[SOLVED] My 1-year new computer started to shut down randomly while gaming

Flourix

Commendable
Jan 4, 2020
9
0
1,510
Hello!
I have been browsing the web and different threads to find someone with the same problem as I have. I have been playing CoD just fine as well as other heavy games and titles just fine before this started to happen. It begun to happen in the end of November 2019. With this thread I hope to find someone who has the same problem as I do and has managed to solve it or hopefully some experts knowing what to do to troubleshoot this. I just want to say a big thank you in advance for reading through this thread and helping me out.

Setup
Case
: Fractal Design Define R6
Motherboard: Gigabyte X399 Aorus Xtreme
Powersupply: Corsair AX1200i Digital 1200W
CPU: AMD Ryzen Threadripper 2950x 3.5Ghz
RAM: G.Skill 64GB (8x8gb) DDR4, 3200 Mhz CL14
GPU: Asus Geforce RTX 2080 Ti 11GB Dual OC
Storage: 2x Samsung 970 (500gb), Samsung 860(1tb), Samsung 970(1tb), Seagate Ironwolf 7200rpm (6tb)
Cooling: 4x Noctua NF-A14 140mm & CPU Cooler Noctua NH-U12S TR4 SP3.

Temperatures are fine, GPU is never above 75 Celsius, both ingame and Rainmeter/HWInfo measurement and CPU is not even reaching above 60 in AMD Ryzen Master. If you want me to check on any specific core temp or something else, please tell me and I will happily do so.

Whats happening?
My computer shuts down completely without a warning, leaving only the LED-Lights within the case with power/light. It happens randomly within the games, no specific patterns. The Power button does not work and I have to flip the switch at power supply at the back to even get it going again. If I do that too quick, the computer boots just for few seconds before turning off again. So I have to wait for at least 10 minutes or so. Don't really know the exact time, since I have always gone frustrated and went do something else when it happens.

I had my PC at service (INET Computer store service in Gothenburg, Sweden), they are also the ones who built my PC when it was new. They couldn't get the problem to occur in their 24-hour stress tests. They reset and reinstalled BIOS, which I also did myself at home, so I was afraid problem was not fixed. I came home and it worked for some days, but I had very short gaming sessions - problems usually occurs past 1 hour of gaming. However, after a few days it shut down again. It has now been about one time a day minimum where my computer shuts down and first time it happened since service was 27th of December.

Right now I'm really tired and sad. Because it's a expensive computer which you invest into to handle both gaming and heavy work with videos in Adobe Premiere. And when it doesn't work it destroys most of your motivation to work and play. I'm really thinking about quitting the game, since this can't be good for my components really. But then again this also happened once when I was in Windows - after the computer woke itself up from a sleep, which it also does every now and then since I bought it.

I am now afraid this might be a game related issue or at least an issue with something I have at home if not the game - like screens, or something with the power from wall sending spikes putting a safety switch off function in the computer or something. But this is where I need expert help because I'm getting crazy thinking about what it can be.

I just wanted to hear if someone else is experiencing the same problem and if so, how it might have been solved for you? Also from you experts that work with these kind of stuff, do you know what this problem might be?

My last way out of this is to again deliver my computer to INET for a new service. But I'm afraid they won't do a full service thinking the problem doesn't exist just because they do not get it to reappear. And I will continue getting my computer back home without it working. I don't have a car at the moment so getting to INET in Gothenburg is a project by itself with a computer this big - so I would be VERY happy if I can fix this myself at home with the online-help of you guys.

What has been done and other information
In this section I will just mention some stuff that has been done now since the problem started and also some information I noticed.

  • In AMD Ryzen Master there is a value named PPT (CPU), which has a %-value of 180W, Limit 900W. I don't know what this value is, but have noticed that it while gaming reaches close to 100%, and that it gets colored yellow and red and higher percentages. Is this a value that can be raised somewhat, or should be raised? I mean my PSU is 1200W, but it limit says 900W and also why only 180W of those 900.
  • Another Value in AMD Ryzen Master is the CPU Voltage under CPU Voltage Control. This is reaching a lot of different values between 1.2750 and 1.5000. I also have a peak speed of almost constantly 4.100-4.300 GHz. And I have NOT overclocked this CPU anything. Is this some sort of spike voltages boosting speed, and sometimes maybe overspikes and shuts computer off? Can it be a reason for my problem?

