P2L97 and Celeron 766 via S370-DL

[Look Inside]

Hi.

Can anybody tell me if this should work:

Celeron 766 (SL4P6 Coppermine) on an Asus S370-DL slotket adapter into a
P2L97. Note this is the plain old P2L97 and not one of the variants.

Motherboard revision is 2.05. BIOS is 1010.004 beta. I have the P2L97 bus
speed at 66MHz. The S370-DL jumpers are set for Coppermine. I have tried
voltage jumper settings at both 1.7 and "CPU default" without any luck -
power goes on but I get nothing - no video and no beeps. Do I need to do
anything special with the P2L97 multiplier jumpers (they are still at 3.5
from my old P2-233)?

However, if I set the voltage to 2.05 (originally happened by mistake), I
get video, and the BIOS recognizes the CPU as a Celeron 766, but I get a
warning about problems with the power management settings (this pauses the
boot with an "F1 to continue" message) and when I look in the BIOS the CPU
core voltage is bouncing between 2.0 and 2.1 (that seems right) with the
error indication beside it (what does that mean?). And of course I do not
want to run a 1.7v CPU at 2.05v for any length of time.

My suspicion is that the P2L97 won't work with Coppermine voltages such as
1.7, but I don't know this for sure. I thought the S370-DL was supposed to
take care of that, but I guess I was wrong. I've searched Google Groups and
seen claims that this works, but they were always with S or DS flavours of
the P2L97.

I'll check back here, but replies or CCs to johnnybower <at> mapleleafsfan
"daught" (calm) are greatly appreciated.

 

[Robert Hancock]

Hi.

Can anybody tell me if this should work:

Celeron 766 (SL4P6 Coppermine) on an Asus S370-DL slotket adapter into a
P2L97. Note this is the plain old P2L97 and not one of the variants.

Motherboard revision is 2.05. BIOS is 1010.004 beta. I have the P2L97 bus
speed at 66MHz. The S370-DL jumpers are set for Coppermine. I have tried
voltage jumper settings at both 1.7 and "CPU default" without any luck -
power goes on but I get nothing - no video and no beeps. Do I need to do
anything special with the P2L97 multiplier jumpers (they are still at 3.5
from my old P2-233)?

However, if I set the voltage to 2.05 (originally happened by mistake), I
get video, and the BIOS recognizes the CPU as a Celeron 766, but I get a
warning about problems with the power management settings (this pauses the
boot with an "F1 to continue" message) and when I look in the BIOS the CPU
core voltage is bouncing between 2.0 and 2.1 (that seems right) with the
error indication beside it (what does that mean?). And of course I do not
want to run a 1.7v CPU at 2.05v for any length of time.

My suspicion is that the P2L97 won't work with Coppermine voltages such as
1.7, but I don't know this for sure. I thought the S370-DL was supposed to
take care of that, but I guess I was wrong. I've searched Google Groups and
seen claims that this works, but they were always with S or DS flavours of
the P2L97.

I seem to recall that 1.8V was the lowest that the VRM on these boards
would deliver, you may have to use that voltage - may make the CPU run a
bit hotter, but likely not a big deal. That slocket adapter doesn't have
a voltage regulator on it, the motherboard has to supply the correct
voltage.

 

[Look Inside]

Thanks Robert!

I tried the 1.8v setting and it works without any BIOS error about the
voltage. BIOS also confirms that the voltage is 1.8. I also tried 1.75v
and it did not work just like 1.7v. So you hit the nail on the head.

I guess the VRM range on the S and DS P2L97 flavours goes lower than it does
on the plain old P2L97 allowing 1.7v to work? Can anyone confirm that?

However, at 1.8v I see another BIOS message now that I believe is about the
microcode:

<BIOS:> BIOS update data incorrect (CPUID=00000686). Update not loaded.

How much of a problem is this? What are the consequences of ignoring it?

Again, I am using the 1010.004 beta BIOS (latest) and BIOS update setting is
enabled. I have re-flashed the BIOS to be certain. Could it make a
difference if I use the alternate version of the BIOS without LM78 support?

 

[Robert Hancock]

Thanks Robert!

I tried the 1.8v setting and it works without any BIOS error about the
voltage. BIOS also confirms that the voltage is 1.8. I also tried 1.75v
and it did not work just like 1.7v. So you hit the nail on the head.

I guess the VRM range on the S and DS P2L97 flavours goes lower than it does
on the plain old P2L97 allowing 1.7v to work? Can anyone confirm that?

However, at 1.8v I see another BIOS message now that I believe is about the
microcode:

<BIOS:> BIOS update data incorrect (CPUID=00000686). Update not loaded.

How much of a problem is this? What are the consequences of ignoring it?

You'll see this when the BIOS doesn't have a microcode update available
for that CPU model and stepping. If you already have the latest BIOS
there may not be much you can do about this. The microcode update is
basically there to compensate for any CPU bugs that may have been
discovered, but these are usually minor. As well, I believe Windows 2000
and XP will load a microcode update for the CPU if it has one during
boot (this is part of the processor driver).

