My PC is randomly rebooting but only in two games.

[Mouzon]

Hi guys, would be amazing if someone is able to help me get to the bottom of a problem I'm having that is killing gaming immersion because of random reboots.

Before posting I've really tried to diagnose the problem according to other peoples help in multiple forums but none of the things that seem to have fixed their problems are working for me.

I'll start with my PC specs:

PSU: Thermaltake TR2 Heavy Duty Low Noise 800W
Motherboard: ASRock Asus Z77 Extreme 4
CPU: Intel i7 3770k 8 core 3.5Ghz overclocked to 4.4Ghz
CPU Cooling: Corsair H80 Water Cooling
GPU: Palit nVidia Geforce GTX 680 2GB
Hard Drive: 240GB OCZ Agility SATAIII SSD + 2 x 2 TB Seagate SATAIII HDD
RAM: 32GB 1600 Mhz DDR 3 G.Skill
OS: Windows 8

Now the issue I'm having:

I am getting utterly random reboots (no freezing or frame rate degradation, no blue screen or error messages) when I play just two games, Wildstar and now GTA V.

I'll be doing anything in either game, just walking around, indoors or outdoors, action or no action and the computer will just restart. It doesn't turn off or shut down, screen is black for one second and then it's booting Windows again.

And here's the thing, it happens for no other game. So I can be playing heavily modded Skyrim, Farcry 3, WoW raiding, nothing else does it. It has also never done it in any other program, so not in Pro Tools (audio editing software) even when I'm running the CPU at 90% + capacity, doesn't skip a beat. It doesn't happen when video editing and rendering in Premier Pro either.

I've tried to narrow it down by stress testing each component but none of them seem to budge. I've downloaded coretemp to monitor the CPU and it rarely nudges 60 degrees. I've run Furmark at full settings for 20 minutes + to try to fault the GPU but that's not happening either, won't crack 77 degrees. I've also stress tested the CPU with Furmark while monitoring the temp for extended periods and nothing goes wrong there either.

I've done all the memtest86 tests and the RAM also checks out with no errors.

I've also turned off the CPU overclocking in the BIOS to cut that possibility out of the equation.

In Windows settings I've clicked the box that says 'Don't restart on critical errors' which did nothing. I also set the CPU to only use 99% capacity at the suggestion of a Wildstar dev.

Someone suggested plugging the PSU directly into the wall in case it is the power board overloading but I've done that and no difference either.

I'm skeptical that it would be the PSU because it can go for days without a restart and then just restart randomly when only playing either Wildstar or GTA V.

Any suggestions are more than welcome, this is killing me!

[yurano]

Maybe the PSU rail voltages dip/spike out of spec in high intensity games (CPU+GPU), tripping a voltage protection in the CPU and/or motherboard.

A simple but tedious way to test this is to buy a new PSU from a big box store with a good return policy (return it if it doesn't solve the problem). Seasonic PSUs (Seasonic, Antec, or XFX) are an excellent choice, but they're rare in retail. If you can't find Seasonic, I recommend Corsair's TX line or better. Thermaltake's TR2 line isn't very good.

Recomended PSUs

Misc notes:
1. Memtest86 usually isn't strenuous enough to detect RAM faults. HCI Memtest is more effective. Even so, RAM errors cause blue screens and not black-screen retarts
2. GPU problems usually cause no display, or display errors rather than restarts
3. In these types of scenarios, the main fault with the motherboard is VRM overheating which causes speed degradation (e.g. AMD-8350 on a poorly cooled motherboard) rather than random shut downs.
4. Since you've ruled out the most common cause of black screen restarts (CPU overheating), the PSU seems the most likely source of the problem.

[Mouzon]

Thanks for the reply, Yurano.

Well, I hope it is the PSU! As it's going to be cheaper than anything else to fix. I will give it a go and let you know here if it solves the problem, fingers crossed!

[Deleted]

Run a cpu stress test that doesn't use your gpu at all for a while, like prime95, then run a gpu stress test that doesn't bother your cpu much, like furmark. If all is fine so far with no errors, run both of them at once. If this then causes your computer to reboot it's almost certainly the PSU.

[Aqua]

This sounds like exactly what happened right before my PSU croaked and I had to replace it. Funnily enough it was while I was testing wildstar also.

Slayer of PSU's that game.

[Shakadam]

I had a similar issue with my previous computer and it turned out to be the motherboard simply not liking several RAM sticks. Just like you it would just randomly reboot in some games. Removing 2 sticks of RAM solved the issue completely. I tried several times to run 4 sticks of RAM again while tweaking RAM voltage, timings etc. I tried pretty much everything but just couldn't get it to work with 4 RAM sticks.
I should also point out that this computer passed through memtest with 4 sticks of RAM without any problems, because the issue wasn't the RAM itself, rather the interface or connection between RAM <-> Motherboard.

