Random Crashing

[Rich]

Very often when I an running EU it shuts down without notice and comes up with an error message about how Entropia Universe did not shut down properly. I also have a more serious problem, which is that my pc shuts down and resarts randomly again, without notice. I have defragmented, error checked, virus checked and anything I could think of but it still happens and has been going on for a while now. It shuts down always when I try and open .pdf files but it also does it completely at random too.

In case it helps here are my pc specs as it says when I run dxdiag.
Processor: AMD Athlon XP 3000+, MMX, 3dNow, - 2.2GHz
Memory: 512 RAM
OS: XP Proffesional
Graphics card: NVIDIA GeForce FX 5600XT 256MB

If anyone can solve this I would be very grateful, and if theres any more info needed just shout.

 

[onyx oZ Zombie]

the only times i've seen machines just outright shut down / restart, it has been:

1) some really, really crappily written software (for instance, the Charter Security Suite, when installed, will cause a machine to reboot after entering Cisco VPN client username and password)

2) a problem with the power supply, lack of sufficient power supply wattage, or overheating

since it's happening outside of EU randomly, and also while opening pdf's, i would look to your power as being the most likely cause of your problem

 

[Rich]

Thanks I'll give it a go and see if thats the problem

 

[Nadie]

Also, check all your fans are clean and working properly and nothing is overheating. I also had this problem when my old graphics card was on it's way out. I changed the card and no more crashes or system reboots. Just some suggestions

 

[Rich]

I think it may be overheating as today was a very hot day and I couldn't even start it up. I don't know if this is hot or not but when it crashes the CPU has a temperature of 57°C but when it starts up after cooling it has a temperature of 23°C. I'm not sure what to now except get the fan replaced

 

[endermigne]

Is it possible that EU or something in DirectX or OpenGL or whatever they use stops the overheat-beeping of the board?

Because if I'm doing something which keeps the CPU busy 100% (i.e. not playing EU) I get a very annoying beeping from the board "informing" me that it's a bit warm in there. If I play EU I don't get this beeping and the PC just shuts down after a while.

Thinking this further the beeping is a step 1-scenario for protecting the hardware from serious damage. And if it's getting too hot then the PC just shuts down. Meaning, step 1-scenario allows the user to save things and such.

Actually when I get the beeping I stop the 100%-thing, keep the PC running for like two minutes and the beeping is gone.

But again, that beeping occurs only when the CPU is busy and I DON'T play EU. If I play EU there's no beeping and the PC shuts down without prior warning.

Also a sidenotes about this issue:

DON'T remove the cover from the chassis as the chassis itself works as a heat conduction and the air flow inside the chassis would break. Also it can get damn dusty in there.

At least that's what I read somewhere...

As for me I'm planning to visit a PC store tomorrow if I have the time and ask them about this. Maybe I learn more about it which I can share with you folks.

 

[Rich]

Thank you No beeping just restarts. Temp was 60 today when it crashed so I think i need another fan

 

[Bogette]

60c is on the hot side, but I wouldn't have thought it should have shut down, I would check the BIOS and see if there is setting that you can adjust. Some motherboards have a "shut down" temperature setting, I'm not sure on yours. I think the max recomended die temperature of that processor is around the 75c mark, so you have a bit of room to play with.

I would suspect the video card over the CPU, and taking the side of your machine would be where I would start. If the case is built right the airflow should be better with it shut, but I have rarely seen a computer built to properly direct airflow.

The heat from the processor will be in turn heating up the video card, take it out and clean it down, some modern graphic cards have crap standard fan/heatsink setups, particularly when a bit of dust gets in there. I burnt my finger on one because of the dust build up, and no air was getting through.


Shaz Monaro

 

[endermigne]

Some more info from my experience which may apply to others:

- after my latest crash I checked the BIOS and the temperature was around 58c which shouldn't do anything bad. But I noticed that the power unit and the graphics card were hot.

- for me I decided that from now on I won't TP to Amethera anymore as all the crashes occurred there if I remember correctly (Nea's Place, Shinook, Memorial Island)

- I remember that I crashed once while I was checking auctions. As this isn't a complex graphical thing (I think) I'm now thinking about concurrent graphic commands (if this is possible at all) EU sends to the card which it can't handle for some reason. I mean like maybe the global swirlies were about to be shown or somebody TP'd into the mobile center.

So currently I'm blaming my graphics card (ATI Radeon9800 Pro 128MB) to cause all the instabilities. Therefore I have some questions:

- Is there a way to check the RAM on the graphics card if it's still working properly?

- Is there a way to check several (if not all) functions the graphics card supports directly (hardware) and log the outcome somewhere? So like I could say this or that command makes my PC crash.


PS: I know that the initial post was about the GeForce FX 5600XT but I'm just thinking out loud. Maybe some of what I write helps anyone.

 

[onyx oZ Zombie]

Is there a way to check the RAM on the graphics card if it's still working properly?

ATI may have a diagnostics program at their website ... maybe check there?

Is there a way to check several (if not all) functions the graphics card supports directly (hardware) and log the outcome somewhere? So like I could say this or that command makes my PC crash. ...

not sure if it's what you're looking for, but dxdiag will allow you to run DirectX hardware and software accelerated tests ... you may find something useful with the outcome of the tests (then again, you may not lol)

 

[Golun]

Richi wrote:

I also have a more serious problem, which is that my pc shuts down and resarts randomly again, without notice.

some time ago, i submitted a bug reprort with exactly the same problem.

just a screen flicker and the pc is off!! full powerdown. and restart after that.

Thought it was me, bogus program. wrong hardware, u know.

MA adviced me to keep good track of the cooling of the pc and graph card. I did install a new fan and the problem on first glance seem to be reduced.

still happens though.

 

[endermigne]

Tx, onyx, I'll look into that.

Yesterday made a theory of mine vanish in no time. I had a power off close to the sweat camp N of PA...

Maybe I shouldn't play EU anymore........

 

[Theren Strange]

If the fan is clogged up it can overheat the processor and will shut down in the same manner as if the electrical supply was pulled.

Putting a new fan on the heat sink (if this IS the problem) will correct the issue - but be warned. The fan needs to be an exact replacement - fans are very different, be 100% sure its the same model as before or it may not cool the heatsink enough. The Heatsink needs to put back on the board exactly as it was before removing. The silicon gel needs to be re-applied(i think silicon) if origionally present or the heatsink wont stick to the processor enough and if theres the slightest wobble, or tiniest air gap, your problems will continue. Sometimes fans come with new heatsinks as well, and if the HS is new the gel wont be already applied but should come in the box with the hardware.

I learnt this because it was what happened to me on my last comp -

 

[Arrun]

Just come across this thread and thought - coincidence or conspiracy??

My laptop (Alienware M7700) worked fine till the second 8.12 download then the gfx card (Geforce 6800) blew (typical mem leak problem - screen went pinstriped, booted with no harddrive and got same problem) - Alienware have sent a replacement which I have fitted but now the new gfx card overheats (and it really does get fecking hot) and shuts the laptop down. Lasts about 20 - 30 mins before dying.

No problem running 'ordinary' gfx-lite apps like Outlook, Word, EF etc.

AW are sending another card so I'll see how that goes. It has occurred to me that now there's no heat-transfer-gunk between the new card and heat-sink, perhaps a factor? Certainly none supplied with the card. I dunno if just the lack of this gunk would have such a dramatic effect. I have pulled the heatsink off before to reseat the original card so any seal (as discussed above) would have been broken anyway plus the heatsink screws down onto the chassis pushing the gfx card into its slots so there 'should' be good contact between them.

Strange though that others have 'overheating' problems starting at the same time - as I say - coincidence or conspiracy??