Sudden shutdown

[NSA]

After nearly a year of successful operation, Vista Home Edition will
suddenly and unexpectly shut down and then automatically restart. Upon
restart, the message appears to the effect that shutdown was not normal and
providing choices to start up normally or safe mode. I have not installed
any recent new applications that can be associated with this behavior. The
system has 4 GB memory, nVidia 7600 GT video card, Intel dual processor.It
may be a coincidence, but this behavior did not commence until after
downloading and installing several Vista security patches from MS.

N. Schwartz

 

[Rick Rogers]

Hi,

After a year, the heat sink and fans may have become clogged. The system may
be powering off in response to an overheating condition. Shut down and open
the case, inspect and clean these areas.

--
Best of Luck,

Rick Rogers, aka "Nutcase" - Microsoft MVP

Windows help - www.rickrogers.org
My thoughts http://rick-mvp.blogspot.com

 

[Earle Horton]

You can try uninstalling the security patches, one by one, in reverse order
of installation. You may be able to identify the one that is causing the
problem. Once you find it, you will then have to choose between leaving a
security hole and having a computer that shuts down all by itself. The
above assumes of course that a security patch, or a combination thereof, is
causing the problem.

 

[the wharf rat]

After nearly a year of successful operation, Vista Home Edition will
suddenly and unexpectly shut down and then automatically restart. Upon

Almost certainly a hardware problem. See if something's logged
in the event log and/or turn off automatic restart on error. Watch the
pattern of reported errors.

If you see a positive indication of a problem such as "error
in paging area" or "disk read error" you'll know what's wrong (although
easy stuff usually just hangs or blue screens rather than rebooting).

If the pattern is random errors with no real pattern suspect bad
ram first, a bad video/pci card second, motherboard third.

If the pattern is no error at all but the problem can be reproduced
semi-reliably by starting high demand apps like games or image editors
replace the power supply.

It's probably not a cooling problem because you restart immediately.

 

[Peter Foldes]

Power Supply can be suspect along with overheating and also as Rick posted in this thread

 

[NSA]

Thanks for the help. The power supply is becoming a likely culprit. A friend
suggeted disabling the automatic restart with the thought that the blue
screen will appear and provide some more details.

NSA

Power Supply can be suspect along with overheating and also as Rick posted
in this thread

 

[Ken Blake]

Thanks for the help. The power supply is becoming a likely culprit. A
friend suggeted disabling the automatic restart with the thought that the
blue screen will appear and provide some more details.

I agree with your friend. The default of automatically restarting is a poor
one, and should always be turned off for everyone.

 

[westom1]

Thanks for the help. The power supply is becoming a likely culprit. A friend
suggeted disabling the automatic restart with the thought that the blue
screen will appear and provide some more details.

Important information is in the system (event) logs. Changing to
eliminate automatic restart might make future failures obvious. But
all that information on past failures is already in that log.

Power supply: if it was always defective (and defective power
supplies can still boot computers), then a multimeter will report that
now as it would have always done. Simply boot the computer to execute
tasks to all peripherals. IOW download from the internet while
searching the hard drive, while playing complex graphics on the video,
while playing a CD while playing sound while ... Now you are ready
to read numbers from the power supply. In your case, the most
critical numbers come from any one purple, red, orange, and yellow
power supply wires. Those numbers must exceed 3.23, 4.87, and 11.7.
If the power supply numbers are good, then move on to other suspects
and don't look back.

If dust caused a failure (especially after one year in a 70 degree
room), then your computer has a hardware failure. That computer must
work just fine while full of dust in a room of 100 degrees F. Heat is
a diagnostic tool to find other defects. If heat created by dust
causes a failure, then you have a severe and probably getting worse
hardware defect - not a heat problem.

 

[the wharf rat]

Important information is in the system (event) logs. Changing to
eliminate automatic restart might make future failures obvious. But
all that information on past failures is already in that log.

We wish If the shutdown is due to power or certain disc issues
there almost certainly won't be anything in the log (because power problems
cause it to die before anything can be logged and an unwriteable disc
will obviously prevent a message from being logged).

Bad drivers and devices can so scramble the kernel data structures
that things just stop dead... And so on.

