[TowerTalk] Balun question shield leakage

[Roger Parsons]

The leakage from a coaxial cable is only indirectly
related to the braid coverage and/or 'hole size'. It
does not leak out of the holes like water through a
sieve! What happens is that the coaxial mode signal is
continually transferred to a single wire mode signal
on the outer of the coax over a considerable distance
- the so-called G-wave. This happens even when the
coax has a 100% shield that is 100% conductive.

The standard test for measuring transfer impedance is
highly flawed as it uses a 1m length of cable with a
closely spaced probe. At less than high vhf
frequencies this is far too short for the single wire
mode signal to emerge and be properly measured.

The only method of which I am aware to correctly
measure transfer impedance (and hence leakage and 
coupling to another cable) is to set up a long length
of cable in a tunnel with non-conductive walls and
take a very large number of fs measurements. Even this
test is complicated when the frequency is above the
waveguide mode cut off frequency of the tunnel -
perfect measurements could only be done in space!

Some quite interesting facts emerge:

The relationship between braid coverage and leakage is
by no means simplistic. It is not at all true to say
that 'less braid = more leakage'. Some of the very
cheaply made cables with low braid coverage have quite
good shielding, although most don't. Some cables with
high braid coverage have relatively high leakage;

It is quite possible to design a cable with less
longitudinal attenuation and more leakage. The
flexible leaky feeder developed by British Coal
(T3529), which is pretty much the world standard for
this lype of cable, was originally based on URM57
(RG11), and has slightly lower longitudinal
attenuation and about 60dB more leakage than that
cable;

Discontinuities in the coaxial mode do not in
themselves create more leakage, but set up reflected
waves that make a complex situation more complex. Of
course, as others have mentioned, a bad connection in
the cable outer will change the cable into an antenna;

Discontinuities in the single wire mode have a
dramatic effect, just as they do with a wire antenna.

Transfer impedance remains constant with frequency
within the limits of the cable design.

A short length of badly shielded cable at hf will not
cause measurable leakage.

It is possible to deterministically calculate transfer
impedance for any given cable, but it is quite a
complicated sum involved multiple differential
equations.

73 Roger
VE3ZI

Former Head of Communications Group, British Coal
Mining Research and Development Establishment.




		
___________________________________________________________ 
Inbox full of unwanted email? Get leading protection and 1GB storage with All New Yahoo! Mail. http://uk.docs.yahoo.com/nowyoucan.html