Support NeoGAF

[mkenyon]

$100 says it's the memory or the PSU, but my money is on the PSU. RAM typically would give you a BSOD, the straight shut down absolutely sounds like a PSU issue.

 

[pestul]

It's probably the PSU for instant shutdowns..

Did you ever push you cpu overclocking / overvolting really hard? You could have caused damage to the chip. Have a close look at all the capacitors on your motherboard too.

 

[Resilient]

All, thank you for your responses. Overwhelming few suggest that it's a PSU. Here is my plan of attack.

1. Order a new PSU. If anybody has any recommendations, please let me know. I'd prefer full modular. I already have my eye on a few.
2. While I wait for it (and before I install it) - test/reconfirm the following:

a. Ensure all cables are firmly plugged in (did this the 1st time, will reconfirm).
b. Swap SATA ports around.
c. Remove WiFi PCI-e card

I'm leaving the OC on because I feel that it's stable, and not a CPU issue at all tbh. If the solution is to turn off the OC after all the above, then that means the CPU is at fault. So I'll attempt all these fixes first, and if the problem persists, with those eliminated, I'll switch my views to faulty CPU/RAM with an extremely bizarre problem that never showed up in stress testing/everyday use. It seems most unlikely.

3. If the problem reappears, we can assume 2 didn't do anything. So here, I'll install a new PSU, and use everything like normal. Hopefully this point is the solution.
4. If it still persists. I'll then start the process of item elimination. This could take weeks, but I'll start it, and report back as I move through each piece.

Thanks again the for replies. I've responded to a few of these as I have some questions, and also for reference sake, for anybody else who may have a similar problem in the future.

It happened during read/write operations sometimes.

You shouldn't go into details, just swap SATA ports on your motherboard, 20 seconds. It should have no effect if it is not the solution anyway.

OK. Will try this.

Give everything a good dusting with some air spray. It sounds very much like a heat problem. PSU would be my second guess.

I did a massive clean of it the first time, and it wasn't extremely dusty when I anyway. then the probelm happened again,

Are you using XMP settings for your RAM? Maybe try the other two alternating slots to rule that out. Not sure how Skylake controls memory, but my 5820k has a less than stellar IMC and I can't use the XMP profile on my 3200Mhz RAM, PC would freeze or just shut down unexpectedly so I slowed the speed down and tightened timings to compensate.

No OC on the RAM because of this issue.

I've fixed 2 or 3 computers where the issue was bad ram and memtest hadn't found anything. I have a feeling that it probably isn't the ram, but I would try using your pc normally and playing games for a bit with each stick of ram on their own. I know you ran memtest like this. Maybe you did this too, but I feel it's worth a shot before you spend any money.

I will try after an attempt with a new PSU.

Go to the bare minimum that you need to run the pc.

First thing, stop with the overclock shit if you've got problems with your pc.

Update your bios. I don't see a January update for your bios here.
https://www.asus.com/us/Motherboards/Z170-PRO/HelpDesk_Download/

http://dlcdnet.asus.com/pub/ASUS/mb...0318.pdf?_ga=1.239210820.631535397.1479846624
Your power supply isn't listed as supported. Usually isn't a problem but something to look into.

http://dlcdnet.asus.com/pub/ASUS/mb...0316.pdf?_ga=1.211423314.631535397.1479846624
Your ram isn't listed as supported either. Try running it at 2133 instead of 2400?

Beyond that...

Wifi card, 1TB drive, printer, whatever, all out. Run something that you know makes the pc fuck up. Does it still do it with the bare minimum? Then it's got to be something in there.

It's a matter of elimination. Just start pulling things one by one till your down to the absolute minimum.

Agreed I need to eliminate everything. The impatience in me will start doing this once I rule out faulty PSU.

Instead of tearing stuff out of your system there's still some basic tests you can do. Fire up OCCT and do the PSU test, it basically runs Linpack and Furmark to stress both the CPU and GPU.

I'll try this also. I want to replicate the issue so that there is something I can do to confirm I've solved it, directly.

Does it crash at stock speed also?

Yep, it did. That's what started this.

Undo any and every overclock you have on CPU, GPU or memory. Change nothing else for days and test.

If the problem goes away, re-enable one of the above overclocks and test until you find the culprit. If the issue continues after disabling overclocks, almost certainly this is either a PSU issue or a heat problem with your CPU. Run utilities to check core temps on GPU and CPUs. Look into a new PSU.

