computer keep shutting down/restarting itself

Status
Not open for further replies.
C

catalzzy

Weaksauce
Joined
May 13, 2006
Messages
99
Hello,

I am kind of pissed off with my current rig which is now starting to have occasional shut downs and restarts. I mainly use this for 3d modeling and rendering. Rage kicks in when it shut down! i feel like its gonna damage the hardrive or sdd because of sudden shutdown.

Sometimes it would shut down and power itself on while on the screen its blank, keyboard lights and mouse is off. So i have to hold the power button to fully turn off then on again.

Sometimes Restart its like it just shut down by itself and restarted back to normal.

I replace the psu with coolermaster vs650 gold certified psu 3 months old and a full tower case with lots of fans so its noisy.

The voltage readings from the hw monitor and the asus bios says :-

3.3volt rail :- 3.344V
5volt rail :- 5.00 and sometimes it droop to 4.960v
12volt rail :- 12.096v

Sometimes i get artificating while web browsing on chrome or firefox.

I am either suspecting its the motherboard or the psu.

I have spare z87 motherboard kept at storage and i cant be bothered to move it around because of the original windows 8.1 key which so far i think its locked to the same motherboard which is the H97.

I overclocked my i7 4770k delidded cpu by not much. Just at 4000ghz at 1.18v with c states and EIST turned off.

Idle temps average 42. I am using coolermaster seidon120XL. On load such as rendering, it max at 72c average.

Thanksssss.

Specs is on the sig. :)
 
Last edited:
Run Memtest overnight, unless you've already done the obvious and didn't say.

Set the CPU back to stock clocks.
 
honestly:

1. remove all overclocks, does it still shutdown randomly?
2. Artifacting sounds like overheating

Your reported max temperature: http://ark.intel.com/products/75123/
TCASE 72.72°C

"Case Temperature is the maximum temperature allowed at the processor Integrated Heat Spreader (IHS)."

3. Test the RAM use a tool like memtest (recommend at least a 24 hour run if you can, although saying that if the RAM is faulty it normally only takes an hour or so)
 
Did you even check that your PC is actually stable before using it. Because if not, what did you expect?

Go with the suggested mem test first. If that passes, run 24 hour stress test loading all cores using prime or similar, and auto record your max temps. Rather than manually checking temps while the PC is working and then claim that to be your max temps it would be better to measure temps just before your PC dies.

Does your PC reboot? If so, then sort it out using basic reasoning and a bit of trial and error. For example if your PC is overheating, then adjust cooling, reduce overclock slightly, whatever. If it looks like memory error, then remove all the sticks except one. etc. etc. etc. Rather than taking a complete guess and blaming one of your components with no analysis.

Redo. And keep doing this until your computer is stable for over 24 hours.

(and why is this even in the PSU section?)
 
Last edited:
Test at stock first then...

Both Chrome and Firefox are GPU accelerated browsers. If you're getting screen corruption, I'd find out whats up with the video card first. I don't think this is a PSU issue.
 
you could try disabling hardware acceleration in chrome settings......artifacts could be related to the power play issues the 7900 series has, specifically the memory speed dropping so low them jumping to full clock.i have noticed some artifacts here and there myself (with chrome open)......since its winter anyway i disabled power play and don't recall seeing it again.....good thing those brand new omega drivers fixed everything lol

As far as your pc restarting or freezing i would focus running stock clocks until you get it stable.....possibly just needs a bios setting adjusted since your rails seem fine. My question is when are you getting theses sudden shutdowns? is it completely random or mainly when pushed?
 
Right hello all,

I just have done the mem test but its 8 hours and no errors. i know 24 hours should be more than enough but its like too long haha.

I have set all to stock. I havent started to stress my cpu yet. So far i am waiting if its gonna shut down suddenly.

When i got the shutdown, its random like be it 8 hours or 6 hours or 3 hours and so on. There is no BSOD.

Also, when it shut down, and it turns on itself. The fans were spinning but the cpu pump is not. So i have to unplug and plug again to make it working.

I might be thinking that the h97 is not fit for overclocking even for a little overclock which has to do with the phase power feed or the BCLK speed.

I feel like i want to change that mobo to my spare one z87 but i feel so lazy cos i have like tons of storage there. :( Plus the windows 8.1 oem key too.

