Computer switching off, no errors, suspect either CPU or Power Supply

[RAtki1]

Hi all,

I'm new to the community, but didn't know what else to do, I can build computers reasonably well, but have difficulty dianosing issues.

I have had issues in the past with Intel stock heatsinks popping off on one corner, causing the computer to power off without warning, but the heatsink is fine in this build.

My specs are as follows:

Corsair Obsidian 750d case with good air flow

Intel Core i7 4790k Haswell Refresh

ASUS Maximus VII Hero motherboard

32GB Corsair Vengeance Red DDR3 1600MHz memory

EVGA GTX980 SC 4GB graphics card

(Doesn't seem to be a point to discussing storage, I don't think that storage is to blame)

Corsair RM1000 1000W power supply (this is separate for a reason, which I will explain below.

I have read that Core i7 4790k processors can overheat resulting in the system switching off, but in the "stickied" thread on this forum it states that overheating occurs very quickly, usually 1 minute into operation. My system runs for hours under load before switching off. At the weekend when I am off work, I like to watch Netflix and play Minecraft, typically I do both at once as I have a dual monitor setup, full screen Minecraft on one monitor and full screen Netflix on the other and my computer seems to be able to handle this just fine for several hours, but then will randomly turn off without warning.

I am able to restart the system immediately and it runs for several more hours without issue, again with two full screen applications running simultaneously.

There is a possibility that I have just discovered, apparently there is an issue with the RM series of Corsair power supplies relating to their Zero RPM fan mode, at about 40% load the fan in the power supply is supposed to kick in and cool the power supply down, but the RM series power supply can (depending on when manufactured) have a design flaw that prevents this from happening (so they overheat and switch off), I have tested 2 RM1000 power supplies and the same random switch off occured with both of them.

Is there any possibility that my Core i7 4790k might be overheating? I know a lot of people recommend after-market coolers, but I don't overclock and everything is set as default by my motherboard BIOS.

Can you think of anything else that might cause of switch off after several hours of operation under load?

Really appreciate any help or suggestions. Thank you.

[ppara5]

I've never seen such a situation, but I know how I would debug it. I would find a tool which allows me to create log files of temperatures. I have no idea if that tool is Intel Desktop Utilities, Speccy, or another one, free or paid, but some tool has to have logging capability. Also, I would look in Microsoft's Event Viewer for the events preceding the shutdown.

And I would not limit myself to the PSU and CPU. It is possible for memory, graphics card, motherboard, SSD, or anything with device drivers to have problems over time. If you have four RAM sticks, remove one pair and run for a while. If the problem persists, swap the pairs. Also, remove the graphics card for a day or two and run on Intel graphics. If the problem disappears, it is probably the EVGA card.

Is this system running with a UPS? If not, then you cannot rule out power problems.

Have you run IPDT? Keep in mind that its real-time messages are more detailed than the summary displayed at the end.

https://downloadcenter.intel.com/download/19792 Download Intel® Processor Diagnostic Tool

As for aftermarket coolers, I would not run a 95W or greater CPU, let alone an i7, with a stock cooler. And I don't play games at all.

[RAtki1]

Thank you for the quick reply.

Given the unusual nature of the problem, this has had me confused, but I cannot sit by and let my computer keep switching off (in case it's doing damage).

The reason that I have been confused is because my computer can run graphically intensive games (like Rise of the Tomb Raider, on high settings) all day without switching off.

It's only when I started to run a game and Netflix at the same time that it started to switch off, which is what led me to a power supply issue, logically the more power your computer needs to draw from the power supply, the hotter the power supply would get and once I found out about the potential Corsair RM series power supply design flaw, I thought I had pinned the problem down.

When I have some time off work, I will attempt what you have suggested, I have some other things to try first however.

I have downloaded version 26.6 of Prime95 and will stress test my CPU this weekend, I will also run the IPDT this weekend too and update my motherboard BIOS to the newest version.

I am hoping that I can sort this out without being too invasive, I would rather not dismantle my computer if I can help it, because it needs more time than I can spare at the moment.

It's just strange that I can run a full screen game on high settings all day without issue, clearly there is some point where something is getting too hot or drawing too much power and it's switching off, but I have to really stress it for several hours to make this happen.

Further information, I do not have a UPS (I have never needed one) and I am using 4 sticks of DDR3 1600MHz memory, but will try it with 2 sticks when I have the time.

As for the stock cooler, it would difficult for me to install an after-market liquid based cooler because I would have to take out the motherboard and fit a back plate behind it to attach the cooler to, this is option I considered while building it and it might need to be back on the table. To be honest however, with the cost of the Intel Core i7 4790k being so high, I trusted Intel to supply a heatsink/fan that would do the job appropriately (especially since I am not overclocking).

Thanks again.

[ppara5]

"I would have to take out the motherboard and fit a back plate behind it to attach the cooler"

Yes, that's the usual problem. In the future, only buy cases with a cut-out in the tray to allow access to the back side of the motherboard.

