Extra-frequent unwanted power off while using ACPI

I have a Toshiba Satellite 1800-314, and I have had this problem for
more than a month. I don't remember doing anything special that could
cause this. In the preinstalled Windows XP, the computer very often
powers off automatically (not a normal shutdown but an instant power
off). I think the frequency of the event has increased greatly later.
I normally use GNU/Linux instead, and this has never happened under
it. I have finally discovered that the shutdowns are caused by ACPI.
In (my) Windows, ACPI is used. In Linux, it was normally disabled but
when I enabled it to try, it immediately started powering off just
like in Windows. With APM it works well.

I didn't take this seriously before and thought it was probably a
Windows bug, but often I need to use a Windows-only program and the
power-offs are so frequent that it is usable. Often it happens
immediately when Windows starts. That sometimes happens five times in
a row. I lose a lot of time because of this. I had to disable the disk
write cache in Windows completely in order to decrease data loss.

So, can someone explain why this occurs? It is acceptable for me to
make Windows use APM if possible (I don't know how yet), but I think
it would be much better to find the source of the problem, because
ACPI has some advantages over APM and more importantly because I might
want to sell the computer in the future
and this would be a problem.

Possibly helpful info:

It happens always when the load (probably CPU) is rather high. By
high, I mean just a few browser windows or maybe a few megabytes of
file in Notepad. Also frequently happens when Windows starts (the load
is high then). The computer runs mostly for about 5 to 40 minutes, but
usually less, and power off.

I nearly always run in AC powered mode, but occasionally on battery
too, and I don't remember this happening on battery power.

The problem began at a time later than the last summer.

acpi is an efficient use of sharing irqs. nothing to do with power
management. right click on the battery icon and select always on. go into
control panel and power options and select always on or desktop or
something.

--
Ian Lincoln Independent I.T Consultant
"Sinan Imamoglu" <> wrote in message
news: om...
> I have a Toshiba Satellite 1800-314, and I have had this problem for
> more than a month. I don't remember doing anything special that could
> cause this. In the preinstalled Windows XP, the computer very often
> powers off automatically (not a normal shutdown but an instant power
> off). I think the frequency of the event has increased greatly later.
> I normally use GNU/Linux instead, and this has never happened under
> it. I have finally discovered that the shutdowns are caused by ACPI.
> In (my) Windows, ACPI is used. In Linux, it was normally disabled but
> when I enabled it to try, it immediately started powering off just
> like in Windows. With APM it works well.
>
> I didn't take this seriously before and thought it was probably a
> Windows bug, but often I need to use a Windows-only program and the
> power-offs are so frequent that it is usable. Often it happens
> immediately when Windows starts. That sometimes happens five times in
> a row. I lose a lot of time because of this. I had to disable the disk
> write cache in Windows completely in order to decrease data loss.
>
> So, can someone explain why this occurs? It is acceptable for me to
> make Windows use APM if possible (I don't know how yet), but I think
> it would be much better to find the source of the problem, because
> ACPI has some advantages over APM and more importantly because I might
> want to sell the computer in the future
> and this would be a problem.
>
> Possibly helpful info:
>
> It happens always when the load (probably CPU) is rather high. By
> high, I mean just a few browser windows or maybe a few megabytes of
> file in Notepad. Also frequently happens when Windows starts (the load
> is high then). The computer runs mostly for about 5 to 40 minutes, but
> usually less, and power off.
>
> I nearly always run in AC powered mode, but occasionally on battery
> too, and I don't remember this happening on battery power.
>
> The problem began at a time later than the last summer.

I have a similar problem with my Sony Vaio GRZ660. However I think my
problem is related to the laptop getting too hot and shutting down to
protect the cpu. I just haven't got round to calling Sony yet.

I tried a tool called motherboard monitor with no luck on my machine. It
might work on yours to give you some idea of temperatures to see if that
mnay be the problem. You should be able to find it searching for MBM or
motherboard monitor.

