On Fri, 23 May 1997 00:38:28 +0000 John Payson <supercat@MCS.COM> writes:

>I have done DTMF decoding using just a zero-cross detector and a 20MHz
>PIC.  Depending upon whether tone generation is required as well, the
>routine uses about 40-70% CPU time [it has two parts: an interrupt
>routine
>that uses about 40-50% CPU, and a calculation routine that needs to be
>called about 60 times/sec].
>
>Unfortunately, because I developed the routine for my employer I
>cannot
>release it on the net, at least not presently.  On the other hand, I
>will
>say that it is possible to do a good job of detecting DTMF or other
>such
>tones, using a PIC, even in the presence of other audio (such as dial
>tone).  My routine looks for eight tones simultaneously, and is
>selectable
>between two groups of eight (one set up for DTMF, one for
>call-progress
>and special tones).
>


        Interesting!  You're detecting simultaneous tones!  I'd always
thought of splitting the tones into high and low groups, then timing the
period of each.  Are you doing this with no band splitting?  Just a
comparator or schmitt driving a port?

Harold