I see on the Asus website forum that other users of this motherboard have
the same problem. Basically the PC switches itself off (not a controlled
shutdown) as if the power plug has been pulled. The machine can be provoked
into failure by intensive memory access.

I have eliminated all other components by substitution and wondered whether
anyone in the UK had managed to find a fix for the problem?

Ian

I use this m/b for 4 years now and never had this problem.
However from my experience with other m/b when this
fault of switching off or rebooting itself the culprit was PSU.
So swap PSU and you'll see if my "diagnostic" is correct...

--
Jdr
.................................................. .................................................. ..
http://homepage.ntlworld.com/josephd.ross/index.html
"Ian" <> wrote in message
news:...
>I see on the Asus website forum that other users of this motherboard have
> the same problem. Basically the PC switches itself off (not a controlled
> shutdown) as if the power plug has been pulled. The machine can be
> provoked
> into failure by intensive memory access.
>
> I have eliminated all other components by substitution and wondered
> whether
> anyone in the UK had managed to find a fix for the problem?
>
> Ian
>
>

"McGrath" <> wrote in message
news:J0Rij.48359$...
>I use this m/b for 4 years now and never had this problem.
> However from my experience with other m/b when this
> fault of switching off or rebooting itself the culprit was PSU.
> So swap PSU and you'll see if my "diagnostic" is correct...
>
> --
> Jdr


>>I see on the Asus website forum that other users of this motherboard have
>> the same problem. Basically the PC switches itself off (not a controlled
>> shutdown) as if the power plug has been pulled. The machine can be
>> provoked
>> into failure by intensive memory access.
>>
>> I have eliminated all other components by substitution and wondered
>> whether
>> anyone in the UK had managed to find a fix for the problem?
>>
>> Ian
>>

Jdr

Thanks for your suggestion, I have tried two other borrowed PSUs and also
bought a brand new 500Watt one after being told that the PSU was the most
likely suspect. The behaviour was the same with all of them. I have removed
all add-on cards and simplified the machine as well.

Whatever is happening on this faulty board triggers the power supply to
switch off instantly rather than the way XP normally does it.

Just to add interest... I tried another old hard disk I had lying around (so
as not to risk losing data on the original C: drive). This old disk still
had Win 98 on it and when the P4C8000 tries to boot with it switches off
everytime just as the win 98 title screen comes up. At least I now have a
consistent means of testing!

Ian

Ian wrote:
> I see on the Asus website forum that other users of this motherboard have
> the same problem. Basically the PC switches itself off (not a controlled
> shutdown) as if the power plug has been pulled. The machine can be provoked
> into failure by intensive memory access.
>
> I have eliminated all other components by substitution and wondered whether
> anyone in the UK had managed to find a fix for the problem?
>
> Ian
>
>
Have you checked the capacitors on the board to make sure that one of
them is not failing? I recently had to RMA my P5P800 because of random
reboots because of a swollen capacitor near the power supply.

Ian wrote:
> "McGrath" <> wrote in message
> news:J0Rij.48359$...
>> I use this m/b for 4 years now and never had this problem.
>> However from my experience with other m/b when this
>> fault of switching off or rebooting itself the culprit was PSU.
>> So swap PSU and you'll see if my "diagnostic" is correct...
>>
>> --
>> Jdr
>
>
>>> I see on the Asus website forum that other users of this motherboard have
>>> the same problem. Basically the PC switches itself off (not a controlled
>>> shutdown) as if the power plug has been pulled. The machine can be
>>> provoked
>>> into failure by intensive memory access.
>>>
>>> I have eliminated all other components by substitution and wondered
>>> whether
>>> anyone in the UK had managed to find a fix for the problem?
>>>
>>> Ian
>>>
>
> Jdr
>
> Thanks for your suggestion, I have tried two other borrowed PSUs and also
> bought a brand new 500Watt one after being told that the PSU was the most
> likely suspect. The behaviour was the same with all of them. I have removed
> all add-on cards and simplified the machine as well.
>
> Whatever is happening on this faulty board triggers the power supply to
> switch off instantly rather than the way XP normally does it.
>
> Just to add interest... I tried another old hard disk I had lying around (so
> as not to risk losing data on the original C: drive). This old disk still
> had Win 98 on it and when the P4C8000 tries to boot with it switches off
> everytime just as the win 98 title screen comes up. At least I now have a
> consistent means of testing!
>
> Ian
>
>

Have you examined the board for bulging or leaking electrolytic capacitors ?
The top of the caps is supposed to remain flat. When they leak, there may be a
brown stain near the base of the cap.

