[sdiy] New Pulse Module, other uses for the AS3341

Ben Bradley ben.pi.bradley at gmail.com
Thu May 20 21:08:50 CEST 2021

This goes back to an idea I did many years ago (circa 1980). I made
some cheap oscillators using 555 timers, with the control voltage
going through a resistor to the timing capacitor of the 555. This made
a crude VCO. One thing I did was have the output trigger a fixed-delay
monostable 555 circuit that had a time delay shorter than the period
of the frequencies I was generating. This gave a constant-time pulse
width output regardless of the frequency/pitch of the input. The sound
was remarkably "resonant" as the pitch varied, though it was only a
rectangular waveform output, and I never would have thought such a
waveform could sound so "resonant" if I hadn't heard it.

My thought with the AS3341 is to use its current output to the
capacitor of a one-shot like the appropriately wired 555 to do
something similar. If you have the AS3341's control voltage at a fixed
value, you get the situation described above. If you have the control
voltage tied to the same control voltage as the oscillator, the pulse
will get proportionally shorter at higher frequency, thus the output
would be the "standard" pulse witdth of modern oscillators, a fixed
percentage of the waveform period. If you have the control voltage
track at less than the oscillator's control voltage (like many synths
allow for filter tracking - the Miinimoog has two switches for 1/3 and
2/3 tracking, or both on for full tracking. I recall the Micromoog has
a pot for 0 to 100 percent tracking), then the pulse percentage will
vary with pitch, giving an in-between sound from the constant-time
pulse width and the fixed-percentage pulse width.

There's the question of what to do when the oscillator's period is
shorter than the pulser's time period. Having it retrigerrable
wouldn't be very interesting, as it would only give a very short (and
perhaps barely audible) output pulse. The 555 is (or at least I
happened to have had it wired this way, I forget) non-retriggerable,
and so it didn't regrigger untile the next pulse after it was done.
When the pitch went above this, the output would be an octave lower,
or as the input pitch went up higher, an octave-and-a-fifth or two
octaves lower than the input pitch. I'm not sure if this is "ideal"
operation, but it's at least interesting. But this is sort of
"abberant operation" and not as interesting as when the period is
lower than the pulse time.

I've had other more traditional" things I've wanted to make such as a
triangle-core oscillator, and having read through so many designs and
schematics over the years, it's clear the exponential converter is
biggest source of pitch problem in oscillators, so having everything
done (presumably) well in the AS3341 would be a big advantage in
making a good oscillator.

>mark verbos <markverbos at gmail.com>
7:15 AM (7 hours ago)
to me, synth-diy at synth-diy.org
>I saw the announcement of this chip on the Alfa site, but I couldn’t think of a use for it. What would one do with that? I guess replace the whole expo section (summing opamp, matched pair, temperature compensation, servo) of an existing VCO?


More information about the Synth-diy mailing list