Computer intermittently locks up and requires manual restart.

[Kyle Darden]

Okay guys, my computer randomly freezes and displays the last screenshot of what I was working on before the computer crashes. The computer stays running with all the fans and everything the only thing the screen is frozen and I cannot use mouse/keyboard. I don't think this is a monitor issue because the sound cuts off as soon as it freezes.

This has been happening for a very long time now and I am really stumped. Fry's looked at it and recommended upgrading to windows 10, but it did not fix the problem. I'm not getting a blue screen and I am not finding and useful log files.

What I have tried:
BIOS - most current version

GPU- Tried 2 different GPU, both with current driver installed
Tried a different pci slot and connection to PSU
32C idle/70C under load

RAM-memtest has no errors, I have cycled RAM in and out to see if I have a bad stick

CPU- 11C idle/54C (while running prime95)

OS-reformatted at least a dozen times. I have tried win7 and 10.

I have ran virus/malware scans and have no infections. I'm pretty safe online I really don't think this would be the problem any way.

Throw out ideas at this point I'll try anything honestly.

System Specs:
AMD Fx-9590
corsair H100 cooling
ASUS crosshair V formula-z
EVGA gtx 660 (currently installed)
ASUS gtx 970 (currently getting RMAd)
4x2g corsair 1333mhz
250gb samsung ssd
corsair 750W PSU
corsair 780t graphite series case

TLDR-computer freezes and I've ran out of ideas.

 

[Lutfij]

RAM - with memtest86, did you run the test for at least 10 passes before ruling out faulty ram?

OS - What version of Windows 7 were you using(have tried)?

PSU - Which series of Corsair PSU are you using?

GPU - Did you use DDU to remove your prior GPU drivers and reinstalled the latest revision? Inf act have you tried reverting to a much older version of Nvidia's GPU drivers?

Freeze - What sort of apps do you use to laod the system and cause the freeze? Or is it happening in desktop when idle as well?

 

[Kyle Darden]

Hey thanks for the reply!

I ran the memory test on windows 7 prior to reformatting, but I cannot recall how many passes I did. I'm installing Memtest86 now and I'll see what comes up from the 10 passes. I did swap the memory out 2 sticks in at a time and the problem persisted.

I was using windows 7 pro 64bit, I am now using windows 10 home or basic 64 bit

Model number is CMPSU-750TX, I believe it is the TX series.

I have not tried using DDU to remove the prior GPU drivers, however prior to yesterday the only card that had been in the system was the 970. Do you think trying an older-version of drivers would be a possible solution? I use nvidia geforce experience app to keep my drivers up to date. (I am not familiar with DDU)

I notice the computer freezing more when I'm playing games, but it has also froze watching youtube videos as well as just sitting on the desktop. I'm not sure if the games are causing the crash or if that's just what I am usually doing so its most of the PC's up time.

I also just reset the BIOS to default settings and reset the CMOS (unsure if these are the same thing, I forget) I was running prime95, Valley Benchmark and a game and the computer is holding up still. I'm wondering if this is because the temperature has gone down out side quite a bit since I was last using the PC.
'

 

[Lutfij]

You're welcome 😉

How old is the TX unit you're working with? I'm going to guess that would be a good place to start by swapping it out for a unit that is newer and is branded of at least 600W.

Yes, irrespective of what brand/camp of GPU you use it's customary to remove your old drivers before dropping in an older or newer GPU. I suggested the older drivers because the GTX660 is already 3 generations old since we're on the GTX1060 from Nvidia. True that newer drivers help address game compatibility but you're also going to need to realize that Nvidia are looking to phase out older GPU's from their driver support list.

Speaking of games, what are the titles?

HANG ON A MINUTE! Are you on 4 sticks of Corsair ram at 1333MHz? Are they all identical or were they mixed and matched to give you a cumulative 8GB's? Mind sharing your ram's specs?

I just noticed your cooling solution for the CPU, what are your temps when in BIOS?

If under load subjection the system doesn't freeze then it can be a driver issue.

DDU = Display Driver Uninstaller.

 

[Kyle Darden]

I've been leaning towards testing a new PSU, but I'm not wanting to spend money if its not the problem 🙁 I should just suck it up I think. I got this PSU in 2010 it has served its uses.

