Sign in with
Sign up | Sign in
Your question
Closed

Experienced computer builder having a really strange problem.

Tags:
Last response: in Windows 7
Share
March 9, 2012 4:08:46 AM

Hi Guys

Setup
Windows 7 64 Bit
Intel I5 - 760 Quad Core @ 2.8 ghz
Hitachi 1TB 7200 rpm SATA HDD
WD 320gb KS HDD Sata 2 (OS is installed on this drive)
8 gigs Corsair 1600 MHZ DDR3 (PN CMX4Gx3M2A) (4x2gig)
Antec Kuhler H20 620 CPU cooler
Gigabyte GA-H55M-USB3 Mobo
Bitfenix Shinobi Case (no window)
Thermaltake 600W PSU
Gigabyte HD6970 Radeon 2gig
OCZ TXT Memory Cooler

[Nothing overclocked except video card fan speed]

So i come to you guys as a last result after spending large amounts of time and money trying to troubleshoot my own issue. Before i dive into details, im 95% sure that my issue is caused by overheating, i just dont know what and why.

As of roughly 3-4 months ago i began having issues with my computer randomly shutting off while playing high intensity games (BF3 on medium which caused my HD5670/CPU to run at ~90-100% capacity). It wouldnt blue screen it would just turn off. Upon shutting down it wouldnt boot for between 15 minutes to even 4-5 hours (where i would pull the whole thing apart in exasperation and upon putting it back together it would boot). So over the next few months i did the following in an attempt to fix/stop the issue (dot pointed so you guys dont TL:D R) [After all these changes the issue is still persisting, occasionally during low intensity work but mainly with high intensity or on hot days]

Hardware Changes
Monitored CPU/GPU temperature (while doing alot of googling)
Thought CPU temp was to high using stock cooler so replaced it with an Antec25 Kuhler H2O 620 (CPU maxes now at ~60 degrees)
Thought power supply might be faulty - Replaced with Thermaltake 600 watt
Thought case wasn't supplying enough air flow - Replaced with a Bitfenix Shinobi (currently running 4 Fans and isolating PSU airflow to its own loop through bottom mounting)
Replaced HD5670 with HD6970 Gigabyte Edition (yesterday)
Replaced MOBO battery

Stats- (monitored with Realtemp/PCZ)
Full Load GPU ~ 80 degrees, typical of high end GPUs.
Full Load CPU ~ 60-70 degrees, typical.
Dont have a ram thermoprobe but it has a mounted OCZ ram cooler on it.

To this day the only thing that seems to stop the random resets is placing a large fan pointing into the CPU case and even this doesnt work on typical 35 degree brisbane summer days (especially if im an idiot and try to play BF3)

When the PC does crash it generally wont boot again unless i wait at least 25+ minutes and then plug an old 420W PSU into the Mobo (only) boot it (light turn on) turn it off and reconnect the 600W after which it functions normally until the next crash.

I have run kombustor/prime95 with the fan running into it on hot days and without on cooler days without any issues. Currently am about to run a memtest96 but i doubt this is the issue.

Now the obvious cause is overheating im just unsure of what, is there a possible inbuilt cut off on my MOBO when the GPU/CPU passes a default temperature which is inherently lower then the temperature they can operate safely at?
Ive turned all the bios fan settings to maximum and clocked the cards fan up to 100% with no avail.

Sorry about the wall of text but any help you could give me would be a huge aid, ill check back on this post often to run any tests or provide any extra info you guys might need to help me out.

Many Thanks
Felice
March 9, 2012 4:13:37 AM

Your CPU is hotter than normal.... even use stock Fan the CPU Temp should below 45C.... I would think is bad contact or you apply too much thermal paste....

remove the CPU FAn again .... clean up all thermal paste... and re-apply on it .... see what's going on... if you don't overclock it...fully load should around 60C
Score
0
March 9, 2012 4:39:04 AM

i reapplied the paste as you said and will have to wait to test it as currently the PC is being stuborn and wont even boot, no lights nothing
Score
0
Related resources
March 9, 2012 4:47:09 AM

So what fan you're using..... after that also check wheather they have any new BIOS... As I remember couple of MB have some problem reading the true tempature...
Score
0
March 9, 2012 7:54:08 AM

60 degree under load isn't really that extreme. Infact none of your temps are particularly bad. How old is that HDD with the OS on it? You've replaced or tested everything else and your temps are far from meltdown scenario. I'd be looking at the HDD now

*Edit to add*

In your shoes now I would think that HDD is running hot. I would put the OS on the other HDD (completely disconnect the 320GB) then do something that normally causes the crash.
Score
0
a b $ Windows 7
March 9, 2012 8:06:09 AM

Yea, like wr said 60 degrees is not that bad.

