Help - boot up problem

[John]

Problem: about two days ago, during boot up my pc started giving me
the error message "no 80 conductor cable installed" followed by "disk
boot failure insert system disk and press enter". I replaced the CMOS
battery because that seemed to me the most likely problem. However, no
luck, so I tried a number of things that I saw recommended - unplug all
the IDE cables and reseat them. I kept them in the same config as
before the problems started. When that didn't work, I tried connecting
using the old 40 pin IDE cable and then the 80 pin (UTMA?) cable to
channel 0 and 1. Next, I tested using a spare system disk I knew
worked and went through the same steps. During all of this, I did not
touch the SATA drive and its cable because that drive was recognized
each time during boot up. It was listed as being on Channel 3, IIRC.

For awhile I thought the problem was a dead hd (the original hd may be
dead; I've set it aside FWIW), but I had the same boot problems using a
system hd that I knew worked. I finally tried reinstalling WinXP Pro
but got the same error messages and had a different problem arise - the
pc would not shut off and restart when XP got to that point in the
reinstall process.

My pc: Biostar P4M80 mobo w/ P2.4 Ghz P4 cpu and 2Gb RAM
WD 40Gb HD (system/boot disk)
Seagate 500Gb SATA HD (data disk)
floppy
DVD
CD/R
I've had this set up trouble free for three or four years

So, any suggestions or comparable experience that resolved the problem?
And how does "no 80 conductor cable installed" lead to a "disk boot
failure" problem? I do remember one comment suggested changing a BIOS
setting related to hdd delay timing, but did not understand the
rationale.

Thanks,

John

 

[Jan Alter]

Problem: about two days ago, during boot up my pc started giving me
the error message "no 80 conductor cable installed" followed by "disk
boot failure insert system disk and press enter". I replaced the CMOS
battery because that seemed to me the most likely problem. However, no
luck, so I tried a number of things that I saw recommended - unplug all
the IDE cables and reseat them. I kept them in the same config as
before the problems started. When that didn't work, I tried connecting
using the old 40 pin IDE cable and then the 80 pin (UTMA?) cable to
channel 0 and 1. Next, I tested using a spare system disk I knew
worked and went through the same steps. During all of this, I did not
touch the SATA drive and its cable because that drive was recognized
each time during boot up. It was listed as being on Channel 3, IIRC.

For awhile I thought the problem was a dead hd (the original hd may be
dead; I've set it aside FWIW), but I had the same boot problems using a
system hd that I knew worked. I finally tried reinstalling WinXP Pro
but got the same error messages and had a different problem arise - the
pc would not shut off and restart when XP got to that point in the
reinstall process.

My pc: Biostar P4M80 mobo w/ P2.4 Ghz P4 cpu and 2Gb RAM
WD 40Gb HD (system/boot disk)
Seagate 500Gb SATA HD (data disk)
floppy
DVD
CD/R
I've had this set up trouble free for three or four years

So, any suggestions or comparable experience that resolved the problem?
And how does "no 80 conductor cable installed" lead to a "disk boot
failure" problem? I do remember one comment suggested changing a BIOS
setting related to hdd delay timing, but did not understand the
rationale.

Thanks,

John

Detach the SATA drive and try connecting the IDE drive only to see if it's
recognized.
I've heard of conflicts between SATA and IDE drives before when it comes to
booting.
Use a different IDE 80 pin cable.
In the bios menu set it for default options

 

[John]

Problem: about two days ago, during boot up my pc started giving me
the error message "no 80 conductor cable installed" followed by "disk
boot failure insert system disk and press enter".
-edited-<

My pc: Biostar P4M80 mobo w/ P2.4 Ghz P4 cpu and 2Gb RAM
WD 40Gb HD (system/boot disk)
Seagate 500Gb SATA HD (data disk)
floppy
DVD
CD/R
I've had this set up trouble free for three or four years

Replaced cable with an UDMA cable. No warning message about the "no 80
conductor" but am still getting the "disk boot failure" message after
system starts "building DMI Pool". Other stuff I've read suggests one
or more problems could cause this - BIOS, mismatch between info in BIOS
vs what system says, hd

John

 

[Paul]

Replaced cable with an UDMA cable. No warning message about the "no 80
conductor" but am still getting the "disk boot failure" message after
system starts "building DMI Pool". Other stuff I've read suggests one
or more problems could cause this - BIOS, mismatch between info in BIOS
vs what system says, hd

John

Have you checked the drive on another computer ?

