Sudden computer shutdowns
NeonSpyder
"Das est NTLDR?" Join Date: 2003-07-03 Member: 17913Members
in Off-Topic
<div class="IPBDescription">Tell me something I want to hear</div>Hey.
This is like the fourth time I've gotten help from you wizards to help fix my computer. Everytime my problem has been vast and perplexing, frequently defying any sanity or reason. Today might be the exception!
My computer is awesome, I love it. It's worked faithfully for me for years now, churning away happily.
Today it hates me. I had a fleet operation scheduled in EVE that I was supposed to be commanding and TODAY is the day my computer decides to crash on me.
the symptom:
Computer gives a hard shutdown with no warning while running the EVE client (potentially any other game, I haven't checked yet). It seems to be working fine and then all black with no thermal warning screech from the case, no BSOD, nothing.
Analysis:
So I suspect a thermal issue and after blowing a few dust bunnies out of the case attempt to run the client again. Runs for a few minutes, long enough for me to get the fleet going before it crashes again. I try a third time after getting some more dust out of the PSU and running SpeedFan to monitor temperatures but the damn computer crashes the second I log into the game!
I have been running windows 7 without any games for the past 15 minutes and it hasn't crashed. Interesting? but what does it mean!
Speedfan tells me my CPUs are running idle at 30-35c, extremely cool as I understand. I haven't yet checked what they are under load, but it can't be that much more.
GPU temperatures are a comfy 52c but have never been historically low. When I boot up and before the fan kicks in the temp can be as high as 70c without any load so I've been watching this but I don't suspect the GPU is overheating. Occasionally I get a glitched out driver that causes... confusing effects on the computer but that's another (unsolvable) problem. Bottom line: I don't think this has anything to do with the GPU.
Conclusion:
Power supply might be dying? If I get a sudden draw from the GPU when a game begins to run, could that be responsible? It might explain things.
What else could it be? Tell me it's just the PSU and I'll go buy a new one. That would be a best case scenario as far as I'm concerned.
Help me Natural Selection Off Topic, you're my only hope.
This is like the fourth time I've gotten help from you wizards to help fix my computer. Everytime my problem has been vast and perplexing, frequently defying any sanity or reason. Today might be the exception!
My computer is awesome, I love it. It's worked faithfully for me for years now, churning away happily.
Today it hates me. I had a fleet operation scheduled in EVE that I was supposed to be commanding and TODAY is the day my computer decides to crash on me.
the symptom:
Computer gives a hard shutdown with no warning while running the EVE client (potentially any other game, I haven't checked yet). It seems to be working fine and then all black with no thermal warning screech from the case, no BSOD, nothing.
Analysis:
So I suspect a thermal issue and after blowing a few dust bunnies out of the case attempt to run the client again. Runs for a few minutes, long enough for me to get the fleet going before it crashes again. I try a third time after getting some more dust out of the PSU and running SpeedFan to monitor temperatures but the damn computer crashes the second I log into the game!
I have been running windows 7 without any games for the past 15 minutes and it hasn't crashed. Interesting? but what does it mean!
Speedfan tells me my CPUs are running idle at 30-35c, extremely cool as I understand. I haven't yet checked what they are under load, but it can't be that much more.
GPU temperatures are a comfy 52c but have never been historically low. When I boot up and before the fan kicks in the temp can be as high as 70c without any load so I've been watching this but I don't suspect the GPU is overheating. Occasionally I get a glitched out driver that causes... confusing effects on the computer but that's another (unsolvable) problem. Bottom line: I don't think this has anything to do with the GPU.
Conclusion:
Power supply might be dying? If I get a sudden draw from the GPU when a game begins to run, could that be responsible? It might explain things.
What else could it be? Tell me it's just the PSU and I'll go buy a new one. That would be a best case scenario as far as I'm concerned.
Help me Natural Selection Off Topic, you're my only hope.
Comments
Windows 7 64bit
Intel Core2 Quad Q6600 @ stock speeds
6gb ram on three chips 3x2gb
ATI Radeon HD 4800
Stock fan for the GPU, super awesome enhanced heatsinkage for the cpu.
Check with other games if you haven't already.
I assume you don't have (m)any spare parts you can swap between to narrow the issue down?
Kernel-Power event
"The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly."
Five of 'em today. Which strongly reinforces my PSU concerns. Well, maybe not strongly, but I don't suspect any of the other components.
I'll test another 3d application to see what happens. I would run a cpu/ram stress test for longer but I'm suspecting that both of those are fine. If it were the ram it would probably give me a BSOD when it stopped working.
