[SOLVED] New PC Dies when CPU hits 70c. Which part is at fault?

[taikellyevans]

PC suddenly cuts out when under load. HWMonitor and AMD's Ryzen Master application both show that this happens whenever my CPU hits 70c (It's rated to handle 95 safely). Is the problem here with the PSU, CPU, or motherboard? Can I fix it myself or do I need to replace something? There's no CPU thermal protection controls in my BIOS. My PSU fan turns on during POST but does not turn on when testing load, not sure if its because the PSU wasn't hot enough or if there's a fan issue there.

My Specs:
PCPartPicker part list / Price breakdown by merchant
CPU: AMD - Ryzen 5 2600X 3.6 GHz 6-Core Processor (£189.98 @ Aria PC)
Motherboard: MSI - B450M MORTAR Micro ATX AM4 Motherboard (£98.12 @ Amazon UK)
Memory: G.Skill - Flare X Series 16 GB (2 x 8 GB) DDR4-3200 Memory (£195.59 @ Amazon UK)
Storage: Western Digital - Blue 1 TB 2.5" Solid State Drive (£134.84 @ Ebuyer)
Storage: Seagate - Barracuda 2 TB 3.5" 7200RPM Internal Hard Drive (£54.79 @ Aria PC)
Video Card: Gigabyte - Radeon RX VEGA 64 8 GB Video Card (£409.79 @ Overclockers.co.uk)
Power Supply: Corsair - RMx (2018) 650 W 80+ Gold Certified Fully-Modular ATX Power Supply (£87.98 @ CCL Computers)
Operating System: Microsoft - Windows 10 Pro OEM 64-bit

 

[Redneck5439]

PC suddenly cuts out when under load. HWMonitor and AMD's Ryzen Master application both show that this happens whenever my CPU hits 70c (It's rated to handle 95 safely). Is the problem here with the PSU, CPU, or motherboard? Can I fix it myself or do I need to replace something? There's no CPU thermal protection controls in my BIOS. My PSU fan turns on during POST but does not turn on when testing load, not sure if its because the PSU wasn't hot enough or if there's a fan issue there.

My Specs:
PCPartPicker part list / Price breakdown by merchant
CPU: AMD - Ryzen 5 2600X 3.6 GHz 6-Core Processor (£189.98 @ Aria PC)
Motherboard: MSI - B450M MORTAR Micro ATX AM4 Motherboard (£98.12 @ Amazon UK)
Memory: G.Skill - Flare X Series 16 GB (2 x 8 GB) DDR4-3200 Memory (£195.59 @ Amazon UK)
Storage: Western Digital - Blue 1 TB 2.5" Solid State Drive (£134.84 @ Ebuyer)
Storage: Seagate - Barracuda 2 TB 3.5" 7200RPM Internal Hard Drive (£54.79 @ Aria PC)
Video Card: Gigabyte - Radeon RX VEGA 64 8 GB Video Card (£409.79 @ Overclockers.co.uk)
Power Supply: Corsair - RMx (2018) 650 W 80+ Gold Certified Fully-Modular ATX Power Supply (£87.98 @ CCL Computers)
Operating System: Microsoft - Windows 10 Pro OEM 64-bit

That is very strange and should not happen. I've had my R7 2700X to 73C while running Prime 95 and have seen on may forums where people have pushed 85C (although I would be uncomfortable at that temp). I would wager that the problem is thermal in nature but isn't your processor. I would wager that your problem lies with your VRMs. What CPU cooler are you currently using? How may case fans do you have? Is there any cooling near your RAM and VRM heat sinks? When your processor is nearing 70C that forces a shut down what are your motherboard temps?

The CPU cooler is designed to cool the CPU but I have seen in many "extreme" rigs where people have gone as far as a custom liquid loop while having virtually nothing to help the RAM and VRM heatsinks dissipate their heat. Sometimes the cheaper blow down CPU coolers actually do better for the VRMs because at least there is air circulating over them. Case fans also help a lot with VRM temps as the more airflow you have the better your system will run. Most people only care about processor and GPU temps (not to say they aren't important) but VRM temps are just as important and widely overlooked. I personally have 6 140mm fans moving air through my system and all the components get adequate cooling, even the hard drives (again overlooked for cooling all the time). I would monitor your motherboard temps along with the CPU temp so you know how hot your equipment is getting. You may also want to see how hot your GPU is getting as well. HWmonitor is a great application and will show the temps of anything in your system with a heat sensor. I would suggest looking at all the temps as your system nears 65C to see if there is anything reporting abnormally high temps and if needed get more case fans to move air over all your components.

