Here I'm collecting all the information I have about perfect in the hopes of one day having it work "perfectly".
The motherboard is amazingly free of useful logos or names. I do know that the BIOS is an American Megatrends AMIBIOS with ID 40-E300-001437-00111111-121593-GREEN-H. It has V1.4 marked on it in the upper left hand corner (just over the keyboard socket).
Armed with this BIOS id, I took a look at the AMI BIOS numbers page. The 1437 code is for Hsing Tech Enterprise Co., LTD. and, interestingly, the full BIOS code is listed as being ??? (MREX ? (made in China)) (Hsing Tech). Unfortunately, searching for Hsing (or Hsin) Tech did not produce anything useful. It turns out that Hsing have some kind of a relationship with PC Chips, but their web site also had nothing useful for my quest.
However, I had a look down the list of complete IDs and found that the following companies also were listed with code 1437
Taking a look at the Amptron site first I found something that looked close to my motherboard - the DX6900. However top left corner seems to say V1.3 and the card slots are the wrong colours. And, looking back to the BIOS numbers page, the DX6900 does indeed have a different BIOS in it. But we were close. Maybe even close enough for the jumper settings to work?
On second inspection is looks like it is a DX6900 since two of the chips are labelled DX6900A-1 and DX6900A-2.
It looks like after all the the card may be by Elitegroup (ECS) who also have an american site. Still no useful info yet though. For future investigation, lots of information can be found through this directory listing. I used the Q&A Index to get at the manuals.
Another clue - while two of the chips are labelled with DX9600, another one has the UMC logo on it with the ids UM9515-01, 94460 and R929H1.
Just found this AMI BIOS Survival Guide - may turn out to be useful.
And Signal 11 is exactly the signal I get when I try the machine with 48M of RAM as opposed to 32M.
Perfect has an Adaptec 1542CF.
The drivers/software page seems to indicate there is an updated BIOS for this card - wonder if that would fix our lockups. The update for the 1542C specifically mentions our CDROM
This is an HP 2247-300 SCSI disk.
Currently we're at
This is an Fujitsu M2694 ES
Looks like it's not set right at the moment, 3-4 and 5-6 are open on terminal CNH10 which seems to be undefined. I can't figure out which setting is for unterminated.
The CDROM is a Toshiba XM3401B.
One suggestion about helping this cdrom and the SCSI card get along can be found here.