My guess is if you OC'd your CPU in the BIOS/UEFI you had to obviously adjust the core clock values. I would gather this isn't playing well with the set CPU core voltage levels. Most modern overclockable CPUs will actually stretch and OC well on Auto set core cpu voltage settings. Keep voltage settings stock and Auto and play with just CPU core values until you find the right balance. But only do this after testing with your OC turned off and the CPU running at stock speeds.

I'm going to look into a new PSU and rule that out. I disabled Auto as it was upping the voltage too high, I have them at about 1.275 or 1.8 can't remember, but the OC is stable as stress tests never threw up any errors or crashes (once the OC was at the right voltage).

I would suggest running something like MSI Afterburner and monitoring the temps and cpu/gpu usage while playing. See what happens when it crashes. Although if it was from temps the PC should shut down completely and not freeze.


Someone mentioned having similar problems with their SATA port. I know on my ASUS motherboard there are marvel and intel Sata controllers. The marvel ones are shit and although they say they are rated for 6gb/s they are not getting near that and are very sporadic in speed. I had a problem with a super slow SSD until I realized it was the marvel sata ports being shit and when I switched to the intel ports they worked flawlessly.


Sounds like a motherboard or power issue.

I'm going to try this before the PSU. Thanks.

If it were the RAM you'd get the freezes doing stuff other than gaming, too. My PC used to exhibit this behavior after I had overclocked it with unstable settings (using Intel Extreme Tuning or however the tool is called), and the problems went away when I took the OC down 100-200 mhz. I suggest trying without OC for a while. And this isn't a temperature issue, it's a power issue.

Yep, this is what is driving me crazy. However, the freezing happened before I had any OC on.

If it were a driver issue, the PC wouldn't sometimes work an hour and sometimes four. This PC freezes because it either has some component physically failing or because he's overclocking too high with inadequate power settings/supply, or perhaps even a combination of both.

Yep, agreed. I'm going to order a new PSU. Recommendations by any chance?

When it freezes, is the hd led on? It could be bad sata ports( you can swap your drives around to check), or just a bad mobo, or the psu. If you have a spare psu, swap it in to check.

good point. I can't remember, when it happens next I'll check. I'm going to swap the SATA ports before testing a new PSU (and see if the error appears).

$100 says it's the memory or the PSU, but my money is on the PSU. RAM typically would give you a BSOD, the straight shut down absolutely sounds like a PSU issue.

Agreed. That's what I had found from majority of my research before this thread.

 

[neurosisxeno]

I would say probably PSU or VRAM issue. PSU is obviously cheaper to replace so I recommend starting there.

 

[Mareg]

I would say probably PSU or VRAM issue. PSU is obviously cheaper to replace so I recommend starting there.

A quality PSU like an EVGA G2/G3 is not cheaper than ram. I too, have run into my fare share of anomalies due to cheap PSU that would drop power out of nowhere.

I would say in the PC world, PSU are the single most important part that is being neglected. A faulty PSU will destroy your stability. No amount of power phase will save a board from a cut in current.

 

[Netherscourge]

My rough guesstimate - your GPU overheating.

Either the HSF/paste is not secured properly on it or the PC tower case is not ventilated enough.

EDIT - if it happens with a different GPU too, then it might be the CPU's HSF/paste. Basically, anything that overheats will trigger the thermal-throttling and/or a hard-emergency shutdown.

 

[Resilient]

My rough guesstimate - your GPU overheating.

Either the HSF/paste is not secured properly on it or the PC tower case is not ventilated enough.

EDIT - if it happens with a different GPU too, then it might be the CPU's HSF/paste. Basically, anything that overheats will trigger the thermal-throttling and/or a hard-emergency shutdown.

the thing is, it's the PC freezing, not shutting down. It's still on with fans spinning. everything is frozen though.

 

[LilJoka]

If you get an audio loop, it generally means the GPU driver crashed. That could be due to a faulty factory GPU overclock.

Try lowering the core and memory clocks by 100Mhz with MSI Afterburner and see if the problem still occurs.

If fans are still spinning and it didn't instant shutdown/reboot then I very much doubt it's the PSU.

Drop the XMP profile and CPU OC while you test.

Corsair RMx650 is a solid PSU, if you consider that route.

 

[Resilient]

If you get an audio loop, it generally means the GPU driver crashed. That could be due to a faulty factory GPU overclock.

Try lowering the core and memory clocks by 100Mhz with MSI Afterburner and see if the problem still occurs.

If fans are still spinning and it didn't instant shutdown/reboot then I very much doubt it's the PSU.

Drop the XMP profile and CPU OC while you test.

Corsair RMx650 is a solid PSU, if you consider that route.

