Computer Shuts Off Randomly (Please read)

[Jooe]

I'll start by explaining this from the beginning. I purchased a pre-built PC from Newegg around late 2013. I didn't have so many problems with it. It ran perfect and there would be times (though they were rare) that it would shut off randomly, but not like this. My PC had a huge problem prior and I just replaced the graphics card and while it did resolve the issue, there was another I was not aware of. The problem my PC had, was that there was no image showing. This was most definitely due to a power surge occurring. I connected my TV to a power outlet and when I did there were sparks everywhere and it was pretty bad. TV was fried and anything that was connected via HDMI messed up. I use an HDMI splitter for pretty much anything and anything that was connected to it during the power surge, it messed up the HDMI for them. My PS4 was connected via HDMI and while it did work via remote play on the Vita, it showed no image on the TV whatsoever. Same case w/ PS3 (had to use component cables to show an image.) After not using my PC for a while I noticed my PC had the same issue. I tried everything from removing the CMOs battery, waiting five minutes and putting it back in, to changing the slot where the GPU went. Nothing. Just recently I bought a GTX 960 2GB and after installing it, an image showed, but now what happens is that it seems the computer shuts off entirely out of the blue. I downloaded Windows 10 right after it worked so I wasn't able to deduce if it may be a problem with the OS. Regardless, I went and downloaded a ton of games on Steam, Origin, and Uplay. Started downloading it League and Guild Wars 2 and next thing you know it the PC shuts off. Played Blacklist after it finished downloading and it shut off at a particular loading screen. I was able to play it once, but once I was back to the loading screen it shut off. It seems like when the PC is doing a lot of activities at once, it seems to randomly shut off. I just played Tomb Raider maxed out and it was running perfect for two hours until it shut off when I pressed CTRL+TAB to go back to the Desktop screen while playing. Before that though, I left it on overnight to download games and it didn't shut off at all. I genuinely can't figure out the problem. Whether it's an issue with the motherboard, the power supply, or the CPU. I know I wrote quite a lot, but I sincerely thank anyone who's read up to this point. Here are the specs along with the link to the Newegg item. 

 

Processor: AMD FX-8350 4.0GHz

Memory: 16GB DDR3 1600

Storage: 1TB SATA III 7200RPM

Optical Drive: 24X DL DVD+/-RW Drive

Graphics: NVIDIA Geforce GTX 960 2GB GDDR5

Power Supply: 650W 80plus bronze certified power (I tried looking up what the PSU was, and it says "GreenMe"? I did a bit of research and I found the PSU: http://www.thinkcomputers.org/reviews/inwin_greenme_650w/5.jpg )

Operating System: Windows 10 64-bit

Motherboard: Chipset AMD 970

CPU: AMD FX-8350

 

Newegg: http://www.newegg.com/Product/Product.aspx?Item=N82E16883258050

[tim81517lh]

I understand what you mean by your pc shuts off randomly after a while. I've had that problem before, but I honestly do not know for certain if I actually fixed the problem. Nevertheless, here are my suspicions:
 

1. Based on your previous problem with the power surge, there may have been a problem with not simply your graphics card but your PSU. It may also have screwed up the motherboard, but that is not for certain. With a PC shutting off randomly, the case is most likely due to the motherboard or PSU because it has to do with power delivery. Make sure you log the temperatures, CPU/GPU usage throughout your intense gaming up to the point of failure, this will help you diagnose the problem.
2. If possible, try borrowing a powersupply from a friend of yours, or bring it into local computer dealers that may be able to help you with the problem. Replace/swap your psu to find out if it's a problem with the motherboard. IMO it's the most obvious method to go about diagnosing this problem of yours.
3. Try running FurMark and Prime95 test both at once and each individually to determine if the problem is a CPU or GPU problem. Based on your experience, however, it might be CPU/mobo/PSU related. But FurMark will still help you draw lots of power from the PSU.

