exxos's Atari, Amiga & retro forum

You can find source code, KiCad project, gerbers, BOM and binaries under my github:

https://github.com/arkadiuszmakarenko/RISKYMSX

I have compiled first release. It has been tested on following titles
Nemesis 1, Nemesis 2, Metal Gear, Outrun, Fantasy Zone 2 and Knightmare.

Now I will spend some time testing on different hardware.

Great work. 8-)
Probably not needed as it is the most basic/common MSX, but I've a HX-10 to test on if needed.

I recently found out about and got a WonderTANG, not sure if you have seen that project? https://github.com/lfantoniosi/WonderTANG
Don't know if anything on there would help you code wise.

Higgy wrote: 18 Jan 2025 11:35 Great work. 8-)
Probably not needed as it is the most basic/common MSX, but I've a HX-10 to test on if needed.

I recently found out about and got a WonderTANG, not sure if you have seen that project? https://github.com/lfantoniosi/WonderTANG
Don't know if anything on there would help you code wise.

I will definitely need some people to test it. I will have like 3 carts to give away for this purpose.
First, I need to fix my TFMSX. I managed to destroy cart port. It had fair share of use recently.
I can't test carts which I build because of it.

@arkadiusz.makarenko if you struggle to find anyone near you to test the carts then I'll be happy to help, I presenty have a Panasonic FS-A1 & CF2700 and a Sony HB201.

I have hooked it up to my MSX HX-10 . I managed to load 32KB cart, but mapped carts don't seems to like it.

I assume time to do some optimizations.

EDIT: Nope it appears that my MSX is misbehaving. I it never enters Write cycle, so either Slot RD is shorted or there is issue with slot WR.
32KB ROMS just need SLT and CS12 to work, they never look at RD or WR.

Nothing wrong with my MSX. I just need to debug firmware more. TFMSX is VERY forgiving.

Issue was that TFMSX asserts CS12 for reads and writes. HX-10 doesn't. So I need to rewrite my bus conditions a little.
On top of it there are some weird writes to unexpected addresses (can't hook up fully LA to find out which one exactly) which caused my firmware to crash. Need to filter them.

I made some more progress today. I have added SCC emulation. SCC is too quiet in comparison to MSX sound, but it works.
This is on non-optimised version of code. I need to use it for longer to make sure it is solid and stable. If I get some stability issues then I will need to implement at least some of optimisations.

I have released v0.2 beta.
Release notes:
- Abandon CS12 check - at least some MSXs don't use this as TFMSX.
- Added additional checks on bus during write
- Added SCC emulation (you need jumper between TX and RX)

This is standard firmware, only gcc optimisations applied. If there will be issues reported I will apply low level optimisations. I am avoiding it now as it difficult to read the code.

I have standard mappers working very solid, unfortunately when SCC emulation is on ,then cart is temperamental for some reason. I assume that sometimes adding value to circular buffer overlap with next cycle and it crashes MSX.
It is very random. I will try to review assembly code and try to shave some cycles. Maybe it will be enough.
If not then I will have to look into faster chip and more expensive chip.

I trimmed few bits a little and it appears that that was enough. Still stress testing it, now. So far no crash... fingers crossed.