Hey LilJoka. you actually helped me when I was first experiencing this problem with the faulty RAM

OK, I'll try that tactic. But, the same crashing problem occurred with my old GPU. Does that seem like more than coincidence to you? I had read that this is symptomatic of the PSU lapsing in the voltage it needs to supply, causing everything to shit the bed. but the fact that it only occurs when I'm gaming is telling.. thoughts?

 

[LilJoka]

Hey LilJoka. you actually helped me when I was first experiencing this problem with the faulty RAM

OK, I'll try that tactic. But, the same crashing problem occurred with my old GPU. Does that seem like more than coincidence to you? I had read that this is symptomatic of the PSU lapsing in the voltage it needs to supply, causing everything to shit the bed. but the fact that it only occurs when I'm gaming is telling.. thoughts?

It could be coincidence, the audio looping is a really good sign for GPU issue really - that is if it's still a symptom.

If the PSU voltage sags, there should be enough capacitance in the components to survive this unless it's substantial and for long enough time. It could be the cause.

I would isolate with a controlled test. E.g. Don't use games, but a benchmark like Unigene Heaven on loop. Make sure fps are uncapped and tune settings till your pushing 120fps+.
Keep an eye on temps and of course turn off the overclocks on the CPU and RAM. It rules out a potential issue/bug with the game/s you are running.

For CPU testing, run prime95 v28.10 blend at stock clocks to make sure it's rock solid.

For RAM use Memtest, you'll need to use multiple instances of the program to utilise almost all your ram, you want to test around 80%.

If any of these fail we can isolate to one component. If none fail then it might be the PSU when both CPU and GPU are loaded. In which case you can test by running heaven and prime95 (run on 5 threads rather than all 8). Now of this fails we could be pointing at the PSU.

Make sure we aren't still getting any latency spikes with LatencyMon the system is idle.

 

[Resilient]

It could be coincidence, the audio looping is a really good sign for GPU issue really - that is if it's still a symptom.

If the PSU voltage sags, there should be enough capacitance in the components to survive this unless it's substantial and for long enough time. It could be the cause.

I would isolate with a controlled test. E.g. Don't use games, but a benchmark like Unigene Heaven on loop. Make sure fps are uncapped and tune settings till your pushing 120fps+.
Keep an eye on temps and of course turn off the overclocks on the CPU and RAM. It rules out a potential issue/bug with the game/s you are running.

For CPU testing, run prime95 v28.10 blend at stock clocks to make sure it's rock solid.

For RAM use Memtest, you'll need to use multiple instances of the program to utilise almost all your ram, you want to test around 80%.

If any of these fail we can isolate to one component. If none fail then it might be the PSU when both CPU and GPU are loaded. In which case you can test by running heaven and prime95 (run on 5 threads rather than all 8). Now of this fails we could be pointing at the PSU.

Make sure we aren't still getting any latency spikes with LatencyMon the system is idle.

Thank you.

Some questions.

1. What constitutes a pass for the Uniengine test?
2. What constitutes a pass for the prime85 28.10 test? Which test settings should I run?
3. How do I decide the PSU is good when running heaven and prime? I'm running these at the same time right?
4. I already ran memtest with each stick individually for 8 hours per stick. Do I need to run again?

Not Understanding the LatencyMon stuff but I'll google. Thanks again mate.

 

[LilJoka]

Thank you.

Some questions.

1. What constitutes a pass for the Uniengine test?
2. What constitutes a pass for the prime85 28.10 test? Which test settings should I run?
3. How do I decide the PSU is good when running heaven and prime? I'm running these at the same time right?
4. I already ran memtest with each stick individually for 8 hours per stick. Do I need to run again?

Not Understanding the LatencyMon stuff but I'll google. Thanks again mate.

1. The longer you run, the less likely it's purely a GPU issue. 4-6hours to be fairly certain.

2. Same as 1. Except for CPU.
Torture test, click blend, click run

3. Yep, running at the same time would stress the CPU, only after testing separately. But when running the prime95 blend test, select 2 threads, so 2 threads are free to feed Unigene Heaven.

4. No I'd say that's enough.

With latencymon, install it and just click start (play icon). Close all programs before doing so, such that the system is idle. Run for a few minutes, we shouldn't see any spikes and the bars should remain green.

 

[XenIneX]

If it's the PSU, you might be able to force a crash by loading up the CPU and GPU (Prime95 + Furmark) and then inserting a disc in your DVD/BD drive. I find the current draw from spinning up an optical drive is enough to tip a marginal PSU over the edge surprisingly often.

