Replacement PSU needed?

Rik89 · 29 May 2014

Background info: Last week, having not used the system for a month, I removed the accumulated dust from my desktop for the first time in over a year, having noticed it was much hotter than it used to be, and in doing so removed the heatsink. Having put it back on, I didn't realise that it'd snapped a plastic prong so it wasn't properly connected. My i7 processor, I presume, then repeatedly shut down the system before I realised what I'd done.

Having bought a new heatsink my processor is now operating at 40'C during normal relatively idle use, but it is now periodically powering off and restarting (so I assume it's distinct from the problem above).

I put the system together in April 2010 and it's using a 'Corsair TX 850W ATX SLI Compliant Power Supply'.

How likely is it that this is where the problem lies at the PSU needs replacing?

Thanks

badger906 · 29 May 2014

have you monitored the temps of the cpu when it powers off?

cpus will shutdown when they hit thermal temp limits, but this doesnt rule out they may cause damage to themselves.

the only real test for a psu issue is to test with another, obviously not everyone has a spare laying around so this does become an issue.

linnemeyerhere · 29 May 2014

Somehow I doubt the issue is with your PSU. I would look more closely at the CPU and it's socket for damage and the proper install of thermal paste and the evenness of the coolers bond to the CPU itself?
Could you have over applied TIM and it's oozed over the side and into the socket? Remember when applying TIM "less is more" !

Rik89 · 29 May 2014

I can't say I adopted the less is more approach and it is conceivable that it oozed off the die (I did have to clean the edges). But if the processor is operating at 40'C, why would it restart? Incidentally, with the first problem, where it was the processor, the machine would power off and not restart, whereas now it does.

As you say Badger, that would be difficult. I was hoping you guys would just confirm it probably is the PSU so I could limit the expense / hassel to that, rather than taking it in and paying for someone to look at it for uncertain results.

It is on a cheap extension wire that I have replaced with an equally cheap extension (same result: spontaneous restart, even when idle) - could this be an issue?

Might the CPU have suffered damage on those 8-10 power ups before I realised that the heatsink was properly attached? It seems to be otherwise working fine though.

Britton30 · 29 May 2014

Less isn't really more but the proper amount IS needed, than being a drop of material about the size of a grain of rice in the center of a thoroughly clean CPU, only on the CPU. The heatsink mating surface should be equally cleaned then tightened down evenly.

I would make a guess and say you've likely damaged the CPU when you had the heatsink mounted wrong by overheating it. What new cooler did you get?

This free program will monitor and show your PSU voltages. http://www.hwinfo.com/

linnemeyerhere · 29 May 2014

Do you have access to another properly sized PSU you could borrow and then try? It can get expensive replacing things in hopes of solving these kinds of problem....speaking from past experience. Corsair is a well respected PSU, and my gut just says try making all the PSU connections again, remove and re-install the ram, double check your cooler and all mobo connections even boot into bios and make sure all is properly configured. Physical connections are always the first things to illuminate then move on !

Could you have done some CPU damage, it's possible but not likely unless a pin is bent or missing or that paste got into the socket.

Also updating your spec's and even pictures help in understanding what you're up against.

Rik89 · 30 May 2014

I really appreciate everyone's thoughts.

Having taken linnemeyerhere advice re testing out all the connections I noticed that one of the new heatsink's prongs wasn't fully fastened in. It still isn't fastened as such as I think I might have broken something with this one too, but it is now firmly in place at least, which has reduced the amount of noticeable leeway when any slight pressure is put on it. The CPU is now running a couple of degrees lower, so that's promising.

Hopefully the problem now goes away (happened once in the past 24 hours)!

I forgot to mention but the Event Viewer calls in an event 41 Kernal-Power, but I assume that's just a generic 'system has rebooted without cleanly shutting down first' event.

Akasa AK-CC4007EP01 Nero 3 CPU Cooler is my new heatsink btw. Thanks again! Hopefully I won't need to bump this thread again...

linnemeyerhere · 30 May 2014

Keep us in the loop and let's hope that helps.

Rik89 · 31 May 2014

It just happened again... whilst I was installing an update on my iphone, which is now forcing me to go through a restore, losing some important data I won't be able to get back. Thought I'd solved it, damn. Really gutted

badger906 · 31 May 2014

just googling 41 kernal power throws up loads of results that all end with psu failure!!

id suggest ordering a replacement psu, testing it, if it persists rebox it and sell it as ex display or ex demo, a few hours use isnt going to hurt!