Computer shuts off randomly while gaming or intense use

March 5, 2014 at 09:03:47
Specs: Windows 7, 2.3GHz Core 2 Quad Q6600 4GB RAM DDR2 800
Troubleshooting I have done:

Speedfan and OCCT Installed

Swapped Power Supply and Graphics Card

Removed and Cleaned Heatsink and reapplied thermal paste

Temps dropped during stress test after these tasks from 87C to 50C but unit still crashes after only a few minutes. Please help.

Intel Q6600 Core 2 Quad
Asus P5N32-E SLI
4GB DDR2 800 Patriot Extreme Ram

See More: Computer shuts off randomly while gaming or intense use

Report •

March 5, 2014 at 10:02:49
Try running HWMonitor instead. Monitor both the temperatures and the voltages. Voltages should run +-5% of the stated value. Post back the results. Get HWMonitor from the link below.

Did you thoroughly clean off all the old thermal paste? See the second link below to learn the correct application method for your CPU. You don't need to use Artic Silver to use those instructions.

Report •

March 5, 2014 at 10:41:22
Temperature and Voltage HWMonitor
I may or may not have followed those instructions perfectly while reapplying the thermal paste, especially because this Zalman tech heat-sink doesn't like to sit still while you screw it down. The computer has been running perfectly for over 5 years until today, just crazy to think how it happened so abruptly. Let me know if you can see the screencast, if not I will embed a photo of the temperate/voltage results from HWMonitor.


Report •

March 5, 2014 at 12:14:29
If I click to full screen I can read the values OK. Is that speedfan or HWMonitor?

There is one negative temperature listed so the program isn't picking up all the sensors right.

The Radeon is getting up there but my understanding is that the GPU can run that hot.

Your voltages are within the 5% range.

What power supply are you using?

I would next check the RAM. Try snapping each RAM module in and out 4 or 5 times while the computer is unplugged. If you still experience the same symptoms then run memtest86+ from a boot device. Get memtest86+ from the link below.

Report •

Related Solutions

March 5, 2014 at 12:39:48
That was HWMonitor. Hmmm, anything I can do to correct that sensor? Got the same reading from Speedfan too. I had just finished playing a round of a game before I took this as well, could explain the temp on the GPU.

I was using a Thermaltake 700w. When I believed the PS was the issue, I swapped it with a BFG 750 SLI Ready Supply and still had the same result. Will perform the tasks suggested and report back to you.

Report •

March 6, 2014 at 09:02:48
Running memory tests now...this screen share feature is terrible though. Is there anyway for me to just upload a picture to the thread?

Report •

March 6, 2014 at 10:50:58
No there isn't. You can link to an online site to post photos like photobucket or dropbox. I don't need to see the results though. No errors are acceptable, so if you get any, you should pop the modules in and out as I suggested above then run the test again.

If you STILL get errors then you need to test each module, one by one with all the others removed. Lets see how it goes first. You need to run more than one pass though. Let the program finish scanning all the addresses.

Report •

March 6, 2014 at 12:01:23
The memtest+ has been running for just shy of four hours and shows 3 passes and 0 errors. Assuming this can isolate that the RAM is not the problem?

Report •

March 6, 2014 at 15:28:45
Well the next thing to check is your hard drive. Do you know the brand?

Report •

March 6, 2014 at 18:06:07
I have a Samsung 830 SSD as the my main and 2 additional HDDs.

Report •

March 7, 2014 at 04:09:49
I don't believe Samsung has a fitness test of their own. You can use the Hitachi fitness test.

Do you have SMART enabled in the BIOS?

Have you refit the Heat sink yet?

message edited by OtheHill

Report •

March 7, 2014 at 06:38:59
No I do not have any of the smart features enabled and will be refitting the heatsink today. I actually work for Samsung, so I could reach out to our product there any specific I should ask from them?

Report •

March 7, 2014 at 13:07:49
If the drive passes the SMART check then it is probably OK. SMART has a series of checks that it performs. The fitness tests do the same thing only in more detail. If any of the tests are out of range they are reported.

I never asked before but have you checked the BIOS settings for when to shut down the computer. There is usually a temperature setting for the CPU, for instance. If you exceed that setting the computer may shut down.

You need to monitor the temperatures and voltages in real time by having HWMonitor open in the tray. You can then see what happens when stressed. The temps might spike or the 12V rail could fall. Either condition might cause a reset.

Going back to the memory. Every component in your computer uses different blocks of RAM. If there is an issue with hardware the computer may not reset until that hardware/memory is accessed.

Your specs state 4GB of RAM. Is that 1, 2 or 4 sticks. If more than 1 are they perfectly matched?

Report •

March 7, 2014 at 16:26:27
I ran a stress test through OCCT after cleaning and reapplying thermal paste. I played a GPU intensive game for roughly 35 minutes, than ran a GPU test which typically would forced a shutdown in 2-3 minutes...System ran for 15 minutes just fine so I switched the test to Power Supply and within a couple minutes it shutdown...I did notice the processor core temps spiked (relative to the GPU test or Actual Gameplay) but they never went above 46C and according to Intel they should be good up 71C.

Each RAM stick is 2GB, same brand and model. Will check the BIOS settings and report back. Thanks again for your assistance thus far.

Report •

March 7, 2014 at 18:58:51
Note on the memory, two of the same sticks do not comprise a matched set, purchasing a matched pair in one package does. To get the best performance and least problems, it is important to purchase memory in matched sets to run in dual channel mode correctly.
Run Memtest from a bootable CD to test your memory and this will also prove that some other components are without issues as well.
When running stress testing within Windows, keep HWMonitor open to see if any of the voltages drop as well as look for temperature spikes.

You have to be a little bit crazy to keep you from going insane.

Report •

March 7, 2014 at 19:40:47
Both RAM Sticks were in the same package when purchased, yes. I ran the memtest yesterday with successful passes. Will run a test again to watch for voltage changes, only temp spikes were on the processor cores as stated above. Thanks!

Report •

March 8, 2014 at 06:31:59
You watch for temperature and voltage changes when using the computer, not when testing the RAM. Check the settings in the BIOS for shutdown and warning temperatures. Also, enable SMART.

Report •

Ask Question