MGCP FXO Disconnect

Unanswered Question
Aug 31st, 2009

I am experiencing an FXO disconnect problem when i configure the gateway as MGCP. Is there some special settings i can set in the callmanager service or somewhere else for MGCP to work.I have CCM 7.02 and 2821 with IOS 12.4T.

I realise that FXO disconnect is problematic on Loop Start FXO Ports.

If PSTN calls IPT Ext and hangs up, the Internal IP handset continues to ring. There are no obvious config changes on Call Manager Gateway config pages.



I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (1 ratings)
iwearing Tue, 09/01/2009 - 08:17

Thanks for the response.

I will get a copy of the cfg and debug asap.


benzhao98 Mon, 08/31/2009 - 06:48


Below way are workiing well in most scenarioes, please try.

1.try below IOS version.


2.add below commands into voice-port

voice-port 0/0/0

supervisory disconnect dualtone mid-call

cptone CN (CN is for china, please change to the your local cpton)

timeouts interdigit 2

timeouts call-disconnect 1

timeouts wait-release 1

iwearing Tue, 09/01/2009 - 08:20


I tried the commands on the FXO Ports. Unfortunately I still experience the same issue.

If PSTN caller hangs up before IPT Phone is answered, the IP Phone continues to ring.


iwearing Thu, 09/03/2009 - 07:55


Apologies for the delay. I have no remote acces to the site which is a 2 hour drive away.

The attachment contains Config and debug VPM signal during a failed call.

IOS Version 124-15.T9



pcameron Fri, 09/04/2009 - 13:16

Are you sure you are using Ground Start operation ? Your CPtone is set to UK and GS signaling is virtually unknown out of North America.

voice-port 1/0/7

supervisory disconnect dualtone mid-call

supervisory custom-cptone DISCONNECT

supervisory dualtone-detect-params 1

signal groundStart

cptone GB

timeouts interdigit 2

timeouts call-disconnect 1

timeouts wait-release 1

timing hookflash-out 50


You also have a custom tone class configured - how did you arrive at these particular frequency/cadence settings ?

If you can get a wireshark capture of the G711 audio stream between the gateway and the phone after the call has been disconnected we can do an analysis and confirm the tone details. If possible, do this and attach to this thread. We only need around 10 seconds of packet capture, just as long as it has some of the tones.

a.gooding Thu, 09/10/2009 - 17:22

PAUL is correct with everything he is saying as i know only too well with this particular issue :).

this is what i do for all sites we deploy usin analog (even the one we did yesterday :).

first off your scenario is exactly what occurs on my side.

below is my config but along with that we normally need to ask the CARRIER to enable cut off and disconnect (COD). This means that the disconnect happens on the carrier side rather than having our FXO ports do the detection of the disconnect signals (in our case there really isnt any hence the issue).

this is a basic config that works flawlessly for me As mentioned, the tones on your side may need to be investigated but nontheless here it is

voice class dualtone-detect-params 100

freq-max-deviation 50

freq-min-power 35

voice class custom-cptone TEST

dualtone busy

dualtone ringback

frequency 500 632

dualtone reorder

frequency 500 631

cadence 500 500

dualtone disconnect

frequency 630

cadence 500 500

supervisory disconnect dualtone mid-call

no battery-reversal


echo-cancel coverage 24

cptone C1

timeouts initial 5

timeouts interdigit 2

timeouts call-disconnect 5

timeouts ringing 10

timeouts wait-release 1

timing hookflash-out 50

timing sup-disconnect 100

its worth a shot,hope it helps a bit

Jeff Bankston Mon, 12/03/2012 - 18:53

I used this config on my voice gateway, MGCP using FXO-M1 here in the States, and it works near perfectly. When the call is up from my lab IP phone to my cell phone over the FXO, and I hang up my cell phone, the IP phone is still seemingly on the call. 10 seconds after my cell phone is hung up, the IP phone gets a fast busy. Then, 30 seconds later the FXO port resets and becomes available for use again.

Perhaps not perfect, but at least this is a viable workaround for one of my clients, thanks for the post still being here.


a.gooding Mon, 12/03/2012 - 20:04

Great stuff,

the COD part is something you might want to ask the carrier for as well. there are defintely a few ways for this but on my side of the world, COD is the key along with the above.

great to know it at least helped.

Kazkommertsbank Mon, 12/03/2012 - 22:18

I had this problem too. I just changed MGCP to H323 for analog ports. PRI trunks remained on MGCP.

And problem was resolved. Everything woks OK.

Ilya Geraskin Thu, 08/29/2013 - 06:03

Hi guys,

If anybody have the same trouble - try to use:

timeouts initial 0

timeouts call-disconnect 1

timeouts wait-release 1

on voice-port

It helps if the disconnect works on h323 or SIP gateway but doesn't work on MGCP.


This Discussion