You can check the partition table in the MBR with PTEDIT32. The first
picture is a screen shot, of the tool being used to display the four potential
primary partitions. The second partition in the example, has the 0x80
boot flag, so that is the partition that boots. Much the same info
can be had in Disk Management, if it'll look at the drive.

http://www.vistax64.com/attachments...n-partiton-recovery-dell-xps-420-dell-tbl.gif

ftp://ftp.symantec.com/public/english_us_canada/tools/pq/utilities/PTEDIT32.zip

You can grab the MBR with a tool like "dd". Then read the data with
a hex editor. There isn't a lot to see there, except if the 446 bytes
of code happen to have been wiped out. In the command line example
here, I'm copying a single 512 byte sector from the very beginning of
Harddisk0. And the data is being copied into the file "my_mbr.dd" in
the current working directory. Using a hex editor, I can look at the
512 bytes contained in the my_mbr.dd file.

http://www.chrysocome.net/dd

dd if=\\?\Device\Harddisk0\Partition0 of=my_mbr.dd bs=512 count=1

There is a second bit of boot code, in the boot sector of the
C: partition. That can be viewed with TestDisk. You can't really
repair the boot sector in TestDisk, but you can at least look
at it. TestDisk is a tool, that can reconstruct an MBR, based
on scanning the disk for partitions. Its ability to "repair"
the partition boot sectors, is limited to copying the
duplicate boot sectors at the end of the partition, and they're
not always loaded with anything.

http://www.cgsecurity.org/wiki/TestDisk_Step_By_Step

So it is possible, with free tools, to have a quick look around.

You can also run a read-only disk test utility. You'd check the
WD site and see if they have a test utility.

For example, I can see various versions of "Data Lifeguard" here.

http://support.wdc.com/product/download.asp?groupid=502&sid=2&lang=en

In terms of repair, the "fixmbr" and "fixboot" commands, in the
Windows Recovery Console, can be used to fix simple problems.
"Fixmbr" puts back the 446 bytes of code, without changing
the 64 bytes occupied by the primary partition table entries.
"Fixboot" reloads the partition boot sector on the C: partition.
(I used that, when I copied my WinXP C: partition with Robocopy,
which doesn't copy the partition boot sector, and then I used
"Fixboot" to put the necessary stuff to make a real C: partition
of it.) Recovery Console is an option of a WinXP installer
CD, such as a retail or OEM one you might buy at your local
computer store.

You can use CHKDSK, to check the partition while the drive
is connected to another computer, but doing so isn't guaranteed
to fix things for you. I haven't been blown away by the
merits of CHKDSK.

So perhaps a little investigation, using another computer,
will tell you whether your drive is fried or not.

Paul

 

[Jan Alter]

Have you checked the drive on another computer ?

You can check the partition table in the MBR with PTEDIT32. The first
picture is a screen shot, of the tool being used to display the four
potential
primary partitions. The second partition in the example, has the 0x80
boot flag, so that is the partition that boots. Much the same info
can be had in Disk Management, if it'll look at the drive.

http://www.vistax64.com/attachments...n-partiton-recovery-dell-xps-420-dell-tbl.gif

ftp://ftp.symantec.com/public/english_us_canada/tools/pq/utilities/PTEDIT32.zip

You can grab the MBR with a tool like "dd". Then read the data with
a hex editor. There isn't a lot to see there, except if the 446 bytes
of code happen to have been wiped out. In the command line example
here, I'm copying a single 512 byte sector from the very beginning of
Harddisk0. And the data is being copied into the file "my_mbr.dd" in
the current working directory. Using a hex editor, I can look at the
512 bytes contained in the my_mbr.dd file.

