Sudden shutdown and power off

[ppbjno1]

Win XP Home. Installed Ulead Video software successfully (printed the manual).

Days later, computer shuts down to complete power off shortly after
startup. I could get to Safe Mode and operate without shutdown for a long
time. I cannot get to Safe Mode now. I am able to operate the CDROM without
shutdowns.

Assuming a virus, I reinstalled the OS. Still the shutdown occurs at
startup. I reinstalled the former hard drive. Still shutdown at startup.

There IS dust in the cabinet. There IS a groan that may be coming from the
cooling fan but the fan appears to operate satisfactorily.

Is this a heat problem only?

 

[Malke]

Win XP Home. Installed Ulead Video software successfully (printed the
manual).

Days later, computer shuts down to complete power off shortly after
startup. I could get to Safe Mode and operate without shutdown for a long
time. I cannot get to Safe Mode now. I am able to operate the CDROM
without shutdowns.

Assuming a virus, I reinstalled the OS. Still the shutdown occurs at
startup. I reinstalled the former hard drive. Still shutdown at startup.

There IS dust in the cabinet. There IS a groan that may be coming from the
cooling fan but the fan appears to operate satisfactorily.

Is this a heat problem only?

There is no way anyone who can't do hands-on testing can answer that. It
could be heat, power supply, motherboard, other components. Here are some
general hardware troubleshooting steps:

http://www.elephantboycomputers.com/page2.html#Hardware_Tshoot

Testing hardware failures often involves swapping out suspected parts with
known-good parts. If you can't do the testing yourself and/or are
uncomfortable opening your computer, take the machine to a professional
computer repair shop (not your local equivalent of
BigComputerStore/GeekSquad). If possible, have all your data backed up
before you take the machine into a shop.

Malke

 

[db.·.. >]

perhaps, the graphics
driver is corrupted and
would cause the instability.

however, such proprietary
drivers are only provided
from the homesite for your
machine or the motherboard.

that is why, the reinstall may
have failed to address the
problem.

incidentally, if you have
added a video board to
your system, then it too
has proprietary drivers.


--

db·´¯`·...¸><)))º>
DatabaseBen, Retired Professional
- Systems Analyst
- Database Developer
- Accountancy
- Veteran of the Armed Forces

 

[westom1]

Is this a heat problem only?

If hardware is OK, then a computer with massive dust balls will work
fine in a 100 degree F room. Heat is a way of finding defective
hardware. Those who understand so little will quickly blame only what
they understand - heat - rather then the defect that heat is making
obvious.

You _assumed_ a virus, then fixed it. There is the first problem.
You fixed something only on assumptions? You did not first ID the
suspect, then more on to repair it. Based upon what was posted, the
better educated cannot provide any useful replies. Your replies will
only be as useful as the fact you have provided. IOW they say this
even in CSI, "Follow the evidence". You are not doing that. Instead,
you are shotgunning.

How would you fix doors in a house? Would you immediately start
planing them? Or would you first confirm the foundation is good?
Your attitude is to immediately plane doors. Instead, start with the
foundation. That is why the fewer and more responsible computer techs
use a 3.5 digit multimeter to get numbers from a computer's foundation
- its power supply 'system'. Yes, a system that is more than just a
power supply.

If you post numbers, then the better informed can provide useful
replies. Best to first load up the computer for maximum loading. IOW
play complex video graphics (ie a movie), while downloading from the
internet, while searching the hard drive, while playing sound, while
reading a CD-Rom, while ... Now you are ready to take VDC readings
with the meter.

What are those numbers on any one of red, orange, yellow, purple,
green and gray wires from power supply to motherboard?

No have no idea of the massive number of facts in those numbers -
yet. Without those numbers, the better informed cannot help you.

Moving on - every responsible computer manufacturer provides
comprehensive hardware diagnostics for free. Execute them.
Inferior manufacturers mean you must download diagnostics from each
component manufacturer or third party. Welcome to why those less
responsible manufacturers are also cheaper.

