> Tom, my experiences with RX antennas over the last few days supports the
> isolation theory. My new Pennant was showing reasonable directivity on
> in-band signals, but seemed to be overly susceptible to computer birdies
> and power line leaks. I finally decided that the feedline was the
problem.
All antennas that are non-symmetrical or non-symmetrically installed with
respect to the environment around them, and less than perfectly grounded at
the feedpoint, have this problem to a greater or lesser extent. Even if the
feedline isn't conducting noise from the shack to the antenna, it may be
contributing unwanted signals and sky-wave or ground wave propagated noise
that deteriorates directivity.
The Pennant should behave like any other antenna with a feedpoint somewhere
in limbo between balanced and unbalanced. Most antennas fit this category,
including verticals and Beverages that have less than perfect grounds. It
is a problem of varying degree, sometimes noticeable and sometimes not, but
always there to some extent.
Since my Beverages use ground rods at the feedpoint, and have very small
radial systems I isolate the feedpoints (including grounds at the
feedpoint) from the cable shield. The Pennant or an EWE would be even more
susceptible to potential problems.
> The transformer I started with was a 9:1 trifilar on a FT50-43 core.
Looked
> great on the bench for match, etc. I wound a new one with separate
windings
> with a 4 to 1 turns ratio. It looked pretty bad on the bench, but what a
> difference on the Pennant!
With a conventional xfmr, 73 material is about the optimum material I've
found (when power levels are low). Core materials are NOT in order by
number, so some seventy something cores are much less useful. 75 material,
for example, is way out of step. It is a VLF core.
The transformer requires about one or two inches of hole depth (or length
parallel to the winding INSIDE the core) with 73, and nearly twice that
with 43, for a two turn 50 ohm primary.
Impedance goes up by the square of turns, doubling turns just like
quadrupling core hole length or stacking four cores.
If a high mu at low frequencies with high loss tangent core is used at HF,
impedance can increase by LESS than the square of turns increase.
> It sure seems like feedline isolation is a critical parameter of any high
> feedpoint impedance RX antenna. Because of the very high impedance of the
> pennant and flag antennas. I suspect that the usual 1:1 "current" balun
or
> choke might not be as effective as you would want.
That's because it is a series Z shunt Z problem. A small loop has a high
shunt Z for common mode unless it is grounded in the electrical center.
That's what the shield on a loop antenna really does, even though it is
incorrectly described in many texts as an E-field shield. There is no
reason for a shield, if you decouple the feedline correctly. That sometimes
takes a LOT of series common mode Z on the feedline, as you noted. A
Faraday shield does nothing at all to help, unless it has its own
independent ground or balancing system.
> The alternative is the type of transformer that you described. In my
> experience, albeit limited, I have never had any luck with the binocular
> core transformers at HF. I am beginning to believe that the cores I have,
> which were spoda be old Indiana General "H" material, are really Q2.
The best cores I've found for this application generally reach "Q"=1 around
the lowest frequency you want to use. If you have an analyzer that directly
reads R and X, that's the frequency where R is equal to X when measuring a
wire passed through the core one or several times, whatever it takes to get
a good reading. The length inside the core from one end to another end of
the hole is much more important than cross sectional area.
73 has X=R at 2 MHz. 77 at ~1 MHz. 75 far below the BC band. 43
up somewhere in mid-SW bands.
Q2 material core X=R occurs in lower VHF, I used that core in a CATV
amplifier design.
Q2 is good for broadband VHF, not HF!
So many things go into transformer design, it's tough to give universal
answers. As a rule for a given core design and winding style, 73 material
works best in 1 to 30 MHz broadband applications. I'd be the first one to
disagree with Fair-rite corp and others if I thought they were wrong in
their recommendations.
"73" Tom
--
FAQ on WWW: http://www.contesting.com/topband.html
Submissions: topband@contesting.com
Administrative requests: topband-REQUEST@contesting.com
Problems: owner-topband@contesting.com
|