Ahhh, those IBM s**tty BIOSes... I have similar Acer V66XA board from Aptiva 2139 and this 5555. error is due to BIOS lacks any microcode for the CPU i.e. the CPU is not 100% "compatable" with mobo. I think I found info on some dedicated to old IBM machines website, that there is some DOS utility from IBM for updating of the CPU microcodes in these BIOSes, but it was untraceable for me.

Yeah.... the annoying PS/2 mouse not found is another gripe.... I notice I've manage to botch the model number.... correct numbers;

Aptiva 217261A
Model AAD39D2
BIOS 6M2EN5R

giantclam wrote on 2023-09-10, 08:09:

Yeah.... the annoying PS/2 mouse not found is another gripe.... I notice I've manage to botch the model number.... correct numbe […]
Show full quote

Yeah.... the annoying PS/2 mouse not found is another gripe.... I notice I've manage to botch the model number.... correct numbers;

Aptiva 217261A
Model AAD39D2
BIOS 6M2EN5R

Check if there is an option in BIOS for disabling the PS/2 mouse. There was in mine V66XA BIOS.

As for the IBM BIOS these utilities for microcode update, serial number update/change after (re)flash, etc. - I couldn't find any on the internet. I only have BIOS flash utility for these IBM BIOSes - AFLASH.EXE (I think I have two different versions).

So, the only option to get rid of this ".5555" error is to install one of the CPU models supported by your motherboard's BIOS.

Nah, nothing in BIOS to do much at all really ... there's a bit of stuff on os2site, and a couple of threads out there indicating the current BIOS version also works on 2170.

The only manual I can find is this (in case anyone needs it) ... https://mega.nz/file/7ZJ0zLLa#wHEOvvrdnwC5rs0 … jBPmP4Yh7DAXwJ4

....I'll do a bit more sleuthing ... else things are headed for the hexeditor =)

giantclam wrote on 2023-09-10, 10:30:

....I'll do a bit more sleuthing ... else things are headed for the hexeditor =)

Wish you luck. I do not know what kind of BIOS is this, but definitely it's not related to standard Award or AMI. Maybe some kind of modified Phoenix, but I'm not sure and I can not confirm that at all, since I'm not able to find any info or tools for modding this IBM BIOSes.

analog_programmer wrote on 2023-09-10, 10:58:

Wish you luck. I do not know what kind of BIOS is this, but definitely it's not related to standard Award or AMI. Maybe some kind of modified Phoenix, but I'm not sure and I can not confirm that at all, since I'm not able to find any info or tools for modding this IBM BIOSes.

...luck versus the corporate tomfoolery wrt what these things were.. 2172-61A is the odd-bod number (these days you'd think '61' + 'A' somehow snaked to here in AU), but back then?...not so sure. I just went out and poked at it with a stick....

....I would say it's an IBM style BIOS, identity covert ...BIOS chip itself bears an Acer Inc sticker and coded v3.2R5-M0 ... which suggests I'm seeing original BIOS ...but other things don't jibe ...ie; this mobo has an extra 5 multiplier config setting the manual doesn't cover, reset jumper is reserved .... as is the fact this multiplier switch does nothing. CPU always identifies as what it is ... so a random guess now is ... "this is what happens when you put a P2-350 in the place the BIOS expects to see minimum P3-450" <grin>... and this for sure will be what the BIOS is whining about...

....I'll prolly just leave it for the next owner to have that 'magic moment' of it booting cleanly with a P3 installed ~ that said, the current setup (replete with matrox G200/creative Voodoo2/SB Live!) works fine with the P2-350 ... which I suppose says something about the backwards compatibility of the board...

Kudos for your comments, made me look a little harder 😎

The 2172 is one of the Acer Aptivas, which does beg the question - would it have used an IBM BIOS or whatever Acer would use (presumably Phoenix/Award/AMI)?

I had a 2137 which was a lower end, one or two year older, Acer Aptiva, and I don't recall the BIOS indicating anything other than IBM. Very different from, say, Dells that would have had very clear Phoenix language in that era. That being said, when something boots with a giant IBM logo... that might actually be a sign they're trying to cover up non-IBMness.

giantclam wrote on 2023-09-10, 12:02:

...luck versus the corporate tomfoolery wrt what these things were.. 2172-61A is the odd-bod number (these days you'd think '61' […]
Show full quote

analog_programmer wrote on 2023-09-10, 10:58:

Wish you luck. I do not know what kind of BIOS is this, but definitely it's not related to standard Award or AMI. Maybe some kind of modified Phoenix, but I'm not sure and I can not confirm that at all, since I'm not able to find any info or tools for modding this IBM BIOSes.

...luck versus the corporate tomfoolery wrt what these things were.. 2172-61A is the odd-bod number (these days you'd think '61' + 'A' somehow snaked to here in AU), but back then?...not so sure. I just went out and poked at it with a stick....

....I would say it's an IBM style BIOS, identity covert ...BIOS chip itself bears an Acer Inc sticker and coded v3.2R5-M0 ... which suggests I'm seeing original BIOS ...but other things don't jibe ...ie; this mobo has an extra 5 multiplier config setting the manual doesn't cover, reset jumper is reserved .... as is the fact this multiplier switch does nothing. CPU always identifies as what it is ... so a random guess now is ... "this is what happens when you put a P2-350 in the place the BIOS expects to see minimum P3-450" <grin>... and this for sure will be what the BIOS is whining about...

....I'll prolly just leave it for the next owner to have that 'magic moment' of it booting cleanly with a P3 installed ~ that said, the current setup (replete with matrox G200/creative Voodoo2/SB Live!) works fine with the P2-350 ... which I suppose says something about the backwards compatibility of the board...

Kudos for your comments, made me look a little harder 😎

Since this section of forum is flooded with mostly non-sense topics, I usually avoid it, but still I try to share my experience to be a little of use to someone.

One more thing I can tell you for sure (from my own experience). These Acer mobos for IBM Aptivas have same type of BIOSes even if they exist as no OEM parts under the brand Acer. I have successfylly flashed my V66XA board with Acer's board BIOS and it's same type as IBM's, but newer in terms of supported CPUs, then finally flashed it with Siemens mechine BIOS for the same V66XA board and no surprice - it's the same as IBM's and Acer's, but newest available. So, Aptiva's mobos are not bad, but available BIOSes are total s**t.

analog_programmer wrote on 2023-09-12, 05:50:

Since this section of forum is flooded with mostly non-sense topics, I usually avoid it, but still I try to share my experience to be a little of use to someone.

One more thing I can tell you for sure (from my own experience). These Acer mobos for IBM Aptivas have same type of BIOSes even if they exist as no OEM parts under the brand Acer. I have successfylly flashed my V66XA board with Acer's board BIOS and it's same type as IBM's, but newer in terms of supported CPUs, then finally flashed it with Siemens mechine BIOS for the same V66XA board and no surprice - it's the same as IBM's and Acer's, but newest available. So, Aptiva's mobos are not bad, but available BIOSes are total s**t.

It is appreciated =) Iirc this was an ex-govt. machine, and typically 'that makes sense' whenever one can't find any reference to the actual model/part numbers ~ such info wasn't released into the 'domestic' sector. I just recently had the same experience with x86 thin-clients (Dell, HP)...you can find info out there for the domestic models, but the business/govt. class constructs are completely different.... but the upside is these machines spend their life in air-conditioned comfort, which extends their working life. Totally concur wrt available BIOS images tho' ... about as useful as tits on a bull..... m000 😎

find the latest bios on the website

https://www.os2site.com/sw/hardware/ibm/aptiva/index.html

use a blank floppy to create a bios update disk
boot from the created floppy.
bios will update and microcode will update.
reboot and enjoy

ala_borbe wrote on 2023-09-12, 06:32:

find the latest bios on the website […]
Show full quote

find the latest bios on the website

https://www.os2site.com/sw/hardware/ibm/aptiva/index.html

use a blank floppy to create a bios update disk
boot from the created floppy.
bios will update and microcode will update.
reboot and enjoy

Been there, done that ...and no, nothing changes =) The BIOS wants a P3 and not the P2 currently in the slot...

Just got a reply to some emails ~ arranged to swap an AMD Phenom II x4 CPU I have spare here, for a P3-550 ... so we each get what we want for cost of postage ..."it's not what you know, but who you know", he said =)

So just to footnote this thread, there seems no way to get around that BIOS warning on these if you fit a slot1 P2 when they were (BIOS) configured for P3, and now the aforementioned P3-550 is here and in the slot (need to replace fan, horrible noises), no more warnings, nice quiet POST pause and load win95 from IDE<>CF ...probably the quickest booting/loading '95 box I've ever put together really....

....the BIOS has other oddities... PCI slot2 is different from slot1 and slot3/16bit ISA shared. I did notice this putting the voodoo2 in, it wasn't recognized in slot2 so I moved it to slot3 (don't have any need for ISA). Didn't think much of it at the time, until it came to putting the sblive! (SB0100) in, and likewise it wasn't recognized in PCI slot 2 either...move it to slot1, all good.... and you think, 'dead slot?'...

....ex govt. machine, WOL connector header onboard....BIOS I can't disable the onboard EssSOLO1 onboard stuff... thought that was going to be a bother, but seemingly not... I've got both, 2 gameports, all working properly, so what's PCI slot2 for? Has to be a NIC/modem card ~ hunt through boxes, it doesn't want to recognize a couple of NE2000 cards I found, but then found a few 3Com 3C905B-TX cards, and sure enough these get detected properly by device mangler, and cause the BIOS to throw another 5555. code qrt to unknown resources (irq/port/dma all blank)...and it all went to pot installing the 3com drivers, and '95 got it's knickers in a knot (BSOD), remove card, safemode remove device from mangler, back to square 1...

So it appears PCI slot2 is reserved for modem/nic pci device ids and the like (perhaps other but not the voodoo2/sblive/adaptec2906 scsi .. cards I tried =), and it dawned on me later reading the user manual, that onboard header is defined as 'modem WOL connector' ... no mention of NICs anywhere. This likely makes sense, iirc this being an ex-dept of education machine, at a time when lots of smaller schools in AU didn't actually have a LAN, so it was dialup & fax operations only. I think what 'makes sense', is to disable onboard USB in BIOS and free-up IRQ 10 ...(USB port is only USB1.0 anyhow)...

//...later...

...the advanced->pnp menu in BIOS seemed 'locked' (in manual mode), tried the usual reset CMOS tricks, still locked...notice sw5 on the multiplier block not really slid all the way to OFF position (password protect... it hadn't been prompting me for password)...slide it to ON, back to OFF...power it up again ... pnp BIOS menu now set to auto, groovy, now getting somewhere...

....the matrox G200 AGP card has been running hot, somewhat too cluttered for it's passive heatsink with the sblive right beside it .... pull the G200 thinking to replace thermal paste, to discover said heatsink is fused on, so no joy there, but seeing as the BIOS is back in auto mode, decide to plop the sblive in pci slot 2, and see if it auto-detects now.... and it does.... so, m'kay...slot2 not locked down, just a coincidence of BIOS switch...(??)...

....stick a low height rtl-8139D NIC in pci slot1, giving the G200 room to breath...card's detected, but drivers aren't interested....hmmm. Turns out you can't disable onboard USB, so disable one of the serial ports instead to regain IRQ 11, and when it restarts the rtl-8139 is still complaining .... and I notice the sblive sb16 function is now flagged...

...sigh...so pull the rtl-8139 out, try ne2000 card again and same issues, and finally pop in the 3C905B-TX again, and it all starts working (sort of)...at least I can ftp to my linux box for files now, but the outside world seems fractured ...and I don't care =)

All fun and games, but I'm glad to get over it