Archive of the former Yahoo!Groups mailing list: MOTM

previous by date index next by date
previous in topic topic list next in topic

Subject: Re: [motm] Comparators

From: media.nai@...
Date: 2002-02-14

>Maybe some combination of:
>
>comparators
>one-shots
>AND/OR/XOR/NOR universal logic
>NOT
>
>Can anyone think of anything else that should be on this list?

Ironically, there was an old response to an earlier thread that I never
sent that I deleted while cleaning my outbox last night. I decided it
wasn't worth posting due to current backlog of designs.

In short, my idea was that simple boolean logic wasn't up to motm standards
-- what are you going to wrap $2 in CMOS chips in $100 of hardware?? A
nice example of this would be the Serge logic module -- obviously useful,
but easily built on perfboard.

Rather than consider everything above a certain threshold low, and
everything above a certain threshold high -- giving two state logic -- how
about a module that was truly analogue?? For example:

AND -- produces the sum of the inputs if all are over zero
OR -- outputs the highest input voltage
XOR -- outputs the highest input voltage if the others are zero
NOT -- outputs the absolute value of the input voltage minus five volts

Consider these as "single-sided" examples where negative voltages are
considered zero (ie. blocked with a diode), notice the results of these
analogue functions are the same as their digital counterparts for inputs of
0 and 5V, but would provide all sorts of useful analogue functions for the
voltages in between.

(Granted, the AND function would need some sort of output limiter to keep
things safe. Perhaps instead of the sum, it could output the highest
voltage if all inputs are over zero. That would still work, and may be
more useful since a voltage mixer is now part of the system.)

You could also add:

NOR -- outputs the lowest input voltage
window comparator -- a comparator with two thresholds

Unlike a boolean version, these "analogue gates" could be used to sort note
priority, invert envelopes, process LFO's, process drum triggers with
"velocity", etc., and to a limited extent process audio (eg. using an OR
gate as a "ring mod").

Perhaps the previously discussed "analogue shift register" could be altered
to perform the function of a one-shot (a digital shift register), creating
a pulse-divider that retains the amplitude of the input signal.

Anyway, I've played around with the "analogue gate" idea by "shadowing"
op-amp circuits with CMOS gates and muxes, so it feasible without using a
PIC or AD/DA converters. Regardless, modules that could perform "analogue
logic" would be much more useful than modules that simply perform "digital
logic".