Major random freezing problems with new Core i3 build

[spikey_richie]

Try WhoCrashed from here http://www.resplendence.com/whocrashed

When you reboot from a crash, load that and it should help you diagnose.

[inferniac]

  On 30/04/2011 at 18:30, Fourjays said:

Just stumbled across what you were probably looking for: http://forum-en.msi.com/index.php?topic=104805.0;wap2

Indicates that the HD5670 requires 26A. I'll admit I'm not too good when it comes to understanding the finer points of PSUs in computers so I have no idea how this would effect the computer. Wouldn't I get random reboots with insufficient power?

I had an insufficient PSU in my PC, the results were, when playing graphic intensive games the PC would just switch off.

[Fourjays Veteran]

  On 30/04/2011 at 22:00, Seizure1990 said:

You said the problem seems to happen most on cold boots. Have any of the memory tests you've done so far been cold boots?

Not really. I've tried leaving it off for an hour or two to see if that made a difference but it hasn't. But both yesterday morning and this morning it did it right away. Will leave the memtest disc in tonight and see if it turns anything up on boot tomorrow.

Would memtest pick up on issues caused by loose RAM sticks? The DIMM slots on this board only have a clip on one end (see pics here) and it isn't exactly a tight fit (not as tight as when there are two clips anyway). Just wondering if it is possible that moving the case around, opening it, changing bits around, etc keeps "shifting" the sticks around in their slots (and therefore varying the severity of the freezing)?

  On 30/04/2011 at 22:05, spikey_richie said:

Try WhoCrashed from here http://www.resplendence.com/whocrashed

When you reboot from a crash, load that and it should help you diagnose.

Will give it a go, in case I've got/had two separate issues causing the freezing. Not showing anything at the moment as there are no crash dumps.

[Seizure1990]

From what I've seen anyways, insufficient or damaged PSUs just end up with the machine shutting off, not hanging.

Have you tries using a different video card or the built in yet like you mentioned?

Also, thought I'd mention that one time, I was working on a computer that had a similar problem. After a while of testing, it turned out that one of the PCI slots was bad, and if a device was plugged into it (any device), it would cause the computer to hang early on in the boot up process.

[V9s]

  On 30/04/2011 at 23:20, Fourjays said:

Not really. I've tried leaving it off for an hour or two to see if that made a difference but it hasn't. But both yesterday morning and this morning it did it right away. Will leave the memtest disc in tonight and see if it turns anything up on boot tomorrow.

I had a similar problem with memory incompatibility when building my i3 system with an intel motherboard, ran memtest a thousand times - no errors......left the computer off a few hours and started it back up again - BSOD/freezing.

So i left my pc off a few hours (5-6 hours) and then booted it up to memtest and. lo and behold, hundreds of errors came up.

Try doing a memtest on a coldboot and check your ram. I had to get the ram replaced to a compatible brand listed in the motherboard manual.

[zeta_immersion]

if you have a voltmeter try to see each plug what gives and if indeed gives the right voltage/amperage

also, try to load a linux cd, without gui and see if that crashes your pc (should be light enough as not to draw a lot of power for any component including the HDD)

also, if you can, and have another PC handy, separate parts or the PSU from one to power parts of the computer ie, let one PSU power the mobo and the other PSU power the components (that way you split the power draw and can at least give it enough juice to do whatever (short of getting and buying a new PSU)

[Fourjays Veteran]

Just ran 3 passes of memtest on the first boot this morning and it had no errors. Booted to Windows, ran BurnInTest again to push the all the components to their max, and then plugged in a PS3 controller for charging (to push the PSU). Nothing. Closed the test, clicked shut down and it froze again (HDD light on as usual). :/

Will leave it off for a while and then try running it without the graphics card. Despite there being no errors my instinct and everything I read says it is probably the memory so if I don't get any further I'll probably see if I can get some other memory to try.