 

[Look Inside]

Thanks once again.

I guess I'm out of luck. I thought that the 1010.004 BIOS would have
support for the 766 since apparently the 1010.004 BIOS for the S and DS
versions does.

I tried to install XP and after it boots from the install CD and goes
through all kinds of loading of stuff, it hangs at the Running Windows
point.

The existing Linux installation boots up fine, however, although I've not
tried to use it for anything.

 

[Robert Hancock]

Thanks once again.

I guess I'm out of luck. I thought that the 1010.004 BIOS would have
support for the 766 since apparently the 1010.004 BIOS for the S and DS
versions does.

I tried to install XP and after it boots from the install CD and goes
through all kinds of loading of stuff, it hangs at the Running Windows
point.

The existing Linux installation boots up fine, however, although I've not
tried to use it for anything.

Actually, now that I recall, there seem to be some problems with
Coppermine-core CPUs on LX chipset motherboards that cause them not to
run properly in Windows 2000 or XP. Not sure why that is, but it seems
to be the case..

 

[Paul]

Thanks once again.

I guess I'm out of luck. I thought that the 1010.004 BIOS would have
support for the 766 since apparently the 1010.004 BIOS for the S and DS
versions does.

I tried to install XP and after it boots from the install CD and goes
through all kinds of loading of stuff, it hangs at the Running Windows
point.

The existing Linux installation boots up fine, however, although I've not
tried to use it for anything.

You can use CTMC from heise magazine, to add microcode to a BIOS.
As long as the BIOS is Award and has the BIOS hook specified by
Intel.

Path to CTMC:
http://groups.google.ca/[email protected]

A recipe:
http://groups.google.ca/[email protected]

If you are curious about Microcode and BIOS updating, see page
306 and 311. This is not the latest spec, but is a bit more
informative:
ftp://download.intel.com/design/Pentium4/manuals/25366814.pdf

HTH,
Paul

 

[Look Inside]

Thanks. I'm going to look into this. At the very least it sounds like a
fun project. If I'm successful I'll post my results back to this thread.

 

[Look Inside]

Well - that seems to have worked - I no longer get the BIOS update error
(but XP still won't install). But when I saved the "new" BIOS with AFlash
and then used CTMC to look at it, I didn't see the new microcode in it -
it's contents were exactly the same as the BIOS it was based on. Is there
another way to create a BIOS with the new microcode in it? Or will I always
have to use CTMC to write it directly if I ever need to do this again?

 

[P2B]

Well - that seems to have worked - I no longer get the BIOS update error
(but XP still won't install). But when I saved the "new" BIOS with AFlash
and then used CTMC to look at it, I didn't see the new microcode in it -
it's contents were exactly the same as the BIOS it was based on. Is there
another way to create a BIOS with the new microcode in it? Or will I always
have to use CTMC to write it directly if I ever need to do this again?

CTMC just writes the microcode update into one of the NVRAM slots
provided, it does not modify the BIOS. A page on my site describes this
in more detail, along with a recipe for permanently adding the new
microcode to the BIOS. It's written for Tualatin processors on the
P2B-DS, but is generally applicable to all Award BIOSes of that era:

http://tipperlinne.com/bios6b4.htm

With regard to your XP installation issue, I've seen similar behavior
when attempting to install on a P2B-AE with an Adaptec 2940UW SCSI host
adapter. The solution in that case was to upgrade the SCSI BIOS - so you
might try removing/relocating/updating any PCI cards to see if the
symptoms change.

HTH

P2B

 

[Look Inside]

Thanks.

I was able to create a BIOS successfully.

Still no luck with XP, but I am led to believe (from internet research) that
this is an XP problem that has no solution - LX chipset with Coppermine CPU.
That means I'll have to get rid of the Celeron 766 and opt for a
pre-Coppermine CPU - back to square one. Anyone have a Mendocino core
Celeron 533 they'd like to swap for a Coppermine 766?


Thanks again to all who helped out.

Well - that seems to have worked - I no longer get the BIOS update error
(but XP still won't install). But when I saved the "new" BIOS with AFlash
and then used CTMC to look at it, I didn't see the new microcode in it -
it's contents were exactly the same as the BIOS it was based on. Is there
another way to create a BIOS with the new microcode in it? Or will I always
have to use CTMC to write it directly if I ever need to do this again?

CTMC just writes the microcode update into one of the NVRAM slots
provided, it does not modify the BIOS. A page on my site describes this
in more detail, along with a recipe for permanently adding the new
microcode to the BIOS. It's written for Tualatin processors on the
P2B-DS, but is generally applicable to all Award BIOSes of that era:

http://tipperlinne.com/bios6b4.htm

With regard to your XP installation issue, I've seen similar behavior
when attempting to install on a P2B-AE with an Adaptec 2940UW SCSI host
adapter. The solution in that case was to upgrade the SCSI BIOS - so you
might try removing/relocating/updating any PCI cards to see if the
symptoms change.

HTH

P2B