Is the Disconnect-Time adjustable? 2

[bryanfarrish]

My new Partner R6 is holding on to phone and intercom lines (especially intercom) for a long time after hangup. I.E., after a user checks his voicemail and hangs up, the system keeps his intercom light engaged (and thus a voicemail port used) for 30 to 60 seconds afterwards. It does it also with outside lines, just not as long.

Is there a setting to tell the partner to release the lines and intercoms sooner, after hangup?

Bryan

 

[Dexman]

Hey Chris

I guess it will be a wait & see when it comes to VOIP. Since VOIP is being accepted more & more, things can only get better (I hope).

I hope with all the posts, Foozy has a better understanding of what direction to head in to resolve the issue.

Paul

 

[foosy]

Dexman,

Great input so far.
I need to call AT&T to find out what signaling exists at the client location. I will post back on this either later today or tomorrow.
Thanks again!

 

[Dexman]

Just wondering if Foosy found out what type of signaling is being used by the system.

 

[itpphoneguy]

Hi Paul,
I see no reason why "the box" doesn't have disconnect supervision no matter the signaling. Either way the box is told to release the connection from the far end (CO). The signaling on a PRI is out-band, T1 is in-band, they are both valid. The only diff. I can see is the possibility of glare (remote).
Am I wrong here? After all, a native machine will release instantly on T1 or PRI (like a 100DCD on a Magix for example). LS or GS does not seem to matter.

-Chris

 

[Dexman]

Hey Chris,

It all goes back to the signaling bits.

With loop start signaling, you do not get a positive disconnect if the at the other end of the call hangs up before the party at the CPE end does.

As strange as this may seem, it is true. It is simply a weakness in combining loop start signaling with T1 service.

That is why, for example, if you dial a toll-free number that terminates on a dedicated T1 circuit that utilizes loop-start signaling, (not ISDN, SS7, T1E&M or ground start), and the CPE employs some sort of auto attendant, and you, the caller disconnects while the auto attendant is doing it's thing, the auto attendant keeps right on playing until there is either timer in the CPE that cuts the call off due to no response from the far end, or the dead call is sent to a "0/covering" extension where a person answers the phone, hears dead air and hangs up.

This is not a problem with T1E&M or ground-start signaling as the signaling bits act differently and will provide a positive disconnect to tell the CPE to release the call when the far end party hangs up. Naturally with out of band signaling methods such as ISDN and SS7, messaging, sent along either a "D" channel or on links, tells the CPE to drop the channel when far end party hangs up.

The only time I won't blink at loop start trunks on a T1 circuit is if the customer will be using the circuit to originate calls, but not accept incoming calls (think telemarketer with 24 stand alone analog single line telephones). In that scenario, positive disconnect will always be generated from the CPE and cause the central office to release the call.

To completely explain the differences in how the signaling bits act for loop start, ground start and T1 E&M would take a vast amount of typing + visuals and is probably beyond the scope of this message board.

(Sorry about the large amount of text I've typed above. Explaining signaling, even a simple gloss over, can involve alot of typing).

Paul (Dexman)

 

[itpphoneguy]

Hi Paul,
Thanks for that. I'd like to learn more about that if you can PM with a link, I'll do the reading. We only ever order PRI's (ISDN) for voice service up here. That's why I have never run into this on T1 service. Looks like the extra channel is just not worth it.
-Chris

 

[Dexman]

Hi Chris!

I'll see if I can find something online. I googled "Loop Start vs. Ground Start" and it came back with some links, but I haven't had the chance to read through them all. There are other keywords that could be tried.

If things are slow at work on Monday, I'll see if I can find something good. There has got to be some good articles online that discuss the various types of signaling.

Paul (Dexman)

 

[foosy]

First of all I owe all of you an apology for not getting back to you with more details. The situation got hairier than I expected causing a state of panick...

So, last I left you with an intermittant problem with lines not always disconnecting. AT&T showed up to check the channel bank and decided to change the lines from '2 seconds battery removal signal' to '0.5 seconds battery reveral signal'. They checked that lines worked and left only for the client to discover that calls were not getting through: if someone was calling, the line would get disconnected as soon as someone would pick up the phone. Unfortunately it took AT&T a day and a half to send their tech back to restore the old settings. During this time I saw many red faces... to say the least ...

Technically the Partner was set to 450msec which should have been OK in both cases. Also, I looked at the channel bank: it was set to use Tandem for signaling.

So I was back to square one. It got me thinking and I found the problem somewhere else...

I figured that what users where attributing to 'line not disconnecting' was really a Partner issue. If somebody was on a call and then hung up, when they picked up the line again -there was no dial tone. They were clicking the switch and thought they were releasing the hung line that had not disconnected from their previous call. In reality every time you 'click' or pick the handset, the Partner hands you a DIFFERENT line. So the 'dead' line would stay dead in the system until the Partner went round-robin through all the lines and handed it again to someone else. Since there are 11 lines and one was 'dead'- it explains why they saw this intermittantly.

This was also the cause for the other problem I had posted; 'line rings indefinitely and auto-attendant does not pick up'. While we blamed the auto-attendant, it was not even ringing internally so neither the receptionist nor the aa would 'see' the call.

The 'dead' line was causing BOTH problems: when someone picked that line internally- they would get dead silence. When someone called that line externally- they would hear indefinite rings (as if no one was answering). I found the 'dead' line and I wanted to see if the problem was with the line itself or the phone system. I hooked a regular phone in parallel to the line and I was both able to dial out AND able to hear it ring on that phone while it was not ringing on the partner phones.

I reset the system and the 'dead' line problem went away- however this is really the worst possible solution. I would like to know what causes the Partner to freeze some lines and if it can be avoided. Thanks for your patience reading through my ordeal.

 

[itpphoneguy]

Hi foosy,
Check your other thread for my response.

-Chris

 

[Dexman]

A dead line will do it every time. (lol).

I've never heard of using a tandem as a signalling source. It is customary to use a T1 as your Timing Source, but timing and signalling are 2 completely seperate issues.

I must admit that I've never seen a Partner system freeze up a line before.

 

[itpphoneguy]

Hi Dexman,
I've seen Partners do all kinds of fun things. Station port lockup is more common, but I have seen one case of CO port lock up. I always do a reboot before condeming a card on a Partner, or if it's doing silly things.

Two common factors with this. Electrical storm and bad AC. I count high electrical noise on station wiring as electrical storm. These things tend to mess up the program in memory.

-Chris

 

[Dexman]

I found something that briefly discusses loop start and ground start signaling. It's not very deep, but it's pretty good.

http://www.silabs.com/public/documents/tpub_doc/anote/Wireline/Silicon_DAA/en/an30.pdf

 

[itpphoneguy]

Thank you Paul!
This is the sound of me reading ....

I didn't see that in my searching.

-Chris

 

[Dexman]

Hey Chris!

There is alot of other stuff mixed in there, but the "good" part is on page #3.

Paul