| Jeff McAdams on 14 Dec 2004 16:30:23 -0000 |
|
Doug Crompton wrote:
> Well it is an AC signal and it is balanced in that equal and opposing
> current flows in two adjacent wires. Why would this be any different than
> the <3.5K analog audio portion? Twisting and/or shielding helps with
> ingress and egress. If it were not balanced there would a a lot of low
> frequency RF bouncing around in those 500 pair cables out on the poles!
Actually, DSL and voice audio are not balanced signals, they each use a
single transmit and single receive signal. Balanced being, that there
are two transmit signals and two receive signals, each with the same
signal, but with opposite polarity from each other. So, you have a
positive tx and a negative tx, and they are sent on pins that twist
around each other (thus the name "twisted pair"). The signal is
measured at the far end as the difference between the two pins, rather
than the signal on a pin from ground. So, if you get an induced current
in a non-balanced signal, it will change the signal as received on the
far end (because the voltage relative to ground has changed). If you
get an induced current in a balanced signal, it will (in theory, at
least) induce the current in both wires equally, with the result that
the difference in voltage between the two wires will remain the same, so
the signal as determined by the far end remains the same.
So, with a balanced signal, like ethernet, T1, etc. you can run that
twisted pair cable right past that fluorescent light and the signal will
come out ok...telco audio and DSL, however, will get ripped apart by the
induced current from the fluorescent light. Not that I recommend
running ethernet or T1 signaling right past a fluorescent light, but
they should be much more resilient to interference like that than telco
audio and DSL are.
Oh, and the issue that you raise about telco audio and DSL running down
the bundles of copper up on the telephone poles is very real. The
problem of cross-talk in telco bundles is a very real problem and one
that is only dealt with through a rather considerable engineering effort
on the telcos' parts. Basically, they put considerable effort into
ensuring that there isn't too much current being sent down lines in
bundles given the length of the lines. This is actually the biggest
factor limiting the bandwidth that DSL can achieve...if they could pump
more current down the wires they could get more bandwidth out of it, but
they would have more cross-talk between cable pair in the bundles, so
they walk a fine line there. If you think back to V.90 and the other
PCM modulations (x2, K56flex, etc.) they were limited to 53kbps rather
than the 56kbps that they originally claimed they were going to achieve.
Its the same issue...the PCM frames that they used to achieve that last
3kbps of bandwidth ended up pushing the power in the copper wire too
high, so they limited it, in the modem "protocol" to 53kbps so that it
wouldn't go over the power limits.
> This is the argument against BPL. A totally unbalanced RF signal sitting
> on a great antenna! Crap technology!
Yeah, I don't know enough about BPL to comment on it, I assume there is
*some* sort of mechanism in it for dealing with those issues, but I
don't know what it is.
--
Jeff McAdams
"They that can give up essential liberty to obtain a
little temporary safety deserve neither liberty nor safety."
-- Benjamin Franklin
Attachment:
signature.asc ___________________________________________________________________________ Philadelphia Linux Users Group -- http://www.phillylinux.org Announcements - http://lists.phillylinux.org/mailman/listinfo/plug-announce General Discussion -- http://lists.phillylinux.org/mailman/listinfo/plug
|
|