If I remember correctly, two short beeps on an AMIBIOS indicates a RAM parity error and 8 beeps would indicate a missing or defective video card. (At least to what I could find online, I couldn't find anything regarding 8 long beeps, only for 8 short beeps)

Another thing, I'm pretty certain that this PLCC socket is for a 387SX coprocessor, given the PSU didn't even fire up with the 286 in the socket as it's shorting the VCC pins of the socket.

Vynix wrote on 2022-03-16, 17:19:

If I remember correctly, two short beeps on an AMIBIOS indicates a RAM parity error and 8 beeps would indicate a missing or defective video card. (At least to what I could find online, I couldn't find anything regarding 8 long beeps, only for 8 short beeps)

This is also what I found for an AMIBIOS... that's why I tried with another video card (a Cirrus Logic CL-GD5422) and with other SIMM modules but without success.

Vynix wrote on 2022-03-16, 17:19:

Another thing, I'm pretty certain that this PLCC socket is for a 387SX coprocessor, given the PSU didn't even fire up with the 286 in the socket as it's shorting the VCC pins of the socket.

I was also thinking that this 286 didn't belong there... I have no idea why it was put there and hope it didn't break anything.

Your SBC has 286 processor in that 387SX, remove that 286 *right NOW*! This socket is *only* for 387SX math coprocessor or compatible.

Cheers,

Alright this gets a little bit confusing now as from all the tables I could find about AMI beep codes list two different entries for the two beeps, either RAM parity error or POST failure:

The 286 on the other hand, there's a huge chance that it got fried given that some of its data pins are placed where the VCC pins on a 387SX would go, suffice to say, I'm not holding any hope for it 😟

286:

387SX:

----

Another possibility is that the Dallas RTC chip battery went flat (given the age of this SBC I wouldn't be surprised if it did), causing the BIOS settings to be lost... In turn causing the BIOS to lose its marbles during POST and go back to its default settings, the only way I can see the BIOS tripping up during this step is if, by chance the default setting tells the BIOS to look for a CGA-compatible card and initialize it if found.. Although by now 386 bioses should be expecting a EGA/VGA card by default, making this a very unlikely possibility.

Thar's all I got unfortunately, I'll let the SBC experts chime in as I don't know much about pre-So7 hardware, let alone SBCs like these 😒

Thank you all for your answers... I forgot to mention that I had already changed the Dallas and it's not a big deal if the 286 fried.

Since the previous owner put a 286 chip where it shouldn't have, I wonder if he also messed with the jumpers. But without any documentation about that board, it's impossible to say.

I found an eBay auction (https://www.ebay.com/itm/134022945361) which looks a lot like this board (it's not 100% identical but very very close). The seller can't get it to boot either. I'll ask him for a clearer picture of the jumpers.

I contacted the seller politely to ask for pictures but all I got in reply was a "no"...

But it's not a big deal, I was able to make some progress anyway: since it seemed to be a parity problem, I thought there might be a contact failure in the memory sockets... I gave it a good brush and now it's starting! 🥳

Can you please show the POST string ?
This might help identify the motherboard with ultimateretro 😀

Unfortunately, I tried to turn it back on today and the problem reappeared. Maybe I got lucky last time. 😞

There really must be a problem with the parity circuit. But at least, I know it's not totally beyond repair.

I forgot I made a backup of the BIOS, inside the binary file, I can see "X0-0100-000000-00101111-060692-386sx".

Thanks

Unfortunately, the POST string has the manufacturer code zeroed 🙁
Now I see the board uses a SARC chipset, which is mostly used by cheap and obscure chinese boards, so I guess that's as far as we can get for the identification right now 😒

odd that that near exact bios string is in the newer MAME list but is a split bios:
"ROM_SYSTEM_BIOS( 2, "a3286a3886", "A3286/A3886-01 COMP V4.0")
ROMX_LOAD( "386-a3286-a3886-01-even_32k.bin", 0x10000, 0x8000, CRC(56ed3332) SHA1(9d113e57228ee596c0c24eabb193d3670fb9a309), ROM_SKIP(1) | ROM_BIOS(2))
ROMX_LOAD( "386-a3286-a3886-01-odd_32k.bin", 0x10001, 0x8000, CRC(9dbe4874) SHA1(589379055cfedd4268d8b1786491e80527f7fad5), ROM_SKIP(1) | ROM_BIOS(2))
// 3: CPU/FPU: 386SX/486SLC - Chipset: ALD 93C308-A (93C206 ???)
// BIOS-String: X0-0100-000000-00101111-060692-386SX-0 / CC-070794-P01"

Another interesting thing is that similar bios string points to this: SARC RC2016A8 chipset on PCChips M396F
of course none of it helps but found it interesting....

Listed as Aton SC386SX40 Шасси (Main Board): SB196 SC386SX40
https://remont-aud.net/dump/kompjutery_noutbu … 0/268-1-0-67780

Never seen Aton as a brand yet. An other ebay listing says this thing is MiTAC.
https://www.ebay.com/itm/134022945361

I'll dig a bit more later on.

Hmm have heard of Aeton but not Aton.

There's also AAEON which is a company that specializes in IPCs.