[Mouzon]

Thanks for the further replies all. Just to update I've tried as suggested running both prime95 and furmark side by side but it just won't reboot. As the screenshot shows, both the CPU and GPU were pulling close to 100% and I even opened other programs at the same time to see if it would push it over but to no avail, it just refuses to budge except when I'm playing Wildstar or GTA V.

Sounds too good to be true Shakadam but I'm going to give it a go! I'm really not very knowledgeable in this area so I'll have to watch a Youtube video about which sticks to take out.. that would be a far easier option in my situation than going out to get a new PSU.

http://imgur.com/QCL1gwG

[Afrospinach]

Thanks for the further replies all. Just to update I've tried as suggested running both prime95 and furmark side by side but it just won't reboot. As the screenshot shows, both the CPU and GPU were pulling close to 100% and I even opened other programs at the same time to see if it would push it over but to no avail, it just refuses to budge except when I'm playing Wildstar or GTA V.

Sounds too good to be true Shakadam but I'm going to give it a go! I'm really not very knowledgeable in this area so I'll have to watch a Youtube video about which sticks to take out.. that would be a far easier option in my situation than going out to get a new PSU.

http://imgur.com/QCL1gwG

Try updating your drivers too. My computer used to reboot without so much and an error log whilst playing flash videos back in 2012. I don't know what exactly fixed the problem but I think it was nvidia related. Either way it has never happened since with no hardware changes

[Mouzon]

Damn, it wasn't the RAM sticks Shakadam, just took two of them out and played for about 5 minutes before it reboot the computer.

Drivers are all up to date, but thanks for the suggestion Afrospinach. The latest Nvidia drivers seemed to actually be specifically for GTA V.

So frustrating...

Before buying a new PSU, are there any ways to test it that don't require an electrician?

- - - Updated - - -

Just another update, I read that DX11 can cause the problems and to run the game in DX10.1 and gave it a go but no luck, same thing happened after a few minutes.

[Yxiomel]

Have you tried pulling out the otehr two RAM sticks?

[Christan]

i haven't seen this mentioned yet but...
https://forums.wildstar-online.com/f...-possible-fix/

I reviewed your Speccy report and it does look like your processor is running at higher than stock speeds. I believe it's not actually over-clocked, but the processor is utilizing the built-in Turbo Boost option, which will raise the cores speeds based on load. This is a feature of the CPU model you purchased, marked with the (k) for Intel or the "AMD Black Edition" for AMD. Any CPU's of these series are able to run faster than intended, automatically, without you doing or changing anything. This can often lead to freezing or crashing in applications that are not expecting this to occur.

Since this situation simulates over-clocking, I'd still like to try to disable this, and for that I have included some instructions for a work around below:

Go into Power options (Control panel -> Hardware and Sound -> Power Options) and click “Change plan settings” near your plan.

Then click on “Change advanced power settings.”

Scroll down to “Processor power management” and expand it.

Expand “Maximum processor state” and modify the setting to 99%. Be sure to also set the minimum processor state to something below the maximum. I recommend 5%.

Then simply click “Apply” and “OK.”

After this has been done try launching the game again. When you do, the turbo boost feature should refrain from kicking in and that may resolve the issue you've been experiencing.

seems the turbo feature will change the processors speed mid game, and they haven't been able to code it to take that into account, or something...idk

[Mouzon]

I haven't tried taking these two RAM sticks out and replacing them with the others but I will give that a go. I don't think the RAM idea though was that there was faulty RAM, the suggestion was just that the MOBO might not like having all 4 RAM slots filled at once.

Thanks for the suggestion from the Wildstar forum Christan, I had actually read that post when the reboots first started a few months back playing Wildstar and still have those settings in place, so unfortunately that isn't the problem either.

[Mouzon]

Update: After watching so many videos on hardware yesterday, I'm pretty sure I could put together a computer with a blindfold on. So I've learned a lot from this experience!

One thing I didn't realize is that PSU's suck air from UNDER the PC case. I'm pretty fastidious with cleaning the inside of my computer and the fans, but I wasn't aware of that one, and I've been running it now for 3 years. I checked it out and there was so much dust and lint built up that it couldn't have been sucking in any air whatsoever.

I cleaned it out and put it all back together again and started up GTA V. Played uninterrupted for a good 5 hours or so, normally it will shutdown anywhere between 5 and 30 minutes of play.