  On 01/05/2011 at 11:21, zeta_immersion said:

if you have a voltmeter try to see each plug what gives and if indeed gives the right voltage/amperage

also, try to load a linux cd, without gui and see if that crashes your pc (should be light enough as not to draw a lot of power for any component including the HDD)

also, if you can, and have another PC handy, separate parts or the PSU from one to power parts of the computer ie, let one PSU power the mobo and the other PSU power the components (that way you split the power draw and can at least give it enough juice to do whatever (short of getting and buying a new PSU)

Not got a voltmeter, but I've got an old 400W PSU. Doesn't have the 8-pin ATX12V (is from a P4 system) but I could plug the drives into it.

[zeta_immersion]

  On 01/05/2011 at 11:26, Fourjays said:

Just ran 3 passes of memtest on the first boot this morning and it had no errors. Booted to Windows, ran BurnInTest again to push the all the components to their max, and then plugged in a PS3 controller for charging (to push the PSU). Nothing. Closed the test, clicked shut down and it froze again (HDD light on as usual). :/

Will leave it off for a while and then try running it without the graphics card. Despite there being no errors my instinct and everything I read says it is probably the memory so if I don't get any further I'll probably see if I can get some other memory to try.

Not got a voltmeter, but I've got an old 400W PSU. Doesn't have the 8-pin ATX12V (is from a P4 system) but I could plug the drives into it.

yes, try that, leave the PSU in the comp powering the mobo and CPU only, and use the other 400W PSU powering the components

[ZakO]

  On 01/05/2011 at 11:26, Fourjays said:

Just ran 3 passes of memtest on the first boot this morning and it had no errors. Booted to Windows, ran BurnInTest again to push the all the components to their max, and then plugged in a PS3 controller for charging (to push the PSU). Nothing. Closed the test, clicked shut down and it froze again (HDD light on as usual). :/

Will leave it off for a while and then try running it without the graphics card. Despite there being no errors my instinct and everything I read says it is probably the memory so if I don't get any further I'll probably see if I can get some other memory to try.

Not got a voltmeter, but I've got an old 400W PSU. Doesn't have the 8-pin ATX12V (is from a P4 system) but I could plug the drives into it.

I've had bad memory which hasn't shown up errors in memtest until 12-24 hours of testing, it would probably be worth running memtest overnight (if you don't have any other memory to try).

[Fourjays Veteran]

Well I was just checking the internal workings again, making sure all plugs are connected, etc and noticed one of the 4 heatsink pins/clips (that hold it down) was not in as far as the others. It was a very difficult one to get in when I built it but I eventually succeeded as far as I could tell. I've just taken the motherboard out of the case to check whether the pins had spread on the other side and while the other 3 were in tight this one pin wasn't. I just tried pushing it in harder but it wouldn't go.

So I've now taken it off and checked that the pins are working. Something wasn't quite right with that one (took a lot of fiddling for it to "lock"), however I have now noticed that the heatsink compound (the Intel stuff on the stock cooler) isn't spread across the whole surface of the processor (as it did for older systems when I've taken off stock coolers) and is just a splodge in the middle. Despite this the BIOS reported temperatures of 37- 40'C idle and no higher than 55'C under load (via an ASUS tool) which seems ok. Is this something to worry about or should I trust the temperatures reported?

[zeta_immersion]

should be ok, but you could remove clean and reapply pate (don't overdo it otherwise it will increase the temp or even crack the cpu heatshield)

also, locking should be quick and without any problems, the cpu goes in the mobo by itself and in only 1 way, no other way will work ... and once you remove the block you can set it back although it is recommended you clean it and reapply paste as now it has air holes in it which is detrimental to heat transfer

[Fourjays Veteran]

  On 01/05/2011 at 12:12, zeta_immersion said:

should be ok, but you could remove clean and reapply pate (don't overdo it otherwise it will increase the temp or even crack the cpu heatshield)

also, locking should be quick and without any problems, the cpu goes in the mobo by itself and in only 1 way, no other way will work ... and once you remove the block you can set it back although it is recommended you clean it and reapply paste as now it has air holes in it which is detrimental to heat transfer

Thanks for the quick reply.

