[sdiy] Ring Mod (was Re: Hadamard Transform Network)

[Neil Johnson]

Hi,

Ben Bradley wrote:
> A "4-quadrant balanced modulator" describes a function, and it can be
> done with several different circuits. Perhaps the most ubiquitous
> example (that's NOT a ring modulator) is the MC1496 chip.
>
> A "ring modulator" does that too, but the narrow use of the term is
> for a specific circuit with four diodes connected in a ring
> configuration. A ring modulator isn't "as good" electrically as many
> other circuits, because the forward drops of the diodes add distortion
> to the output (there's no output until the input signals exceed the
> diode forward drop), but like so many faults and imperfections in
> sound and music generation/reproduction, people get used to the sound
> and end up liking it more than other, "cleaner" circuits.
>
> There are two articles on Wikipedia, both showing the ring modulator circuit:
> https://en.wikipedia.org/wiki/Ring_modulation
> https://en.wikipedia.org/wiki/Ring_modulator

Hello - if you look at the history and talk pages of the wikipedia
entry for the ring modulator you can see some of the
confusion/controversy surrounding this seemingly simple circuit (I'm
"Nejo17" BTW).  The original wikipedia page on ring modulation ("ring
modulator" was a redirect to that page) described the ring modulator
almost exclusively in the context of electronic music, as if that was
the *only* application of ring modulators.

I would wager that if we counted the total number of things that are
called "ring modulators" in existence, the majority are diode ring
modulators used in telecomms and radio comms.  There are literally
thousands upon thousands of real ring modulators out there in
telecomms and radio comms systems.  A vanishingly tiny percentage of
what are called "ring modulators" are not actually ring modulators but
other circuits based on Gilbert-cell multipliers or VCAs or other
4-quadrant multipliers.

Ironically, all this talk about diode drops introducing distortion
completely misses the point of how a ring modulator works.  Yes,
mismatches in the forward drops introduce some small imperfections,
but easily a couple of orders of magnitude smaller than anything
caused by the total forward voltage.  The carrier should/must be a
very sharp square wave which quickly turns on/off the corresponding
diodes at each half cycle.  Anything less than perfect square waves
introduces unwanted components into the output.  For example:

https://electronicspost.com/ring-modulator-for-the-double-sideband-suppressed-carrier-generation/

Another way of looking at the diode ring modulator is that it is a
type of chopper modulator. Once you start googling for chopper
modulators, Cowan modulators, and so on you'll find a lot more
information about these devices, very little of which has anything to
do with electronic music.

As regards the "technically shit but we like how it sounds" point,
yeah that seems right ;)

Neil
-- 
http://www.njohnson.co.uk