Topband: Problem with compression F connectors on Quad RG-6

[Paul Christensen]

> The standard drop cable is a bonded foil with single braid, it has been 
> that way since the late 1970's. The current best grade is "Brightwire" by 
> CommScope. Any good cable will far exceed FCC specs without a quad shield.

It remained that way with Comcast & AT&T Broadband at least until 2002 when 
I left AT&T corporate engineering, shortly after the merger.  I doubt much 
has changed since then.

Concerns are a bit different in the broadband world where downstream leakage 
must be minimized between 50 MHz - 1 GHz.  The upstream path is in HF 
region, but no sane cable operator uses spectrum below about 10 MHz.  The 
only services anywhere near that area of spectrum are used for data 
transponders and IPPV from the set-top box where modulation is almost always 
QPSK.  Telephony and DOCSIS cable modem service is all relegated to an area 
above 20 MHz.  In the Jacksonville Comcast system, it's now all above 30 MHz 
which was made possible through the use of tighter diplexer specs in the 
system amps and fiber-optic nodes.  if you could see a spectrum analyzer 
display of the return path back at the headend, you would be amazed that the 
return path works at all.  Any point of ingress results high levels of 
interference, most notably SWBC.

I recall writing up a paper in the mid-'90s that predicted a need for better 
return path certification and an upward move of the lowest usable frequency 
to well above 20 MHz.  I pointed to the 11-year sunspot cycle as an 
important driver.  The non-ham engineers in our group didn't get it.  But 
the CTO of MediaOne was a ham, and he did get it.  The public did not know 
it, but there was a real fear between 1995 and 2000 that return path 
broadband technology would never work.  When you consider all the points of 
potential failure, especially on a power-passing system, it truly is a 
miracle that it works at all.  Consider this: The typical fiber-optic node 
services between 200-500 home passings.  From the comfort of your living 
room, and with an RF signal generator, one can wipe out an entire service 
area when the return path frequencies of the system are known.  This isn't 
theoretical, I demonstrated the impact to a sober group of engineers with an 
Eico generator.  That potential still exists today.

Concerning SANS connectors and wiring, I would take the lead from the cable 
operators.  They cannot afford to have unreliable cables anywhere between 
the customer equipment and the headend or hub facilities.  When you've got 
thousands of miles of cable plant and interfacing hardware, that becomes the 
most important piece in the network.  If a router, modulator or fiber amp 
fails in the system, the fix is easy with money.  But if you deploy bad 
cable and hardware into a system, you'll feel the pain a long time as it 
affects long-term service call volume, unhappy customers, and angry 
government leaders who generally hate the cable operators.

Paul, W9AC