4.1b Release
28-Feb-95 22:27:47
Sb: #67928-4.1b Release
Fm: George Steeg 72560,3176
To: LESZEK T. INGLOT 74442,1336
Leszek, Caller ID stopped working for me when the local telephone company
switched from number-only (standard) to name+number (deluxe) service, even
though I did not upgrade from standard to deluxe. If your DTS suddenly stopped
working, it may be a change in your local service, not DTS. Following sent to
PPI forum, describing my experience:
>>Caller ID stopped working for me when the local telephone company (Bell
Atlantic = BA) switched from number-only (standard) to name+number (deluxe)
service, even though I did not upgrade from standard to deluxe. BA swears
they did nothing to interfere with standard service. My exchange uses the AT&T
5ESS switch, which had a previous software change that interfered with some
Caller ID software. BA claimed at that time that they didn't cause the
problem, but I had waveform measurements that proved they did.
Early modem chips with Caller ID are not supposed to work on deluxe format.
When my PPI 14.4 stopped working, I sent it back to PPI who checked it out as
working to Caller ID spec, that is: OK on standard, no good on deluxe. Most
modem suppliers claim their new 14.4 and 28.8 chips work with both formats. I
upgraded to 28.8. It works as promised.
You can check out modem response in Windows Terminal. Enter AT followed by the
Caller ID enabling string, e.g. AT%CCID=1, for your modem, then CR. You'll get
OK. Then call yourself on another line (or have someone call you). You should
get a first RING, followed by ID data, and then additional RINGs. If you get a
first RING followed by RING, and you know the modem is OK, then the local
telephone company Caller ID service is inhibiting the chip function.
The problem is that for the deluxe service, the name is sent first. Both name
and number data is frequency shift keyed (FSK). I'm guessing that sending the
name adds something to the waveform that prevents the chip from functioning on
the number. BA insists their waveform is to specification, and that the
problem is with vendors' implementations — but they are double checking. If I
hear more, I'll follow up on this thread.<<
Regards, George.