Tricky Shutdown/pwr cycle/CMOS issue

[gorignak]

I've looked around and seen similar issues but I felt this was still unique enough to warrant a new post. Sorry if you disagree, Not trying to annoy or offend. Well to start here are my specs:

1 x COOLER MASTER HAF 922 RC-922M-KKN1-GP Black Steel + Plastic and Mesh Bezel ATX Mid Tower Computer Case
1 x ASUS P6 X58-E PRO LGA 1366 Intel X58 SATA 6Gb/s USB 3.0 ATX Intel Motherboard
2 x XFX HD-685X-ZNFC Radeon HD 6850 1GB 256-bit DDR5 PCI Express 2.1 x16 HDCP Ready CrossFireX Support Video Card with Eyefinity
1 x CORSAIR Enthusiast Series CMPSU-850TX 850W ATX12V v2.2 / EPS12V v2.91 SLI Certified CrossFire Ready 80 PLUS Certified Active PFC Power Supply
1 x Intel Core i7-950 Bloomfield 3.06GHz LGA 1366 130W Quad-Core Processor BX80601950
1 x G.SKILL Ripjaws Series 12GB (3 x 4GB) 240-Pin DDR3 SDRAM DDR3 1600 (PC3 12800) Desktop Memory Model F3-12800CL9T-12GBRL
1 x Western Digital Caviar Black WD1002FAEX 1TB 7200 RPM SATA 6.0Gb/s 3.5" Internal Hard Drive -Bare Drive
1 x LITE-ON DVD Burner - Bulk Black SATA Model iHAS124-04 - OEM

Running Windows 7 ultimate 64 SP1

So here is the problem. Random shutdowns, and when I say random I mean seriously random. The shutdown is just that, bam off, like someone yanked the power cord, no BSOD, No errors nothing.
It's not a heat issue, all temps are below 30C at load. ( Partial water cool with an aftermarket cooler, Corsair H50)
CPU is okay, no bent pins or errors.
Not GPU.

Now for the REALLY tricky bit. When it cuts off, if i try to power it back up, it enters an infinite 5 second power cycle. Now if I slightly unplug the 24 pin connector, the power cycle stops and it powers up, but wont post or boot UNLESS.... I clear the CMOS. Checked CMOS battery with multimeter its fine, also this is the second MOBO, first was ASUS Sabertooth, same issues. Also No overclocking.

From here it boots normally, no errors. In fact the only error is that event viewer says the system suddenly lost power and shut off. Which is obvious.

Now here are some tricky bits, I've tested the PSU with a multimeter both voltages and current and they are all nominal. But two things make me think it could be a PSU issue, first that adjusting the 24 pin connector ends the power cycle, and that when it cycles power only comes on to a few areas... the mobo, the HDD, PSU cooling fans, and only 2 case fans ( all these are fans that are directly connected to the PSU along the same modular cable). Not all the fans come on.... might just be that cable is bad? or good?

The other tricky thing is that the RAM is DDR3 1600 but BIOS runs it at 1066 for some reason.... its not on the QVL for the mobo but it came as a bundled pack and I doubt it would have been if it was incompatible. Plus the Ram seems to work fine, and how could it cause shutdowns, and power cycles that require resetting the CMOS?

Now the shut downs are VERY Random. Like I could have one after every boot, or it could take like 36 hours. Sometimes I can go months without issue and then bam every day. I havent changed anything in loooong while. Built this rig in June, worked great till like December, got a new mobo ( the P6) after trouble shooting, worked fine till like last week, shut off like twice. Just re-installed windows again thinking it was software related since I didnt do a fresh install when I put the new board in. Now its happening more frequently rather than less...

I haven't replaced the PSU yet because I'm still not sure that's the issue, and I don't want to buy a new PSU only to have the issue persist. Not trying to rebuild my whole rig, I don't really have the money plus im an Electrical Engineering student, shame I didn't do Computer Engineering... then I might have figured out the issue....

Does anyone here know what the problem is? And if you do think its a PSU issue, do you mind recommending a good modular PSU.... Bare minimum I need is 621 watts... but Id prefer something stronger, rather have power to spare than not enough...

Thanks in advance.

 

[Raiddinn]

Download a program called Memtest86+ and run it while you are sleeping.

I am kinda worried about the RAM thing. I am not the biggest fan of G.Skill RAM, but I don't hear often that RAM from a major manufacturer with a fairly good track record can't be recognized at its regular speed.

Also, you may want to make sure your computer isn't set to restart instead of showing a BSOD. The default is to restart without displaying a BSOD.

To check this setting you right click on computer, properties, advanced system settings, startup and recovery, settings, and see what it says under "Automatically Restart" if it is checked or not. If there is a check you probably want to get rid of it. Then maybe you could see a BSOD if there are any.

