Strange Issue

[bandieramonte]

I also tried this, setting the voltage of the memory modules to the most stable level according to manufacturer specs. But system is still very unstable.

I just received a blue screen saying: Page_Fault_In_Nonpaged_Area

I'm researching onto this. At least now the system is detecting 4 cores from the Q6600.

[bandieramonte]

Okay so if it was shorting against the case then it might have been enough to damage the motherboard. At this point all you can really try to do is reset the CMOS, drain the caps, and start it back up.

Every time I reset the CMOS, and restart I receive a message saying: CMOS Cheksum Bad (is this supposed to be normal?), and I have to press F1 to go to BIOS setup and set again the memory voltages to 2.1v (according to OCZ)

And how can I know for sure if the motherboard is really damaged? There are no bulging capacitors, and would a damaged motherboard even allow system to boot to windows even for 5 minutes before freezing?

[Major_A]

I really don't know how to fully test the board.

Anecdotal story:
I had a P4 that I ran for 5 years. I bought an aftermarket cooler to put on it and when I went to try and boot the PC it gave me all kinds of errors. Somehow in the process of removing the stock cooler and replacing the aftermarket cooler I killed the motherboard. It would randomly boot but the BIOS was corrupted. I pulled the CPU, RAM, etc... and everything was fine but the motherboard was toast.

[bandieramonte]

You know what? Immeadiately after my system began to boot successfully, it never crashed. But at that time, I was using the stock heatsink, so I changed it to the aftermarket one, and the crashes begun immediately and only 2 cores were being detected.

So I switched back to the stock heatsink, and the same happened. I reseated this heatsink and the 4 cores were being detected but the crashes persist. So I got a very similar anecdote to yours.. do you think that I damaged my motherboard just because I switched to the aftermarket cooler?

I see I maybe will have to replace a component here, but I'm still unsure on whether if the motherboard is damaged, or if the processor is damaged (because it had only 2 cores detected, and now 4, does this mean that the processor is damaged?) but maybe none of them are damaged. I must find a way to be completely sure on what is damaged and what is fine.

I just did the capacitor draining you suggested, and so far my system has not crashed for almost an hour... I'm doing massive backups of all files, of almost 40GB and my system has not crashed. Maybe the draining of capacitors solved this?

[Major_A]

Maybe. I guess only time will tell. Did the aftermarket cooler you used have a backplate?
Example:
http://www.newegg.com/Product/ImageGall ... U%20Cooler

Sometimes the insulation isn't installed correctly and can short against the motherboard. I don't think that your CPU is damaged but there is really only one way to find out. You'll have to try it in another LGA 775 board and see how it behaves.

[bandieramonte]

How exactly a backplate? Does it go at the back of the motherboard? In that case, it does not use a backplate (I got the Gemini II cooler)

Right now I'm using the stock heatsink, and system has still not crashed, and I'm doing an AVG scan flawless.

And I forgot to say that I did something very very stupid 2 hours ago. When system was only detecting 2 cores, I turned on the system without any cpu cooler installed, to go directly to the BIOS and see if the 4 cores appeared somewhere and to see if the heatsink had something to do with this issue. But when I went to the hardware monitor section in the BIOS, I saw that the CPU was at 95 degrees celsius and my system immediately shut down. Could this have caused any damage to the processor? It's a GO revision.

[Major_A]

Likely not, it is designed to turn it's self off to prevent damage. Was any part of the Gemini touching a component on the motherboard? Maybe rubbing against a capacitor or something? Otherwise I have no idea why it would've hurt the motherboard unless it was overtightened somehow.

[bandieramonte]

Likely not, it is designed to turn it's self off to prevent damage. Was any part of the Gemini touching a component on the motherboard? Maybe rubbing against a capacitor or something? Otherwise I have no idea why it would've hurt the motherboard unless it was overtightened somehow.

The Gemini II was not touching a single part of the motherboard, in fact, I used rubber between every screw and the motherboard to separate every metal part of the cooler with the motherboard. I have no idea what happened.

After 2 hours of stable use. I restarted the machine, and again it's crashing. There must be something going on, I doubt that the motherboard or processor is damaged if I just could use the machine for 2 hours, doing an antivirus scan, watching videos over the web and backing up 40GB of files.

[Major_A]

Are you recording the blue screen errors and looking them up on Google? I've diagnosed an intermittent problem I had that way to a faulty IDE cable.

[bandieramonte]

Yes, and I did not like the description of this one PAGE_FAULT_IN_NONPAGED_AREA. It says that it can be due to faulty hardware, probably RAM related. So I'm doing memtest runs with each memory module individually. The problem with this is that switching memory modules is causing my system to stop booting again just as before, and would not read the DVD drive where I got the memtest ISO. So I'm restarting the system many times till it can randomly boot correctly and run memtest. So far, it passed with the first memory module. I'm trying to boot to run memtest with the second ram memory.

If the test passes too, then I would have no clue on why wasn't it booting with that memory module if it passed memtest.

[Darkstar]

you can run memtest with all the mem in the system, it will tell you which one if any is bad.

[Wifuzzy]

Every time I have had to fix a comp with the problems you are describing...it has allways been the motherboard. Random wierd boot errors, does not restart... just shuts off. Usually followed by video errors or lockups, then more ramdom problems. Honestly if I had to guess, when you put in the heatsink, something happened. All it would take is a liitle static discharge and the motherboard is morally wounded.
the spark you see and feel when you touch a door knob is around 25,000 volts...a 100-500 volt discharge you would not even know happened.
Or it could just have been time for the MB to bite the big one. Time for an RMA or a new one.

Just my opinion

[bandieramonte]

Ok I ran memtest with each of the 4 memory modules individually, and they passed. I ran in dual channel mode and passed too. I ran with 3 of them, and pass. But when I run with all them 4 at the same time, it never passed..

And each time I touched the motherboard, I always grounded myself by touching a metal which goes to the floor, a really doubt whether if I damaged the motherboard with a static discharge, but maybe a could have without realizing it. Another anecdote I have is that 2 years ago, when I was first building this machine, I had the exact same problem, that I solved by repositioning the motherboard in the case a thousand times till it stop freezing randomly and worked fine for two years. It seemed to me that at that time it was shorting or something with some part of the case, till I managed to randomly prevent the shotings in one of the "repositionings". But this time I got the motherboard outside the case, over an ample cardboard paper, and I can't see why it's freezing.

I mean, is there any way to be completely sure on whether if the motherboard is damaged or not? Why would the ram modules return errors when the 4 are installed at the same time?

[kenc51]

Try increasing the chipset/nb voltage a tad. Sounds like 4x dimms are putting strain on the chip set.
Have you the memory timings set to auto/by spd in the bios?
You might have to increase the timings / latency a little with 4 sticks.