For the past few weeks, the PSU on my PC has been very loud (one of the fans wasn't spinning properly), but otherwise it would w**k fine even for extended periods of time. But on Tuesday night I was in the middle of typing a long e-mail and the whole thing just shut down and wouldn't restart. I figured the CPU had probably overheated - after leaving it for 3 hours to cool down, I turned it on again, it booted up fine and seemed to w**k ok - I switched it off overnight and decided to get a new PSU yesterday. After fitting the new PSU, the machine turned on fine, but after half an hour or so it turned itself off again and now it won't turn on at all. I can't see any loose connections inside the machine, there are no obvious problems with any of the components and there's definitely power going to the motherboard.

I'm going to try swapping the leads to the reset button and on/off button on the motherboard, just to make sure it's not a dodgy power switch, but if that doesn't w**k I don't know what more I can do to find the problem without taking it into a shop. Anybody got any ideas?

As you may very well know, 9 out of 10 times for intermittent shutdowns its a failing power supply - But as you've said you've already replaced it and the problem still continues. The second biggest cause is heat. Have you upgraded your PC lately? New graphics card or anything? moved your base unit from its regular position? Could even be due to this recent hot weather we've been having, assuming its been like that where you are. If so then run your machine with the base unit out in the open, maybe even with the side panel(s) off and a regular house fan cooling it at close range, if there's no problems then you know its heat. Do you currently have a good system flow? Intake at the front and exhaust at the back? If so then maybe try some fans with higher rpm's, or some side/top fans if you've got the drill holes or don't mind a bit of DIY drill w**k. Maybe even a new case

Is your new PSU powerful enough? did you check that there's enough wattage to go around?

Just out of curiosity, what CPU do you have and what temp does it run at and also what's your ambient temp if you can find out?

Well, it's not the on/off switch as it still wouldn't turn on even when I swapped the leads around. It does seem to be totally dead now - it's not a question of intermittent shutdowns anymore, but rather it just won't turn on.

Rob - it's an Athlon XP 2800+ running at 2.08 GHz, so no overclocking. I'm not sure what the normal operating temp was because I accidentally removed the desktop monitor program a while ago and never got round to replacing it. I think normal operating temp was about 50 °C from memory, but I could be wrong. And besides, the 2800+ is rated up to 85 °C so it would have to get seriously hot inside before the safety shutdown kicked in. However, when I examined the processor, I noticed that all the thermal conducting compound had been displaced around the actual processor chip and the top of the chip was in direct contact with the bottom of the cooling fan. I doubt this happened recently, and there didn't appear to be anything wrong with the chip, but it was just something I noticed. Ambient temp was normal RT, about 22 °C.

Before replacing the power pack the last upgrade was to add more RAM, but that was some months back. The old power pack was 400W and the new one is 550W so no problems there. System flow was fine - the fans were a little dusty, but there was a good air velocity coming out the back.

Ouch! a CPU in direct contact with the heatsink, could kill a chip dead. :fear:

Do you have another CPU you could try? Sounds like a process of eliminations in order :ermm:

I don't have another socket A-compatible processor, no. Most of the chips at w**k are either Intel or Athlon 64. Sounds like I'll have to try and find a PC repair shop <_<

It's nice to know that all the helpful "open 24 hours a day 7 days a week" computer repair services advertised in the yellow pages are actually not open 24 hours a day and also refuse to answer their phone on a saturday <_<

But judging by the scarcity of socket A processors, even on ebay, I think I may just bite the bullet and get a cheapo socket 939 processor/mobo. I had my eye on a dual core processor, but at £250 I think it's a little too expensive for now. Bloody computers.

I think just about everyone into computers have got their eye on an AMD dual core, but for £250+ :lol:

I had a spare low end Sempron not long ago, could have sent you it if only your machine broke earlier :rolleyes:

Have you tried plugging it in?

Or Turning it off and back on again?

:lol:

Sorry couldn't resist.....

Grrrrr sodding Novatech selling me a dud psu :mad:

It's w**king now, although not particularly well. I downloaded a cpu monitor program from t'internet and basically I can't do anything which requires intensive processing power over more than 5 mins. I haven't tried playing a DVD or watching the latest downloads yet, fingers crossed I can still do that. But I think it's time to save up for a dual core :)

AAAAAAAAAAAAAAAAAAAAAAAAAAAAAHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!



The fucking thing has died again! That's 3 PSUs in 3 weeks it's killed!

I don't understand why it's suddenly going through so many PSUs. Since the new PSU was fitted, the CPU temp has been around 52°C on about 30% usage and the HDDs are about 35 °C each - yes I know the weather's been warm, but there's been a good airflow, I've had the side panel permanently off and there's been a desktop fan on for the past 3 days. Plus it didn't have this problem last year, or the year before, on a lower-rated PSU.

Thankfully the PSU is still within the 30 day money-back period so hopefully I'll just get a new one. I'm wondering if there's a deeper-lying fault somewhere in my PC which is causing them to short out? But if that was the case, why did it take 2 weeks to go?

Oh, and the dual core 3800+ retail package is now down to £190.

I was told by an expert when I had temp problems a while ago that its better to have the case closed but with correct airflow rather than the side off and a fan (sounds daft I know).

I don't think the temp is defintely the problem - I'd be concerned with a 52 but not panicky. Sounds more fundamental (ie M/B) to me.

[vagueness]Got very limited motherboard knowledge on the fault finding side of things, but a simulated fault I once saw sounds not unlike yours and was all due to a faulty voltage regulator (another perfect source of intermittent power failures, and perhaps overheating if its on the right bus and sending its fault elsewhere? The overheating part could be complete bollarcks though, just a thought :ermm: )

Might even be worth taking a look for hairline cracks around your P1 port? You might see something obvious? :ermm:

Also, its very unlikely, but a quick BIOS flash might even solve your troubles, assuming its mobo related :ermm: [/vagueness]





Or maybe ...just maybe, its a problem with your flux capacitor, causing your dilithium crystal to overheat :ermm:

Ha,

Thats easy to fix just give it a few zaps with your sonic screwdriver and reverse the polarity of the neutron flow and your computer will be fine and dandy....

Hmm, ok, thanks Rob and Neil. And Dan, although I don't have a sonic screwdriver - anyone fancy lending me theirs for a day? ;)

I'll ask the Novatech techies when I take the broken PSU in later. I really hope it's not the mobo, I don't have the cash to upgrade to S939 yet :(