Bios Rom checksum error

3 minutes should be enough time for a flash update, and your EEPROM is
256KB, so the thing isn't that big. Still, I'd be careful with pressing
the reset button too quickly.

Did you investigate the command line options for the flash program at
all ? For example, if you can still boot to DOS, can you use the flash
program to make an archival copy of the current contents of the
flash chip ? Then, examine the file with a hex editor, and see what
is in there. These are some examples of what you might see in the existing
256KB archival file.

Address 0 - Main BIOS
Code modules delimited with "-lh5-"
On the PCChips M811LU V3.1 board, the module names I can
see are "a.BIN", "awardext.rom", "ACPITBL.BIN", "AwardBmp.bmp",
"awardeyt.rom", "_EN_CODE.BIN", "via_pxe.lom" (for network boot).
Each code module is compressed with LHA. The main BIOS body
is checksum protected. As far as I know, the boot block won't
jump into the main code, unless the checksum is good. The
first steps to using the code would be, decompression, followed
by shadow copy into system RAM.

Address 0x27240 - A blank area. I don't see DMI and ESCD default values in
the rest of the file. On an archived file from a working
computer, this area of the EEPROM could have been overwritten
as a normal part of the BIOS POST process. DMI and ESCD are
updated according to the hardware present. This area may also
contain text strings, like a serial number put there by the
system builder (using a DMI utility). Ideally, it would
make the most sense, if these areas were aligned to a
block boundary, as the flash could be segmented, and to update
DMI or ESCD, you'd want to erase some number of blocks, to
make room for the info.

Address 0x3E000 - Boot block starts here, and runs to 0x3FFFF. This is an
8K piece of code, which is what is currently being used
to boot from the floppy.

For a hex editor, you can use a program like this one. Their free edition.

http://www.softcircuits.com/cygnus/fe/

The M811LU V3.1 motherboard has a manual, even if it is a bit skimpy.
Unzip to find a PDF manual. Skip to page 31.

http://download.ecsusa.com/dlfilepcc/manual/m811/811s31.zip

On page 31, you can see "BIOS Flash Protect" and the default value
is [Disabled]. If you clear the CMOS, then the default value should be
put back there. That is the only thing I see in the manual, that could
prevent the flash update from happening. The M811LU doesn't appear
to have a jumper to prevent flash updating.

When flashing, always make sure the flash file is the right size. If the
board documentation says it is a 256KB EEPROM, then a 256KB (262144 byte)
file should be used to flash the board. Similarly, when you make your
archive copy of the current rom, it should also weigh in at 262144 bytes.
If I look at the options for the award flash program, pn means "No
flash programming" and sy means "backup original BIOS". What I'm not sure
about, is whether the syntax has to have two file names (even though we're
not flashing here), or just the filename that will be used for the
archival copy.

awdflash.exe backup.bin /pn/sy

And checking here, that does appear to be a usable syntax.

http://www.hardwaresecrets.com/printpage/41/1

When you look at the BIOS flash options, there are these:

Sb = Skip boot block
Wb = Write boot block

Writing the boot block is dangerous, in the sense that you
may not be able to boot into DOS any more, if this goes wrong.
That is how motherboards get bricked, by the boot block being
corrupted (erased but the programming fails to complete).
Skipping the boot block, means the area from 0x3E000 to 0x3FFFF
should not get updated.

Some BIOS flashers beep when the flash update is complete.
But I don't know enough about all of them, to say what to
expect in this case. There are later Award Flasher programs
with slightly more screen information displayed, and that can
make it easier to see how the thing is going. This web page
shows an example of a more elaborate GUI.

http://support.asus.com/technicaldocuments/technicaldocuments_content.aspx?no=532&SLanguage=en-us

Paul

 

I have a sneaky feeling that the wording implies that the BIOS is set to
be flashable, but the opposite is true.

That would also explain the very rapid transition to a blank screen when
the flash utility was run.

