Maybe the copy you downloaded is corrupted. Here is the archived Supermicro page with 1.35a (DLS2081) :
http://web.archive.org/web/19991013112219/htt … T/BIOS/bios.htm (this download works)

Bit older Supermicro page that lists 1.34C (DLS8101) :
http://web.archive.org/web/19981205124303/htt … T/BIOS/bios.htm

Current Supermicro page that has 1.34C and 1.35a available for download:
https://www.supermicro.com/wdl/Bios/oldbios/440LX/

Seems to be something of a recurring theme with the P6DL family of boards

Supermicro P6DLS BIOS modification

https://www.supermicro.com/support/faqs/faq.cfm?faq=1574

HAHAHAHA. There is a reason why the computer store I worked at from 2000 to 2001 decided to refuse to sell any more Supermicro motherboards. They literally had a 90% DOA rate and they got tired of replacing them a bunch of times before the customers would actually get a fully working board.

On the flipside, Biostar motherboards, which we ordered directly from Biostar, had a less than 1% failure rate.

We sold multitudes more Biostar motherboards than all other brands combined and we had less Biostar boards come back than any other single brand of board that we sold at the time.

I take it there is no <Ctrl><A> option at boot to enter the scsi bios?

I thought LX-GX with scsi had jumpers to enable/disable scsi.

The LX did have power circuit issues. Shortly after this gen very few SM's ever failed. I built many 10 year 24/7 systems since.

Thanks very much for all the replies, I wasn't aware the dls8101.rom was interim for the P6DLS as the Supermicro download archive doesn't list the BIOS's for each respective board.

However, on checking the label on my BIOS chip it states DLSC221 which incidentally is listed in that same archive. I can't find any reference if this was infact any early BIOS for the P6DLS - mine is a rev3 board.

Comparing the ASCII DLS221 and my current BIOS code they are in fact identical - could it be that someone flashed the wrong BIOS to the board or swapped out an incorrect chip, or even it shipped like this and although it 'works' its not correct and its screwed the flash process as all the BIOS's seem to share common code for P2 and the 440LX chipset.

If I could find out what board the DLSC151 and/or DLSC221 BIOS's were for this would be helpful.

ketma wrote on 2021-06-13, 18:29:

Thanks very much for all the replies, I wasn't aware the dls8101.rom was interim for the P6DLS as the Supermicro download archiv […]
Show full quote

Thanks very much for all the replies, I wasn't aware the dls8101.rom was interim for the P6DLS as the Supermicro download archive doesn't list the BIOS's for each respective board.

However, on checking the label on my BIOS chip it states DLSC221 which incidentally is listed in that same archive. I can't find any reference if this was infact any early BIOS for the P6DLS - mine is a rev3 board.

Comparing the ASCII DLS221 and my current BIOS code they are in fact identical - could it be that someone flashed the wrong BIOS to the board or swapped out an incorrect chip, or even it shipped like this and although it 'works' its not correct and its screwed the flash process as all the BIOS's seem to share common code for P2 and the 440LX chipset.

If I could find out what board the DLSC151 and/or DLSC221 BIOS's were for this would be helpful.

The P6DLS was the only LX chipset board Supermicro made to use DLS in the model name, so it's reasonably safe to assume that any of those files in their archive are appropriate for your board.

As far as I can gather the BIOS revision numbers look something like this...

DLS9301 Rev. 1.2c
DLSC221 Rev. 1.33
DLS8101 Rev. 1.34C
DLS2081 Rev. 1.35a

I can't find specifics on the other three files - DLS2161, DLS9181 & DLSC151, so you'd need to try to install them to get more info off the POST screens.

TBH the board / BIOS revision combinations seems all over the place - yours is a rev3 board with a rev 1.33 BIOS, but another user here at VOGONS (@Robert B) found a rev2.1 board with rev 1.34C marked on the chip and the thread I linked above - Supermicro P6DLS BIOS modification (by @i386) - was a rev3 board with rev 1.2C marked on the chip. Here's a more in-depth link to his project

https://translate.google.com/translate?sl=aut … 17/kt1701.shtml

Suppose it's possible that your problem is related to either BIOS corruption or a somehow incompatible BIOS version / faulty chip (Supermicro were big on BIOS chip swaps when things went wrong).

