tri 2 saw modualtion

[James Joseph Clark]

> 
> A triangle-saw sweep is the integral of a square-pulse sweep (if all
> of your pulse waves have the DC properly stripped out), so the nulls
> are in the same place and all.  But the spectrum is tilted 6dB/oct.

But in this case the amplitude of the pulse would need to grow as it
got thinner (as the tri-saw approaches saw) so it's not quite the same
as a pulse width sweep..

> 
> Another for the synth list: the Matrix-N family has saw/tri modulation
> on the audio oscs.  I've never looked at the waveform, though.

The Matrix-6,1000 use a wave shaper on a variable amplitude sawtooth to
generate triangle-saw sweeps of constant amplitude and pitch. The only
problem is that between the triangle and sawtooth waveforms is a
trapezoidal wave, and not a skewed triangle.

It goes something like this:

INPUT TO SHAPER:    (note: I use a variable pitch due to limitations of ASCII art!)
                / 
          /|   /|
     /|  / |  / |
 /| / | /  | /  |
/ |/  |/   |/   |

SHAPER TRANSFER FUNCTION (V_OUT vs. V_IN):

   /\
  /  \
 /    \
/      \_______


SHAPER OUTPUT:



 /| /\  /\  /\
/ |/  |/  \/  \_

I don't know if this makes sense or not. Details can be found in the
CEM3396 data sheet which can be found on Paul Schreiber's web page.

Jim Clark
clark at cim.mcgill.ca
http://cim.mcgill.ca/~clark/mxed.html
http://cim.mcgill.ca/~clark/cylonix.html

> 
> -- 
>      Eli Brandt  |  eli+ at cs.cmu.edu  |  http://www.cs.cmu.edu/~eli/