Random restarts

[cRazy-]

Lately I have been having random restarts (usually when im playing Neocron). I disabled the automatic restart feature in WindowsXP and now instead of restarting the computer just freezes to a blank black screen. I've tried reinstalling Windows XP and my motherboard, GFX drivers, but too no avail.

I dont think its a heat issue, my CPU temperature is at a constant 60 degrees C and my GFX card is also low. Recently however its been happening every minute whenever I log into Neocron and I cant see whats wrong at all.

Any help is appreciated thanks.

System specs for this Computer:
Windows XP
AMD Athlon 2000+ XP
1GB DDR ram
Motherboard: Gigabyte GA-7VAXP
GFX card: Nvidia GeForce 6600

[J45]

tho you do have an amd, i think 60 degrees is to high personally

maybe get a good heatsink and fan

a clean install of windows will help too

[Panhandler]

most likely your vid card is getting too hot, slap a volcano heatsink on it and all your problems are gone

[cRazy-]

Im starting to suspect its a CPU overheating problem. I'll try repasting it and fixing a new heatsink and fan, see how it goes.

[cRazy-]

Update:

Whenever I go into Neocron my CPU temperature goes up too around 68-70 degrees Celcius. Plus Im also getting a lot of quick buzzing, bleeps and screeching every couple of seconds.

Im fairly sure its the heat problem now, but if anyone has any suggestions let me know.

[WaR eYe]

i Got almost the same pc, only an amd athlon 1700+
same gforce, 1gig ram, but i use allot of things to keep it at
40 degrees Celcius. cause i had the same problem when it goes to 60-70 degrees Celcius.

So keep it cool :D

[Toxen]

Yar its a heat problem, should be idling around 40 peaking at 50, whats the air flow through your system like.

[cRazy-]

Pretty good. The case has 3 fans total, one at the front takes air in, the 120mm at the back and the fan at the side sucks it out.

Its been this way for ages but something must of happened to cause it to start overheating, probably just general wear and tear over the months has made the cooling system much less efficent then it once was.

[Toxen]

Most likely dust buildup on the cpu heatsink and fan, pop it out give it a good clean, clean out the dust else where. Rember used compressed air.

[Tratos]

Might be a daft question but has dust and crap managed to build up over time to gradually cause the overheating?

[cRazy-]

I have. I cleaned everything out a couple of days ago when the problems first started, still happening.

[[TgR]KILLER]

shut down the comp. let it cool down.

then open up the thing take off your cpu cooler its pretty easy cus its an amd <3

clean the cpu and the heatsink. apply a new thin layer of pretty cooling paste ( right word ? :x )

put everything back together and see what it did.

amd shouldn't ever get that high. my old amd 2200+ ran a 200 mhz OC and never got above 50 even when it was damn hot in the room. my new amd 3700+ runs 40 idle to and runs on 2.53 ghz 300+ mhz oc

but the above should fix your problem. had alot of comps @ work with the same heat problem and most of the times it was wrong applied paste to much/ to little or none @ all.

btw i think the bios would say comp shut down @ 80 degree's so yea if it is running that hot its no wonder why it restarts.

[CMaster]

Meh, my 2200+ Idles at around 60 and thats with a quality cooler on and a fair pile of case fans. That generation of athlons (thoroughbred cores) just seems to run real hot. Never had any overheating issues though.

[pottburter]

@crazy :

I'm not a technican, but as a w2k sysadmin i have some experience with the various reasons why windows crashes.

Having your CPU overheated will mostly result in a bluescreen for the CPU starts making errors in functions with increasing heat.

Given your random restarts i would bet my money on a memory fault.
And that your systems freezes after you deactivated the "Automated reboot after memory error" function instead of rebooting is another good hint.

With the above activated, windows restarts after it addressed a faulty memory space and generates a corresponding error report for a technican.

Have you tried running a stress test on your computer ?
There's some very good freeware out there to do this. PM me if you would like have a link.


he, who hopes this helps

[cRazy-]

@crazy :

I'm not a technican, but as a w2k sysadmin i have some experience with the various reasons why windows crashes.

Having your CPU overheated will mostly result in a bluescreen for the CPU starts making errors in functions with increasing heat.

Given your random restarts i would bet my money on a memory fault.
And that your systems freezes after you deactivated the "Automated reboot after memory error" function instead of rebooting is another good hint.

With the above activated, windows restarts after it addressed a faulty memory space and generates a corresponding error report for a technican.

Have you tried running a stress test on your computer ?
There's some very good freeware out there to do this. PM me if you would like have a link.


he, who hopes this helps

I did consider it might be a memory problem, but this is the first time I have seen the CPU temp go anywhere near 70c so I just assumed that was the problem, if not then its a very strange coincidence.