I actually think something is going on with your motherboard.
When I overheat, my computer turns off, but it comes back up within seconds because cpu's cool off relatively quickly when they are off.

Either that or a psu issue, I personally don't think you are having a overheating problem.
Have you hooked up a real voltmeter to the psu to monitor real volts as it's running?
Score
0
March 9, 2012 8:12:05 AM

Could as well be the motherboard, since you replaced the PSU, that's my best bet. :) 
Score
0

Best solution

March 9, 2012 8:12:36 AM

He tried a new PSU though. I think he should shove the OS and BF3 on the other HDD. Totally disconnect the 320GB and see if it crashes that way.

If that still crashes it rules out the HDD running hot (the only temp he cant really monitor) and then only really leaves the mobo. I've never had a Gbyte mobo go bad if it is that hopefully he is in the warranty period (3 year for MANUFACTURE date)
Share
March 9, 2012 8:19:50 AM

Both Speedfan and HDTune show HDD Temps.

Since all temps are quite ok and since I doubt the airflow in his case manages to heat up the HDDs, but nothing else, it must be some kind of hardware defect.

A shutdown due to a faulty HDD would still boot up the PC ... smells like a motherboard issue.

However, it could still be overheating of the motherboard itself (the passive cooled parts, like northbridge, etc). Speedfan also shows those temps, if your motherboard supplies them (it should, even my 5year old noname crapboard does).
Score
0
March 9, 2012 8:25:35 AM

Gigabyte GA-H55M-USB3 Mobo - what bios version do you have. latest is F11
i take it the PC works fine just at windows desktop?
can you disable intel turbo boost feature in the bios?
Score
0
March 9, 2012 8:25:37 AM

I've always found those HDD reading less than meaningful. Main reson I think he should sniff at the HDD and try the other is he says he can stress test the setup on hot or cold days with no issue. Worth trying i think before RMA'ing the mobo and ending up with a pile of components for however long it takes for the new board to appear.
Score
0
March 9, 2012 8:30:31 AM

wr6133 said:
I've always found those HDD reading less than meaningful. Main reson I think he should sniff at the HDD and try the other is he says he can stress test the setup on hot or cold days with no issue. Worth trying i think before RMA'ing the mobo and ending up with a pile of components for however long it takes for the new board to appear.


Worth trying it is. However, I still don't see why it wouldn't boot up after shutdown if it was just for a faulty HDD. :) 
Score
0
a b $ Windows 7
March 9, 2012 8:33:27 AM

In my experience, when hdd's/ssd's get hot/go bad, blue screens are the name of the game. Unless it's something physically shorting out inside the drive, in which case the psu would shutdown.... but in that case the, hdd doesn't come back either....
Score
0
March 9, 2012 8:38:05 AM

I agree with you but its an odd one if its the HDD it should still post after it canes itself but if its the mobo the stress test should probably cause the crash. I'd try the HDD first just as its less hassle (though admitedlly still a ballache) then if it that shows no joy RMA that board.

If both of those dont resolve it then we both eat our hats lol
Score
0
March 9, 2012 8:43:35 AM

Or a power issue from the wall socket? Sounds silly but maybe the OP needs to try another socket or does he have an insane number of plugs daisy chained on strips from 1 socket? I know when i was in Iraq we lived off generators and the current wasnt stable it would constantly mess up our comms kit.
Score
0
a b $ Windows 7
March 9, 2012 9:07:31 AM

How hot does the PSU get? You say it has its own air; is its fan running? How hot is the air coming out of it?
Score
0
March 9, 2012 10:26:45 AM

Hey guys, firstly thank you so much for the quick intelligent responses, im unused to this in forums and have obviously been hanging around in the wrong places. Secondly sorry for the late response i just finished work haha.

