> Mike,
>
> As others have said, there is noting wrong with your concept as long as
you
> understand how it works.  I did this once with a simple music program.  A
> timer was setup to interrupt the main loop. In the isr, the flag was
> cleared and a jump to the main code was executed.  It went on from there
to
> play the next note.  The isr set the duration of the note.  Yeah, the
stack
> continually wrapped - so what?  This prevented me from needing to test end
> conditions in the sound generating code. The code was extremely simple as
> was the intended result.  If it works for you, don't hesitate to do it.
>
> Good luck!
> Tom

This reminds me of a trick I once did on an 8008 (yep, the predecessor to
the 8080!) to build a DDS synth.

On this chip, when an interrupt occurred the processor would assert an
interrupt-acknowledge signal and then execute whatever 1-byte opcode was
found on the bus. Normally external hardware would jam in a 'RST'
instruction, which was a one-byte call to one of 8 predefined locations.

For this project, I jammed in a return instruction instead. This way, the
synth could run without having to check for any exit conditions. This made a
significant difference in the performance.

Bob Ammerman
RAm Systems
(btw: this was in 1975 or 1976)

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads