Hi Michael,
this is the CEPAC-6502 you mentioned. I call this thing SYMilar as it consist of a NMOS 6502, 6532, RAM, and ROM (but no 6522s) similar to the SYM-1.
It uses bit banging for serial I/O (like the SYM). And the monitor program is very similar to the SUPERMON.
Attachment:
cepac.jpg [ 4.8 MiB | Viewed 7524 times ]
I added the RAM and the TTL<=>RS232 converter and used a 74LS138 (with 3 diodes forming an AND gate!!) to override the memory map. Now there is 24K (3x 8K) RAM, then 8K I/O (= 32x the 6532), and 32K ROM (last burned April 1996). The RAM, decoder, and RS232-level translator are built on a perfboard.
Attachment:
perfborad.jpg [ 5.55 MiB | Viewed 7524 times ]
There I used a wire pencil to connect the RAM to the VG-connector. This sort of wire is very thin and coated with a tough isolation that only melts away when soldered. This is a quick and dirty way of building a circuit and a very nasty one if you need to change something
Attachment:
mow.jpg [ 4.61 MiB | Viewed 7524 times ]
In that time I managed to disassemble SYM-Basic V1.1 so far that I could transfer it to this board. Currently it is heavily computing
prime gaps Once I found the article to this board once published in the German monthly c't somewhere (bitsavers?) but I saved a PDF (sorry only German).
Attachment:
epac65.pdf [3.63 MiB]
Downloaded 247 times
Cheers.
edit(1):
+ There is a bug in the schematic around the 74xx74 dual-FFs. Michael post below shows the corrected version!
+ I used a 74LS13
9 (not 138 as stated above) to generate a /CS (via 3 diodes wired AND) with one half LS139 and a qualified /WE with the other half.