PSU - A multimeter won't tell you anything you don't already know. The correction circuits inside the device can detect and fix a problem in output wattage within milliseconds. That is faster than multimeters can read that there is a problem in the first place.

You would need an oscilloscope and a data collector to measure problems exhibited by today's PSUs (ones that will turn on, anyway).

In any event, the PSU maker is Corsair which is generally good and it is from the TX line which is also generally good. I am not too overly worried about that at this moment.

I would try the Memtest86+ thing and if that doesn't come up with errors then try booting into safe mode and trying to crash it however the most reliable way you have found to do so is.

 

[bf2player1978]

i say run memtest, if all is ok id see bout replacing the power supply. i would go with a good 650-750w psu. corsair is good and i have had antec and they are good too..hope its your ram and not psu.

 

[gorignak]

Already Ran memtest during initial troubleshooting and it came back fine... I will run again and definitely check the automatically restart setting.

The one major issue though, i really really really really really mean RANDOM. Months of testing and I have found no way to make it happen, it just does. Randomly. Thats why its been such a tricky issue, there doesnt seem to be any rhyme or reason. I mean the other day I was stress testing for 16 hours with no issues. Today it happened about 3 hours ago.

 

[gorignak]

Memtest came back with no errors at all... sooo probably PSU? waiting for hopefully a blue screen that will tell me the problem now that I turned off autorestart

 

[Raiddinn]

If you are running stress tests just fine and ram tests just fine, that says to me that it is probably a recurring scheduled event causing problems.

If it is, a BSOD should be able to give a clue as to what sort of thing this is.

Hopefully it won't be too long before the next one happens.

 

[gorignak]

If you are running stress tests just fine and ram tests just fine, that says to me that it is probably a recurring scheduled event causing problems

If its a recurring scheduled event then it would show up in even viewer right? Plus wouldn't a fresh install of windows fix that? Also it doesn't explain why it just started happening out of the blue. Literally I changed nothing and it just started happening. Anyway I'll keep my eye out...

 

[Raiddinn]

Not all the scheduled events are handled through the event scheduler.

Many programs have such functionality built into their code and as long as the program is running then it will check every so often if there is an update available for it. That could potentially be a problem you wouldn't see in the event viewer.

It could be something simple that everybody uses like Acrobat Reader or something. Java has an auto updater, and so does many other things people consider basic system components like anti-virus and firewall programs. Even the Windows OS itself has an auto-update function.

Which firewall is it that you use, the basic one that comes with Windows 7 preinstalled?

You could try going through your installed stuff and turning off auto-update settings, but it would be easy to miss and leave some on too, so you could get a false negative reading due to this.

If you could handle the reduced functionality, you could just blow away the whole Windows OS and reinstall it. Then slow roll installing new stuff.

That way you would have next to nothing other than Windows itself that was doing pretty much anything so if it still crashed you could mostly rule out software as the cause of the issue.

If you were going to go that route you would want to first thing go download Microsoft Security Essentials after the OS was installed and then run Windows update over and over with restarts between to get all the patches and then just sit around to see if it crashes.

It could take a pretty long time.

Also, I just wanted to ask, do you have the computer plugged into a surge protector? Just in case it is faulty wiring in your house causing these symptoms. Power problems not even inside the computer could potentially be playing games with your hardware.

If you don't have one or even if you do, you could try moving the computer to a different outlet just for testing sake.

 

[gorignak]

I have tested with 3 different types of surge protectors and in multiple different outlets, i had the same idea. Also most testing was done with installs of win 7 so nothing else was installed.

 

[Raiddinn]

We can go ahead and ignore the autoupdate stuff for now then.

Did you get one of the BSODs or restarts since you unchecked the box?

 

[gorignak]

Update: So was running fine since last post only I had just 1 GPU installed. Decided hey its been fine so far let me install the 2nd again. Put the second in everything was good, used it for a few hours then went to bed. Used it again today for a few hours and BAM shutdown. Only it was a little different today, it didnt power cycle it just shut off, but when I turned it on it seemed to be fine except nothing would boot. Reset CMOS and it booted for minute then the screen went blank had to turn it off and unplug/replug my HDMI cable into my GPU's to get it to boot and show the standard mobo screen and post messages.

So would you say its some weird with having both GPU's in crossfire that is causing problems? No BSODs or error messages just off like the power went out.

 

[gorignak]

Still shutting off occasionally, however now sometimes I can just turn it back on without clearing CMOS. Only 1 GPU installed. I know this kinda died because I haven't posted but like I said it doesn't happen consistently enough to really test in a set time frame.

I'm starting to think it is an issue with my GPU's but I don't want to buy something until know for sure because I have already tried that route and all I have to show for it is less money.

Please doesn't anyone know whats up??