http://www.chrysocome.net/dd

dd if=\\?\Device\Harddisk0\Partition0 of=my_mbr.dd bs=512 count=1

There is a second bit of boot code, in the boot sector of the
C: partition. That can be viewed with TestDisk. You can't really
repair the boot sector in TestDisk, but you can at least look
at it. TestDisk is a tool, that can reconstruct an MBR, based
on scanning the disk for partitions. Its ability to "repair"
the partition boot sectors, is limited to copying the
duplicate boot sectors at the end of the partition, and they're
not always loaded with anything.

http://www.cgsecurity.org/wiki/TestDisk_Step_By_Step

So it is possible, with free tools, to have a quick look around.

You can also run a read-only disk test utility. You'd check the
WD site and see if they have a test utility.

For example, I can see various versions of "Data Lifeguard" here.

http://support.wdc.com/product/download.asp?groupid=502&sid=2&lang=en

In terms of repair, the "fixmbr" and "fixboot" commands, in the
Windows Recovery Console, can be used to fix simple problems.
"Fixmbr" puts back the 446 bytes of code, without changing
the 64 bytes occupied by the primary partition table entries.
"Fixboot" reloads the partition boot sector on the C: partition.
(I used that, when I copied my WinXP C: partition with Robocopy,
which doesn't copy the partition boot sector, and then I used
"Fixboot" to put the necessary stuff to make a real C: partition
of it.) Recovery Console is an option of a WinXP installer
CD, such as a retail or OEM one you might buy at your local
computer store.

You can use CHKDSK, to check the partition while the drive
is connected to another computer, but doing so isn't guaranteed
to fix things for you. I haven't been blown away by the
merits of CHKDSK.

So perhaps a little investigation, using another computer,
will tell you whether your drive is fried or not.

Paul

From WD download their Data lifeguard diagnostic tools. You can even
download the program as an ISO and make a boot startup disk so as to check
the drive on your original system. Check your original 40 Gb drive with the
program to see if it passes. Of course change the boot order so as to start
from a CD rather than a hdd.
..
If it passes then I would disconnect all peripherals (including CD, DVD,
floppy, SATA) and try booting again from the original HDD. If the problem
persists I would try resetting the bios (usually a jumper on the mb or
pulling the battery for 20 min).
Things that hit me to do if the disk passes are to take the drive to
another computer, (back up the partition and its data to an external hdd
first) and see if it will allow XP to be installed on it. If it does then I
would suspect bad RAM. Of course if you had another spare hdd you could try
installing XP on your original system with that and see if you get errors
during the install for verification that it's RAM.

 

[John]

Thanks for the suggestions. I have used them where I could but am
hampered because I only have the one floppy (installed on the pc giving
me problems) and my other pc is a laptop with CD burner.

So to recap what I have done:
1. disconnected all devices, reset CMOS (again)
2. inspected mobo for any sign of bad capacitor (puffed out dome) but
didn't see anything there
3. d/l Western Digital LifeGuard Diagnostics and made bootable CD
4. booting pc with only video card, k/b and mouse behaved normally FWIW
5. each of the hard drives passed both the quick and extended disk tests
6. connected only the boot HD. pc boot process reaches screen that says
"building DMI pool" and then I get disk failure message - insert system
disk
7. connected the cable to the optical devices - Windows Setup disk is
seen at boot up and Windows setup procedure starts. I stop there
because no hds connected yet
8. set CMOS to boot from CD, added connector cable for the hd, and
started Windows Setup. At setup screen, chose option to install new,
then selected hd (only the one), delete existing partition and install
new one and format. Windows Setup proceeds and eventually displays the
screen saying to remove any floppies and that it will reboot the pc in
15 seconds. However, it does not reboot and AFAICT it's as if all the
prior setup work is lost and am back at square one.

I still haven't figured out if my problem(s) is a CMOS/BIOS problem,
problem somewhere else on the mobo or conflict between BIOS and Windows
config (because it hangs at point "building DMI pool").

