2 modules ideas

[Don Tillman]

   Date: Wed, 16 Aug 2000 01:30:12 -0700 (PDT)
   From: Don Tillman <don at till.com>

   I think this frequency-multiplying-a-given-triangle-wave approach is
   not optimal.  Sure it's possible, but it's a lot of work, and any
   inaccuracies are going to show as uneven waveforms, and errors will be
   worse for the higher harmonics, leaving a lot of the fundamental in
   each harmonic.

   It's a lot easier to build N sync'd triangle wave oscillators, all
   running with a common current source, and with capacitor values of c,
   c/2, c/3, c/4, etc.

On second thought, both approaches have their advantages.

The N sync'd triangles is nice because you can have N outputs and
adjust them to be whichever harmonics you'd like.  But any frequency
inaccuracy will be taken up at sync time, while with the frequency
multiplying approach any frequency inaccuracy is adjusted over the
whole cycle of the fundamental.

  -- Don

-- 
Don Tillman
Palo Alto, California, USA
don at till.com
http://www.till.com