3

I've been fighting with this problem for the last 3 weeks where my machine will just instantly reboot. No BSOD, and when i checked the event log all that was reported was the generic "Kernal-power" error with the detailed information pointing to a hard crash.

This is a machine that was working for 18 months before these crashes started happening. When they started happening is after I added 3 HDs in a RAID-5, upped the memory to 12gb, moved to a new house, added a SSD and added about 5 case fans.

I have thus eliminated the RAID, and determined that the SSD was not the cause (because it was still crashing even though the ssd wasn't connected).

I've run memtest several times over night with no memory problems showing up.

I've run IntelBurnTest to max out the cpu to see if it was a heat issue and at full tilt after 20 min it was only at 85C and the machine didn't crash. I also took a look at the voltages during this test, with a screenshot at the bottom of this post

I've ruled out a software issue by reinstalling windows 7 ultimate x64 a total of 5 times, but even during that the install it crashes. Happens sometime during file copying at the beginning, or during uncompressing files, or sometimes during running windows update.

The only discernible pattern i can see is that it seems to crash when hard disks might be spinning up or when they are accessed heavily from large file transfers.

My current guess is that it is probably an issue with the MB, PSU or the power coming through the outlet.

Any suggestions of what i could try to troubleshoot or what may be wrong?

Specs PSU: Seasonic M12 700w Mem: 12gb CPU: i7-920 with stock heatsink MB: Asus P6T HDs: 3 green WD and 1 Corsair force 3 120b with 1.3.3 firmware

UPDATE: I've reseated everything save the cpu on my motherboard and replugged in everything as well. Now my machine can only cold boot, when it does boot. Other times it just sits here with power and doesn't even POST or do the little beep right before it POSTS. And twice it just sat there continuously beeping at me. Unfortunately my asus manual has nothing to saying about what the beeping could mean.

Also on two occasions when it did crash, it wiped the settings in the BIOS.

I'm really starting to wonder if something is just busted on the mb.

Here is the event data from the event log though from what i've found it seems having all zeros means windows doesn't know why it crashed.

EventData

BugcheckCode 0 BugcheckParameter1 0x0 BugcheckParameter2 0x0 BugcheckParameter3 0x0 BugcheckParameter4 0x0 SleepInProgress false PowerButtonTimestamp 0

UPDATE 2: I ran memtest multiple times with no problems whatsoever. Regardless I decided to remove the most recent additional memory I put in and now no more cold boot issues, computer is FASTER than before AND going on 3 days with only one reboot. Which makes me think it is a motherboard issue, and might be a northbridge controller overheating (though I don't know why it would do that).

Also, when it was rebooting constantly, I noticed that it was always the same HD that had a "Write error" in SMART. So that drive is also suspect. I think I have/had mulitple issues which is making this so hard to diagnose.

Running full tilt voltages Running full Tilt

Idling Voltages
Idling Voltages

Update 3:

So I've replaced the PSU, GPU and now the MB and the machine still crashes. So i popped out all but 1 memory module and it continued to crash, but then I found that it continuously crashed on 3 of the 6 memory modules that I had. So even though Memtest wasn't reporting any problems, it turns out there was some problems with my memory...

Geuss the moreal of the story is to trust the old school hardware checks and to take software checks with a grain of salt.

9
  • Good post, I get 2 feelings from this, the chipset (not the cpu or memory), and the power. When you say no BSOD, is it set to autorestart on a crash? Because that could certannly lean the clues one way or another. Your into discovery, for 1 try and put fan blowing on chipset, or any adjust on the voltage of the chipset. for the other reduce power needs to test. That just leaves jacking up the voltages on the cpu and memory to test again.
    – Psycogeek
    Nov 16, 2011 at 0:25
  • And the SSD, I dont have enough experience with SSD, but there are many many reports where the SSD in combination with SATAIII and if your using a different Port and Controller Because of that. Which goes along Much better with the actual problems your having of it occuring during disk activity. When you removed the SSD, Do you still have different ports being used for the HD, you did not use before?
    – Psycogeek
    Nov 16, 2011 at 0:47
  • Yes when I removed the SSD I was still getting crashes. Currently the SSD is attached but, I'm not using it as the OS drive just to make it easier to troubleshoot this problem. The disk activity crashes I am getting was for transfers between two relatively new WD drives. My board doesn't have SATAIII so that should be ok. I have also have it set to no autorestart on crash. But it hasn't crashed since made that change Nov 16, 2011 at 1:01
  • By the way, I dont allow my own cpus to get that hot, the intel will throttle before breaking . but. . After being around many "overclockers", we would never use the word "only" for 85*C :-) The temps on your burn test are too high for me. For little extra funds for a good heat sync and some good thermal goop, you should be able to address that.
    – Psycogeek
    Nov 16, 2011 at 1:23
  • 1
    @MarkHosang the i7 is only rated to 72C. The T-Case rating is only 67.9C, and from the datasheet on this processor generation, the entire i7 line is due to go up no more than 72C. You're also way over the rated TDP for that chip, so I'd tone down your voltages/clock speed unless you want to buy a new CPU within a year. At the end of the day, it's your call, but those temperatures are way too hot. Dec 20, 2011 at 12:32

2 Answers 2

2

A Seasonic 700W should be plenty for that setup. If you get hard crashes during Windows install, in various spots, and the drives have been checked as good, I'd say it's about a 98% change it's a RAM problem. Try removing the new 12GB you added and see how it behaves for a while, and/or test the RAM for 24+ hours.

3
  • i've done 8 hours scans of my memory about 3 times already with no errors in the memory at all. Nov 16, 2011 at 5:51
  • I've done additional memtest scans on my memory and still no problem there. Are there times when the memory is bad, but it doesn't create a problem in memtest? Nov 20, 2011 at 23:37
  • In the end I ended up popping in one memory stick at a time and found out that two of the three i recently put in were bad. So got rid of those sticks and no more reboot issues. May 22, 2014 at 15:36
1

Just about any piece of hardware connected to the PC or motherboard can cause a spontaneous restart, I had a buggar a few years back doing the same thing, it was a case fan causing it, its a process of elimination to find the cause.

If it only does it during a file copy to disk then I suspect hardware related to hard drives or a hard drive itself. Also a PSU overloaded or failing can do this.

Run Memtest86+ for at least 24-48 hours to stress test the memory.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .