Full shutdown while gaming, but why ?

[Dreadsncatz]

Hello Guys, im new to the forum cause i'm looking for help with my gaming rig.

First of all my Setup:

OP: Windows 10 Pro

CPU: AMD FX 8350 @ factory clock

Mainboard: Gigabyte GA-990FXA-UD3 rev 1.2 at Cmos version FE

GPU: MSI Geforce GTX 1060 6gb

RAM: Corsair Vengeance 2x8GB DDR3 1866MHZ 10-11-10-30
CPU Cooler: BeQuiet! Dark Rock Advanced C1

Case Cooling: 4x bequiet Pure Wing 2 - 120 and 140 in Front, 140 at Back and 140 at top

PSU: FSP Aurum 650w Gold+

Case: BeQuiet! Pure Base 600
Storage: Seagate 120gb SSD and 2x 1TB Western Digital 

 

I experienced random Freezes or Full shutdowns (Black screen PC turns off) While Playing Subnautica or For Honor so i made some Temp readings and testings.
I Fired up Aida64 stresstest wich froze the PC after about 20min, this where the Temp Readings:

CPU: 79°C (Max 82°C)

MOBO: 52°C

GPU: 73°C
In HWMonitor i also measured TMPIN2 wich is VRM i guess wich had 93°C

 

Someone has an idea for me ?

Thanks in advance.

 

[PineyCreek]

What's an Amrum PSU?  Never heard of it, can't find it, doesn't pop up as a PSU brand on Google.  Based on that I'd say your PSU is garbage.  You're not overheating.  Maybe heat expansion causing a short or a PSU that sucks.

[Dreadsncatz]

Sorry was a Typo, the brand is FSP the series is Aurum. Its a pretty old PSU but the new one are still sold pretty often here in Germany.

[Dreadsncatz]

Update: Tested with BeQuiet! Straight Power 10 600W PSU. PC Shutdown after about 15min of Aida64, second test endet in Freeze (No shutdown) after 9min.

[TetraSky]

This might seems crazy, but have you tried with a different power cable from the wall to the PSU ? (preferably one that's 14 to 16 gauge (also known as "AWG" or American wire gauge))

I've had a similar issue where I used a "cheap" power cable, the kind that's barely enough for a 65W monitor, to power a computer and the moment I put any sort of load to it, it would shut itself down, changing the cable fixed it.

https://www.amazon.com/C2G-Premium-Universal-IEC320C13-Compliant/dp/B00009OLLS/ref=sr_1_4?ie=UTF8&qid=1529188416&sr=8-4&keywords=computer+power+cable+14

 

[Dreadsncatz]

As far as my knowledge goes we only have one diameter of Power cables in Germany. Our whole Power system works different than the US. 

But i thing i figured it out here's the story for the record:

Since the Mainboard ran into issues recognizing the RAM as 1866 i tried tweaking it for Hours and Hours with al the Bad stuff happening, like no boot at all, Cmos reset and so on. 
The board only wanted to boot as 1333 RAM, setting RAM to 1866 made it not boot due "RAM overclocking issue" and reset to 1333. 
Setting RAM Multiplier to 9,33 by and setting the Timings by hand resulted in the Same issue. Doing the same and setting North Bridge Voltage from 1,100V to 1,110V made it boot as 1866 RAM. Finally but unstable (Same issues as described before)

After hours and hours of testing, switching RAM Modules to older spare ones and so on, if figured out, and don't ask me why or how, it made the System run stable for now when i set the RAM Timings to 10-11-10-29 instead of the 10-11-10-30 like written on the RAM and bump the Voltage from 1,5V to 1,55V and NB to 1,11V.

AIDA64 ran for about 45Min straight. Tomorrow i may have time to do a longer run like 3 hours or so to confirm stability now.

So story short, i feel like the Motherboard is garbage and maybe the RAM modules are like 1 of 1000+ that are a bit faulty.

Edit: only thing is BIOS says system power not optimized, strange thing is that if i uncheck the box that says ist 1866 RAM the voltages are the same and are optimzied but it will boot as 1333

[aezakmi]

Quote

CPU: 79°C (Max 82°C)

In HWMonitor i also measured TMPIN2 wich is VRM i guess wich had 93°C

That thing is on fire, that 93 hwmonitor reads is probably the northbridge... but that temp is abnormally high since you're not even overclocking it, right? I remember my old board hit 93 and even 95 while using the stock heatsink and a frequency bump of 10%, just don't pump more voltage into it as you might fry it

 

Max safe temp for that CPU is 70, it shouldn't even hit 70 with the stock cooler so something's wrong with your cooling, just replace the thermal paste

 

Your mobo is OK as you're using stock clocks, 52C is also pretty high if you ask me, you might want to remove that rear fan and leave only the top one or add 2 top exhausts if the case allows it

 

RAM should be OK, I'd roll back to 1333

[Dreadsncatz]

Yea i thought that the HWmonitor read is the NB. I don't overclocked anything, thermal paste is new, changed it today for testing but it changed nothing.

I could put the rear fan also on the top and make another test.

 

Btw here Temps while browsing: CPU 45°, MoBo: 45, TMPIN2 (maybe NB) 42

 

What should be the purpose of the RAM back roll ? I mean its 1866 Ram and CPU and MoBo supports it.

 

Maybe its simply time to upgrade to Ryzen...