 

[teh_pwn]

If you've run memtest for 24 hours (out of hundreds of systems I have seem memory fail so slightly that bit errors occurred every day or so), then the problem is this in order of likelihood:
1. Bad PSU
2. Bad motherboard (caps out of spec, someone fucked up the board and it's missing a resistor/capacitor)
3. Bad CPU. I have seen this once out of hundreds of systems - a CPU actually was the root cause and replacing it resolved everything.
4. GPU if it's only in games. I missed that detail one way or another.

This really is the biggest problem with building custom PCs - i you don't have parts to use for debugging, you'll get stuck.

 

[Resilient]

If you've run memtest for 24 hours (out of hundreds of systems I have seem memory fail so slightly that bit errors occurred every day or so), then the problem is this in order of likelihood:
1. Bad PSU
2. Bad motherboard (caps out of spec, someone fucked up the board and it's missing a resistor/capacitor)
3. Bad CPU. I have seen this once out of hundreds of systems - a CPU actually was the root cause and replacing it resolved everything.
4. GPU if it's only in games. I missed that detail one way or another.

This really is the biggest problem with building custom PCs - i you don't have parts to use for debugging, you'll get stuck.

Thanks for your input. How would you be able to tell if it's a bad motherboard? Just process of elimination for all other elements? So you're saying it's least likely to be GPU?

 

[LilJoka]

Dunno if you fixed it yet, but here's my thought. I have a completely different system from you BUT I had the same issue. Random freezes. It would literally just power down immediately out of nowhere. 1 hour after turning it on, 5 hours after turning it on, or hell, even 40 hours after turning it on (I don't turn it off). My issue came down to my PSU. Apparently it was on it's way out and would randomly drop voltage I'm assuming, or maybe boost it. Who knows. I could never monitor it as it would crash randomly. Replacing my PSU alleviated ever issue though. But also, your temps seem kinda high. I hit 70C but I'm on an FX8320 that's heavily overclocked. Check that too if you havent yet.

It's fine for Intel temps, they throttle at 100c, AMD danger zone starts at 65c.

 

[hodgy100]

check temps run a ram test and check your hard drive for errors.

 

[Javaman]

I had weird crashes and lockups a while back due to a USB wifi stick. Even several brands and different USB ports caused it. I never did figure out how the fix the root cause so I've been running on Ethernet ever since. Another time the cause for random crashes was the power supply. Like others have said try that first.

 

[Xelios092]

It's fine for Intel temps, they throttle at 100c, AMD danger zone starts at 65c.

Huh, guess you learn something new everyday. I always heard to avoid getting above 70C. My CPU freezes up once it hits 72-75C and the whole computer locks up.

As far as PSU goes, I personally own the EVGA SuperNova G2 850W. Probably overkill though, but the lower wattage models are all solid. I see the G3 series came out too, so that should be a solid bet as well.

 

[Resilient]

OK, so an update.

Tonight, I took out the wireless network card.
Then I swapped the SSD and HDD SATA cables (I switched them at the SSD/HDD side, because the SATA ports were blocked by my video card and I didn't want to take them out.

I was just playing FF14, and the game crashed to a black screen, except this time it just restarted (WTF???), 15 seconds later I'm at the W10 login screen and now I'm posting this.

What the hell does any of that even mean?

Event 41
The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly.

 

[accx]

99% certain it's PSU related.
If running furmark only it crashes, possibly gpu drivers/gpu. Test in another system.
If running furmark + prime95 and it crashes, likely PSU. Test another PSU. EVGA G2/G3 are highly recommended. Check http://www.jonnyguru.com/
If you're uncertain of what unit to chose.

Also, when troubleshooting it's wise to disconnect everything except the essentials.
any network cards or sound cards or weird usb stuff, just get rid of it. ssd, ram, gpu and power only.

If crash, remove gpu and run on integrated.
If crash on integrated when trying to run games, likely PSU.
RAM checks out fine, temps checks out fine (i presume?), driver unlikely (now that i've given it some more thought) since you fresh installed OS.

Sorry this post is all over the place, so hungover.


EDIT

If it's the PSU, you might be able to force a crash by loading up the CPU and GPU (Prime95 + Furmark) and then inserting a disc in your DVD/BD drive. I find the current draw from spinning up an optical drive is enough to tip a marginal PSU over the edge surprisingly often.

That's actually hilarious.

 

[Resilient]

99% certain it's PSU related.
If running furmark only it crashes, possibly gpu drivers/gpu. Test in another system.
If running furmark + prime95 and it crashes, likely PSU. Test another PSU. EVGA G2/G3 are highly recommended. Check http://www.jonnyguru.com/
If you're uncertain of what unit to chose.

