Sounds very do-able.

How many do you plan to build?  If the number is small, contact a company
called 'Zetron' (I can find a phone number / address if you need) - they
make a box that does most of what you want.  You will have to process the AC
signals into DC voltages suitable for their A/D inputs but otherwise, the
box does everything (dial-up, modem data, hang-up; detect ring, get
password, modem data, hang-up).

If you need many of these boxes, build is the way to go.

I'd use a 16c73 for the brains, talking to a standard external modem using
the 'AT' command set for control.  What modem speed do you want?

The AC signal conditioning is only slightly involved.  Get the signals into
a reasonable range to work in (lets say 5V peak max); bandpass filter them
to limit the frequency response, convert to DC.  Do you need RMS conversion
or is Average responding adequate?

If all this sounds complicated, ask for help on SPECIFIC parts.  You may
want to consider hiring a consultant - there are many on this list.

dwayne


>Basically what I'm trying to do is to use a PIC (16C74?) device to monitor
>two ac voltage inputs. If either I/P exceeds it's predefined threshold over
>x successive samples, then the remote monitoring device would establish a
>modem link to a controlling point and report this.
>
>It is envisaged that the controlling point would poll each of the remote
>outstation at least once daily, no matter if any of them had alarmed back.
>It is also the intention that the control point operator would be able to
>select any outstation and be able to monitor the current status of thee
>I/Ps.
>
>The two I/Ps being monitored are very different and any suggestions as to
>how I could best interface them to the pick would be appreciated.
>
>Input 1: Very small ac noise voltage present in a telecommunications
>circuit. The
>voltage level is normally within the range 0.1 to 0.5 mV ac RMS and is
>terminated in 600ohms. The signal bandwidth is 0.2KHz to 3.0Khz. This is
>required to be put through a psophmetric filter prior to being sampled.
>
>Input 2: The second signal is again an ac V but the magnitude ranges from
>40V ac
>rms to 400V rms with the alarm limit being 60Vac rms. This voltage is again
>taken from a telecommunications circuit and is due to induction from
>overhead power transmission lines. This voltage is normally measured with on
>leg of the two wire circuit connected to Earth, and as such the i/p
>impedence required is 1Mohm. The bandwidth required is 0.5 Hz to 5Khz
>although this is under review.
>
>The remote station must be capable of being addressed.


Dwayne Reid   <dwayner@planet.eon.net>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(403) 489-3199 voice     (403) 487-6397 fax