[sdiy] Using MIDI internally in a polysynth

[ASSI]

On Freitag 31 August 2007, Tom Wiltshire wrote:
> How easy/difficult is it to talk to multiple voices with a SPI
> connection?

There is quite a palette of solutions, you'll have to decide which one 
fits best.

Point-to-point for each voice is conceptually easy, but it scales quite 
badly and somewhat defeats the purpose of going serial.  Bonus points 
for allowing multiple voices to communicate back to the master without 
an arbitration protocol.

THen you can have an address bus and wire the SPI in parallel to each 
voice.  This allows you use the full bandwidth for a single voice and 
it's pretty easy to poll the voice from the master for any events.  
Voice initiated communication is more bothersome and requires more 
lines for implementing some sort of arbitration.

Daisy-Chained connection is expandable, but requires some sort of frame 
protocol to address the voices.  If you close the chain to a loop at 
the master, you can create a token ring system where each voice can 
replace an ampty slot with their own message which then wanders back to 
the master.  Handling of the frame protocol eats some resources at the 
individual voices, but the wiring complexity is minimal.  Given the 
speed of SPI I'd opt for this scenario, but you may have different 
requirements


Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk]>+

Wavetables for the Terratec KOMPLEXER:
http://Synth.Stromeko.net/Downloads.html#KomplexerWaves