does heat effect PSU stability , is there way to stresstest?

Got a shopping cart of parts that you want opinions on? Get advice from members on your planned or existing system (or upgrade).

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
soulie
Posts: 28
Joined: Mon Jul 17, 2006 12:19 pm
Location: BE

does heat effect PSU stability , is there way to stresstest?

Post by soulie » Sun Jun 27, 2010 7:52 am

I bought a new pc like a couple of months ago. On my previous build i might get a bluescreen once every 6months or so playing WOW. On my new build the frequency is quiet a lot higher , and i having a hardtime to actually specify what is causing the errors.

Spec :

MSI Hawk 5770
Gigabyte P55A-UD3
Intel 750 stock and stock cooled
4GB geil black dragon
530W be quiet

Temperatures wise i kept a eye on it and kept it logging temperatures the CPU 750 was hovering most of the time around 35°C and on high peeks 45°C
GPU temp 40-50°C and peek 58-60 °C

In the end i noticed , that on warm days like 30°C outside it would happen easier or when the drain on the system was higher.And now durign the summer months it isn't a bluescreen but [b]just a auto restart[/b] if it crashes. So was wondering if the PSU maby wasn't stable enough for the GPU . Is there actually a way to test PSU reliability like stresstest your pc.

And can a lack of power to GPU result in a bluescreen?

BlackWhizz
Posts: 266
Joined: Wed Mar 31, 2010 8:19 am
Location: OV, The Netherlands

Post by BlackWhizz » Sun Jun 27, 2010 9:59 am

it could be your drivers too. First try that.

line
*Lifetime Patron*
Posts: 338
Joined: Wed Sep 28, 2005 10:07 pm
Location: Israel

Post by line » Sun Jun 27, 2010 1:44 pm

Reset the BIOS to defaults. Run memtest86++ overnight to see if the system is stable. (If you get an error, you should definitely investigate it.) Memtest doesn't stress the GPU but it's a good test for stability and it runs outside the OS so if you see an error you know that it is not caused by Windows or a driver.

Also look in Control Panel -> Administrative Tools -> Event Viewer -> System. See if you can find info about system errors and let us know.

soulie
Posts: 28
Joined: Mon Jul 17, 2006 12:19 pm
Location: BE

Post by soulie » Mon Jun 28, 2010 3:23 am

Well drivers i always updated to the latest ATi drivers(ATI-site) , P55A chipset drivers from gigabytes site , and updated bios everytime after 3th new bios i also just stucked with default bios settings.

Memory test i ran a couple of times always with no errors. But yeah i can give it a try and run some memtest true the night.


UPDATE : http://95.211.18.179/log.png
screenshot of some bluescreencaptures and event log
most bluescreen were dxgmms1.sys , but also have had file mentioning in blue screen like : fltmgr.sys , ntoskrnl.exe , Npfs.SYS , usbhub.sys , usbuhci.sys , USBD.SYS

Two most popular files mentioned during blue screen are by far dxgmms1.sys and ntoskrnl.exe



PS to hot 35C , 90ish F for you american , laaaaaaaaaaaaame laame lame lame. I stop fucntioning properly at 27C. Think goign to lay in the shade and puf

b_rubenstein
Posts: 273
Joined: Tue Aug 04, 2009 7:03 am
Location: Brooklyn, NY

Post by b_rubenstein » Mon Jun 28, 2010 4:18 am

OCCT can be used to stress, the CPU, GPU or both together to stress the PS. If your computer can get through an hour of Linpack, the the problem is either with the video card or the drivers.

soulie
Posts: 28
Joined: Mon Jul 17, 2006 12:19 pm
Location: BE

Post by soulie » Fri Jul 02, 2010 7:56 am

Well i downloaded OCCT , not sure what i had to run but i ran CPU:OCCT test , and well first time i ran after 1 minute it stopped saying error on core 1 , and second time it ran 2 minutes and gave error on core 1. But no clue what to do now.

Anyway nothign is overclocked tbh.

b_rubenstein
Posts: 273
Joined: Tue Aug 04, 2009 7:03 am
Location: Brooklyn, NY

Post by b_rubenstein » Fri Jul 02, 2010 3:30 pm

If the CPU test crashes after 1 or 2 minutes you have a problem and it's prpbably hardware. The CPU test isn't very severe; Linpack is the heavy hitter.

Try running only 1 stick of memory at a time and see if one of the sticks is stable. Check in the BIOS to see if the memory voltage is set to the right value for your memory. If there is no difference running only 1 stick of RAM, then you could try using a different PS. It's more likely a problem with the MB, but it's easier to hook up a different PS.

