Sign in with
Sign up | Sign in
Your question

(strange) problem with my system

Last response: in Systems
Share
February 21, 2006 11:38:29 AM

Hi everybody, need some help with my PC - thank you very much in advance...

-The system: (built in summer 2005)
CPU: AMD AthlonXP 2500 (Barton) 512k L2
MB: Asus A7N8X (nForce2, dual channel, no SATA)
RAM: 2x512MB Corsair DDR400 XMS
Video: Asus Ati Radeon 9550GE, 256MB DDR

Others, in short: Seagate 160GB parallel ATA, ASUS DVDRW, older Creative Live sound card
Running XP Professional

-The problem: from time to time, the system keeps restarting spontaneously (as if the reset button is pushed); from time to time means once every few days. After rebooting, I get the usual option from Windows to send an error report; after completing the report, the page that opens gives one of the following messages:
1) Problem caused by a device driver.. unable to determine exact cause
2) Problem caused by a fault in the RAM
(sometimes one of them, sometimes the other one)

I've tested and re-tested and re-re-re-re-tested the CPU, RAM and hard drive, with absolutely perfect results (no hardware errors whatsoever)

I can live with it, but it's certainly annoying... I am simply out of ideas for the cause and/or solution of this problem.
Please.. ANY suggestions or ideas are welcome.

Thank you very much in advance!

More about : strange problem system

February 21, 2006 11:57:37 AM

Power supply perhaps?
February 21, 2006 12:04:29 PM

Quote:
Power supply perhaps?


OOOPS... Forgot to state that my first power supply was getting the Asus PC Probe monitor to display some alerts about the voltages, so I changed it.
Also, according to Asus PC Probe, all the temperatures are OK (CPU = maximum of 46-48 degrees Celsius, MB = around 30 degrees).

The strange thing is that it APPARENTLY, it restarts in a random way - not dependant on the amount of stress. It may restart when I'm playing a game, but sometimes it restarts even if nothing more than, say, Winamp is open... so it doesn't seem to depend on how hard I use it.
Related resources
February 21, 2006 1:19:38 PM

I've had this problem with my ATI 9800 AIW, and it was the drivers. Try updating to the latest if you haven't already. If you have, try rolling back to a previous driver and see if the problem goes away. I had to stay using the version 5 Catalysts for awhile before I was able to get one of the version 6's to work. Hope this helps.
February 21, 2006 1:39:24 PM

Yikes, these are never fun...

Things you may want to try. (If you have the means to)

1) Swap the RAM into a different machine, boot using a bootable linux CD, and try running memtest on the ram. (I know you have tried this, but this is a MAJOR cause of the problem you have explained.)

2) Video card, if the card is starting to give the voltage flucuations caused may be enough to trip the circut causing a reboot.

3) Video Drivers could be a major issue.

4) Improper BIOS seetings (AGP apature, etc...)

5) Motherboard Drivers, although the nForce2 setup is pretty stable.

6) Chipset cooling, if it's actively cooled, make sure the fan is running free of impeadance, as a hot chipset will definitely cause this.

Those are some major areas to look.

I hope it helps.
February 21, 2006 2:35:03 PM

@Penguin_D: well, about the memtest issue -> indeed, I've run the test, but I've done so on my own machine, I haven't tried that on other PC's.

Sorry if the following is a dumb question, guess you'll just have to forgive a newbie: Why run the test under Linux? Is it more reliable or precise that way?

Anyways, thanx a lot guys, there are some suggestions that I haven't thought of so far. Be sure I'll try them...
February 21, 2006 3:14:29 PM

I've always found it has produced better results/more accurate.

I've had sticks of ram that would memtest fine from Windows, then blue screen 2 minutes later. Those same sticks under Linux, failed memtest miserably. Stuck em back in the windows box and they would pass memtest again...

Personal preferences I guess, I've always prefered linux for troubleshooting hardware issues...

Other than that I prefer Windows, but I play games so that's the reason...
And linux is just a pain in the arse.
February 21, 2006 4:51:43 PM

I had the same problem 16 months ago with a p4 @ 2.8 GHZ..
The problem was in BIOS
Check the bios with something like this: On push on/off button, remind last action...
Sorry i'd like to enter in my bios, but i've sold my old pc 12 months ago...
Hope you'll find it. Good Luck!

