Location of error files

I recently cleaned out my computer case and removed a parallel port
card. Now I am getting unexpected shut-downs, usually when viewing
something in a browser. Computer abruptly is OFF; no warning or
procedural shut-down. WinXP SP3.

I'm thinking it may be heat related and the bios is ordering an
immediate shut down as a protective measure.

Motherboard is an MSI 760GM-E51 (FX) with an AMD 3.0 GH quad core processor.

Two questions:

Is there a logfile written upon reboot from an unscheduled shutdown that
might give me a clue? What is its name and where would it be stored?

Are there utilities that could read CPU temperature and display it so I
could test whether it gets near limits?

Rich

| Are there utilities that could read CPU temperature and display it so I
| could test whether it gets near limits?
|

http://www.cpuid.com/softwares/hwmonitor.html

The download doesn't seem to be working right
now, but HWMonitor and CPUID have been around
for some time. I expect it's probably a temporary
glitch. Both programs are handy to have.

On Thu, 26 Feb 2015 08:32:31 -0500, Rich Hare
wrote:

I recently cleaned out my computer case and removed a parallel port
card. Now I am getting unexpected shut-downs, usually when viewing
something in a browser. Computer abruptly is OFF; no warning or
procedural shut-down. WinXP SP3.

I'm thinking it may be heat related and the bios is ordering an
immediate shut down as a protective measure.

Motherboard is an MSI 760GM-E51 (FX) with an AMD 3.0 GH quad core processor.

Two questions:

Is there a logfile written upon reboot from an unscheduled shutdown that
might give me a clue? What is its name and where would it be stored?

Are there utilities that could read CPU temperature and display it so I
could test whether it gets near limits?

Rich

I cant help with the logfile, but I had a computer that would act goofy
when I installed or removed any cards. I found that if I pulled out the
memory, and reinstalled it, the problem would go away. I came to the
conclusion that there was a bad contact, or maybe a loose solder joint.
But I also noted that the motherboard seemed to lack enough support on
the end where the memory and the cards go, and would flex too much when
I plugged in cards. I found a rubber bumper in my hardware box, which
perfectly fit under the MB. I jammed it between th case and the MB
about midway between the front and back of the MB. I used that computer
for several years afterwards and never had that problem again.

Take out your memory, clean the contacts with some rubbing alcohol. Let
it dry real well, and reinstall it. That may or may not fix the
problem.

"Rich Hare" wrote in message ...

I recently cleaned out my computer case and removed a parallel port card.
Now I am getting unexpected shut-downs, usually when viewing something in a
browser. Computer abruptly is OFF; no warning or procedural shut-down.
WinXP SP3.

I'm thinking it may be heat related and the bios is ordering an immediate
shut down as a protective measure.

Motherboard is an MSI 760GM-E51 (FX) with an AMD 3.0 GH quad core
processor.

Two questions:

Is there a logfile written upon reboot from an unscheduled shutdown that
might give me a clue? What is its name and where would it be stored?

Are there utilities that could read CPU temperature and display it so I
could test whether it gets near limits?

Rich
Look inside again and see what is close by that p-port card you removed that
you may have touched or moved slightly while removing it.
If nothing is obvious, unplug and re-connect all connectors. Remove and
reinsert all other cards, including video and memory sticks. Sometimes that
is all it will take.
The programs mentioned to monitor the cpu temp, etc are excellent. Never
choose an automatic install, always go custom, just to make sure some other
programs (adware etc) are not automatically installed besides the one you
want.
--
Buffalo

Rich Hare wrote:
I recently cleaned out my computer case and removed a parallel port
card. Now I am getting unexpected shut-downs, usually when viewing
something in a browser. Computer abruptly is OFF; no warning or
procedural shut-down. WinXP SP3.

I'm thinking it may be heat related and the bios is ordering an
immediate shut down as a protective measure.

Motherboard is an MSI 760GM-E51 (FX) with an AMD 3.0 GH quad core
processor.

Two questions:

Is there a logfile written upon reboot from an unscheduled shutdown that
might give me a clue? What is its name and where would it be stored?

Are there utilities that could read CPU temperature and display it so I
could test whether it gets near limits?

Rich

I'm replying to my own posting with more information.

I thank several people for their help and I THINK I have the problem
solved. It continued to plague me after several attempts to cure it. I
tried a memory tester and it failed each of several tries. I re-seated
the memory sticks to no avail.

Today I had more severe problems, including having the USB mouse freeze
intermittently and the whole computer bog down. It finally shut down
with an error message to the effect of "Error writing to E:\$Mdt". I
rebooted and went to "My Computer" to see whether any message had been
written there (I have three SATA hard drive plus a SATA DVD R/W). Only
TWO of the hard drives were showing up!

Opened the case and reseated all of the SATA data cables and the drives'
power cables. Booted it up and all three drives were shown and labelled
"Healthy" by WinXP. Perhaps I had unseated one of those SATA cables
when I was installing the parallel port.

Anyway, I set up the memory tester (Memtest86) and let it run for four
hours. It completed six iterations with no errors.

