[sdiy] VCO reset time

[Magnus Danielson]

From: Don Tillman <don at till.com>
Subject: Re: [sdiy] VCO reset time
Date: 03 Jun 2004 22:24:36 -0700
Message-ID: <m24qprj3or.fsf at till.com>

Don,

>    > Date: Wed, 02 Jun 2004 21:21:56 +0200
>    > From: Rene_Schmitz <uzs159 at uni-bonn.de>
>    > 
>    > > What is a good ramp reset time ?  How fast have you seen
>    > > ramp reset be ?
>    > 
>    > Zero. (A special triangle core ;-))
> 
> 
>    > Date: Thu, 03 Jun 2004 00:14:30 +0200 (CEST)
>    > From: Magnus Danielson <cfmd at bredband.net>
>    > 
>    > I consider triangular cores a very interesting concept and they
>    > might compete very well with sawtooth cores IMHO. 
> 
> 
> I've been questioning the whole point of sawtooth cores for a couple
> years now, and I'm thinking that triangle cores are a lot better.
> 
> Here are three advantages of triangle core VCO's:
> 
>   1. No reset detuning.
> 
>   2. You can get each of the basic waveforms in just one step:
> 
>      triangle -> saw
> 	      -> square
> 	      -> sine

Actually, with a triangle core you get the square waveform for free, since you
need the square waveform to change the charge-direction, so that state is
there.

The triangle together with it's squarewave makes it easy to do the sawtooth.
If you try to generate a squarewave from the triangle such that it is suitable
to aid in the sawtooth generation, then you got yourself into much greater
trouble than you might first think.

>      With the sawtooth you need two steps to get to the sine wave (and
>      possibly the square wave depending upon how you want PWM to
>      work):
> 
>      sawtooth -> triangle -> sine
> 	                  -> square
> 
>   3. Any glitches in the triangle-to-sawtooth conversion will be less
>      noticable because the sawtooth has more harmonic content to mask
>      them, while any glitches in the sawtooth-to-traingle conversion
>      will be more noticible as there is less harmonic content to mask
>      them.

4. A Pulse-Width Modulated output based on triangle does not have the phase-
modulation that a sawtooth based variant has. This may or may not be a good
thing depending on what you are after.

> The only advantage I can see to a sawtooth core is in visualizing the
> waveshaping.  That is, your waveshaping transfer function looks like
> your output waveform.

We bend our minds because we like to bend our minds. Choosing one design which
may be worse than another design just because some other part will be easier to
come up with may not be the best of rules to go by. Besides, you can still do
things with sawtooth if you want to. Many of the triangle waveshapers we might
want to consider is fairly easy to figure out anyway.

> Of course a trapezoid core VCO is better still.  :-)

Let's just put the VCS3/Synthi-A envelope generator into turbodrive then, shall
we? ;O)

Cheers,
Magnus