Bizarre overheating(??) problem

[Reynard]

1. Overview of the problem
   Leaving my computer on and idle for extended periods of time causes it to croak.
   
   
2. Full description of the problem and symptoms
   When I leave my new computer on overnight or while I'm at work, there's a fair chance (but not always) that at some point while I was away, things went wrong. I return to a computer that is giving no signal to my monitor, but still clearly has power because the lights are on. Most tellingly, all of the fans are going what seems to be full tilt - the first time this happened was the first time I had actually been able to hear the fans since I built the machine a month ago. I have not been able to reproduce this. Rebooting just comes up with the generic "Windows didn't end properly do you want safe mode" screen and then works fine. Most recently I popped open the case and felt a heatsink, which was nowhere near dangerous (barely even warm), so I don't even know whether it actually overheated and the fans cooled it down post-crash or not.
   
   
3. Did it work normally at one time, or has the problem always existed?
   The first three weeks after I built it were peachy-keen. This has only been happening the past week.
   
   
   
4. Is the problem consistent and repeatable, or entirely random, or semi-random?
   Semi-random. It's only happened five or six times, but every time was during a long period (3+ hours) of no input from me, and even then it's maybe a 40% chance of happening. It has happened with programs running in the background; uTorrent and Pidgin for certain.
   
   
   
5. I already tried these steps:
   I downloaded a little heat monitor called SpeedFan that I'm not certain to trust or not (it's giving me some wacky values). It's telling me my system is generally running at 40C right now, and that fan 1 is going at approx. 1000 RPM. (Though Fan2, Fan3, and Fan4 are at 0, even though I just checked and more than one fan is definitely moving.)
   
   
   
6. My software:
   • Vista Home Premium, SP1 (Build 6001)
   • Antivirus and firewall are just the Vista-provided ones. I should probably download ZoneAlarm or something, but I've never had a problem with viruses.
   
   
   
7. My hardware
   • Mobo: Dammit, I don't recall exactly. I know it's a Gigabyte P35 model. D3SL might be it, but I looked at so many while building this it's all a blur.
   • CPU: Core 2 Duo E8400, 3.0ghz, no overclocking
   • GPU: ASUS Radeon 4850
   • RAM: G.SKILL 2 x 2GB 1066 DDR2
   • PSU: Corsair 750W
   • Cooler: XIGMATEK HDT-S1283 120mm. This monster suggests otherwise against the overheating theory. The fan is definitely turning.
   • Case: Antec P182

 

[robisbell]

try disabling the power saving for all hardware, and check to make sure the mouse, keyboard, video card, monitor, hdd, optical do not have the option enabled for the syswtem to turn them off when it believes they are not in use. when you leave, manually turn off the monitor and leave the system running and when you return check to see if everything is good or not. the power setting for the system to turn off hardware is under device manager, each piece of hard pull up their properties and check to make sure nothing in power management is checked.

 

[Webbterfuge]

if its not a OS power saving problem, try removing any dust from the case, make sure the heatsink is seated properly, and if theres alot of cable clutter clean it up. the heatsink could be great, but if its not completely in contact with the CPU then its useless. It could be your powersupply is dying, sounds kinda like something that happened to me a long time ago, but it doesnt seem like your setup would really load a 750W PSU. BTW with Speedfan the reason it doesnt show all the fan speeds is because each fan needs to be plugged into the motherboard with a special connector, so the 0 RPM isnt something to worry about.

 

[Reynard]

Huh. Windows actually managed to "recover" from the latest kaboom (it still crashed, just a bit more gracefully, I didn't have to press the power buttons), stating that something be up with my Realtek PCI-E thing. I'm not certain whether this is related to me installing the latest drivers for my Radeon 4850. I went on Windows update and installed an optional update (un)related to the ethernet thinger in question ("Realtek RTL8168B/8111B Family PCI-E Gigabit Ethernet NIC (NDIS 6.0)"), so I'll just have to wait and see if that was the problem. It seems pretty clear it's my video card overheating that is the problem... I felt the heatsink on that and it has always been too hot to touch... But I'm given to understand that the 4800s generally do run hot. How can I monitor my video card's temp? The following is Vista's little diagnosis. Oh, and it seems the optional update has nothing to do with my video card:

Problem
Video hardware error
Description
A problem with your video hardware caused Windows to stop working correctly.

Problem signature
Problem Event Name: LiveKernelEvent
OS Version: 6.0.6001.2.1.0.768.3
Locale ID: 1033

Files that help describe the problem (some files may no longer be available)
WD-20080711-0103.dmp
sysdata.xml
Version.txt

Extra information about the problem
BCCode: 117
BCP1: FFFFFA8005534010
BCP2: FFFFFA60024314F8
BCP3: 0000000000000000
BCP4: 0000000000000000
OS Version: 6_0_6001
Service Pack: 1_0
Product: 768_1
Server information: 2948d406-dcd1-482e-9162-25d5e98beb8b

 

[robisbell]

a nifty program called PC Wizard will allow you to monitor the temps and voltages, plus much more.