Gaming rig shutting off during stress test/BF4.

I'll start off with my entire rig. Everything has been purchased within the last year, except the processor, which is a few years older.

i7 920 D0
Intel DX58OG Mobo(With latest BIOS)
CoolerMaster V6GT
G.Skill Ripjawz 12GB 1333(3x4GB)
Galaxy GTX 760
Corsair CX750M PSU
OCZ Agility3 120GB SSD
WD 320GB HDD
320GB External HDD
Creative Sound Blaster Recon3D Fatal1ty Pro
DVD-RW/CD-RW
(5) 120MM LED Case Fans
(2) 140MM LED Case Fans
(3) Cold-Cathode Lights
All wrapped in a Corsair C70, running Windows 7 Ultimate 64 bit(real licensed copy).

It will crash during Prime95 Blend Test, within 30 seconds of running it. The Small FFT option I ran for 5 minutes without it crashing. I can run this longer to see if it crashes if needed, but not sure if this means anything. It also started crashing during very intense moments of Battlefield 4(large close quarter firefights). When it does crash, my computer will just shut off. No BSOD, nothing in the event log, and not overheating. I always keep Core Temp running, and it's never gone above 65*C on any core, even right before it crashes. It literally just powers down, spend about 3 seconds off, and then turns itself back on. Temps are typically 30-40*C at idle, 50-65*C while gaming. I did try a live boot into Ubuntu, and installed/ran mPrime(Prime95 for Linux), and it crashed there as well. Unsure of temps since I barely know Linux and couldn't get my LMsensors working. So Windows instability can be ruled out. Also ran MemTest86+ for 18 hours overnight, passed 8 full tests no errors. I know this can't rule RAM out entirely, but I don't believe it to be a RAM issue either.

Processor is not overclocked, and BIOS is set to default, minus fast-boot options. System is connected to a new surge protector with a high Joules rating. This problem has occurred in 3 different houses. I installed a new PSU(Antec NeoEco 620W) a year ago, and it continued to happen, though just during stress test, not gaming, so I ignored it. I just installed the Corsair CX750M yesterday, since I thought my new 760 was drawing too much power with BF4, and same thing. So I do not believe this to be a PSU issue, as I highly doubt I received two DOA PSU's(though knowing Newegg, this isn't all that unlikely). I just reseated the CPU with artic silver 5 yesterday as well, when I redid my entire fan configuration.

I will note, my patience for my personal computer problems is VERY thing. I run a computer repair store, and deal with silly issues day in/day out. Typically home/office computers, very rarely gaming systems, but the last thing I want to do is come home to work on my own. With that said, I ordered an i7 930 off eBay this morning to eliminate my processor being the culprit. I completely hope it is the issue, so this can stop. It's the last piece of my system that wasn't purchased brand new since the beginning of 2013, so I figured I'd start there with ordering parts.

I have no idea what's going on, and it's got me pretty confused. Any ideas/suggestions to try? I'm nervous that when I plug the 930 in, it's going to do the same thing. At that point, I would have replaced everything in my computer at least once since my initial rig that was doing this.

 

It may sound strange, but is that external 320GB drive connected via USB and plugged in during the stress test?

 

Strange is where the best solutions come from! Unfortunately that was not one. I unplugged all USB devices, except mouse/keyboard, and it still crashed on the blend test. Though it took somewhat longer to crash than usual. Normally it will crash by the time it heats the CPU up to like 61(Which is under what it gets during gaming, to eliminate the overheat possibility), but this time it made it to 65.

The i7 930 comes in tomorrow, I'll be installing it after work and hoping that fixes this issue once and for all. If not, mobo is the only thing I can think. Should still be under warranty, as it's only a year old.

 

Have you thought about perhaps getting an AIO (All In One) water cooling kit, for the CPU? That is my best guess at the moment on your issue. Something like this: http://www.newegg.com/Product/Product.aspx?Item=N82E16835181031

I'm using the H90: http://www.newegg.com/Product/Product.aspx?Item=N82E16835181036
Though, I'm not overclocking.

 

My first thoughts are heat, malware/virus activity, PSU or other unknown at this point in time, hardware/software issue. I'd start by ruling out heat. I assume your heat-sink/fan is properly installed and you've used a quality thermal compound. What are your CPU temperatures before your system crashes? Are you logging temperatures and voltages? If not do this you can use something like CPUZ at the very least to monitor on screen what's happening.

I'm sure you've already done this but if not run a full/complete scan for malware/virues use your A/V, use Kaspersky's boot disc if needed, use Malwarebytes etc.

