BSOD when stressing CPU out of the blue

Ok, I will run Memtest tonight with RAM on the set 1.50V, DDR-1600, Timing 2N. Had Vcore back to auto, but will set it to 1.15V. If it passes I'll bump IMC up and keep HT and Turbo disabled. What am I going to test with these settings if I may ask? I think the result will be high temperatures again eventually.

I already ran Memtest a week ago on default settings for 10 hours, passed without errors.

I'll keep you posted tomorrow again!

Ran Memtest for 7.5 hours, no errors found. It only says 5 passes, but I think it ran 15 or 25 and doesn't show the 1 or 2.



I just ran IBT again with IMC bumped to 1.30V and the rest of the mentioned settings. I already noticed temperatures were high on booting. When I ran IBT temperatures went up pretty high again, peaking at 87C. Eventually just after the 2nd run I got the error again

My thought is still the cooler. Those temperatures are way to high for stock clocks.

Yeah I am still going to take everything out and remount the cooler, probably towards the weekend.

But still, on default settings I have pretty average temperatures and things like Battlefield 3 have definitely been influenced in a way. My system had also run fine for 3 weeks till the BSOD's started.

Also when I play BF3 it doesn't immediately shoot up to high temperatures, this goes pretty steady. I would say if the cooler wasn't on properly it would shoot up within an instance.

Anyway, it seems like my last hope, so I'll be trying it ASAP!

The temps wouldn't shoot up instantly. I had a core i3 540 @4.0ghz with a stock Intel cooler and one of the push pins popped out and my temps would start slowly rising from idle of 36c to a max of 85c while gaming. I also think the lack of airflow in that case isn't helping much plus a 570 runs pretty hot.

I had a nzxt h2 which has horrible airflow and by just switching to a better case my overall temps dropped by 10c.

Yeah ok, does sound plausible.

The airflow is really bad indeed. I would like to use push/pull cooling but my Corsair Vengeance RAM is too high to place a cooler above. Eventually I should get a new case and maybe some low profile RAM as well.

Actually a push pull configuration doesn't help all that much. By just getting a newer case with better airflow would be a lot better.

Also what leads me to think its the cooler is the fluctuations in the max core temps. Maybe an air bubble in the TIM might be causing an uneven displacement of heat.

How much are you looking to spend on a newer case?

Yeah maybe. I evenly spread a thin layer of TIM with a bankcard. Always done it this way, but read other people advising to just stick the cooler on a small pea sized dot and then even it out. This is almost impossible with the Scythe Mugen 2, since I have to mount it with the motherboard upside down. I've placed it with a thin layer 3 years ago and it ran perfect for almost three years.

Was thinking of seeing if remounting the cooler would have any effect.
Then if I would have to go for a new cpu/mobo then I would get a new case together with it. I guess if the problem would be solved after remounting, it's still really worth the investment. I think somewhere around 120 euros (about $150), with some good future-proof options, like maybe front panel USB3? I guess a bigger type of case would be nice as well, not having to fiddle about so much in a small space!

OK, remounted the cooler. Couldn't see anything wrong with the cooler when I had the mobo out. Cleaned CPU/cooler and re-applied TIM. Three-double checked it was mounted on properly. Built it back in and booted up.

Idle temps after settling were 39-42C with Vcore on auto, IMC on 1.15V, DRAM on 1.50V, HT and Turbo on.

Ran IBT for 10 runs on Very High and it passed, but the temperatures weren't good. Near the end it actually hit 95C. I'm suprised it actually kept going.



Now there is is something what I've found on the motherboard. A bent ferrite choke (Yageo R68).



From what I can google, it has to do with the VRM circuit. This would sound plausible I guess, also with the initial BSOD 0x00000124; increase Vcore/VTT.

I didn't notice if it was bent or not when I received the motherboard. I gently tried to bend it back, but it really wouldn't budge. There would have to been some serious force applied to have bent it like that, and I'm sure I haven't done that.

The problem is, this is my 2nd motherboard that I got from a guy from another forum. The first one was already an RMA he still had lying around. A mosfet burnt through on that one after I had used it for 1.5 month, and I had to send it back to him. He went back to the store and got a 2nd RMA, which is my current Asus P7P55D-E. It would be insane to tell him now, a bit over a month later, that this one needs replacing as well.

It ran fine for a little over three weeks, so I'm not even sure if this is actually the main issue. I am still doubting if the previous mosfet burnout has damaged my CPU as well. I've always had the feeling my CPU was acting up over the last year (jerkyness in BF3) and now it's only gotten much worse.

Either way, the temperatures I get on almost stock are way too high. There is something wrong for sure and I don't really know what to do now.

Any more advice?

Your CPU is getting far too hot. It's your heatsink. I can almost 100% guarantee it. You either have a bad mount, bad thermal paste application, warped surface, or a combination of the three.

What kind of TIM are you using?

I am pretty sure it's not a mounting problem. I already checked it when I took the cooler off. After putting it back on I took 10 minutes inspecting it from all different angles and it is properly mounted. You can't really go wrong with the Mugen 2 either I think, since it's connected with 4 screws. They are either properly connected and tightened or not.

The surface of the cooler is still as smooth as it's ever been. It ran good for 3 full years till my problems started acting up. I never had taken the cooler out or changed the setup in that whole period.

I am using Cooler Master High Performance Thermal Compound

http://www.hardwaresecrets.com/imageview.php?image=34282

Do you still have a stock Intel cooler?

peanutbudder said:

Do you still have a stock Intel cooler?

Click to expand...

Unfortunately not That mighy be a good idea tot test though, maybe get one that also is compatible with a future upgrade. Any recommendations?

Just to let you know, I went out and got an i5-3570K with a Gigabyte Z77X-D3H, a Mugen 3, some new RAM and a new Corsair 500R case to put it all in. Massive improvement! The case cools really well too.