What does the system (event) logs and Device Manager report? Again,
long before trying to fix anything, first get facts. Do you have
something that reports CPU temperature? If so, then why did you not
provide that number. To get a useful reply always means numbers.

... but the fan appears to operate satisfactorily.

says nothing useful. Where are numbers such as fan RPM, CPU
temperature, etc. Comprehensive hardware diagnostics would even
provide those facts. Without numbers, your only reply with be, "if
could be this or could be that or could be the other thing". Wild
speculation. Instead, follow the evidence.

 

[Leythos]

If hardware is OK, then a computer with massive dust balls will work
fine in a 100 degree F room. Heat is a way of finding defective
hardware. Those who understand so little will quickly blame only what
they understand - heat - rather then the defect that heat is making
obvious.

Not true WTom, dust will block air-flow over heatsinks, and that will
cause Shut-down. Many tech's have seen the problem in the real world
many times.

Even in a 60 deg F room, the presence of blocked vents, blocked fins on
heat-sinks, can cause a computer to overheat.

 

[John Doe]

(e-mail address removed) says...

Not true WTom, dust will block air-flow over heatsinks, and that
will cause Shut-down. Many tech's have seen the problem in the
real world many times.

But w_tom isn't from the real world.

By the way. If heat is being considered, since it's a power related
issue, I would also consider momentary household power interruptions
that might be worsened by a heavier load. Or a bad power supply. A
voltage regulator (line conditioner) or UPS can help solve household
power problems.

But doesn't the CPU run at full power when outside of Windows?

 

[westom1]

Not true WTom, dust will block air-flow over heatsinks, and that will
cause Shut-down. Many tech's have seen the problem in the real world
many times.

Dust balls still leave holes for air flow. Those holes with big
dust balls must be more than sufficient to cool any computer even in
room temperatures well about 70 degree F.

Meawhile if dust had completely blocked every hole in a heatsink,
then CPU temperature numbers would report that. If the CPU
temperature is anywhere below it maximum number, a computer must be
working just fine - with big dust balls. Heat is a diagnostic tool to
locate a hardware defect. Computer failed only due to dust balls has
a hardware defect elsewhere. Minor temperature increase due to dust
balls identifies a hardware defect.

 

[Leythos]

Dust balls still leave holes for air flow. Those holes with big
dust balls must be more than sufficient to cool any computer even in
room temperatures well about 70 degree F.

Meawhile if dust had completely blocked every hole in a heatsink,
then CPU temperature numbers would report that. If the CPU
temperature is anywhere below it maximum number, a computer must be
working just fine - with big dust balls. Heat is a diagnostic tool to
locate a hardware defect. Computer failed only due to dust balls has
a hardware defect elsewhere. Minor temperature increase due to dust
balls identifies a hardware defect.

Wrong again, as normal use will cause CPU temperature as well as Video
CPU temperature to increase and decrease under loads. This means that
the temp of either, and in the case, will change as the computer is
used.

One must not reach the "Max" for a item to fault, only reach that
specific items Max, which may not be the same as the vendors specified
Max.

So, again, Dust coating a Heat sink reduces the ability to dissipate
heat, it also changes airflow in the system as dust coats vents and
fans.

So, again, WTOM, you're missing how things work in the REAL WORLD.

 

[westom1]

Wrong again, as normal use will cause CPU temperature as well as Video
CPU temperature to increase and decrease under loads. This means that
the temp of either, and in the case, will change as the computer is
used.

With large dust balls inside, in a 100 degree F room, and with those
routine temperature changes as loads increase and decrease, a system
must work just fine AND tempertures must remain below maximum spec
temperatures. This is why engineers - not computer assemblers -
design reliable machines. The machine is designed theoretically (with
numbers) and confirmed with experimentation. Without numbers from
both, then the machine is not properly designed.

A machine with defective hardware that works just fine in a 70
degree F room can begin failing when dust balls begin to expose faults
in defective hardware. Heat is a diagnostic tool; finds actual
failes. Only the naive would simply cure symtpoms - fix it by only
removing dust. Once the actual problem is fixed, then also remove the
dust balls.