  • I have cleaned the computer since the problem occured. It was not very dusty, it has been cleaned about one time each Quarter, and no, not with a vacuum cleaner.
  • I have let the computer stay on outside of Call of Duty, like when sleeping, to see if problem occurs. It does not. Only time it did was once the computer itself woke up from a sleep while I was at work.
  • I have tried to play less demanding games, and it works well. But those game sessions was not that long really. Only been for 2+ hours into Football Manager 2020. I'm thinking about patching up Battlefield V again and play and see if problems occurs there as well, since it's like almost equally as demanding game as Call of Duty.
  • I did render in Adobe Premiere without problems for about 30 minutes. The PPT CPU value did though reach high % values just as it did while I'm gaming.
  • I have updated some drivers for motherboard that was some version old, like sound and wifi driver.
  • I have updated Windows to the latest autumn patch.
  • I have updated NVIDIA drivers, but it was only game ready drivers for other games, nothing major.
  • I have myself done stresstest like Furmark and Unigine at home without any problems. I have not done any CPU test at home though - so if you have a suggestion here what to run, I gladly would try it.
Images
Photo from Ryzen Master when rendering video in Adobe Premiere:
EFeXYV4.jpg

Photo from Ryzen Master during Idle time in Windows
Suv7e48.jpg

A Series of photos taken from Ryzen Master between each game in Call of Duty to show temperatures looks fine.
wityFSn.jpg

vjSPPZ4.jpg

8HzUUIE.jpg

Yfqqy1Z.jpg

jUakmOb.jpg

FhAs9L0.jpg

ZP5C8qG.jpg

jrogZwY.jpg

OlVNK5k.jpg


Once again, just tell me if you need any information or values to troubleshoot this and again Big thanks in advance for any help!
-Robin
 
Last edited:
Solution
So, I have tried to play some Battlefield V now. Maximized settings with DXR and everything.
Temperatures were all fine with GPU around 75 Celsius and CPU around 50 Celsius.

What struck me was that CPU voltage now looked stable and didn't spike as much up to the 1.4 and 1.5v. So the peak speed was at about steady 3.9-4.0GHz

nUliycd.jpg

xR4mQ2P.jpg

gAhHRfo.jpg

I was wondering if it was a good sign since it has never been like that before. And the only thing I have done so far since the opening post is actually to switch the PSU-cable to another power socket in the wall - because I thought of it could be my powersockets being old or dusty/damaged...
Normally, random restarts like that, at load, indicate an issue with the PSU. The Corsair AX is a great PSU, but it could be failing. Can you get a lend of one to swap out and try or have a back up anywhere you can use to test?
Otherwise bring to a local repair store and get them to swap it out and test. It might only cost about $20.

You have a pretty power hungry system, so you would need something with 850w+ and possibly two eps connectors for the TR CPU to swap with.
 
First off, that is the best written and most informative "PC problems" post ever. Also being your first post is awesome!

On to the issue at hand. Echoing the post above, I'd test with a new PSU. Before that, though, I'd check another thing. Even though it doesn't appear to be a RAM problem, run memtest86 overnight to check for errors. Any error is bad.

Another thing to do is use Ryzen Master to reset any changes you made to the system to default. Then uninstall the program. Next, reset to defaults in BIOS.

I would also load up BF1 as it really uses a lot of CPU and RAM.
 
  • Like
Reactions: Roland Of Gilead

Flourix

Commendable
Jan 4, 2020
9
0
1,510
Normally, random restarts like that, at load, indicate an issue with the PSU. The Corsair AX is a great PSU, but it could be failing. Can you get a lend of one to swap out and try or have a back up anywhere you can use to test?
Otherwise bring to a local repair store and get them to swap it out and test. It might only cost about $20.

You have a pretty power hungry system, so you would need something with 850w+ and possibly two eps connectors for the TR CPU to swap with.

Thanks for replying to me!
Sadly I only have a Corsair TX 750W in my old computer available here at home. Maybe I have some friend who do have a stronger one laying around, I can check around.

If I find someone to drive me to Gothenburg for a new warranty service, do you have any suggestion of what to tell them so that they test what's need to be tested? I am not sure if they actually did test PSU last time, I asked what's inside their 24h stress test and he said they have their own setup of tests, including Unigine, Furmark, Prime95, Memtest and also playing game if that's what customer reported as a trigger of the problem. I know one thing, he told me service technician did test the GPU seperately in another rig.

