Maybe PSU or Mobo.. but not sure

[Riot_Light]

*Edit: think i posted this to the wrong forum. mods please move*

I have been having crashes with my computer for some time now. I have already attempted to fix the problem but it has returned and is now getting worse than ever. This thread is very similar http://www.tomshardware.co.uk/forum/id-1646906/tech-assistance-restarting-running-games-applications.html

but I will explain my problems as briefly and precisely as possible;

Bought this pc back in 2009.
Main specs:
intel i7 920
3 X gskill DDR3 2GB PC10666 1333MHz HK series.
HD4870 1Gb radeon made by 'powercolor'
MSI X58 platunum crossfire motherboard
Coolmaster PSU silent pro 700w
2 x 500gb HDD
DVDRW

So every now and again I would get random BSODs or freezes/crashes with no bsod all pointing to different faults.
BlueScreenView shows the BSODs date back to 2010
(2 in 2010, 4 in 2011, 1 in 2012, but many crashes were now not even bsods, just freezes, and 20 in 2013. Again with many of those not listed)

BSODs pointing to atikmpag.sys or dxgkrnl.sys became the main reason though.
However these crashes only occurred shortly after a cold boot. Occuring within the first 30-60 mins. Untill the graphics card bsods became the main cause just recently in Grid2.

The BSODs pointing to the graphics card became so frequent I borrowed an HD5870 from a friend and the system was much more stable (I say much more stable as I did get one crash but that, I thought was due to overheating the graphics card, as the fan on the graphics card had stopped working and it was being cooled by a case fan right next to it. Obviously it could not cool it enough to handle GRID2...).
When it did crash using that HD5870 after running Grid2 it failed to boot, instead seemly trying to power up then turn off, after several attempts I shut it down and left it 15 mins.
The computer reported “Warning!! the previous overclocking has failed...defaults will be loaded”

THIS COMPUTER IS NOT OVERCLOCKED however!!

Having all the BSODs gone while using this HD5870(without over stressing it) I thought it was my old graphics card so bought a new GTX660 only 1 or 2 months ago.

System was stable for quite some time. Being able to run GRID2 on High settings for hours with no problem.

The cold boot crashes have returned, again after 10-20 mins up time (Even with no activity. I turn on go afk for 20 mins come back and its frozen)
Sometimes I get a Bsod (the most recent and most common bsods are Page fault in non paged area) but many are just freezes with possible 'looping' sound.

I ran memtest86+ from a USB over night. 9 hours, 7 test complete, 0 errors.
Tried again the following day from cold boot, 10 – 20 mins in restart, reboots to “Warning!! the previous overclocking has failed...defaults will be loaded”

Since then the crashes and freezes are getting worse by the day, when it can crash up to 2 hours after boot, until yesterday it crashed after 12 hours up time. Then after several more crashes in quick succession (when gaming) I again got the “Warning!! the previous overclocking has failed...defaults will be loaded” error.
Again today it crashed shortly after boot just playing music and downloading, and then crashed while gaming.

I have kept an eye on my Temps since I was using the test graphics card. They system doesnt seem to be over heating with the cpu normally around 45, IOH 68 (apparently normal for this Mobo), system 47. Little change under stress but well within acceptable ranges.
When restarting form the “Warning!! the previous overclocking has failed...defaults will be loaded” temps are normal too.

I read about the capacitor plague but none of my capacitors look damaged. So am unsure if it is Mobo.

I do not have a PSU to test with. I will ask the IT department at work so see if they have one but I doubt it.

It would be great to hear any opinions you may have on the problem and if I can some how diagnose which it is without buying something I don’t need.


I am kinda hoping it is just PSU as that is a lot simpler to fix..

Thanks in advance for any help and advice you can provide.

 

[Eric Fisk CGD]

I read your question twice just to make sure I didn’t miss anything.

My father in laws computer was having a similar problem which became progressively worse as time wore on. Turns out that one of the major problems he had was the lint quilt that was on the processor’s heat sink that kept it nice and toasty warm. Obviously, that’s great for us on a winter’s afternoon, horrible for computers! By the time I got to it, the damage had been done. He cooked the CPU and may have also done damage to the mother board – when I opened the box all I could smell was the smoke of burnt components.

Which leads me to ask, how’s your ventilation? Have you ever taken a vacuum cleaner and mildly sucked out or blown out all the dust? PC’s that sit on the floor suck up a lot of dirt, dust, and pet hair like you can’t imagine – even in just a few months. Dust, dirt and lint can hide between the heat sink and the processor, so keep an eye on that – just in case. Double check on anything that might have accumulated in the graphics cards – especially the fans – and the power supply. Anything with a fan WILL collect performance sapping gunk.

With that said; since you’re keeping tabs on the temps then I gotta say that your power supply is the most likily cupret. I suspect that it’s simply gone bad and you might want to replace it. We bought a new power supply for $75 dollars and that made a world of difference in one of our units and solved some automatic shut-down problems we were having. If you get a new power supply and that doesn’t solve your problems, save it for when you build your home brew.

 

[Riot_Light]

I read your question twice just to make sure I didn’t miss anything.

My father in laws computer was

having a similar problem which became progressively worse as time wore on. Turns out that one of the major problems he had was the

lint quilt that was on the processor’s heat sink that kept it nice and toasty warm. Obviously, that’s great for us on a winter’s

afternoon, horrible for computers! By the time I got to it, the damage had been done. He cooked the CPU and may have also done

damage to the mother board – when I opened the box all I could smell was the smoke of burnt components.

Which leads me to ask,

how’s your ventilation? Have you ever taken a vacuum cleaner and mildly sucked out or blown out all the dust? PC’s that sit on the

floor suck up a lot of dirt, dust, and pet hair like you can’t imagine – even in just a few months. Dust, dirt and lint can hide

between the heat sink and the processor, so keep an eye on that – just in case. Double check on anything that might have

accumulated in the graphics cards – especially the fans – and the power supply. Anything with a fan WILL collect performance

sapping gunk.

With that said; since you’re keeping tabs on the temps then I gotta say that your power supply is the most likily

cupret. I suspect that it’s simply gone bad and you might want to replace it. We bought a new power supply for $75 dollars and that

made a world of difference in one of our units and solved some automatic shut-down problems we were having. If you get a new power

supply and that doesn’t solve your problems, save it for when you build your home brew.

I regulary clean the dust of off

all fans and heatsinks, so thats all good.

I couldnt wait any longer and took it to a PC repair shop. They tested it for 2 days

without 1 single crash!! I get it home.. crashed in 10 mins..

So seems as if its somthing not computer related.

I am now bypassing

the "surge protector" and it is a lot better, but i have just tried to run Rome Total War 2 and it just completely restarted it

self

So am now thinking it is more likely power supply comming from the wall.

My dad, who was an electrician, told me to run a lamp

at the same time and see if it flickers at the same time as the computer crashes.. I am going to do this now and then also try a

different wall socket!