 

[Leythos]

With large dust balls inside, in a 100 degree F room, and with those
routine temperature changes as loads increase and decrease, a system
must work just fine AND tempertures must remain below maximum spec
temperatures. This is why engineers - not computer assemblers -
design reliable machines. The machine is designed theoretically (with
numbers) and confirmed with experimentation. Without numbers from
both, then the machine is not properly designed.

A machine with defective hardware that works just fine in a 70
degree F room can begin failing when dust balls begin to expose faults
in defective hardware. Heat is a diagnostic tool; finds actual
failes. Only the naive would simply cure symtpoms - fix it by only
removing dust. Once the actual problem is fixed, then also remove the
dust balls.

WTOM, the fact remains that even in a cool room, a dusty heat-sink is
less able to dissipate heat and can cause thermal shutdown as well as
damage of some components that don't have other thermal protection built
into them.

Heat is a THREAT, not a diagnostic tool. Heat that is not removed is the
source of many failures of system.

Your limited experience inside that little box you live in is not
helping anyone.

The problem could easily be JUST DUST COATED SURFACES failing to provide
the proper cooling to the protected component(s).

 

[westom1]

Heat is a THREAT, not a diagnostic tool. Heat that is not removed is the
source of many failures of system.

Too many without technical experience assume that heat caused
failure by damaging semiconductors. Yes, it a semiconductor is so hot
as to remove skin. But then only one between us really has such
technical experience having learned this even when semiconductors were
once sold on land that later became the WTC.

It is only too hot when touching it leaves skin. Temperatures below
only cause temporary operation failures. But those only eductated in
popular myths would not know that.

Heat created by dust balls in a 100 degree room does not get anywhere
near to those destructive temperatures. Diagnostic temperatures
created by dust in a 100 degree F room or by a hairdryer on high must
not even cause a program crash.

How to find defective hardware? Those diagnostic temperatures cause
defective hardware to crash program execution. We who actually learn
this stuff (who ignore poplular myths promoted without numbers from
Lythos) routinely use heat to find hardware failure.

A computer chock full of dust balls must work just fine in a 100
degree F room. If not, the trained computer technician locates and
replaces the defective hardware.

Heat is not the threat so often _feared_ by myths. Heat is a
powerful diagnostic tool to locate defective hardware. What is burn-
in testing? Executing hardware at those same temperatures that
Leythos calls a threat. Heat only becomes a threat when touching a
semiconductor leaves skin. A valid concept even long ago when
semiconductors were not so robust.

Use heat to find defective hardware. The naive even want to cure
symptoms with more fans rather than learn from manufacturer
datasheets.

 

[Leythos]

It is only too hot when touching it leaves skin. Temperatures below
only cause temporary operation failures. But those only eductated in
popular myths would not know that.

Heat created by dust balls in a 100 degree room does not get anywhere
near to those destructive temperatures. Diagnostic temperatures
created by dust in a 100 degree F room or by a hairdryer on high must
not even cause a program crash.

And yet we can have "intermittent" failures, as described by the OP,
because of excessive heat caused by dust.

Again, you're showing your completely lack of understanding in anything
except a very narrow scope WTOM.

 

[westom1]

And yet we can have "intermittent" failures, as described by the OP,
because of excessive heat caused by dust.

Owner would call us in to actually fix it. We discovered that
Lythos kept fixing it by removing dust. Turns out one memory was
defective the day that computer was purchased. As trivial dust balls
developed, Lyethos would clean the dust - and defective memory
remained. Computer would crash intermittently - more often as the
dust balls increased. We finally fixed it permenantly one summer day
by running it in a 100 degree rooom and locating the defective part.
Had the computer been tested with comprehensive hardware diagnostics
in a 100 degree room on the first day, then the computer would have
never failed again and Leythos was never needed.

The computer industry is chock full of Leythos computer assembler
who always know how to fix things but never bothered to first learn
these concepts. If not obvious to others, Leythos and I are
constantly having these discussion about old and well proven
technologies and concepts that he never learned and still denies
today.

