spontaneous PC power losses - HELP!

Your DVD drive could be defective.So could that second power drive too.One of the worst things to do with a computer is to physically unplug that computer without properly shutting the computer down first. cuhulin

Unplug/replug all power and data connections. That way, the errant loose one will be fixed and your problem will be resolved.

Thank you for the tip. I did this, but no luck.

A couple interesting new developments: one one occasion when it wouldn't start, holding down the power button kept the PC on, for as long as I held it. That made me wonder if it was the switch, but the effect was never repeated. Another time it wouldn't start, and I disconnected all the devices but the hard drive, and then it started. Yet other times I have disconnected every single device, and the fan won't even spin and try to start.

Strange...

Thank you! I appreciate the helpers among the trolls. :)

I had a similar thing which drove me crazy. Happened just after fitting a new graphics card. Which was just coincidence. Turned out to be the main processor overheating and shutting down. Its fan was working normally and no error messages generated. Removing the heatsink, cleaning and applying new thermo conductive paste sorted it. Now have a spare power supply...

In message , "Dave Plowman (News)" writes

I had the same problem with the PC shutting down when the processor was working hard. Its heatsink was full of dust. Soon afterwards, I again had shutdowns. This time it was because the PSU fan had seized solid (probably dry bearings). Freeing-off and re-lubrication with WD40 has worked for the last 6 months (but I will fit a new fan one day).

In the front of the case, I have since fitted an additional fan which sucks the air in, and also blocked off some unused holes (to ensure that most of the air entering passes through the fan). In front of the fan is a crude filter made out of thin scouring pads. Every couple of weeks, I wrench off the plastic front panel of the computer, and clean a VERY dirty filter.

Finally, as this is an Asus motherboard, I have installed 'Asus PC Probe' - a monitoring program which tells you things like the temperature of the processor and hard drives, the voltage rails, the fan speeds etc. As all is now working OK, I don't use it much, but it is interesting to watch the processor rise from 28C to 45C during a prolonged spell of 100% activity!

Clear the BIOS and then re-input the parameters.

You didn't say how old the PC is. It might be the onboard battery. First, try to reseat it; if that doesn't help, I'd try replacing it w. a new one. Also, check carefully that the MB is not occasionally shorting out to the chassis (that all insulating spacers are in place).

radiant snipped-for-privacy@outgun.com wrote:

You didn't say how old the PC is. It might be the onboard battery. First, try to reseat it; if that doesn't help, I'd try replacing it w. a new one. Also, check carefully that the MB is not occasionally shorting out to the chassis (that all insulating spacers are in place).

radiant snipped-for-privacy@outgun.com wrote:

Mine was relatively new and the heatsink clean. Plenty of fans, too, all working. It's a home assembled one which worked fine for a year or so before developing the symptoms. I used the paste which came with the heatsink originally and that had gone hard - dunno why.

In message , "Dave Plowman (News)" writes

Unfortunately, in this house, there seems to be an inexhaustible supply dirt and dust. God know where it comes from. You really would NOT believe how filthy my home-made filter gets.

Actually, I forgot to mention also that the clamping of the heat dissipater/ fan assembly to the processor was rather loose. This would definitely have made things worse. I disassembled everything, cleaned off the dust and thermal paste (which was still soft), did a bit of judicious bending so the clamp was firm, put on new thermal paste, etc. It was like a new machine when I had finished.

The heat dries out most pastes ! That is why the move to phase change materials was made. Also the phase change material allows for a much thinner interface improving thermal conductivity and thus a more effective heat transfer.

I agree. Shutting a computer down by simply unplugging it (a method I call "brute force") rather than using the correct shutdown procedure can and often does corrupt the operating system, requiring that the entire OS be reinstalled--not to mention all your application programs you may have installed at the time. Most of the time the computer will go into safe mode if it is shut down other than by the recommended procedure, but I wouldn't count on it. I don't even remember how many times my first Windows computer (running Win95) crashed and corrupted the operating system when I tried to shut down the system with the brute force technique. (I don't do that anymore with my current Win98SE system unless I have to, say if the computer freezes and absolutely cannot be shut down normally; live and learn.) Another problem with shutting down simply by pulling the plug is that the hard drive may have errors on it when the system is restarted; many if not most Windows computers are set up by default to scan the hard disk as soon as the system boots after a crash. The system can be set to repair automatically any errors thus found.

