Looked at the roms: IC62 is a character set rom. It contains (from analysis with Neil Corlett's NanaWin): 0x0000-0x07ff: Character set, 8x8, one character per 8 bytes, top to bottom of character; 000-127 is normal IBM codepage 437 characters, 128-255 is a non-codepage 437 graphical set 0x0800-0x0fff: Blank. 0x1000-0x17ff: Top half of an 8x16 (or 8x12 or 8x14, unclear) font, same character set as above, i.e. 'half codepage 437, half graphical'. 0x1800-0x1fff: Bottom half of the 8x16 (8x12 or 8x14 of that is used) font at 0x1000-0x17ff 0x2000-0x27ff: Character set, 8x8, one character per 8 bytes, top to bottom of character; All 256 characters are the standard IBM codepage 437 characters. 0x2800-0x2fff: Blank. 0x3000-0x37ff: Top half of an 8x16 (or 8x12 or 8x14, unclear) font, codepage 437 0x3800-0x3fff: Bottom half of the 8x16 (8x12 or 8x14 of that is used) font at 0x3000-0x37ff 0x4000-0x5fff: same as 0x2000-3fff 0x6000-0x7fff: same as 0x2000-3fff
Ah interesting, I didn't try to look at them graphically, my tool of choice there is a Java app called "Tile Molester."
The other two roms are supposed to be interleaved; IC39 is EVEN and IC40 is ODD. (so the cpu must be 16 bits wide like 8086 and 80286, and the rom likely lives at 0xf0000-0xfffff as on any ibm-compatible PC) Other than the first byte of each rom being weird (you'd expect them both to be 0x43 0x43 to make the copyright message work, but they're 0x50 and 0x9a, which may have been intentional to make some sort of ROM Checksum work) they look pretty normal, I didn't see any obvious corruption.
I get that they were intereleaved, and I even wrote out a new set that had the wrong character corrected with no go.
Do they read out consistently?
I've verified them a few times and it was fine. I could see activity on the data pins on the chips hitting some of the data and address lines on a scope, so there is some acitivity but it's my guess it's stuck in a loop or something. The machine has a dip switch that switches the video between external (Herculeas or CGA) or internal LCD. No matter the position of the dip switch it's always coming up external, and I traced back the dip switch and it's read by software I think. It had leaky caps -- I repleaced those and cleaned up the board. There could still be damage, but things look and test ok. When I've seen them up for sale 3/4ths of them seem to have the same always external video/nothing comes up issue. - Ethan