[IUman]

Hello BannikUK,

Thank you for joining the Intel communities.

Have you seen this happening with any other application?

For example: If you run any other application or a website other than Netflix or playing any games.

Have you checked the temperature in your system when running these apps?

Try the IPDT that paramountain suggests. The IPDT will check your processor for errors

The best way to test if this is an issue with your processor is to test your processor in another system to see if you get the same behavior.

Best wishes,

Ivan

[RAtki1]

Thanks for the reply, I just thought I would post an update, in case it helps.

When I got home from work I updated my motherboard BIOS to the latest version (nothing special about the BIOS, it just said "improved system stability").

I also ran a program called Core Temp while running the Intel Processor Diagnostic Tool and at 100% load on all 4 cores, the temperatures did not exceed 75 degrees per core and my processor passed with flying colours, the system did not switch itself off. Looks like my CPU is OK.

I have just installed a new graphics driver too, and I'm considering the possibility that the problem might be software related (i.e. the Netflix Windows 10 Metro app), but I'm not sure how I would test this. Of course the problem could also be Minecraft since that's what I run at the same time.

I received a full refund from Amazon for the second power supply, they didn't even question it, perhaps they knew about the RM series design flaw.

I don't know if I should purchase another model of power supply, I don't want to keep spending money changing parts trying to work out which part is defective.

I'll keep researching and trying to figure it out, meanwhile if you can think of any other causes, I would be grateful and thanks again for your time.

[ppara5]

Speaking of PSUs, I believe the best manufacturer is Seasonic, using nothing but Japanese capacitors. Corsair used to OEM from Seasonic -- I have a 5+ year-old HX 650W which is really good -- but now it OEMs from vendors which use lower-quality Chinese parts. The RM is made by CWT (Channel Well). The Corsair OEM for PSUs below RM is Great Wall.

[IUman]

About issues with software related, you can try reinstalling your drivers to make sure all drivers have been correctly installed; you can download the drivers for your motherboard model at:

https://www.asus.com/Motherboards/MAXIMUS_VII_HERO/HelpDesk/ https://www.asus.com/Motherboards/MAXIMUS_VII_HERO/HelpDesk/

NOTE: This link is being offered for your convenience and should not be viewed as an endorsement by Intel of the content, products, or services offered there.

Try using the onboard graphics instead of the graphics card to see if the video card is not causing the problem, if you connect the monitor to the onboard you can update the graphics driver for the Intel® HD Graphics 4600, you can download it here:

https://downloadcenter.intel.com/download/25749/Intel-Beta-Graphics-Driver-for-Windows-7-8-1-10-15-40-?v=t Download Intel® Beta Graphics Driver for Windows® 7/8.1/10* [15.40]

[RAtki1]

A further update - success!

I never thought this would be the case, but my CPU IS over-heating, BUT ONLY with Minecraft, and I'm not the only one with the issue.

Apparently, Minecraft depends more on the CPU and less on the graphics card, I just played Minecraft (on its own) and I can play for 2-3 hours before the switch off happens.

I believe I have my answer, the problem is not my PSU, or my CPU technically (as it only happens with Minecraft), it's Minecraft pushing my CPU temps too high and causing the shut off.

My options are:

1) Stop running Minecraft full screen (windowed Minecraft doesn't cause the switch off).

2) Stop playing Minecraft altogether (wouldn't like that).

3) Reduce the graphics settings in Minecraft (experiment, could be acceptable).

4) Keep the settings the same and try fitting an after-market cooler (like a Corsair liquid cooler, my case more than has room for it).

Thanks for all your help!

[ppara5]

At the risk of sounding like an anti-Corsair person -- I'm not -- I would recommend a cooler from Noctua or Prolimitech. I have both. If you look at comments for Noctua, the only complaints are from people who prefer black and chrome. Noctua is expensive and worth the money. Prolimitech is almost as good and their products often come with no fan so users can choose their own (Noctua makes the quietest fans). Glad you discovered the problem with your system.

[RAtki1]

You got me, I'm a massive Corsair fan, I have a Corsair PSU, Corsair RAM, Corsair keyboard and mouse, several Corsair gaming headsets.

However, I am willing to consider alternatives, I'm now looking at dual fan Noctua air coolers, yes they're ugly, but they're good, I know they are.

Only thing that concerns me is properly applying thermal paste to the CPU, however I have watched enough YouTube videos to have a decent go.

It's just that every time I watch a PC build on YouTube, they almost always have a Corsair liquid cooler, but I believe that they are only good for so many years.

To be honest, my PC is fine for the moment, I will purchase the cooler, but what I am considering now is an upgrade.

I'm considering upgrading to a new motherboard with an Intel Z170 chipset and an Intel Skylake Core i7 6700 (non-K), because I don't overclock.

I'll probably keep everything else the same and just upgrade motherboard, CPU and cooler, then I'll have a little more future proofing in my system.

It may seem like over-kill, but I've been planning it for some time.