Solved freezing( locks up) every few hours, Event ID 41

[gagnon0vincent]

I will ignore that error then. I did what Koolkat said about removing and putting back some components and so far my computer has not crash for 3 days. I will continue to do some testing. Thanks for your help everybody.

 

[gagnon0vincent]

It still freezes.

 

[koolkat77]

In safe mode?

 

[gagnon0vincent]

When I went in safe mode I saw a lot of event ID 7001 and 1014 in the Event Viewer. I am still in the process to see if the computer freezes in safe mode .

 

[Vir Gnarus]

When you tried to install Windows on the HDD and test stability with that, did you actually go ahead and disconnect the SDD? You'll need to make sure to do that because even with it being present it'll be enough to trigger instability, regardless if the OS is installed on it or not. Also, have you tried with the SSD only? Who knows, it could be the HDD that's bad here.

As for Prime95, how did you run it, and for how long? Prime95 usually should be run for 7+ hours straight on Blend settings, and then an additional 7+ hour run on Large FFT settings. Depending on the stability of either test we can get a better idea if it's the CPU at fault itself, or something in the CPU-to-RAM crosstalk.

Also, when you say that the system starts freezing up once you start using it but remains stable when idle, what activity are you actually trying to do that triggers the freeze? Sometimes it's best to log what you were doing at the time of a system freeze as there may be patterns involved.

For additional hardware testing, I recommend running Hwinfo on Sensors only option and set it up to log onto a file and let it continue logging till it crashes. Do this twice and send both logs. We might be able to see if there's temperature or voltage instabilities taking place. Also, if you motherboard has onboard graphics, try removing the video card and test stability by running graphics from the onboard video.

It is evident we're dealing with hardware failure, it's just a matter of ascertaining what piece. Unfortunately sometimes this ends up resulting in a process of elimination and slapping potentially bad parts into a stable system to see if it bugs things up.

 

[gagnon0vincent]

Yes I did unplug the ssd before installing windows. I tried on the ssd only, on the hdd only and it would still crash. My pc does crash even when idle it just take a little longer. I did put my graphics card in another system. The second pc was fine but mine was still freezing on the integrated graphics . I am doing the tests you ask me I will report back when done. I just have one question. Some people say that gigabyte does not keep the latest driver on their website. So where can I get the latest?

 

[gagnon0vincent]

I think I fix my pc driver issue because I am capable of being on idle for days now but when I put prime 95 it freezes in about 20 min with temp of almost 100c. I will go get a h100 and try again.

 

[Vir Gnarus]

You don't even need to go that far with liquid cooling, but it's your call. Usually stock coolers are not sufficient enough high loads, and if you were using a 3rd-party cooler, either it was just a really really bad one or most likely you did not install it correctly, like proper application of the thermal paste.

Evidently what we're not dealing with is your graphics, so don't worry about that. Just work on the CPU first. Hopefully there has not been permanent damage to it, otherwise you'll just have to replace it.

 

[gagnon0vincent]

well I bought the H100i to replace the stock cooler. Now, i get around 50c at load but the computer still freezes with prime 95 after one hour on the windows side.I just did a test on my OS X drive with Prime95 and did not have any problem(i ran it for 9 hours).I will try reinstalling windows again. The attachment is just the temps of my cpu.

 

[Vir Gnarus]

Hmm, this is odd. There's gaps in the log where occasionally a sensor is reporting 0, with a number of them doing this. It may just be a software fluke, but I'm a bit leery about it. All other readings look pretty solid from what I've viewed so far.

Which drive is the OSX drive, the SSD or the HDD?

 

[gagnon0vincent]

OS X is on the ssd and windows on the 2 TB hdd.

 

[Vir Gnarus]

I'm not sure what to tell you, skip. There's no hard data we can pull from this, and hardware tests so far either come up clean or are showing behavior we can't trace a pattern from. Which makes me think either you got a bad CPU, a bad mobo, or a bad PSU. I call them the Trio of Trouble because they are very difficult to diagnose and usually show very random and untraceable behavior. Unfortunately unless Prime95 tests show something definitive (in this case they don't) it's had to resort to replacing each of those three parts until the symptoms disappear with previous cases involving one of these bad components. I'm sorry I cannot be more accurate in my analysis.

I'd start with the CPU, since previously it was made aware that you had some heating issues with your CPU and it may have caused permanent damage to the CPU before you could get a better cooler on in time.

 

[gagnon0vincent]

Well I tried windows 8 and found that my CPU clock speed keeps changing under load. What does it mean?

 

[Britton30]

Well I tried windows 8 and found that my CPU clock speed keeps changing under load. What does it mean?

Sounds normal, how much variance is there and what is loading it?

 

[gagnon0vincent]

Well when I was running prime 95(100% load), the clock speed was changing from 4.06 to 4.10 ghz.

 

[Britton30]

That is fine, prime exerts varying stress level depending on the calculations it does. The CPU can cut back too if it "thinks" all the clock speed isn't needed.

 

[gagnon0vincent]

I found the problem. I fell so stupid. I guess this kinda stuff only happens on the first build. The bios default vcore was wrong so I just changed it to 1.14 voltes and now everything works(9 hours on prime 95). I will do some more testing then if everything is good I will change the status to solved. Thanks for your time.

 

[Britton30]

Good to hear you found the Vcore issue. This sort of thing can happen on the 9th or 10th build too.

 

[Vir Gnarus]

I know that some mobos have a default setting that is an OCed variation that they setup under the belief that it's still within safe parameters for any typical system setup. However, it's still designed on speculation, so it still has the potential to bug things out. Usually there are default 'safe' settings available on the BIOS as well. However, do be aware that sometimes existing hardware problems can actually just end up manifesting and being exacerbated by even the most miniscule alterations that would otherwise be stable to healthy parts. Be vigilant to ensure that's not the case.