You can try running memtest86+, to see if it is memory related.
Memtest86+ boots from a floppy or from a CDROM.

You can also try Prime95 (assuming the thing stays up long enough).
If the machine dies, just as Prime95 is starting the torture test,
that could be some kind of Vcore power problem.

I'm not familiar with all the ways that a shutdown can occur. These
are some I know of:

1) Shutdown via the normal software path. Obviously, there is a well
known address somewhere in the Southbridge or SuperI/O chip, that
causes the machine to shut off. That would be by deasserting PS_ON#.
Any (corrupted) software that happens to access that hardware,
could trigger it.

2) Processor overheat (THERMTRIP) could shut off the computer. That
protects the processor from damage. On older processors, the
trigger temp is rather high. On newer ones, it might get triggered
at 90C or so.

3) The Vcore regulator will have overcurrent limiting, but I don't think
it is necessarily tied to PS_ON#. It is possible for the regulator to
protect itself, without shutting off all power. The fans would run
but the monitor screen might be blank.

4) Motherboards with "AGP warn" (red LED located near the AGP slot),
have the ability to switch off PS_ON#. A damaged AGP warn circuit
could do something like that. Note that Asus removed that circuit
from later generations of boards. My P4B board is the first
I know of, to use the circuit. To save a few pennies, they removed
the circuit, on the assumption that erroneously keyed 3.3V only
AGP cards no longer existed. I also have a suspicion that some boards
still have the transistor circuit, but no red LED for a display.

This thread has one mention of a matching symptom. Running for 1
minute before switching off.

http://vip.asus.com/forum/view.aspx?...Language=en-us

More cases here:
http://vip.asus.com/forum/view.aspx?...Language=en-us

Interesting here - "The house fan seemed to stop it from shutting down."
http://vip.asus.com/forum/view.aspx?...Language=en-us

Note that there is a problem with latchup failure of ICH5 (affecting
a wide range of brands and models of motherboard), but the symptoms
of a major latchup failure are a burn mark on the top of the ICH5,
and the computer will not POST again. A minor failure results in loss
of usage of the USB ports, but the computer still boots. So your
symptoms don't seem to match what is known about ICH5 problems.

Paul

On 15-01-2008 01:44:15, Paul wrote:
> Ian wrote:
>> "McGrath" <> wrote in message
>> news:J0Rij.48359$...
>>> I use this m/b for 4 years now and never had this problem.
>>> However from my experience with other m/b when this
>>> fault of switching off or rebooting itself the culprit was PSU.
>>> So swap PSU and you'll see if my "diagnostic" is correct...
>>>
>>> --
>>> Jdr


> 2) Processor overheat (THERMTRIP) could shut off the computer. That
> protects the processor from damage. On older processors, the
> trigger temp is rather high. On newer ones, it might get triggered
> at 90C or so.