I can try to install an old driver after this memtest.

Really all I'm playing now is WoW, but I had had the same problem with DOTA 2 and Skyrim(I think).

The ram, while bought in two 2x2g sets are all identical 1333 mhz

I can double check the BIOS temps, but I think they are similar to what CPUID is displaying.

The computer sometimes runs games perfect, other times it will freeze up. It's too hit or miss to really say if there freezing under load or not.

 

[Lutfij]

You can always borrow a PSU from your neighbor/friend and see if the system works without a freeze though they wattage and reliability need to be present otherwise you can damage everything or only the PSU in the process.

Mind sharing a link to the ram kit? Provided they are identical...while the temps you say are fine, we are yet to learn what the temps are. Mind sharing that too?

From that last line, it can be anywhere between drivers, ram or PSU.

*Those are esports game titles(except for Skyrim) so you're not taxing your system to any high degree. Just to ask out of curiosity, which BIOS version are you on?

 

[Kyle Darden]

I can try to use my brothers PSU, I think his is over 600W.

I'm pretty sure this is the ram kit,
https://www.newegg.com/Product/Product.aspx?Item=N82E16820145251&ignorebbr=1&nm_mc=KNC-GoogleAdwords-PC&cm_mmc=KNC-GoogleAdwords-PC-_-pla-_-Memory+%28Desktop+Memory%29-_-N82E16820145251&gclid=CNrVip2T8dECFQSBaQodULAArg&gclsrc=aw.ds

So using CPUID, the temps on CPU range from 11C to 54C (when running prime95), but in BIOS I'm getting cpu temp at 35C. GPU is reading from 32C to 70C in CPUID.

Current BIOS version is 2210.

Memtest is running now.

 

[Lutfij]

Don't just drop in an XYZ PSU just because it's got the aforementioned wattage! What brand/make and model is it and how old is it? What do you mean pretty sure, is the linked kit the same as what you have now?

There's no version 2210 on Asus's support site, only 2201. You may want to reflash the BIOS. The temps look off since the CPU you have is rated at 220W and that should leave you above 60C provided your AIO is on full blast. What are you ambient temps?

 

[Kyle Darden]

Oh of course not, I'll make sure it will be compatible before trying to place his PSU in my computer. I don't remember the exact model right now, but I can post it on here tomorrow.

That RAM I'm pretty sure is mine, it looks identical and is 1333 2gb which mine are. The timing on mine are also 9-9-9-24. (EDIT: my RAM is PC3-10600 DDR3 1333mhz /9-9-9-24/)

Forgive me, it is getting late and I mistyped. My current BIOS is indeed 2201, not 2210.

"You may want to reflash the BIOS. The temps look off since the CPU you have is rated at 220W and that should leave you above 60C provided your AIO is on full blast. What are you ambient temps?"

Can you clarify this a bit more? How are you relating the wattage to temp/I'm not sure what AIO is, but I'm assuming you mean the cooling system. The temp in the room is maybe 15-18C?

 

[Lutfij]

The processor is rated at 220W TDP which means tha it's heat dump will be mush more than say an i7-4770K. The temps are looking off since reviewers when subjecting the same processor under load scenario's(like Prime95) found a lot of heat being dumped into the system/loop/case/environment. AIO = All-In-One which is a title we folks have coined the pre-assembled closed loop coolers from the likes of Asetek.

Further reading.

At those ambient's, you're fine.

 

[Kyle Darden]

So ran memtest64 overnight and came back with about 60 errors. I didn't think to write it down, but most of the error codes were not the expected codes.

I'm assuming it is safe to say the RAM is bad?

 

[Lutfij]

Now comes the hard part, you need to test with each stick of ram, each for 10 passes to rule out your slots haven't gone bad.

 

[Kyle Darden]

That does sound to be the hard part.. if I try known good RAM in the slots and the problem does not persist, could that rule out the slots as well?

Or is it a best idea to run the memtest first? Also, should I try the same stick in every slot or how would you position it?


I really appreciate you helping out also.

