[asterisk-users] X100P Burnouts

[Brent Davidson]

Thought I would post this experience to the list so it's archived for 
posterity...  My company is deploying Asterisk-based PBX's to all of our 
branch offices.  Each office has 2 analog Voice lines and a fax line.  
We didn't want to go to the expense of using TDM400's in the servers 
(which run asterisk and Hylafax) so we opted for 2 X100P cards in each 
box.  So far they have worked fine at all but one office.  The system at 
the office in question would work perfectly for an entire day once it 
was set up.  The next morning, however, one of the 2 phone lines would 
appear to be dead and the X100P would be in Red alarm.  Plugging a phone 
into the X100P's pass-through connection would show dial-tone on the 
line, and the phone worked perfectly.  It was as if the X100p lost it's 
ability to see the audio on the line and nothing would revive it.  Tried 
restarting the zaptel module, rebooting the server completely, complete 
power down, unplugging the phone line and even connecting up a phone 
line simulator and moving the card to another server. The card never 
works again.  This went on for three days.  Burned out an X100p every 
night.  I called the telco (Verizon) and they sent out a couple of guys 
to run tests on the line, but found nothing.  Their Demarc is properly 
grounded and has surge suppression modules attached, the cable that runs 
from their demarc to our punch-down block is in grounded metal conduit 
and does not run near any power source.  The cable that runs from the 
punch-down block to the wall jack also does not run anywhere near 
anything electrical, and everything is twisted pair all the way from the 
wall jack to the demarc.  To further eliminate the possibility of echo 
or other noise, I ran twisted pair carrying both voice lines from the 
wall jack to the server, approximately 3 feet in length.  Now here is 
where things get interesting...  That 3 foot cable run passes behind a 
21" monitor that was connected to the server.  When the line tests 
showed everything OK, I decided the monitor might be a long shot but I 
could understand how the degaussing coil coil could possibly induce a 
surge on the phone line if the monitor was somehow degaussing nightly, 
so I unplugged the monitor's power cable and left everything else as it 
was.  So far so good.  X100p #4 is still working this morning, so it 
looks like the problem is solved.  Hope this helps someone else later on.

Thanks,
Brent Davidson