Any other FXO solutions?

[cgallery]

I have (I guess) a v1.0 Grandstream 4104 gateway. I have the latest firmware for it, yet the caller ID is only 50-50 (sometimes I get nothing). As others have complained about this, I assume it is just a Grandstream issue.

The other items supported according to the 3CX page are a Patton 4114 and the Sangoma card.

Are there any other tested FXO gateways people here have used? I'd sort of like to keep to an Ethernet (external, not a card) solution, if possible. But I have tried the Grandstream, and the only other choice is the Patton. So looking for more options.

 

[eagle2]

Cisco/Linksys SPA-3102 (1x FXO + 1x FXS ports) or SPA-8800 (4x FXO + 4x FXS) is a good option.

However the problem with Caller-ID is insufficient delay before answering the call - the Caller-ID is transmitted between the first and the second ring, so waiting 3 seconds before answering the line is usually safe to obtain the Caller-ID.

Regards

 

[cgallery]

Thanks for the feedback.

I currently am "sharing" the lines with an older Nortel Norstar phone system that doesn't do caller ID. I'm doing this as I test and learn about 3CX.

The Norstar phones always ring first. There is a slight delay (one ring) and then the 3CX phones ring. But the caller ID is often missing.

I always assumed that having the Nortel connected at the same time was okay because one can have multiple "simple" phones connected to a line and get caller ID (I'm right on that, aren't I? I don't know much about caller ID, I've never had it before). But I should test by disconnecting the Nortel from the lines and test calling them, to see if it somehow interferes with the ability of the Grandstream to get the caller ID.

This is interesting stuff.

 

[leejor]

Having multiple sets connected to a line might cause the caller ID data to come thorough at a lower level (higher capacitance) and to not be recognised at times on the FXO, but this usually isn't the case. A simple test is to connect a stand alone caller ID box across the line, see if the caller ID shows up each time on that. If it does, then you know that you have an issue with the Grandstream. As eagle2 pointed out, there needs to be a sufficient answer delay (3 to 4 seconds) before a call is presented to 3CX to ensure that CID info has been captured successfully.

 

[lneblett]

I recently installed a 4104 on AT&T lines (2) and I have not experienced a problem. I do notice that caller-I'd is not instaneous, you have to wait for at least a ring, but it always comes through. The same delay is evident on my PSTN cordless AT&T phones at the house, so I do not believe this to be a Grandstream issue. I had never used GS before, but I accidentally plugged in the wrong transformer which blew a couple of electrolytic caps. I replaced the caps and the device functioned perfectly. not sure that I helped, but I do not have the issue.

 

[eagle2]

The Caller-ID should appear after the first rring and before the second.
The time between rings is 5 seconds (standard). If the line is very long or have many loads (phones attached to it) the second and further rings may be delayed.

To get the caller-ID it is normally safe to omit the first ring and wait at least 3 seconds or more until answering the call. This is easy to program in SPA-3102, should be possible with Grandstream too.

Regards

 

[leejor]

If the line is very long or have many loads (phones attached to it) the second and further rings may be delayed.

Long lines, or too many phones (high ringer load) won't delay a ring. It may cause the ringing voltage to be much lower than usual, especially long cable pair length (distance that you are from a central office). Too many ringers may cause a ring trip or ringing overload (being detected at the central office), in which case , you may not get any (proper) ring at all.

As far as the PSTN answer delay...I have never had any problems, with CID delivery, using 3 seconds as a setting, on the SPA3102 (this is for North America, other regions may vary).

 

[eagle2]

Hi Leejor,

probably you are right for ring delay - no reason for this.
I've experienced such delays (up to 10 seconds) on long lines or with several REN with old public exchanges -- could be a side effect like the loop resistance is influencing the period. I've also seen varying intervals between different rings. However, this is not important. I mentioned this only because if there is a ring delay, 3 seconds may be not enough.

3 seconds delay for SPA-3102 is OK also in Europe. I've tested this in most countries (I have some Pan-European customers). Not sure about France, Spain, Portugal.

Regards.

 

[leejor]

3 seconds delay for SPA-3102 is OK also in Europe. I've tested this in most countries (I have some Pan-European customers). Not sure about France, Spain, Portugal.

I haven't had "real life" experience using CID devices outside of North America so I can't comment on what works elsewhere. I do know that there are a number of different timing and delivery methods used in many countries, and even several within each country, depending on the provider of your service.

I've experienced such delays (up to 10 seconds) on long lines or with several REN with old public exchanges -- could be a side effect like the loop resistance is influencing the period.

Ringing cadence is determined by the central office equipment and the make of switch being used will cause different results. It can also be influenced by the ringing load at a particular time. If a lot of phones are being rung at one time, (some event causes a lot of calls), some sets may experience a delay so as to not overload the ringing generators. This is more common in smaller offices where load balancing of lines is more difficult because of the limited number of subscriber terminations. This can also happen where a neighbourhood, district, or building, is served by a "remote" from a larger central office.
Rural areas can be a different matter, and present another set of problems, altogether.