Sign-in / Sign-up
Your question
Solved

Repeated BSODs with no overheat or bad RAM?

Tags:
  • Homebuilt
  • RAM
  • Systems
Last response: in Systems
March 10, 2012 4:43:38 PM

I have finished building my new system with the following components:


ASRock G41M-S3 Motherboard
2 Patriot PSD34G13332 4GB 1333MHZ RAM Sticks
Intel Pentium D 945 3.4GHZ Dual Core CPU
400W PSU

While inside Windows 7 64-bit, I get BSOD's at total random, anything from "SYSTEM_SERVICE_EXCEPTION"
to "KMODE_EXCEPTION_NOT_HANDLED".

After running "WhoCrashed" almost every crash at first was from files from some driver, first being from the graphics card, then LAN, then finally my Hauppauge WinTV HVR-950q TV Tuner Card.

I reseated the RAM and ran MemTest86+ and no problems reported there.

I also checked the temp of the CPU is around 66C, which seems high but after reading other forums, many claim their CPU runs at that temp no problem.

I've tried updating every driver on the system, no go.

Also from past experience and double checking on Google, symptoms of overheating normally seem to be freezes to random reboots with no BSODs.

More about : repeated bsods overheat bad ram

Best solution

a c 78 B Homebuilt system
March 10, 2012 9:23:29 PM

The difference between random reboots and BSODs most of the time is a windows setting that lets you choose which one you want to do.

If you choose to get the BSOD, you will, if you choose to just restart, you will do that instead.

That being said, when your PSU is getting too hot internally and can no longer supply enough wattage for the PC to function normally, yes it does often cause a freeze with no BSOD or restart.

However, PSU problems can disguise themselves as problems with anything else.

What is your PSU maker and model? If it is a generic brand that has wattage listed for ambient temperatures much lower than room temperature then it could very easily be the problem. Especially so if it is a top mount PSU and pulls all the heat from the computer into itself during normal operations.

Also, you shouldn't assume newer drivers are more stable than older drivers. The oldest ones are usually the most stable. Newer drivers usually bring new capabilities and reduce stability.

Updated firmware (as with the BIOS) does, however, often increase stability especially when used with newer hardware components.

The first thing you should do is restart and go into safe mode and try hard to crash the computer.

If you can crash it in safe mode, it is probably failing hardware. If you cannot, then you should consider installing old drivers and trying to crash it inside regular Windows.
Share
March 11, 2012 4:22:03 AM

It appears that the problem may have just been the heatsink.

While attempting to diagnose the machine running Linux for a while to see if Linux would crash in a similar manor, all the sudden the heatsink unconnects on one side complete (the two push-pins).

So I unscrewed the motherboard and attempted t make sure 100% that there was no way the CPU heatsink was not on right this time, by examining the space between the CPU and heatsink and also looking at the push-pins underneath the board to see how the heatsink is making contact. All seemed well.

Suddenly even while just in the BIOS H/W Monitor, I noticed the CPU was around 20C less in temp.

So I am going to see what happens now.
m
0
l
Related resources
a c 78 B Homebuilt system
March 11, 2012 5:07:36 PM

If the heat sink was just hanging off in space, that could cause overheating problems, yes.

I am not a fan of the push pin concept from Intel. These things fail at like 50x the rate of the AMD brackets.
m
0
l
March 12, 2012 7:30:41 PM

It appears it was the heatsink the entire time.

I noticed for some reason the first 2 tries the heatsink was not actually making the clicking noise it should make when it makes full contact.

This time it did and I could see underneath the board that all 4 plastic pieces appeared exactly the same depth in to the motherboard.

When I boot it up now, it went from it's normal 60C+ down to about 45-50C very rarely touching 60C unless under extreme load and I barely ever seen it go to 61C and normally when it hits these higher temps the CPU fan seems to go faster (RPM) and the speed is usually knocked back down again.

I so far have been watching XBMC videos and YouTube in HD and even though Flash was able to somehow use 74% of a 3.4GHZ Dual Core 64-bit CPU, no BSOD or any issue. Apparently Flash player loves resources, because even XBMC at 720p, doesn't even come close to those resources. LoL.
m
0
l
March 18, 2012 12:48:16 AM

Best answer selected by redpenguin2.
m
0
l