I'm not saying that's the answer, cause it might just stuff up again tonight, but fingers crossed that was the problem and if it was then that's a good lesson to anyone to keep ALL of your fan filters clean!

[blackblade]

Any suggestions are more than welcome, this is killing me!

next time it happens, first thing you should do is take a peek at event viewer and see if a cause of shutdown is listed.

It could be any number of things. feedback from event viewer will significantly narrow it down (and give you a starting point on the google)

[OhpUldum]

Update: After watching so many videos on hardware yesterday, I'm pretty sure I could put together a computer with a blindfold on. So I've learned a lot from this experience!

One thing I didn't realize is that PSU's suck air from UNDER the PC case. I'm pretty fastidious with cleaning the inside of my computer and the fans, but I wasn't aware of that one, and I've been running it now for 3 years. I checked it out and there was so much dust and lint built up that it couldn't have been sucking in any air whatsoever.

I cleaned it out and put it all back together again and started up GTA V. Played uninterrupted for a good 5 hours or so, normally it will shutdown anywhere between 5 and 30 minutes of play.

I'm not saying that's the answer, cause it might just stuff up again tonight, but fingers crossed that was the problem and if it was then that's a good lesson to anyone to keep ALL of your fan filters clean!

When troubleshooting PCs it is quite often the simplest things that are overlooked. Especially odd since it was only occurring with the two specific games but I can chalk that up to something specific that those two games do. Hopefully the PC keeps on rocking for you dude.

[Mouzon]

Okay so this problem has actually continued, cleaning out the PSU fan fixed it for just the one day which is kind of maddening as it still doesn't help me locate the cause of the issue whatsoever.

Windows event log can see that it's crashed but the event (31 from memory?) is super generic when you google it and could be caused by anything, so no help there.

I've switched the RAM around to a bunch of different combinations so I can safely say that it's pretty well ruled that out.

I'm just going to replace everything one piece at a time starting with the cheapest thing, the PSU.

Can anyone who knows their stuff please tell me, based on my PC (link in signature) what a good replacement would be that is available in this PC store so I can just order it today.

http://www.pccasegear.com/index.php?...772870fd804f57

Thanks for all of your help on this one, everyone!

[Afrospinach]

http://www.resplendence.com/whocrashed

try this, it can be more informative than event viewer.

[mordrac]

Okay so this problem has actually continued, cleaning out the PSU fan fixed it for just the one day which is kind of maddening as it still doesn't help me locate the cause of the issue whatsoever.

Windows event log can see that it's crashed but the event (31 from memory?) is super generic when you google it and could be caused by anything, so no help there.

I've switched the RAM around to a bunch of different combinations so I can safely say that it's pretty well ruled that out.

I'm just going to replace everything one piece at a time starting with the cheapest thing, the PSU.

Can anyone who knows their stuff please tell me, based on my PC (link in signature) what a good replacement would be that is available in this PC store so I can just order it today.

http://www.pccasegear.com/index.php?...772870fd804f57

Thanks for all of your help on this one, everyone!

you have an overclocked 4(8) core cpu and a 680gtx. both eat a lot power from the 12v lines.
run something like furmark to thress the gpu and prime to stress the cpu at max.
if your psu runs hot, gets loud and the pc just shuts down/restarts, it for sure is the psu that fails.

you can also test you gpu and cpu at 100% each for some hours and see if the pc runs stable, so you are sure that the cpu and gpu are fine.
i cant realy suggest you a better psu, just tell you that you need to look at how much ampere your 12v lines offer.

its labeled on its side. if you like post it here.
good psu have up to 4+ 12v lines with 24 ampere+ each.
the wattage doesnt say to much about a psu. you can buy 750w psu for 40$. and you can be sure that it would explode when it would be used in your pc.

edit:
http://www.pccasegear.com/index.php?...ducts_id=26073

i myself use one of those. i used serveral bequiet psu before and this one realy silent and cool, and it got a lot of power.
just look at the 12v lines. i bet you psu only have something like 18amps each line or something 32 in a single one.

[Mouzon]

Update! I bought and received the PSU that was suggested in another thread, and had an absolute blast putting it into the system. I've never put anything into a computer so it was lots of fun to learn how it all works. It actually looks even better than the previous one did due to being able to selectively choose which cables to use and which not to use.

And here's the thing. It would appear as though it has completely solved the problem, no more restarts! So thanks for all of the advice! Hopefully someone else with such a specific and maddening problem might try this as their solution.

[Deleted]

Have you tried changing shoes?

In GTAV when i changed Trevor's boots to sandals, what a difference, calm and cool.