Computer questions

[HiTech]

This 'puter of mine keeps freezing up-- primarily during gaming, but at times even during a simple pgm. like MS Word. Memory timing is fine, Event Viewer doesn't indicate anything wrong when it happens. MEM test also shows good results, even w/ Legacy USB disabled. What do you think of these temps that I logged immediately after it locked up:

Motherboard Name: Asus P5N-E SLI
Motherboard 34 °C (93 °F)
CPU 37 °C (99 °F)
CPU #1 / Core #1 48 °C (118 °F)
CPU #1 / Core #2 37 °C (99 °F)
GPU 51 °C (124 °F)
SATA HD 34 °C (93 °F)

Cooling Fans
CPU 1566 RPM
Chassis #2 1527 RPM

Voltage Values
CPU Core 1.25 V
+3.3 V 3.36 V
+5 V 5.03 V
+12 V 12.22 V
+5 V Standby 5.05 V
VBAT Battery 3.23 V

Next question is the 'puter sometimes boots from the D drive (IDE) instead of from the C drive (SATA1). It does this randomly. BIOS checks show things to be in order... that is until it happens and then the BIOS doesn't list the SATA drive!!!. To remedy this I unplug the IDE drive, reboot so the SATA is recognized, power down and reconnect the IDE drive and then reboot again. Got any ideas for a permanent fix?

 

[Sceadwian]

They temp immediatly after a lockup isn't going to show you much, even the time it takes the machine to get into bios or boot into an OS that lets you see the temperatures any chips that may have been over heating are going to have cooled to normal or almost normal temperatures. All those temperatures are very reasonable, in fact I'd say they're damn good. My PC runs much warmer and it doesn't lock, the video card goes as high as 70+ degrees under load (it's fanless case airflow only)
I would lean towards a failing video card, I had an almost identical problem about a month ago and after replacing the video card with a lower end one I had kicking around the system was rock solid stable again, the vid card never overheated at all it just started hard locking the system.
I'd go into full debug mode and start with the video card, find somethign else to replace it with for testing and run the machine on it for a little bit. Even if it's not up to snuff for games run it as hard as you can and see if you get the same kinds of lockups, if you do then it's safe to rule the video card out, at that point I'd go into bios and underclock the processor and memory bus to as slow as you can get it to go and do the stability test again. What's the vid card you have?
For the HD problem I'd check to make sure you have an updated bios, I've heard some systems having trouble with IDE and SATA drives running together (My motherboard is supposed to have this issue but I don't use the SATA drives) I have an Asus A78NX-deluxe. Probably just a fault with the way the way ASUS implemented the SATA controller on some of their boards, check google for something like "P5N-E SLI IDE SATA problem" or the like.

 

[Torben]

I know you've more than likely checked this a few times, but the last time it happened to me I was pulling my hair out until I checked the seating on all my mobo cards. Problem was the video card just needed to be re-seated.

Just a thought.


Torben

 

[dknguyen]

Never had a video card freeze up my comp- only cause mmy computer to reboot when it overheated. Does the problem go away when you take your case off? If so, probably overheated...videocard (what else?)

 

[jpanhalt]

I did a fresh install of XP a few months ago on a new drive plus a supposedly clean old drive. For whatever reason, XP decided to install some of the startup files on the old drive and the rest on the new (maybe the old ones weren't completely wiped?). That gave similar symptoms. I just removed the old (small drive). Put it in permanent storage. Formatted the new drive again, then installed a second, really new drive for the backup. Install went normally, so far as I know.

As for locking up, I have noticed the same thing with Word as well, after the fresh install. I am using Word 2002 SP3. It seems worse if I fire it up and then try to do something right away, like auto-adding the date. It's more than a bit aggravating. My work-around
is to start the body of the document first, then go back and add the date, etc. In my case, it is Word that locks up, not Windows itself.

