P5PE-VM - Problem installing XP SP1 - Crashes at first reboot

I'm trying to help a friend get XP installed on a P5PE-VM system.
The install appears to be going fine, with the HD being formatted
successfully, and the files are copied ok. But when the system
reboots to the hard drive for the first time as part of the install
process, it hangs up.

The XP flash screen comes up ok, and the little left-to-right lights
dance like they should, but nothing happens after that. No HD
activity, no CD-ROM activity. The NUMLOCK key no longer functions.
Sometimes it may get a bit farther, with a blue screen, a cursor, or
maybe an hourglass showing, before freezing. There was actually one
time when I got all the way to the screen that estimates how many
minutes are left in the install process - before crashing there.

This is his standard XP-SP1 cd-rom. The processor is a 3GHz Pentium
D, and the RAM is 2 each Kingston KVR400/512R. The HD is a 15GB
P-ATA.

Between the two of us, we've established:

1. Installs of Linux and Win98SE work just fine on this system.

2. My friend says both the processor and the RAM work fine in
another system running XP, but with a different model MB. But when
the processor and RAM from the other system are put into this one,
it still crashes.

3. By process of substitution on working systems, I've confirmed
that the power supply, the HD cable, the CD-ROM cable, the CD-ROM
drive, the keyboard, the mouse, and the display are good.

4. The same problem occurs whether both RAM sticks are used, or
either is used alone in the first slot.

5. Using his HD, HD cable, and XP disk, XP installs just fine on my
computer - an older system running an AMD Athlon64. So as far as I
can tell, all that stuff is good.

6. The voltages look fine.

7. I've tried every conceivable variation of BIOS settings, to no
avail. I've tried flashing the newest BIOS, to no avail.

8. My friend has another of these P5PE-VM motherboards, and it
behaves the same way.

So it seems to me that there's an incompatibility of some kind
involving the motherboard, the processor, the RAM, and/or SP1.
Either that, or both of his motherboards of this model are bad,
which I would think is unlikely.

Can anyone suggest what might be going on here, or what we might do
to solve this problem?

George wrote:
> I'm trying to help a friend get XP installed on a P5PE-VM system.
> The install appears to be going fine, with the HD being formatted
> successfully, and the files are copied ok. But when the system
> reboots to the hard drive for the first time as part of the install
> process, it hangs up.
>
> The XP flash screen comes up ok, and the little left-to-right lights
> dance like they should, but nothing happens after that. No HD
> activity, no CD-ROM activity. The NUMLOCK key no longer functions.
> Sometimes it may get a bit farther, with a blue screen, a cursor, or
> maybe an hourglass showing, before freezing. There was actually one
> time when I got all the way to the screen that estimates how many
> minutes are left in the install process - before crashing there.
>
> This is his standard XP-SP1 cd-rom. The processor is a 3GHz Pentium
> D, and the RAM is 2 each Kingston KVR400/512R. The HD is a 15GB
> P-ATA.
>
> Between the two of us, we've established:
>
> 1. Installs of Linux and Win98SE work just fine on this system.
>
> 2. My friend says both the processor and the RAM work fine in
> another system running XP, but with a different model MB. But when
> the processor and RAM from the other system are put into this one,
> it still crashes.
>
> 3. By process of substitution on working systems, I've confirmed
> that the power supply, the HD cable, the CD-ROM cable, the CD-ROM
> drive, the keyboard, the mouse, and the display are good.
>
> 4. The same problem occurs whether both RAM sticks are used, or
> either is used alone in the first slot.
>
> 5. Using his HD, HD cable, and XP disk, XP installs just fine on my
> computer - an older system running an AMD Athlon64. So as far as I
> can tell, all that stuff is good.
>
> 6. The voltages look fine.
>
> 7. I've tried every conceivable variation of BIOS settings, to no
> avail. I've tried flashing the newest BIOS, to no avail.
>
> 8. My friend has another of these P5PE-VM motherboards, and it
> behaves the same way.
>
> So it seems to me that there's an incompatibility of some kind
> involving the motherboard, the processor, the RAM, and/or SP1.
> Either that, or both of his motherboards of this model are bad,
> which I would think is unlikely.
>
> Can anyone suggest what might be going on here, or what we might do
> to solve this problem?
>
>

Experiment with Safe Mode ? Enable boot logging while in Safe Mode ?

http://support.microsoft.com/default...&Product=winxp

Have you tested the hard drive with the disk drive manufacturer's diagnostic ?