Power supply: if it was always defective (and defective power
supplies can still boot computers), then a multimeter will report that

Lol. Almost no one has a multimeter, and what if the power supply
USED to be good but ISN'T any more? All those capacitors and things inside
degrade over time, y'know.

Simply boot the computer to execute tasks to all peripherals.

Pfffft.

That won't tell you much. Even if you COULD manually coordinate
all that and still test all 5 leads at once it's only a point of time
capture and doesn't mean much. You really need an active load tester
(http://www.chromausa.com/apsts.htm) but it's way easier just to buy a new
power supply and see if that fixes the problem. If not you always needed
a spare, anyway.

If dust caused a failure (especially after one year in a 70 degree
room), then your computer has a hardware failure. That computer must
work just fine while full of dust in a room of 100 degrees F. Heat is
a diagnostic tool to find other defects. If heat created by dust
causes a failure, then you have a severe and probably getting worse
hardware defect - not a heat problem.

Oh, great Ghu... Dust clogs fans, interferes with airflow, and
insulates the parts it coats preventing efficient heat exchange. It's
not a diagnostic tool. It's poor hygiene. Blow it out and if all the fans
are spinning the thing's probably fine. Anyway, this guy doesn't have a
heat problem because he crashes once then reboots and runs fine. Heat
problems you either crash and can't boot for a while (until the hot parts
cool off) or crash reboot and crash very quickly again. (IOW it doesn't run)
Well, most of the time anyway...

A dust story: when I set up my first home office the small house
we owned didn't have any place to put it except in the spare bedroom we'd
dedicated to the cats. It had 2 litterboxes and cat trees and stuff. After
a few months my Pentium 90 stopped working, and when I opened it up all that
cat litter dust had coated the inside like cement. After I cleaned it out and
replaced the gummed up fans I put filters over the intakes and it was fine.

 

[westom1]

    We wish  If the shutdown is due to power or certain disc issues
there almost certainly won't be anything in the log (because power problems
cause it to die before anything can be logged and an unwriteable disc
will obviously prevent a message from being logged).

Same type failure also does not display any useful BSOD
information. Long before changing anything, the informed diagnostic
technician first gets the facts. System logs will even report a
failure long ago that would eventually start causing shutdowns today.
Many problems causing a crash today were being detected and 'worked
around' long ago while also stored in the system logs. Information
stored so that the informed tech will first learn about a failure long
before it was causing crashes.

Dust causing a crash in a one year old computer in a 70 degree room
means the informed technician locates the defective hardware - as we
even did 40 years ago. The naive would instead cure the symptom -
remove dust - and declare dust as the problem. Then 6 months later,
the same hardware defect gets even worse. Unfortunately the hardware
defect that once could have been covered by the warranty is now the
responsiblity of the computer onwer. Too many experts confuse 'curing
symptoms' with 'solving problems'.

Any properly constructed computer works just fine in a 100 degree F
room with large dust balls inside. Nothing new where the computer
tech first learns what is wrong before fixing it. BTW, this reality
is the minority opinion because most self proclaimed computer experts
don't even know why heat causes computer crashed. They just saw the
computer crash in a warmer room, then somehow automatically know heat
is the only reason for failure. Experts in curing symptoms.

 

[the wharf rat]

Long before changing anything, the informed diagnostic
technician first gets the facts. System logs will even report a
failure long ago that would eventually start causing shutdowns today.

It's not really possible for a home user to keep complete
diagnostic equipment handy, though, is it? And you simply don't
always get a nice graceful failure with useful warning messages logged
for a week beforehand.

Dust causing a crash in a one year old computer in a 70 degree room
means the informed technician locates the defective hardware - as we

Like the fans that stopped spinning cause of the dust? Geez.

even did 40 years ago. The naive would instead cure the symptom -
remove dust - and declare dust as the problem.

40 years ago, huh? Got much Univac experience on your resume,
you phony? May Ghu, the great, save me from poseurs.

Any properly constructed computer works just fine in a 100 degree F
room with large dust balls inside.

Is that why machine rooms are carefully maintained at 70 degrees
or less and include air handlers and filters? And sticky blue mats at the
doors to get the dust off your shoes?