Assuming it isn't heat or malicious software. I'd disconnect and remove all non essential items from your machine, USB devices, internal lights, drives and other hardware, start up with just the bare basics needed to get operational e.g. CPU, RAM, GPU and a HDD and go from there. If you still have issues swap out the GPU, PSU and even the RAM. Eventually you'll get a hit on what's causing the issue.

Good Luck

 

Heatsink is installed with Artic Silver 5, with the vertical line method as recommended by Artic.

Temperatures sit around 35-40*C idle, 50-60* while gaming on BF4, and the highest I've seen it get on this setup is 65* during 1 blend test of Prime95 that lasted longer than normal.

Voltages I never thought of paying attention to. Is it possible for the CPU to be spiking in voltage and causing this? The overclocking section of my BIOS is untouched, and still set to factory settings. I currently use CoreTemp to monitor my temps(it does show voltages too), but I don't know if it has a logging feature. Is there a program that will actually keep a live running log file of temps and voltages, similar to how you can set FRAPS to record FPS in an excel sheet?

Viruses are definitely not my issue. I've been doing virus removal since I was 10/11(When I discovered that a music file isn't a .exe), and currently run the second store of a local IT company. My computer is spotless as far as that goes, since I'd be the laughing stock of the company if I was to get infected. My girlfriends computer on the other hand rolleyes

Last night I tried unplugging all USB devices, and running the Prime95 blend test. Same results, except it ran for a couple minutes, as opposed to 10-15 seconds, before crashing. Now I still did have all my fans plugged in, along with my sound card, and CCL's.

GPU I want to say is definitely not it, as this was happening with 3 separate video cards, two of which are in other systems performing flawlessly. The latest card was brand new.

While I'm not ruling RAM out completely, 18 hours on MemTest86+ with not 1 error leads me to believe its okay.

CPU is where my money is, and I should have an i7 930 showing up to my shop any minute now. I'll be swapping that into the system, and immediately be running a Prime95 blend test. IF it crashes then, I'll be trying a single new RAM stick. If it crashes then, look's like I'll be RMA'ing my motherboard.

Let's hope the new CPU fixes it, since that is literally the only piece in my system that isn't >1 year old, and purchased brand new.

I appreciate the well thought out response and input!

I do have plans to overclock, and while I love the look of liquid cooling, I am not personally for it. My maintenance on my computer just involves bringing it down to my shop, and spraying the dust out with my compressor. Takes longer for the compressor to charge up than it does for me to actually clean it. I would be the guy that lets the coolant stay in the system for far too long, which I have seen first hand the problems that causes. Had a girl buy a customer gaming rig offline, with liquid cooling. She had no idea you were supposed to flush the coolant out every so often, and it ended up clogging the pump, breaking a seal right where it mounts to the CPU, and leaking all over her mobo. Fried her mobo, cpu, and GPU. That would definitely be me when I put off flushing the system for months.

 

Okay, just got the i7 930 installed, but problem didn't go away :cry

It doesn't fail instantly on the stress test like the 920 was doing, but it still crashed. So I tried unplugging everything. Unplugged all fans, except the two attached to the cpu cooler, unplugged all lights, the CD drive, my backup drive, took out my sound card. Literally everything but the bare essentials(CPU, GPU, Boot SSD, RAM). Still crashed. So I don't think this is a power issue at all.

Though I will note, I had Prime95 Blend running for about 10 minutes, and was on the phone with a coworker getting excited that it may just be stable. Prime95 started to list that it was starting a test for each worker/core, and once it hit the last core(Which was Worker 1, the top worker, since it started from the bottom and went up), it shut off. The first time it crashed, it was running for like 5 minutes, I went to open Chrome during the test, and it crashed. I'm at a loss as to what this can be, and it's really starting to aggravate me.

I'm thinking mobo issues, which sucks because RMA takes forever to get back. I'm going to test it playing BF4 to see if it crashes during that anymore. If it doesn't, all is well and dandy since I could care less about stress tests, just if BF4 is stable. Unfortunately, BF4 causes this to happen so randomly, it's tough to test it just by playing the game.

Just wanted to clarify, still no BSOD, or anything in the event log. Just a sudden power loss and restart.

 

Humor me for a few hours.. what do you have to lose? You've everything to gain!

Remove your current hard drive, take a spare mechanical drive and install a clean copy of windows on it and retest.. see what happens. If you get the same results, set BIOS to default values. Rerun the test again and see what happens. Same thing.. try with just your CPU, GPU, 1 stick of RAM and remove everything else, lights, fans, sound cards etc everything that isn't critical to make your computer work. If you get the same results. I know you already did this but do it fresh to ensure nothing is missed.