[amp88]

Could you try running HWMonitor for a while (maybe an hour or two) and show us a screenshot of what the Voltages section (the one which includes things like CPU VCORE, +3.3V, +5V etc) looks like, please? My thinking is that the PSU could have been damaged and it's no longer properly regulating the voltages, which manifests in erratic behaviour.

[Jooe]

 

I understand what you mean by your pc shuts off randomly after a while. I've had that problem before, but I honestly do not know for certain if I actually fixed the problem. Nevertheless, here are my suspicions:

 

1. Based on your previous problem with the power surge, there may have been a problem with not simply your graphics card but your PSU. It may also have screwed up the motherboard, but that is not for certain. With a PC shutting off randomly, the case is most likely due to the motherboard or PSU because it has to do with power delivery. Make sure you log the temperatures, CPU/GPU usage throughout your intense gaming up to the point of failure, this will help you diagnose the problem.
2. If possible, try borrowing a powersupply from a friend of yours, or bring it into local computer dealers that may be able to help you with the problem. Replace/swap your psu to find out if it's a problem with the motherboard. IMO it's the most obvious method to go about diagnosing this problem of yours.
3. Try running FurMark and Prime95 test both at once and each individually to determine if the problem is a CPU or GPU problem. Based on your experience, however, it might be CPU/mobo/PSU related. But FurMark will still help you draw lots of power from the PSU.

 

 

Could you try running HWMonitor for a while (maybe an hour or two) and show us a screenshot of what the Voltages section (the one which includes things like CPU VCORE, +3.3V, +5V etc) looks like, please? My thinking is that the PSU could have been damaged and it's no longer properly regulating the voltages, which manifests in erratic behaviour.

I honestly thank the two of you for the advice. 

@Tim, just a few questions. Is FurMark and Prime95 safe? I was suggested to download Core Tempt to monitor the temperatures for the CPU, but after doing some research I found many users stating it was flooded with viruses. Moreover, are the applications you mentioned free? I'll try and see what the programs tell me and if worse comes to worse I'll buy or borrow a PSU. 

@Amp, more or less the same question is if HWMonitor has any malware since I was recommended Core Temp, but found that it had viruses. 

[amp88]

HWMonitor (when downloaded from an official mirror) will contain no malware/viruses/trojans etc. It's a very common and popular diagnostic tool, and I use it myself. Similarly, stress testing tools (including Prime95 and FurMark) are widely used and safe. However, if you feel uncomfortable about installing them then you shouldn't feel forced into doing anything.

[Jooe]

HWMonitor (when downloaded from an official mirror) will contain no malware/viruses/trojans etc. It's a very common and popular diagnostic tool, and I use it myself. Similarly, stress testing tools (including Prime95 and FurMark) are widely used and safe. However, if you feel uncomfortable about installing them then you shouldn't feel forced into doing anything.

Sorry, I just wanted to be certain. I'm downloading the the programs as we speak. (: 

Would you say, that this could all be a fault with the OS? I was considering going back to 8 and see if the problem still persists. 

[amp88]

Sorry, I just wanted to be certain. I'm downloading the the programs as we speak. (: 

Would you say, that this could all be a fault with the OS? I was considering going back to 8 and see if the problem still persists. 

 

Honestly, I would be surprised if it was a software problem. I think that at least one component in the system (the PSU or motherboard, perhaps) has been damaged and is causing the instability. I could be wrong, however!

[ApolloX75]

I'm inclined to agree with @amp88. This definitely sounds more like a damaged PSU.

The HWmonitor readouts should confirm. Make sure you keep it running in the background and use your pc normally for an hour or more. Hopefully we can draw a good conclusion from that.

[0x1e]

I agree, sounds like a broken PSU.

CPUs, GPUs, motherboards and so on are mainly digital. Once broken they are dead. PSUs on the other hand are mainly analogue. Some will limp on even if they are damaged. However they show their damage when you ask for a high load. 

[Jooe]

http://pasteboard.co/2DV4aTXj.png

[Jooe]

I agree, sounds like a broken PSU.