I am now hopeful that the computer is back to its previously reliable
state. It is interesting that such a little thing can so easily happen
to disable an otherwise reliable machine.

Rich

Rich Hare wrote:
Rich Hare wrote:
I recently cleaned out my computer case and removed a parallel port
card. Now I am getting unexpected shut-downs, usually when viewing
something in a browser. Computer abruptly is OFF; no warning or
procedural shut-down. WinXP SP3.

I'm thinking it may be heat related and the bios is ordering an
immediate shut down as a protective measure.

Motherboard is an MSI 760GM-E51 (FX) with an AMD 3.0 GH quad core
processor.

Two questions:

Is there a logfile written upon reboot from an unscheduled shutdown
that might give me a clue? What is its name and where would it be
stored?

Are there utilities that could read CPU temperature and display it so
I could test whether it gets near limits?

Rich

I'm replying to my own posting with more information.

I thank several people for their help and I THINK I have the problem
solved. It continued to plague me after several attempts to cure it. I
tried a memory tester and it failed each of several tries. I re-seated
the memory sticks to no avail.

Today I had more severe problems, including having the USB mouse freeze
intermittently and the whole computer bog down. It finally shut down
with an error message to the effect of "Error writing to E:\$Mdt". I
rebooted and went to "My Computer" to see whether any message had been
written there (I have three SATA hard drive plus a SATA DVD R/W). Only
TWO of the hard drives were showing up!

Opened the case and reseated all of the SATA data cables and the drives'
power cables. Booted it up and all three drives were shown and labelled
"Healthy" by WinXP. Perhaps I had unseated one of those SATA cables
when I was installing the parallel port.

Anyway, I set up the memory tester (Memtest86) and let it run for four
hours. It completed six iterations with no errors.

I am now hopeful that the computer is back to its previously reliable
state. It is interesting that such a little thing can so easily happen
to disable an otherwise reliable machine.

Rich

With regard to your "losing a SATA drive" issue.

1) SATA drives have no effective hardware reset subsystem.
This is my first hand experience, having a drive "go missing"
as yours did. On the old IDE drives, there is a hardware "RESET#"
signal, and when the motherboard starts up, a pulse on that line
causes "sanity" to return to IDE drives. If they have an issue,
that hardware path should restore them (barring completely failed
hardware, and not just a transient issue).

The thin SATA data cable, has no such function. There are too few
wires for a discrete RESET signal. I'd hoped there would be an
out-of-band JK code that could be sent, but the lack of sanity on
my drive, suggests they don't have such a capability built in.

2) OK, so what caused my SATA drive problem ? It was a power problem.
I put too many drives on the same power cable. If the 12V rail or 5V
rail are slightly out of spec, the drive can attempt to reinitialize
itself. And sometimes, the thing just gets hung up. Some cheap Molex
4 pin power cables, don't fit tightly enough (the original Molex
branded ones were excellent, while the knockoffs suck).

Check that you haven't used too many third-party Y-cables or
drive-power-connector-type adapters in a row.

As soon as I re-routed power, and removed my wiring rats next,
the "disappearing drive" issue went away.

3) There are several generations of SATA drive cablings. The first
generation did not have positive retention features. The power
connector could fall off. Later generations got some "metal bits"
to aid in retention. They work great, except... some WD drives have
too wide a gap near the connectors, for the retention to have
"something to bite onto". WD did this, because WD sells a WD-branded
cable that snugly fits in there.

So if you ever reboot a computer, and one of the drives (even the
boot drive) "disappears", a simple power cycle will fix this. For
longer term satisfaction, examine how the power is distributed
inside the PC.

I've never seen an IDE here, disappear on me. Not in the same way
as SATA. And I've had several of these SATA occurrences, so this is
something that a lot of people are going to experience at some
point. Just don't panic... :-)

*******

In your description, I didn't see any "100% CPU loading" test cases.
You need something to peg all the cores on your processor, then
watch CPU temperature. The cooling system on your CPU is supposed
to take that, and keep the CPU cool. If the PC shuts off while
running such a test, then fix it. Either use fresh thermal paste
on the CPU heatsink, or examine whether the CPU cooler is really
intended for the power levels of that processor. And so on.
A computer that is idle now (and running cool), you could
go to bed, some Windows Update starts to run, the CPU is pegged,
the CPU heats up, the PC shuts off, the file system is corrupted
and a CHKDSK needs to be run. Always make sure your PC is sufficiently
fit, to run with the CPU pegged at 100% for extended periods of
time. Because the PC can enter that state on its own, when
no one is around to watch it or help it. Or put ice cubes
on top of the chassis (to cool something).

Even a brand new computer, that you just got from Dell, should
receive a few tests. You should run memtest86+ for a few passes.
You should run a program that pegs the CPU (I like Prime95 from
mersenne.org and the included torture-test). That's to make sure
the cooling system is working right. The Dell fan should speed
up, and a temperature measurement program should show nice temps
(like 65C Tcase perhaps). In any event, you want the CPU to be
below the Throttle temperature. If the CPU hits the Throttle
point, you're effectively not getting the full clock rate
from it. And that would be a waste of your money (for the
fancy CPU).

Paul