HP nx9030
2 GB DDR RAM
40 GB HDD
Audigy 2 ZS
February 21, 2006 5:21:56 PM

First off I'd turn off the Windows Auto-reboot on crash "feature" if you haven't done so already (System Properties, Advanced, Startup and Recovery).

That will show whether or not a program is crashing and taking XP with it.
If it still reboots on you, I'd look towards a hardware issue.

If you're overclocking, turn things back to normal.

You can run knoppix or some other bootable from cd OS, and play around. If it restarts in there, then you probably have a power problem.

What brand and wattage is your PSU?
February 21, 2006 6:13:54 PM

I was having the same problem with my rig.....system would reboot on its own or screen would go blank.
I did a reinstall.....had suspect might be catalyst drivers....so i did not reinstall the catalyst contriol center....ati video driver only......and this solved my problem........has run great ever since.

ps:I have a 9800 pro
February 23, 2006 4:51:38 AM

@nobly: My PSU is from a manufacturer called Premier (never heard of it personally, 'el cheapo' I guess) - but I measured the parameters with a digital multimeter (including parameters under load), and it doesn't SEEM to be the cause, as the voltages are pretty stable.

As for overclocking it, I never did that up to now; it IS a distant thought in the back of my mind, but since I am not a hardcore gamer, the system satisfies my current requirements.
So I wouldn't really think about overclocking, at least until I get it to run 100% stable at the regular parameters...

@to all guys suggesting problems with the video card drivers: when I first fired up my computer, I installed the ATI drivers from the supplied CD; after a few weeks, seeing that it restarts once in 3-4-5 days, I installed the latest Catalyst version I found on ATI's website. Unfortunately, I noticed no improvement.

The trouble is that diagnosing this is a very slow process, since as I stated, sometimes there go 2 weeks without a single spontaneous reboot. So after every change that I do, I have to wait, and wait... :) 

Right now I disabled the Auto Reboot on crash feature, as NOBLY suggested.
Newbie retard question: If it were a software related issue, how should it manifest itself with auto reboot disabled? By locking up the PC? or...?

Anyways, a big THANK YOU goes out to all people who took the time to answer to me. I guess there are nice people everywhere after all...
February 23, 2006 6:41:16 PM

Quote:
@nobly: My PSU is from a manufacturer called Premier (never heard of it personally, 'el cheapo' I guess) - but I measured the parameters with a digital multimeter (including parameters under load), and it doesn't SEEM to be the cause, as the voltages are pretty stable.

....

Right now I disabled the Auto Reboot on crash feature, as NOBLY suggested.
Newbie retard question: If it were a software related issue, how should it manifest itself with auto reboot disabled? By locking up the PC? or...?

Anyways, a big THANK YOU goes out to all people who took the time to answer to me. I guess there are nice people everywhere after all...


Basically w/ Auto Reboot on, if the system goes down due to a software crash or a Windows issue, your computer automatically restarts. Which really sucks because you typically don't know what program caused it. But with it off, it'll blue screen, and you might be able to glean some info from that.

Well, since you measured the voltages under load, I guess the PSU would be excluded. Except that its random. We can't be sure because we didn't measure it when it rebooted. You know, I have a computer w/ an almost exact setup like that. Do you have ASUSprobe running? It'll throw up a warning and log any voltage errors, etc, that may happen. I think its on your mboard CD and you can probably download it somewhere.

Heh, you might want to double check the reset pins on your motherboard while you're at it..never know, might be crossing! :p 
February 24, 2006 2:23:55 AM

Im experiencing a rather similar problem. Almost identical. I built a computer about 3 weeks ago.

Pent D 920 (w/ Zalman 9500)
Intel brand 945 P mobo (the one w/o raid support)
Seagate 250 SATA
eVGA 6800 GS
2 gig Corsair Value Ram
XClio 450 Watt PSU (had decent reviews on newegg, and wasn't super cheap....apparantly the OEM Antecs as well...)

I had the computer for about a week and a half and hadn't noticed a crash. Because of higher temps i installed the zalman (just as a precaution) about 2 weeks after the first build. Then about 2 days after installing the zalman (and OCing the graphics card with it) I got a random reboot when opening HL2. This was actually on Saturday. Now its thursday evening, and the same thing happened again after playing a little CS and then surfing the web.

No errors or anything, just as he said....its like hitting the reset button.

So after seeing this post, i decided to restore my Video card to default... ( i oc'd it from 425/1000 to 485/1.07 and it appeared to be stable) and see if i get any more reboots.....if the problem persists, i guess its time to check the ram/HD/Mobo or PSU?

Oh, according to Speedfan....my voltages are as follows (not at load)
+2.5: 1.55 V
Vcore: 1.25 V
+3.3: 3.32V
+5: 5.10V
+12: 12.38 V

after running 3d mark 05, the +12 dropped to 12.25 (if that means anything....)

Anyways, im not trying to steal the post...but it just made sense to post here considering it was a similar issue
February 24, 2006 2:29:04 AM

Oh...a little more info.

Im running XP Home, and upon reboot i do NOT get the option to send an error message. Its as if the computer just restarted normally.

Also, i've heard the occasional click inside my computer case lately...not consistent...i don't know if thats normal or just my paranoia. But its a distinct click, almost as if something is being triggered (if that makes any sense)
February 24, 2006 9:59:49 AM

Don't worry, no problem whatsoever w/ stealing the post! After all, that's why I posted in the first place, to see any possible information..

I definitely DON'T hear the click in the case, and as I said before, I get restarts both when gaming or when performing much less stressful tasks.

What I DO hear is that, in the last second, maybe half a second before a reboot, the sound seems to "stick" (you know, like when a section of music a millisecond long repeats over and over again) -> kinda like ddrrrrrrr.... and then it reboots (sorry for the NON-scientific :)  explanation).

But I don't think it's due to sound card drivers or smth like that, because it manifests itself exactly the same both when using the onboard nForce audio chip and when using my old Creative Live PCI card.

Right now I have disabled the auto-reboot feature that NOBLY was talking about, and I'm waiting to see what happens.

Also, to whom asked me if I have Asus PC Probe, it's running and all parameters seem stable (except that 2-3 days ago I changed the thermal grease on the CPU cooler cause I thought it was low quality, and after I changed it, the temp actually INCREASED which means that my new thermal grease is sh*t... but highest temp is around 50-53 Celsius full load, nothing too serious.)
February 24, 2006 11:32:29 AM

Here's a stupid-simple thing to check, maybe the reset switch is bad/shorting. Unplug the leads to it for a few days and see what happens.

Look for a program called Bluesave, I think it's from Sysinternals. When it's installed and the system bluescreens, it saves a txt file in the windows directory of all the blue screen info. Very handy to have.
February 27, 2006 9:04:28 AM

Right, now some INTERESTING information...

I installed Motherboard Monitor last Friday, and what do I see???

The "CPU" temp that Asus PC Probe was reporting was, in fact, the temp of the CPU socket, and NOT from the CPU diode.

So, when CPU socket temp was around 48-50 Celsius, the CPU diode temp was, in fact, 68 - 70 Celsius!

WTF!!?? The Asus PC Probe isn't supposed to read info from the CPU diode?? It's set by default to shut down the PC when the temp is around 80! If the CPU socket is around 80, than the CPU diode is around 100!! That means very close to (if not) frying your CPU if you listen to PC Probe!

Anyways, I took apart the CPU and HSF, removed the grease that I had on, and replaced it with some heavy duty thermal grease that we sometimes use at work; it's made by Dow Corning and it's originally intended for power transistors, not CPU's, but I figured it may work for now. After all, a high power transistor dissipates much more power than a CPU, so this grease should be able to handle a fair amount of heat.

Right now, my CPU diode temps, according to Motherboard Monitor, are around 50 idle and 58-60 full load.
(went briefly to 62-63 when running 3Dmark05, but dropped to 56-57 a few seconds after shutting down 3Dmark). I'm NOT overclocking anything, so what do u think, are these temps safe?

Anyways, I'm sick of "thermal" emotions, so even if not overclocking, I'm looking for a serious cooling solution for socket A (Barton 2500+).
!