CPUs, GPUs, motherboards and so on are mainly digital. Once broken they are dead. PSUs on the other hand are mainly analogue. Some will limp on even if they are damaged. However they show their damage when you ask for a high load. 

 

I'm inclined to agree with @amp88. This definitely sounds more like a damaged PSU.

The HWmonitor readouts should confirm. Make sure you keep it running in the background and use your pc normally for an hour or more. Hopefully we can draw a good conclusion from that.

 

Honestly, I would be surprised if it was a software problem. I think that at least one component in the system (the PSU or motherboard, perhaps) has been damaged and is causing the instability. I could be wrong, however!

It took quite some time to get this screenshot, but I finally got it. More times than not when the CPU was at 81 C it would shut off. It made getting the screenshot rather tedious. Here's the link. It went as high as 82-83 C. At one point, once the PC shut off it failed to turn on for quite some time. Few seconds to be frank. 

Gigabyte: http://pasteboard.co/2DVqYaUZ.png

CPU: http://pasteboard.co/2DV4aTXj.png

[Jooe]

GPU: http://pasteboard.co/2DVm5ZqJ.png <- For whatever reason it'll show the CPU screenshot unless you copy and paste. Don't know why.

[amp88]

Gigabyte: http://pasteboard.co/2DVqYaUZ.png

 

That wild fluctuation in the -12V readings (Minimum -16.256 volts, Maximum -3.2 volts) indicates to me that the PSU is faulty. A slight variation under heavy load would be expected, but that's a huge variation. The Maximum CPU VCORE reading of 1.44 volts is also quite worrying, and that's probably a contributing factor to the high CPU temperatures you mentioned.

 

In my opinion that's enough to consider your PSU the cause of the instability, and I'd replace it as soon as possible.

[ApolloX75]

The PSU readings are pretty stable other than the -12v being "high" but it's a DC negative so it's probably a common return of your 12v and 3.3V rails.

The positive readings are the important ones to watch and they look pretty stable. Your cpu temp of 80+ on the other hand is the likely culprit for your random restarts since it's hitting thermal shutdown.

Either replace your cooler with an aftermarket one if you haven't already or its time to dust out your case and change the thermal paste.

*Edit. I was thinking maybe the 5v vcch being low could be a problem, but apparently it's a commonly misreported reading for many power supplies. Same with abnormal negatives.

I'd stick with the overheating as a root cause for now. Fix your Temps and see if the problem goes away.

[Jooe]

That wild fluctuation in the -12V readings (Minimum -16.256 volts, Maximum -3.2 volts) indicates to me that the PSU is faulty. A slight variation under heavy load would be expected, but that's a huge variation. The Maximum CPU VCORE reading of 1.44 volts is also quite worrying, and that's probably a contributing factor to the high CPU temperatures you mentioned.

 

In my opinion that's enough to consider your PSU the cause of the instability, and I'd replace it as soon as possible.

 

 

The PSU readings are pretty stable other than the -12v being "high" but it's a DC negative so it's probably a common return of your 12v and 3.3V rails.

The positive readings are the important ones to watch and they look pretty stable. Your cpu temp of 80+ on the other hand is the likely culprit for your random restarts since it's hitting thermal shutdown.

Either replace your cooler with an aftermarket one if you haven't already or its time to dust out your case and change the thermal paste.

*Edit. I was thinking maybe the 5v vcch being low could be a problem, but apparently it's a commonly misreported reading for many power supplies. Same with abnormal negatives.

I'd stick with the overheating as a root cause for now. Fix your Temps and see if the problem goes away.

