View Full Version : Will a component explosion always create a burn?

Jon Gerow
10-17-2006, 06:56 PM
Let's say I turned a PSU on and there was a pop and a bright white flash of light.

Shouldn't I be able to see a "burn mark" coming from the source of the light if I were to open the unit up for investigation?

Let's just say "I'm curious."

10-17-2006, 07:09 PM
Pop+light=an arc of current. In an inert atmospher there would be no blackened burn mark, just a tiny creater from the transfer of conductant. Kind of like deplating.
So yes, with careful inspection something should show. I'm old and would need good light/magnification. Use great care not to wipe away any burn mark. It the Trio acting up already. ;)

Oklahoma Wolf
10-17-2006, 07:18 PM
Most of the time there should be a mark. Might be a little hard to see at times though.

10-17-2006, 07:34 PM
If a cap were to blow there'd also be fluff from the electrolytic material. I've blown several caps in my car stereo :D For a transistor it's easy, just look for the transistor with the body blown open and possibly a leg sticking up at right angles to the other two. Yes, I've blown a few of them up too.

Jon Gerow
10-17-2006, 09:11 PM
Not a cap. It was a sharp pop. Like when a transformer blows. ;)

But I can't see any damage.

Oklahoma Wolf
10-17-2006, 09:15 PM
I would look carefully at anything attached to a heatsink... chances are, you'll find the culprit there somewhere.

Jon Gerow
10-17-2006, 09:39 PM
Everything attached to the sinks seem to be in tact... oh wait.. "theoretically."

10-17-2006, 11:10 PM
It all depends. If you have a large POP and a flash of light, it could be a lot of things. Depending on whati t aws, it could be a small short, a fuse, transformer, etc.

When I had a maxtor 200gb give up the magic blue smoke, it went POP, and let off the smoke. Aside from the slight pock marks on the IC and a bit of melted plastic, there was no burned material, no burned pcb, and only a lot of heat was created. Enough to discolor the casing of my Seagate.

I'd try looking all around the transformer. It might be the PCB side which is why you cant see any damage.
Then maxtor replaced the drive with a 300gb.

Super Nade
10-17-2006, 11:35 PM
It all depends. If you have a large POP and a flash of light, it could be a lot of things. Depending on whati t aws, it could be a small short, a fuse, transformer, etc.

A fuse does burn out brightly. Blew a measely 2A one today. :D

10-23-2006, 06:30 PM
Turn it on again. lol

I've heard pops from flashovers but you should never have a flashover in a PC power supply!

10-25-2006, 05:04 PM
The pop means that there was vibration/movement in the air. The light means that there was a conversion of energy to light. Since almost all controlled conversions, never mind unexpected conversions, are not efficient I would expect that there should be some trace of the event (smoke, smudge etc). The problem is whether it is enough to be observable. A different but close example is static electricity. At a previous residence my wife and I and were able to generate huge sparks which we inflicted on each other. These were audible (extremely so when discharged to the ear lobe) and could actually be seen flashing quite a distance. I may not remember correctly but I believe that some of them jumped nearly if not over one inch. Not once did we ever see any signs of the event on the metal light switch or other items that got zapped. (This is pre-computer - how else do you explain having the time to do this kind of mindless stuff.) My guess is that since it was high voltage with very low amps it could not create anything seen by normal means but possibly an electron microscope might have. If the same is possible in this case you may not see it. My gut reaction though is that there should be some trace. Since there was visible light I believe that it would have to have an opening to the air. Possibly it occurred near the edge of a components cover and was not violent enough to leave a mark.

11-01-2006, 11:17 AM
I had a Enermax 620 which blew and took out the RCD in the house, I reset the RCD and tried again a big blue flash from inside the PSU and the RCD had to be reset again, I repeated this around 6 times and coudn't see where it was comming from I couldn't see what was happening and I couldn't take it to pieces because it was under warranty I returned it and swapped for a OCZ GameXStream 700 then I read it has very high ripple so I'm thinking of swapping it out now.

Any idea's what was causing mine to blow the RCD? Also it didn't destroy any other bits in the PC I'm pleased to say.

11-03-2006, 06:10 PM
You should see something if it was bad enough to spark and kill the supply.

Now i had an easy one: It blew components off the PCB board. Thank god i had unhooked it from my system first. I never did figure out why it would do that, but i learned never to use a cheap PSU from that :D.

03-06-2007, 08:46 AM
dont forget the nice burning smell

Mad Machinist
04-15-2007, 07:07 PM
Just finished fixing a Klipsch 2.1 sub with dead power transistors on the right channel final amp........no burn marks and took me about 20 minutes with a magnifying glass and a led flashlight to see where the case of the transistor had ruptured.......

04-17-2007, 05:51 PM
I've seen cases where the package of a component just blew apart with no smoke or other traces other than the component itself being broken into several pieces.

Other times though failures can be more catastrophic. I've seen transistors internally short and heat up to the point where smoke comes out of it's plastic casing, to be followed by an arc at one of the pins melting away the metal pin much like an arc welder, until the gap is big enough the electricity stops arcing. The end. Such failures are extremely easy to spot obviously.

05-25-2007, 07:57 PM
I believe some of these psu's use MOVs for over current/over voltage protection. The sound like a cap gun going off if they exploded. The mylar will often blow a hole in the plastic casing.

07-22-2007, 04:17 PM
Assuming the PS still works, there could have either been a small piece of solder, or a trace that was too close to something else and 'self clearanced'. There could similarly have been a small stray wire that could vaporize and leave barely a trace.

I have had a couple components that have popped and flashed, but no magic smoke. Took them apart, inspected everything, power on and worked fine.

The most interesting example was an Alpine PDX 1.1000 amplifier (small 1000 watt car amp) that I recieved. The first time I powered it up there was a bright white flash from inside the unit and it blew the 4 20 amp fuses on the front panel. I took everything apart (quite a challeng with that one) and could find no damage. Replaced all the fuses and have been using it now for about 6 months without any signs of impending doom. . . Yet . . .