Timing jitter seems to be a result of a lot of factors. I've done what I can to prevent it, but there's no saving a slow laptop. In my design, the YM has a dedicated oscillator, so midi timing issues won't cause audio jitter per se. Because I'm not using a proprietary midi driver, I've decided to receive the data at a faster rate than spec, and wait for each data byte. That should resolve dropped bytes and buffer overflow. That said, I'm sure someone will find a way to break it.

As to Mac FTDI Midi drivers: I'm using unhacked FT232R drivers and bumping up the speed with a helper app to resolve OSX & LInux stress.

If you have them id love to know some details/Links of the ftdi midi drivers on the mac. Iv been thinking about making some USB midi stuff for a while and just never got fully up to speed.

I think this is a great project, are you planning to keep it breadboard style? Breadboard stuff really appeals to me as an easy way to 'get my hands dirty'.. 'but not too dirty'

Also, I've been wondering about how usb-to-chip solutions (e.g. Hardsid) compare in terms of bandwidth and jitter to midi-to-chip solutions (if that's what you are planning). I know sometimes the timing jitter is quite bad on pc midi interfaces-- have you thought about whether that might limit the precision of sequencing this thing?

Sweet

I thought so, but I just wanted to check.

The chip has 3 out ports with a tone channel on each and noise on all three. The old ST mod made a left tone, right tone and 2 channel tone. You can pan noise to all 3 spots as well. No soft sweeps, but plenty of room for L/R pinging or 1 channel sounds.

Awesome! Are the stereo effects external to the chip?

Hey all,
Just an update about the project:
More info on my blog here: http://www.straytechnologies.com/ym_min … irst-look/

Got the first prototype PCB design finished, and the test board built: Running pretty well. It's about a third the size of an iPhone for scale.

This will be a kit for the open source firmware for those who don't want to source their own parts. It's set up to appear as a midi port via USB on OSX, WIN or Linux via a Java based driver. I saved some $ on the USB chip, by using an external FTDI (usb to serial) cable. This leaves a pin port so that a real midi port and external power supply can be added later as a plug-in module.

More Audio coming soon.

AY-3 8910 yes...   AY-3 8912 not yet. The pinout is different as the 8912 is a 28dip vs. the 2149's & 8910's 40dip package. That said... the driver sequence looks to be the same.

If someone could send me a couple or tell me where to find a couple, I'd happily add a second pinout mapping to the same board, so both could be used by using a different socket.

ZX Spectrums never got to the states, and the Sinclair 2068's are rare enough that I'd be pressed to part with the cash to gut one.
(especially since they have NTSC composite out & I do video as my 1st love).

edit: let me re-phrase that! YOU could definitely figure out the pin conversion and use the same driver!

Awesome work! Very exciting. Love the sounds so far

Can one use an AY-3 8912 in place of the YM2149?

Sounds awesome, I think its a good compromise, I shall be on board once its available. Thanks again for doing it!

It'll work the same as the A_SID, but it uses an actual YM2149. Look up the register, plug in the value.
I'll likely move all of the registers to midi CCs, which won't preclude processing.. but will allow Renoise, trackers & midi control as well.

It's really going to need the PCB. I have about 15 lines running off of the YM chip. that sucks on a breadboard. But the rest should be pretty flexible.

(I really just want someone to make a song with it using keykit: http://nosuch.com/keykit/  !)
Alright, maybe not... but you get the idea.

This is really exciting, Will it be a chip based emulator of the YM2149 in the same way the A_Sid emulates the sid? So we can look up the registers on the Datasheet and code it in processing?

Thanks all for the support. more to come..

Right now, I just want to stabilize the hardware.
I'll be happy for a second set of eyes, as soon as I get the hardware stabilized, and start the midi driver. It's currently at the same state as the A_Sid, but way more bang-for-the-buck as there's only 16 registers and 3 channels out.

If you're game, I'll dump you a parts-list, pinout, layout, and register code, when it gets that far. I suspect from the workshop, you'll have a midi-pump from Max/MSP an hour after pulling the breadboard out.

The oscillator hung me up for a few days.. I didn't have a 2mhz osc (as apposed to a crystal) in my parts box. grab one if you don't.
Thanks!

very excited for this!

Nice!