I also would like to figure out why the pc does not perform the "soft"
reboot at the end of the Windows Setup Install process. I read
somewhere that that was probably a result of BIOS settings, but the
stuff I've read is usually as clear as mud describing the settings.
These were the two clearest articles:

http://www.dewassoc.com/support/bios/awardfaq.htm

http://www.buildeasypc.com/sw/bios_setup.htm

John

 

[Paul]

Thanks for the suggestions. I have used them where I could but am
hampered because I only have the one floppy (installed on the pc giving
me problems) and my other pc is a laptop with CD burner.

So to recap what I have done:
1. disconnected all devices, reset CMOS (again)
2. inspected mobo for any sign of bad capacitor (puffed out dome) but
didn't see anything there
3. d/l Western Digital LifeGuard Diagnostics and made bootable CD
4. booting pc with only video card, k/b and mouse behaved normally FWIW
5. each of the hard drives passed both the quick and extended disk tests
6. connected only the boot HD. pc boot process reaches screen that says
"building DMI pool" and then I get disk failure message - insert system
disk
7. connected the cable to the optical devices - Windows Setup disk is
seen at boot up and Windows setup procedure starts. I stop there
because no hds connected yet
8. set CMOS to boot from CD, added connector cable for the hd, and
started Windows Setup. At setup screen, chose option to install new,
then selected hd (only the one), delete existing partition and install
new one and format. Windows Setup proceeds and eventually displays the
screen saying to remove any floppies and that it will reboot the pc in
15 seconds. However, it does not reboot and AFAICT it's as if all the
prior setup work is lost and am back at square one.

I still haven't figured out if my problem(s) is a CMOS/BIOS problem,
problem somewhere else on the mobo or conflict between BIOS and Windows
config (because it hangs at point "building DMI pool").

I also would like to figure out why the pc does not perform the "soft"
reboot at the end of the Windows Setup Install process. I read
somewhere that that was probably a result of BIOS settings, but the
stuff I've read is usually as clear as mud describing the settings.
These were the two clearest articles:

http://www.dewassoc.com/support/bios/awardfaq.htm

http://www.buildeasypc.com/sw/bios_setup.htm

John

That's pretty weird.

In terms of resources, your BIOS chip looks like this.

Main BIOS code (consists of a number of compressed read-only code modules
It's like a miniature file system.)

DMI - re-writable segment
ESCD - re-writable segment
microcode cache - re-writable segment

Boot block (initial startup, runs checksum over main BIOS code)

Generally, the Boot block should never be reloaded. If you never attempt
to erase it, it is always available as a way to bootstrap a recovery of
the computer. But a lot of flash upgrade programs, end up erasing the
boot block, which can leave the computer "bricked" if there is a failure
while programming.

Now, looking at that list of contents, say for example, that your BIOS
has been writing DMI and ESCD, every day since you got the computer.
It might do that, if some hardware in the computer is not identified properly,
and causes the information to be updated on each boot. (DMI or ESCD should
only be written, if something changes.) While the BIOS flash chip is good
for many many cycles, it is possible the flash operation isn't working, and
something is getting stuck there. The BIOS actually contains code, that
knows how to flash itself, and as long as separate segments are used,
the main BIOS code would not get touched. (The main BIOS code is protected
by a checksum, so changes to that code can be detected.)

The CMOS RAM, is a separate volatile 256 byte thing inside the Southbridge.
The contents are held in RAM, as long as the CMOS battery is there as a power
source. By volatile, I mean, the contents can be lost if all power is lost
(including the battery). The "clear CMOS" jumper is just one way of
triggering an emptying and refilling of the CMOS contents (return to
BIOS defaults).

So information necessary for operation of the computer, is stored in
a couple places. The CMOS holds BIOS settings (such as your boot order
perhaps). But the BIOS also prepares certain information that is
stored in DMI/ESCD, and that involves writing to the BIOS chip itself.

Changing out the BIOS chip is a possibility. Many boards have that
chip mounted in a PLCC socket, so another one can be installed.