Otherwise a comprehensive explanation from Paul.

 

Before I go any further let me run something by both of you. I installed
the Syntax mother board myself back in 2005 and was able to use my old CPU
which is 1GHz. I purchased the board from Tiger Direct for $39.95 and got a
29.95 rebate so I really don't have much money tied up. The mother board has
worked fine for the five years. Since my computer trouble shooting and
proceduers knoeledge are very limited do you think I would be better off
just upgrading my computer or buy a new one? If both of you are still
willing to deal with my lack of knowledge I am still willing to learn. It
just might take a while. I do like to fix things.
William

 

If you are willing to play around a little, I don't mind ! We both
might learn something !

Can you get into the BIOS settings screen ? Usually "Del" or "F2" on
boot up.

If so change the "BIOS write setting" to "Enabled" and re-try the flash
utility. Don't make any changes ! Just read the instructions and if
there is one any "readme" file on the floppy.

Make sure that you understand them !

Next step is to make a copy of the existing BIOS. The flash utility will
give you this option. Give it a name that you can recognise. You may
need another blank floppy for this. The one that you are using should
be write protected or a copy of the original. The ideal is to use a
copy, then it doesn't matter if it gets damaged.

Since you can boot a floppy its a simple matter to make one.

 

Thanks Baron. Here is where I am now. I can't get to BIOS. When I push and
hold down the delete key or the f2 key on boot up the floppy is read but
does not go to (A) prompt. The flash utility I have only has one option and
that is to enter file name for flash. Where can I get a flash utility with
more options? Something else, computer will not turn off or reboot unless I
use switch on back or reset. If I to turn it off by the switch on back I
have to turn it on and off several times before it will boot at all and
sometimes I have to use reset. Would it be better for me to remove all
perifials and all cards except video card? I made the msdos startup disk I
use to boot on my other machine from windows. Was that OK?
William

 

Ok ! Based on what you have just said I would seriously consider biting
the bullet and junk the main board. Its unlikely that you would get a
board that supports your CPU and ram ! So I would look at a main
board, CPU, Ram combination to replace it.

The Nvidia 6100 series main boards with built in video, sound, network
card, USB, SATA/IDE and floppy are a good choice. These support an AM2
dual core CPU and 4GB+ ram. Gigabyte is a favoured manufacturer of
mine since it uses solid dielectric capacitors on the main board. Its
also 100% Linux compatible.

 

Thanks for all your help. I also appreciate the sugestion on the mother
board. Just for my satisfaction take a look at this mother board and give me
your openion.
http://www.tigerdirect.com/applications/SearchTools/item-details.asp?EdpNo=2487805&Sku=P459-1226
My CPU is AMD Athlon Duron Socket A 462-pin 1GHz. My Memory consists of 2
unbuffered 2.5v 184 pin slots and supports DDR SDRAM PC2100/PC1600.
Thanks again.
William

 

Learn from your experience. If my posted two minute procedure had
identified the error, then a flashing program may have only added more
BIOS damage.

If BIOS was truly defective, then computer never once powered on
without reporting that checksum error. If it powered on only once
with no checksum error, then flashing the BIOS may have only created
another defect. As I read it. sometimes computer powered on without
reporting a checksum error. That means checksum was not the problem -
only a symptom.

Is the motherboard defective? Maybe not. But until measurements in
that two minute procedure are made, then defect could be anywhere (and
maybe not motherboard).

Never flash a BIOS if power is unstable or uncertain. However some
BIOS may be recovered if unstable power is first identified and
corrected.