In response to your guys ideas, i can rule out the PSU as its done the same thing across 3 different supplies.

ATM im thinking more likely it's the mobo, as i said in my OP blowing a fan onto the GPU (and really the mobo) seemed to remove the issue (except for really hot days), furthermore ive had experience with other overheating computers and they all turn on within 5 minutes, mine seems to not even respond to power from anywhere between 5 minutes to over 8 hrs (ive left it all night while sleeping and it still didnt boot, had to do the little PSU shenanigan i said before), and today even that PSU swapping didnt seem to make a difference.

Now im quiete happy (metaphorically its a pain but ill do it for thoroughness) to redo the OS, quick (kinda nooby) question, any way i can keep all my rainmeter customisation etc across to the new OS apart from just backing up the entire C drive? The drive in question (340gig) is something like 5 years old so it wouldnt suprise me either if this was the problem.

The PSU is brand new and the air exiting it doesnt seem to be anything warm, i have noticed since i got the new GPU the air exiting the case is a fair bit warmer (i expected this) but since it was happening before hand i dont see how this would matter that much.

As for power sockets, i have done some small amount of testing, hardly thorough, running the PC through a different wall socket and through different slots on my 8 slot surge protector and still encountering the error.

So while i wait for any last responses before i format the drive and transfer the OS ill swap to a diff wall plug and run a prime95 test to give you guys some up to date numbers, if that runs stable ill run a kombustor as well :) 

Many many many thanks
Felice
Score
0
March 9, 2012 10:55:52 AM

davidgermain said:
Gigabyte GA-H55M-USB3 Mobo - what bios version do you have. latest is F11
i take it the PC works fine just at windows desktop?
can you disable intel turbo boost feature in the bios?


Sorry i forgot to reply to this, bios is F11 :)  checked all my driver updates as the very first thing
Score
0
a b $ Windows 7
March 9, 2012 7:54:34 PM

Hmmm, your mobo appears to have all solid caps on it, so it isn't likely bad caps.
Score
0
March 9, 2012 8:50:14 PM

It usually does work fine on desktop though i have had it happen once or twice with low load on the CPU/GPU, (FB and maybe dwarf fortress)

I just finished the OS install on a parition of my newer HDD, ill run prime95/kombustor while im at work while recording temps with realtemp and report back when i get home :)  (around 6:30 EST)

Thanks guys
Score
0
March 10, 2012 7:00:29 AM

Hey guys, so i installed the new OS as i said before and left a prime 95 test running while i was at work.
It ran from 9:15 am until 1:28 pm, whereby it crashed and at this time won't boot.
CPU temps on the 4 cores were stable with readings of 75/76/72/74 (+- 2 degrees) for the cores respectively at 100% load. GPU temperature was a stable 53 degrees. This is the part that has me stumped. There is no overheating reason whatso ever for the computer to have crashed, it ran for over 4 hours, through the hottest part of the day, with no issue before crashing randomly. At this time it still won't boot. Having ruled out a dodgy harddrive, the only real thing left to do is RMA the mobo, any more ideas before i send it off?

Many Thanks
Felice
Score
0
March 10, 2012 8:17:33 AM

Guess I best eat my hat. I think via process of elimination the only thing left is to RMA the board (if you bought from a local shop it may be covered by them and faster).
Score
0
March 10, 2012 9:45:03 AM

wr6133 said:
Guess I best eat my hat. I think via process of elimination the only thing left is to RMA the board (if you bought from a local shop it may be covered by them and faster).


Since it wasnt booting, weirdly enough when i removed a ram stick it booted up, thinking it might be a ram issue i missed i shut down to test them 1 by 1 and now it won't boot haha. This has me stumped so im thinking its time to RMA the board and go from there, thanks everyone for your input :) 
Score
0
March 10, 2012 9:46:27 AM

Best answer selected by drueller.
Score
0
a b $ Windows 7
March 10, 2012 10:31:25 AM

This topic has been closed by Nikorr
Score
0
!