[VoIP] 709 Andrew Green online? (VF BW limitations)

[Steph Kerman]

For practical purposes, except for stray shunt capacitance in the bank 
wiring and cabling, the high frequency response is not limited in a 
purely passive analog transmission path within an electromechanical 
switch using the standard capacitor-inductor transmission bridge .  In a 
small switch the stray capacitance is probably insignificant and the BW 
probably goes up to at least 15KHz.  The receiver and transmitter 
responses roll off substantially below that.  You may find a response 
curve for a 500 set if you look around on line.

Low end response is limited by the value of the series capacitors 
(usually 2 uf each side) and inductance of the current feed device.  
This roll off is considered desirable to avoid propagation of 60Hz power 
line noise and its harmonics.  When passive analog trunks are connected 
in tandem, LF response may suffer due to the transmission bridge 
capacitors in each trunk circuit effectively being connected in series.  
Sometimes 4 uf capacitors are used instead of 2 uf capacitors to 
minimize this effect.  Trunk circuits may also contain repeat coils (VF 
transformers) which may also limit HF response well below the raw HF 
bandwidth of the switch itself.

Steph

Mark Rudholm wrote:
> Speaking of codecs, I've been using g722.  It goes up to 7.1kHz
> and I have to say, it sounds delicious.
>
> I suspect a standard 500 type set could go up a bit higher than 4kHz,
> but I don't know what the response curves of the transmitters and receivers
> are.  Where else in the path (in a mechanical switch where the
> voice channels are not digitized) are there likely to be frequency response
> limitations?
>
> Andrew Green wrote:
>   
>> I guess you aren't using g726 on Asterisk 1.4, I'll change the
>> settings in iax.conf to allow gsm and ulaw as well. Thanks for
>> notifying me of the problem :)
>>
>> -Andrew
>