Power supply started smoking

Thread Starter

KeithW

Joined Mar 24, 2013
15
I have one of these:

http://i.ebayimg.com/t/30V-5A-110V-...x2/$(KGrHqJ,!lgFELCCQ8rgBRKI)2eD2g~~60_57.JPG

The other day my basement smelled distinctly "melty" and after hunting around for a it I noticed a very thin stream of smoke coming from the fan vents (not a big obvious black cloud, just a difficult to find wisp).

I was only running it at 2.0V and drawing about 4-5A. It rated for 30V and 6A. It was running in constant-voltage mode, meaning the current wasn't even maxed out.

Obviously, I turned it off, unplugged it, and haven't used it since. I opened it up, saw a few large capacitors, waited a day and then hunted around inside it a little but couldn't find any obviously ruined parts.

What would be the procedure for diagnosing and potentially repairing the problem?

Thanks.
 

crutschow

Joined Mar 14, 2008
34,284
The worst-case power dissipation for a linear regulator is at low (or zero) output voltage at a high current. In such a situation the internal series control element has to dissipate maximum power.

So I expect you zapped the series control component (likely a transistor on a heat sink) so I would look there.
 

studiot

Joined Nov 9, 2007
4,998
basement? Mouse/spider (fried) inside?

Incidentally the power supply is working hardest at low voltage settings.

So it will be dissipating near max at 2volts, 5 amps output, probably something like (36-2) x 5 = 68 watts.
 

Stuntman

Joined Mar 28, 2011
222
Is this a switched or linear supply? Was the power supply still operating correctly when you shut it off?

My first rule of thumb would be look for discolored solder/thermal compound.

Next, I'd be very temped to plug the unit back in and trace down the smoke. However, realize, it may not smoke immediately. If you can confidently stay clear of mains voltage, your finger can do a great job of finding hot parts.

Consider, the part that was smoking is almost certainly already zapped, so your only goal is to find it before it de-laminates any pads/traces or smokes the solder mask.
 

Thread Starter

KeithW

Joined Mar 24, 2013
15
The worst-case power dissipation for a linear regulator is at low (or zero) output voltage at a high current. In such a situation the internal series control element has to dissipate maximum power.

So I expect you zapped the series control component (likely a transistor on a heat sink) so I would look there.
Ah, of course. Now that you point it out, that makes sense (that low voltage is actually the most expensive scenario, especially coupled with high current).

I'll take that into account.

Thanks.
 

Thread Starter

KeithW

Joined Mar 24, 2013
15
Is this a switched or linear supply? Was the power supply still operating correctly when you shut it off?

My first rule of thumb would be look for discolored solder/thermal compound.

Next, I'd be very temped to plug the unit back in and trace down the smoke. However, realize, it may not smoke immediately. If you can confidently stay clear of mains voltage, your finger can do a great job of finding hot parts.

Consider, the part that was smoking is almost certainly already zapped, so your only goal is to find it before it de-laminates any pads/traces or smokes the solder mask.
Thanks, I'll gingerly consider such an approach.
 

Thread Starter

KeithW

Joined Mar 24, 2013
15
Those are linear supplies, and they can get pretty hot especially at low voltage & high current.
So, if I bought a replacement rated for a higher current, 10A, (with the intention of still using at in my original current range of 4-5A), does that increased overhead make the replacement less likely to die in a similar fashion? Does that extra current capacity render the entire unit more capable of withstanding a given voltage/current scenario?

Alternatively, I suppose I could add a second fan (cut a hole in the top, install a fan, and wire it parallel to the existing fan).

I guess I need a lower voltage PS. I bought a 30V thinking the extra voltage capacity would give me lots of overhead, but maybe that just means it really really doesn't like to work at low voltages, whereas perhaps a lower max-voltage PS would simply operate better for my needs. Does that logic pan out?
 

MrChips

Joined Oct 2, 2009
30,712
A few lessons to be learned here. When you see or smell smoke, of course turn off the power immediately. But I would also open the unit right away and strike while the iron's hot. You want to be able to identify the burning component by sight, smell and touch. Don't wait until the obvious evidence is lost.
 

studiot

Joined Nov 9, 2007
4,998
Did you find a mouse inside?

I would be more inclined to look first for arcing internal connections. Many of these type of supply are poorly constructed, we have just been discussing another one that was not earthed at all, let alone properly earthed.

Look first for evidence at external lead connections.
Is the fuse screwed home properly?
Then look at internal termianls and connections.

Finally look turn it on and look inside from a safe distance. You may be able to spot the arcing.
 

Thread Starter

KeithW

Joined Mar 24, 2013
15
A few lessons to be learned here. When you see or smell smoke, of course turn off the power immediately. But I would also open the unit right away and strike while the iron's hot. You want to be able to identify the burning component by sight, smell and touch. Don't wait until the obvious evidence is lost.
Capacitors scare me. :) Point taken however.
 

Thread Starter

KeithW

Joined Mar 24, 2013
15
Did you find a mouse inside?

I would be more inclined to look first for arcing internal connections. Many of these type of supply are poorly constructed, we have just been discussing another one that was not earthed at all, let alone properly earthed.

Look first for evidence at external lead connections.
Is the fuse screwed home properly?
Then look at internal termianls and connections.

Finally look turn it on and look inside from a safe distance. You may be able to spot the arcing.
Whoa, "spot the arcing". You mean I might actually see an air-electrical connection? That's both awesome and terrifying. I'll consider it.
 

crutschow

Joined Mar 14, 2008
34,284
So, if I bought a replacement rated for a higher current, 10A, (with the intention of still using at in my original current range of 4-5A), does that increased overhead make the replacement less likely to die in a similar fashion? Does that extra current capacity render the entire unit more capable of withstanding a given voltage/current scenario?

Alternatively, I suppose I could add a second fan (cut a hole in the top, install a fan, and wire it parallel to the existing fan).

I guess I need a lower voltage PS. I bought a 30V thinking the extra voltage capacity would give me lots of overhead, but maybe that just means it really really doesn't like to work at low voltages, whereas perhaps a lower max-voltage PS would simply operate better for my needs. Does that logic pan out?
A higher current transistor will likely make little difference as it's the heat sink that is likely unsized.

Adding a second fan blowing directly on the heat sink would definitely help.

Yes, with a linear supply, a low voltage unit will dissipate less power at low voltages as compared to a higher voltage unit. Some high voltage units have dual voltage ranges (dual output windings on the transformer output) so it dissipates less power for low voltage usage.
 

studiot

Joined Nov 9, 2007
4,998
I think we should hang fire on suggesting modifications to a mains powered unit until we know for certain what is wrong with it, and that the OP has the capability to modify mains equipment.

Looking at the wiring inside the photos I linked to I think it far more likely that the burning is a result of a poor connection somewhere or as I said an electrocuted unwelcome visitor.
 

Thread Starter

KeithW

Joined Mar 24, 2013
15
Yeah, even though the PS in that thread looks vaguely similar to mine on the outside (we obviously got similar Asian PSs off ebay), mine is quite different some respects. For one thing, it has a three-prong plug. For another, when I opened it up, I believe the grounding is wired quite rationally. The ground banana plug on front (yellow) goes straight to the transformer chassis and then straight to the ground outlet plug...which I believe makes sense.

Since that entire discussion is mainly about the lack of a three prong power connection, I doubt it is too relevant to my situation. These PSs, which are packaged in similar external cases, seem to be quite different on the inside depending on where you get them.

I dunno, I'll keep digging around, see if I can make some sense of it.
 
Top