Sawtooth generators

[Don Tillman]

   Date: Sat, 16 Mar 1996 14:16:31 -0800 (PST)
   From: "Duane R Balvage" <dbalvage at ptdcs2.intel.com>

   I noticed a small problem with sawtooth generators that use a capacitor
   discharged through a cmos gate or FET...

What VCO is this? 

     ----------------------------------------------V+
	     /|            ----
	    / |          /    |         
	   /  |         /     |      /|    /|
	  /   |        /      |     / |___/ |  
     _______________________________________________gnd
       mid freq        low freq      hi freq

	   as long as the f in ~= discharge rate of cap, you get nice
    saw waves. BUT, at low frequencies, the cap charges up to max, and then
    stays there @ that voltage until the gate turns on again, resulting
    in waveform B (not a very good saw wave, although it may sound
    interesting). 

No, this won't happen.  When the saw hits the trip point, it'll reset;
it won't keep going.  You didn't say how you were measuring the
signal, where the signal was coming from, or what the circuit was, but
it sure looks like you're looking at the signal sometime later in some
other circuit that's clipping the saw.

    At high frequencies >> discharge rate, the cap is never allowed to charge
    completely, resulting in a low output, inverted version of waveform
    B. (see waveform C).    

Same thing; the saw won't reset if the trip point isn't hit.
Typically what happens for frequencies high enough that the reset time
is a factor are two things:
  1. the wave becomes less like a sawtooth and more like a triangle.
  2. the pitch is lower than it ought to be.

  -- Don