frequent shut downs

[overclocking101]

so at STOCK SETTINGS my pc recently started randomly hard shutdown no bsod just looping sound. Now it is happening more frequent. can it be my memory is dieing?? I increased imc volts to 1.2 over the stock 1.1 and it still does it. im hoping its not a cpu or SSD problem.
it happens at desktop, or watching movies, or browsing the net. usually not while gaming not hard shut downs. specs are up to date the memory is about 6/8months old since I bought them used so probably 1/1.5yrs old. cpu is 3/4months old. everything is withing 3/4months except psu and ram.

Right now im going to run memtest and see what happens.

[erocker]

Originally Posted by overclocking101

Right now im going to run memtest and see what happens.

Looping sound, I am thinking so. You could try increasing voltage to your RAM as well. If that doesn't do it and/or your RAM checks out good, try increasing voltage to other components. Have you checked for vdroop under load and all that?

[Arrakis+9]

check your event log in windows for clues of what might be happening

start > right click computer > manage > event log > windows logs > system

Sort by level and look for all the ERROR events, check the time and date and try and match when it was happening

[overclocking101]

all the events from when it does bsod point to ram. I ran memtest for 2 full passes and no problems. now the weird thing, I upped vcore a small amount and now it works, no freeze in 4 hours. that sucks that means my 655k may be degrading before it would run for an hour 2 max then shutdown

[overclocking101]

ok well I ran with the increased vcpre fpr 29hrs then hard froze while watching movies. is it possible my "movies" drive is screwing up and it hard freezes when it hits bad sectors??

[cadaveca]

Yes, for sure a bad drive can cause shutdowns.

How long does it take for your system to POST...3 seconds....or like 10 or more?

[overclocking101]

like 10 or more seconds. im starting to think its ULPS on my gpu's. and i cant seem to turn it off for some reason.

[overclocking101]

well it is not ULPS I used RBE to edit those clock out of the bios, and 45min later bam freeze loud sound for like 10-15 sec, then restart. it has to be either disk related or O/S related.

[overclocking101]

just hdtuned my mechanical drives and somehow it says every drive is in warning status wtf??

[cadaveca]

hdtune isn't exactly accurate with SMART sometimes. Use the drive maker's tool.


If your system is taking that long to post, I'd be looking at memory too. I am in my OS within 10 seconds(of course, the SSD makes it fast).

more devices and more ram can make post take a bit longer, but I have 5 physical drives, and even wit h8GB, it doesn't take very long at all. one thing you may want to try is disabling teh "vanity" logo that is shown when hte board boots, so you can see it going through the POST, etc. if it hasn't started loading the OS within like 5 seconds, something is up. both a bad drive and bad memory can make POST take a bit longer too.

[overclocking101]

no it takes like 10-15sec to RE-BOOT from the freeze meaning it hangs for like 15 sec before resetting even on a hard reset. my o/s drive is ssd to get to windows from post takes 10 sec. but theres no error message no nothing.

[cadaveca]

Still seems a bit lnog to me.

Anyway, let's start at the beginning. Remove the power plug from the PSU. Remove CMOS battery. Clear CMOS(30 second, no less, more is better).

Replace battery, power cord, and boot up. Go in bios and set up drive controller, etc, but leave mem and cpu at stock.

Get into windows. Run IBT @ standard settings for 50 runs. Make sure to set number of cores, too.

Then, run IBT at "MAXIMUM", again, being sure to set # of cores. 50 runs as well...will take couple of hours.(50x about 400sec).

If those pass, then I'd look at the drive itself. If those two fail, look at cpu/mem, but do not exclude the drive...be sure to test using the drive maker's tools, but there is no point in testing with a system that has stability issues.

I need to update my specs. I aam running the same Mushkin redlines, I think...mine are ridgebacks though.

[TheMailMan78]

PSU problem? Do you have a spare you can test?

[overclocking101]

Originally Posted by cadaveca

Still seems a bit lnog to me.

Anyway, let's start at the beginning. Remove the power plug from the PSU. Remove CMOS battery. Clear CMOS(30 second, no less, more is better).

Replace battery, power cord, and boot up. Go in bios and set up drive controller, etc, but leave mem and cpu at stock.

Get into windows. Run IBT @ standard settings for 50 runs. Make sure to set number of cores, too.

Then, run IBT at "MAXIMUM", again, being sure to set # of cores. 50 runs as well...will take couple of hours.(50x about 400sec).

If those pass, then I'd look at the drive itself. If those two fail, look at cpu/mem, but do not exclude the drive...be sure to test using the drive maker's tools, but there is no point in testing with a system that has stability issues.

I need to update my specs. I aam running the same Mushkin redlines, I think...mine are ridgebacks though.

did all that before posting the last time, everything checks out, hell I even left the cpu-vcore on auto everything even imc and it passes ibt, prime95 etc etc. but after watching movies or sitting at desktop for a long while it freezes with a loud ass noise hangs there for 15sec then restarts and boots fine rinse and repeat.

Originally Posted by TheMailMan78

PSU problem? Do you have a spare you can test?

I have another I could test it with, this 950w isnt very old.