A semiconductor is harmed only when it gets to hot as to leave skin
when touched.

 

[Leythos]

A semiconductor is harmed only when it gets to hot as to leave skin
when touched.

And yet there are thermal shutdown protection circuits in CPU's and on
Video cards and even some motherboards have them for different chipsets
- all facts you choose to ignore.

So, when something gets hot it can SHUTDOWN before damage, causing
intermittent problems, all caused by HEAT, without a single component
being defective.

WTom would have you believe that a component has to FAIL in order for
you to experience an Intermittent heat related problem, or that the
default must be defective in some manner - while a defective device can
fail for heat related issues, many devices cause intermittent faults
BECAUSE OF HEAT without the device being damaged.

If your computer is on the floor, not 20+ inches above the floor, there
is a very good chance you have Dust inside it. If you clean out the dust
every so often (once every 6 months or so, depends on your environment)
you will have a cooler operating computer system and be less prone to
intermittent heat issues.

WTom only sees one side of issues, the limited things he's been exposed
to or read in blogs, he's really ignorant of the real-world issues that
plague most computer users. Yes, components can be bad, yes they can
fail, not one person has stated otherwise - but the Dust can also cause
devices/components to go into thermal shutdown or thermal limiting to
prevent damage to them - this ends up looking like an intermittent
problem to the user. Removing the possibility of thermal issues is cheap
and a first step when there are "intermittent" problems.

 

[Twayne]

Too many without technical experience assume that heat caused
failure by damaging semiconductors. Yes, it a semiconductor is so hot
as to remove skin. But then only one between us really has such
technical experience having learned this even when semiconductors were
once sold on land that later became the WTC.

It is only too hot when touching it leaves skin. Temperatures below
only cause temporary operation failures. But those only eductated in
popular myths would not know that.

Heat created by dust balls in a 100 degree room does not get anywhere
near to those destructive temperatures. Diagnostic temperatures
created by dust in a 100 degree F room or by a hairdryer on high must
not even cause a program crash.

How to find defective hardware? Those diagnostic temperatures cause
defective hardware to crash program execution. We who actually learn
this stuff (who ignore poplular myths promoted without numbers from
Lythos) routinely use heat to find hardware failure.

A computer chock full of dust balls must work just fine in a 100
degree F room. If not, the trained computer technician locates and
replaces the defective hardware.

Heat is not the threat so often _feared_ by myths. Heat is a
powerful diagnostic tool to locate defective hardware. What is burn-
in testing? Executing hardware at those same temperatures that
Leythos calls a threat. Heat only becomes a threat when touching a
semiconductor leaves skin. A valid concept even long ago when
semiconductors were not so robust.

Use heat to find defective hardware. The naive even want to cure
symptoms with more fans rather than learn from manufacturer
datasheets.

westom...


Wow; may I assume you have no fingerprints left? I suppose you're
trying to write to the lowest common denominator, but ... IMO it's
didn't work, is pretty ineffective and fraught with using facts to
confuse and then miscellaneous generalities to convince.

I think you would do well to listen to Leythos and others with similar,
more accurate and applicable real world viewpoints. "Leaving skin"
behind is such an erroneous temperature indicator that it's anything
from very occasionally a little useful to completely useless depending
on what is being evaluated. Additionally, carcass temps are only
general indicators; it's the temp at the junctions/internal components
that matters, not what's at the carcass' surface.

Fortunately at the end you mention useing a manufacturer's datasheets
although most people will have trouble finding them. You speak in °F
in a couple places but all spec sheets etc. speak in °C, and you did not
bother to mention any functional temps where they could be useful. RS
has a nice $12 temp probe that's great for that kind of thing too,
rather than a finger, which actually has me laughing to myself when I
picture it. I assume delta-t is known for the finger touch too?
It reminds me of the guy in school who put a soldering iron up to
his cheek to see if it was heating, and actually touched his cheek with
it by mistake. Yes, it was completely heated up. And had melted solder
on the tip.