Still another potentially fatal (for the computer) problem created by the act of shutting down without following the correct procedure is that the read/write heads of the hard disk will literally crash into the disk, again with the potential for disk errors when the system is rebooted. These crashes will eventually ruin your hard drive, so it is best to avoid them if possible. Occasional disk crashes are inevitable (power outages, for example), but here I am speaking of unnecessary hard disk crashes caused by incorrect shutdowns; for maximum trouble-free life of your hard disk, please, for gosh sakes use the Windows-recommended shutdown method. This can be accessed directly from the Start menu by clicking on "Shut Down" and selecting the option of the same name from the menu which appears, or by doing the so-called "three-finger salute", i.e. pressing Ctrl+Alt+Delete while the computer is powered up.

Jeff Strieble, WB8NHV Fairport Harbor, Ohio USA

The heads on modern HD's automatically retract when power is removed, it's a function of the design. They do not crash into the platters.

Jerry

That may be true for today's hard drives (used in new laptops and desktop systems), but I was referring to the hard drive in my IBM Aptiva 595 system which was new eight years ago. Were the read/write heads on HDs of that vintage designed to retract on power down as well? If so, I'm sure I was worrying for nothing every time the disk crashed on that system and my first Windows computer circa 1997 (AST Adventure! model 200).

Jerry Peters wrote in news:QoOOj.221484$cQ1.29403 @bgtnsc04-news.ops.worldnet.att.net:

...'should' automatically ... 'should not' crash ....

Unfortunately, design and reality diverge at times. Also, on modern drives, the heads don't retract, they move to a 'parking area'. On older drives, the heads actually were removed from the area of the platter when the drive was off.

Now, the heads never leave the platters. They actually land on the platters when the platters stop spinning. They take off and fly when the platters spin.

I don't know how they avoid 'sticking'.

I have seen platters with bands of magnetic coating scrapped completely off due to head crashes.

In the late 70's, early 80's I actually replaced platters and heads and aligned drives.

Now, the heads are so small that I can barely see them!

In any case, since win 95/98 and later systems WRITE to the hard drive during start up, while running and during shut down, thus are updating the directory almost constantly. It is very likely that killing power without proper shut down will write garbage to the directory tracks. Once that is done, files are lost. If you are UNLUCKY, it is your data files that get corrupted, but you don't notice until you need them. If you are LUCKY, the system won't boot and you must repair the op system from the CD and repair disk you made when you installed the op system.

My first computer, bought in 1989, came with a 79MB IDE HD with a voice-coil actuator. It's an intrinsic part of the VC design that the heads retract when power disappears.

The old "head parking required" drives used a stepper motor actuator; when the power went down, the heads would be left wherever they were last positioned.

Jerry

When I worked for a large aluminum company, I had a 14" platter from a crashed drive, the aluminum substrate had been gouged by the heads in at least a 1/2" wide circle.

Again at the large Al company we had some IBM 3380 drives where the head actuators retracted against a rubber bumper. Over time the rubber degraded and became sticky, the IBM CE's actually replaced or covered the rubber bumpers with new ones, while the disk were spinning! Turns out that the disk assemblies while having a clear plastic surround, were not really sealed.

Not in my experience. Given the number of times W9x locked up with no alternative but to do a hard reset, I have had few problems. Never had to do a reinstall because of this. Depending on how and what NTFS journals, XP should be good in that regard also.

Jerry

The last time I saw a hard drive that needed to be parked was in the mid

Jerry Peters wrote in news:_18Pj.117388$D_3.86716 @bgtnsc05-news.ops.worldnet.att.net:

I have one of the 14" platters here, now.

One crash was caused by the rubber bumpers in the top of the removable disk pack shrinking and dropping onto the hub of the spinning disk. The friction caused the rubber tow abrade and the dust crashed the heads.

Another crash was due to the velocity transducer coming loose. The head loading current kept increasing because there was no indication of the heads moving. The carriage made a max velocity seek toward the hub. Sounded like a shotgun going off. Broke off the head travel stop.

....

If the system is locked up, the hard reset is necessary. No choice.

It is less likely to cause damage than just killing power to the computer in the middle of normal boot up, running or shutdown.

Well as the department's 'fix it' person, I see the results of bad luck. Often they tell me 'the power was shut off accidentally' or some such.

And I have managed to crash a few, on my own, by ignoring the 'proper shutdown' rules.