Boo! (Korg Poly61)

[Fraser, Colin J]

> -----Original Message-----
> From: Stewart Pye [mailto:stew at uq.net.au]
> Sent: 04 September 1999 18:53
> To: synth-diy at mailhost.bpa.nl
> Subject: Re: Boo! (Korg Poly61)
> 
> 
> Hi...
> 
> There may be a much neater solution than putting analog switches in
> parallel with the existing switches....
> 
> Scan the row output of the synths micro. Then output the code 
> corresponding
> to the keys  pressed for that row, to the column input. This 
> would have to
> be done before the synths micro reads the columns. That could 
> be achievable
> as todays micros are  much faster than the ones used in most 
> polysynths. It
> will probably require different code for different synths, 
> but it will be
> similar. 

There is a neater solution that doesn't need a much faster micro in the
interface than the scanning cpu in the keyboard.

I built my own interface for the Polysix that is very similar to the one at
http://www.mcs.com/~syzygy/webdoc/mpx61.html although it was based around a
6502.

Instead of using the analogue switches, I used the latches directly to drive
the matrix.
The outputs of each latch are connected to the row inputs of the scanning
circuit via a diode on each output so that they act like open collectors -
the outputs can only pull their row input down to Ov.
The column outputs from the synth drive the output enable lines on the
latches.

61 diodes should be cheaper than 16 4066s.

If I get the time, I'll do a PCB, and port the software for the Atmel 2051 -
this is an 8051 micro in a 20 pin dil package with a built in uart.
This would be ideal in this sort of circuit. The board should only comprise
9 x 20 pin dil chips, the diodes, and a handful of other components, so it
would be a lot smaller than a Z80 or 6502 solution.


Colin f