[Asterisk-Users] X100P Cards have gone belly up?

[Steven Critchfield]

If you read to the end, the guy is not having trouble if he connects to
the PSTN via the first digit. I would be concerned that either the DTMF
detection in asterisk then was too strict where the PSTN would otherwise
detect it. So while it is possible, it seems odd that Digium wouldn't
have noticed it right off.

Of course maybe as a test, Ryan should try using a smaple.call file to
dial out and see if all the digits dial out. If so then Eric probably
has hit on a likely problem. 


On Mon, 2004-02-09 at 15:19, Eric Wieling wrote:
> Your problem has the classic symptoms of using inband DTMF and a
> compressed codec.  If you are using a codec OTHER than ULAW or ALAW you
> MUST set the DTMF mode on the PHONE AND in Asterisk to be rfc2833 (or
> INFO in some cases).  The classic symptom is that random DTMF is lost
> when dialing.
> 
> On Mon, 2004-02-09 at 14:59, Ryan R. Fligg wrote:
> > Alright, 
> > 
> > I have quite a problem on my hands and even the digium engineers are
> > stumped.  First my system layout
> > 
> > Asterisk CVS-01/15/04-16:44:03 built by root at tele on a i686 running Linux
> > 
> > X101P cards: 3
> > 
> > SNOM200 Phones: 5
> > 
> > 2 outgoing lines and 1 incoming line (DSL)
> >  
> > Okay, so one day our systems stopped working all together.  I know this is
> > very vague but I assure you that I did not change anything.
> > 
> > Our problem was that we were unable to dialout.  Receiving calls works just
> > fine.  Let me break this down.
> > 
> > Asterisk was up and running perfectly with no problems.  The X100P cards
> > were detected upon bootup and seemed to be loading just fine.
> > 
> > In our dialplan I have setup several contexts in which our callers are
> > limited to Local, Long Distance and emergency/immediate numbers.  A user
> > will 
> > 
> > Dial a 9 before any call and the pattern matching will interpret what kind
> > of call it is, as is defined in my dialplan.  When a user would try to
> > initiate ANY call the response from the console was that the X100P card was
> > dialing out, but the user would hear nothing on the Snom200 end even though
> > I could verify from the console that the X100P card was bridging the call to
> > the Snom200 phone.  
> > 
> >  
> > 
> > I then hooked up an analog phone to the X100P card and found the the X100P
> > was only dialing about ½ the number that was entered in the Snom200 phone.
> > I was a little stumped here so I called the digium tech support and the
> > ssh'd into our machine and were stumped.  They used zapbarge to listen on
> > the X100P card remotely and verified what I had heard.  Their fix to the
> > solution was to ditch all my contexts and use one universal context 
> > 
> > exten => _9,1,Dial(${OUT}/)
> > 
> > Now all my users dial 9 and then get tone to an outside line to dial.
> > 
> > This is fine and dandy for our office of 5 at the moment but the
> > scaleability of our system has been GREATLY compromised.  All the future
> > accounting practices that I was going to implement with MySQL and other
> > features are not available now.  
> >  
> > Here are the steps I took after getting the Digium solution:
> > 
> > 1)       Built a new Asterisk Box and tried the cards: Results: Same
> > 
> > 2)       Put a brand new card in the new asterisk box with current CVS:
> > Results: Same
> > 
> > 3)       Tested our lines and found that the current was about 75 mA, built
> > resistor packs and lowered it to 23 mA, industry standard
> > 
> > 4)       Used same hardware configuration as in 2 but with a new X100P card,
> > with lowered lines: Result: Same
> > 
> > 5)       Took configuration in 4 to a residential location with decent line
> > amperage and tested: Result: Same
> > 
> >  
> > 
> > So after determining that reordering new X100P cards from Digium would be
> > useless because I got the same results with our spare X100P, I am at a loss
> > as what to do next.  I will include my any files any of you would need to
> > look at my configuration upon request because I don't want to make this
> > e-mail lengthy with my rather large extensions.conf file and others.
> > 
> > Thank you in advance for all your help.
> > 
-- 
Steven Critchfield <critch at basesys.com>