 

In terms of the economics, if you were to go to badflash.com and
order a new flash chip, to stick in the socket where the old one
is located, that would cost $30. A new motherboard, at $39, is
not much more expensive. So from that perspective, not a big deal.
But when I look at the specs, I notice the Tigerdirect FSB spec
is 266,333,400, and there is no FSB200 for the older processors.
But I did find the CPU support page for M848A V5.0, and the Duron
is down at the bottom of the page. So the new board looks OK and
the Duron should work in it. (You might want to download the
manual and read all about the motherboard, before buying it.)

http://www.pcchips.com.tw/PCCWebSite/Support/CPU_Support_Model.aspx?detailid=341&MenuID=1&LanID=2

For the flash program, the extension on the end of the file name,
might have been important. Some programs want to see .bin on the
end of the file name, and others look for .rom . But since
it doesn't seem to be booting that far any more, it'll be
a bit more difficult to recover. (If it won't touch the floppy
drive at all, that is when you know that even the boot block
is gone.) If you can still get the flash program to work, then
you could continue to play with it.

If the new M848A doesn't boot properly, you can inspect the power
supply.

There is no reason for a flash chip to fail, based on a power supply
problem. If so, then many other things could be damaged in the
computer at the same time. Like hard drives can be pretty sensitive
to a power supply problem. But I don't see a reason for *just* the
flash chip, to be the only thing affected in the computer. But flash
chips can fail on their own, when they get old enough. On motherboards
around 10 years old, this is called "bit rot", and has affected the
odd 440BX board. But it shouldn't prevent the flashing program, from
trying to reflash the chip.

As for the suggestion of removing excess hardware, yes, that would
also be an interesting experiment.

HTH,
Paul

 

Learn from your experience. If my posted two minute procedure had
identified the error, then a flashing program may have only added more
BIOS damage.

If BIOS was truly defective, then computer never once powered on
without reporting that checksum error. If it powered on only once
with no checksum error, then flashing the BIOS may have only created
another defect. As I read it. sometimes computer powered on without
reporting a checksum error. That means checksum was not the problem -
only a symptom.

Is the motherboard defective? Maybe not. But until measurements in
that two minute procedure are made, then defect could be anywhere (and
maybe not motherboard).

Never flash a BIOS if power is unstable or uncertain. However some
BIOS may be recovered if unstable power is first identified and
corrected.

w_tom
I appreciate any input to help me solve this problem. I don't mean any
disrespect to you but your post seemed to me more of a reprimand than a
help. You are right in that concerning power supply outputs the numbers are
important. I do not belive any damage has been done because everything is
the same. I am 67 years old and have bad eye site and I am not a technician
so I am reluctant to do certain things. I am not going to be in any hurry to
purchase a new mother board.
Have a good day.
William

 

Thanks Paul for all your help. I will not be in any hurry to make any
purchases. I will do some more testing and I will read the specs on the new
mother board before I make a dicision. I did down load another flash utility
and that did not work either. Have a good evening.
William

 

Hi William,
That board looks like it should be Ok. Its virtually identical to the
one you already have.

When it comes to swapping over the CPU make sure that you use a good
quality heatsink compound. I use "Artic Silver 3" when I do
CPU/heatsink cleanups.

 

I would hope that William checks out the PSU before he goes ahead with a
main board replacement. Particularly since you can get a bare bones
machine for $100 and I've seen notebook machines for $200

I agree !

It would, but only in terms of minimising PSU loading !

 

Thanks Baron. I just think I have'nt lost much if it don't work. Also I
don't think I can remove the BIOS chip anyway. Where do I "Artic Silver 3"?
There is some kind of white paste on the bottom of the CPU. Where can I get
that?
William

 

Except I was not discussing damage. Discussed are failures that do
not cause hardware damage - implying the motherboard can be fixed once
the real problem is identified and eliminated. When voltages are
marginal, chips fail (nobody said anything about damage). However if
voltage is marginal and BIOS is flashed, then strange data
(temporarily destructive) can be saved - more checksum errors. That
is only two types of failures. One that occurs only this time.
Another that causes bad data. And a third (which I never suggested)
that is hardware damage - replace the motherboard.

If a power supply is not verified, the many symptoms in the OP's
posts are also associated with marginal voltages - also known as a
defect elsewhere (ie power supply system).