Paul

"Sinan Imamoglu" <> wrote in message
news: om...
> I have a Toshiba Satellite 1800-314, and I have had this problem for
> more than a month. I don't remember doing anything special that could
> cause this. In the preinstalled Windows XP, the computer very often
> powers off automatically (not a normal shutdown but an instant power
> off). I think the frequency of the event has increased greatly later.
> I normally use GNU/Linux instead, and this has never happened under
> it. I have finally discovered that the shutdowns are caused by ACPI.
> In (my) Windows, ACPI is used. In Linux, it was normally disabled but
> when I enabled it to try, it immediately started powering off just
> like in Windows. With APM it works well.
>
> I didn't take this seriously before and thought it was probably a
> Windows bug, but often I need to use a Windows-only program and the
> power-offs are so frequent that it is usable. Often it happens
> immediately when Windows starts. That sometimes happens five times in
> a row. I lose a lot of time because of this. I had to disable the disk
> write cache in Windows completely in order to decrease data loss.
>
> So, can someone explain why this occurs? It is acceptable for me to
> make Windows use APM if possible (I don't know how yet), but I think
> it would be much better to find the source of the problem, because
> ACPI has some advantages over APM and more importantly because I might
> want to sell the computer in the future
> and this would be a problem.
>
> Possibly helpful info:
>
> It happens always when the load (probably CPU) is rather high. By
> high, I mean just a few browser windows or maybe a few megabytes of
> file in Notepad. Also frequently happens when Windows starts (the load
> is high then). The computer runs mostly for about 5 to 40 minutes, but
> usually less, and power off.
>
> I nearly always run in AC powered mode, but occasionally on battery
> too, and I don't remember this happening on battery power.
>
> The problem began at a time later than the last summer.

Sinan Imamoglu wrote:
> I have a Toshiba Satellite 1800-314, and I have had this problem for
> more than a month. I don't remember doing anything special that could
> cause this. In the preinstalled Windows XP, the computer very often
> powers off automatically (not a normal shutdown but an instant power
> off). I think the frequency of the event has increased greatly later.
> I normally use GNU/Linux instead, and this has never happened under
> it. I have finally discovered that the shutdowns are caused by ACPI.
> In (my) Windows, ACPI is used. In Linux, it was normally disabled but
> when I enabled it to try, it immediately started powering off just
> like in Windows. With APM it works well.
>
> I didn't take this seriously before and thought it was probably a
> Windows bug, but often I need to use a Windows-only program and the
> power-offs are so frequent that it is usable. Often it happens
> immediately when Windows starts. That sometimes happens five times in
> a row. I lose a lot of time because of this. I had to disable the disk
> write cache in Windows completely in order to decrease data loss.
>
> So, can someone explain why this occurs? It is acceptable for me to
> make Windows use APM if possible (I don't know how yet), but I think
> it would be much better to find the source of the problem, because
> ACPI has some advantages over APM and more importantly because I might
> want to sell the computer in the future
> and this would be a problem.
>
> Possibly helpful info:
>
> It happens always when the load (probably CPU) is rather high. By
> high, I mean just a few browser windows or maybe a few megabytes of
> file in Notepad. Also frequently happens when Windows starts (the load
> is high then). The computer runs mostly for about 5 to 40 minutes, but
> usually less, and power off.
>
> I nearly always run in AC powered mode, but occasionally on battery
> too, and I don't remember this happening on battery power.
>
> The problem began at a time later than the last summer.

OEM XP is typically set up with automatic restart on a system error. The
system error is displayed, but many times the shutdown is too fast to
see it. From Device Manager, select Advanced Tab/Startup and Recovery
Settings button. Under System Failure, uncheck the box Automatic
Restart, and check the first box Write an event....

You can check if error messages are written to Control
Panel/AdministrativeTools/Event Viewer.

Ascertain that the problem is not a system failure.

From Device Manager, look for any device with a yellow or red warning
icon. If there is one, it might be related.

