[sdiy] Discrete OTA

[David G Dixon]

> W dniu 2014-04-09 23:48, David G Dixon pisze:
> >
> > The fact that I have never published the schematic for the Rubicon 
> > should give you a hint.
> >
> > I'm not really sure what point you are trying to make.  
> > Please clarify.
> >
> >
> clarification:
> actualy I do have few new designs of TZ-VCO in my copybook, 
> and 1 of them is triangle core. Invented before I knew 
> Intelligel exists.
> Well my first TZ-VCO was playing in 1999, but I can't 
> remember if I ever posted the schematics somewhere.
> Those creations will be published sometime eventually.
> 
> Roman
> I know, I said I'll shut up on this in last email. Now I do.

OK, that's what I thought.

Here, for all posterity to ponder, is how the design of the Intellijel
Rubicon Through-Zero FM Triangle Core VCO came about:

I was sitting on the can thumbing through a textbook that I had checked out
of the UBC library.  I can't remember exactly what book it was now, but I
think the author was Joseph Carr.  Anyway, there was a simplified triangle
wave oscillator (not a VCO, just an O) schematic on page 300-something, and
I started looking at it.  Just by looking at it, I realized that this
oscillator could be the basis of a nice tricore VCO using the 2164 as expo
converting current amp -- this became the Intellijel Dixie.  It also
occurred to me how to make it do TZFM right there, in that very first minute
-- that became the Intellijel Rubicon.  Note that the Rubicon concept, while
successful, was a bitch to get right, and required many revisions before
launch.  The Dixie design and product came together rather easily.

So, why was I interested in a tricore design?  Because I had just designed
and built a 2164-expo-based sawcore design, and it was very difficult to get
right (my trials and tribulations with it are well documented in the SDIY
archives).

I later discovered (through this list) that there is a similar design to the
Dixie core which had been made public.  It is an LFO design by our very own
Neil Johnson.  However, there are still many significant differences between
the two designs.  I am not aware that anyone has ever implemented TZFM in
the way the I have.

I learned of the 2164 tempco scheme from Sam Hoshuyama, and I think I earned
my "street cred" on that front by actually doing the mathematical analysis
of the situation, and calculating what the proper tempco VC pin voltage
should be, given the true gain ratio of the 2164 (information I will happily
share with anyone who is interested).  I only learned much later that Roman
Sowa did it first.

I have never seen, nor was I aware of before today, Roman Sowa's TZFM
schematic.  I have seen Ian Fritz's, and the one in Electronotes, and I had
a chance to play with a Zeroscillator (although I think mine was broken,
because it didn't work the way I thought it should).  I can say that my
design is fundamentally different from all of those.  Ian has seen my
design, and will concur.  I shared it with him after that whole "Blue
Lantern" drama.  I can't remember ever looking at any of Roman's schematics.
I don't often look that closely at other people's schematics when designing
something.  I take a "systems design" approach to circuit design -- I think
through what I want the circuit to do (usually by drawing waveforms on my
office chalkboard, or by deriving transfer functions), and then I put
together a sort of "block flowsheet" to achieve that, and then I assemble
the circuit from previously existing circuit blocks which I have used before
and with which I am sufficiently happy.  Some of these blocks I have
"borrowed" from other sources.  For example, my "go-to" full wave rectifier
is from a webpage by Rod Elliott (Figure 6 from here:
http://sound.westhost.com/appnotes/an001.htm -- he has pubished a ton of
tutorial information, and I assume it's because he wants people to use it,
and it's all really good).  The circuit block I have used probably more than
any other is the linearized 2164 VCA circuit published in EDN Magazine by
Mike Irwin, and even that I have significantly improved and altered.  (Phil
Gallo pointed out a dumb blunder in my original expo-linear VCA circuit, for
which I am eternally grateful.)  Other circuit elements I have gleaned from
textbooks, websites, etc.  I never use anything without confirming its
performance on a SPICE simulator, and I almost always make my own tweaks
after I understand it.  I have cloned a few things for my own use (such as
Ray Wilson's ADSR), but I have never copied a whole synth circuit for a
commercial product, and I don't expect that I ever will.  The closest I came
was with the Korg MS-20 filter: I developed my own filter based on the
transfer function, which I derived, manipulated, and then generalized into a
completely new 2164-based product which I believe is infinitely better than
the original.

That's how I roll.