[sdiy] Matrix keyboard idea

[The Old Crow]

On Mon, 19 Mar 2001, Colin Fraser wrote:

> > http://pages.hotbot.com/edu/syncon/p6retro.html
> 
> Has anyone else tried this circuit ?
> It looks to have some problems to me (no disrespect to the designer, if it
> works it's a very sleek design)
> 
> The 16F84 doesn't have a UART, so it must be using software midi
> reception. The keyboard scanning pulses on the polysix are only about
> 1 ms long - shorter than a burst of 2 or more midi messages. While the
> PIC is receiving midi data, it is unlikely to be able to respond to a
> scanning pulse and output the correct data on the port. This would
> suggest that sometimes the scanning processor may get the wrong data,
> or see sudden erroneous changes in the state of keys. Or that
> sometimes the software UART may miss bytes.
> 
> The designer may have found ways round these problems or may not have
> used the keyboard in circumstances where they would show up. I would
> want to test the device under heavy midi bandwidth conditions for
> reliable keyboard output before I used it in anger.

  My modification to this idea would be to modify the 8049's program
slightly to provide a 'keyboard scanning' status output on the unused I/O
pin.  Much easier than hoping my timing was good enough. ^^

  That, and using either a 16F628 or 16F873 as the PIC part that has
a hardware UART.

Crow

/**/