Overclock.net banner

1 - 11 of 11 Posts

·
Registered
Joined
·
743 Posts
Discussion Starter #1
i'm building another computer. i left the computer running after getting it all setup after going to bed to see if it would still be on when i wake up. well i wake up (7:40am) noticing it was indeed still on but sitting at a windows login screen. i thought that was odd because i left it already logged in. so after logging in i admittedly check event viewer. the only "errors / critical's" i had was two reports, one stating the previous shutdown at 2:22am was unexpected with a log time stamp of 3:11am and windows wasn't shutdown cleanly this could be caused by crash, power loss, malfunction with a log time stamp of 3:11am.

no bsod was recorded. i used bluescreenview to check for any bsod crash dumps and none where found.

i checked task manager and it stated an uptime of only 4 hours when it should indeed have been 9 hours.

what i found interesting is that apparently according to windows the power loss happened at 2:22am, and the boot up time from where windows "recovered" was 3:11am. so the machine was off for 50 minutes before it somehow, magically, turned itself back on.

first thing i did was to check if i actually did have a power loss to at least my room. my imac was still on with an uptime of 2 days. my file server was still on with an uptime of 3 months. my mac is connected into the same surge protector as the computer i'm putting together so i figured if it was the wall socket my mac should of turned off as well.

if it was the motherboard, processor, ram, gpu, i figured a bsod would of occurred. so i'm left here baffled. i don't understand how it randomly turned itself back on after a "power loss." i'm leaning towards it being the power supply. i'm still leaving the motherboard as a possibility but seeing how it was off for near an hour before its resurrection it leaves me thinking its the power supply.

right now i'm putting some load on the computer by running heaven benchmark to see how it runs.

specs of the machine:
intel 6700k (no oc, turbo off, stock 4ghz, voltage set to offset with a -0.080, c-states 1 - 6 enabled, 7 and 8 off)
gskill tridentz 32gb ddr4 3000 (xmp profile used)
asus maximus viii gene
gigabyte xtreme 980 ti
evga p2 750
sound blaster zx
 

·
Moderator
Joined
·
13,090 Posts
Well a couple things could have happened:

1) PSU crash - This was likely not captured by Windows. If your BIOS was set up to either "Turn on after AC loss" or "Return to last state after AC loss", this would cause it to reboot. Once it rebooted, maybe it did some updates or attempted Start-up Repair in that 50 minutes then restarted again which gave you the uptime counter you are seeing.

2) Hardware driver crash - The video driver crashed hard enough, or memory error, chipset driver error. Any of those can be bad enough it might skip the BSOD completely in some cases. For some reason the computer restarted itself to recover from the error and either took 50 minutes to recover before restarting, or went through start-up repair or updates like #1

3) Your computer actually had a scheduled restart from another program, of updates, or something else that just caused it to restart itself.
 

·
Registered
Joined
·
743 Posts
Discussion Starter #3
Quote:
Originally Posted by axipher View Post

Well a couple things could have happened:

1) PSU crash - This was likely not captured by Windows. If your BIOS was set up to either "Turn on after AC loss" or "Return to last state after AC loss", this would cause it to reboot. Once it rebooted, maybe it did some updates or attempted Start-up Repair in that 50 minutes then restarted again which gave you the uptime counter you are seeing.

2) Hardware driver crash - The video driver crashed hard enough, or memory error, chipset driver error. Any of those can be bad enough it might skip the BSOD completely in some cases. For some reason the computer restarted itself to recover from the error and either took 50 minutes to recover before restarting, or went through start-up repair or updates like #1

3) Your computer actually had a scheduled restart from another program, of updates, or something else that just caused it to restart itself.
i didn't think about the "turn on after AC loss." I have to check the bios to see if that was indeed enabled.

it could be possible the machine had hard locked and sat there for 50 minutes before some how unlocking and rebooting itself. i don't know if thats possible for it to randomly power itself off from a hard lock.

at first i thought it could of been updates but i made sure to disable windows auto updates via gpedit, and that doesn't explain the power loss and ~50 minutes in between said power loss and booting back up into windows.
 

·
Moderator
Joined
·
13,090 Posts
Quote:
Originally Posted by orlfman View Post

Quote:
Originally Posted by axipher View Post

Well a couple things could have happened:

1) PSU crash - This was likely not captured by Windows. If your BIOS was set up to either "Turn on after AC loss" or "Return to last state after AC loss", this would cause it to reboot. Once it rebooted, maybe it did some updates or attempted Start-up Repair in that 50 minutes then restarted again which gave you the uptime counter you are seeing.

