Built a rockin' new system and dual boot vista 64 and win xp sp2. Specs include new nVidia graphics card, Q6600 cpu, eVGA 680i mobo, and 2 sticks of 2gb ocz 800 mhz ram. However it's unfortunately not stable and I can't figure the source of the problem. The crash is a system seizure (freezes). But the screen will freeze first and the sound will keep playing for all of about 5 seconds, and then turn into a buzzing noise. The crash occurs with both Vista and XP (vista has nothing installed). The frequency is about once an hour. It will occur closely after a previous crash. (this lead me to think it was overheating, but it's usually running at 55 celsius which is normal for a quadcore). I've tried underclocking the processor and the memory but it makes no difference. I have had one BSOD which mentioned the realtek onboard HD audio with my eVGA 680i mobo. Currently I'm thinking I can try updating the mobo bios to try to fix whatever problem there was with the onboard sound. I can also try using just one of the sticks of ram and try different sockets. The reason I'm posting here is that while it is very hard to locate a system instability, this crash does seem strange in that the sound keeps on playing, also the crash is not affected by usage of system resources (ie it doesn't immediately crash when running a game or a burn-in program like Prime95). It does seem tied in with usage of the USB ports. If anyone could point me in the direction of a guide to narrowing down what is causing a system instability, or has any ideas, please let me know. Thanks!

To help us understand the problem you'll need to list your system specifications. CPU: Motherboard: Graphics Card: Sound Card: RAM: Hard Drive(s): CD/DVD Drive(s): Power Supply: Operating System: Driver Versions: i.e. sound card, graphics card, chipset

Are you sure you have the ram in the right slots? E.g. dual channel ram must be installed in specific slots, often specific pairs of slots, the pair being of the same sizes and a matched pair bought as a pair. See your mboard manual. Your ram must be compatible with your mboard, particularly it's main chipset. If you haven't already done so, go here and look up which specific ocz modules are compatible with your mboard: http://www.ocztechnology.com/suppor... If your mboard make and model is listed, but the ID no's of the ocz modules do not include the ones you bought, whether they will work properly in your mboard is a crap shoot - they may not. If your mboard make and model are not listed, a mboard with the same main chipset will do, but that will probably take some detective work to figure out. Even if your ram is compatible, it may have connection problems because it is not fully seated in the slots, or there may be something on it's contacts even if you can't see anything visible on them. E.g. even just fingerprints on the contacts can cause problems, and you can have problems with brand new ram. See response 2 in this - try cleaning the contacts on the ram modules. http://www.computing.net/hardware/w... If you use a ram testing utility, such as memtest86, do that AFTER you have checked the ram is properly seated and you have cleaned it's contacts, otherwise any errors it finds may not be valid. If you DO get errors after that, "bad" ram is extremely RARE - it is a lot more likely the ram is incompatible with your mboard, or you have it in the wrong slots. Is your power supply used? Are you sure it has enough capacity? If it was new, what brand and model is it, and how much did it cost? El-cheapo PSs are a lot more likely to cause you problems. Failing power supplies are common and can cause your symptoms - e.g. they can certainly overheat. Check your PS. See response 4 in this: http://www.computing.net/hardware/w... Did you load the drivers for your mboard, particularly the main chipset drivers, after Windows Setup was finished? If you didn't your mboard probably won't work properly. You must load the drivers for the main chipset so that Windows has all the proper drivers for and information about your mboard.

RAM voltages should be set accordingly in the BIOS to match the volt specifications of the RAM itself. If voltages in your BIOS match the voltages that your RAM is rated to run at safely then just leave the mobo model that way we can tell if the RAM is compatible with the mobo. Check the thermal compound job. Did you use the thermal tape that comes with the CPU or did you use your own thermal paste. You'll want to make sure you only applied a paper thin layer of the paste either by using the size of a bb amount and placing the heatspreader onto the CPU or you can spread it with your fingers to make it paper thin but you're not supposed to do it with your finger as this can cause small air pockets which can disrupt thermal dissipation. Too much thermal paste can cause memory errors and random restarts leading the operator to believe the RAM is faulty, the RAM is a non-compliant part, or that the issue is PSU related. After leaving your mobo model also tell us where you bought the RAM or tell us the RAM model number so we can see the specifications. After ruling that the RAM is compatible and the voltages are set in your BIOS to match the voltage of the RAM, we may be able to proceed to increase the voltage of the RAM in the BIOS until we possibly regain stability of the system.

When you instal a CPU you should really use a special guide or tool,are you shure you havent damaged the cpu?

Run memtest86 or memtest86+. Test the sticks individually, then as a pair. Errors are unacceptable & means the RAM needs to be replaced. I believe your CPU runs at 1066MHz FSB (266MHz), correct? Then your RAM should also run at 266MHz (DDR2-533). If you're running your RAM at 400MHz (DDR2-800), it's out of sync with the CPU & performance isn't optimal.

"When you instal a CPU you should really use a special guide or tool" Oh really? Most CPU sockets are what are known as ZIF (Zero Insertion Force)...in other words, the CPU just drops right in place. No force or tool or guide are necessary.