Dying PSU or CPU/Mobo

Semple

Semple

Semple

Soldato
Joined
5 Mar 2010
Posts
12,634
So my PC randomly decided to reboot tonight - not a blue screen, so confident it's not a software issue. The problem now is that it doesn't stay on long before rebooting again.

System spec is in sig, it's coming up to about 5 years old, although it's not been running at 4Ghz for at least 3 years. Was originally bought as an overclocked bundle from OCUK, but warmed the room up too much in the summer, and it never needed running at 4Ghz, so i dropped it down to factory defaults at 2.8Ghz, which has been running stable without issues for the last 3 years. So it definitely cannot be an issue with the config. I did make a note of the Vcore voltage at 1.23750 volts.

I've also loaded the motherboards fail-safe config, which drops the multiplier a little so it's running at 2.00Ghz. However it still didn't stay powered up for long. Temperatures seem to run between 35C and 50C so definitely not overheating.

I'm wondering (and hoping it's the PSU over the mobo/cpu...) if the CPU is finally on its last legs and can no longer provide a stable voltage which is causing it to trip.

I've got a 450W PSU from my server i can hopefully try out tomorrow evening and see if the issue gets resolved or if i'm looking at a mobo/cpu issue.

The next question if it isn't a PSU problem is finding out if it's the motherboard or the CPU causing the problem.
 
5 years is a long time and warranty already expired.

What CPU cooler you used?

When the last time you applied CPU thermal paste?

Think it is PSU struggled to stayed powered up. What is OCZ 600W PSU model number?
 
If it were just once - it might be everything. Lower voltage in a network/micro-short blackout, for example. I've had something similar once - all PCs made a reset, except mine PC (OCZ Z-Series 850W "Gold"), but I was able to see very short light fade off (bulbs). TV & monitor turned off, TV-SAT made a reset :-)

Another time, all PCs in my house behaved strange. It was a day when Microsoft released major security update ;-) I don't know what was a relation between them, but all PCs (including mine) made a reset. Not at the same time - this is strange, but it was during one,same day. Including laptop. Lol. No light fade off has been observed whatsoever.

And there's something more - plenty users would say they experienced some occasional resets in whole life (PCs, of course). It's why "rock solid" sentence regards to some motherboards, and for others - doesn't ;-)

But in your case I'd say it's PSU. Try another one and see how it goes. But I'm not sure how "old" 450W PSU would keep up with your config. In the meanwhile - you might inspect motherboard looking for convex capacitors.
 
Last edited:
AthlonXP1800 said:
5 years is a long time and warranty already expired.

What CPU cooler you used?

When the last time you applied CPU thermal paste?

Think it is PSU struggled to stayed powered up. What is OCZ 600W PSU model number?
Click to expand...

CPU cooler is a Titan TTC-NK85TZ Fenrir.

Last time thermal paste was applied would have been right at the start, although no change in temperatures over the last few years, leads me to believe it's unlikely to have overheated.

PSU = ModXstream Pro 600W, OCZ600MXSP-UK
 
Semple said:
CPU cooler is a Titan TTC-NK85TZ Fenrir.

Last time thermal paste was applied would have been right at the start, although no change in temperatures over the last few years, leads me to believe it's unlikely to have overheated.

PSU = ModXstream Pro 600W, OCZ600MXSP-UK
Click to expand...

Accorded to CPU thermal paste manufacturers, you should reapply thermal paste on CPU once every year. But you had never cleaned the CPU and reapplied thermal paste in 5 years so the thermal paste would been completely dried up 3 years ago because of CPU overclocked. 6 years ago I applied Arctic Cooling MX-2 thermal paste on Phenom II X4 940 CPU, I never thought of reapply thermal paste once a year but 3 years later when I was in processed to removed AMD motherboard upgraded to Intel motherboard, I unclipped AMD CPU cooler and to my surprise the cooler didn't sticked to CPU it just slipped off and noticed the thermal paste was completely dried up. So I learnt my lesson now cleaned up and reapplied Arctic Cooling MX-4 on Corsair CPU cooler once a year.

I checked the PSU model, it has dual 25A +12V rails combined 50A total so I think one of +12V rail failed with your PC struggled to stayed powered up for long with just 1 25A +12V rail.
 
Have you tried booting up with the Gpu disconnected?

If yes, then just disconnect the Gpu and try with the 450W Psu. And see what happens.
 
Thanks both.

I will try without the GPU tonight before swapping over the PSU.

If the 12v rail failed on the GPU then i'd imagine the system would stay online, but have no display output. I think there's a 12v rail for the CPU as well, which might be the one that's failing.
 
Update from last night, i disconnected 4 HDDs and 2 monitors, leaving just the single monitor and SSD. Booted into windows which stayed up for the few mins, so i set the clocks back to default speed and then ran Prime95 under heavy load for an hour. System stayed up without any issues, CPU was running around 50-55c.

Ran heaven benchmark on the GPU, and within about 5 mins it lost power again. Temps reached about 80-85C. I've removed the card to give it a clean, and when i can get the silly cover off, i'll reseat the cooler and try again.

So at this stage i can say it's unlikely to be the CPU/mobo, and is either the GPU heating up too much, or still the PSU not being able to provide a stable voltage when the GPU increases demand for it.
 
You must log in or register to reply here.
Back
Top Bottom