[Gllug] ADSL on BT phone wires

[Christopher Hunter]

On Wednesday 09 Aug 2006 10:40, Chris Bell wrote:

>    The ADSL must be tapped off using at least one capacitor, rather than
> just a transformer which would act as a DC short circuit, and I assumed
> that non-electrolytic capacitors would be physically larger but more
> reliable. It then depends on whether the signal polarity really matters,
> and that would require deeper knowledge of the system.
>    The problem occurred on an existing phone line that was upgraded for
> testing connections to AOL, not the main BT ADSL, and is connected via an
> extension. It worked at one end of the extension, not the other end. I was
> not involved, but am suspicious about the explanation given.

You've been fed nonsense from a BT "engineer" that doesn't know a thing about 
ADSL (generally they don't - the standard of "training" of the average BT 
monkey is laughable these days).

A BT line has approx 48 Volts sitting across it when quiescent - 80 Volts AC 
is applied to signal ringing.   A capacitor in the "master" socket (typically 
1.8 uF  at 250 Volts) is used to send the ringing (A.C.) voltage to the 
bells, bleepers, flashing lights or opto-couplers that signal the arrival of 
a call.  

When the line is busied, the D.C. voltage drops to about 8 to 10 Volts (a 
telephone, answering machine or whatever loads the line with roughly 300 
ohms) and the current flow is used at the exchange to show "in use". 

Instrument polarity is entirely immaterial - the equipment is designed to be 
polarity insensitive, and would fail approval if there was any polarity 
sensitivity.   (I used to design telephone equipment for a living!).  
However, it is necessary that each of the three wires used at each socket are 
connected to the right terminals.  They made it as easy as possible - 
terminal 2 connects to all the other terminal 2's, terminal 5 connects to all 
terminal 5's, and the same with terminal 3.

You probably had a dodgy IDC connection, but the monkey tried to baffle you 
with BS because he has to try and justify he silly amounts the charge for a 
call out!

ADSL gear just appears across the line as a high impedance A.C. load / source, 
and is capacitively coupled to the line.  Polarity has no effect, and "phase" 
is insignificant - the phase shifts down a BT line at the frequencies used 
for ADSL will be horrendous, and are compensated for in the PQSK modulation 
scheme. At worst they may (marginally) reduce data transmission speed, but 
will (in reality) have no noticeable effect, and will be a characteristic of 
the twisted pair from your place to the exchange rather than anything in your 
home!
Chris

-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug