On 3/2/06, Wheaton, Simon put forth:
>So, how about turning those clock outputs into an LFO? (without the
>re-trigger/tune freq as close as you can method)
I've given this some thought in the past. I couldn't figure out a
way to do it using CMOS without using a ridiculously massive amount
of chips. However, it can be done using a microprocessor. It could
use a counter reset by leading edges at its input to measure the
number of internal (microprocessor) clock pulses between external
(sequencer) clock pulses, then divide its maximum output by half that
number, and then increment another counter by that amount until it
reaches that maximum output, then decrement by that amount until it
reaches zero again. Run through a DA converter this would give you a
triangle "oscillator core" that could be plugged into circuitry
similar to the MOTM-320 to give sine, square, and saw waveforms, all
with analogue shape control.
Alternatively, the square and saw could be generated directly by the
microprocessor as well using similar counter programs. An AD
converter could be used to add PWM. A sine output could also be
generated by the microprocessor if its programming language included
that function.
Caveat that I'm not an EE, so there might be a much better way to do
it. However, I would be surprised to learn that there is a way that
doesn't require a microprocessor.