So presumably when you backupd the original BIOS you got this backup in a file of some sort?
(probably .BIN or perhaps some sort of ASCII/HEX dump?)

When you reflash that original backup to the chip, then read it again to a different file...
Does the original backup file and the one just read from the chip match?

If no, then your programmer is not writing the chip correctly (config?)

If yes, if you re-flash it with different but known good BIOS, and read it back to a different file...
Does the file you used to flash the known-good BIOS match the one you just read back?

If yes, then either your "new" BIOS is not at all the right one, or the problem is
something else (not the BIOS)

If no -- does it match the original BIOS file?
If yes, you programmer is not flashing the chip at all...
If no, the chip itself may be defective.

If you have the ability to hex-dump the files at look at the content,
you should be able to see various BIOS text strings.. perhaps that would
help confirm that the files are likely correct and different versions.
(keep in mind that if the backup file is not binary, appears to contain
mostly printable characters (likely '0'-'9' and 'A'-F') it's probably not
a directly dumpable binary image, and you'll have to convert it first)

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

Hi!

Thank you for your answer.
I have verified my T48 programmer with a tiny EEPROM AT93C66. It's OK, I can write some stuffs, read it several times (the same output!), erase it and so on...
So I'm quite confident with the T48.

I have quickly compared my backuped BIOS with some official releases found on The Retro Web and it is quite close. Data begin at the same address, I can read some ASCI characters in the beginning of the data block. After it's a garbage of ASCI symbols.

Here below you will find an archive with the backuped BIOS (biosbckp.bin) and another official V7.0 from The Retro Web (A6309VMS.700).
Both files seem to be close with HxD :
- From the beginning to offset E21F, it's empty (only FF....)
- From E220 to 37FFF, both data block have the same content
- After data blocks are different

I heard about "capsule files" for bios data which can't be transfered directly with a programmer. Do you think it is the case for these files ?

Cheers,
JB

Ok, your BIOS files are exactly 256k in size and appear to contain binary
date - so I think these are exact binary images of the ROMS.

I scanned the images looking for "reasonable" strings, most of them exactly
match (even the addresses) -- but there appears to be an "extra set" of some
of them in your BIOSBCKP

In fact: 38000-3BFFF is all FF's in A6309VMS and has code in BIOS_BACK
3C000-3FFFF has code in both and it matches!
-- So this suggests that you BIOSBCKP is the same as A6309VMS and is not
damaged (it has everything the the other one does)
but... The code from 38000-3BFFF in BIOS_BKCP does not appear to match!
.. although it has the same strings!
.. so perhaps your BIOSBCKP has "extra stuff", or a different version of the
.. BIOS as well as the same one as the other... or maybe it has "junk" in that part (although it seems odd
to me that a manufacturer would have used a ROM so much bigger than needed!)
Given that everything else matches the known good one, I would probably proceed with the idea
that the BIOS chip is not damaged and the failure is somewhere else.

A6309VMS.700
00024380"MSI "
000243A1"MS-6309 "
0002440D"MSI "
0002442E"MS-6309 "
000244A3"Uknown Chassis Manufacture "
000244C4"Version 1.00 "
0002463A"DIMM1 "
0002468F"EDIMM3 "
000246C5"PCI1 "
000246F4"PCI2 "
00024723"PCI3 "
00024752"PCI4 "
00024781"PCI5 "
000247B0"AGP "
000247DF"ISA1 "
00024806"SMBIOS Support from AMI "
00024827"Tested by James "
00024852"USB "
00024873"Def "
0002489A"System Management BIOS from Atlanta"
000248BE"SMBIOS from AMI "
000248F6"English "
0003FB53"JKCTVIA 82C694x BootBlock Table"
0003FB8A"MASK / Reg.DataPRHSa"
0003FC58"Mask / Reg.DataABCGH"

R:\BIOSBCKP.BIN
00024380"MSI "
000243A1"MS-6309 "
0002440D"MSI "
0002442E"MS-6309 "
000244A3"Uknown Chassis Manufacture "
000244C4"Version 1.00 "
0002463A"DIMM1 "
0002468F"EDIMM3 "
000246C5"PCI1 "
000246F4"PCI2 "
00024723"PCI3 "
00024752"PCI4 "
00024781"PCI5 "
000247B0"AGP "
000247DF"ISA1 "
00024806"SMBIOS Support from AMI "
00024827"Tested by James "
00024852"USB "
00024873"Def "
0002489A"System Management BIOS from Atlanta"
000248BE"SMBIOS from AMI "
000248F6"English "
000380B1"MSI "
000380D2"MS-6309 "
0003813E"MSI "
0003815F"MS-6309 "
000381D4"Uknown Chassis Manufacture "
000381F5"Version 1.00 "
0003829A"Intel "
000382BB"Intel Celerontm "
0003836B"DIMM1 "
000383C0"TDIMM3 "
000383F6"PCI1 "
00038425"PCI2 "
00038454"PCI3 "
00038483"PCI4 "
000384B2"PCI5 "
000384E1"AGP "
00038510"ISA1 "
00038537"SMBIOS Support from AMI "
00038558"Tested by James "
00038583"USB "
000385A4"Def "
000385CB"System Management BIOS from Atlanta"
000385EF"SMBIOS from AMI "
00038627"English "
0003FB53"JKCTVIA 82C694x BootBlock Table"
0003FB8A"MASK / Reg.DataPRHSa"
0003FC58"Mask / Reg.DataABCGH"

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

Some news...

I checked the activity on the data and address pins of the bios chip. It's dead calm, no activity.

I plugged a diagnostic POST card in a PCI slot. Nothing on the digital display but voltages as well as clock are corrects.
I noticed that when I apply some pressure on the right bottom corner of the motherboard, I get the reset LED of the diag card lighting on...