The psu is five years old and has more then it's fair share of dust, I'm starting to think the poor old thing is on death's door and is just begging for mercy already. It's not because of an imbalanced systems ( I don't have anything that's not critical that I can take out of the system, no extra HDD or anything) so I'll have to replace it. I'm pretty sure it's 500w but it might be 700w. Will have to check before getting a new one.
Now I suppose I have to fix it somehow. I'll continue to explore other possibilities and testing the PSU for problems but if AVP or Supcom doesn't cause it to drain the available power, EVE certainly wouldn't either so... it's probably not the PSU. Which is both good news and bad news, good because I can keep the PSU I have and don't have to shell out for a new one and bad because WTF the game is hardcrashing my computer?! Way to program CCP.
Linky:
<a href="http://www.futuremark.com/benchmarks/" target="_blank">http://www.futuremark.com/benchmarks/</a>
And since numerous other applications haven't caused so much as a hiccup in the computer I can't consider it to be anything else then the application.
I've already run a dozen different programs of varying graphics levels and intensiveness for varying lengths of times and nothing has caused a crash except EVE. In fact I just did an EVE re-install (steam) and it crashed in the first second of loading the main game window after logging in. The steam install had no remnants of the previous cache and user settings (as it had to redownload the game cache and it had default settings and no memory of log ins)
So... CCP programmers hate me and my computer. It's the only logical conclusion.
I'll continue to experiment but I've sent CCP support a petition, I guess they'll get back to me whenever they feel like it.
I'm somewhat pleased and somewhat infuriated but at least I don't have to buy any new components, but still... hard crash? Come on. That's just not nice.
Thanks for your quick help everyone. I'll let you know how it turns out and if anything new crops up.
--Scythe--
Just try downloading any windows updates for now that might help.
--Scythe--<!--QuoteEnd--></div><!--QuoteEEnd-->
Here's one I suggest: Run the full test suite. Then run the three hour bitfade test. Then the full test suite again.
Also, I advise caution: Just because memtest86+ finds memory errors, you can't be certain your memory is broken, and that replacing it will solve the problem. In one case I had accidentally undervolted my memory, which, while perfectly harmless to the hardware, did cause happy error funtime. Setting the correct voltage fixed that problem. In another case, the GPU of all things was faulty, and since it does some fairly low-level stuff with the north bridge and the north bridge also handles memory...
Anecdotally, mine went to bluescreen while doing a core dump, then rebooted. If you have the core dump disabled but have automatic restarting enabled it could reboot without ever showing you the bluescreen. However, a bluescreen should not lead to automatic shutdown. The computer should either stay on the bluescreen until the user manually intervenes, or reboot automatically.
It was the power supply.
I'm not sure why it didn't crash yesterday on doing anything but EVE, but it mislead me as to the true cause. I figured out what it was in reality because the next morning when I booted the computer up it did the same power-off crash that it's been doing, but this time it did it for a different game so that threw me 'eve is glitched' theory out the window.
So I open the case to test the ram one by one to rule out the possibility and lo and behold, the computer doesn't even want to boot up anymore! Of course if it was a problem with the ram I was testing, I would still boot but get a mobo error report and since the mobo manages to get to FF on the display (fully functional) I figured the mobo was probably fine too. So I go and get a new PSU (on sale woo!) and plug everything in. Oh my god psus are so hard to take out and also to install. The damn cables get caught on everything and the plugs are so hard... hhr, either way. Everything works now.
Yay! Thank you SentrySteve for your valuable PSU-related suggestion.
And of course everyone else for your help and clues, it helped narrow down the problem. You guys are my favouritest ever!
Just wait until you have to swap out a mainboard. You practically have to pull the entire computer apart.
<img src="http://msinfluentials.com/blogs/jesper/Disable%20Automatic%20Restart.jpg" border="0" class="linked-image" />
(Disable automatic restart on system failure)
Sometimes the bluescreen pops up so fast you dont see it and it will appear that your computer shuts off without giving an error, I see it occasionally at my job working on laptop hardware. Could also be power supply failure, a short in any part that gets power from the power supply/motherboard, overheating (check your fans/heat sinks for dust), overclocked hardware, broken USB port shorting out, incorrect settings in BIOS, might need BIOS/driver updates, theres still a slew of things that can cause it.
I highly recommend getting PC check (boot disk diag software), it will fully check every part of your hardware and let you know about a majority of problems, though note that it can false-positive while running tests that a piece of hardware does not support (new dual core processors have this problem a lot)
I once had a computer with shut down problems like that, turned out that the GPU fan was not working and the card also had bad memory.