Voip + Voip = Cableone.net + C*Net

[Babybearjs]

well, I finally did it.... switched over to cable one for my telephone and internet. Centurylink can't provide me with any faster speed then 20GB. with cableone, I have 100GB. Plus, the sound quality is soooo much better. I was surprised. My entire system is working just fine and the phones ring with no problem. I can add a second line for just an additional $20/Month and I think the taxes are lower.... (but won't know until Dec.) who else has had a good experience with changing over to a cable based phone system? the only downside to all of this is that the Voip System DOESN'T RESET, so if you put a call on hold, and its abandoned.... the hold circuit continues to keep the line tied up. I noticed that with the C*net system, and now the Cableone System.... is there a way around it? who else has had this issue....

[unbeldi]

well, I finally did it.... switched over to cable one for my telephone and internet. Centurylink can't provide me with any faster speed then 20GB. with cableone, I have 100GB.

I think you mean Megabit/s (Mbit/s), 1000-times slower.

[unbeldi]

... the only downside to all of this is that the Voip System DOESN'T RESET, so if you put a call on hold, and its abandoned.... the hold circuit continues to keep the line tied up. I noticed that with the C*net system, and now the Cableone System.... is there a way around it? who else has had this issue....

It should not "reset".  The call on hold should either be hung up automatically, or the adapter should briefly ring the telephone to alert the user that a call is on-hold.  That has been the proper protocol for a long time, but perhaps many of today's ATAs or eMTAs neglect that.  I would complain.
Hook flash should get you connected to the call, or sometimes a star code does.

[poplar1]

OK, I'll admit I know next to nothing about these fake dial tone producers. However, on a landline, a holding bridge from a 1A2 key system appears to the CO as an off-hook condition. If a line is off-hook, how can the c.o. send ringing current over the line, since the line is already tripped and still off-hook?

On CO line, in the case of a held call, if the party on the other end disconnects, the loop toward the held line is momentarily opened, which causes the hold relay to release and the line to go on-hook, ready for another incoming call.

John, you might want to read up on 100 Key System and see how WE handled locked in ringing signal in order to prevent ringing to continue all night after a caller abandoned the call. I don't recall now whether they also dealt with abandoned calls on hold.

[unbeldi]

In a VoIP system, the local loop only exists between the ATA and the telephone set, so sending a ring splash is easy to do when the user erroneously hangs up and abandons the call on hold.  IP phones may have a visual indicator instead, but a plain old analog phone has only one way for alerting.

Key systems at least had visual indicators, the lamps, and these were also controlled locally by equipment at the customer premises, not remotely. So this is not very different actually. Key systems also don't use the ringing signal from the central office to alert the stations, they only use it as a trigger and generate alerting locally.

This is taking the concept of common control one step further in key systems and yet another step in voice over IP where no more knowledge of the physical method of alerting is needed by the switching system, it's just a digital command.

[unbeldi]

VoIP endpoints, as well as the servers at the softswitch, automatically monitor the exchange of speech on the connection.  The standard transmission protocol for speech in form of packet voice is the Real-Time Transport Protocol (RTP), when no speech is exchanged, no packets flow, absent other features (comfort noise), and both ends of the connection have an opportunity to terminate the call when the configuration rules determine that a call was ended or should be ended.

[poplar1]

What I was saying is that a c.o. line has to be idle before ringing current can be applied. By definition, then, the line is not (no longer) on hold.

So are you saying that with an ATA, the ring splash can be sent from the ATA to an analog phone, all while maintaining the line on hold? I do remember that electronic key systems can do this; I think it's known as "hold recall." The installer can change the length of time before an audible reminder is sent to the station that put the call on hold, or in the case of the Tie 616, before it rings every phone in the system.

1A2 Key Systems maintain common audible ringing and lamp flashing during the "silent interval", usually 4 seconds between the 2 second rings, until the A lead is grounded, which interrupts the ringing and lamp flash even if the interrupter motor has not yet returned to "home" position.

[unbeldi]

Oh, I am beginning to understand the issue...

John is connecting a key system in front of the VoIP adapter and having a problem with disconnects.

Check out whether there are configuration options for modifying the critical time for flash signals and hangup supervision.  The line wink of the key system may be incompatible with those settings.