Paul

"George" <> wrote in message
news:qDhOi.13395$...
> I'm trying to help a friend get XP installed on a P5PE-VM system.
> The install appears to be going fine, with the HD being formatted
> successfully, and the files are copied ok. But when the system
> reboots to the hard drive for the first time as part of the install
> process, it hangs up.
>
> The XP flash screen comes up ok, and the little left-to-right lights
> dance like they should, but nothing happens after that. No HD
> activity, no CD-ROM activity. The NUMLOCK key no longer functions.
> Sometimes it may get a bit farther, with a blue screen, a cursor, or
> maybe an hourglass showing, before freezing. There was actually one
> time when I got all the way to the screen that estimates how many
> minutes are left in the install process - before crashing there.
>
> This is his standard XP-SP1 cd-rom. The processor is a 3GHz Pentium
> D, and the RAM is 2 each Kingston KVR400/512R. The HD is a 15GB
> P-ATA.
>
> Between the two of us, we've established:
>
> 1. Installs of Linux and Win98SE work just fine on this system.
>
> 2. My friend says both the processor and the RAM work fine in
> another system running XP, but with a different model MB. But when
> the processor and RAM from the other system are put into this one,
> it still crashes.
>
> 3. By process of substitution on working systems, I've confirmed
> that the power supply, the HD cable, the CD-ROM cable, the CD-ROM
> drive, the keyboard, the mouse, and the display are good.
>
> 4. The same problem occurs whether both RAM sticks are used, or
> either is used alone in the first slot.
>
> 5. Using his HD, HD cable, and XP disk, XP installs just fine on my
> computer - an older system running an AMD Athlon64. So as far as I
> can tell, all that stuff is good.
>
> 6. The voltages look fine.
>
> 7. I've tried every conceivable variation of BIOS settings, to no
> avail. I've tried flashing the newest BIOS, to no avail.
>
> 8. My friend has another of these P5PE-VM motherboards, and it
> behaves the same way.
>
> So it seems to me that there's an incompatibility of some kind
> involving the motherboard, the processor, the RAM, and/or SP1.
> Either that, or both of his motherboards of this model are bad,
> which I would think is unlikely.
>
> Can anyone suggest what might be going on here, or what we might do
> to solve this problem?
>
>
Had a lot of problems with this board a few months ago, which I initially
put down to the graphics cards I was trying to install (FX5200 & FX5600 AGP
graphics cards). It was that bad one board got RMA'd, but to start off the
replacement was no better, but eventually managed to get up running.

I no 'expert' but there appears to be a memory problem with these
motherboards (not sure whether it is an addressing problem or a problem with
the default memory settings in BIOS). On both boards with 2 sticks (in my
case 2 x 1Gb, but also tested with 2 x 512mb) of matching RAM (Asus approved
Kingston RAM) the boards would not boot with either of the above graphics
cards installed. Replaced the two sticks of RAM with 1 x 512 mb stick in
DIMM Socket 1 and it booted straightaway (obviously in single channel).

Once I had got the machine working with the one stick of RAM I started a lot
of testing (didn't want to RMA a second board) and one thing I did discover
is that in memory settings in BIOS the default setting 'Configure DRAM
timing by SPD' is [Enabled] and defaults to CL2.50 even when the installed
RAM (Asus approved Kingston RAM) is CL3.00. Changed this setting to [Manual]
and changed the DRAM CAS Latency to 3, the DRAM RAS Precharge to 3, and the
DRAM RAS to CAS Delay to 3, reinstalled the 2 x 1Gb RAM and the system
(including both graphics cards) started and apart from one unassociated
problem no problems since.

So if you have two sticks of RAM installed, try removing the one in DIMM
Socket 2 and at the same time check the speed of your RAM (if it's CL3.00
change the settings to Manual in BIOS). Might not work for you but it
certainly did for me.

Just a word of advice, a CMOS reset reproduces the problem exactly and you
will need to start from scratch (mine would not boot to POST until I had
changed the memory settings in BIOS).

Hope this helps

Geoff

In article <>,
says...