Don't know if that helps, but I feel your pain. John

 

[Torben]

Other things to check are the power supply and mobo battery. I don't see the mobo battery causing the freezes, but it might be responsible for the BIOS forgetting where you want to boot from.


Torben

 

[HiTech]

All good points of advice given by you guys but some of it still has me wondering-- Just last night the sound locked up on me while everything else worked fine. The sound is mobo integrated, Realtek AC97. I have checked and updated all necessary drivers and the BIOS is current (except for the very latest which adds Chinese language to it and I'm not reflashing the BIOs for that useless crap). Everest Ultra performs no less than 10minute stress tests to memory, CP, GPU, and FPU showing a graph and results pane. The interesting thing is that if I stress all 4 items at once the system hard-locks quickly. If I only choose 3 of the four, it runs fine forever, with no reported issues or errors. I'll reseat the cards again.
- The power supply is a dual 12v rail 600watt (that should be enough power for the system!?!?)
- Vid card is EVGA 8600GT w/ 1gig on-board RAM
- Memory is Corsair and accoring to their tech. RamGuy, the emory is not the issue. I do have another 2gigs (PC6400 DDR2) in the mail as I type this to make for 4gigs total.
- Both GPU and CPU have Arctic Silver paste applied to them. The Northbridge chip has a large fan-cooled heatsink.

The problem is determining if it's a flaky mobo, memory, PSU or GPU card. The other possibility is a bad OS disk that I created by slipstreaming SP3 to the original XP-Pro-32bit. The old early version of XP-Pro didn't include SATA/RAID drivers, hence why I slipstreamed a new version. The BIOS is set to AUTO for the memory's Command Rate which should be 2T. If it somehow was defaulted to 1T that could corrupt OS installation files. I may need to do a fresh install again, and then load SATA/RAID drivers independently when it prompts for them (F6). I do have VISTA Home Premium which I can load, but I'm not a VISTA fan (yet).

 

[Sceadwian]

Again, remove the video card, put in a known working card even if it's a lower speed one, stress test the system. It doesn't matter if it's a PCI card or whatever. Start with the video card first because they will hard lock a system with no other symptoms when they lose it. You have to have a working PCI card floating around somewhere. Modern video cards have more transistors in them than modern CPUs silent hard crashes are a direct sign that your vid card is on the fritz, especially with a high end model.

 

[HiTech]

A high end card would be more likely to mess up than a cheapo one?? -- that's weird!! I guess the more parts the more chances of component failure. I have a new nVidia card (lower model) arriving early this week for my son's PC. I'll put that in and stress the system. When I stressed just the vid card for 20 minutes at 100% it worked fine even running at 60­ degrees C.

 

[Mikebits]

A bad OS would not cause your BIOS to not recognize a drive. You sure your supply is up to the job?

 

[ArtemisGoldfish]

A 600 watt is fine for a single-video card system, even if it's a modern card like the 8600-series. No one has seemed to mention, but rarely, it's the CPU. I've had a cracked processor (Resulted from reseating the heatsink/fan). It doesn't happen often, but it's a possibility. As others have said, try swapping out the video card. Also, go into your BIOS and check the voltage monitor.

 

[Sceadwian]

Yeah, cheapos are slow and usually based on old technology, slow as hell but there's not that much that can go wrong. Some of the higher end video cards out there use 100+ watts of electricity, and that's almost 100% heat. Even mid end cards require extra power connectors because they'd violate the card slots current limits. Heat management is a major concern for high performance chips of every kind. It could be a mini bubble in the heat sink interface has caused a tiny portion of the chip to be too hot over time but not hot enough and close enough to the chips thermistor to register. Over time the transistors under that particular portion will age rapidly compared to the rest of the chip, presto whamo one day the chip just goes... "nahh I don't think so" and starts rebooting, or hard locking or any number of a thousand other minor glitches very infrequently. They get more frequent to the point where you notice it and by then it's too late and the glitches will make the machine unuseable . That's my theory at least.