Q

"Quaoar" <> wrote in message news:<n6iKa.14705$Bg.9583@rwcrnsc54>...
> OEM XP is typically set up with automatic restart on a system error. The
> system error is displayed, but many times the shutdown is too fast to
> see it. From Device Manager, select Advanced Tab/Startup and Recovery
> Settings button. Under System Failure, uncheck the box Automatic
> Restart, and check the first box Write an event....
>
> You can check if error messages are written to Control
> Panel/AdministrativeTools/Event Viewer.
>
> Ascertain that the problem is not a system failure.
>
> From Device Manager, look for any device with a yellow or red warning
> icon. If there is one, it might be related.
>

I disabled Automatic restart as you suggested, Write an event was
already enabled. After doing that, I also installed Motherboard
Monitor as Paul Cook suggested. In that session there was no problem.
I restarted to see if was solved but it had became even much worse. I
had to boot 15 times and it didn't go for more than a minute. Then I
deleted the Motherboard Monitor SYS file from Linux (probably I made a
wrong choice while installing during setup) and up to now it seems to
work well. While deleting MBM I also removed the TI Connect (links the
computer to calculator) because in one of the blue screens during
booting I saw the message PAGE_FAULT_IN_NONPAGED_AREA and ticalc.sys.
This might be related to the problem. During my usage of this and
other calculator linking programs I had several power-offs before. But
this doesn't explain why Linux has a very similar problem when ACPI is
enabled. Maybe it is just a coincidence. I looked at /var/log/acpid
and the following was at the end of it:

[Wed Jun 25 23:13:34 2003] starting up
[Wed Jun 25 23:13:34 2003] 1 rule loaded
# the following two nine times
[Wed Jun 25 23:17:31 2003] received event "battery BAT1 00000080
00000001"
[Wed Jun 25 23:17:31 2003] completed event "battery BAT1 00000080
00000001"
# the following two four times
[Wed Jun 25 23:17:32 2003] received event "thermal_zone THRM 000000f0
00000001"
[Wed Jun 25 23:17:32 2003] completed event "thermal_zone THRM 000000f0
00000001"

I don't understand it but maybe the thermal_zone is related? Looking
at other threads, an idea came to me that it could be related to the
ever-increasing temperature. Anyway, I will post here if any
significant change occurs. As I said, it looks OK now. I thank you all
very much for helping.

On 24 Jun 2003 08:50:08 -0700, (Sinan Imamoglu) wrote:

:>I have a Toshiba Satellite 1800-314, and I have had this problem for
:>more than a month. I don't remember doing anything special that could
:>cause this. In the preinstalled Windows XP, the computer very often
:>powers off automatically (not a normal shutdown but an instant power
:>off). I think the frequency of the event has increased greatly later.
:>I normally use GNU/Linux instead, and this has never happened under
:>it. I have finally discovered that the shutdowns are caused by ACPI.
:>In (my) Windows, ACPI is used. In Linux, it was normally disabled but
:>when I enabled it to try, it immediately started powering off just
:>like in Windows. With APM it works well.
:>
:>I didn't take this seriously before and thought it was probably a
:>Windows bug, but often I need to use a Windows-only program and the
:>power-offs are so frequent that it is usable. Often it happens
:>immediately when Windows starts. That sometimes happens five times in
:>a row. I lose a lot of time because of this. I had to disable the disk
:>write cache in Windows completely in order to decrease data loss.
:>
:>So, can someone explain why this occurs? It is acceptable for me to
:>make Windows use APM if possible (I don't know how yet), but I think
:>it would be much better to find the source of the problem, because
:>ACPI has some advantages over APM and more importantly because I might
:>want to sell the computer in the future
:>and this would be a problem.
:>
:>Possibly helpful info:
:>
:>It happens always when the load (probably CPU) is rather high. By
:>high, I mean just a few browser windows or maybe a few megabytes of
:>file in Notepad. Also frequently happens when Windows starts (the load
:>is high then). The computer runs mostly for about 5 to 40 minutes, but
:>usually less, and power off.
:>
:>I nearly always run in AC powered mode, but occasionally on battery
:>too, and I don't remember this happening on battery power.
:>
:>The problem began at a time later than the last summer.

This is a known issue with the 1800 series (at least the Toshiba America models,
and probably yours too). The original ACPI BIOS had problems communicating with
any ACPI enabled operating system. The situation wasn't helped by the fact that
the 1800 series uses a desktop CPU chip which are designed to shut down if they
get too hot. This was fixed (for about 95% of the cases) on the US models by a
BIOS update (currently 2.10 for the US models). Check the Toshiba site for the
latest BIOS and try installing that. Hopefully that will fix your problem as it
has for the majority of the people with your problem.

Good luck,
OscarT

--
Please note that any opinions stated in my message are strictly my own
and are not to be taken as official statements either from my employer
or the manufacturer of any mentioned product