I thing you have trouble here.
/skibber

"Paul" <> wrote in message news:fmgvek$acf$...
> Ian wrote:
>> "McGrath" <> wrote in message
>> news:J0Rij.48359$...
>>> I use this m/b for 4 years now and never had this problem.
>>> However from my experience with other m/b when this
>>> fault of switching off or rebooting itself the culprit was PSU.
>>> So swap PSU and you'll see if my "diagnostic" is correct...
>>>
>>> --
>>> Jdr
>>
>>
>>>> I see on the Asus website forum that other users of this motherboard
>>>> have
>>>> the same problem. Basically the PC switches itself off (not a
>>>> controlled
>>>> shutdown) as if the power plug has been pulled. The machine can be
>>>> provoked
>>>> into failure by intensive memory access.
>>>>
>>>> I have eliminated all other components by substitution and wondered
>>>> whether
>>>> anyone in the UK had managed to find a fix for the problem?
>>>>
>>>> Ian
>>>>
>>
>> Jdr
>>
>> Thanks for your suggestion, I have tried two other borrowed PSUs and also
>> bought a brand new 500Watt one after being told that the PSU was the most
>> likely suspect. The behaviour was the same with all of them. I have
>> removed all add-on cards and simplified the machine as well.
>>
>> Whatever is happening on this faulty board triggers the power supply to
>> switch off instantly rather than the way XP normally does it.
>>
>> Just to add interest... I tried another old hard disk I had lying around
>> (so as not to risk losing data on the original C: drive). This old disk
>> still had Win 98 on it and when the P4C8000 tries to boot with it
>> switches off everytime just as the win 98 title screen comes up. At least
>> I now have a consistent means of testing!
>>
>> Ian
>
> Have you examined the board for bulging or leaking electrolytic capacitors
> ?
> The top of the caps is supposed to remain flat. When they leak, there may
> be a
> brown stain near the base of the cap.
>
> You can try running memtest86+, to see if it is memory related.
> Memtest86+ boots from a floppy or from a CDROM.
>
> You can also try Prime95 (assuming the thing stays up long enough).
> If the machine dies, just as Prime95 is starting the torture test,
> that could be some kind of Vcore power problem.
>
> I'm not familiar with all the ways that a shutdown can occur. These
> are some I know of:
>
> 1) Shutdown via the normal software path. Obviously, there is a well
> known address somewhere in the Southbridge or SuperI/O chip, that
> causes the machine to shut off. That would be by deasserting PS_ON#.
> Any (corrupted) software that happens to access that hardware,
> could trigger it.
>
> 2) Processor overheat (THERMTRIP) could shut off the computer. That
> protects the processor from damage. On older processors, the
> trigger temp is rather high. On newer ones, it might get triggered
> at 90C or so.
>
> 3) The Vcore regulator will have overcurrent limiting, but I don't think
> it is necessarily tied to PS_ON#. It is possible for the regulator to
> protect itself, without shutting off all power. The fans would run
> but the monitor screen might be blank.
>
> 4) Motherboards with "AGP warn" (red LED located near the AGP slot),
> have the ability to switch off PS_ON#. A damaged AGP warn circuit
> could do something like that. Note that Asus removed that circuit
> from later generations of boards. My P4B board is the first
> I know of, to use the circuit. To save a few pennies, they removed
> the circuit, on the assumption that erroneously keyed 3.3V only
> AGP cards no longer existed. I also have a suspicion that some boards
> still have the transistor circuit, but no red LED for a display.
>
> This thread has one mention of a matching symptom. Running for 1
> minute before switching off.
>
> http://vip.asus.com/forum/view.aspx?...Language=en-us
>
> More cases here:
> http://vip.asus.com/forum/view.aspx?...Language=en-us
>
> Interesting here - "The house fan seemed to stop it from shutting down."
> http://vip.asus.com/forum/view.aspx?...Language=en-us
>
> Note that there is a problem with latchup failure of ICH5 (affecting
> a wide range of brands and models of motherboard), but the symptoms
> of a major latchup failure are a burn mark on the top of the ICH5,
> and the computer will not POST again. A minor failure results in loss
> of usage of the USB ports, but the computer still boots. So your
> symptoms don't seem to match what is known about ICH5 problems.
>
> Paul

Paul

I appreciate your very detailed response, I had read all the messages on the
Asus forum and had decided that nobody had actually found a solution. To
give a bit more background info, this pc is used specifically for video
editing. Normally it has PCI cards including SCSI controller and a dual
monitor Matrox card as well as the main 'Velocity' video editing hardware
card (which also has its own dedicated SCSI bus and external Raid drives).
The machine has been in use for several years until the original P4C800 (not
deluxe) died and the deluxe board was recently purchased to replace it so
that the original components would just transfer. With the deluxe installed
the machine works as it did previously (apart from the shutdowns) but in the
course of investigating the problem the machine has been simplified to a
bare minimum spec and PSU, processor, memory, drives etc have all been
substituted with known good items.

I do not have the schematic of the board but assume the 'PS_ON#' you mention
is the 'power good' line between the board and the PSU? I had toyed with the
idea of inserting a switch in whatever signal line that controls the PSU as
a temporary diagnostic aid, I may be wrong but feel that the machine would
carry on running correctly (software wise) if it the power stayed on.

I am sure the fault is not heat related as it switches off within a few
seconds from cold start, its just that something in the machines inner
workings is incorrectly telling it to shut down.

Ian

Ian wrote:

>
> Paul
>
> I appreciate your very detailed response, I had read all the messages on the
> Asus forum and had decided that nobody had actually found a solution. To
> give a bit more background info, this pc is used specifically for video
> editing. Normally it has PCI cards including SCSI controller and a dual
> monitor Matrox card as well as the main 'Velocity' video editing hardware
> card (which also has its own dedicated SCSI bus and external Raid drives).
> The machine has been in use for several years until the original P4C800 (not
> deluxe) died and the deluxe board was recently purchased to replace it so
> that the original components would just transfer. With the deluxe installed
> the machine works as it did previously (apart from the shutdowns) but in the
> course of investigating the problem the machine has been simplified to a
> bare minimum spec and PSU, processor, memory, drives etc have all been
> substituted with known good items.
>
> I do not have the schematic of the board but assume the 'PS_ON#' you mention
> is the 'power good' line between the board and the PSU? I had toyed with the
> idea of inserting a switch in whatever signal line that controls the PSU as
> a temporary diagnostic aid, I may be wrong but feel that the machine would
> carry on running correctly (software wise) if it the power stayed on.
>
> I am sure the fault is not heat related as it switches off within a few
> seconds from cold start, its just that something in the machines inner
> workings is incorrectly telling it to shut down.
>
> Ian
>
>

PS_ON# is an open collector signal on the main power connector. What you
suggest would be an interesting experiment. If you kept PS_ON# shorted
to COM (i.e. logic ground), that would maintain the power in a running
state. It would also not damage the motherboard driver, since open
collector logic can be shorted to ground. (That assumes that designers
follow rules, and don't take shortcuts by using non-OC logic to drive
the signal.) So, for the most part, your experiment is a worthy one.

A downside though, is it defeats the real THERMTRIP feature. Say your
CPU heatsink fell off while you were doing the experiment. Then, you'd
no longer have that safety feature. I probably wouldn't let that stop
me though :-)

As for a schematic for the board, you aren't likely to get a real schematic
for a retail board. Intel does, in fact, have a schematic for the 875P
chipset, which is available for public download. What is wrong with it ?
Well, it uses proprietary stuff for portions of the board. I had high hopes
it would be a real education, but it is only partially satisfying in
terms of what they did. To my knowledge, Intel hasn't released a
schematic like this, since this one.

http://www.intel.com/design/chipsets...ics/252812.htm

One thing you may notice, is there is a big difference between the
"power train" on tbe Intel schematic, and what you find on an Asus
board. Intel appears to have put all sorts of protection in the
power paths. Judging by the simpler looking Asus implementation,
Asus considered some of that to be unnecessary. The Intel schematic
would be an attempt to "do it right", while Asus engineers are
more in tune with "do what's necessary". When you make millions of
motherboards, stripping off components is part of your profit margin.

If you need pinouts and wire colors for ATX, you can find them here.

