🔒 Locked

Computer Instability

Started by Buzzy Jan 11, 2005 at 8:34 PM 20 replies 2.7k views

Original Post

Jan 11, 2005 at 8:34 PM

Over the last month and a half my computer has been getting more and more unstable. Initially I attributed it to Windows being dumb, but things have been getting worse. A month ago I tried updating my video card drivers to the new(est) official NVIDIA Forceware 66.93. Half-Life 2 and some other games refused to run, so I downgraded back to 61.77. I was getting BSODs complaining about the driver, occasionally memory (page fault) errors, and sometimes some interrupt errors. After a while, I figured it could be a problem with my memory, so I pulled out one (of two) stick of memory (the older, smaller stick). This seemed to fix that problem. About three weeks ago, I got a new processor. I installed it myself, with a new heatsink/fan. Again Windows became very unstable around this time. So I remounted the HSF making extra sure to do everything right. I've run stress test benchmarks to try and induce a crash to see if it was the CPU, but the temperature never gets above 53 degrees celsius (~128 deg farenheit) and it hasn't crashed when I tried to make it. I'd like to assume it is not the problem. Again I thought it could be the memory. So today I bought two more DIMMs and plugged them in. When I tried to turn it on, nothing. Wouldn't even boot. I messed around, trying each stick individually, changed an IDE cable (when the hard drives started complaining), changed around the power connectors on hard drives, etc. In the end, after over a half an hours work and much frustration, it worked... In the exact same configuration that I started with... Now after numerous reboots, it's still not entirely stable. In fact I'm sure it's very unstable. I talked to my dad who has been working with computers for the last 30 odd years, and he suggested it could be either my power supply or my motherboard. I thought I'd see if anyone here has any idea of what it might be. So what do you think? Feel free to ask if you need to know anything else. P.S. Nothing in this computer has been overclocked... well except the video card, but that was using the built in (coolbits) overclocker that tested itself to make sure it was stable. And I never left it overclocked for very long anyway (30 minutes max). Computer Specs: Asus A7V-133 Mobo 300W PSU AMD Athlon XP 2100+ (Palomino) 512MB (x3) PC-133 SDRAM (Unbuffered, non-ECC, CL3) MSI Geforce FX5200 AGP Video Card 80GB UATA100 Hard Drive (7200RPM, 8MB Cache) 200GB UATA100 HDD (7200RPM, 8MB Cache) SBLive Value, Two IDE Controllers, Realtek 100Mbps NIC Two case fans running at 5000RPM, plus HSF on the CPU DVD-RW & CD-RW Computer is ALWAYS ON, and has been for years... :P --Buzzy

Raduprv

Jan 11, 2005 at 8:35 PM

What about the GPU temperature?

NeonCheese

Jan 11, 2005 at 8:52 PM

I had similar problems about 6 months ago when trying to play certain first-person shooters on my PC (BSODs during gameplay, random reboots, etc.). It took me awhile to find out the root cause. I reformatted my hard drive and reinstalled everything only to find that the problem had not gone away. I doubled my RAM, upgraded my processor and even tried reconfiguring portions of the BIOS.

At the time, I had a 350W power supply, too. I kept reverting back to the idea that my power supply was sub-par for all the hardware I was running. The guy at the computer shop insisted that couldn't possibly be the reason behind my PC issues. Taking his advice with a grain of salt, I went ahead and bought a 450W power supply.

I haven't experienced the above problems since.

Buzzy

Jan 12, 2005 at 12:36 AM

Quote:
Original post by Raduprv
What about the GPU temperature?

I don't actually know how to find this out. I'm not sure if my GPU has a sensor on it or not, but I don't see anything in the driver settings, and I don't see anything in SiSoft Sandra. I don't think this would be the case, but it wouldn't hurt to check.

@NeonCheese: I'm not sure if my power supply is not powerful enough to run everything I have in my case now, or if it's just starting to give up the ghost from almost constant use for three and a half years, but I think I'll try getting a new one of them and see how that goes.

Salsa

Jan 12, 2005 at 12:51 AM

I feel your pain. I recently moved all of my hardware to a new case (a Shuttle XPC), and since my computer randomly decides to completely lock up for absolutely no reason. No symptoms whatsoever.

etothex

Jan 12, 2005 at 12:57 AM

Quote:
Original post by Buzzy
Quote:
Original post by Raduprv
What about the GPU temperature?

I don't actually know how to find this out. I'm not sure if my GPU has a sensor on it or not, but I don't see anything in the driver settings, and I don't see anything in SiSoft Sandra. I don't think this would be the case, but it wouldn't hurt to check.

@NeonCheese: I'm not sure if my power supply is not powerful enough to run everything I have in my case now, or if it's just starting to give up the ghost from almost constant use for three and a half years, but I think I'll try getting a new one of them and see how that goes.

There should be something in the NVIDIA control panel, I can't tell you exactly since I'm not in windows right now, but it's somewhere in there on my 5700 Ultra so there should be one on the 5200 also.

Though, with all the stuff crammed in there 300W might not be enough. If you've got a crummy power supply it might be straining. There's nothing particularly draining about your system, it's just there seems to be a lot of components.

Try running the system with some stuff like one of the HDs powered off. If it helps you need a bigger PSU, prob. 350W to 400W, and a higher quality one too. Once you've ruled out overheating and software, PSU is the next guess. After that, RAM, but I have never seen a quality stick of ram fail before, and as you said, you replaced it.