Also, when troubleshooting it's wise to disconnect everything except the essentials.
any network cards or sound cards or weird usb stuff, just get rid of it. ssd, ram, gpu and power only.

If crash, remove gpu and run on integrated.
If crash on integrated when trying to run games, likely PSU.
RAM checks out fine, temps checks out fine (i presume?), driver unlikely (now that i've given it some more thought) since you fresh installed OS.

Sorry this post is all over the place, so hungover.


EDIT


That's actually hilarious.

I'm gonna run these tests tomorrow.

Temps are fine.
RAM is fine.
Drivers unlikely, this is the 3rd fresh install.

Disconnect the USB connections to the Motherboard? That huge ass USB 3.0 one yeah? How about the little power cables for lights/USB and stuff?

Thank you for your post! sorry to hear about your hangover

I ended up purchasing the Corsair RM750x Modular 80 Plus Gold Power Supply as I couldn't find the G2/G3 in stock near me. Should be fine?

 

[accx]

I'm gonna run these tests tomorrow.

Temps are fine.
RAM is fine.
Drivers unlikely, this is the 3rd fresh install.

Disconnect the USB connections to the Motherboard? That huge ass USB 3.0 one yeah? How about the little power cables for lights/USB and stuff?

Thank you for your post! sorry to hear about your hangover

I ended up purchasing the Corsair RM750x Modular 80 Plus Gold Power Supply as I couldn't find the G2/G3 in stock near me. Should be fine?

Good luck!
I mean, i just usually disconnect everything... you'll need the small powercables for the power button to work but everything else could go. It's not really necessary though... I was talking more about case fans, additional hard drives, optical drives and so on..

I'm getting mexican food from my SO soon so hangover likely to be cured!

Yea it looks like it has a great score on johnnyguru
http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story&reid=452

I'm wary of Corsair just because they had once a great track record (TX-series for example), but then started releasing V2's of these units, which broke down like... all the damn time. So users essentially got tricked into buying the "The new and improved V2's" because people on forums didn't distinguish them from V1s. This lead to a lot of less tech savvy peoples computers broken down.
Then they released the extremely shitty low budget CX-series which also everyone less tech savvy bought (because hey, it's cheap), which again, lead to peoples computer breaking down.
After that shit i just stay clear away from them. They made life miserable for people like me, that was at the time, acting tech support for EVERYONE!
Ugh.
It's ironic that the PSU i've got in my main computer is a Corsair TX-750w. Still going strong, almost 10 years now.

 

[Resilient]

Yup, just happened again, this time no reset, froze at the game screen, audio micro-stuttering (just garbled). Had to power it down manually (5s hold down). This is without the OC on, everything in the BIOS set to Default. Stress testing tomorrow.

 

[Engell]

I ended up purchasing the Corsair RM750x Modular 80 Plus Gold Power Supply as I couldn't find the G2/G3 in stock near me. Should be fine?

750watts is prolly overkill... but sure it's a good PSU

sucks that it is still doing shit.
had similar problem many years ago, where winXP would hard crash. turned out i could provoke the error in memtest86 test7(and only that test for some reason).. but ultimately it was actually a defective memory controller on the CPU that was causing it.(but CPU problems are very rare, but it can happen)
The problem was very hard to explain in regards to warranty etc, but in the end i got it exchanged to a new CPU for free.

 

[Rudimental]

Do you have any crash dump files in C:\Windows\Minidump?

 

[Resilient]

Good luck!
I mean, i just usually disconnect everything... you'll need the small powercables for the power button to work but everything else could go. It's not really necessary though... I was talking more about case fans, additional hard drives, optical drives and so on..

I'm getting mexican food from my SO soon so hangover likely to be cured!

Yea it looks like it has a great score on johnnyguru
http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story&reid=452

I'm wary of Corsair just because they had once a great track record (TX-series for example), but then started releasing V2's of these units, which broke down like... all the damn time. So users essentially got tricked into buying the "The new and improved V2's" because people on forums didn't distinguish them from V1s. This lead to a lot of less tech savvy peoples computers broken down.
Then they released the extremely shitty low budget CX-series which also everyone less tech savvy bought (because hey, it's cheap), which again, lead to peoples computer breaking down.
After that shit i just stay clear away from them. They made life miserable for people like me, that was at the time, acting tech support for EVERYONE!
Ugh.
It's ironic that the PSU i've got in my main computer is a Corsair TX-750w. Still going strong, almost 10 years now.

good to know. I went with it cause, like you said. reviewed well. thanks for your help man. take it easy, i'll let you know how it goes tomorrow.

