ve2pid wrote:
Hi to all

In a document from Roy Lewallen, we read this:

When a balanced antenna is fed with coaxial cable (Fig.1),
the outside of the shield appears as an extra, separate
conductor connected to the right side of the antenna at the
feed point. (page 157. What They Do And How They Do lt)

That's _Baluns: What They Do and How They Do It_, in the _ARRL Antenna
Compendium_, Vol. 1, and also at
http://eznec.com/Amateur/Articles/Baluns.pdf.

My question is this: Can we say the same thing for a dummy load (i.e
can we replace the words 'balanced antenna' with 'dummy load'?) so the
dummy load would 'see' three conductors?

Not if it's a good dummy load. A good dummy load is shielded so that all
the current from the center conductor stays on the inside of the shield.
This current then returns on the inside of the coax shield.

We are trying to explain why some dummy load/Tx line systems radiate.
Perhaps, tthe current outside the coax (common mode current) exits,
while being very small, so the radiation could come from a leaking
coax... ?

Ok, let's look at a poorly designed dummy load, say a 50 ohm resistor
connected with pig tails to the coax feedline. Remember that the current
on the outside of the inner conductor is *always* equal and opposite in
phase to the current on the inside of the outer conductor. Let's call
that current I1. So the lead connected from the coax center conductor to
the resistor has I1 flowing. If the pigtails are very short with respect
to a wavelength, the current in the whole coax - pigtail - resistor -
pigtail - coax loop is essentially the same, I1. So where the other
pigtail connects to the coax shield, it carries I1 which all has to flow
on the inside of the shield (so that the current on the inside of the
shield equals the current on the center conductor).

But a dummy load radiating 10 uW, say, would drive a nearby receiver
bonkers. If you put 100 watts into the dummy load, that means only
1/10,000,000 of the applied power has to radiate, and this can be hard
to prevent. The pigtail - resistor - pigtail loop length isn't exactly
zero, so there will be a little difference in the current at one end
than the other. And this difference current will end up on the outside
of the coax shield. In addition, there's probably some small amount of
coupling from the resistor and pigtails to ground, the feedline, and
other conductive objects. This will result in a displacement current
that causes current imbalance and consequent current flow on the outside
of the feedline. The only way to prevent this is to completely shield
the dummy load and transition to the feedline, and have no seams, holes,
or cracks which could let any current get from the inside to the outside.

I've had limited experience with leaky coax, but have found a
substantial difference between single and double shielded cable in the
amount of signal radiated. Connector assembly is also critical -- it has
to be done just right, or you'll end up with one of those sneak paths
from the inside to the outside, regardless of how good the shield is.
You usually won't have to deal with this, though, unless you need a
pretty high value of attenuation.

Roy Lewallen, W7EL