Swap out PSU for known good part and test again.
Same result, swap out GPU for known good part and test again
Same result, swap out RAM, one stick at a time and retest each time.
Same result, swap out M/B for known good part.

Try the above one step at a time.. this takes time, it's boring.. it's not BF4 but it will allow you to get to the cause of your computer problems without throwing more expensive hardware at your computer.

I hope you get your issues resolved :major

 

Also, are ye getting any minidump files under c:\windows\minidump? If so, zip them up and attach them here so we can analyze them.

 

Better still, get us some additional data to check through as well, carefully follow the instructions here, you may need to split the resulting data across 2-3 zips to attach them here.

Run HWiNFO64 sensors-only and paste up some screenshots from it, it creates a long page, scroll down to get all the data showing. Don't run any tests the first time, just get some baseline info to begin with.

 

Oh I have no issue spending hours working on a computer. I mean shit, I do it 45 hours a week, every week. This is far from boring as well. I'm one of the lucky people to say I love my job, and as frustrating it is to work on these phantom problems, when you finally figure it out and get it workin, it's so satisfying!

I have no idea why I didn't try this sooner, but I booted into Safemode to remove a couple of SuperAntiSpyware registry entries that were giving me my only Event in the event log at startup, and decided to run a stress test in safe mode. I left it running in my shop, and went to run an errand, which took at least a half hour, if not longer. When I came back, the computer was still doing the stress test. I wanted to :banghead I don't know why I didn't try this much much earlier, as this shows there is clearly a driver/OS issue.

I'm currently working on it now, just ran a SFC scan, which did repair things(Will gladly provide the CBS file if wanted). I'm going to run chkdsk /f and /r, and then boot in with a diagnostic startup, and see if it crashes then. If it does, looks like it's reload time :cry and redownloading and installing over 300GB of the games I like to keep installed. Luckily I don't play many single player games, so not many gamesaves to worry about it.

No minidump files to be had. As I said, this isn't your typical BSOD crash. This thing just POOF, power off. No logs of it besides the "Window shutdown or loss power unexpectedly."

Before I bounce to your suggestions, I'm going to see if I repaired the issue within Windows that I mentioned after the first quote. Since today's my day off, I'm going to end up reloading it if the repairs don't work. If after that it's still crashing, I'll be posting those logs for you!



Thank you all for your help and suggestions so far. :major

 

Okay so I've got it narrowed down to services. Ran another stress test in safe mode while I showered, no crash. Booted into Windows in Diagnostic mode, no crash. Booted in with all services enabled, no startup items, crash. Closed all services EXCEPT Microsoft ones, no crash. Enabled my usual startup items-Which consist of 5 things. The two GeForce Experience programs, DisplayFusion, Sound Blaster Control Panel, and my mouse application. No crash then. So it has to be one of my services that are running. I understand that enabling one at a time will solve my issue, but maybe these logs will say something. I will be working on one at a time in the meantime.

http://i.imgur.com/zDFhzt9.jpg
http://i.imgur.com/vy0BwIF.jpg

The Performance Monitoring test is in the second zip as well.

Satrow, quick question. That first link with instructions had me download two things, but only run one. Does that program it had me run require the first program, or did I miss something by not running that first program(A).

And just to clarify, I purposely have no anti-virus, and UAC off. I have not ran an a/v software since I built my first computer in high school. Personal choice, I know the risks involved.

 

Forgot to actually insert the attachments.

View attachment jcgriff2.zip

View attachment Windows7_Vista_jcgriff2.zip

 

:banghead:banghead:banghead:banghead:banghead:banghead:banghead:banghead:banghead:banghead:banghead

Fresh reload using a Win 7 Ultimate SP1 disc. Ran Windows updates, got all my hardware installed, base programs that I'm definitely going to have on the system, all the latest drivers, Java, Adobe. Formatted the drives prior to reload. Aaannnddddddd it still crashes:cry. Crashed once, and I had the page file set to system auto managed on my backup drive(to save writes to the SSD), and on restart it said crash dump initialization failed. So I went ahead and put it on system auto managed for both the ssd and backup drive, ran stress test again. Made it to test 5 on all 8 workers, then poof crashed, and no minidump/dump/BSOD/events.

I did image my drives before doing all of this, so it wasn't a complete waste 6 hours backing it up and reloading, since I can revert back to where I was fairly easily.

Hopefully everything I posted earlier says something to someone. I'm going to leave it on the fresh reload for now.

Thanks to everyone actually reading these posts, I know I ramble a bit :-o

 

Sorry, Skeetzy, I'm not seeing anything definitive there, very little in the way of clues at all.