24 pin
http://www.formfactors.org/developer...public_br2.pdf

20 pin
http://www.formfactors.org/developer...X12V_1_3dg.pdf

Paul

"Paul" <> wrote in message news:fmi380$ene$...
> Ian wrote:
>
>>
>> Paul
>>
>> I appreciate your very detailed response, I had read all the messages on
>> the Asus forum and had decided that nobody had actually found a solution.
>> To give a bit more background info, this pc is used specifically for
>> video editing. Normally it has PCI cards including SCSI controller and a
>> dual monitor Matrox card as well as the main 'Velocity' video editing
>> hardware card (which also has its own dedicated SCSI bus and external
>> Raid drives). The machine has been in use for several years until the
>> original P4C800 (not deluxe) died and the deluxe board was recently
>> purchased to replace it so that the original components would just
>> transfer. With the deluxe installed the machine works as it did
>> previously (apart from the shutdowns) but in the course of investigating
>> the problem the machine has been simplified to a bare minimum spec and
>> PSU, processor, memory, drives etc have all been substituted with known
>> good items.
>>
>> I do not have the schematic of the board but assume the 'PS_ON#' you
>> mention is the 'power good' line between the board and the PSU? I had
>> toyed with the idea of inserting a switch in whatever signal line that
>> controls the PSU as a temporary diagnostic aid, I may be wrong but feel
>> that the machine would carry on running correctly (software wise) if it
>> the power stayed on.
>>
>> I am sure the fault is not heat related as it switches off within a few
>> seconds from cold start, its just that something in the machines inner
>> workings is incorrectly telling it to shut down.
>>
>> Ian
>>
>>
>
> PS_ON# is an open collector signal on the main power connector. What you
> suggest would be an interesting experiment. If you kept PS_ON# shorted
> to COM (i.e. logic ground), that would maintain the power in a running
> state. It would also not damage the motherboard driver, since open
> collector logic can be shorted to ground. (That assumes that designers
> follow rules, and don't take shortcuts by using non-OC logic to drive
> the signal.) So, for the most part, your experiment is a worthy one.
>
> A downside though, is it defeats the real THERMTRIP feature. Say your
> CPU heatsink fell off while you were doing the experiment. Then, you'd
> no longer have that safety feature. I probably wouldn't let that stop
> me though :-)
>
> As for a schematic for the board, you aren't likely to get a real
> schematic
> for a retail board. Intel does, in fact, have a schematic for the 875P
> chipset, which is available for public download. What is wrong with it ?
> Well, it uses proprietary stuff for portions of the board. I had high
> hopes
> it would be a real education, but it is only partially satisfying in
> terms of what they did. To my knowledge, Intel hasn't released a
> schematic like this, since this one.
>
> http://www.intel.com/design/chipsets...ics/252812.htm
>
> One thing you may notice, is there is a big difference between the
> "power train" on tbe Intel schematic, and what you find on an Asus
> board. Intel appears to have put all sorts of protection in the
> power paths. Judging by the simpler looking Asus implementation,
> Asus considered some of that to be unnecessary. The Intel schematic
> would be an attempt to "do it right", while Asus engineers are
> more in tune with "do what's necessary". When you make millions of
> motherboards, stripping off components is part of your profit margin.
>
> If you need pinouts and wire colors for ATX, you can find them here.
>
> 24 pin
> http://www.formfactors.org/developer...public_br2.pdf
>
> 20 pin
> http://www.formfactors.org/developer...X12V_1_3dg.pdf
>
> Paul

Paul

Absolutely brilliant! You have given me some really good info which I am
going to follow up on.

I will not have access to the machine for a few days yet to experiment. If I
do as you say (holding pin to ground) will the XP still be able to shut down
the machine, or will I need to mount an external switch to disable the mod
to power down?