750watts is prolly overkill... but sure it's a good PSU

sucks that it is still doing shit.
had similar problem many years ago, where winXP would hard crash. turned out i could provoke the error in memtest86 test7(and only that test for some reason).. but ultimately it was actually a defective memory controller on the CPU that was causing it.(but CPU problems are very rare, but it can happen)
The problem was very hard to explain in regards to warranty etc, but in the end i got it exchanged to a new CPU for free.

provisioning to plug audio gear etc in..didn't want to skimp on it. it is on the high end though. good to hear that you actually found the problem! fingers crossed on my end.

Do you have any crash dump files in C:\Windows\Minidump?

that folder doesn't exist for me

 

[Rudimental]

that folder doesn't exist for me

Ah. You can enable dumps then by going to Control Panel > System > Advanced System Settings (on left) > Advanced (Tab) > Settings (under 'Start-up and Recovery' heading)

Then tick the boxes under 'System Failure' and select 'Small memory dump (256k)' under the 'Write Debugging Information' heading. It should put in the directory automatically.

edit: you'll probably need to reboot for it to take effect

 

[jonnyp]

OP, do you have Killer network card by any chance?

 

[Apt101]

I should change the thermal paste? But are those temps not good (and under stress testing too)?



I don't have a spare CPU around so I'd have to buy one - not the cheapest option around here for 1151 socket either see above - why is the CPU most likely the culprit ? It's confusing because it works fine for extended periods or short periods in the same games and then just dies.

If a CPU overheats doesn't the system shut down?

Oh shit I meant PSU. I blame my lack of coffee.

 

[LilJoka]

OP, do you have Killer network card by any chance?

Intel® I219V, 1 x Gigabit LAN Controller(s), GameFirst technology

good to know. I went with it cause, like you said. reviewed well. thanks for your help man. take it easy, i'll let you know how it goes tomorrow.

provisioning to plug audio gear etc in..didn't want to skimp on it. it is on the high end though. good to hear that you actually found the problem! fingers crossed on my end.

that folder doesn't exist for me

RMx Series is good, Corsair have a bunch of series that are rubbish now.

 

[jonnyp]

Intel® I219V, 1 x Gigabit LAN Controller(s), GameFirst technology



RMx Series is good, Corsair have a bunch of series that are rubbish now.

OK, the problem sounded like the same issue I was having with killer networking driver and windows network data usage monitoring service causing a memory leak and would freeze my PC after a few hours. Drove me nuts until I found a solution

 

[i3allistic]

OK, the problem sounded like the same issue I was having with killer networking driver and windows network data usage monitoring service causing a memory leak and would freeze my PC after a few hours. Drove me nuts until I found a solution

shit, i've been having freezing problems too...and i have a killer E2200 network card with my MSI MB ...im downloading the update now ....

 

[soco]

Drivers could still be an issue. Fresh installs don't protect against buggy drivers. Try checking searching for the gfx card version to see if others are having issues.

i seem to recall some rare cases where bad drives could cause stalls and halts like that.

does it happen if you take the OC down? overclocking can cause weird and infrequent and subtle issues.(nevermind just saw the update in the op)

 

[Baron Aloha]

Did you try uninstalling GeForce Experience yet? Assuming you have it installed?

 

[PaNaMa]

In my experience, that's usually a ram issue. But I've also seen it as a power issue. Disable your CPU OC entirely, do not use XMP mode (for now). Set CPU and RAM to just "auto" and run everything at base. Save those settings, then shut down... Assuming you have sticks of RAM, try operating on just the one stick and then boot into windows and play a game. 
Does it crash?
If so, shut down, take out the other stick of ram and swap. Boot again, play a game.
Does it crash?

It can also be power. It could actually be your power outlet. I had Seasonic send me a new PSU because I thought the PSU was shot - cause my system was behaving exactly like yours.. Never crashed during normal operations, only under gaming load. It wasn't an issue with the PSU as the same thing happened with the replacement.  

While you are gaming you are drawing a lot of power. Even a momentary dip will cause the system to crash. I think it was my outlet, or the electrical circuit in the house servicing that outlet, + the baseboard heater, + room lights etc was overloaded. 

I ended up buying a UPS to ensure clean power to my system at all times, and prevent even the slightest power interruptions. No more system crashes. UPS prevented brownouts to the computer, and I never had the problem again. Your mileage may vary, but it's worth a try.