2) Hardware driver crash - The video driver crashed hard enough, or memory error, chipset driver error. Any of those can be bad enough it might skip the BSOD completely in some cases. For some reason the computer restarted itself to recover from the error and either took 50 minutes to recover before restarting, or went through start-up repair or updates like #1

3) Your computer actually had a scheduled restart from another program, of updates, or something else that just caused it to restart itself.
i didn't think about the "turn on after AC loss." I have to check the bios to see if that was indeed enabled.

it could be possible the machine had hard locked and sat there for 50 minutes before some how unlocking and rebooting itself. i don't know if thats possible for it to randomly power itself off from a hard lock.

at first i thought it could of been updates but i made sure to disable windows auto updates via gpedit, and that doesn't explain the power loss and ~50 minutes in between said power loss and booting back up into windows.
It could definitely hardlock for 50 minutes before giving up and restarting like a regular error recovery from BSOD's. Maybe it was actually having a hard time dumping from your RAM to Disk because of a chipset driver or something affecting your memory and/or disk I/O. Finally it gave up and just restarted which is why you had no crash dump to inspect.

It wouldn't hurt to maybe run a full MemTest on your RAM and a full SMART test on your main boot drive from outside the OS using a Livedisk or something.
 

·
Registered
Joined
·
743 Posts
Discussion Starter #5
Quote:
Originally Posted by axipher View Post

Quote:
Originally Posted by orlfman View Post

Quote:
Originally Posted by axipher View Post

Well a couple things could have happened:

1) PSU crash - This was likely not captured by Windows. If your BIOS was set up to either "Turn on after AC loss" or "Return to last state after AC loss", this would cause it to reboot. Once it rebooted, maybe it did some updates or attempted Start-up Repair in that 50 minutes then restarted again which gave you the uptime counter you are seeing.

2) Hardware driver crash - The video driver crashed hard enough, or memory error, chipset driver error. Any of those can be bad enough it might skip the BSOD completely in some cases. For some reason the computer restarted itself to recover from the error and either took 50 minutes to recover before restarting, or went through start-up repair or updates like #1

3) Your computer actually had a scheduled restart from another program, of updates, or something else that just caused it to restart itself.
i didn't think about the "turn on after AC loss." I have to check the bios to see if that was indeed enabled.

it could be possible the machine had hard locked and sat there for 50 minutes before some how unlocking and rebooting itself. i don't know if thats possible for it to randomly power itself off from a hard lock.

at first i thought it could of been updates but i made sure to disable windows auto updates via gpedit, and that doesn't explain the power loss and ~50 minutes in between said power loss and booting back up into windows.
It could definitely hardlock for 50 minutes before giving up and restarting like a regular error recovery from BSOD's. Maybe it was actually having a hard time dumping from your RAM to Disk because of a chipset driver or something affecting your memory and/or disk I/O. Finally it gave up and just restarted which is why you had no crash dump to inspect.

It wouldn't hurt to maybe run a full MemTest on your RAM and a full SMART test on your main boot drive from outside the OS using a Livedisk or something.
oooo, ok that makes sense. i'll run a memtest and a smart test (the boot drive is a samsung evo 850 250gb ssd) after finishing this heaven loop. i'm letting it run for an hour. i figured if it is the power supply, and it crashed with computer at idle with a barely any load, putting on a mild load like heaven would at least cause it to crash again. am i correct in assuming putting a mild load on the psu could help show its instabilities if there are any?

i don't know if this makes a difference but the ram requires 1.35v's, and in the bios 1.35v's was indeed inputted, but both the bios and hwinfo reports 1.34v's. the gene does have its latest bios flashed.
 

·
Moderator
Joined
·
13,090 Posts
Quote:
Originally Posted by orlfman View Post

Quote:

oooo, ok that makes sense. i'll run a memtest and a smart test (the boot drive is a samsung evo 850 250gb ssd) after finishing this heaven loop. i'm letting it run for an hour. i figured if it is the power supply, and it crashed with computer at idle with a barely any load, putting on a mild load like heaven would at least cause it crash again. am i correct in assuming putting a mild load on the psu could help show its instabilities if there are any?

i don't know if this makes a difference but the ram requires 1.35v's, and in the bios 1.35v's was indeed inputted, but both the bios and hwinfo reports 1.34v's. the gene does have its latest bios flashed.
The RAM voltage shouldn't be an issue if you are running the RAM at its stock settings/timings.

PSU's are a weird thing though. Some power supply designs are actually less efficient when they aren't above 50% of their rated power so running the PSU below 50% of it's max can actually cause some fluctuations on the different voltages. It can also depend on how much power the 5V rail is pulling vs. the 12 V rail under certain conditions.

I had a power supply die last year with a near perfect 12 V rail, but the 5 V would constantly go out of spec under different loads. So it could also just be one bad part in the PSU.
 