However i'll take a look into it and try running Memtest for a bit, thanks for the heads up.

[[TgR]KILLER]

Meh, my 2200+ Idles at around 60 and thats with a quality cooler on and a fair pile of case fans. That generation of athlons (thoroughbred cores) just seems to run real hot. Never had any overheating issues though.

Read what i said above..

and same for you crazy its the damn cooling paste cus it should NEVER run that hot..

and cmaster.. i had a 2200+ and it was a thoroughbred core to.. it never ran 60 even when it was 40 degree's in my room in the summer it would stay halfway 55 degrees and thats when it was OCed and ran games and stuff... a pile of casefans in the wrong order can make the air flow even worse then better to btw.. but a cooler without or with wrong applied cooling paste runs 20 degree's more then it should :p

dno what quality cooler you got.. but got a Thermal Take silent boost k7 on my amd 2200+ pure copper heatsink. and have the same cooler but a k8 version on my 3700+ now runs like a dream.



Having your CPU overheated will mostly result in a bluescreen for the CPU starts making errors in functions with increasing heat.


every bios has a setting to auto reset the comp when the cpu gets to hot.. last time i had to rush to a company were the comp get resetting to. after taking the side off the case i could see there was just a wire between the blades of the cpu fan that made it stop going. thus the cpu din't get cooled like it should and the comp kept resetting.

[pottburter]

every bios has a setting to auto reset the comp when the cpu gets to hot

Never saw that one. Only know temp alarm and temp shutdown.

[[TgR]KILLER]

every bios has a setting to auto reset the comp when the cpu gets to hot

Never saw that one. Only know temp alarm and temp shutdown.

uhm...

if the cpu gets to HOT aint that a temp shutdown ? :P at least shutdown on that board it just kept resetting when it got to hot.

anyway the temps of the proc just aren't right @ all should be way lower.. maybe more is broken to but that at least aint right lol.

[Lodar]

I myself had a similar problem. Be careful as the constant rebooting eventually damaged my harddrive (lack of shutting down properly). In the end it turned out my fan was so clogged full of dust that it was actually insulating the heatsink, making it far hotter than it should have been.

[DIS]

*just noticed he had a gigabyte mobo and an athlon 2000+ at 60 degrees constantly, could be the board is miss reading CPU temp, try updating the bios.

IMO the CPU is running a tad hot, however i recently took a look at a friends PC as he was having issues burning CD's he had an athlon 2000+, noticed his CPU temp was idleing at 60 degrees in the bios. Added silver compound solution, cleaned up the heat sink and it made very little difference, could just be an issue with those types of CPU's, saying that another friend has a 2400+ and that sits at 40 degrees....

If I were you i'd still buy a better heat sink and fan, apply some decent silver compound like Akasa or artic silver etc, run memtest from www.memtest.org to make sure the RAM is ok.
If that passes check the voltages in your bios ensure they are not erratic, ensure they are roughly within .5 of a volt, I.E if your 12v=12.5 then this could be causing stability issues.

If all else fails run prime95 on the system, if it starts flagging errors you could have a faulty CPU, an overheating CPU can be the sign of a failing CPU, could also be board related but the only way to tell that is run the CPU in a different mobo with prime95 and see if you get errors, make sure you run it under a torture test for a good hour.

A goosed board or CPU wont last 5 mins under prime 95.

[[TgR]KILLER]

if its the cpu he has a faulty one.. i had "exactly" the same cpu even ran a 200 mhz oc for more then 2 years on it and it always ran below 40 idle never got even near 60 ;(

[Argent]

KILLER']if its the cpu he has a faulty one.. i had "exactly" the same cpu even ran a 200 mhz oc for more then 2 years on it and it always ran below 40 idle never got even near 60 ;(

There are 2 (or 3 if you count Thrb A and B) different versions of 2000+ core. Palomino, which I also have, easily runs at 55-60 degrees (might actually run lower, but the sensors aren't too good, +-5 degrees inaccuracy isn't anything unusual) and is perfectly stable at that temperature as well. Throughberd generates less heat though (max: 60w vs 70w), and should run at bit lower temperatures.

[cRazy-]

lol holy crap.

I baught some £8 thermal paste, reapplied (the previous paste was on rather thick, not sure who did it) now im at 44c, TWENTY degree's lower then before.

Good stuff.

[Heavyporker]

Whoa! A 20 degree reduction just from re-pasting? That's pretty damn good.

[cRazy-]

http://www.akasa.co.uk/akasa_english/spec_page/thermal/spec_ak_5022.htm

Tempted to do some overclocking now tbh.