I've re-attached the heatsink now and it was much easier. First time around one of them was quite difficult. Will see if this makes any difference. I've got some old, cheap thermal paste but I'll probably order something better (or maybe even a different cooler) once I figure out if there is anything to replace under warranty.

[zeta_immersion]

so, have you solved the issue yet?

still crashing?!

[Fourjays Veteran]

  On 01/05/2011 at 12:34, zeta_immersion said:

so, have you solved the issue yet?

still crashing?!

Nope. :( Ran it for 30 mins to get the heatsink/compound settled a bit then left it off for around 20mins. Booted it up just now and it's frozen on "Starting Windows".

Will see if I can reliably reproduce this (seems to be doing it more today) then try running it on the Intel HD graphics and/or two power supplies to diagnose that aspect of it.

[Fourjays Veteran]

So far no freezes with it running on the built-in graphics (although sometimes it takes a while). Will have to see how we go, but if this goes ok then it would either be the video card, lack of power, or the PCI-E slot, right?

I can test the video card with a known working motherboard and a known working card. Power could be tested by doing using another power supply to supplement the current one.

[Seizure1990]

  On 01/05/2011 at 15:20, Fourjays said:

So far no freezes with it running on the built-in graphics (although sometimes it takes a while). Will have to see how we go, but if this goes ok then it would either be the video card, lack of power, or the PCI-E slot, right?

I can test the video card with a known working motherboard and a known working card. Power could be tested by doing using another power supply to supplement the current one.

Don't think it's the PSU, but you seem to be on the right track. Hopefully, it's the video card, and you can wrap this up soon.

[Fourjays Veteran]

  On 01/05/2011 at 16:44, Seizure1990 said:

Don't think it's the PSU, but you seem to be on the right track. Hopefully, it's the video card, and you can wrap this up soon.

Well I think I just narrowed it down to the memory. It froze running on the built-in graphics, which rules out the graphics card and should also rule out it being a power issue? At the time it froze it was doing 4 downloads, which goes with all the other freezes (during data transfer, which to my knowledge is heavy on the memory).

What do you guys think?

[Seizure1990]

Only one way to find out for sure. Got a friend with compatible RAM?

[Fourjays Veteran]

  On 01/05/2011 at 18:19, Seizure1990 said:

Only one way to find out for sure. Got a friend with compatible RAM?

Nah, all my friend's computers have DDR2, and this is my first build that uses DDR3. :/ Think the only way to find out is to buy some new RAM and test it. Then return the current RAM if it is definitely faulty.

Just froze again sometime in the last hour while I left it idle.

[RiverCampher]

It does appear to be a memory issue. You have a similar problem to http://forum.corsair.com/v2/showthread.php?t=89751 Take one module out and try them each in turn to reproduce the problem to confirm which module is faulty.

[Fourjays Veteran]

I think I've got it narrowed down to a specific module now. Ran with one in last night and it instant-froze. Gave it another go and it froze several times while idling. Now got the other in and no issues yet. *fingers crossed* :p

[Fourjays Veteran]

Dangit, just happened again. :( Took a lot longer for it to happen though. Are both sticks bad, or does this now point to something else?

[Seizure1990]

Is it possible to buy some new RAM and return it if you don't need it?

[Fourjays Veteran]

  On 02/05/2011 at 14:47, Seizure1990 said:

Is it possible to buy some new RAM and return it if you don't need it?

Not sure. Contacted the retailer at the moment to see how/if they can help because at the current rate I could spend the next few weeks buying and returning bits until I eventually hit on the right one. :/

I find it odd how it seems to get better or worst depending on which bits I've got in the machine. Feels like there is more than one thing wrong, although I know that is incredibly unlikely (read that fault rates are only 1%).

[Seizure1990]

Yea, seems like you've got a real puzzle on your hands. As irritating as it may be, swapping parts may be the quickest way to resolve this.

Good luck, let us know how it goes!