[RFI] U-Verse problem update

[Jim Brown]

On 2/21/2011 7:22 AM, Martin Ewing wrote:
> Older, cheaper
> coax is much less well shielded and permits both ingress and "outgress" of
> signals.  (Old connectors are weak points, too.) There's no reason Cat-5/6
> ought to be better than well shielded coax*.

These statements assume that the coupling mechanism is DIFFERENTIAL 
mode.  But it very well may be a COMMON MODE problem, coupling via a Pin 
One Problem, which could explain why using CAT6 made things better.

Another point with regard to shielding.  Shielding effectiveness is a 
function of RESISTANCE and of the UNIFORMITY of the shield. Yes, quad 
shields help with uniformity, but all the coax I've seen that is made 
for CATV uses very light weight shields with fairly high resistance. 
Those shields are effective at VHF and UHF, for which those cables are 
optimized, but they aren't great at MF and low HF, where that cable 
modem is working.

And finally, it is extremely '40s (that is, 1940s) to describe coax 
using RG numbers. There are dozens of RG59s in the Belden book, all very 
different from each other.  Different center conductors, different 
shields, different insulation, different jackets, and so on. The best 
COAX for use in this application would be one that has a BEEFY COPPER 
BRAID shield to provide low resistance, and other characteristics to 
maintain the needed bandwidth for TV signals up to 1 GHz.  Does that 
sound like LMR400 or Commscope 3227?

There's a tutorial Q&A on coax and stubs on my website.

http://audiosystemsgroup.com/publish.htm

73, Jim K9YC