VDP wait examples?

Page 2/2
1 |

By Grauw

Ascended (9578)

Grauw's picture

28-07-2018, 16:15

Agree the more practical problem will be that the machine timing is off-spec, running a certain percentage faster than a normal MSX (10-20%), which some software can relies on (PSG samples, screensplits, etc.). A turbo function is nice, and the M1 wait is rather stupid, but you’ll want a standard configured Z80 mode as a basis, and I don’t think those few percent are worth that.

Although the only reason for M1 cycle is memory instruction fetch speed so it must impact some memories, probably most things will be fine indeed, given how most ROM cartridges work fine at 7 MHz (I’ve never had one that doesn’t), though there are a few RAM expansions with issues.

But really I’ve no clue what amount of ROM/RAM cartridges will have issues, removing the M1 wait has not been something I’ve seen someone do before, so I don’t think there’s any real data about it. Already the VRAM corruption is quite surprising to me, I didn’t know the BIOS was so tightly timed. Then again screen 0 is very timing sensitive on the V99x8, more so than the TMS9918, even though for all the other screen modes the V99x8 is much faster, so it’s no huge surprise either.

By msd

Paragon (1412)

msd's picture

28-07-2018, 16:47

i removed it in a 8245. Only some to fast vdp register access. With some extra logic that could be fixed

By lintweaker

Champion (349)

lintweaker's picture

28-07-2018, 20:44

With the /M1 wait circuit in place I have no more VRAM corruption issues. So I'll add it to my next PCB version. Now on to fix / investigate my order remaining issues Smile

Page 2/2
1 |