If the problem is "wearout", reflashing the BIOS chip wouldn't change
anything, which is why I haven't emphasized that as a diagnostic approach.
There is no evidence at the moment, that reflashing the thing would
do anything positive.

If you reflash a BIOS, and the flashing tool erases DMI/ESCD, then
DMI/ESCD are written on the very next power cycle. So the BIOS is
still going to attempt to "update the DMI pool", even after
the BIOS has been flashed. It is one reason, if you flash a BIOS,
and then later, make an archival copy of the BIOS, the contents won't
match the original BIOS flash file.

A company like badflash.com provides chips to order, where they
sell new chips, you tell them what file you want flashed in the
chip (like some version of BIOS), and then they ship that to you.
There are other small firms, which might be closer to you,
providing that kind of service.

The BIOS chip itself (an empty one), is relatively cheap. Mouser
sells some, and they're generally less than $5. But you'd need
a mechanism to program the chip, to make it useful for your computer.
(There are tiny programming adapters, but they're not that cheap.)
While "hotflashing" is one technique for loading a file into the BIOS
chip, that requires pulling the existing chip after the computer
has booted a DOS flasher, then plugging the empty chip into the socket.
That technique can work, as the BIOS may be "shadowed" in RAM, and
that allows the chip to be pulled. I wouldn't have a problem recommending
that with DIP (dual in line packages), as they're a bit easier to
manipulate and get into place. But juggling 32 PLCC chips is not
for the squeamish - they bounce all over the place. Who is to say,
whether a power pin or a data signal makes contact first, and so on.

(A socketed 32 PLCC)

http://www.badflash.com/images/new_IO_bios.jpg

(Tool for pulling a 32 PLCC - gets underneath the diagonal corners of the chip)

http://www.radioshack.com/product/index.jsp?productId=2062619

Another way to do that for a 32 PLCC, is the usage of a BIOS
Savior from ioss.com.tw . With those, you have to figure out which
model of product to buy, and they don't have hardware to sell for the
most recent motherboards. It's possible, if your motherboard is five
years old or more, that a BIOS Savior would work. The BIOS Savior
has its own BIOS chip soldered to the assembly, and has a PLCC
socket where you insert the existing BIOS chip. That gives you a
total of two BIOS chips, plus a slide switch to select which one
is electrically "the boss". That allows hot flashing (sliding the
switch after the computer is booted), such that the empty chip is
then the target for the flasher program. So a BIOS Savior is
a clean way of doing a hot flash, in that the chip is already seated
and not bouncing around.

(A BIOS Savior dual-BIOS adapter, for socketed 32 PLCC - different models
have different BIOS chips soldered underneath. The right model, out of
a number of possible ones, must be selected, so that the chip can be flashed
by existing BIOS tools.)

http://www.euclideanspace.com/tech/pc/shuttle/mods/biosSavior/front500.jpg

http://www.euclideanspace.com/tech/pc/shuttle/mods/biosSavior/back500.jpg

*******

Have you verified your boot order in the BIOS setup screen ?

I leave mine: floppy, CDROM, hard drive

and that covers most possibilities. I don't expect that is why
you're having a problem with the Windows install, but you could
check it.

Paul

 

[Jan Alter]

Thanks for the suggestions. I have used them where I could but am
hampered because I only have the one floppy (installed on the pc giving
me problems) and my other pc is a laptop with CD burner.

So to recap what I have done:
1. disconnected all devices, reset CMOS (again)
2. inspected mobo for any sign of bad capacitor (puffed out dome) but
didn't see anything there
3. d/l Western Digital LifeGuard Diagnostics and made bootable CD
4. booting pc with only video card, k/b and mouse behaved normally FWIW
5. each of the hard drives passed both the quick and extended disk tests
6. connected only the boot HD. pc boot process reaches screen that says
"building DMI pool" and then I get disk failure message - insert system
disk
7. connected the cable to the optical devices - Windows Setup disk is
seen at boot up and Windows setup procedure starts. I stop there
because no hds connected yet
8. set CMOS to boot from CD, added connector cable for the hd, and
started Windows Setup. At setup screen, chose option to install new,
then selected hd (only the one), delete existing partition and install
new one and format. Windows Setup proceeds and eventually displays the
screen saying to remove any floppies and that it will reboot the pc in
15 seconds. However, it does not reboot and AFAICT it's as if all the
prior setup work is lost and am back at square one.