I will try and call them and say my problems have re-appeared and ask if they did test the PSU I guess.

Is there some way I can test PSU without changing it to another one? I mean any software tests recording errors pointing in the direction towards PSU or anything. And if there are some settings that can be changed in BIOS to "tweak" the PSU or something to maybe fix the problem.



First off, that is the best written and most informative "PC problems" post ever. Also being your first post is awesome!

On to the issue at hand. Echoing the post above, I'd test with a new PSU. Before that, though, I'd check another thing. Even though it doesn't appear to be a RAM problem, run memtest86 overnight to check for errors. Any error is bad.

Another thing to do is use Ryzen Master to reset any changes you made to the system to default. Then uninstall the program. Next, reset to defaults in BIOS.

I would also load up BF1 as it really uses a lot of CPU and RAM.

Thanks for those words! I'm getting very tired of the problem and want it solved as quick as possible and if possible also solved here at home without the need of a new service. So a good opening post I know is always important, as much information as possible and you - the experts - can value what's needed and not in the troubleshooting process :)

I will check and run Memtest86 and report back the results.

I have not changed anything with Ryzen Master or in BIOS since I got it back from service where they did a reset of BIOS and reinstalled F5j version of my BIOS. I've only used Ryzen Master because I've read it is the best program to track real and accurate temperature in the Threadrippers CPUs.
But I can try reset everything just in case - because I did react when I got PC back from last service - it boot up as if it did a shut down (longer boot time with max fan speed in the beginning) and also when I was back into Windows my fans where still going with my own manually set fan curves. Usually when BIOS is updated or reseted - all the fan curves disappear, so unless the service technician did set fan curves again I was a bit suspicious they did not reset BIOS. The default fan curves are horrible - with my Noctua fans far too close max speed making a lot of noise.

Finally I see that I wrote wrong information in opening post - It's Battlefield V I have here on the PC, Battlefield 1 was on my PS4. But I guess it's equally as demanding - it was a game that worked perfectly fine to play after the release for me at least so if it makes computer shut down as well - I know for sure problem is within my computer at least right :) So I will test this and try run a 2 hour+ gaming session and see how it goes.

Once again, thanks to all your support!
 

Flourix

Commendable
Jan 4, 2020
9
0
1,510
So, I have tried to play some Battlefield V now. Maximized settings with DXR and everything.
Temperatures were all fine with GPU around 75 Celsius and CPU around 50 Celsius.

What struck me was that CPU voltage now looked stable and didn't spike as much up to the 1.4 and 1.5v. So the peak speed was at about steady 3.9-4.0GHz

nUliycd.jpg

xR4mQ2P.jpg

gAhHRfo.jpg

I was wondering if it was a good sign since it has never been like that before. And the only thing I have done so far since the opening post is actually to switch the PSU-cable to another power socket in the wall - because I thought of it could be my powersockets being old or dusty/damaged whatever causing spikes.

However, the computer did shut down after all. And it was at the exact 2 hour mark of the gaming session.

When playing Call of Duty it's about the same time it shuts down and the only way I have been able to get past it and play for more is if I have been quitting and restarting the game like every hour. It seems to postpone the shut down.

I might go for a try in a less demanding game now and see if it will be a shut down around 2 hour mark again - for example in World of Warcraft. I already know a almost no graphical game like Football Manager 2020 gets past 2 hours.
If it happens, what conclusions can we make?

I'm also going for the reset settings, uninstalling Ryzen Master, resetting BIOS tests tonight - and after I've done that I will try play a new game session in Call of Duty.

And for the night run the Memtest86 as suggested and report results back tomorrow morning.

Edit: Is too many shut downs like this and the type of testing that we do dangerous for my components and hardware in any way? I mean can we continue try and fix this with our changes or should it be handed in to warranty service as quickly as possible?
 
Last edited:
So, I have tried to play some Battlefield V now. Maximized settings with DXR and everything.
Temperatures were all fine with GPU around 75 Celsius and CPU around 50 Celsius.