 

[taikellyevans]

I would wager that your problem lies with your VRMs. What CPU cooler are you currently using? How may case fans do you have? Is there any cooling near your RAM and VRM heat sinks? When your processor is nearing 70C that forces a shut down what are your motherboard temps?

I'm using the stock cooler that came with 2600x and I have 2 case fans, 120mm front and 140mm rear. When my CPU hits 69c during testing, all other hardware temperatures monitored in HWMonitor remain in the 30s (MOS was 44). So it doesn't appear as if anything else could be overheating.

 

[Redneck5439]

Have you updated to the newest bios for your motherboard? If you haven't check the official web page for MSI and download the latest bios. After flashing the latest bios I usually recommend doing a CMOS reset and then load any optimizations you made in bios manually, don't load a saved profile (in fact after a bios update those saved profiles can be erased and then save your new one that you can utilize until your next bios update).

 

[taikellyevans]

Updating my BIOS and drivers were the first thing I did after installing the OS. BIOS is using the default settings.

 

[Redneck5439]

Have you overclocked anything? What do you have your RAM set at? It could also be possible that your memory could be the issue you may want to run Memcheck. It could be that on AUTO your SOC voltage is going to high and overloading your memory controller, you could try setting that at 1.1V.

Other than that or possibly a faulty temperature sensor and your processor is actually running much hotter than you think I just can't think of any hardware issue that would cause this. If there is some bios related setting that is causing this somehow and setting optimal defaults doesn't work I would suggest pulling the battery on your motherboard, leave it out for a few minutes and depress the power button a few times to fully discharge the capacitors in your power supply. After just a few minutes replace the battery and see what you get when you load into Windows.

 

[drivinfast247]

I'd say it's the CPU. Try reseating the CPU in its socket. Clean of old thermal paste and apply new.

Another thing to try is to reset the BIOS .

You have all good components that should work without issue.

 

[ohiou_grad_2006]

Also look in the bios, some of them have options for alarms at certain temperatures. Perhaps there's a setting in the bios that turns off the system.

I also would suspect the vrms. On my cheaper ASRock board, it has vrm heatsinks to the side and above the socket. I'm noticing your board seems to only have them towards the one side.

 

[taikellyevans]

Have you overclocked anything? What do you have your RAM set at? It could also be possible that your memory could be the issue you may want to run Memcheck. It could be that on AUTO your SOC voltage is going to high and overloading your memory controller, you could try setting that at 1.1V.

Other than that or possibly a faulty temperature sensor and your processor is actually running much hotter than you think I just can't think of any hardware issue that would cause this. If there is some bios related setting that is causing this somehow and setting optimal defaults doesn't work I would suggest pulling the battery on your motherboard.

Nothing is overclocked, all default. RAM is at 2400, higher speed XMPs or manual settings don't help. A previous Memcheck turned up clean, Manually setting SOC didn't help either. Since it can't handle playing games on low settings that it should easily be able to handle on high, there is definitely a serious fault somewhere even if under-reported temps were the case.

My current plan of action is to borrow another PSU to test, if that isn't the problem then try to return both the CPU and motherboard as I have no way to figure out which is faulty.

 

[ohiou_grad_2006]

Good idea. Also, it should not be, but could it be the power supply? I think Vega can pull some juice.

https://www.anandtech.com/show/11717/the-amd-radeon-rx-vega-64-and-56-review/19

If I'm reading that right, under load Vega 64 can pull 410 watts. Maybe that psu just can't handle it. I'd be interested if say a gtx 1050 or something were temporarily placed in the system, if it dies.

Also, where are you placing the system? On a desk? On carpet? If you have it on carpet, try it on a desk or put a plank of wood under the PC. You might find it stops.

Back in the day when I was a kid and the power supplies were almost all top mounted, I had trouble with a PC shutting off. The long and the short of it was that I didn't have an exhaust fan under the power supply, and the computer would shut off because the power supply was overheating. Installed a little 80mm fan under the power supply to pull hot air out of the case, and the shut downs vanished. Computer was good for a while after that.