[Top] [All Lists]

Re: [TowerTalk] Balun Recommendation

Subject: Re: [TowerTalk] Balun Recommendation
From: Ian White GM3SEK <>
Reply-to: Ian White GM3SEK <>
Date: Sat, 21 Apr 2012 10:17:24 +0100
List-post: <">>
Gerald wrote:
>Hi Ian and others:
>Is common mode current the same as rf current flowing on the coax
>shield because it is connected to one half of the antenna?

Broadly speaking, yes...

For coaxial feedlines, "common mode current" is the name applied [1] to 
RF current flowing on the outside of the shield, regardless of the 
reason. The outside of the shield then behaves as an unwanted 
single-wire antenna, connected to one side of the antenna you thought 
you had. If the CM current is large enough, this can lead to all the 
well-known problems of RFI and increased noise levels.

But let's back up a little, so that everyone can join in at the same 
point. First recall that, for RF, the skin effect makes the inside and 
the outside of the shield behave as two entirely separate conductors. 
This means that coax is actually a THREE-conductor transmission line!

For "TowerTalk purposes" the coax shield can be considered completely 
solid and impenetrable to RF, so the only place where current can flow 
between the inside and outside surfaces is at the open end - at the 
antenna feedpoint.

The most obvious cause of CM current on the outside of the coax is the 
one that you identified: one side of the antenna has an exclusive 
connection to the center conductor of the coax; but the other side of 
the antenna is connected to both the inside of the shield AND the 
outside. This unavoidably creates a launching-point for CM current to 
flow down the outside of the shield... unless you DO SOMETHING to 
prevent it.

Electromagnetic (radiated) coupling between the antenna and the feedline 
is another cause of CM current, but this is generally less severe except 
for pathological cases involving a grossly asymmetrical layout of 
antenna, feedline and ground, and/or a resonant length of feedline.

>If so, no wonder the feed point is unbalanced; there are two return
>current paths for any voltage present on the shield side of the antenna.
>The "hot" side of the antenna reflection sees the only the coax center
>conductor and associated impedance. The other side of the antenna
>reflections sees the inside of the coax and associated impedance in
>parallel with the outside of the coax and it's associated impedance. The
>total impedance then has to be less than the "hot" side.
>So, there is a balanced antenna feedpoint with say 50 ohms referenced
>to ground on one side and maybe 35 ohms on the other side reference
>to ground. The voltage would have to divide 35/50 in that case.
>Providing a means to reduce the current on the shield has to improve
>the voltage ratio on the two halves of the antenna as that makes the
>impedance referenced to ground more nearly equal. Series choke
>losses also have a parallel equivalent that would tend to degrade the
>balance. So, it seems the choke can help balance to a large degree but
>never completely restore balance.

That last point is correct. If we choke off the CM feedline current at 
the antenna feedpoint, it will equalize the currents at the two 
terminals of the antenna. But a choke at the feedpoint cannot enforce 
symmetry across the entire antenna, and it obviously cannot "cure" any 
asymmetry in the surrounding environment. Something has to give, so the 
current distribution across the whole antenna has to change: the CM 
choke forces the current to become more symmetrical close to the 
feedpoint but the current distribution remains quite asymmetrical 
further out.

One of the biggest fallacies in ham radio is the "symmetrical balanced 
antenna". That foolish notion bears as much resemblance to reality as, 
say, a pencil remaining forever balanced on its point. It's one of those 
situations that you can imagine easily enough, but in real life it never 
happens because the real world always tends to create UNbalance [2].

So in real life, ALL antennas are asymmetrical. The only relevant 
questions are "How much?" and "How much do I care?"

At the one extreme we have obviously asymmetrical antennas like OCFs and 
verticals close to ground, which will have strong CM feedline currents 
(a large fraction of the wanted antenna current) unless we use feedline 
chokes to prevent it. As we well know, some of these extreme cases can 
be very tough to cure, and may require very high impedance chokes in 
more than one location to make any significant impact on the RFI 

At the opposite extreme, we have horizontal yagis on high towers where 
both the antenna and its environment are pretty close to symmetrical. CM 
current on the feedline will usually be a much smaller fraction of the 
wanted antenna currents, so these antennas present a much "softer" type 
of problem which can often be handled by quite inferior CM chokes like 
the random coil of coax.

>Any corrections to this understanding are cheerfully accepted.

Well, not so much "corrections" as filling out the broader picture and 
filling in the details.

Even more details:

[1] The concept of "common mode current" was originally developed for 
twin-wire lines, on which signals are carried by equal-and-opposite 
differential currents. Any inequality or phase error between the 
currents on the two wires can be represented as a separate current 
flowing equally and in phase on BOTH wires - which is what the word 
"common" was originally meant to imply. In other words, common-mode 
current makes a twin-wire line behave like a single wire. Contrary to 
popular belief [2], both the differential (transmission-line signaling) 
currents and the common-mode currents will coexist freely on the same 
twin-wire line.

Engineers have borrowed this label "common mode" from twin-wire 
transmission lines, and have also applied it to any current flowing on 
the outside surface of a coax shield, for any reason. This isn't 
accurate usage because the geometry of coax is completely different from 
twin-wire, effectively giving us three conductors instead of two... but 
RF engineers seem to revel in twisting the English language until it 
points newcomers in entirely the wrong direction.

What engineers don't appreciate is that twisted language turns around 
and bites you! What comes out of your mouth goes right back into your 
mind and muddles your own thinking too (see "ERP", "gain", "ground" etc 
:-)   Worst of all is that students and hams have to dig down several 
technical levels before they can begin to understand that their mentors 
are often saying one thing but meaning quite another... and often 
without even realising it.

[2]  "Balanced feedline" is equally mythical. It's another of those 
things that is easy to imagine but never occurs in real life. Parallel 
line is not self-balancing - the differential mode coupling between the 
two wires is quite weak compared with the unbalancing effects of poor 
installation and asymmetrical connections at its two ends. Therefore you 
cannot buy "balanced feedline" straight off the drum, no matter what the 
label may say!  When you install it, YOU have to create the conditions 
that will make it balanced... and good luck with that too...


73 from Ian GM3SEK

TowerTalk mailing list

<Prev in Thread] Current Thread [Next in Thread>