Hi!

On Mon, 27 Sep 2004 14:40:58 -0700, Andrew Warren wrote:
> 
>     I guessed that A was amplitude, and tried various values; I set
>     initial X1 to 1.  X0 output just cycled through four values, no
>     matter how I adjusted ToneFreq, SampleRate, A, and X1... So I
>     assume that I'm missing something important.

It's a standard algorithm, and a very good one to implement
using fixed point arithmetic.

It's simpler to describe it this way:

     x(0) = Amplitude
     x(1) = x(0) - e/2
     
     x(n) = (2-e) x(n-1) - x(n-2) = 2*x(n-1) - e * x(n-1) - x(n-2)

Then, the frequency is: f = sampleRate * acos(1-e/2) / (2pi)

I once run a simulator to search "best values" of "A" and "e",
using fixed point arithmetic. Using 8.8 bits (ie, 16 bit add and
mult), I found it to be really stable.

        Daniel
_______________________________________________
http://www.piclist.com
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist