4 ms·
Surprised to see this as a barrier, it's not hard to generate composite video in hardware/software and there's plenty of composite to VGA/HDMI converters out th
by zh3 5y ago
Surprised to see this as a barrier, it's not hard to generate composite video in hardware/software and there's plenty of composite to VGA/HDMI converters out there (e.g. search Amazon for "composite to HDMI"). I've written plenty of code for PIC micros that generates full-colour PAL/NTSC signals (text and graphics); have to think a bit about how to make it memory-mapped from an 8-bit MCU, some sort of cycle-stealing approach seems feasible (brute-force approach would be to use dual-port RAM or an FPGA, but that doesn't seem in the spirit of 8-bit retro hacking).
- whartung 5y agoYea, I don't know what the barrier is. But there is some barrier. I would think that by now, and this hobby of make 8-bit CPUs is by no means "new", there would be some "oh, Frank did that and has a circuit" so that, at least, everyone makes a stab at using Franks video subsystem. But, nothing.
- zh3 5y agoCould be just that this kind of thing lives on the boundary of electronics, software and precision timing (e.g. for PAL it's necessary to either clock a deterministic MCU at a precise multiple of the colour subcarrier, or use a pile of dedicated logic - lots of chips for TTL or cheat by using an FPGA). To me it just feels 'wrong' to use modern tech to implement a memory-mapped display on a Z80 (I still have a working Nascom 1 [0], which managed to fit a memory-mapped text display into a system with a total 2K of RAM). Happy to discuss further, email is in profile. [0] https://en.wikipedia.org/wiki/Nascom_(computer_kit) https://en.wikipedia.org/wiki/Nascom_(computer_kit)