Login
Back to forumSee the full topicGo to last reply

Posted By

TLC
on 2011-10-07
07:08:16
 Re: 256K

Well, IMHO it's a dead end in its current form. Many possibilities, but none of them optimal.

First, some of you asked whether Csory's expansion logic could be included in an enhanced (or "final") version. Well, siz kindly sent me Csory's 256k Turbo Assembler ROM images. Inspecting the code revealed that it makes heavy use of two other (one of them previously undocumented) features of the expansion circuitry. First, TED memory fetch, when done from ROM, reads from bank 0, regardless to the active ROM bank. Second, a small slice of the expansion ROM is mapped in to address $fe00 permanently. There's no way TASM could run without those features implemented. Regardless to say, that made me less than enthusiastic about the whole matter. Above the added complexity, added incompatibility (...who said that no cartridges would ever try using their own character sets from ROM?...), there's a small, additional problem: implementing the first one needs trimming some traces on the motherboard, which is IMHO plain evil. Unless those problems could be sorted out, with, say, a configurable expansion module, I'd not play with the idea, at all. ...But a configurable expansion logic board would be different level, with its added complexity.

So, I can see a couple of options - with only the first two ones to be real in short terms IMHO.

- Re-designing the Solder memexp logic board that you've seen on the photos, as it was, and ordering a small production run. That's IMHO still within reach, only constrained by available time.
- Re-designing the above board, but with SMD components. Benefits: smaller board. Relocating some components on the motherboard would no longer be neccessary, ie. simpler installation process. Drawbacks: more tricky to design. Old, contemporary chips could no longer be used. Probably also withing reach nevertheless.
- Forget about discretes and SSI logic, and move to PLDs. Put a big-ass CPLD on a small board that replaces some of the chips (maybe the 6529B, maybe the FPLA), and "code". Benefits: small size. "Anything" could be implemented (within certain but rather far limits). Drawbacks: startup from next to 0, I'm currently not into PLDs at all. Consequently, time until it could come into "form" is slightly non-predictable. Looks a bit anachronic. Using contemporary components, or style, for that matter, is (by far) out of question.
- A fully new design (an external expansion module). By far the most complicated and doubtful if ever to be realized. Occupies the expansion port. Higher current consumption. Definitely only possible (if at all) by using PLDs. Benefit: no installation process...



Back to top


Copyright © Plus/4 World Team, 2001-2024