[Asterisk-Users] Asterisk - fax - spandsp

[tim panton]

On 15 May 2005, at 17:26, Steve Underwood wrote:



> Michael Welter wrote:
>
>
>
>
>>
>>
>>
>>
>>> What is in the bug tracker helps make things clearer to people  
>>> who know what they are doing. What we need is something that  
>>> makes things clear to laymen. Saying internally and externally  
>>> clocked doesn't cut it. It needs to be made clear to laymen that  
>>> clocking internally is rarely right; that slaving to a PBX, when  
>>> there is also a PSTN connection is almost never right; that the  
>>> PSTN is the grand master of the telecoms universe, and all  
>>> Asterisk boxes should be its slave.
>>>
>>>
>>>
>>>
>>
>> Maybe the confusion is because some pieces of the explanation are  
>> being left out.  If you look at the leads on a synchronous RS232  
>> interface it is very clear:  Rx clock and Tx clock, and people  
>> understand that.  It's not as clear-cut with T1/E1.
>>
>> I'm not the person to explain this, but someone (Rich Adamson  
>> perhaps, or a Digium engineer) needs to explain phase-locked loops  
>> and the recovery of the timing signal from the data stream.   
>> Perhaps understanding that the timing comes from the data stream  
>> itself might lead to better understanding.
>>
>>
>>
>
> Put the term PLL into an explanation, and 99% of the people reading  
> will switch off.
>
>

I think of it in terms of cost of hardware. The most expensive kit  
generally has the
best clock source.

Make a list of all the devices that are talking to your T1/E1 cards,
rank them in order of price - most expensive first.

99 times out of a 100 the one at the top should be use as the   
primary clock,
the next one down as secondary etc...



>
> Regards,
> Steve
>
>
>

Tim.