Get the hardware down to the bare minimum, turn off anything unused in the BIOS (RAID controller, Serial/Parallel ports, sound card - even USB3) ensure your internal drive(s) are connected to the Intel SATA controller, connect any external USB to USB2, if you really must connect it.

Remove the Creative card during testing - really, if you can live without it for a few days or so, pull it!

Ensure you have the latest firmware installed for the OCZ boot drive and that TRIM is active. Check and double-check all cables/connections internally, replace them with new, if you can. Disconnect most of the internal fans (not CPU/GPU )and leave the case open with a household fan pointing in there.

Test ...

 

I've got it stress testing in safe mode right now, no idea if it's still going, since I'm downstairs at my shop. It had already been running longer while I was getting ready for work, than any other time. Might have a chance to run up and check it around lunch time. It crashed last night with everything enabled, and when I rebooted with it running only Microsoft services, no startup items or non-MS services. Which I find peculiar, because it seems to be stable in Safe Mode and during Diagnostic Startup.

I'll double check the firmware on the OCZ drive, as that is something I don't believe I've updated during this frenzy. TRIM is active, CDisk confirmed that for me last night. All my cables are just as new as everything else.

I've got a plan of attack for tonight, which will go 1 of 2 ways, depending on how the safe mode stress test pans out. If it's still running and "stable" in safe mode, then I'm chalking this up as one of Microsoft's services is what's messing with my computer, since the only time it seems stable is when they're also disabled. At that point, I'll be reloading it to Windows 8.1 and trying my luck there.

If it crashes in Windows 8.1 OR the safe mode stress test has failed, than I'll be leaning more on the hardware side of the fence. At that point, I'll be updating the SSD firmware if I haven't already, and disconnecting everything off the computer. I'll be running it with the true bare minimum: 1 brand new stick of RAM, onboard video(so no GPU), CPU+cooler, and the cd drive.

Once again, thank you very much for the suggestions and confirmation that I've got a mystery problem. I really wish I could just get a BSOD during this, or some point towards what exactly is making it crash. At this point, literally everything in the computer has been replaced at least once with brand new parts. I'm reaaallyyy starting to think motherboard, but the only way I'll be able to test that is by RMA'ing this one. That's an absolute last resort, as I really don't want to go without a computer for upwards of a month.

 

Just wanted to update you guys as to where I've gotten today with it.

Did the bare bones boot, in Diagnostic mode, safe mode, and normal mode. All 3 modes failed the Prime95 test, as I anticipated. Work got busy, so didn't get to really even work on it till now. But noticed when I was taking my RAM out to put a brand new stick in to test, that I was running in Channel B. Now not a big deal, but it clicked in my head I did this on purpose. So I plugged the new stick into Slot 1 Channel A aaaannnndddd 3 beeps+LED POST Code. :dancer:dancer:dancer

I almost cried with tears of joy, as this was the first error I've gotten, and finally points to something. I popped the RAM stick over to Channel B Slot 1, booted no problem. So finally nailed it down that theres bad RAM slots on the motherboard. Ran a Prime95 test with the brand new stick, and it crashes. So I'm like 95% certain it's a bad motherboard, especially after seeing a few reports of similar problems with the RAM slots on this motherboard.

Contacted Intel, and super lucky me, my motherboards in warranty until the 27th of this month. I've got it imaging back to my clone I made before this whole reloading fiasco, and am going to enjoy one more night of BF4 before I RMA the motherboard tomorrow. More than likely going to RMA the RAM and processor while I'm at it, just in case. Better safe than sorry.

Well with that, we should have this problem solved. I will know in a couple weeks.

 

Glad to see a resolution to your issue!

 

Thansk for the help! But, fights not over yet. With my luck, it's still going to crash after everything rolleyes

Hoping for the best though. Good thing I ordered the parts to fix my Xbox's cd tray, cause it's going to be getting a lot of use the next two weeks lol.

 

New motherboard came in today, just had Prime95 blend running for 4 hours, no crash!!!! :dancer:dancer:dancer:dancer:dancer Highest core temp was 66*C, and thats with fresh Artic Silver 5, so hopefully shed a few degrees once its cured.

Word of wise, keep your plastic CPU socket covers. Intel made me wait for them to ship me a socket cover, before I shipped them my motherboard. Still got my motherboard back when I anticipated, so wasn't too bad, but still little silly. They also sent me JUST the socket cap in the same box that my motherboard came back in. Total waste of resources, but whatever it was free.

Now on to overclocking this 930! Thanks for all the help guys!