So we replace that motherboard. then have a same (marginal) power
supply defect. Then what? Another example of replacing parts without
first learning the problem; also called shotgunning.

Again, failures here are also consistent with a marginal power
system. Nothing says a power supply system is bad because numbers are
not provided. Those failures say zero about hardware damage or a
damaged motherboard. Those checksum errors are also consistent with a
power system defect. If that problem exists, then flashing can even
contaminate the BIOS or simply fail. To eliminate that failure reason
takes less than 2 minutes. Discussed were failures - not hardware
damage. Curiously, William has repeatedly posted symptons consistent
with a defective power 'system'.

 

William Inscribed thus:

Good Morning William,
That white paste is the heatsink compound that I referred to ! The
white paste stuff is just not good enough for modern CPU's. Hence the
recommendation for artic silver.

You would need to ask locally to you. The last tube, 3 grams, I bought
cost me 75p UK, a little over a Dollar. It also lasts a long time, 20
or 30 applications in a tube. The advantage is that it doesn't dry out
with the heat like the white paste does.

Re BIOS chip. I think on that board it is a surface mount device that
you will not be able to remove without specialised tools. On better
quality main boards it is usually in a socket.

As far as w-toms advice is concerned, it is only valid under some
conditions ! Also w_tom forgets that we are not all engineers
experienced in measurement diagnostics. His assumptions and belief are
based on "All computer failures are PSU based" ! Now I am sure that
few will disagree that PSU problems are a major cause of computer
failures, but not all computer failures.

Indeed in my years as an engineer a high proportion of failures are due
to damage from failing to take proper anti static precautions when
handling these very sensitive components. I see many machines that
have failed or give unpredictable behaviour simply because sufficient
care has not been taken during component handling and machine
construction.

 

Static created failures and (more common) manufacturing defects are
major reasons for semiconductor failure. But there is no reason to
believe semiconductor failure exists Measuring voltages is so
*complex* that multimeters are sold to Kmart shoppers. If measuring
voltages are so difficult, then a person must never own a cell phone
or Ipod - which are more and too complex.

Power is not the most common source of damage. But no reason to
believe damage exists. Fixing with unstable power means anything can
appear defective. 30 seconds to measure voltage reports so much.
Bios will not flash? Again, symptoms of a power problem. Checksum
error occurs intermittently? Another symptom of defective power.

Well it might be a bad chip, or might be too much heat, or might
be ... that is called wild speculation. Is defect in the power
system (which is more than a power supply)? 30 seconds to have a
definitive answer - no speculation – no shotgunning.

Intermittent checksum error does not come from heat or from
defective BIOS code. That intermittent is symptom more common to other
problems. A most common reason - bad power system - can be identified
or exonerated in only 30 seconds. Nothing identifies a suspect faster
or more definitively. One who cannot switch a meter to VDC, touch
probes, and then read a number has no business trying to fix a
computer OR using a cell phone. Measuring voltages is that simple.
Nothing posted says the motherboard is defective. Nothing posted says
CPU heat has significance.

30 seconds to identify or exonerate a common reason for all those
symptoms. So simple that only fear could impede that act. Not
relevant to checksum errors is CPU heat. Fixing a checksum error with
thermal compound is only wild speculation. Known facts do not suggest
CPU heat.

 

Hey Baron or Paul
I installed new mother board and after a few tries I had to set the
bios on the fail safe setting. Message came up "Windows could not start
because the following file is missing or corrupt"
"\windows\system32\config\system"
"You can attempt to repair the file using the orignal setup CD-Rom". Is this
because of the new mother board?What would you suggest I do?
William

 

Unless you have done a clean install of Wins you will get these
problems. Wins doesn't like it when you swap mainboards on it !
I always do a clean install on a new mainboard.

 

Thanks Baron. Just two more questions. Should I run the setup disk that
came with the mother board? If I do a clean install do I have to register
windows xp again and how about the password and will I loose all my other
programs? Thanks.
William