@Amp. I see. I was possibly considering borrowing or buying a PSU and see where to go from there. I don't know which PSU would be the best, but I was considering this. 

http://www.amazon.com/EVGA-Warranty-100-W1-0500-KR-Replacement-computer/dp/B00H33SFJU/ref=sr_1_5?s=electronics&ie=UTF8&qid=1439317598&sr=1-5&keywords=600+watt+power+supply

http://www.amazon.com/EVGA-BRONZE-Warranty-Supply-100-B1-0600-KR/dp/B00EON40CS/ref=sr_1_2?s=electronics&ie=UTF8&qid=1439317598&sr=1-2&keywords=600+watt+power+supply

Could you please recommend some. The most I'm concerned about is getting a good, affordable PSU for my current rig. I don't intend to upgrade any time soon.

@Apollo, I'll definitely consider that, but do you feel there's no way the PSU is at fault? The most I've done for the rig is just upgrade the GPU and upgrade to Windows 10. Everything was left the way it was. Any coolers that are there are the ones that came in the pre-built here:  http://www.newegg.com/Product/Product.aspx?Item=N82E16883258050

I did switch the ram around and move removed the CMOs battery for five minutes and put it back in. That resets the BIOs I'd imagine, right? I was thinking that maybe the BIOs just needed to be updated, but probably not. I'll definitely change thermal paste and see where it goes from here. My biggest concern was that as a result of the power surge, possibly some of the components messed up and I'd have to replace either the PSU, CPU, or motherboard.

Edit: Apollo, I found a cooler that I'm considering purchasing. What are your thoughts? http://www.amazon.com/Cooler-Master-Hyper-212-RR-212E-20PK-R2/dp/B005O65JXI/ref=sr_1_1?s=pc&ie=UTF8&qid=1439318926&sr=1-1&keywords=cpu+cooler

[ApolloX75]

@Amp. I see. I was possibly considering borrowing or buying a PSU and see where to go from there. I don't know which PSU would be the best, but I was considering this. 

http://www.amazon.com/EVGA-Warranty-100-W1-0500-KR-Replacement-computer/dp/B00H33SFJU/ref=sr_1_5?s=electronics&ie=UTF8&qid=1439317598&sr=1-5&keywords=600+watt+power+supply

http://www.amazon.com/EVGA-BRONZE-Warranty-Supply-100-B1-0600-KR/dp/B00EON40CS/ref=sr_1_2?s=electronics&ie=UTF8&qid=1439317598&sr=1-2&keywords=600+watt+power+supply

Could you please recommend some. The most I'm concerned about is getting a good, affordable PSU for my current rig. I don't intend to upgrade any time soon.

@Apollo, I'll definitely consider that, but do you feel there's no way the PSU is at fault? The most I've done for the rig is just upgrade the GPU and upgrade to Windows 10. Everything was left the way it was. Any coolers that are there are the ones that came in the pre-built here:  http://www.newegg.com/Product/Product.aspx?Item=N82E16883258050

I did switch the ram around and move removed the CMOs battery for five minutes and put it back in. That resets the BIOs I'd imagine, right? I was thinking that maybe the BIOs just needed to be updated, but probably not. I'll definitely change thermal paste and see where it goes from here. My biggest concern was that as a result of the power surge, possibly some of the components messed up and I'd have to replace either the PSU, CPU, or motherboard.

Edit: Apollo, I found a cooler that I'm considering purchasing. What are your thoughts? http://www.amazon.com/Cooler-Master-Hyper-212-RR-212E-20PK-R2/dp/B005O65JXI/ref=sr_1_1?s=pc&ie=UTF8&qid=1439318926&sr=1-1&keywords=cpu+cooler

That cooler is excellent. The 212 is a great balance between performance and price. As for your PSU it's more than possible a surge could have damaged it, so if you can borrow a known good PSU and do a test then by all means please do, it will help narrow things down immensely. Just because the voltages are stable doesn't mean the unit has no issues, since a problem may not present itself until a damaged component heats up for instance. And software monitoring is never 100% accurate.

 

And yes the CMOS battery being removed resets the BIOS to defaults, but it is not likely that a BIOS update will solve this type of issue.