I still haven't figured out if my problem(s) is a CMOS/BIOS problem,
problem somewhere else on the mobo or conflict between BIOS and Windows
config (because it hangs at point "building DMI pool").

I also would like to figure out why the pc does not perform the "soft"
reboot at the end of the Windows Setup Install process. I read
somewhere that that was probably a result of BIOS settings, but the
stuff I've read is usually as clear as mud describing the settings.
These were the two clearest articles:

http://www.dewassoc.com/support/bios/awardfaq.htm

http://www.buildeasypc.com/sw/bios_setup.htm

John

I admire your diligence and tenacity in continuing to carry through
suggestions. As Paul suggests possibly changing the bios chip, since you
continue to get a lack of completion of the DMI information. I might
consider that, but would be very thoughtful to necessary cool and steady
hand coordination to actually changing a bios chip while the computer is on.
He has additional recommendations to changing the chip and they may be
viable. However, I'm still unsure if that's the problem. Though mentioned
that it probably wouldn't work I would go to the degree of just checking the
mb maker's website to see if there is a newer bios available than the one
you have programmed on the CMOS chip. It might be worth flashing the bios,
thinking that the current bios is partially corrupt. You did mention that
you were having trouble with your floppy drive. I would not even consider
flashing without having a reliable one. If another is available exchange
them, since a bad flash would render the cmos chip to be totally useless.
It would seem though that at this point you could rule out the hdd from
being part of the problem. I'm not sure how much RAM you have onboard, but I
believe one needs 125 mb to install XP. If you have two chips and you still
have 125 mb left after removing one of the chips then try the installation
again. I know that it's a long shot but bad RAM can cause quite a bit of
mysterious occurances to happen. If the same incomplete install happens
substitute the other chip and try the install again. If none works and
you've gone through the flashing of the bios successfully then I would go
the route of replacing the mb (that would be for myself talking now) rather
than replacing the bios chip on an older machine.

 

[Sjouke Burry]

Thanks for the suggestions. I have used them where I could but am
hampered because I only have the one floppy (installed on the pc giving
me problems) and my other pc is a laptop with CD burner.

So to recap what I have done:
1. disconnected all devices, reset CMOS (again)
2. inspected mobo for any sign of bad capacitor (puffed out dome) but
didn't see anything there
3. d/l Western Digital LifeGuard Diagnostics and made bootable CD
4. booting pc with only video card, k/b and mouse behaved normally FWIW
5. each of the hard drives passed both the quick and extended disk tests
6. connected only the boot HD. pc boot process reaches screen that says
"building DMI pool" and then I get disk failure message - insert system
disk
7. connected the cable to the optical devices - Windows Setup disk is
seen at boot up and Windows setup procedure starts. I stop there
because no hds connected yet
8. set CMOS to boot from CD, added connector cable for the hd, and
started Windows Setup. At setup screen, chose option to install new,
then selected hd (only the one), delete existing partition and install
new one and format. Windows Setup proceeds and eventually displays the
screen saying to remove any floppies and that it will reboot the pc in
15 seconds. However, it does not reboot and AFAICT it's as if all the
prior setup work is lost and am back at square one.

I still haven't figured out if my problem(s) is a CMOS/BIOS problem,
problem somewhere else on the mobo or conflict between BIOS and Windows
config (because it hangs at point "building DMI pool").

I also would like to figure out why the pc does not perform the "soft"
reboot at the end of the Windows Setup Install process. I read
somewhere that that was probably a result of BIOS settings, but the
stuff I've read is usually as clear as mud describing the settings.
These were the two clearest articles:

http://www.dewassoc.com/support/bios/awardfaq.htm

http://www.buildeasypc.com/sw/bios_setup.htm

John

Check the bios for a bootrecord protection, and if found, switch
that off.

