mini moog clone

[Haible Juergen]

	>At 01.09 18/09/00 -0400, KA4HJH wrote:
	>>>I'd avoid using the Fairchild uA726 Temperature Controlled
Differential
	>>>Pair,
	>>>for starters....
	>>
	>>No way, that's a fundamental part of the *Moog sound*, dude. 
	>
	>The easiest way is to keep the old osc board design. IMHO the
necessary
	>additions  are: PWM instead of fixed squares, a dedicated LFO, VCO
sync,
	>an overall octave switch (takes all 3 VCOs up or down), four stages
	>envelope (ADSR),  rotary switch to address vco or noise signal to
the ext
	>in, different resistance value for the cut off and emphasis pots
for a more
	>linear response (not like real mins where nothing happens from 0 to
8 and
	>it is all between 8 and 10), and (why not?) an envelope for VCO's
pitch (so
	>I can play Hoedown on it).
	>ciao
	>riccardo

First of all, I think the 726 argument was clearly a joke.

Second, I remember a very intelligent person making the bold statement that
the
wooden finish has to do with the minimoog sound as much as the filter or
oscillators (I'm paraphrasing). While it was never literally true, it still
makes a
lot of sense.

Now seriously, in my personal opinion, substituting a set of fixed waveforms
with 
PWM will - change the sound !
Not the sound of a certain setting or waveform of course, but in the
practical use
of the instrument. And I'm not speaking of technical differences here. Let's
asume
that the MM's original pulse waves are a full subset of the PWM circuit
you're
going to build. So where's the difference ?

The difference is that you are not likely to choose these exact pulse
waveforms 
during playing or programming. The case might be different on a synth that
can store
patches (but then again, just the procedure of storing patches will have a
similar
effect on the real time sound ...)
On a Minimoog, you will pick a combination of waveforms.
On a synth with PWM instead of the fixed pulse waves, you will choose
something
different.

It's questionable if I am really talking of "sound" here. People tend to
split an instrument's
features into categories like user interface, controllers, whatever, and the
"pure sound".
But this is misleading at best. There is nothing like "pure sound" other
than under
laboratory conditions. In real life "sound" is what comes out of the
instrument,
and this is affected by the approach the user has to take to get a certain
sound out
of it.

So much for philosophy (;->).

As a practical suggestion, keep the fixed waveforms and *add* the PWM rather
than
replacing anything.


Speaking of PWM, I recently noticed that the Yamaha CS-series synth's PWM is
quite different from other synths, and that may be part of the reason why
even the
single VCO per voice models sound so "rich". The Modulation range is
*clipped*
(as opposed to "limited") to approx. 10% and 90%. So you can overdrive the
PWM
without modulating out of range. 
The effect is that the spectrum will not be continously animated as in
normal PWM,
or momentarily disable the wave like in normal over-modulated PWM, but you
get
a periodic "part time" change of spectrum instead. Does this ring some bells
?

If ordinary PWM resembles two saw waves beating / phasing against each
other,
the CS-series PWM resembles two saw waves phasing against each other *and*
the animation of this beating being partly masked by ... *filter overdrive*
!

Do some experiments and see for yourself ...

JH.