System crashes after 10-15 mins of stress test

heyfranky

Reputable
Jun 21, 2014
3
0
4,510
Alright, This is a long and drawn out story so I will give as much details as possible.

First off the build:

AMD FX(tm)-8350 Eight-Core Processor, 4000 Mhz
MOBO: asus m5a97 R2.0
DDR3 8 GB
NVIDIA GeForce GTX 750 TI
PSU: EVGA 600B 80PLUS Bronze Certified ATX12V/EPS12V Active PFC 600W Power Supply
WD Blue 1 TB Desktop Hard Drive: 3.5 Inch, 7200 RPM
Windows 10 upgrade from win8.1

Ok lets start here I built this about a year ago this time around. Everything was fine. about 3 weeks ago, my comp started randomly freezing. Since I just finished updating League of Legends to its previous patch (5.23) I thought maybe it's because I needed to run it on compatibility. after 3 more tries, no luck.

I figured it might be overheating or something similar to that since the other methods of using windows 8 compatibility mode of win7 didn't work.

I started watching the temps on my cpu using HWmonitor and as I noticed it would hit 55-58c while playing league.

I remembered that I never got around to getting a better heatsink than the stock one for the 8350. Went to frys picked up a bigger heatsink. picked up the evercool hpq-12025.

This is where it gets weird. I take off the old heatsink, and the thermal paste (cooler master, red lable around it, grey color. LINK: http://www.technogog.com/wp-content/uploads/2013/05/seidon15_thumb.jpg) gets on the mobo socket and the 6-8 of the cpu pins as I remove the cpu to see if the thermal paste went into the mobo socket. Me removing it definitely did get some on the cpu pins as well as the mobo cpu socket as stated above, not much though. Looked online, found other victums such as myself and got some 91% ispropyl alchohol, an art brush soft enough to remove the paste off with some alchohol lightly added to it.

2 hours later, I got most of it off, I would say about 98% off in my eyes. Put everything together and everything seemed fine as it was booting. I got a screen that said "overclocking failed" or something similar to that asus logo came on for a couple of seconds. I went to bios, set everything back to default and rebooted. Everything seemed fine, once again, booted and got to the desktop.

I used HWmonitor and watched the temps and everything was fine, started researching about the voltages and everything and it decided to freeze again. just browsing. I then decided to just upgrade to WIN10 to see if it was my system files or anything related to that.

After lengthy upgrade...

monitored closely using HWmonitor, I downloaded prime95 to see if the heatsink would work well.

20mins into stress test and kept at 52c. no issues until 28min and then got BSOD. WHEA_UNCORRECTABLE_ERROR.

I then started to research the error, thought it was the thermal paste, so I took it all apart and cleaned it once more, put it back together and then back to stress test and monitor the voltages.


This is where I was before it froze 30 mins ago (its 12am PST this happened at 1130ish.)

according to HWmonitor (even though I read these things arent accurate)

ardware monitor ITE IT8721
Voltage 0 12.21 Volts [0xED] (+12V)
Voltage 1 5.06 Volts [0xEC] (+5V)
Voltage 2 1.10 Volts [0x5C] (CPU VCORE)
Voltage 3 1.66 Volts [0x8A] (VIN3)
Voltage 4 1.54 Volts [0x80] (VIN4)
Voltage 5 3.12 Volts [0xD0] (+3.3V)
Voltage 6 0.64 Volts [0x35] (VIN6)
Voltage 7 1.54 Volts [0x80] (VIN7)
Voltage 8 1.69 Volts [0x8D] (VIN8)
Temperature 0 32°C (89°F) [0x20] (CPU)
Temperature 1 27°C (80°F) [0x1B] (Mainboard)


I ran a stress test, boom, happened again (no BSOD this time) after 14 mins at 53c. All the cores running on 100%. I used the CPUZ stress test this time.

I got my power supply tester and went to work since I knew I had one lying around somewhere. I bought it when I first bought all of the parts for the build.

This is what I got when I connected the 24pin and the pcie 6-8pin connector into the psu tester:

-12v 11.7
)))))))))))))))))))))))))))))))))))))
+12v2 12.0
))))))))))))))))))))))))))))))))
5vdb 4.9
)))))))))))))))))))))))))))))))))))))))
pg 260m
)))))))))))))))))))))))))))))))))))))))
+5v 5.0
)))))))))))))))))))))))))))))
+12v1 12.1
)))))))))))))))))))))))))))))))))))))
+3.3v 3.3
))))))))))))))))))))))))))))))))
I am all out of ideas to why this is happening. Any feedback or suggestions are welcome. Thanks for reading and Let me know if you guys need more details.
 

superstition

Distinguished
Feb 27, 2009
150
0
18,710
Look at your VRM temps in HWINFO64 when under load. You may need to put a fan on the VRMs because your board is a weak 4 phase.

Your freezing could be due to the VRMs getting overloaded from a lack of active cooling. 8 core chips really like having six to eight phases and a fan on those.

If your VRM temps are OK then you may have a bad PSU or bad RAM. My bet is on the PSU.

Try to swap the PSU with a spare and see if things improve.

Also run 8 instances of Windows memtest.

If not, you may need to replace the board because of getting paste into the socket. Be absolutely sure your CPU pins are clean.