[TowerTalk] Coaxial Moxon
steve at karinya.net
Tue Aug 5 03:54:21 EDT 2008
All that Dan says, plus......
...... the Reflector has the Inner and Braid shorted at both ends. There
is no way for differential-mode currents to be excited in the Reflector
coax so there can be no "velocity-factor shortening". The only thing
that will cause any shortening (maybe 5%) is the insulating jacket of
the coax. By cutting the Reflector to 66% of normal size, but with no
velocity factor effect to compensate, the Reflector resonance will be
way high and it will not operate like a Moxon.
This is not just my "theorising" - I measured these things in my search
for a broadband Hexbeam Reflector. Using RG58 did just as I expected - a
slight shortening due to the jacket, and slightly more bandwidth because
it was "fatter".
Even the "inventor" can't decide how the Reflector should be connected -
compare the HamUniverse diagram with the same antenna on the Moxon site:
Finally - lumped component loading at the feedpoint produces much better
results - you can build a coil with less loss than the stub without
really trying! You also get better bandwidth because the coil reactance
doesn't shift as quickly with frequency as the stub does. You'll
sacrifice the convenience of having the inductive loading integral with
the dipole arms - on the other hand you don't have heavy coax to contend
with. Oh, and the SWR will get worse - a sure sign that the antenna is
By all means build one if you want to learn about antennas. As many
others have done, you'll quickly learn that this one doesn't work!
Dan Zimmerman N3OX wrote:
>> I have looked at the pics and dont see why the
>> antenna would not work.
> It does work, in a way.
> The radiating elements are the *outside* of the coax shield.
> The *inside* of the coax shield and the center conductor form a coaxial stub
> that's in series with the outside.
> Imagine building a dipole from aluminum tubing that's 66% normal length.
> Now, you want to put some pretty big loading inductors at the feedpoint to
> get it to resonate, and you don't feel like winding coils, so you take some
> coax and make a couple of shorted stubs about 1/4 wave long (which have very
> high, resistive impedance) and stick them in *series* between either side of
> your feedpoint and the aluminum elements. The shield of the coax is
> connected to the aluminum element side and the center conductor is
> connected to the feedpoint side. Exactly 1/4 wave stubs are no good, very
> high impedance, but if you trim just a little bit off those coax stubs,
> they'll go off resonance in the inductive direction.
> At some point, you've trimmed the stubs so they show an inductive reactance
> that just cancels the short aluminum elements' capacitive reactance and you
> have a resonant antenna... but you've got these dumb stubs dangling down
> near the feedpoint. So you stuff them inside the aluminum antenna. Well,
> the space between the outside of the coax stub shield and the inside of the
> aluminum element shield has no electromagnetic field inside of it, it's
> shielded from the fields inside the coax and it's shielded from the fields
> outside the aluminum element, so you just magically fuse the outside of the
> coax shield and the inside of the aluminum tube together with metal to keep
> the wasps out.
> Now you've used magic and a lot of work to come to the realization that the
> *outside surface of the aluminum tube radiating element* and the *coax stub
> outer shield's inner surface* could have been two sides of the same pipe to
> start with.
> So you start the next one with coax cable as *both* the inner stub and the
> outer element instead... same thing as a "velocity factor" dipole. The
> trick is that if you're just a little short but *close to resonance* with
> your stub, you can find *any value* of inductive reactance you want, so you
> discover that there's always a little tiny bit of trimming you can do at the
> end of an "electrical quarter wave" worth of coax to make it resonant.
> But it's a terrible way to load an antenna. It's very, very lossy.
> Give it a try if you want, but objectively measure it against a reference
> dipole antenna for *gain* ...
> TowerTalk mailing list
> TowerTalk at contesting.com
More information about the TowerTalk