I posted it on this because i thought it could be powersupply related hence the sudden shut downs.

Thanks
 
I posted it on this because i thought it could be powersupply related hence the sudden shut downs.
Click to expand...
Reason enough to post it here in my opinion....the turning on itself is possibly a bios setting like in many bioses that are usually defaulted to off.(like turn back on after power failure) I suggest download and run intel burn test and set to use nearly max ram and see if it can make 10 or 20 passes. you can substitute ibt for linix with latest binaries or even try occt power supply stress test. But just to be sure stress the cpu and gpu at max overnight at least just to make double sure the power supply is not a problem. If you get shutdowns like as if someone hit the power switch then i would further into the supply but if it looks more like a crash then it could be unstable even at stock, or a piece of software running (even some antivirus apps) causing the crash. crashes will and can cause corruption in windows that can be hard to pinpoint and fix so a fresh reinstall might be something to think about.
if you do make a new install make frequent backups thru the built in back up (file history). I prefer the system image backup for this but everyone has there own favorites. after you have the first problem restore the perfect backup and it might point to an app causing the crashes (if its software related)

install core temp, hardware mon pro or you favorite choice to monitor temps since if it crashes while the temps are low, we can rule out overheating. keep us posted and good luck sorting the issue, i know well figure it out. Oh and lastly reseat all cards and ram and your idea about using one stick at a time is not bad idea at all......could very well be just a bad stick:) Dont give up...once we confirmed the suppy is good you can have the thread moved by paul to intel section and believe me theres some very intelligent people there who will get you on the right path (to many people to name but its true)

Software wise HWinfo64 can monitor every detail, temp voltage your board has and could help at least rule certain things out, but its not required since everyone has their own favorites
 
Last edited:
it could be sleepmode problems or something.
or your GPU is wonky..
could be GPU drivers..
bad heatsink mount on gPU..
oy..
I was having same problem with an install a while back.
..restarted from scratch.
 
What else do you have running off the PSU? MB+GPU+SSD+HDs+FANS ... who knows you could have gotten a bad PSU to replace your old PSU. It happens sometimes. How much juice does the OC+GPU+4 HDs use? I never went beyond two HDs (pre-SSD days) but I'm an over-buyer when it comes to PSU max capacity vs actual usage.

But like everyone else has already said, stability first before any O/Cs.
 
Yeah... you're not going to get far overclocking by the BCLK. If you're lucky, you can hit 110 BCLK. Typically, you will get instability above 105. You bought the wrong board for overclocking. Actually, you bought the wrong chipset for a K processor.

Try HCI memtest. It is quicker at detecting memory errors in my experience. Then try CPU stress tests. Also try a GPU stress test like OCCT GPU test (I prefer OCCT because of automatic error checking).
 
Tsumi said:
Yeah... you're not going to get far overclocking by the BCLK. If you're lucky, you can hit 110 BCLK. Typically, you will get instability above 105. You bought the wrong board for overclocking. Actually, you bought the wrong chipset for a K processor.

Try HCI memtest. It is quicker at detecting memory errors in my experience. Then try CPU stress tests. Also try a GPU stress test like OCCT GPU test (I prefer OCCT because of automatic error checking).
Click to expand...

I didn't know some of the info you knew, but we both agree on the troubleshooting steps (i use occt for general checks on mine) and it warms up my room as benefit as well lol...so to me sounds like you gave good advice. It also points out how important research is for planning any builds, giving the info you gave;)
 
I know this is an old thread, but I'm having the same issues as the OP:

Random restarts
Artifacts in Chrome browser

The PC has never been over clocked, and I've had the same hardware in it since I initially built it a few years back.

CPU temp is good, all case fans work as does the fan on the HD5850 GPU.

I did a fresh win 7 install and the reboots haven't gone away.

I'll be testing the ram later today

Any other suggestions besides what's already been stated above?
 
Best to start you own thread. Gets more attention that way, and doesn't get bad attention from the mods (necro-posting is a big no no).

Best way is to stress test each component one by one. If you're unable to catch it that way, start replacing components with known good ones.
 
Rarely PSU related. Start this off in General Hardware and don't necro threads.
 
Status
Not open for further replies.
Back
Top