Random Computer Restart

MattHewitt · 07-30-2003, 06:43 PM

I am building a computer... PIV Processor, 3.06 Gigahertz, L4VXA2 Motherboard, DDR RAM... Nvidia G Force 4200 Videocard. Anyway, the problem is... I'm installing Windows XP on it, and when I get to about the part where you enter in the authorization code (about 10 minutes into the install process) the computer restarts all by itself. I don't think this is a fan problem, because all of the fans seem to be working AOK. Also, it seems to restart around that time, but not at the exact same time... Also, before it restarts completely, the moniter seems to shut on and off... It goes black, then the screenc omes up, then goes black again... Sometimes before it restarts, a error message flashes really fast, but there is not enough time to read what it says. Any help would be greatly appreciated. I want my beast up and running.

Wizzard~Of~Ozz · 07-30-2003, 08:35 PM

If you have 2 sticks of memory try reversing them (switch their places) and see if it still reboots, If you only have 1 stick, try setting it in the BIOS to a more conservative timing. (try a higher number on the CAS setting)

random reboots are consistantly a problem with Memory for me, if you reverse them and the computer doesn't reboot until much later on, it would indicate a bad chip, Alternatively you can download a program like Memtest86 (don't have link for it, but Google works

)

Best of luck.

Idigguns · 07-30-2003, 08:39 PM

I concur with WoO.... and here's that link... http://www.memtest86.com/

Oops, I almost forgot... Welcome to TechIMO!

MattHewitt · 07-31-2003, 02:41 PM

Haven't tried the memory thing yet, but last night, by some miracle of god, I actually got XP installed. I tried formatting the hard drive once again, started the install, and it made it all the way through. After that I booted up, got online, checked windows updates... Installed Service pack one... Restarted like instructed... Now when I restart, it's on for like 5 minutes and then it restarts all by itself. Frustrating. But fun at the same time, I suppose. No error message this time, no flickering of the screen, just blackness then POST. Fun fun fun. Gonna try that memory thing though, if that doesn't work, any new ideas witht he new symptoms I gave ya? BTW, thanks for the help and quick replies! Oh yeah, also, only one stick of memory. Does it matter which slot the memory goes in? I've got it in the far left one? I mean it detects memory, so it should be fine unless there is a problem with the stick like you guys said. It's brand new, so I'm gonna be taking it back if there is a problem. Everyone else I talk to tells me overheating... I don't think that's it. It just doesn't seem possible. Same with the power supply thing. Brand new power supply, but I guess something could be wrong with it. Frustrating...

I was gonna try and do the higher number on the CAS setting, and then I realized I don't even know what a CAS setting is or how to change the number? It's in BIOS? I don't see it? Any elaboration for this newbie would be greatly appreciated.

SiliconJon · 07-31-2003, 08:01 PM

I the BIOS there should be a way to reset the settings to the default [failsafe, not optimized or performance]. This should reset your BIOS settings to conservative, but operational settings. And it does sound like it may be overheating. A crappy PSU [power supply unit] could be causing issues as well. But first, check the heat. If you have a multimeter with a thermo, check the heatsinks on the mobo, CPU, and video card. Temps of 60C + are getting high. Temps of 80C+ on a video card will crash a system. If you don't have such a tool, run the monitoring program for your motherboard, and get one for that card if it has one.

What is the model of PSU you are using?

Also, leave the first PCI slot empty to give your AGP card room to breathe. I have had this alone overheat a video card (though with proper cooling, it shouldn't happen...but not all of the cards these days are coming with appropriate cooling).

MattHewitt · 07-31-2003, 08:17 PM

The temperatures aren't reading too high. 40-50 Degrees... Tried the BIOS failsafe thing... That didn't seem to help too much. My power supply is a 400W HERCULES® ATX Power Supply w/P4 (+12V) connector... If that means anything to you. The company I bought if from said it would be fine with a P4 system.

I did indeed think that maybe my video card wasn't breathing, because it was mighty close to my other cards, so I did give it some room, and that didn't help either. Now that the computer has restarted so many times without being shut down properly, it constantly wants to run scan disk and sometimes locks up in scandisk. Annoying. Frustrating. Thanks for the help, and keep teh suggestions coming. :-)

SiliconJon · 07-31-2003, 10:03 PM

I've had so much bad luck with cheap PSU's, that I want to tell you go get an Antec. Not only do they cause headaches, but I believe them to be the culprit in many other hardware failures as well (they are, after all, supplying the power to every device, and if it's providing BAD juice, it can fry just about any device it's powering). And using a hardware monitor is only as reliable as the motherboard's censor circuits, nor does the software take measurements nearly often enough to catch deadly changes in a rail's voltage.

I would very much try to find someone to loan you some parts. I would swap out each piece with a replacement known to work flawlessly. Do each item: RAM, PSU, AGP...and well, you can go all the way to the keyboard & mouse. Start with whatever pieces you can get your hands on first.