EDIT:
running ibt "maximum" 4 cores, xtreme burn mode AND furmark dual gpu no prob right now, psu meter is only pulling 700w at the wall

[TheMailMan78]

Originally Posted by overclocking101

did all that before posting the last time, everything checks out, hell I even left the cpu-vcore on auto everything even imc and it passes ibt, prime95 etc etc. but after watching movies or sitting at desktop for a long while it freezes with a loud ass noise hangs there for 15sec then restarts and boots fine rinse and repeat.


I have another I could test it with, this 950w isnt very old.

EDIT:
running ibt "maximum" 4 cores, xtreme burn mode AND furmark dual gpu no prob right now, psu meter is only pulling 700w at the wall

I would just test it with a different PSU. That way at least you know the power is good.

[cadaveca]

Originally Posted by overclocking101

running ibt "maximum" 4 cores, xtreme burn mode AND furmark dual gpu no prob right now, psu meter is only pulling 700w at the wall

I have specific reasons for tell you what I did. Adding in the furmark testing invalidates the IBT test. And you very specifically need to run the tests the way isaid, and for how long I said, or I cannot suggest any alternatives. Prime95 needs to run near 10 hours minimum for you to declare "Prime95 stable", and that would be the next test after IBT.


It is also possible that you page file or superfetch file is corrupt. google how to delete them, and see how it goes.

And again, scan the drive with tools from the drive maker. Do not use Windows testing tools. Each can assign SMART data differently, as as such, other SMART tools may give incorrect readings.

If you ahve an etr adrive, I'd suggest also re-installing windows. Random crashes ca ncause registtry corruption too, and you problem may just simply be a corrupt OS.

[overclocking101]

Originally Posted by cadaveca

I have specific reasons for tell you what I did. Adding in the furmark testing invalidates the IBT test. And you very specifically need to run the tests the way isaid, and for how long I said, or I cannot suggest any alternatives. Prime95 needs to run near 10 hours minimum for you to declare "Prime95 stable", and that would be the next test after IBT.


It is also possible that you page file or superfetch file is corrupt. google how to delete them, and see how it goes.

And again, scan the drive with tools from the drive maker. Do not use Windows testing tools. Each can assign SMART data differently, as as such, other SMART tools may give incorrect readings.

If you ahve an etr adrive, I'd suggest also re-installing windows. Random crashes ca ncause registtry corruption too, and you problem may just simply be a corrupt OS.

ill run ibt alone, and get back at you, ill also check the pagefile, come to think of it, It started happening when I changed my cache from 1024mb to auto,ill try that to, what is an etr adrive??

[cadaveca]

extra drive, sry...kinda typing things in a rush today...water started falling from the ceiling of my house...from light fixtures. Guys are working in the house, pulling the damaged ceiling down and making repairs.


Anyway, if you have an extra drive, I'd wipe it clean, install a fresh copy of windows on it, and then test from that. If windows is corrupt, that might affect test results.

[TheMailMan78]

I would hate for you to do all that and it be a bad PSU.

[cadaveca]

meh. swapping PSU ofr another doesn't 100% confirm that the PSU is bad...I suppose it's possible, but you know what....


Proper testing isn't easy. there is no 100% quick answer to any stability problem...test it all, eliminate causes, and you find the real issue.

[TheMailMan78]

Originally Posted by cadaveca

meh. swapping PSU ofr another doesn't 100% confirm that the PSU is bad...I suppose it's possible, but you know what....


Proper testing isn't easy. there is no 100% quick answer to any stability problem...test it all, eliminate causes, and you find the real issue.

Well I always go simple first. First check power source and work your way up. 9/10 times its the simple crap that gets you. Tech guys like us always look to complicate problems instead of seeing if its plugged in.

[cadaveca]

True enough. I mean, to test a psu, you do what?

PLug in a meter, load it up, watch for volt sag?


Or so you get a proepr load tester, hook the PSU and test?

Or do you pull the system apart, pull the psu, replace with another, and then test?

Of course, if the problem is due to a corrupt windows install, switching PSUs might show nothing...or the same problem...hard to say...

Still gotta go with a fresh OS install.

[TheMailMan78]

Originally Posted by cadaveca

True enough. I mean, to test a psu, you do what?

PLug in a meter, load it up, watch for volt sag?


Or so you get a proepr load tester, hook the PSU and test?

Or do you pull the system apart, pull the psu, replace with another, and then test?

Of course, if the problem is due to a corrupt windows install, switching PSUs might show nothing...or the same problem...hard to say...

Still gotta go with a fresh OS install.

Oh sure do the OS first. Then the PSU.

[RuskiSnajper]

Looping sound = definitely unstable RAM voltage/clock/ = bad settings/bad ram

OEMs are (IN)famous for bad bios settings , especially your local OEM from the town (non international)

[overclocking101]

ran "IBT 50 passes stable. all maximum settings. ill try to run smart checks via dos, I know seagate has a dos program but im unsure about WD. the raptor is only a few months old, and hdtune says is on its way out, I find that hard to believe

I really dont want to do a fresh install if I dont have to, I really dont want to have to redo all my shit.