I wish I could maintain facilities at 100F. I'd save millions.

 

[NotEvenMe]

LOL, 40 years ago I was replacing vacuum tubes to make the computer work...
They were fairly easy to spot, turn off the lights and look for the dark
tube.

<snip> Dust causing a crash in a one year old computer in a 70 degree room
means the informed technician locates the defective hardware - as we
even did 40 years ago. <snip>

 

[the wharf rat]

LOL, 40 years ago I was replacing vacuum tubes to make the computer work...
They were fairly easy to spot, turn off the lights and look for the dark
tube.

With all due respect, the last computer to use vacuum tubes was the
UNIVAC I, circa 1958. By the 1960's vacuum tubes had been abandoned -
because of HEAT - and machines such as the IBM 1400 were transistorized.
By 1968 - 40 years ago - the integrated circuit had been adopted.

So if you used tubes then it was well more than *50* years ago

And you can't really just look for the dark one, can you? I remember
tube testers, sometimes the tubes would light up just fine but still fail.

 

[FBonServer2008X64]

With all due respect, the last computer to use vacuum tubes was the
UNIVAC I, circa 1958. By the 1960's vacuum tubes had been abandoned -
because of HEAT - and machines such as the IBM 1400 were transistorized.
By 1968 - 40 years ago - the integrated circuit had been adopted.

So if you used tubes then it was well more than *50* years ago

And you can't really just look for the dark one, can you? I remember
tube testers, sometimes the tubes would light up just fine but still fail.

Yeah...a good trouble shooter could turn the equipment on and based on
what is was doing, or not doing, replace the appropriate vacuum tube(s).
The last piece of equipment I worked on that had vacuum tubes was from
Texas Instrument(circa '62), and it had only two vacuum tubes.
All the rest was printed circuit boards.
That piece of equipment was used in one of the systems that was a
precursor to the current GPS.

 

[westom1]

It's not really possible for a home user to keep complete
diagnostic equipment handy, though, is it?

All Dells come with a comprehensive hardware diagnostics - and for
free. What other diagnostic equipement is available? Screw driver,
maybe some wrenches, and a multimeter. All standard equipment to fix
computers, electrical applices, cars, building wiring, etc. What is
needed to obtain useful answers here without doubt? Same tools. No
more 'it could be this or could be that' useless answers.

Naysaying provided the OP with nothing useful and suggests how
little that naysayer knows. Dust should never cause heat problems in
a one year old computer. That computer full of dust in a 100 degree
room should still work just fine. If not, heat has probably
identified some other defective hardware. Heat is a diagnostic tool
as it was 40 years ago when vacuum tubes had long been replaced. Just
another tool powerful diagnostic tool that anyone already has.

Provided are the solutions that result in useful replies - no more
useless "it could be this or could be that" answers.

 

[fatsteve]

What other diagnostic equipement is available? Screw driver,
maybe some wrenches, and a multimeter.

Wrenches? For a computer? LOL There is no use for a wrench when
working on computers. Let alone using one as a "diagnostic" tool

 

[the wharf rat]

Wrenches? For a computer? LOL There is no use for a wrench when
working on computers. Let alone using one as a "diagnostic" tool

Sure there is. You use it to diagnose if a standoff is riveted
or just stuck: apply wrench, if standoff unscrews it was stuck if it snaps
it was riveted.

 

[fatsteve]

Sure there is. You use it to diagnose if a standoff is riveted
or just stuck: apply wrench, if standoff unscrews it was stuck if it snaps
it was riveted.

Can't you just look at it to tell, err diagnose, if it is screwed or
riveted?

 

[westom1]

Can't you just look at it to tell,  err diagnose, if it is screwed or
riveted?

Or one could take 30 seconds, use a meter, and then get useful
replies in this discussion.

Or once could execute the comprehensive hardware diagnostics that
are only provided by more responsible computer manufactuers.

Either way, then the next post would actually say something useful
and significant.

 

[the wharf rat]

Can't you just look at it to tell, err diagnose, if it is screwed or
riveted?

Don't be silly. A good technician always uses the proper tool.