Mike Hagans wrote:
> There are several inexpensive, reasonably fast, small, dtmf receivers out
> there.  A 14pin DIP package I've used recently with great success, as a
> peripheral to a 16c84, is the motrola mc145436 dtmf decoder.  It does the
> filtering, detecting, and decodiing for you.  It is very friendly to
> interface, has a large (accurate) dynamic range, and rarely if ever
> falses.
All above sound attractively ;)

> Although it would be an accomplishment, doing this in software on a pic
> (no dsp engine) will be difficult, time-consuming and resource eating.
Yes.

> I'm all for making pics do as much as possible, but you'll have a tough
> time making a cost-effective replacement for a <$2 part like this.
The problem in that the all Motorolla components in Russia have very
high prices.
So creating PIC DTMF decoding algorithm are job with great significance

> I've also had success with SSI202 and M8880 varieties.  For the SSI202
> and MC145436, you just monitor strobe and four bits of data... the M8880
> is a transceiver and requires a bit more manipulation.
Hardware solution have certain advantage before software of course.
( At the side of simplicity ;)
But there are some project that demand the minimal resulting cost of
each device,
than less cost of component used than better.

Dmitry. Zelenograd, Russia.

P.S. Has anybody information/link_to_WEB about some specific filter
theory ?
This filter have to pass to output some freqs at the same time, for
example: [Fo,Fo+dF,Fo+2*dF,Fo+3*dF,...,Fo+n*dF] or
[Fo,Fo*n,Fo*n^2,Fo*n^3,...,Fo*n^k],
where Fo - basic frequency, dF - step, n - some coefficient .