actually running too hot compared to the rest of the board ?  if it is then it would seem lifting it would be worthwhile.

[SeanB]

Removing the cover means the little air the fan moves is not directed over the hot transistors and will actually increase the temperature of them.  Darkening occurs as the board sits at over around 70C for long periods, so just raising the resistors on spacers of ceramic will help a lot with cooling. New capacitors as well will help with long term reliability, as the originals are only rated for 1000h at elevated temperature, so now at 5 years on they likely have had 40000 hours of use, and will be in need of replacement.  Sadly the good quality ones are all snap in, so will need short wire leads soldered to the lugs to mount them to the board, or you will have to use a cheaper one with wire leads and shorter life.

[wraper]

Thank you SeanB, wrapper and AndyC_772.   I went back and examined the board. I don't think the darkness is any kind of dust baked on. Looks like its something embedded into the PCB or underneath (as far as color) so it might be that the PCB is just darkening with heat.  I also did look at the caps and it appears all the electrolytic caps are rated for 105C.  None appear bulged but maybe they are slowly failing if indeed that area is getting very hot.

There is a 40mm fan in the back of the unit which is drawing air through the interior. The power transistors for the channels dissipate through the heatsinks on the side but to ensure nothing gets too hot, there is the little fan. Maybe I can just run the units with the top covers off. There wont be anything mounted above each unit anyway.

Of course this is PCB material itself darkened. That 105^o rating usually is only for 1000-2000 hours operation. 85^oC rated quality capacitors live much longer than 105^o rated crap. Also small capacitors don't bulge when fail. Actually they do not have any vent on top to bulge.

They are Tocon caps, so very likely the ESR is now in the ohm range. If i was going to pull the board out to do the resistors on longer leads I would at the same time just replace them at the same time. Use 105c devices, as they then will last at least another decade or two even if they are Crapxon devices. Those look like 85C devices.

85^oC and 105^oC usually look the same unless some gold / silver colored low esr series.

[tino]

Thank you again Sean and wrapper!  I can take the bottom off to look but I'd have to break the epoxy or adhesive seal on top of the transformer nut because its holding the bottom on in addition to a few screws.  Once that's done, I can lift the bottom and replace any necessary parts or even just to look.

I have one of those Infrared temperature guns and I took the cover off and without the cover, after a few minutes, the large resistor settled at around 90C on one and 88C on the other amp. So maybe with the cover on and the little fan going, its sitting at like 80C or high 70's which would jive with what Sean mentioned about the board darkening at 70C for long periods of time.

Would it make any sense for me to replace the resistor with a slightly higher wattage item as well as putting the standoffs or just replace it with a similar one. I'm guessing its a 2W item. I guess I should make sure the inductance is around the same too right?  I guess I could also replace the caps with higher quality brands of the same value/voltage if I intend on keeping the units. They sound great so I really don't want to let them go .. EVER LOL

[tino]

BTW are Nichicon capacitors still considered good? I'd like to make sure if I'm going to take the effort to do this, I get something that actually is well made

Many thanks!

[SeanB]

Replace with the same type of resistor, and stand it off the board so it will run cooler. Just remount the original ones on stiff copper wire leads so they are 10-15mm off the board and they will be cool and not cook the board. Nichicon are good capacitors, just do not buy them off fleabay as they are likely to be fake or used ones.

[Rog520]

Nichicon makes some good caps, but really I'd select a replacement based on ratings. Any of the name brands (Nichicon, Chemicon, CDE, Rubycon, Panasonic, etc.) are going to be reliable replacements.

[AndyC_772]

If it's not broken, don't fix it.

Seriously, there's nothing wrong with your amplifier. Just because a resistor is getting hot, doesn't mean there's a fault. Slight darkening of the PCB material isn't a reason to start taking apart a perfectly good piece of kit.

I have no idea where the idea has come from that there might be anything wrong with the solder joints.

[tino]

So I'm not slowly burning the board to a crisp or going to cause problems with traces around that area or anything?  I'm just a bit paranoid that it might lead to a bigger problem down the road, but if that's not uncommon and its not an issue, then I have no issues leaving it alone. Just paranoid because I seriously love these amps. IMO they sound better than my friends older Bryston 2BLP and he kind of thinks so too.