> Had a lot of problems with this board a few months ago,
> which I initially put down to the graphics cards I was
> trying to install (FX5200 & FX5600 AGP graphics cards).
> It was that bad one board got RMA'd, but to start off
> the replacement was no better, but eventually managed to
> get up running.

> I no 'expert' but there appears to be a memory problem
> with these motherboards (not sure whether it is an
> addressing problem or a problem with the default memory
> settings in BIOS). On both boards with 2 sticks (in my
> case 2 x 1Gb, but also tested with 2 x 512mb) of
> matching RAM (Asus approved Kingston RAM) the boards
> would not boot with either of the above graphics cards
> installed. Replaced the two sticks of RAM with 1 x 512
> mb stick in DIMM Socket 1 and it booted straightaway
> (obviously in single channel).

> Once I had got the machine working with the one stick of
> RAM I started a lot of testing (didn't want to RMA a
> second board) and one thing I did discover is that in
> memory settings in BIOS the default setting 'Configure
> DRAM timing by SPD' is [Enabled] and defaults to CL2.50
> even when the installed RAM (Asus approved Kingston RAM)
> is CL3.00. Changed this setting to [Manual] and changed
> the DRAM CAS Latency to 3, the DRAM RAS Precharge to 3,
> and the DRAM RAS to CAS Delay to 3, reinstalled the 2 x
> 1Gb RAM and the system (including both graphics cards)
> started and apart from one unassociated problem no
> problems since.

> So if you have two sticks of RAM installed, try removing
> the one in DIMM Socket 2 and at the same time check the
> speed of your RAM (if it's CL3.00 change the settings to
> Manual in BIOS). Might not work for you but it certainly
> did for me.

> Just a word of advice, a CMOS reset reproduces the
> problem exactly and you will need to start from scratch
> (mine would not boot to POST until I had changed the
> memory settings in BIOS).

> Hope this helps

> Geoff

Thanks very much for the response. I had thought about RAM
being the problem, but that wouldn't explain why the
computer installs and runs Linux and Win98SE with no
problem, and only screws up during the XP install. And I
tried using each stick alone, but that didn't help.

Well, there's been a further development, which appears to
have solved the problem. Maybe.

If I change the BIOS setting of "ACPI APIC Support" in the
Power section to Disabled (the default setting is Enabled),
then everything seems to work fine. I am able to install XP
in the normal way, with all the other default settings left
unchanged.

So my guess is that this mobo, and perhaps the other one he
has of this same model, are defective in some way that
involves ACPI APIC support, so disabling it in BIOS lets the
thing run properly.

But I don't know what ACPI APIC Support is, or whether it's
useful in running XP. Whatever it is, it appears Linux and
98SE don't need it. Anybody know what it is? Is it
important enough to return the motherboard?

George wrote:

>
> But I don't know what ACPI APIC Support is, or whether it's
> useful in running XP. Whatever it is, it appears Linux and
> 98SE don't need it. Anybody know what it is? Is it
> important enough to return the motherboard?
>
>

Have you tried pressing F5 right at the beginning of the XP installation
to bring up a little window that lets you choose ACPI support? I forget
exactly where in the installation sequence you have to press F5, but
it's VERY near the beginning. There's no specific prompt for it,
though. What you get is a small window with three items showing. It
wasn't obvious to me at first that I could scroll items in that window,
so look carefully. The ACPI option is the third from the top in the
list in the window, I believe. Just scroll up and up and you'll find it.

My problem wasn't that I couldn't install Windows on my old Asus P2B-F
system. It was that once installed, I couldn't shut down the system
using XP. I had to manually press the power button. When I used F5 at
installation and told XP to support ACPI, the problem went away.

