After a minor mess-up with a flipped pinout on a flex pcb, yesterday i was finally able to assemble and test my first prototype. I've got some good news and bad news. Also just to preface/remind everyone: I am not an electronics engineer and I've been doing this out of pure curiosity as a personal project. I don't plan or expect to ever make this a product so I'm probably(definitely) making a tonne of mistakes with my pcb designs
. I hope my findings can be useful for others.
Good News: My approach works and I'm able to select between multiple games by simply changing a dip switch between power cycles. Currently my prototype is just 6 games, as doing the routing for more graphics chips was taking me way too long and I just wanted to test if this even worked to begin with. It was quite an exciting moment being able to switch between each game on a whim.
The design is quite simple. A 4 bit rotary dip is hooked up to the top 4 address lines of 2 larger nor flash chips, 1 for sound and 1 for prog. Each are programmed with a merged rom containing the game data, changing the dip changes which chunk of data the cv1k reads from the chips. For nor flash graphics roms, the dip connects to a 4 -> 16 de-multiplexor that feeds into the CS pin of each, enabling the chip of whichever game the the dip has selected while disabling the rest. The difficult part for me is chaining these graphics chips together in a compact manner, and then connecting the mod pcb to the cv1k. My current solution is connecting the mod pcb via ribbon cables to a revised version of the parasite pcbs that sit in the original chips place, as seen previously in this thread with my singular chip swapping system.
Bad News: yagawa games seems to suffer from graphics corruption and eventual freezes lol. Most definetaly due to poor pcb design causing crosstalk, bad attenuation or something else, messing up however those games load graphic data. In my testing so far this has only impacted pinksweets and MMP, but im yet to test original ibara. The other games I've tested (ibara BL, espgaluda 2, akai katana, and sdoj) all appear to work fine but I have to assume there is something is still going wrong, just not surfacing as badly as those yagawa games. Fortunately my take is that this is just a design issue, so it can defintely be resolved with some more work on my part.
Other points:
- A tiny bit more expensive than I had hoped. Flash chips and flex pcb fabrication are a little bit costly. I guess still relatively cheap compared to other multis, but I cant really guage a total price at the moment.
- I have to reiterate there is A LOT of work is needed to get this built, installed and functional compared to most other multi systems. Unless you pay the huge premium for assembly on some of the pcbs, have a programmer, and have the skills in very fine pitch soldering, probably not something your standard user could build from scratch.
- Reprogramming sucks unless you have a passion for desoldering and cleaning up flash chips.