[AlfBaz]

From the location pointed to by the arrow, it could also be heat generated from the tracks feeding the resistors. They may have routed them too thin in which case you can either solder a wire in parallel with the track or scrape the resist off and add solder to the track

[AndyC_772]

If these amps were mine, I'd make a mental note that the resistor runs hot, then put the lid straight back on, sit back, and enjoy some music.

You have two amps, both are working, and that component gets equally hot in both. That's a good sign, it tells you there's no actual fault causing it to get excessively hot.

If there's a component I worry about when it comes to getting warm, it's power connectors. Where you have bare (ie. unsoldered) metal to metal contact, there's no problem as long as the contact is clean. But once the surfaces start to oxidise, that causes a little heat to be generated at the contact point, which accelerates the oxidation process, and the result can be thermal runaway. The first you know of it is the product stops working, and inside there's a brown and black lump where the connector should be.

[Rog520]

I have no idea where the idea has come from that there might be anything wrong with the solder joints.

Depends on the quality of the board and of the soldering. It's not uncommon for hot components to cause degradation of the board and/or joints, especially with single-sided phenolic boards like the one pictured. I see no harm in giving it an inspection and touching up the joints. Better to spend a few moments checking it out now than for it to cause a more extended failure down the road.

[tino]

Hi everyone, I just wanted to thank everyone for the input again.  I did further testing, so with the cover completely off, the resistor finally settled at 105C.

Deciding that I'll never be comfortable with slowly heating the board (even if its seemingly ok at the moment and no problem), I picked up a couple to measure against the original and replace with the one that seems like its the closest in terms of resistance, capacitance and inductance at least according to my cheap LRC meter. One was a 2W 1K which physically looked identical. One 5W ceramic.

I believe my meter may be way off but at least it was consistent between the resistors.

So I measured .995Kohms, 7.67nF and 1.37mH  for resistance, capacitance and inductance of original 2W resistor
I measured .997K, 7.59nF and 1.38mH for the new 2W resistor
I measured 1.05K, 7.22nF and 1.39mH for the ceramic 5W resistor

The new resistors both claimed 5% tolerance.  I installed the 5W ceramic and left it raised off the board by about 1cm. After 1 hour of the system running, its measuring about 82C with the cover off. Much lower than the 105C for the original 2W.

Now one thing is very odd and I cannot for the life of me figure out is why the amp sounds better. My friend who has better ears than me (he works in broadcast and listens for errors, glitches and stuff like that on commercials and content produced before it goes on air (via decent headphones) also commented.

So before doing the swap, I had the level pots for both amps set via an 8ohm resistor across the speaker terminals, fed 60Hz tone through the preamp into the amp and measured the voltage across the resistor. Did this for both channels for both amps.  We listened to a few songs and both amps sound identical.

After the resistor swap on one - without anything else being changed like the amp level or preamp volume level, the one with the resistor swap to the larger 5W seems to reveal more clarity in instruments especially when things are complicated like classical orchestral music. Bass seems to have a little less overhang as if the woofers are under more control. More snap I guess. I'm not sure how this could be but we swapped back and forth and there is a fairly evident improvement.

I did research online and on some tube amp forums (which of course this amp is solid state), they talk about resistors having noise. This makes no sense to me but I don't know enough. Supposedly ceramic wirewounds having the least noise and sounding the best??

I don't even believe this resistor has anything to do with the signal path because I don't see a corresponding one for the other channel anywhere close to the output transistors. It only exists on the one side as far as I can see. Maybe its some component of the power supply or some other function like protection circuitry?

Anyway I modified the second one with the discolored circuit board  and it now sounds identical to the first one I modified - which now both sound a bit better than the new amp which is stock. Aside from the extra level of clarity I hear, it does not appear to sound dull or bright or different. Just somehow more revealing of instruments, like someone cleaned my ears and I'm listening again LOL.

So I don't know whats happened but I prefer it this way.