[TowerTalk] Balun question shield leakage

[Jim Lux]

At 04:16 PM 6/14/2007, jeremy-ca wrote:


> >> Anything is possible, of course, but coax would have to be
> >> pretty awful for
> >> enough field from the differential signal to cause
> >> significant heating.
> >
> > "Unlikely" isn't strong enough, it would be impossible.
> >
> > I can buy the cheapest radio shack coax, the type that you
> > can see through the shield to the center dielectric, and it
> > is nearly impossible to measure the leakage through that
> > shield unless I use special equipment.
>
>
>
>Playing in a lab somtimes has little bearing in a real world situation. With
>a reactive load as would exist with many commercial and home brew antennas
>leakage is not only possible but very likely.

I do not think that's true.  Can you give an example and the 
mechanism by which anything connected at the end of the coax would 
change the propagation "through" the shield?

I can believe currents induced on the outside very readily.

I can believe some small amount of leakage carried on the surface of 
a loosely woven mesh (why welded mesh or perforated sheet is 
preferred for shielding purposes) particularly if the shield wires 
have a resistive coating on them so they don't make "good enough" 
contact at the junction points.

I can believe coax that is designed to be leaky (as for tunnel radio 
systems), but this has fairly large holes in the shield, and, of 
course, it only leaks at higher frequencies...(e.g. Times Microwave 
FlexRad-600 is specified for use at frequencies >150 MHz)

But none of those would have the leakage changed by what's connected 
to the coax (unless you're talking about a severe mismatch, so you 
have a lot of circulating power, so the voltage and current is higher 
than in the matched case?)  I suppose if you had 10:1 VSWR, you'd get 
100 times the leakage power?

 > The real worry, unless you bundle transmitting and receiving
> > cables next to each other for long distances, is common mode
> > currents or high resistance in shield connections.
>
>
>
>Also incorrect. A case in point on a grand scale follows.
>
>The nuclear carrier USS Carl Vinson CVN-70 had a broadband 5-450 mHz local
>area network installed while still in the shipyard. Designed by the company
>I worked for as a R&D Broadband Development Manager the installing
>contractor had a very detailed set of cable specifications.
>
><snip>
>We quickly discovered that the installing contractor ignored the specs and
>used cheap unbranded RG-11 with what looked to be around 75-80% shield
>coverage.

Hmm. that's up to 450MHz, where the holes in the shield start to be 
significant size (compared to HF) and, is it possible that some of 
the problem might also be crummy connectors and crummy installation practice?

It's also possible to make coax that would be quite leaky (e.g. 
spiral shield wires, not braid) at some frequencies.