A starting place might be to try an 'update' to DLS9301 / Rev. 1.2c and see if that helps, though in the end you may need to try them all (remember that some may lose you some processor compatibility as well)

Thanks for the reply, that information was really useful.

I was hoping the DLSC221 was an incorrect BIOS as that does show as 1.33 on POST, but like you say the BIOS versioning seems all over the place.

I'm assuming that 1.35a was the last BIOS for this board.

Ok, with no success in BIOS flashing I pulled the BIOS chip from the board and attempted to burn the BIOS directly with an EEPROM programmer.

Chip can be read OK but when attempting to write it fails with the following:

**************************************************************
APP Version : 11.00 Model : TL866II Plus
Device : AT29C020
**************************************************************

Pins Detected Passed!
ID: 0x 1F DA ......OK!
Erase ...Succeeded. Time : 0.204 S
Protection Disable......OK!
Programming FLASH ...
Programming ...Failed

Same error after several attempts. The BIOS chip must be faulty so I've ordered some new IC's and hopefully they'll be here soon.

The current BIOS chip remains as is unchanged.

Just to update this thread, I got hold of a new BIOS chip and programmed it to v1.35a and installed it in the board successfully.

All 512Mb detected and working and both CPU's identified ok.

However, the SCSI issue still plagues me. No SCSI BIOS is detected on POST and its not listed in the PCI devices config screen.

The SCSI LED also does not remain lit and the BIOS option is ghosted N/A in the BIOS. The LED flashes for a brief second on power up.

IDE drives work ok and apart from that it seems to be working perfectly. Everything works fine in DOS and Win98 SE, no errors, lockups, strange behaviour.
Need to install Win2000 for SMP testing.

I'm thinking that some element of the SCSI side of things must be dead as I can't see anything physically wrong with the board, very strange though.

Also for ref. these are the BIOS revisions (I went into each BIOS to get the revision):

P6DLx 440LX

ROM FILE DATE VERSION
DLS9181.ROM 18/09/1997 1.2
DLS9301.ROM 30/09/1997 1.2c
DLSC221.ROM 22/12/1997 1.33
DLS2161.ROM 18/02/1998 1.34a
DLS8101.ROM 10/08/1998 1.34c
DLSC151.ROM 17/12/1998 1.34c
DLS2081.ROM 12/02/1999 1.35a

Also undocumented 83MHz Setting:
JC3 - Off | JC2 - On | JC1 - On

The manual states that SCSI shares resources with the PCI slot next to the AGP slot, do you have anything plugged in there? Also did you enable option ROMs in general in the BIOS? I assume you also tried a CMOS reset?

Hi, yes - nothing in that particular slot. I also tried enabling all the options in the BIOS and full CMOS reset, with battery removed.

Just seems very odd that everything is working apart from this one thing on the board. All other connectors, slots, pins are working ok. And there must be some sort of connectivity with the SCSI LED flashing for a split second.

ketma,

Of all the bios files that you have for the board... Do any of them contain any text that would make you think it is part of the scsi bios?
Just going off the assumption here that the actual scsi bios would've been included in the main eeprom together with the main system bios though somehow got lost.
and that it isn't compressed.. which seems very likely from my own quick look at some images.
It also seems the text based modded bios seems to be missing some code at the beginning of the file. Though, no clue if it's required.

Hi, all the BIOS are universal for all the P6DLx range of boards, but all the BIOS's do contain code that reference the SCSI and have the modules present.

I'm not sure how it determines if SCSI is present but it must check for something on post, but doesn't load the SCSI bios as its not detecting the SCSI onboard.

From what I understand the LED is meant to be lit all the time but some part of that area on the board must have failed.