View Full Version : Test suggestions

05-12-2012, 10:47 PM
I am getting a 22 mth old Computer Lounge built I3 PC delivered to me for 5 hours next Sat morning so that I can do some tests on it.

Specs are ASUS P7H55-M/USB3 P55 mobo, I3 540 Clarkdale CPU, Kingston HyperX 2x2GB DDR3-1600, Seagate 1TB HDD, Corsair VX450 PSU, Win 7 64bit.

I have posted about this one before. It just hard crashes every month or two with differing errors but it has done it a couple of days apart. So, with this erratic behaviour there is no way I can have much idea whether I have done anything to fix it.

But I want to give the hardware a thrash to see if anything shows up. I will re-seat all sockets, check for dust etc then run memtest and OCCT (1 hr stress test that provides temps and voltages). I will also run chkdsk and check event viewer and reliability charts etc. I have previously checked the mem dumps with Bluescreen view but will check the last couple I haven't seen yet.

I would appreciate other suggestions or tests to try as this is a bit of a nightmare!


06-12-2012, 01:44 AM
Incase its a hardware problem just swap out different things until it works.

06-12-2012, 07:36 AM
Could try something like PassMark's Burn in Test, and FutureMark's PCMark and 3DMark.

06-12-2012, 08:54 AM
Errr, Slanky, didn't you read where it fails about once every two months or so and I have it for only 5 hours? (It lives 45 min away from my place). But, thanks anyway.

Thanks for the alternate ideas, inphinity. The burn test sounds like a must to run. I'll get these test progs downloaded and ready to go on Saturday.

I'm not greatly optimistic at getting a definitive answer but will keep fingers crossed that something tangible comes up.

Speedy Gonzales
06-12-2012, 10:07 AM
If its crashing install something like bluescreenview. And see whats causing it. And tell us, what AV / security suite programs are on it.

06-12-2012, 10:51 AM
How often is that PC restarted?? Is it left allways on & never restarted ??

Unfortunately, most good test suites cant be run in an endless loop unless you buy the full version
So you have to keep running BIT , 3dMark etc etc manually.
If you find a test suite that will run continuously/looping on Vista/7 & is free, please let us know.

I'd suggest Memtest86+ , that can be looped endlessly, but only tests RAM & faulty RAM can still pass those tests
Perhaps something like HotCPU or Prime to heat/stress test the CPU ??

If its a desktop case, sometime the top cooling vents get covered with the users documents & crap etc
Some cases are very flimsy & the vid card can get partially unseated (its a long shot admittedly)
Good luck :-)

11-12-2012, 03:54 PM
Well, guys, I was pleased to find I could reliably cause it to bsod with a stop 101 (that doesn't refer to a crash course at uni!!). I ran memtest and the RAM looked good. Then I ran OCCT. That crashed the machine after a couple of minutes.

OCCT showed the cpu clocking at 3.6GHz so I reduced the cpu mult from 160 to the std 133 (3GHz). OCCT ran with no problem for over an hour with cpu temps no greater than 61 deg. Put mult back to 160, crashed again. Back to 133, ran fine.

chkdsk was fine. RAM fine. So with everything looking OK, it went home to Levin. Fingers crossed that the occasional crash is now a thing of the past. Only time will tell.

For those interested, the stop 101 is where a cpu watchdog times out. This causes an immediate shutdown, leaving little info to go by. You have to troubleshoot a kernel dump it seems. This problem is said to be fixable by upping vcore a little if you want to maintain the overclock.

Thanks for the input, all.

Speedy Gonzales
11-12-2012, 05:13 PM
Did anyone install any power management software?? If you did that can cause that 101 stop error. These can also cause it

Nvidia Performance Control panel and system monitoring software & ADIA 64. And if youre overrclocking programs like these will crash it. It looks like if the CPU is multi-core you can also get this stop error. Some people have fixed this crash by changing the CPU's in msconfg. And if the latest BIOS isnt on it, I would update the BIOS

Speedy Gonzales
11-12-2012, 05:33 PM
It can also be caused if Shader 3 is enabled and youre using a Nvidia card. If you are, disable Shader 3, if you cant select Shader 2

12-12-2012, 08:39 AM
I'd side with the suggestion that a rare crash from a Windows machine is not anything to be too upset about. More just a signature that you're running Windows!

Also, with such an intermittent fault, the owners are being way stoopid expecting you to be able to diagnose and test in 5 hours for a fault that may not show up in days of use.

12-12-2012, 10:04 AM
Thanks, Speedy. It is running integrated video and no power management software has been added. It is very bog standard.

Must disagree, Paul, that you should expect Windows to crash repeatedly. None of my machines, and many others I look after, do that. I was not hopeful I could fix a rare crash, either, but was pleased I could induce a crash at will and then change things to make it stable.

As I said, I will just have to wait and see!

12-12-2012, 11:02 AM
Did you have to underclock it? Could be a cooling issue, heatsink full of dust, degraded thermal pad etc.