 

[John]

That's pretty weird.

In terms of resources, your BIOS chip looks like this.

Main BIOS code (consists of a number of compressed read-only code
modules It's like a miniature file system.)

DMI - re-writable segment
ESCD - re-writable segment
microcode cache - re-writable segment

Boot block (initial startup, runs checksum over main BIOS code)

Generally, the Boot block should never be reloaded. If you never
attempt to erase it, it is always available as a way to bootstrap a
recovery of the computer. But a lot of flash upgrade programs, end up
erasing the boot block, which can leave the computer "bricked" if
there is a failure while programming.

Now, looking at that list of contents, say for example, that your BIOS
has been writing DMI and ESCD, every day since you got the computer.
It might do that, if some hardware in the computer is not identified
properly, and causes the information to be updated on each boot. (DMI
or ESCD should only be written, if something changes.) While the BIOS
flash chip is good for many many cycles, it is possible the flash
operation isn't working, and something is getting stuck there. The
BIOS actually contains code, that knows how to flash itself, and as
long as separate segments are used, the main BIOS code would not get
touched. (The main BIOS code is protected by a checksum, so changes
to that code can be detected.)

The CMOS RAM, is a separate volatile 256 byte thing inside the
Southbridge. The contents are held in RAM, as long as the CMOS
battery is there as a power source. By volatile, I mean, the contents
can be lost if all power is lost (including the battery). The "clear
CMOS" jumper is just one way of triggering an emptying and refilling
of the CMOS contents (return to BIOS defaults).

So information necessary for operation of the computer, is stored in
a couple places. The CMOS holds BIOS settings (such as your boot order
perhaps). But the BIOS also prepares certain information that is
stored in DMI/ESCD, and that involves writing to the BIOS chip itself.

Changing out the BIOS chip is a possibility. Many boards have that
chip mounted in a PLCC socket, so another one can be installed.

If the problem is "wearout", reflashing the BIOS chip wouldn't change
anything, which is why I haven't emphasized that as a diagnostic
approach. There is no evidence at the moment, that reflashing the
thing would do anything positive.

If you reflash a BIOS, and the flashing tool erases DMI/ESCD, then
DMI/ESCD are written on the very next power cycle. So the BIOS is
still going to attempt to "update the DMI pool", even after
the BIOS has been flashed. It is one reason, if you flash a BIOS,
and then later, make an archival copy of the BIOS, the contents won't
match the original BIOS flash file.

A company like badflash.com provides chips to order, where they
sell new chips, you tell them what file you want flashed in the
chip (like some version of BIOS), and then they ship that to you.
There are other small firms, which might be closer to you,
providing that kind of service.

The BIOS chip itself (an empty one), is relatively cheap. Mouser
sells some, and they're generally less than $5. But you'd need
a mechanism to program the chip, to make it useful for your computer.
(There are tiny programming adapters, but they're not that cheap.)
While "hotflashing" is one technique for loading a file into the BIOS
chip, that requires pulling the existing chip after the computer
has booted a DOS flasher, then plugging the empty chip into the
socket. That technique can work, as the BIOS may be "shadowed" in
RAM, and that allows the chip to be pulled. I wouldn't have a problem
recommending that with DIP (dual in line packages), as they're a bit
easier to manipulate and get into place. But juggling 32 PLCC chips
is not for the squeamish - they bounce all over the place. Who is to
say, whether a power pin or a data signal makes contact first, and so
on.

(A socketed 32 PLCC)

http://www.badflash.com/images/new_IO_bios.jpg

(Tool for pulling a 32 PLCC - gets underneath the diagonal corners of
the chip)

http://www.radioshack.com/product/index.jsp?productId=2062619

Another way to do that for a 32 PLCC, is the usage of a BIOS
Savior from ioss.com.tw . With those, you have to figure out which
model of product to buy, and they don't have hardware to sell for the
most recent motherboards. It's possible, if your motherboard is five
years old or more, that a BIOS Savior would work. The BIOS Savior
has its own BIOS chip soldered to the assembly, and has a PLCC
socket where you insert the existing BIOS chip. That gives you a
total of two BIOS chips, plus a slide switch to select which one
is electrically "the boss". That allows hot flashing (sliding the
switch after the computer is booted), such that the empty chip is
then the target for the flasher program. So a BIOS Savior is
a clean way of doing a hot flash, in that the chip is already seated
and not bouncing around.