·
Registered
Joined
·
743 Posts
Discussion Starter #7
Quote:
Originally Posted by axipher View Post

The RAM voltage shouldn't be an issue if you are running the RAM at its stock settings/timings.

PSU's are a weird thing though. Some power supply designs are actually less efficient when they aren't above 50% of their rated power so running the PSU below 50% of it's max can actually cause some fluctuations on the different voltages. It can also depend on how much power the 5V rail is pulling vs. the 12 V rail under certain conditions.

I had a power supply die last year with a near perfect 12 V rail, but the 5 V would constantly go out of spec under different loads. So it could also just be one bad part in the PSU.
so far its been running stable since it powered back on at 3:11am. heaven so far for an hour now with zero issues. do you think it could of been the psu then? what do you think it was?

edit:
well i reboot and checked bios and i had set "restore ac power loss" to "power off" so now i don't think it was the power supply. if power loss occurred it should of stayed off.

i'm guessing now it probably was some sort of hard lock?

gonna run memtest

any more tips appreciated greatly.
 

·
Moderator
Joined
·
13,090 Posts
Quote:
Originally Posted by orlfman View Post

Quote:
Originally Posted by axipher View Post

The RAM voltage shouldn't be an issue if you are running the RAM at its stock settings/timings.

PSU's are a weird thing though. Some power supply designs are actually less efficient when they aren't above 50% of their rated power so running the PSU below 50% of it's max can actually cause some fluctuations on the different voltages. It can also depend on how much power the 5V rail is pulling vs. the 12 V rail under certain conditions.

I had a power supply die last year with a near perfect 12 V rail, but the 5 V would constantly go out of spec under different loads. So it could also just be one bad part in the PSU.
so far its been running stable since it powered back on at 3:11am. heaven so far for an hour now with zero issues. do you think it could of been the psu then? what do you think it was?

edit:

well i reboot and checked bios and i had set "restore ac power loss" to "power off" so now i don't think it was the power supply. if power loss occurred it should of stayed off.

i'm guessing now it probably was some sort of hard lock?

gonna run memtest

any more tips appreciated greatly.
Once you run MemTest and a full SMART test we will have a better idea. Also maybe Windows CHKDSK, I forget how to make it initiate a full check, but that would at least fix any bad files that may be related to drivers and caused a hard lock.
 

·
Registered
Joined
·
743 Posts
Discussion Starter #9
alright so now my computer won't start at all to even do a memtest. when i press the power button the leds come on, fans kick on, but the motherboard sits at a post qcode of 00 and nothing else.

so i think its the motherboard unless a cpu / ram / psu can cause this.

edit:
since i bought the parts off of amazon i'm just going to send the motherboard, psu, and video card back for replacements. video card is random i know but it did produce very loud coil whine when under load and since i'm returning two things for replacement might as well. replacements should be here tomorrow.

i am sending the ram back to newegg for a refund. switched it with a corsair dominator kit off of amazon. even if its not the ram the dominator kit just matches everything better and should be here tomorrow.

i would send the cpu back for a replacement too but since i bought it off of newegg, and newegg doesn't allow for refunds on processors, if i sent it off today i wouldn't get its replacement until next friday the earliest, the following tuesday the latest.

though from what it sounds like it isn't the cpu so it should be good. unless anyone else thinks it could be it?
 

·
Moderator
Joined
·
13,090 Posts
I doubt it's the CPU, I personally have never had a CPU die on me unless I was overclocking the crap out of it with way too much voltage.

If you are able to just swap out everything under warranty, then I'd say at that point you may as well just do a fresh install of Windows as well.
 

·
Registered
Joined
·
743 Posts
Discussion Starter #11
Quote:
Originally Posted by axipher View Post

I doubt it's the CPU, I personally have never had a CPU die on me unless I was overclocking the crap out of it with way too much voltage.

If you are able to just swap out everything under warranty, then I'd say at that point you may as well just do a fresh install of Windows as well.
yeah as do i. might as well do a fresh windows install since i pretty much have to rebuild it from scratch lol.

question though, it being stuck in "post" with "00" on the debug screen, if that is by chance left for 20 - 40 minutes at that stage, it won't cause any damage to lets say the cpu will it?

i ask because i left it at that stage for 30 minutes wanting to see if it would go past 00 and turn it self fully on.

its weird though. the sound blaster zx red led's light up so power is going to the pci-express lanes at least when i try to power it on. but the windforce logo on the 980 ti xtreme doesn't light up, and neither does the pci-express power connector ports light up since the card has led's on the power connectors to alert you that the cables are plugged in.
 
1 - 11 of 11 Posts
Top