Another issue is that most ATAs have the ringing voltage floating between tip and ring, not referenced to ground.  It was rather late when WECo started producing 1A2 line cards that could detect ringing on floating lines.

[poplar1]

VoIP endpoints, as well as the servers at the softswitch, automatically monitor the exchange of speech on the connection.  The standard transmission protocol for speech in form of packet voice is the Real-Time Transport Protocol (RTP), when no speech is exchanged, no packets flow, absent other features (comfort noise), and both ends of the connection have an opportunity to terminate the call when the configuration rules determine that a call was ended or should be ended.

Can it distinguish speech from music-on-hold? Even if the MOH is actually an announcement on a broadcast radio station?

[unbeldi]

What I was saying is that a c.o. line has to be idle before ringing current can be applied. By definition, then, the line is not (no longer) on hold.

So are you saying that with an ATA, the ring splash can be sent from the ATA to an analog phone, all while maintaining the line on hold?

Yes, what happens between the phone and the ATA is rather independent from the network signaling.  It is rather unfortunate that the consumer market is moving so slowly in terms of endpoint devices,  I don't know of a single cable company that permits the use of true IP phones.  Other ITSPs have provided this feature almost since the industry started and IP phones became available. For business systems this is the rule by now, for a while actually.  Using ATAs there (usually called media gateways in that realm) has only been a method to leverage existing telephones as a transition period to save cost.

I do remember that electronic key systems can do this; I think it's known as "hold recall."

IIRC, the 1980s Panasonic 616 systems also have this.

[unbeldi]

Can it distinguish speech from music-on-hold? Even if the MOH is actually an announcement on a broadcast radio station?

Not if it is generated remotely on the same connection.  In that case it is just like speech, except perhaps sounding nicer.
But often in VoIP systems, MOH is actually played from the ATA as well, possibly via a separate network connection to a special MOH server, which is specialized for that service. The audio path to the user is simply switched over to the MOH connection.   VoIP systems design is increasingly distributed, so that components can be switched in and out at will, and can be geographically dispersed to optimize w/r/t network bandwidth, distance, etc.

[Babybearjs]

the way I used to test the system was to put a line on hold and wait for it to end. with the landline, it would timeout just fine, but with the voip it will flutter, and stay on hold. when the line is picked up, all you hear is a busy signal.... so with my 1A1 there is no way to correct this issue. it may be time to upgrade my system to a newer format.... I do have a 551C KTU on standby.... do you think that should be used? or would the problem continue.

[unbeldi]

the way I used to test the system was to put a line on hold and wait for it to end. with the landline, it would timeout just fine,

That is not the proper way to end a telephone call.  You are relying on a central office fall-back mechanism to resolve a line being out of order, or not being hung up properly.

but with the voip it will flutter, and stay on hold.

I don't understand what that means, it will flutter.  Why would it work differently on VoIP?  Do you mean the hook flash, that is generated by the line card?
Indeed, a hook flash signal is commonly recognized to place a connection on-hold, if you have three-way calling service or call waiting, which is typically provided for free with VoIP packages.   Typically, these services can be turned off by dialing a vertical service code, either on per-call basis or persistently. Consult your providers documentation to turn them off.  Turning these services off, should provide a clean line that is very close to a dumb POTS line, and permit you to hangup the line cleanly.

when the line is picked up, all you hear is a busy signal....

More likely you are hearing the reorder signal, not a busy signal.

so with my 1A1 there is no way to correct this issue. it may be time to upgrade my system to a newer format.... I do have a 551C KTU on standby.... do you think that should be used? or would the problem continue.

You need to properly diagnose the problem and understand how the key system interacts with the features of the ATA.  The typical ATA today is far more complex and feature rich that a POTS line.

PS: Connect a phone directly to the ATA and test the features of the ATA or eMTA, especially the hook flash operation.
Find out whether the device has a configuration web page where features can be customized.
Which kind of device is it?   Is it embedded into the CATV modem or VDSL modem, or a separate ATA?

[Babybearjs]

I was referring to testing only. I'd pickup the line, then put it on hold....because no call was in session, it would timeout and the wink cycle would end.