Yamaha's own MIDI definition

[The Old Crow]

On Thu, 24 Aug 2000, Magnus Danielson wrote:

> When we are on the topic... which microprocessors are there in the DX7 and
> then also in the DX7II/DX7IIFD?

  The DX7 used a Hitachi 63B03 microcontroller and a 6805S
microcontroller. The 6805S (slave) was internally programmed and was
responsible for reading the keyboard, front panel and footswitches as well
as digitizing things like the data slider, breath controller CV, etc.  
The 6303 (main) had the main OS in external ROM and operated everything
else, calling on the slave for input info as needed.  IIRC, the 6805S
operated from a 4MHz clock and the 6303 from a 4.77MHz clock.

> I currently have to hack on it to make it deal with the messy i8088,
> but most simpler MCUs should work just like a beauty.

  There should be info online for 6303s, it follows the 6800 model.

  The OS for the main CPU is easy enough to copy to a file, but the
embedded code for the 6805 is going to be tough to obtain.  Probably
simpler to see how the main CPU works with the slave, then roll your own
code to do the slave functions.  (Of course in doing this, one may as
well do the key-velocities to 127 right there). The DX7 technical analysis
book from Yamaha detailed the slave CPU's tasks pretty thoroughly, to the
point writing your own workalike is not too difficult.

  As for the DX7-II: it also uses a 6303 and 6805S, though the 6303 is the
64-pin 7 mil pin-spacing type (more I/O).  Too bad the DX7-II case and
assembly isn't as robust as that of the original DX7.

  The Godzilla of DXes, the DX1, had 5 CPUs.  Four 6809s: the master CPU,
one each for driving each of the two DX tone engines, one for processing
the polyphonic aftertouch.  One 6803 for managing the roughly 750 LEDs and
display segments.

Crow

/**/