Raven. A homemade Atari-like computer

[agranlund]

mxPlay and Jam plugins for MPU401 compatible midi controllers on ISA port 330:
https://github.com/agranlund/atarisound

Should most likely also work on Milan, Hades and possibly Panther/2, assuming ISA_BIOS is installed and you have some kind of soundcard with an MPU401 exposed on port 330.


It's only been tested on a PicoGUS so far and there it works with both the internal wavetable header and external devices.
Now I need to pick up one of those Gameport->Midi adapters so I can test this on the real GUS and other soundcards too, and some more wavetable daughtercards :)

This is with a tiny Dreamblaster S2 from Serdaco:

pgus_wavetable.jpg

[mikro]

Should most likely also work on Milan, Hades and possibly Panther/2, assuming ISA_BIOS is installed and you have some kind of soundcard with an MPU401 exposed on port 330.

It would be great to test this: if so, that would be the final push for me to release the more system-friendly version of mxPlay. Currently it does a lot of Falcon-specific things, it's on my TODO list for years.

(the bottom line is that currently mxPlay can't run on Milan and Hades as is)

[agranlund]

(the bottom line is that currently mxPlay can't run on Milan and Hades as is)

You mean the base mxPlay application prevents it from working, or just the plugins that come with as standard?
(I can't use any of the default bundled plugins either except for the SNDH one)

I believe artic-wrock has the OPL plugin playing on a Hades with ISA SoundBlaster16 - but I could be misremembering and perhaps he was using the Jam version.

[mikro]

Both. :) Plugins don't know about the extended XBIOS (so they e.g. ignore if the XBIOS works in little endian) and mxPlay is so stupid that it has hardcoded hardware store/restore for Falcon registers: https://github.com/mikrosk/mxplay/blob/ ... routines.S.

[polpo]

I'm the creator of PicoGUS and this is blowing my mind. It's crazy to see the card I designed in an Atari and working! Awesome that you ported pgusinit as well. I just released firmware v2.0.0 so it does change that interface a bit, BTW.

[PhilC]

Very happy to say this arrived yesterday, thank you Mr @agranlund

PXL_20240709_110134546.jpg

[Steve]

Oh I'm envious :) I hope to build one sometimes too.

[JezC]

Oh I'm envious :) I hope to build one sometimes too.

Likewise...but probably something for my retirement rather than the near future...
:sigh:

[Cyprian]

unfortunately same for me...

[agranlund]

I'm the creator of PicoGUS and this is blowing my mind. It's crazy to see the card I designed in an Atari and working! Awesome that you ported pgusinit as well. I just released firmware v2.0.0 so it does change that interface a bit, BTW.

Thanks for the heads up! I grabbed latest and merged with the Atari bits, I love the quick fw mode change on v2.0.0
I've only tested OPL and MPU401 so far and both of these seem fine.
USB Joystick and mouse support will be quite interesting to test too and I think it should be fairly easy to make a TSR for injecting these into TOS bios. The same TSR would likely work for cards with real Gameport too (or serialport, in case of mouse, but who has an old PC serial mouse nowadays :lol: )


In terms of pgusinit and main card control, it's not 100% on Atari yet but can be used.
With the new v2.0.0, pgusinit sometimes have issues receiving data correctly on the picogus control port.
It's not a big issue since your settings are saved on the card, and the emulated device seem unaffected, thus it's a low prio that I'll investigate further.

Specifically; when the card is in GUS or MPU401 mode then pgusinit behaves as expected.
But when in SB or CMS mode it'll sometimes receive wrong response data over the picogus control port;
the displayed firmware string will have some bad letters here and there, and sometimes the wrong data is important for pgusinit so it causes the program to say that the card is not detected -- It'll usually work running the program again, or again.
Since it behaves differently in the different firmware modes I'm hoping I can isolate what specifically might be affecting comms over the control port; perhaps the timings in my setup is just generally on the edge and different cpu or irq utilisation on the pico pushes it over the edge.