NetNews Usenet Archive 1992 #16/NN_1992_16.iso/spool/sci/electron/13301

Text File | 1992-07-25 | 3.0 KB | 58 lines

Newsgroups: sci.electronics
Path: sparky!uunet!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu!sal.wisc.edu!rat!cindy!rmintz
From: [e-mail] (Rich Mintz)
Subject: Ring Signal/Software Problem
Message-ID: <[e-mail]>
Sender: [e-mail] (no news is good news)
Nntp-Posting-Host: cscihp.ecst.csuchico.edu
Organization: California State University, Chico
Date: Sat, 25 Jul 1992 05:26:16 GMT
Lines: 46

I have an application I've programmed that monitors the phone line.
I've had a constant problem keeping track of how many time the phone
has rung, because of all the specific timing problems associated with
checking for the duration of a ring, how long after ringing stops to
reset the count, considerations for passing through the ring-check
procedure in mid-ring, while constantly monitoring if the phone has
been answered (I'm using a voice mail card that can detect a pickup on
the connected telephone) and doing miscellaneous other processing.

Through trial and error, I can sometimes get it to work right, but
when I change to a different type/speed of computer, or change some
aspect of the program that will affect the timing checks, it messes
everything up. The RI serial line also jumps between 1 and 0 very
quickly when a ring is detected, which complicates the routine to
reliably determine the presence or absence of a ring. This indicator
also seems to get falsely triggered sometimes when a pickup or hangup
is made. I had considered constantly polling the Hayes S1 register on
the modem with serial i/o (it keeps track of how many rings have
occurred), but there were problems with polling commands to the modem
getting messed up when the modem would send a ring result code at the
same time as commands were being sent to it, which made that method
unreliable. On top of that, I found that doing serial i/o in Turbo C
isn't exactly fun (ie: pain in the ass).

I have the phone line going into my modem, and then to my voice mail
card, which then software-selectably passes the line to a connected
telephone. I disengage the phone from the line via software at times
when I don't want my phone to ring. If I put a rectifier bridge and a
big capacitor on the phone line going into the modem/voicemail card,
will it (A) still ring my phone when I select to pass the line through
to the connected (electronic ringer) telephone, and (B) register as
ring detect to the modem and produce a continuous "1" signal on the
serial RI line (as opposed to the normal 1..0..1..0 ring signal), and
(C) screen out false '1' "blips" in the signal when I pick up or
hangup my phone? In other words, will just the positive portion of
the AC ring smoothed out with the capacitor so that it appears to the
various ring-detect circuitry as a very elevated DC voltage be
recognized as a ring? If not, is there an alternate configuration
someone can suggest that would do this job?

Note: my modem is internal and I'm using an AT 386 SX @ 16 Mhz.

Thanks a lot . . .

      Rich -> [e-mail]