BlackWhizz
Posts: 266
Joined: Wed Mar 31, 2010 8:19 am
Location: OV, The Netherlands

Post by BlackWhizz » Sat Jul 03, 2010 1:10 am

Most probably its your mobo or ram. There are also some usb driver errors. Can u disable the USB3 controller? And then run OCCT?

soulie
Posts: 28
Joined: Mon Jul 17, 2006 12:19 pm
Location: BE

Post by soulie » Sat Jul 03, 2010 3:16 am

[quote="BlackWhizz"]Most probably its your mobo or ram. There are also some usb driver errors. Can u disable the USB3 controller? And then run OCCT?[/quote]

I did that once before , but dint do much. Anyway i gave it a other try
disabling stuff i don't treally need :

-onboard serial port
-onboard parallell port
-USB3 disabled
-SATA3 disabled

Now i also disabled
-Intel turbo boost
-Swith memory from dual mode to single.

It is running for 10minutes at the moment that is at least already 8 minutes longer then previous and still running.

soulie
Posts: 28
Joined: Mon Jul 17, 2006 12:19 pm
Location: BE

Post by soulie » Sat Jul 03, 2010 3:23 am

sigh , i stop test open firefox surf a bit around , bam blue screen

ntoskrnl.exe

:cry:

b_rubenstein
Posts: 273
Joined: Tue Aug 04, 2009 7:03 am
Location: Brooklyn, NY

Post by b_rubenstein » Sat Jul 03, 2010 8:22 am

soulie wrote: -Swith memory from dual mode to single.
Did you do this by taking out a stick of RAM? Did you run the PC trying both RAM sticks, one at a time?

I wouldn't pay much attention to what file was being used when the PC crashed. That fact that several different ones are being used at the time of the crash suggests that the crash has nothing to do with any specific file and is related to some other root cause.

tim851
Posts: 543
Joined: Wed Aug 13, 2008 11:45 am
Location: 128.0.0.1

Post by tim851 » Sat Jul 03, 2010 11:54 am

My money is on your Northbridge getting too hot. This will usually present with blue screens. An overheating PSU will result voltage drops which should always present with sudden reboots.

Your unmodded (?) Be Quiet PSU should not overheat on your system. And 530W is about twice as much as you'd actually need. So I really doubt it's the PSU.

soulie
Posts: 28
Joined: Mon Jul 17, 2006 12:19 pm
Location: BE

Post by soulie » Sun Jul 04, 2010 4:13 am

yes , nothing is modded and all are stuck i neatly in my antec solo.

-I installed again World of warcraft but for this I pulled out one ram module because the installer kept crashing in random places, I dint believe people that had the same problem and they pulled out one ram module to solve it. And yes that worked strangly enough :roll:

-For single mode i just plugged in the second ram module again but not in a slot that would result in dual mode. Did not configure true bios here.


PS ;p this topic isn't really about PSU anymore so sorry for that

soulie
Posts: 28
Joined: Mon Jul 17, 2006 12:19 pm
Location: BE

Post by soulie » Sun Jul 04, 2010 8:42 am

[quote="tim851"] My money is on your Northbridge getting too hot. This will usually present with blue screens. An overheating PSU will result voltage drops which should always present with sudden reboots.

Your unmodded (?) Be Quiet PSU should not overheat on your system. And 530W is about twice as much as you'd actually need. So I really doubt it's the PSU. [/quote]

Well if you mean the P55 chip since northbridge is now the cpu itself on corei5 (memory controller ,...) , after a full day of playing games etc on a pretty warm summer day (with 2 bluescreens , yippie!) the thing is not hot on thouch really i have to keep a finger on one of the tiny hole for a while to really feel any high heat.


Picture of my pc and poistion of P55 chip pretty close to gfx , but like mentioned before not really runing hot
http://95.211.18.179/04072010009.jpg

PS and why doesnt quote work ;o

BlackWhizz
Posts: 266
Joined: Wed Mar 31, 2010 8:19 am
Location: OV, The Netherlands

Post by BlackWhizz » Sun Jul 04, 2010 10:12 am

U already did a memtest?

tim851
Posts: 543
Joined: Wed Aug 13, 2008 11:45 am
Location: 128.0.0.1

Post by tim851 » Sun Jul 04, 2010 12:05 pm

Ah, you have a Core i5, I carefully missed that info...

