On Feb 11, 2:39*pm, " wrote:
First off - I reckon there is no better place than this to get honest
descriptions of what kind of poor-quality phone wiring installations
there are around the country than asking about it here!

Second off - I need to be upfront and say this is about my day job -
and not diy phone wiring.

Basically, I'm going to be testing some prototype high speed (200Mbps)
modems, and as well as test rigs that match up to standard modern
practice - I also want to build a rig that reflects poor but all-too-
common wiring too.

I'm particularly thinking about Victorian blocks of flats (like London
tenements) that were never built with wiring in mind. So the phone
wiring to the flats might be drop wires bunched together, with more
and more of them added ad-hoc as more phones were connected.

Even more so if the ad-hoc wiring happens to be crammed together with
mains wiring (although wiring regs says in shouldn't be!)

If you can spare a couple of minutes to tell me about it (particularly
if you know the wiring layout around the building) - I'm really
interested!

The only other way I can find out is to peer around blocks of flats,
tracing wiring, making sketches and getting arrested. I'd be very
grateful if you told me about it instead!

-----------

Explanation for those that want to know what's going on:

FYI - these are high speed modems that use the existing copper wire
from your house/flat to the pole (or whatever - the place known as the
CN - convergence node), where there's another box - followed by
bonding of lines back to exchange/central office).

Hence the wiring length in original copper is usually less than 100m,
hence it can support much greater speeds - provided that it doesn't
pick up masses of electrical noise.

For those not familiar with how this stuff works - phone drop wires
consist of a number of twisted pairs - i.e. two insulated wires
twisted around each other into a spiral, before being laid next to the
others and the outer cable jacket applied. This improves the noise
immunity of the wiring because each of the wires in the pair should be
affected equally. Twisted pair wiring is then used with differential
signalling i.e. the signal is the voltage difference between the two
wires in the twisted pair. This works to a degree - if the source of
interference is too close or too big, the signal can still be
excessively affected.


I would think you'd encounter all you've described, plus a great deal
of untwisted pair, plus bad connections, plus long extensions with
their capacitance and a few old fashioned inductive bells and wiring
run next to bell circuits.

I hope you're including plenty of fallback modes, reduction in
throughput is far less likely to cause contract end or dispute.
Failure to implement such fallback modes is a real headache for some
people, and ties up a lot of manhours and creates ill will.


NT