MSX 8235 PS2 Keyboard

By AxelF

Champion (375)

AxelF's picture

11-04-2016, 16:23

I am trying to add a PS2 port to my MSX 8235 so i can use any PS2 PC keyboard.
The keybard matrix of this msx has 9 Y-lines (0 to 8) and 8 data lines (0 to 7)
I was thinking of using a Microcontroller to convert the PS2 signals to the MSX keyboard matrix
Any ideas or tips on how to build this converter are welcome.

P.s.
An example of a PS2 port for ZX Spektrm

Thankx

Login or register to post comments

By Pencioner

Scribe (1373)

Pencioner's picture

11-04-2016, 17:25

look here
http://caro.su/msx/kbd4msx/y_yis503.htm
page in russian but it contains the schematics. Kamil Karimov (caro) is the author. There's 4 versions - for Daewoo CPC 400, YIS503II, YIS503III and YIS805

also
http://msx.gouget.com.br/2014/06/24/ps2-keyboard-to-msx-with...

By AxelF

Champion (375)

AxelF's picture

11-04-2016, 20:23

Thanks, very usefull information Smile

By brawaga

Resident (42)

brawaga's picture

17-04-2021, 03:57

@AxelF, howts goin? Any progress on 8535 PS2 MSX keyboard? Unusual choice for the task among what I saw. Did you chose 8535 for its pin count?

By bsittler

Master (177)

bsittler's picture

17-04-2021, 05:33

I assume 8235 here means the MSX model https://www.msx.org/wiki/Philips_VG-8235 rather than Atmel 8235 ATtiny20 8-bit microcontroller http://ww1.microchip.com/downloads/en/devicedoc/atmel-8235-8... though of course it might be a situation of "why not both?"

By st1mpy

Paladin (738)

st1mpy's picture

17-04-2021, 11:18

I think attiny 20 doesn't have enough pins. (Also depending on the design, you want chips that can do 8 bit ports rather than individual gpio and make sure they have enough interrupt capable pins. And 5v logic tolerant chips would be easier. When I was experimenting with external cartridge slot keyboard design, I found that There are several designs out there, but most of them in code are returning the next row that's being asked in the matrix, which means the chips are too slow (to respond and process) and wouldn't work in games that has its non bios scan routine that doesn't scan all rows in order. I think it's better suited for a cpld atleast, or a fpga.

By Danjovic

Master (184)

Danjovic's picture

17-04-2021, 15:24

st1mpy wrote:

I found that There are several designs out there, but most of them in code are returning the next row that's being asked in the matrix, which means the chips are too slow (to respond and process) and wouldn't work in games that has its non bios scan routine that doesn't scan all rows in order.

Indeed, many designs can not respond quicly enough (around 3.77us) and rely on bios sequential scan of the keyboard. They may work in Basic, but will fail on any software that calls SNSMAT to read specific rows.

Karo's design can answer in roughly 1.5us. My recent design can do in roughly the same (1.25us).

Tiny20 will certainly not fit but it is possible to reduce the amount of interrupt pins required by using few diodes. The 6 pins of ports B or C of an ATmega(4/8/16/32)8 are enough to scan up to 12 rows.
.