A little more consideration for your audience and less of your own mix
of experience/guesses would take you a long ways here and build yourself
some credibility.

Regards,

Twayne

 

[westom1]

I think you would do well to listen to Leythos and others with similar,
more accurate and applicable real world viewpoints.  ...
   Fortunately at the end you mention useing a manufacturer's datasheets
although most people will have trouble finding them.  You speak in °F
in a couple places but all spec sheets etc. speak in °C, and you did not
bother to mention any functional temps where they could be useful. RS
has a nice $12 temp probe that's great for that kind of thing too,

Or maybe learn from someone who actually did this stuff
professsionally - first did the math and then proved it
experimentally. Remember concepts even taught in junior high science?
Leythos doesn't. Learn the science from one who actually did this
stuff for generations (not just decades).

Routine (and misguided) is to blame crashes on temperature. Higher
temperature is how to find defective hardware (and it not destructive
to hardware). But then which one of us has our labors tested even in
outer space? One thing is certain - Leythos posts popular and
erronesous myths.

Remove dust balls every six months? Too many chassis fans also
create excessive dust problems - another hardware defect that
contradicts popular myths. Computers typically operate for years just
fine without removing dust. Too many chassis fans cause exceess
dust. If dust causes computer failure in a 70 degree F room, well,
dust balls clearly are not the reason for failure.

How to find defective hardware? Put that computer in a 100 degree F
(or 38 degree C) room and execute comprehensive hardware diagnostics.
Silly comments about using 'degrees F' suggests only only wants to
argue - to confuse others who would otherwise learn. Doing this
science stuff so long while educating American layman and exposing
Leythos myths .... it is not 'degrees'; it is 'degrees F' or
'degrees C'. One who came from where the work gets done understands
anyone also using F and C actually does know this stuff. Obviously
that is not Leythos who never posts numbers with his accuastions and
subjective reasoning.

Heat has always been a superior diagnostic tool to first locate
actual failures. A technique routine even geneations ago and rarely
understood by those without training. Myth purveyors are easily
identified. They deny this diagnostic tool using only accusations and
insults - and no numbers..

Twayne, where are _your_ numbers? Twayne would do better to first
learn the numbers rather than praising popular myths, accusations, and
outright lies from Leythos. Or simply learn that Leythos has years of
history posting insults based in technical ignorance.

Dust balls cause failures in computers that have other real
problems. A computer must work just fine in a 100 degree F room with
dust balls. A computer failing in a 70 degree room is not 'cured' by
removing dust balls.

 

[Leythos]

Or maybe learn from someone who actually did this stuff
professsionally - first did the math and then proved it
experimentally. Remember concepts even taught in junior high science?
Leythos doesn't. Learn the science from one who actually did this
stuff for generations (not just decades).

And yet you seem to be missing all of the basics from those days of
science class and you're missing all of the rational experience things
too.

Blocking cooling can cause a computer to overheat components that can
cause intermittent faults without causing damage to the components or
indicate that the components are bad.

You FAIL to understand that many modern circuits are designed with
thermal protection, because the vendors knew they COULD be subject to
temperatures that could damage them, so they built in thermal
protection.

A very dusty case vent/PSU is a sure sign that the CPU and other fan
protected devices are also coated with Dust and not providing optimal
cooling - this simple real-world issue has cause more intermittent
faults that you can shake a stick at - WTom would have you ignore this
completely.

 

[westom1]

Blocking cooling can cause a computer to overheat components that can
cause intermittent faults without causing damage to the components or
indicate that the components are bad.

Normal is to design assuming a certain amount of restricted
airflow. Properly designed hardware assumed the computer will contain
big dust balls and must not crash. Many years later, that computer
can have big dust balls and must not crash in a 100 degree F room.
Meanwhile, if it is crashing in a 70 degree F room, then heat is not
the problem. Hardware defect (defective part or defective design) is
the problem.