What struck me was that CPU voltage now looked stable and didn't spike as much up to the 1.4 and 1.5v. So the peak speed was at about steady 3.9-4.0GHz

nUliycd.jpg

xR4mQ2P.jpg

gAhHRfo.jpg

I was wondering if it was a good sign since it has never been like that before. And the only thing I have done so far since the opening post is actually to switch the PSU-cable to another power socket in the wall - because I thought of it could be my powersockets being old or dusty/damaged whatever causing spikes.

However, the computer did shut down after all. And it was at the exact 2 hour mark of the gaming session.

When playing Call of Duty it's about the same time it shuts down and the only way I have been able to get past it and play for more is if I have been quitting and restarting the game like every hour. It seems to postpone the shut down.

I might go for a try in a less demanding game now and see if it will be a shut down around 2 hour mark again - for example in World of Warcraft. I already know a almost no graphical game like Football Manager 2020 gets past 2 hours.
If it happens, what conclusions can we make?

I'm also going for the reset settings, uninstalling Ryzen Master, resetting BIOS tests tonight - and after I've done that I will try play a new game session in Call of Duty.

And for the night run the Memtest86 as suggested and report results back tomorrow morning.

Edit: Is too many shut downs like this and the type of testing that we do dangerous for my components and hardware in any way? I mean can we continue try and fix this with our changes or should it be handed in to warranty service as quickly as possible?

For me, it still points to your PSU. Hence, random shutdowns. Sometimes it's hard to replicate the load, but when it starts randomly, it eventually becomes more frequent with time. The only way to be sure is to test with another PSU. You could purchase a multimeter and test it like that, if you feel you can do it, it's not that hard, and there are online guides to do it. With decent 850w+ Gold 2 x eps connector PSU's available at good price points, i'd prob just go that route. Getting the multimeter will just confirm the PSU issue, and then you still have plump for the cost of the PSU plus the multimeter.

No, too many shutdowns isn't really too hard on the components, except maybe the HDD/SSD's. It can have an effect on the OS though, and can corrupt both the boot loader and the OS at startup, which often results in having to re-install Windows, which can be a pain.
 
Solution

Flourix

Commendable
Jan 4, 2020
9
0
1,510
For me, it still points to your PSU. Hence, random shutdowns. Sometimes it's hard to replicate the load, but when it starts randomly, it eventually becomes more frequent with time. The only way to be sure is to test with another PSU. You could purchase a multimeter and test it like that, if you feel you can do it, it's not that hard, and there are online guides to do it. With decent 850w+ Gold 2 x eps connector PSU's available at good price points, i'd prob just go that route. Getting the multimeter will just confirm the PSU issue, and then you still have plump for the cost of the PSU plus the multimeter.

No, too many shutdowns isn't really too hard on the components, except maybe the HDD/SSD's. It can have an effect on the OS though, and can corrupt both the boot loader and the OS at startup, which often results in having to re-install Windows, which can be a pain.

Oh yeah maybe it is possible to get back to the young age testing with those guides and a multimeter. But the cost, time and energy consumption to do it is maybe not something I have right now with all the work I have at the moment.

I think I better find a friend to drive me to the store for a warranty service again and point out clear to them I want my components tested and not just in the stress tests this time.

Reinstalling Windows is fine for me, but is something I want to avoid ofc since reinstalling all other software is a pain and preparing uploading files to save to a Online Drive.

Anyways, I'll try the BIOS reset and everything else we talked about tonight at least.
 

Flourix

Commendable
Jan 4, 2020
9
0
1,510
Just a quick update.
After turning PC back on after latest shutdown it got stuck in the boot process with left side of RAMs dark (no led lights) and red lights on MBU lit. I was afraid that they were damaged in some way but a new restart and PC booted up again.

I went directly into Memtest after that occurred instead of more gaming testing after the BIOS settings reset. And I'm currently 1h 40m in and one pass, no errors. I'll leave it on for the night now and within the upcoming days try and get computer back for a warranty service and full component check.
 
It might be worth testing with the 750 watt PSU that you have around, to see whether that makes any difference. It might not have an ideal capacity for long-term use, as I could see that system's power draw potentially getting up over 600 watts with both the GPU and CPU fully loaded, but it might be fine for testing for a while, especially since today's games are not going to load most of the cores of that CPU, and I don't think there would be many applications that would heavily load both the graphics card and processor simultaneously either.
 
  • Like
Reactions: Roland Of Gilead