[asterisk-users] X100M never goes on-hook state

[Marlon Dutra]

Hi guys,

I've been experiencing a very strange issue with my Digium Card TDM400
as of this week. It has two FXS and two FXO.

The FXO modules (both of them) never goes on-hook after hanging up in
Asterisk. It had worked perfectly well for over four years.

I put an ammeter in series with the line and the card, and immediately
after plugging the connector to the card, I got 26mA in the circuit and
a dial tone from the carrier, where it should be zero amper (on-hook
state).  If a Dial() something, it works perfectly. I can Hangup() the
call, freeing the channel in Asterisk, but the hardware keeps off-hook
forever, locking the line. If I Dial() again, Asterisk opens the line,
sends the DTMFs normally, but it doesn't work since the carrier thinks
I'm still holding the first call.

It behaves exactly the same way with another analog line. If I plug
either of the lines and my other Digium card (TDM2400), it works ok. The
same with my Brazilian DigiVoice FXO card.

Ok, you all might say: your card is damaged, throw it away. Ok, I could
do it, but now comes the funny part:

If I put an DSL filter in series with the line and the card, IT WORKS
PERFECTLY!!! The filter imposes 25 ohms over the circuit. Maybe that's
causing the card to work. When I put the filter and the ammeter in
series, I get zero amper when on-hook and 26 mA when off-hook, that's
the expected behaviour.

I'm not an expert in electricity, so I really don't know why the card is
behaving that way. What does that resistance make for the card to start
working ok? I know the DSL filter isn't only a resistor. Maybe it has
another electrical component that's helping more than the resistor. Just
a guess.

Tomorrow I'll buy a 30-ohm resistor, take the DSL filter off, and test
the card only with the resistor, to check it out.

In order to isolate the problem even more, I plugged the FXO port in one
FXS port. Immediately after plugging it, Asterisk announced at the
console that someone went off-hook at the FXS port. So, it's not really
a carrier issue. The FXS port is perfectly -48V on-hook, and about
20 mA in the circuit when off-hook, closer than the carrier to the
standard values.

Any clue is welcome.

-- 
MARLON DUTRA
Propus
GnuPG ID: 0x3E2060AC pgp.mit.edu
http://www.propus.com.br/
http://hackers.propus.com.br/~marlon/