[sdiy] Polyphonic keyboard scanner

[Rainer Buchty]

On Sat, 1 Aug 2009, Jerry Gray-Eskue wrote:

> if you allow for 61 keys 6 bits gives you 64 combinations or 6 bits to 
> identify each key if you use a logic filter to "see" only one key at 
> an evaluation pass.

Which would be the interesting part, for the moment I do not see an 
(conceptionally) easier way of doing this than using a 61 bit register 
to sample the keyboard state, then AND this register with a 61-bit 
cyclic shift register where a single 1 is running around, and taking the 
AND output through a mux to get the 64->6 conversion.

Collapsing it to an 8x8 bit scheme (as opposed to the 1x61 bit) 
will require far less components, of course.

> Like I said this is almost a lost art

True, but in this very case this kind of art only increases overall 
complexity and parts cost, as using a microcontroller gives *way* more 
flexibility at greatly reduced parts cost and complexity.

Rainer