We've inherited a pair of mostly identical PBX systems, each with a
TDM400P Rev I boards and 4 FXO modules. The production system is
running Asterisk-Now with 1.4.9, and despite some other issues, it
is able to answer inbound calls just fine. The replacement system is
currently running Asterisk 1.2.28, and is unable to detect incoming
calls, outbound calls work fine.
We discovered later that the analog lines are supplied by Cox Cable
in Los Angeles, apparently Cox is the only telco available in this
office building. I came to the conclusion that Cox is probably
providing analog lines via Cable/VOIP service and that the FXS ports
in their equipment are providing a lower than normal ring voltage.
ztmonitor shows the ring, but Asterisk 1.2 never starts a simple
switch on the zap channel. I did find a posting from the 1.0.X days
where someone had to lower the sample peaks in wcfxo.c from +-32000
to +-10000, but this seems to be the standard in the current 1.2
zaptel source (wctdm.c of course). Clearly the 1.4 zaptel driver is
doing something different, but I'm not sure what since the values
look the same.
We've been avoiding Asterisk 1.4 because of some serious stability
issues we've seen in several versions. 1.4.19 seems to have
addressed these issues, so we're likely going to deploy the
replacement PBX with 1.4.19. Just mentioning it because I'm sure
someone will ask ;-).
While I hope to never run into this issue again, it does raise the
question of how could one determine the actual ring voltage, as well
as any other analog line values that would help troubleshoot this
sort of issue? If there are further issues with the analog lines, we
may need the ability to detect and tweak the ring detection parameters.
Thoughts?
Chris