--
Bill Anderson

I am the Mighty Favog

George wrote:
> In article <>,
> says...
>
> > Had a lot of problems with this board a few months ago,
> > which I initially put down to the graphics cards I was
> > trying to install (FX5200 & FX5600 AGP graphics cards).
> > It was that bad one board got RMA'd, but to start off
> > the replacement was no better, but eventually managed to
> > get up running.
>
> > I no 'expert' but there appears to be a memory problem
> > with these motherboards (not sure whether it is an
> > addressing problem or a problem with the default memory
> > settings in BIOS). On both boards with 2 sticks (in my
> > case 2 x 1Gb, but also tested with 2 x 512mb) of
> > matching RAM (Asus approved Kingston RAM) the boards
> > would not boot with either of the above graphics cards
> > installed. Replaced the two sticks of RAM with 1 x 512
> > mb stick in DIMM Socket 1 and it booted straightaway
> > (obviously in single channel).
>
> > Once I had got the machine working with the one stick of
> > RAM I started a lot of testing (didn't want to RMA a
> > second board) and one thing I did discover is that in
> > memory settings in BIOS the default setting 'Configure
> > DRAM timing by SPD' is [Enabled] and defaults to CL2.50
> > even when the installed RAM (Asus approved Kingston RAM)
> > is CL3.00. Changed this setting to [Manual] and changed
> > the DRAM CAS Latency to 3, the DRAM RAS Precharge to 3,
> > and the DRAM RAS to CAS Delay to 3, reinstalled the 2 x
> > 1Gb RAM and the system (including both graphics cards)
> > started and apart from one unassociated problem no
> > problems since.
>
> > So if you have two sticks of RAM installed, try removing
> > the one in DIMM Socket 2 and at the same time check the
> > speed of your RAM (if it's CL3.00 change the settings to
> > Manual in BIOS). Might not work for you but it certainly
> > did for me.
>
> > Just a word of advice, a CMOS reset reproduces the
> > problem exactly and you will need to start from scratch
> > (mine would not boot to POST until I had changed the
> > memory settings in BIOS).
>
> > Hope this helps
>
> > Geoff
>
> Thanks very much for the response. I had thought about RAM
> being the problem, but that wouldn't explain why the
> computer installs and runs Linux and Win98SE with no
> problem, and only screws up during the XP install. And I
> tried using each stick alone, but that didn't help.
>
> Well, there's been a further development, which appears to
> have solved the problem. Maybe.
>
> If I change the BIOS setting of "ACPI APIC Support" in the
> Power section to Disabled (the default setting is Enabled),
> then everything seems to work fine. I am able to install XP
> in the normal way, with all the other default settings left
> unchanged.
>
> So my guess is that this mobo, and perhaps the other one he
> has of this same model, are defective in some way that
> involves ACPI APIC support, so disabling it in BIOS lets the
> thing run properly.
>
> But I don't know what ACPI APIC Support is, or whether it's
> useful in running XP. Whatever it is, it appears Linux and
> 98SE don't need it. Anybody know what it is? Is it
> important enough to return the motherboard?
>
>

ACPI is what allows you to go to "Standby". That is suspend to
RAM, and allows you to put the computer to sleep, then recover
very quickly the next day. The fan goes off, and the power
consumption of the computer will be in the 10W to 20W range.

ACPI also handles power management, like making changes to the
computer settings when the computer is idle, to save power. So
the computer could run cooler with ACPI enabled and working.
If you install with ACPI disabled, then at shutdown, you're
going to see "it is safe to turn off your computer" on the
monitor screen.

Paul

George wrote:

[snip]

> But I don't know what ACPI APIC Support is, or whether it's
> useful in running XP. Whatever it is, it appears Linux and
> 98SE don't need it. Anybody know what it is? Is it
> important enough to return the motherboard?

Paul answered what ACPI is, so I'll stick with APIC. Basically, it
generates more and better handles IRQs, helping reduce or eliminate IRQ
sharing problems.

For in depth info, read here:
http://www.microsoft.com/whdc/system...f/IO-APIC.mspx

RobV wrote:
> George wrote:
>
> [snip]
>
>> But I don't know what ACPI APIC Support is, or whether it's
>> useful in running XP. Whatever it is, it appears Linux and
>> 98SE don't need it. Anybody know what it is? Is it
>> important enough to return the motherboard?
>
> Paul answered what ACPI is, so I'll stick with APIC. Basically, it
> generates more and better handles IRQs, helping reduce or eliminate IRQ
> sharing problems.
>
> For in depth info, read here:
> http://www.microsoft.com/whdc/system...f/IO-APIC.mspx
>
>

I just checked the manual, and yes, that setting is a reference
to APIC (advanced programmable interrupt controller). I don't know
why they put ACPI in front of the name for that. With APIC disabled,
there would be 16 interrupts, while with APIC enabled, there
could be 24 seen in the table of interrupts. As far as I know, there
is no harm done with either setting. Normally, I'd leave that enabled.
(My Nforce2 board might be an exception.)

