Cross Product Sub Octave Module

[Romeo Fahl]

I didn't catch the original message, so I don't know what circuit is being
discussed, but it reminded me of one of my "still-born" modules that is in a
box in the basement:

A few years ago I built a slightly modernized version of Lester Ludwig's
Sub-Octave Cross-Product Module. It takes two VCO inputs both of which go
through their own comparator, then are fed  through a pulse divider, the two
divider outputs through a set of AND gates.  The comparators' outputs, the
dividers' outputs and the AND gates' outputs are fed into individual pots
for mixing, then summed into an output opamp.

The original version used SN74161s for the dividers and I believe a 7408 for
the AND gates, my version used a CD 4520 for the dividing and a CD4081 for
the ANDing.

I think something got lost in the update, because the module worked for a
while, then quit and I was unable to get it working again.  I posted the
schematic here:

http://www.spiritone.com/~8brain/xproductflat.gif

If anyone can see any obvious flaws, please let me know.  It could be
something simple in the connection between the opamps and the CMOS that I
left out... just wasn't sure what would work best.  Or it could be my fuzzy
slippers on the shag carpet were frying the CMOS (-*,

When it did work, it sounded like a ring modulator from hell... or a casio
having a nuclear meltdown, depending on how the pots were set.

There's a good amount of room for modifications, like switching the divider
outputs to different AND gates.  The 4520 also has reset inputs that can be
pinged for extra nuttiness.




----- Original Message -----
From: "Sean Costello" <sean at staccatosys.com>
To: <efm at xavax.com>
Sent: Friday, October 20, 2000 2:22 PM
Subject: RE: [EFM] mmb - lives!


>
>
> > From: Harry Bissell [mailto:harrybissell at prodigy.net]
>
> > Use the 4520 dual binary up counter... 2 - 4 stage counters
> > in one chip. Smaller
> > better...
>
> Good idea. Price is around $.45 at Jameco.
>
> > If you want ring mod, you need exclusive-or gates not  and
> > gates.  OTOH
> > and gates would sound cool also.
>
> Apparently AND gates sound far more "consonant" for dissonant intervals.
The
> nice thing about this effect is that completely noisy signals as inputs
will
> turn into fairly smooth square waves after being divided down by 16. Of
> course, an XOR gate could be swapped in, if it had the same pinout.
>
> I was thinking that this configuration could be made even more powerful,
by
> incorporating a 3rd VCO that is divided down, and the results ANDed with
the
> outputs of the previous AND gate. This would multiply the suboctaves of
all
> three VCOs together, giving a sound that would be exceedingly superhuge.
> Could be done with 3 4024 chips, 2 4081 AND gates, and enough resistors
and
> opamp stages to sum the outputs together as required.  Or, you could use 1
> 4520 and a 4024, or use 2 4520 chips and have provisions for 4 inputs,
or...
>
> Some other fun things I have done with this circuit:
>
> - Run noise through a filter, and use this as one of the inputs. The
result
> is very dense noise, but with some sense of pitch; as the Q of the filter
is
> increased, the output sound gains a much greater sense of pitch.
>
> - Use two or more signals as the input to one of the octave divider
chains.
> The result is nice and chaotic.
>
> I have also found that mixing in the straight, non-ANDed suboctave signals
> greatly adds to the sound, so this should definitely be an option. Some
> experimentation would be useful to determine the best way of combining the
> outputs while using the minimum number of controls.
>
> I will put an MPEG on my website of these sounds in the next few days. The
> sounds were taken from a digital algorithm, but should sound about the
same
> in analog (well, even better, as those square waves tend to alias in a
> digital system).
>
> Sean Costello