BleepingComputer.com: laptop shutting down randomly

I have an HP Laptop that is shutting down randomly. To test it, I have been running it 24/7, and it seems to shutdown on it's own maybe once every five days. After it shuts down, I can turn it back on immediately, and it will stay on, so I don't think it's a heat issue. I also noticed that when defragging, it seems to get stuck on some ADS files (C:\WINDOWS:$SDS:$DATA ?).

The laptop had several viruses, but they have all been removed (but ya never know...).

Any suggestions on how to diagnose this? It's very difficult, since it's so hard to duplicate the problem.

Justin

It could be anything at all causing that. Failing battery, hard drive, memory, any part of the motherboard, or any of a hundred other causes that I can't think of off the top of my head.

I think the first thing I would check would be the hard drive. Run a chkdsk on it from the command prompt just to see if it has any errors or bad sectors. If something shows up there, I'd start with that. If nothing shows in a simple chkdsk, take the drive out to get the manufacturer and then go to there web site and get their own drive testing software. Use their software to test your drive. If nothing comes of that either ........... your guess will be as good as anyone else as to what is causing it.

Sorry, but problems that are 5 days intermittent are very hard to nail down. You could check your Event Viewer before trying anything. It may show something in there. Go into Control Panel - Administrative Tools - Event Viewer. If you need help using the Event Viewer, go to the Action menu once you've opened the Viewer and click Help.

Good luck!!

Yes, I've kept a close eye on the event viewer. A couple of times (about 15min before shutdown), the BITS service reported that a randomly named file c:\windows\temp\GUR***.exe (*** is the random part of the filename) had been terminated. The event viewer also reported some disk errors on the second hard drive (yeah, this laptop has two hard drives... silly), but those were a couple weeks ago, and I haven't seen anything recently. I've run chkdsk with a bad block scan on both drive.

BTW this happens whether the laptop is plugged in to the wall, or on battery power.

Disk errors...would do the job. Such are not to be taken lightly.

Louis

Any disk error showing up in chkdsk could definitely cause the problem. Its really up to you as to what you want to do now. I'd still suggest downloading the drive manufacturer's testing utility and test the drives using that. Most of the time, those diagnostics are reasonably in-depth scans and may tell you a little more.
Otherwise, pull out that second drive and try to run without it in the laptop. If the second drive is your problem causer, it will be easier to replace than the primary.

Thanks for the advice. I downloaded the Hitachi drive fitness test iso, and ran it on both drives. Both passed the Advanced Test. I've also run multiple bad block scans on both drives. Drive 1 has no remapped sectors. I'll have to double-check to see if Drive 2 does. Chkdsk has never reported any errors. If it did, it would have been during boot, and I might not have noticed it.

Last night I ran memtest86 for another 8 hours, still no issues.

The laptop shut itself off twice yesterday. One time when running a virus scan, and another time while running defrag (Ultimate Defrag). The last time I looked at the screen before noticing it had shutdown, it seemed stuck on C:\$SECURE:$SDS:$DATA. This isn't the first time I've seen ultimate defrag get stuck on ADS, but it's the first time it's shutdown while running defrag.

This post has been edited by JustinHoMi: 04 May 2009 - 02:04 PM

FWIW: http://forums.pcpitstop.com/index.php?showtopic=154652

Louis

Well, I've narrowed it down a bit.

I ran Prime95 (stress tests the cpu) in Safe Mode, off a Boot CD, and in normal windows mode. In safe mode and off a boot cd, it will run for a long time, and the temperatures stay low (<50C). However if I boot into windows normally, and run Prime95, the cpu temps quickly jump to 80-95C, and the laptop shuts itself down (takes 10-20min before it shuts down).

I really wonder if windows is reporting the correct temperature.

What could cause this? Are there drivers for the temperature sensors, or do these programs access them directly?

Weather you temperature are on the money or not they are almost twice as high. That's not good. IMHO this indicates that when you hump your system you do not have enough cooling and or the inside of your computer has dirt/dust ect.

But... what would cause them to be higher in a normal boot, as opposed to safe mode? This makes me wonder if the temperatures are really as high as they are reading.

In safe mode your computer is only running the bare minimum of operations/processes. In normal mode many more things are running in the background, and that depends on how you have your computer set up. The more thing running in your computer, weather you see them or not are working your computer more in normal mode than safe mode.

Yes... but Prime95 maxes out both cores at 100% whether booted normally or in safe mode. Correct me if I'm wrong, but isn't 100% = 100%?

Overheating = Premature Shutdown....You have probably diagnosed your problem properly...I would suggest a complete cleaning of your laptop internal components...I would also recommend that you purchase a cooling pad for your laptop...They are relatively cheap and work great on overheating machines...But I definately would have it cleaned professionally first...

Its almost like no one else is seeing the problem of the temps being high in normal mode and not in safe mode. The Prime95 test has nothing to do with drivers (that I'm aware of). It just loads your cpu up for testing. There's no reason I can think of why the temps should be higher in normal mode than when in safe mode.

A simple way to verify the temperature thing. Its kind of backwards thinking but it should work. Get yourself an old style mercury-bulb thermometer, if you can find one (or any kind of good thermometer should work) and put the temp sensor right at the cooling vent exit on your laptop. Make sure the air in the room isn't moving around too (ie. ceiling fans and such need to be off). Let the thermometer lay on the desk to adjust to room temp first. Boot up in safe mode and do the Prime95 test again for a half hour (since you say it shuts down in normal mode in 20 minutes). Get the highest temperature reading on the thermometer for that time. Then, do the same thing in normal mode. If you're seeing a spike in temp by a factor of 2 (80C normal mode vs 40C safe mode), you should see the same or close to it on the thermometer. That will at least verify if the temp is actually that high.
Just for the record, I'm betting that you'll see a spike in the temp for normal mode showing that the actual temperature is being reported by the system.

The next thing to check would be the system processes. Is there any kind of process running that might be causing this glitch? (I have no idea what it might be either, but maybe with your next post put a list of all the processes running in normal mode. Maybe somebody will be able to see something in that).

I hate problems like this. Just trying to give you some way to verify what the system is telling you about the temp thing though. If the thermometer trick proves the temp is actually higher in normal mode, it would probably be best to follow the OldGrumpyBastard's advice and get it cleaned first. Then, get a cooling pad for it and hopefully that will help.

This post has been edited by possumbarnes: 10 May 2009 - 09:42 PM

OldGrumpyBastard, possumbarnes, and I are thinking along the same road,(I think). It's a temp problem. I like possumbarnes idea of a way of getting the temps. A professionally cleaned computer can only help even if it doesn't solve your problem. I have had tower computers stop about mid way when defragging because of a poor cooling fan that stopped.