oh probably. Look for the "skidmark" and we can go from there.

Anything in a PC can smell burnt and then make the system not power on. Until you narrow down the problem there's no way to tell if it can be fixed. I'm assuming you haven't found any obvious burnt spots, or else you'd have mentioned it.

First off, check for any swollen or leaky capacitors on the motherboard, your video card or inside the power supply. That isn't normally the source of a burning smell but it can certainly cause other things to suddenly go that way, and it is the most common problem with boards from that era.

Second, check for any pins, solder joints, chip legs, etc. that are bent or touching anything they shouldn't.

Third (or maybe this should be first now that I think about it...), make sure your CPU heatsink is attached properly. Athlon XPs were a bit of a pain in the butt to install coolers on and I don't believe they knew how to save themselves from burning up if the cooler wasn't installed correctly. I remember boiling water on an old Socket A Athlon or Duron without a heatsink attached.

Almost anything can be saved, especially if it is just a minor short like you describe. It all depends on the level of effort you put into saving it. As far as cost for repair, it should be minimal doing it yourself (I'd assume less than $20 if you need to buy IC components) Time is the most costly -- tracing everything down, desoldering, and testing components to see what is dead. If a trace is vaporized/broke from a high current event, that can also be repaired. If there are no damaged pads or traces, it should be a dead simple repair and you lucked out.

Of course you can save it. It just depends on how you value your time vs the cost of buying a replacement.
Some of course get enjoyment from the repairs!
Perhaps you could get some high res pictures for the eyes on this forum to look over.

I haven't pulled the board out of the case yet to properly examine but I didn't notice any bulged or domed caps. I did have a heat sink properly installed with its fan running during the test and the power actually stayed on when I was smelling burning, I had to cut the power myself. Before the burning smell the computer was running but not posting.

good chance its some capacitors, those socket A boards are around the time of the capacitor plague and bad cheap caps

Could be, on the other hand Gigabyte tended to have better caps than most. I'd say core crunch is at least as likely.

dionb wrote on 2020-02-16, 23:06:

Could be, on the other hand Gigabyte tended to have better caps than most. I'd say core crunch is at least as likely.

Core crunch? Do you mean the CPU die? I know the die getting broken was a problem back in this era...

Both socket A and socket 370 cpu's had no thermal limiting and no LID. If heat sink was not attached properly would very quickly fry the cpu, sometimes in a matter of seconds.

Horun wrote on 2020-02-17, 02:06:

Both socket A and socket 370 cpu's had no thermal limiting and no LID. If heat sink was not attached properly would very quickly fry the cpu, sometimes in a matter of seconds.

I used thermal paste on the die and the heat sink certainly seemed to be touching said die so I don't think it overheated...but I could be wrong.

Guys, are you sure it's a good idea to raise EvieSigma's hopes?

Let us be honest. The board is probably dead for good. If EvieSigma had the skill and the resources to fix it, he/she would not have made this thread.

True, but you don't learn when dont ask questions.

It's not like the matrix where you can just upload skills to your brain.

Yet.

RacoonRider wrote on 2020-02-17, 03:25:

Guys, are you sure it's a good idea to raise EvieSigma's hopes?

Let us be honest. The board is probably dead for good. If EvieSigma had the skill and the resources to fix it, he/she would not have made this thread.

well, the question was if it is even possible ^^
but without knowing exactly what exactly is the cause it's really hard to tell... it might be as easy as replacing a simple component on the board... or as bad as all the chips being be fried.

in any case, pictures and checking the board for burn marks would definitely help.

EvieSigma wrote on 2020-02-16, 23:20:

[...]

Core crunch? Do you mean the CPU die? I know the die getting broken was a problem back in this era...

Yes, that's what i'm referring to. Essentially the die is a bit of brittle glass, and in the Athlon(XP) series it was very exposed and very easy to break a corner off. That generally means the end for the CPU.

Thermal death - as referenced above - is another potential cause of CPU (and frequently) motherboard death, but I'd say it's less common than plain mechanical destruction. Then again, a burning smell implies something got too hot. Have you identified what yet?

I ran one with a broken corner just fine for several years ^^

I have an update! I examined the board and found what seems to be the damaged component. The only problem is that I don't know what this component is, and the damage makes its markings unreadable.

https://twitter.com/EvieSigma/status/12294632 … 8675546112?s=09

EvieSigma wrote on 2020-02-17, 17:53:

I have an update! I examined the board and found what seems to be the damaged component. The only problem is that I don't know what this component is, and the damage makes its markings unreadable.

https://twitter.com/EvieSigma/status/12294632 … 8675546112?s=09

That's a MOSFET in the power regulation, by the look of it for the AGP power supply. Generally when one of these goes it's bad news. MOSFETs are pretty robust, so by the time they fry, there's probably a cascade of other damage. Usual trigger is bad caps messing up stuff, although an incorrectly inserted or shorted-out AGP card could also have been to blame.

That's not to say that the board is by definition beyond repair, but it's definitely not trivial - at the very least you need to SMD-desolder the MOSFET without damaging the board, then solder on a replacement - but I'd check everything between ATX PSU connector and this MOSFET before even touching the soldering iron.

14N03LA MOSFET is what should be populated there. It looks like several different models were used there throughout different manufacturing dates.. probably just depended on what equivalent part they had on hand at the time.

Infineon part # P-TO263-3-2 ... 25V, 30 amp

mosfets don't pop like that without reason (generally speaking)

I see two likely scenarios:
#1: Shorted caps. you should replace all the large electrolytic caps (and the bad part), should be right as rain after that.
#2: You plugged something into the agp port that overloaded that mosfet