ACPI and APIC are two different things. ACPI is software and a power
management spec. APIC is a hardware feature for interrupt control.
APIC allows things like interrupt steering from multiple processor
sockets.

If a person wanted to know more about the (boring) ACPI spec, then
this would be a place to get a download. 624 pages. I see there are
references to APIC in here, as there are some tables passed from
the BIOS, to the ACPI OS software.

http://www.acpi.info/spec30a.htm

There is a Wikipedia page for APIC here, even mentioning some potential bugs.

http://en.wikipedia.org/wiki/Intel_APIC_Architecture

Paul

Paul wrote:
> RobV wrote:
>> George wrote:
>>
>> [snip]
>>
>>> But I don't know what ACPI APIC Support is, or whether it's
>>> useful in running XP. Whatever it is, it appears Linux and
>>> 98SE don't need it. Anybody know what it is? Is it
>>> important enough to return the motherboard?
>>
>> Paul answered what ACPI is, so I'll stick with APIC. Basically, it
>> generates more and better handles IRQs, helping reduce or eliminate
>> IRQ sharing problems.
>>
>> For in depth info, read here:
>> http://www.microsoft.com/whdc/system...f/IO-APIC.mspx
>>
>>
>
> I just checked the manual, and yes, that setting is a reference
> to APIC (advanced programmable interrupt controller). I don't know
> why they put ACPI in front of the name for that. With APIC disabled,
> there would be 16 interrupts, while with APIC enabled, there
> could be 24 seen in the table of interrupts. As far as I know, there
> is no harm done with either setting. Normally, I'd leave that enabled.
> (My Nforce2 board might be an exception.)
>
> ACPI and APIC are two different things. ACPI is software and a power
> management spec. APIC is a hardware feature for interrupt control.
> APIC allows things like interrupt steering from multiple processor
> sockets.
>
> If a person wanted to know more about the (boring) ACPI spec, then
> this would be a place to get a download. 624 pages. I see there are
> references to APIC in here, as there are some tables passed from
> the BIOS, to the ACPI OS software.
>
> http://www.acpi.info/spec30a.htm
>
> There is a Wikipedia page for APIC here, even mentioning some
> potential bugs.
> http://en.wikipedia.org/wiki/Intel_APIC_Architecture
>
> Paul

I'm aware of the fact that ACPI and APIC are two different things and
would not normally be listed together, as you say. I just thought that
info on what APIC is would be helpful to the OP as well as the info you
provided regarding ACPI, but in a more compact form. ;-)

On my nForce2 NF-7 MB system, APIC is disabled as well (ACPI enabled to
allow power options), as I had few devices to worry about sharing IRQs,
and in some cases caused problems with audio if enabled.

On my new system (P5B-Plus, E6600), with many more devices and an Intel
chipset, APIC is enabled (along with ACPI, of course) with no dire
consequences. :-)

In article <470aed52$0$24344$>,
lid says...

> I'm aware of the fact that ACPI and APIC are two
> different things and would not normally be listed
> together, as you say. I just thought that info on what
> APIC is would be helpful to the OP as well as the info
> you provided regarding ACPI, but in a more compact form.
> ;-)

> On my nForce2 NF-7 MB system, APIC is disabled as well
> (ACPI enabled to allow power options), as I had few
> devices to worry about sharing IRQs, and in some cases
> caused problems with audio if enabled.

> On my new system (P5B-Plus, E6600), with many more
> devices and an Intel chipset, APIC is enabled (along
> with ACPI, of course) with no dire consequences. :-)

I appreciate everybody's suggestions on this problem.

It seems pretty clear to me that the failure of XP to
install (but not Linux or 98SE) with APIC support enabled
means that there is almost certainly something wrong with
the motherboard. XP sets up its hardware tables based on
the motherboard showing APIC being enabled, but then when
the install process reboots to the hard drive for the first
time, with APIC now actually being used by XP, it doesn't
work, and the system hangs. But if I install XP with APIC
turned off in the first place, XP knows that, and doesn't
try to use it, and everything works fine.

I've found a reference somewhere on the net when reading
about this to the effect that APIC is required for
multiprocessors. But it's not clear whether that includes
multi-core single-chip processsors. It's clear that at
least one core of the Pentium D works with APIC support
disabled, but I wouldn't want the other core to be unusable
as a result. Anybody know about that, or know of a program
that will show whether both cores are actually working?