Random power-down, and unable to power-up

Hello Geeks!

I've had an interesting problem start recently (within the last month or so). Hope you can help me figure out what it is and what needs to be done. My PC's stats are in my sig. Any other relevant info I'll try to include in the explanation. Here goes...

About a month ago after turning off my computer, it wouldn't restart. I would push the power button, the HD activity light and power light would turn on, the keyboard lights (caps lock, numb lock, scroll lock) would flash, and the fans would start but the monitor would not start and just stay blank (the "no input" signal would flash for a second). I am using a generic LCD flat-screen monitor. After repeated attempts at turning on the PC (sometimes using the front power button, sometimes using the power button on the back of the power supply), it would turn on.

This would happen intermittently whenever I would power-down my computer. It wouldn't affect restarting the computer, only after a power-down.

Within the last week or so (maybe further back than that...some of the symptoms I will describe here would occur when the computer would go into suspension), my PC would reset itself during the night. I usually leave my computer running all day (turning it off when I leave for the weekend or for longer periods of time). I would leave the computer for the night, only to return to it in the morning to the initial BIOS password startup screen. Along with the random restarts, the PC will sometimes lose picture to its monitor and esentially lock up (I have another computer networked to this PC, and I can't access the shares when this happens), forcing me to power-down and returning to the problem in the above paragraph.

This lock up problem will also happen after about a minute into the Windows boot-up screen (after the "welcome" screen, but before all my running processes can start). And it's not just a Windows problem either, about a minute into my Linux start-up, I get the same lock up.

Any ideas what kind of hardware problem this is? Any ideas on how to fix it?

If you need any more info, please ask.

Thanks.

 

Overheating. Check that the CPU fan is working correctly and that the cooling fins are cleaned of dust. Also check the the case fans are also working correctly as they provide the airflow through the PC.

 

When this problem started occuring, had you done anything different to the computer? Add or remove hardware or devices. Did you change habbits on the computer: turning it off less often/ more often? If any little change was made right before this started to happen, let me know. Lastly is there a switch other than the power on/off switch. A small one that you would have to change using your fingernail. Make sure it is at the correct setting.

 

I have suspected overheating as well. I run Seti-at-home on the computer over night. However, this shutting down problem didn't happen immediately after installing S@H...only within the last week. Last night I tried running the "History" feature of my AsusProbe program. This morning, after restarting the computer, I checked the results. They stopped about at about 1:00 in the morning (before S@H is scheduled to run), registering only 42 degrees C. That doesn't seem to be too hot, does it?

Aside from the overheating problem, what do you think causes the problems trying to start up the computer? And why would the computer shut down sometimes when it's not overheating?

 

What powersupply do you have? 42 C is starting to get hot. Good temps would be 28 C to 36 C. Is your fansink secure to the cpu? Dust can collect between the spaces, hindering airflow through the fansink on the cpu. That can heat things up a lot. Even if it doesn't seem hot, your system can shut off as a safety feature when it thinks thinks temps are starting to get dangerous.

 

Not so. See here: http://www.heatsink-guide.com/content.php?content=maxtemp.shtml

 

@mdowns 42C is fine - I've just fitted a new set of cooling fins and a new CPU fan into my own PC just three days ago and my PC's CPU temperature is currently 38C.

If you leave the PC running in Safe Mode for a length of time does it exhibit the same symptoms?

 

Thank you Master Sergeant Yargwel, I was almost certain. (somebody mislead me) But thanks again.

 

No problem Higara7 . Now if anyone else tells you that 42C is too hot you can put them right.

 

OK, guys. I just spent the last 10 minutes or so trying to start my computer. I had shut it off for the afternoon, to let it cool down or whatever.

I turn on the power button, the keyboard flashes, but nothing happens on the monitor (it stays blank). The fans in the computer are running, the DVD drives are spinning (there are no disks in the drives), the HD light is solid, but there is nothing on the monitor. I hit the reset button, and the case beeps once, and then nothing. I try it again, same beep, same nothing. I flip the back power switch on and off and on and off and on and off and on. It finally catches, and starts up the BIOS splash screen. I get to my GRUB screen, and choose Windows. During the initial boot-splash-screen (the one with the big "Windows" and little blue ticker) the screen goes blank. I have to hit the power switch and go through the same routine again. The next time I get into Windows for about a minute before the screen goes dead again and I have to reboot.

This time I choose the MemTest 86 option, rather than Ubuntu or Windows. I run the memory test for about 2 hours with no problems. I reboot and go into Windows. I get halfway through typing this post out before the screen goes blank again and I have to go through the same rigmarole again.

Any ideas? This is driving me crazy. Is it a monitor problem? Is it a Hard Drive problem? Is it a PSU problem (the CPU, by the way, is 350W, which I think is more than enough for what I need)? Is it a CPU problem?

 

Well if the PC gives a single POST beep that generally indicates that all is well. Have you disconnected the monitor from the PC and checked to see if you haven't got a bent pin. Sometimes just disconnecting and reconnecting a connection will cure the problem. But if that appears OK I would suggest a possibly failing video card (maybe you could borrrow one from a friend) or that the PC is trying to switch into a video mode that it doesn't support. To eliminate the second possibility get hold of a Windows 98 boot disk (you can download one from http://www.bootdisk.com and boot from this floppy. If that works OK then you need to reset the video mode.

 

Hey Yargwel. Thanks for the help. I checked for a bent pin. None, so that's good.

I'm a little distressed about this being a failing video card. I bought this 6600GT less than a year ago. I would hate to have to buy another card so soon. I tried uninstalling the drivers and installing the new ones (I had been lax and hadn't updated the drivers for a couple releases), but I'm getting the failing monitor at an increased rate now. I successfully uninstalled the drivers, but have not been able to keep the computer running for long enough to install the new ones.

Do you think this might be a power problem. I'm pretty sure the 350W PSU is adaquate for my uses (it's only recently that the computer has started doing this...it worked well enough since April, when I put it together). Do you think the 12V line is not giving it enough juice? My hardware monitors in Windows and in the BIOS all say that it's giving +/- 12V...maybe they're lying? I do have a spare video card lying around (an old POS ATI 9200SE ). Maybe I'll try that out in the morning...and since I've been unsuccessful in installing the nVidia drivers, that might be a good thing. Man, I really don't want to have to buy a new video card. My mobo is AGP, and I was hoping that everything would last for at least another year or so, before I made a complete system upgrade. What do you think I should do?

 

Try the old video card that you have. If the PC then works fine you'll at least know that it is the "new" video card. You could also just try taking the "new" card out cleaning the contacts and then reseating it just in case it is as simple as that. If the old card does bring everything up OK then boot into Safe Mode and go to the device manager and select and then delete ALL the video cards you see installed. Turn off the PC and then refit the new card and allow it to be redetected and again you might be lucky and it will then come up and you can then reload all the relevant drivers.

 

Yeah, I tried taking the new card out and cleaning the contacts and stuff, but that didn't work.

I tried two things. One, I have a shuttle PC, which has an Nvidia FX5200 card in it, connected to my TV. I dismantled that and took out the card and tried it in the main computer (I tried installing the "possibly faulty" card into the Shuttle, but the 12V power cord wouldn't reach). It worked OK. Didn't seem to have any shut-downs. It played A Tale in the Desert OK, if a little slow.

Next, I tried the POS ATI 9200SE card. It is working fine too (so far). It won't play A Tale in the Desert (computer locks up, forcing a restart), but then again it never could, which is why I got the new card in the first place. I guess I'll leave it in for the weekend and see how it goes.

If there are no random shut-downs or blank screens, I guess we can say that it's the 6600GT which is the problem, huh?

If that's the case, I am seriously depressed, cause I don't want to have to make a decision on whether to buy a new card for this AGP board, or to buy a new board and a PCI Express card. I guess I'd lean towards an AGP card, in that my mobo is socket 939, which, if I want to upgrade in the future will be obsolete, huh?

Oh, well, lets see how this goes.

 

You could also download Everest from Major Geeks link http://www.majorgeeks.com/download.php?det=4181 at take a good look at your voltages as well as much more PSU's are notorious for faulty 5V faults which can cause symptoms like you are having, plus a PSU is an easy swap. Hope that helps

 

Hey, thanks for the tip. I checked Everest, and it looks like the 5V wire/rail is pumping out 5 volts...5.19 to be exact.

I'm going to go ahead and keep the POS card in for the evening, and see if there are any hangups. I'll try putting in the 6600GT again tomorrow, and see if I get the same problems. If so, then I guess I'll have to get another card. Maybe a 7600GS?

 

Bit of an update: I decided to run a Stability Test while I wasn't using the computer. Well, within 20 minutes or so of testing the computer locked up. Tried running it again, and it got about 30 minutes in and then locking up. The rest of the time the computer seemed to be running OK (i.e. not while the test was running). Do you think this is a sign that there's a problem with the motherboard?

I don't want to go off buying new parts when they might not be the right new parts.

 

More updates: The computer locked up a couple times while using the web-browser, so I ended up just turning it off for the night. This morning I took out everthing, and hauled out my old Asus A7V8X-MX motherboard and AthlonXP 2500. I hooked up the 6600GT video card to this along with a couple componants from the box having probmes (RAM, HDs, one of the DVD players...couldn't move everthing since the PSU on the old board is smaller). Everthing is running fine. The computer hasn't locked up, and the video card is working perfectly.

After doing some reading on the internet, I'm guessing that maybe the motherboard is the problem. What do you think?

 

Are you using the same RAM chips or different ones. It is possible that a dodgy RAM chip is causing the lock-ups.

 

The RAM chips are all the same. And they've passed MemTest86 with no problems.

However, that point is now moot. It turns out that the PSU on the main computer was kaputski. After doing some more reading on the net, I decided to swap out the PSUs between my old computer and the main one. The old one had a 250W PSU, so I had to leave out some componants. After hooking it up, though, the computer is working like a charm. So, I guess I'm now in the market for a new PSU.

Thanks for the help in this thread. Even though it ended up being something completely different than what we intitially though was the problem, threads like these are a good way for me to "think out loud" and get troubleshooting ideas.