Ram or Power, almost certainly imo =\

 

[EAPidgeon]

From what I can tell you isolated driver issues by reformatting. So unlikely to be drivers although I have fixed a system with similar issues and such was the case. (Upgrading from 7 to 10 fixed it in this case).

Second guess would be the PSU. I had an old Corsair HX1050 which drooped out when I payed attention to the voltage and I would get random restarts infrequently (no freezing though). I might suggest running Warframe as weirdly this really seemed to consistently and quickly agitate the issue. (oddly not The Witcher 3 for hours on end).

You seem to be tackling it in a timely manner, but finally I did have to replace the PSU and the board in that system. I suspect the faulty PSU burned out a component in the motherboard as I identified small scorch marks on one part of the board.

Just something to keep in mind when repairing or swapping PSUs. Give the motherboard a good inspection as well.

 

[Resilient]

I've just completed a 7.5 hour CPU test in prime95 28.10 - blend, 4 threads.
No OC. Default BIOS settings. XMP off. RAM XMP off.
No tests were failed. No freezing. Temps ran from 48 - 68 on average with a few peaks at 73.

I think this safely rules out the CPU as being the problem. CPU = OK.

Next step is to test the GPU using Unigine Heaven. I'll be doing this tomorrow.

Thanks again for the input, all.

I had about 3-4 crashes/restarts last night while playing FF14 - this was with the OC turned off, default BIOS, no XMP settings applied. So that rules out the OC being a problem I guess..

Drivers could still be an issue. Fresh installs don't protect against buggy drivers. Try checking searching for the gfx card version to see if others are having issues.

i seem to recall some rare cases where bad drives could cause stalls and halts like that.

does it happen if you take the OC down? overclocking can cause weird and infrequent and subtle issues.(nevermind just saw the update in the op)

Yeah, it happens even with the OC turned off

Did you try uninstalling GeForce Experience yet? Assuming you have it installed?

No, I didn't (sorry). How does this affect my driver updates in the future, though? That's what I was a bit confused about.

In my experience, that's usually a ram issue. But I've also seen it as a power issue. Disable your CPU OC entirely, do not use XMP mode (for now). Set CPU and RAM to just "auto" and run everything at base. Save those settings, then shut down... Assuming you have sticks of RAM, try operating on just the one stick and then boot into windows and play a game. 
Does it crash?
If so, shut down, take out the other stick of ram and swap. Boot again, play a game.
Does it crash?

It can also be power. It could actually be your power outlet. I had Seasonic send me a new PSU because I thought the PSU was shot - cause my system was behaving exactly like yours.. Never crashed during normal operations, only under gaming load. It wasn't an issue with the PSU as the same thing happened with the replacement.  

While you are gaming you are drawing a lot of power. Even a momentary dip will cause the system to crash. I think it was my outlet, or the electrical circuit in the house servicing that outlet, + the baseboard heater, + room lights etc was overloaded. 

I ended up buying a UPS to ensure clean power to my system at all times, and prevent even the slightest power interruptions. No more system crashes. UPS prevented brownouts to the computer, and I never had the problem again. Your mileage may vary, but it's worth a try.

Ram or Power, almost certainly imo =\

I've tested the ram using memtest86, for 8 hours, one stick at a time (total test time 16 hours) and didn't encounter any errors, so I'm fairly certain it is not the RAM at this point. Though - if/when I do replace the PSU, and if the error persists - I will isolate the RAM and use it in a game - but surely the memtest86 runs would rule out it being a RAM issue?

The power outlet thing is a worry. I'm plugged directly into a wall, and I've never had an issue with power in this room for other items. But maybe it could be an issue..

From what I can tell you isolated driver issues by reformatting. So unlikely to be drivers although I have fixed a system with similar issues and such was the case. (Upgrading from 7 to 10 fixed it in this case).

Second guess would be the PSU. I had an old Corsair HX1050 which drooped out when I payed attention to the voltage and I would get random restarts infrequently (no freezing though). I might suggest running Warframe as weirdly this really seemed to consistently and quickly agitate the issue. (oddly not The Witcher 3 for hours on end).

You seem to be tackling it in a timely manner, but finally I did have to replace the PSU and the board in that system. I suspect the faulty PSU burned out a component in the motherboard as I identified small scorch marks on one part of the board.

Just something to keep in mind when repairing or swapping PSUs. Give the motherboard a good inspection as well.

I'll inspect the Motherboard first thing tomorrow morning. I'm going to check the capacitors to see if any of them have bulged, but is there anything else in particular I should look out for? Image links would be great.

