[Help] PC abruptly shuts off

[Soylent Joe]

About a week ago I rescued the lonely hull of a custom PC from the landfill at a flea market for $2.

From that I got...
- 3GHz Pentium D dual-core CPU with stock cooler
- Asus P5B-VM SE
- Antec Minuet 300 & 300W Antec slim PSU
- 80mm Antec Tri-Cool fan
- 512MB Corsair DDR2 400MHz

Then I...
- Added a spare 640GB 7200.12
- Replaced the memory with 2GB DDR2 400 from a Dell I have
- Added a SATA optical disc drive

Spoiler

The drive had a pre-existing Win7 install from the Dell it was in, so I hooked it up into the rig and ran that install. Everything worked alright, but it was randomly shutting off. I thought it was from mismatched drivers, so I wiped the drive and installed Win7 so it would match up with the new hardware.

Now that it has a clean install with the proper drivers, it is still shutting off. Sometimes it can be used for an hour with no problem, other times it shuts off 3 times or more before it can even fully boot up. Each time it's like someone is yanking the power cable out of it.

I wiggled the computer around while it was running, with no result. That shows that it's most likely not a loose cable. One thing that tends to set it off is installing Windows updates. Almost every time I click "install updates" and it starts doing it's thing, it dies.

So there are a lot of different things I could try. I'm thinking of moving the guts to my little mATX test bench with another PSU to see if that case is causing a short or that slim PSU is messing up.

What else could I do to get this thing running right? I'd like to get it up for sale soon.

Thanks.

[Jstn7477]

I'd definitely check out the PSU. You probably want to check temperatures too, as it could be hitting some threshold and shutting the system down to protect it from damage.

[Soylent Joe]

Originally Posted by Jstn7477

I'd definitely check out the PSU. You probably want to check temperatures too, as it could be hitting some threshold and shutting the system down to protect it from damage.

Yes the temps were the first thing I thought to check, esp. with that Intel cooler. The CPU runs in the mid 40's, and there's no graphics card that would be overheating.

[lilhasselhoffer]

I believe what was meant was the overheating protection in the PSU.

Dell is notorious for components failing, then generating a boat load of heat. The extra heat from high usage trips the overcurrent protection, and shuts the system down.

This occurs at start-up because fans and disks taken higher amperage to start spinning. The windows update, depending upon what you're installing, functionally makes sure all of the hardware is active by cycling it. This power draw does the same thing as what happens at start-up.

Replace the PSU with something that's got some chutzpah. If the power cycling continues then the mobo or processor is slagged. At that point you can send your "deal" out to pasture given that it will cost more to refurbish than it is worth.

[Fourstaff]

Check other parts for overheating too, like northbridge etc.

[Bo$$]

try running memtest86+
probably PSU problem

[Soylent Joe]

Is it normal for the south bridge to be hot to where it will burn you after touching it for more than about 3 seconds?

[Velvet Wafer]

Originally Posted by Soylent Joe

Is it normal for the south bridge to be hot to where it will burn you after touching it for more than about 3 seconds?

if the airflow is insufficient, yes. no part on the board should be warmer than handwarm (so that its no problem to touch it, even for hours)

[Soylent Joe]

Well it's really hot, even with the panel off and ceiling fan blowing down on it. The only way to cool it down would be to put a fan in there for it.

It's running off a spare PSU I have and has been doing well. I'm going to stress it right now to see if it holds up, hopefully it will.

Edit: Also I remember that it was quite noisy and I couldn't quite tell which fan was making the racket. It's clear now that it was the PSU fan, the whole system is now about half as loud with the other PSU. I guess I'll have to scavenge the net for a replacement power supply.


Are busted caps anything to shake a stick at? There are two just left of the CPU that are bulging a bit.

[Velvet Wafer]

Originally Posted by Soylent Joe

Well it's really hot, even with the panel off and ceiling fan blowing down on it. The only way to cool it down would be to put a fan in there for it.

It's running off a spare PSU I have and has been doing well. I'm going to stress it right now to see if it holds up, hopefully it will.

Edit: Also I remember that it was quite noisy and I couldn't quite tell which fan was making the racket. It's clear now that it was the PSU fan, the whole system is now about half as loud with the other PSU. I guess I'll have to scavenge the net for a replacement power supply.


Are busted caps anything to shake a stick at? There are two just left of the CPU that are bulging a bit.

You can improve the output and stability of a PSU, if you add more fan power to it... i used 2 fans, to bring a generic 400w to life again... still lives today

Busted Caps are very dangerous, if you let the board live with them for too long, they will most probably kill it... done that to a biostar board before, it makes a nice paperweight since then
you can resolder them, but if you dont want to do that, i guess its either: "RMA time" or "time for the bin"