Well, then it does seem to be most likely that it's a memory issue. Run memtest as has been recommend by others.

soulie
Posts: 28
Joined: Mon Jul 17, 2006 12:19 pm
Location: BE

Post by soulie » Sun Jul 04, 2010 1:26 pm

yeah a couple of times but only the one from windows , that starts after your reboot and never a problem.

BlackWhizz
Posts: 266
Joined: Wed Mar 31, 2010 8:19 am
Location: OV, The Netherlands

Post by BlackWhizz » Mon Jul 05, 2010 7:00 am

Run a memtest from (for example) a Ubuntu liveCD. You need the real memtest.

soulie
Posts: 28
Joined: Mon Jul 17, 2006 12:19 pm
Location: BE

Post by soulie » Mon Jul 05, 2010 8:08 am

ran default tests with http://www.memtest.org/ LAST VERSION : 4.10 , made usb bootdisk

It ran true 3 cycles while i was swimmign for 2h30min , i let it run true thisnight i guess. During those 3 loops no errors detected

BlackWhizz
Posts: 266
Joined: Wed Mar 31, 2010 8:19 am
Location: OV, The Netherlands

Post by BlackWhizz » Mon Jul 05, 2010 9:06 am

Hmm.. After 3 loops your mem should be good. Can u try another OS (Linux LiveCD or something like that). Runs your PC stable with that os?

soulie
Posts: 28
Joined: Mon Jul 17, 2006 12:19 pm
Location: BE

Post by soulie » Thu Jul 08, 2010 11:11 am

Found a memory setting in bios that stands default on enhanchement or something and with big red letters "if you overlock put in on standaard to improve reliability!" Anyway put that on standard , no blue screen atm :roll: If it really was that settings very strange defualt settings

BlackWhizz
Posts: 266
Joined: Wed Mar 31, 2010 8:19 am
Location: OV, The Netherlands

Post by BlackWhizz » Thu Jul 08, 2010 11:56 am

soulie wrote:Found a memory setting in bios that stands default on enhanchement or something and with big red letters "if you overlock put in on standaard to improve reliability!" Anyway put that on standard , no blue screen atm :roll: If it really was that settings very strange defualt settings
So you probably found the issue? I think its a little incompatibility between the mobo and memory. Mobo cant read the XMP profile from the memory (XMP contains the settings of the ram like voltage, speeds cas latencies etc). So your mobo used the default settings (because it couldnt read the XMP) setting (what the mobo manufacturer set up in the bios) which was probably a little too less for your memory.

But im glad u found it!

b_rubenstein
Posts: 273
Joined: Tue Aug 04, 2009 7:03 am
Location: Brooklyn, NY

Post by b_rubenstein » Thu Jul 08, 2010 2:44 pm

On a Gigabyte MB that setting has nothing to do with the RAM timing settings (those are found in the Memory settings). I've yet to run into a situation where if the memory and CPU were run at rated speed the Enhanced setting didn't work fine. However, the timing on the all the signals is very critical, and one funky trace on the MB, RAM stick, RAM chip, memory controller, etc., and random errors can be introduced.

soulie
Posts: 28
Joined: Mon Jul 17, 2006 12:19 pm
Location: BE

Post by soulie » Fri Jul 09, 2010 3:09 pm

[quote="b_rubenstein"]On a Gigabyte MB that setting has nothing to do with the RAM timing settings (those are found in the Memory settings). I've yet to run into a situation where if the memory and CPU were run at rated speed the Enhanced setting didn't work fine. However, the timing on the all the signals is very critical, and one funky trace on the MB, RAM stick, RAM chip, memory controller, etc., and random errors can be introduced.[/quote]


Well the setting was M.I.T -> advandce memory settings -> then Performance Enhance , so far no blue screen.

Just find it strange that gigabyte would put a setting on default which they give a warning for themself if anything is overlocked leads to system instability. Which i dint do , if there is anything overclocked it is by manufactures themself. It is not really a board for overclokers/gamers so why not put it on standard as default. Well I hope this what it. I always had stable systems so was extra annoyed why my latest build wasn't. Anyway crossing my fingers , it has run smooth some days after a row before in the beginning :roll:

PS I wished i sticked to my favourite XMS-corsair dominator , but hey the geil one were lower voltage , cooler and cheaper ;p

soulie
Posts: 28
Joined: Mon Jul 17, 2006 12:19 pm
Location: BE

Post by soulie » Tue Jul 13, 2010 4:40 am

again bluescreen but hey at least it lasted 7 days the system , petty ram prices are so high atm else i would just replace them.

Post Reply