 

[Resilient]

Sneaky double post.

I just ran Unigine Heaven for about 0.5 - 1hrs, and it crashed, frozen looking at the dragon. lol. couldn't alt+tab, lost functionality.

CPU at default BIOS settings.
GPU at stock Mhz. It was peaking at around 71/72 according to Unigine.

I'm going to try and test it again, but indicatively, what could this suggest - considering that this freezing/crashing error was occurring with my previous GPU (7850 OC 2GB Gigabyte card).

 

[Rootbeer]

I know you've tried a lot of things. Just wanted to add a small experience of mine.

I was having a lot of system instability issues recently after a power outage. I couldn't figure out what was the problem and ruled out it was not failing components, yet it was still happening. I literally thought the mobo was faulty after the outage. Tried a lot of things you mentioned. One thing finally fixed it for me... I simply went into the BIOS and used the "reset to factory defaults" button and then everything worked again. After booting into Windows and confirming all the problems were gone, I then went back into the BIOS and adjusted settings to the way I wanted them (RAM Speeds, etc), and things still worked perfectly after many reboots. Somehow, the actual settings were corrupted. Yes, even though I removed the CMOS battery and did that whole song and dance prior, I had to manually reload the defaults it to get it instability to go away.

So, I know you said you tried using the default bios settings, but did you try resetting those settings in this specific way? If not, one more thing you can try.

Good luck.

(PS I have a very similar mobo to you, same manufacturer and chipset so yeah)

 

[DieH@rd]

Since the freezes happen only during heavy use [gaming], check if maybe your VRMs are overheating. I had that issue, and it took me a while to locate what is causing framerate spikes during gaming.

 

[Rudimental]

I really suggest turning on those mini crash dumps. Even if your PC is freezing in a way that doesn't make them, it tells you something, i.e. that it is almost certainly hardware related. If they do get made, then you can see whether it is hardware or software related, and in the latter case trace the cause. You need Windows debugging tools to analyse them, if you get the dump files, but you can give them to me or another Gaffer to look at.

 

[Resilient]

Fuck. Recent posts have me thinking I shouldn't replace the PSU just yet...does reset to factory default for the mobo take it alllllll the way back to the original BIOS?

I really suggest turning on those mini crash dumps. Even if your PC is freezing in a way that doesn't make them, it tells you something, i.e. that it is almost certainly hardware related. If they do get made, then you can see whether it is hardware or software related, and in the latter case trace the cause. You need Windows debugging tools to analyse them, if you get the dump files, but you can give them to me or another Gaffer to look at.

I will do this tomorrow. Only reason I hadn't earlier was because it was in the middle of the CPU test. Appreciate it - if some dumps show up I'll PM them to you if you don't mind.

 

[s_mirage]

Fuck. Recent posts have me thinking I shouldn't replace the PSU just yet...does reset to factory default for the mobo take it alllllll the way back to the original BIOS?

No, it just resets it to the defaults for the currently installed BIOS.

 

[Diablos]

Either your GPU is overheating or you have a power supply issue.

 

[jonnyp]

shit, i've been having freezing problems too...and i have a killer E2200 network card with my MSI MB ...im downloading the update now ....

Updating driver didn't work for me. What I had to do was disabling windows network data usage as per https://bluebluewave.wordpress.com/...00-and-windows-network-data-usage-monitoring/

You have to do this every time win 10 updates to a newer version if you are an insider and regularly get new builds

 

[Nostremitus]

My wife's PC was freezing constantly a few months ago. It took me forever to troubleshoot. Turns out it was a storage drive. It was going bad and a program she had running in the background was on it. That program would call the drive, it wouldn't respond correctly, and the whole system would freeze.

I replaced the drive and she's had no issues

 

[KTOOOOOOM]

let me help you.
2 things you need to check.

1- PSU.
write back when you test the new PSU.

2-Motherboard memory controller.
this is a bad one if you can find the memory controller for you Mobo then a little of soldering can fix it if the warranty is expired .otherwise RMA it.

 

[zeomax]

I just ran Unigine Heaven for about 0.5 - 1hrs, and it crashed

Download Nvidia Inspector. Set the GPU Fan speed to 100% and try the test again.

 

[BasicMath]

Sounds like the issue's the PSU. Switch it out.
After that, I'd start testing DRAM modules/slots. I doubt it's your GPU (you basically tested that) or CPU (rare).

Stress testing

Stop fucking around with it until you get the new PSU to test it out. You're playing with fire.

Event 41
The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly.

Yep. I'd say we're at a good 70% chance of it being a PSU issue.