[amp88]

If you can borrow a PSU that would be ideal, as it would hopefully give you good diagnostic value. If you decide to replace the PSU there are a number of options. If your budget can stretch to it, you may want to try and get one from the 'Tier Two Class A' list here, such as a SeaSonic G Series 550 or an XFX XTR Series 550W.

 

I agree with ApolloX75 - the 212 is one of the best CPU coolers out there in terms of price/performance and should be perfect for your needs.

[tim81517lh]

I haven't heard anything bad about FurMark and Prime95 personally. But the general idea is the same if you try other benchmarking/stress test programs. Prime95 kicks up a lot of heat especially with the Haswell-E, sometimes to dangerous levels. But it should be fine with your processor.

 

I don't think 80+ degrees Celsius would cause thermal shutdown, it would cause thermal throttling but shutdown would be higher temps (100+ I think, don't quote me on this though). To assess thermal throttling, use AIDA64. AIDA64 is a decent CPU stress test, it also does GPU stress test so you can try that as an alternative to both FurMark and Prime95.

 

The ideal case would be to borrow a PSU to thoroughly test out your system. Or try to bring in your PSU to Customer Support and indicate to them the problems you have, see if they can identify such problems.

[Jooe]

 

I haven't heard anything bad about FurMark and Prime95 personally. But the general idea is the same if you try other benchmarking/stress test programs. Prime95 kicks up a lot of heat especially with the Haswell-E, sometimes to dangerous levels. But it should be fine with your processor.

 

I don't think 80+ degrees Celsius would cause thermal shutdown, it would cause thermal throttling but shutdown would be higher temps (100+ I think, don't quote me on this though). To assess thermal throttling, use AIDA64. AIDA64 is a decent CPU stress test, it also does GPU stress test so you can try that as an alternative to both FurMark and Prime95.

 

The ideal case would be to borrow a PSU to thoroughly test out your system. Or try to bring in your PSU to Customer Support and indicate to them the problems you have, see if they can identify such problems.

 

 

 

If you can borrow a PSU that would be ideal, as it would hopefully give you good diagnostic value. If you decide to replace the PSU there are a number of options. If your budget can stretch to it, you may want to try and get one from the 'Tier Two Class A' list here, such as a SeaSonic G Series 550 or an XFX XTR Series 550W.

 

I agree with ApolloX75 - the 212 is one of the best CPU coolers out there in terms of price/performance and should be perfect for your needs.

 

 

That cooler is excellent. The 212 is a great balance between performance and price. As for your PSU it's more than possible a surge could have damaged it, so if you can borrow a known good PSU and do a test then by all means please do, it will help narrow things down immensely. Just because the voltages are stable doesn't mean the unit has no issues, since a problem may not present itself until a damaged component heats up for instance. And software monitoring is never 100% accurate.

 

And yes the CMOS battery being removed resets the BIOS to defaults, but it is not likely that a BIOS update will solve this type of issue.

Alright, I just received the new cooler and installed it! Was my first time doing it. Was a bit tedious, but nonetheless I installed it. Have to say it's a night and day difference. The PC never, ever makes obnoxious fan noises anymore. Everything seems cool and no random shut downs so far. When it was idle it stayed around 20 C. I've been playing roughly 30 mins to an hour of Crysis 3 and it's stayed around 38-40 C. Huge difference from before when after only playing Crysis 3 for five minutes it'd reach 83 C and turn off. Thanks so much guys, I greatly appreciate it. Here's the screenshot, btw. http://pasteboard.co/2Hx9teu7.png

[amp88]

I'm really glad to hear that your problem appears to be fixed. The difference in temperatures there is huge. I'm also sorry that it turns out my advice above (that your PSU needed to be replaced) was wrong.

[Jooe]

I'm really glad to hear that your problem appears to be fixed. The difference in temperatures there is huge. I'm also sorry that it turns out my advice above (that your PSU needed to be replaced) was wrong.

No need to apologize! It was a possibility the PSU could have been damaged based off everything I mentioned. I'll get around to replacing it with one of the PSUs you suggested. (: Thank you very much.