This is the method I sometimes have to resort to...sure do wish there was a better way. But so far the only other way I've found is to buy a fortune's worth the testing equipment.

Also, read that manual that came with the motherboard from beginning to finish. It may not help, but it will definately not hurt.

EDIT:

Don't change anyhardware around while the green LED on the Mobo is lit. To get the light off, you have to turn the PSU off. If the PSU does not have an on/off switch, get another one.

Some checks for that board:

JP2: short 2+3 for 533Mhz CPU FSB

Make sure everything is secure, RAM has both retension clips fully closed, all slot cards are fully inserted.

Did the heatsink come with a thermal material attached? I highly recommend removing this puddy like material and using thermal paste. Also, don't forget to take off the paper that is on the thermal material if your are going to use it (yes, I've seen it done).

MattHewitt · 08-01-2003, 04:08 PM

Yeah... didn't know that there was paper on the thermal stuff. Ahem ahem... Havn't checked to see if there was. If there was... is it melted to everything now, or should I still be able to repair? Is it good for the heatsink to be put on and taken off like that? It was a bad boy to get on in the first place.

Read the motherboard manual. I feel slightly more knowledgable now... It did help me understand some stuff. I took out all pieces other than needed ones, and swaped video cards once. That means it's either the hard drive, the ram, the power or the processor. Process of elimination is fun.

The funny thing is heatwise, the computer doesn't lock up until it is definatly installing parts of windows. Before when the screen is DOS ish and no pictures, it does fine. It can be on for hours as long as there are no pictures. What does that mean?

Is the flickering screen thing normal when you're at the part of the XP installation where it says "isntalling devices?" Some people have told me that that's not something I should be worried about, and that the restart around the time after it installs the devices points to a driver problem... But sometiems it doesn't actually restart until after that step and into the next one... entering the region settings and registration code... So I'm not sure.

One of these days it will be up and running and I will do a happy dance.

BTW, all this restarting and trying stuff out isn't going to harm anything will it?

Still clueless.

MattHewitt · 08-01-2003, 10:46 PM

Ok, tried replacing the hard drive... When I go to install windows on the new hard drive it gives me a new wonderful error... (At least I have an error to give this time though.) "Windows could not start because the following file is missing or corrupt! [Windowsroot\system32\hal.dll] Please reinstall a copy of the above file." Isn't that nice? My only option is to format as NTSF if that helps any. Only on partition on the drive. Forget all that other stuff... I just want windows on. Any idea what this message means? I figure once I get windows starting to install with tha graphics portion, I can see if by doing a whole bunch of other stuff, the restart problem has been fixed, but I can't even tell if it's been fixed because I can't get to the part where it hangs up. I guess one thing at time. Thanks again...

MewsicLovr · 08-08-2003, 09:33 PM

I have this exact same problem, exact board and cpu. Here's my full specs:

ECS L4VXA2
3.06GHZ w/533 FSB
2 x 512 MB PC2700 RAM
Maxtor 160 GB 7200 RPM 8mb cache
Samsung 16x DVD ROM
Inno3D GEForce4 Ti4200 128mb RAM

This is the bare minimum I have running with it now; I had more but it still has the problem even with the minimum required to run the machine.

First discovered it when installing a fresh copy of windows, it reboots about 15 seconds after reaching the first breakpoint in the install when it asks for your initial username, and then wants the cd key. Every time, haven't been able to get around it. Also tried booting up a hard drive with windows already installed - it stll reboots right when you would otherwise see the windows loading screen with the scrolling bar.

OK, so here's what I've done:

1. I had a whole bunch of devices hooked up internally, so I dropped it down to just the bare cd drive and single hard drive, maybe it was the power, right? Wrong. Also no possibility of a device conflict.

2. Suspected bad video card, I'd tested it in another system and it had worked but maybe was incompatible. I switched it with another good video card I had. Same problem.

3. Suspected RAM. Didn't really have a way to test this as my old system doesn't take pc2700 memory. Best I could do was to underclock it, maybe it was too fast for the board. Nope, still creates an error. Tried playing with the latency (bumped it to 3 clock ticks) - nope. I had two sticks of RAM already in there, so I tried each individual one in each slot. Nothing.

4. Finally was able to get a hold of some better quality RAM (I originally had generic - yeah yeah, I know). Still didn't fix it.

5. Tried to run memtest on it but it looks like the floppy doesn't work either so I don't think it worked...it just looped over and over the same addresses, didn't even get the program started. Same thing with my old computer which I KNOW is good.

I have yet to flash the BIOS because of the abovementioned floppy problem. I believe, without looking, that i have bios v 1.0e. This is not listed on ECS's website, is this an unstable version? Would I benefit by flashing?

I'm at a point now where it's either the board or the processor itself, with no way at the moment to test either one. Does anyone have any insight?