(A BIOS Savior dual-BIOS adapter, for socketed 32 PLCC - different
models have different BIOS chips soldered underneath. The right
model, out of a number of possible ones, must be selected, so that
the chip can be flashed by existing BIOS tools.)

http://www.euclideanspace.com/tech/pc/shuttle/mods/biosSavior/front500.jpg
http://www.euclideanspace.com/tech/pc/shuttle/mods/biosSavior/back500.jpg

*******

Have you verified your boot order in the BIOS setup screen ?

I leave mine: floppy, CDROM, hard drive

and that covers most possibilities. I don't expect that is why
you're having a problem with the Windows install, but you could
check it.

Paul

Thanks Paul. I think I've reached the end of what I can do
(diminishing returns on effort here) because I build a new pc perhaps
every seven years or so and that really means I look for off the shelf
eqpt. that works with each other and put it together in a case. In
this instance, I'll look at possibly buying a replacement mobo (more
likely because of my budget) or build a new pc.

Even though I didn't fix my problem, it has helped to learn more about
an area that I didn't know much because I had not encountered this type
of problem before.

John

 

[John]

I admire your diligence and tenacity in continuing to carry through
suggestions. As Paul suggests possibly changing the bios chip, since
you continue to get a lack of completion of the DMI information. I
might consider that, but would be very thoughtful to necessary cool
and steady hand coordination to actually changing a bios chip while
the computer is on. He has additional recommendations to changing the
chip and they may be viable. However, I'm still unsure if that's the
problem. Though mentioned that it probably wouldn't work I would go
to the degree of just checking the mb maker's website to see if there
is a newer bios available than the one you have programmed on the
CMOS chip. It might be worth flashing the bios, thinking that the
current bios is partially corrupt. You did mention that you were
having trouble with your floppy drive. I would not even consider
flashing without having a reliable one. If another is available
exchange them, since a bad flash would render the cmos chip to be
totally useless. It would seem though that at this point you could
rule out the hdd from being part of the problem. I'm not sure how
much RAM you have onboard, but I believe one needs 125 mb to install
XP. If you have two chips and you still have 125 mb left after
removing one of the chips then try the installation again. I know
that it's a long shot but bad RAM can cause quite a bit of mysterious
occurances to happen. If the same incomplete install happens
substitute the other chip and try the install again. If none works
and you've gone through the flashing of the bios successfully then I
would go the route of replacing the mb (that would be for myself
talking now) rather than replacing the bios chip on an older machine.

Thanks Jan. I've reached the end of what I feel I can try to fix on my
own and will switch to either looking for a replacement mobo fitting my
budget that would also allow me to re-use my stuff or to (less likely)
building a new pc.

It's certainly been a learning experience because it's something that I
don't deal with on a frequent basis. I think I've only had to replace
a CMOS battery once or twice since the mid-90s. I think I've had to
replace mobo's more often and that doesn't happen that frequently to me
either

John

 

[John]

Check the bios for a bootrecord protection, and if found, switch
that off.

Thanks Sjouke. I'll check that out. Otherwise, I think I've done what
I can with the resources I have and will look for either a mobo to
replace it or maybe build a new pc. Replacing the mobo is most likely
given my budget.

John

 

[Jan Alter]

FWIW getting a board with onboard graphics these days is not a bad route
if your video needs are not demanding. I make it a point to read reviews on
newegg.com of mbs I may be interested in using for a build. Though the
feedback may not always be perfect I've gotten a good sense of which ones
have good reliability and those that don't along with specifics of what is
lacking in a particular mb.

Good luck John.