This concept does not play well where technicians removed dust
balls, the hardware worked, therefore the technician must be an
expert. Part of training a technician is to know that dust balls are
not the problem. Dust balls can be excellent diagnostics tools for
finding the problem.

When does a semiconductor get too hot as to be damaged? When one
touches it and leaves skin. Dust balls would never cause hardware to
get that hot - no matter how many times Leythos says otherwise.

Did Leythos ever yet post any numbers? Is he still posting
subjective assumptions? Numbers were even presented in a form he
would better understand. Hot enough to leave skin. Dust balls don't
make hardware that hot - therefore do not create semiconductor
damage. Another number: a computer with dust balls must work just
fine in a 100 degree F room.

 

[V Green]

Blocking cooling can cause a computer to overheat components that can
cause intermittent faults without causing damage to the components or
indicate that the components are bad.

Normal is to design assuming a certain amount of restricted
airflow. Properly designed hardware assumed the computer will contain
big dust balls and must not crash. Many years later, that computer
can have big dust balls and must not crash in a 100 degree F room.
Meanwhile, if it is crashing in a 70 degree F room, then heat is not
the problem. Hardware defect (defective part or defective design) is
the problem.

This concept does not play well where technicians removed dust
balls, the hardware worked, therefore the technician must be an
expert. Part of training a technician is to know that dust balls are
not the problem. Dust balls can be excellent diagnostics tools for
finding the problem.

When does a semiconductor get too hot as to be damaged? When one
touches it and leaves skin. Dust balls would never cause hardware to
get that hot - no matter how many times Leythos says otherwise.

Did Leythos ever yet post any numbers? Is he still posting
subjective assumptions? Numbers were even presented in a form he
would better understand. Hot enough to leave skin. Dust balls don't
make hardware that hot - therefore do not create semiconductor
damage. Another number: a computer with dust balls must work just
fine in a 100 degree F room.


^^^ the heck they (dust balls) don't.
Dust "balls" (or whatever you want to call them) can STOP CPU
and video card cooling fans completely through physically jamming
the blades.

I think even YOU can understand what will
happen next...

BTW - your statement:
"a computer with dust balls must work just
fine in a 100 degree F room."

is patently ridiculous. There is an almost infinite amount of variation in
hardware and software combinations for any given PC...any of which
can affect cooling. You just can't make a blanket assertment like that.

BTW, given the "race to the bottom", build it as cheaply as possible
mentality of PC mgfrs, I would venture to say that ALMOST NONE of them
design to the standard ("Many years later, that computer
can have big dust balls and must not crash in a 100 degree F room.")
you describe unless it is a very expensive server with redundant everything
or similar mission-critical design.

Mass-market PC's (far and away the largest percentage of the installed
base) are el-cheapo "consumer" devices. Almost considered disposable.
The mgfrs WANT them to fail so you have to buy a new one.

Bottom line - your advice is BAD for the vast majority of folks running an
E-machine or similar inexpensive "home" device. Give it up and move on.

 

[westom1]

^^^ the heck they (dust balls) don't.
Dust "balls" (or whatever you want to call them) can STOP CPU
and video card cooling fans completely through physically jamming
the blades.

That is not dust balls. If dust is so packed heavy after three
years of operation as to stop fans, why is a home computer being used
in an industrial environment? That is why industrial computers are
sold. Why do clone computers have so many (too many) fans causing
massive dust concentrations?

Even heatsinks coated in dust should have holes permitting airflow.
Heatsinks are also selected so that cooling is sufficient even when
coated with dust.

V Green is using Leythos' subjective reasoning to worry about a
problem that must never exist in a properly constructed computer in
typically harsh residential or commercial environment. Another major
reason for dust so heavy as to stop a CPU fan are too many chassis
fans - moving too much air. A hardware design failure often found
where a computer assembler did not learn some simple numbers such as
chassis CFMs.

A naive computer assembler is the butt of a classic Tim Allen joke:
"More Power". Their solution when designing a computer? "More
Fans". Then the computer does not have dust balls. Instead, a
defectively designed computer has packed dust sediment.