 

[Lutfij]

memtest ensures your board is not the culprit or the rams themselves. You can distinguish them and send off for an RMA confidently. Try the slots you normally have them in which should be slots 2 and 4 or it could be 1 and 3(check the manual to see which slots to populate if working with 2 sticks of ram for dual channel.

Nah just helping out man, no worries!

 

[Kyle Darden]

So I'm running a 10x pass on with one stick in.. Should I run 10 pass for every slot or just switch the sticks out in one slot?

I have the stick in the proper slot as according to the manual for 1 stick of ram.

 

[Kyle Darden]

Also, you think the RAM/DIMM slots could be the issue even though the computer loads to OS. Most articles I have read are just attempting to get the PC to post.

 

[Kyle Darden]

I have ran 3 out of 4 sticks of ram in the same slot.
1st stick 4 passes- 0 errors
2nd stick 4 passes- 7 errors
3rd stick 4 passes- 1 error

All of the errors are on the hammering pins section of the memtest. I am going to run the 4th stick overnight.

I also tried two new sticks of ram that are known to be working and the issue is persisting. Since the computer froze with the new sticks, I am wondering if one of those slots are bad. After running this 4th stick, I am going to run the 1st stick in the slot that I am thinking is not working properly just to see if that slot is throwing errors with RAM that has been scanned with no errors.

I am looking into the PSU being bad at this point. After looking at my voltages in CPUID HWMonitor, I am noticing that the +3.3V is steady at 3.132V at idle. Is this reason for concern? I know this is under 5% of ideal value, but just barely. I am going to look around to see if any shops will test the voltages with a multi-meter or buy one myself.

 

[Lutfij]

Hmm, some substantial information since my last outing here. Have you tried each stick on the first slot? I asked since you mentioned sticks 1 through 3 placed in the ram slot closest to the CPU socket. You should also isolate the kits you used. Regarding the new sticks of ram? What are they? As long as they aren't mentioned we're left guessing/assuming.

Pertaining to your PSU, isn't there any way you can borrow a PSU from your friend/neighbor of similar wattage but ofc off a reliable branding?

 

[Kyle Darden]

So I tested all the RAM sticks so far in slot A2 which is second from the processor, but the recommended DIMM for 1 stick by the user manual. I wrote a number on the original 4 sicks of ram 1,4 no errors 2,3 errors.

The RAM I tried was Corsair Vengence 2x8GB 1866mhz in slot A2 B2 as said in the manual.

I guess I could ask my brother to use his, but we live an hour away so I wouldnt be able to until this weekend.

 

[Lutfij]

So you mean to say one slot from each channel off your motherboard is reporting errors? If so then the board is ...on it's way out.

 

[Kyle Darden]

Well I was thinking that was the issue, but I'm on pass 6/10 on the slot that I thought was the problem and I have no errors yet. If the Vengeance RAM was in slots A2 and B2 and there was a lock up of the pc I would think one of this slots would be causing the error. If no error comes up on either of the slots, I am once again very confused as to what the problem is.

After this I think I'll run 10 passes with the Vengeance RAM in slots A2/B2. All of the initial errors from when I had all four RAM installed could be because the board isn't liking ever slot filled, maybe?

 

[Kyle Darden]

No errors on the 10 passes of stick 4 in slot B2, placed sticks 1 and 4 (both scanned and found no errors) in slots A1 B1, computer still freezes. I haven't tried the Vengeance ram yet, I'm thinking of taking it into a shop tomorrow.

 

[Lutfij]

If the dual channel setup gives you an error, the issue is with your motherboard slots since each stick of ram cooperates with the system, that is, when we;re looking at the 4x2GB Corsair XMS3 loadout.

 

[Kyle Darden]

You think I should try running just one of the Vengeance 4g stick and see how it works with only 1 stick of installed? I think I'll RMA motherboard any way I've thought it was the problem the whole time just hard to troubleshoot.

 

[Lutfij]

First you say the Vengeance kit is 2x8GB now you're saying a single 4GB of the same nomenclature, are you working with a 4x4GB kit of Vengeance memory? Your board can support a maximum of 32GB of ram meaning with an even distribution across all memory channels you can have a single stick of 8GB of ram per slot.

Going along the information of the 2x8GB kit, you can try with a single 8GB stick of ram and populate it in A2 and fire up your system.