[sdiy] Communications with voices in a polyphonic synth

Tom Wiltshire tom at electricdruid.net
Thu Mar 11 11:53:49 CET 2010


On 11 Mar 2010, at 02:43, MTG wrote:

> I"m pretty sure you could do a 485 protocol with no transceivers if  
> you wanted. Simply go open-collector (open drain). That's basically  
> the physical layer of I2C. Either one can be driven great distances  
> by adding transcievers/buffering/etc.  Inside the same box you  
> won't likely need any, though slew-rate limiting can save you noise  
> problems down the road.
>
> For me this type of decision comes down to what peripherals do I  
> have access to (UART vs I2C) and what speed do I really need.  I  
> don't think either are going to be screaming fast and it's pretty  
> rare that you need screaming fast (at least with the kind of stuff  
> I do which is human-machine interaction ... damn human is so slow).
>
> I think any flavour of CAN is overkill just like TCP/IP would be.  
> Ditto for I2S unless you have a spare I2S port in your micro.  What  
> micro is this again?

The voices are based on the dsPIC 33FJ128GP802. This is my favourite  
part currently because it's a 28-pin DIP and has a stereo audio DAC  
on the chip.

Thanks,
Tom




More information about the Synth-diy mailing list