Ian

In article <>, "Ian" <> wrote:
>
>"Paul" <> wrote in message news:fmgvek$acf$...
>> Ian wrote:
>>> "McGrath" <> wrote in message
>>> news:J0Rij.48359$...
>>>> I use this m/b for 4 years now and never had this problem.
>>>> However from my experience with other m/b when this
>>>> fault of switching off or rebooting itself the culprit was PSU.
>>>> So swap PSU and you'll see if my "diagnostic" is correct...
>>>>
>>>> --
>>>> Jdr
>>>
>>>
>>>>> I see on the Asus website forum that other users of this motherboard
>>>>> have
>>>>> the same problem. Basically the PC switches itself off (not a
>>>>> controlled
>>>>> shutdown) as if the power plug has been pulled. The machine can be
>>>>> provoked
>>>>> into failure by intensive memory access.
>>>>>
>>>>> I have eliminated all other components by substitution and wondered
>>>>> whether
>>>>> anyone in the UK had managed to find a fix for the problem?
>>>>>
>>>>> Ian
>>>>>
>>>
>>> Jdr
>>>
>>> Thanks for your suggestion, I have tried two other borrowed PSUs and also
>>> bought a brand new 500Watt one after being told that the PSU was the most
>>> likely suspect. The behaviour was the same with all of them. I have
>>> removed all add-on cards and simplified the machine as well.
>>>
>>> Whatever is happening on this faulty board triggers the power supply to
>>> switch off instantly rather than the way XP normally does it.
>>>
>>> Just to add interest... I tried another old hard disk I had lying around
>>> (so as not to risk losing data on the original C: drive). This old disk
>>> still had Win 98 on it and when the P4C8000 tries to boot with it
>>> switches off everytime just as the win 98 title screen comes up. At least
>>> I now have a consistent means of testing!
>>>
>>> Ian
>>
>> Have you examined the board for bulging or leaking electrolytic capacitors
>> ?
>> The top of the caps is supposed to remain flat. When they leak, there may
>> be a
>> brown stain near the base of the cap.
>>
>> You can try running memtest86+, to see if it is memory related.
>> Memtest86+ boots from a floppy or from a CDROM.
>>
>> You can also try Prime95 (assuming the thing stays up long enough).
>> If the machine dies, just as Prime95 is starting the torture test,
>> that could be some kind of Vcore power problem.
>>
>> I'm not familiar with all the ways that a shutdown can occur. These
>> are some I know of:
>>
>> 1) Shutdown via the normal software path. Obviously, there is a well
>> known address somewhere in the Southbridge or SuperI/O chip, that
>> causes the machine to shut off. That would be by deasserting PS_ON#.
>> Any (corrupted) software that happens to access that hardware,
>> could trigger it.
>>
>> 2) Processor overheat (THERMTRIP) could shut off the computer. That
>> protects the processor from damage. On older processors, the
>> trigger temp is rather high. On newer ones, it might get triggered
>> at 90C or so.
>>
>> 3) The Vcore regulator will have overcurrent limiting, but I don't think
>> it is necessarily tied to PS_ON#. It is possible for the regulator to
>> protect itself, without shutting off all power. The fans would run
>> but the monitor screen might be blank.
>>
>> 4) Motherboards with "AGP warn" (red LED located near the AGP slot),
>> have the ability to switch off PS_ON#. A damaged AGP warn circuit
>> could do something like that. Note that Asus removed that circuit
>> from later generations of boards. My P4B board is the first
>> I know of, to use the circuit. To save a few pennies, they removed
>> the circuit, on the assumption that erroneously keyed 3.3V only
>> AGP cards no longer existed. I also have a suspicion that some boards
>> still have the transistor circuit, but no red LED for a display.
>>
>> This thread has one mention of a matching symptom. Running for 1
>> minute before switching off.
>>
>>
> http://vip.asus.com/forum/view.aspx?...d=1&model=P4C8
>00+Deluxe&page=1&SLanguage=en-us
>>
>> More cases here:
>>
> http://vip.asus.com/forum/view.aspx?...d=1&model=P4C8
>00+Deluxe&page=1&SLanguage=en-us
>>
>> Interesting here - "The house fan seemed to stop it from shutting down."
>>
> http://vip.asus.com/forum/view.aspx?...d=1&model=P4C8
>00+Deluxe&page=1&SLanguage=en-us
>>
>> Note that there is a problem with latchup failure of ICH5 (affecting
>> a wide range of brands and models of motherboard), but the symptoms
>> of a major latchup failure are a burn mark on the top of the ICH5,
>> and the computer will not POST again. A minor failure results in loss
>> of usage of the USB ports, but the computer still boots. So your
>> symptoms don't seem to match what is known about ICH5 problems.
>>
>> Paul
>
>Paul
>
>I appreciate your very detailed response, I had read all the messages on the
>Asus forum and had decided that nobody had actually found a solution. To
>give a bit more background info, this pc is used specifically for video
>editing. Normally it has PCI cards including SCSI controller and a dual
>monitor Matrox card as well as the main 'Velocity' video editing hardware
>card (which also has its own dedicated SCSI bus and external Raid drives).
>The machine has been in use for several years until the original P4C800 (not
>deluxe) died and the deluxe board was recently purchased to replace it so
>that the original components would just transfer. With the deluxe installed
>the machine works as it did previously (apart from the shutdowns) but in the
>course of investigating the problem the machine has been simplified to a
>bare minimum spec and PSU, processor, memory, drives etc have all been
>substituted with known good items.
>
>I do not have the schematic of the board but assume the 'PS_ON#' you mention
>is the 'power good' line between the board and the PSU? I had toyed with the
>idea of inserting a switch in whatever signal line that controls the PSU as
>a temporary diagnostic aid, I may be wrong but feel that the machine would
>carry on running correctly (software wise) if it the power stayed on.
>
>I am sure the fault is not heat related as it switches off within a few
>seconds from cold start, its just that something in the machines inner
>workings is incorrectly telling it to shut down.
>
>Ian
>
>
Go into the BIOS and load bios defaults and run like that a bit. Its possible
some setting in the bios is overclocked or underpowered(Memory or cpu voltage
is incorrect?? or overclocked?)