PnP Bios

Jan 12, 2005 at 1:00 AM

Quote:
Original post by Salsa
I feel your pain. I recently moved all of my hardware to a new case (a Shuttle XPC), and since my computer randomly decides to completely lock up for absolutely no reason. No symptoms whatsoever.

bad ground?

Dmytry

Jan 12, 2005 at 1:42 AM

You might need to look at "ultimate boot CD" , you can use it to test your hardware.

Yes, the most probable cause of problems is dying power supply or mainboard. Had same with my brother's PC due to bad mainboard... have noticed that twisting mainboard a bit fixes some of problems [grin].

oliii

Jan 12, 2005 at 3:11 AM

jeez, that's weird. My housemate's two PCs keep crashing, my PC keeps crashing, every body elses PC keeps crashing. I don't believe my machine still suffers from a new years eve hangover (but me on the other hand ...) [grin]

Or maybe I'm being more sensitive because I now also experience the same problem...

I'd check out the PSU. 300 watts could be a tad low and maybe now the PSU is raising a white flag and giving up.

Goober King

Jan 12, 2005 at 3:29 AM

Dude don't go around buying memory when you can test it first. That Ultimate Boot CD linked above should be your first stop. Touching hardware risks hardware.

My guess is your mobo memory controler could be fritzing out. That might explain both why removing one DIMM helped and others fail to work at all. The power supply could also be heading south. Giving you unstable power or what not. Just because something has served you well doesn't mean it cant fail on you.

Raduprv

Jan 12, 2005 at 6:24 AM

Try to touch the GPU just after the computer crashes, see if it is hot or not. And while you are at it, check it's cooler and fan.

Eelco

Jan 12, 2005 at 6:49 AM

i agree with your dad: mobo or power supply. dont know which one though, your symptoms could be caused by both.

try taking something power-draining but non critical out, like your vid-card, and see if that helps.

Dmytry

Jan 12, 2005 at 6:51 AM

also, if many computers crash together, it might be caused by voltage peaks (in wallplug).

Goober King

Jan 12, 2005 at 9:45 AM

Quote:
Original post by Eelco
i agree with your dad: mobo or power supply. dont know which one though, your symptoms could be caused by both.

try taking something power-draining but non critical out, like your vid-card, and see if that helps.

How in the world is a video card NOT a critical part? Many systems wont even turn on without one and its not going to boot. All in addition to not being able to see what the hell is happening even if it could.

Erik Rufelt

Jan 12, 2005 at 11:54 AM

My computer started to crash like every time i tried to play a game recently..
blue screen and reboot..
its kinda old, like 3 years, so i opened it and vacuum cleaned it from all the dust, and after that it has worked perfectly
must have been overheating or something

Buzzy

Jan 12, 2005 at 12:44 PM

Thanks for all the replies everyone. I appreciate it. Also thanks for pointing out the Ultimate Boot CD. I've never really had a need in the past (so I never looked for something like this), but I have wished on more than one occasion that such a thing existed... And it does!

Also, Goober King, I didn't exactly just go out and buy the memory as replacement, assuming it was the memory. I was planning on getting the memory anyway (to fill up my computer), it just happened to be at a (in)conveniant time.

Something similar happened to a friend of mine about six months back and he replaced his power supply which fixed the problem. We bought our computers at the same time, so I know that, time wise, it's not unlikely that it could be the PSU.

Eelco, for the time being, I'm running with my CPU and memory clocked down in the BIOS "safe mode", which has done the trick so far, but I'm not exactly pushing to get a crash right now. I'm gonna try out the Ultimate Boot CD now to see what kind of things that can tell me.

Thanks again to everyone who replied.

--Buzzy

Eelco

Jan 12, 2005 at 1:08 PM

Quote:
Original post by Goober King
Quote:
Original post by Eelco
i agree with your dad: mobo or power supply. dont know which one though, your symptoms could be caused by both.

try taking something power-draining but non critical out, like your vid-card, and see if that helps.

How in the world is a video card NOT a critical part? Many systems wont even turn on without one and its not going to boot. All in addition to not being able to see what the hell is happening even if it could.

all computers ive had over the last years had onboard vga-crap. also, a lot of people still have 4mb pci vga card thingies lying around.

however, downclocking memory and cpu is an easier way to do it, and it seems it has given the needed info. luckily new power supplies arnt too expensive.

Saruman

Jan 12, 2005 at 1:18 PM

Sounds exactly what used to happen to my old system before I upgraded to a nice 480W Antec PSU.

oliii

Jan 12, 2005 at 1:50 PM

I found why my machine crashed actually. Or one of them...

I ran memtest86 all day, and it found a failed memory module. from address

0x000100fd618 , ... 0x000100ffc48.

overall, 10k of dodgy memory. out of 256meg! that's so enraging, I can't do anything about it but replace the module... grrr.

Goober King

Jan 12, 2005 at 2:18 PM

Quote:
Original post by oliii
I found why my machine crashed actually. Or one of them...

I ran memtest86 all day, and it found a failed memory module. from address

0x000100fd618 , ... 0x000100ffc48.

overall, 10k of dodgy memory. out of 256meg! that's so enraging, I can't do anything about it but replace the module... grrr.

I think mine has a small bad spot but thankfuly has caused me no problmes.

Topic Locked

This topic has been locked by a moderator. New replies are not allowed.