This is actually a fair amount of work to come up with 6 models and make sure
they are "clean" at the frequency used and represent specific components.
Also you will find that generalities do not always map well to individual
Modeling feedlines correctly is only exceeded in complexity and gotcha's by
modeling dirt. Here's a how to for a dipole fed with coax.
To model a coax feed for the *essential* effect (not exact) break the dipole
into two wires. Run a vertical wire down from the junction to represent the
coax. Just use #12 for everything to start with. To begin, set the segments to
be 1 segment per foot everywhere and wires even foot lengths. Only change to
non-multiples of a foot and mixing shorter and longer segments when you can
see if it changes anything vs equal. If it changes, take it back to equal
Break the coax wire up as needed to show sloping, running along the ground,
etc. Where the coax is laying on the ground or buried, place the wire .01
feet above ground. Take it all the way to the house. Where the coax is
grounded (as where your coax enters the house?) run a one foot wire from .01
ft above ground to 0.0 height, the length equal to one segment of the wire
coming to it.
Place the dipole source in either dipole half in the SECOND segment from the
center junction. This avoids problems from placing a source in a segment
adjacent to a complex junction. You can play with the % from end to make this
happen. This will not be an exact dipole impedance point, but it WILL place
power in the dipole, show the resultant conductor currents to a couple
significant digits, and allow you to move placements and view conductor
It FREQUENTLY happens in explicitly modeled dipoles INCLUDING nearby
conductors, and slopes, that having the shield on one side or another of the
center insulator makes a big difference. Flip the coax shield by moving the
source to the other side of the junction.
If the land is sloped, model the situation to place the slope in the
conductors, as if you had tilted the land to make it level and left the wires
as-is. Sloping land unbalances a dipole.
Wires coming into complex junctions need to have the same segment length and
diameter at the junction.
To put a choke on the coax add a load into any segment (except next to the
complex junction at the dipole) and set the load to produce the blocking
impedance specified for the particular device at the frequency. If you don't
have that information, you will have to go MEASURE something, or pick a
reasonable value that COULD be constructed. There is NO one value for a choke,
only what is constructed/used. You can put multiple chokes in various places
on the wire representing the coax and see what it REALLY takes to get common
mode current off the shield. Having a single choke at the feed point only
lessens common mode noise getting into the feed
If this doesn't seem like enough information, you have a sense of how goofy it
can be to model this stuff.
Once you have this done, start adding passive conductors for your tower, other
antennas, nearby metallics like gutters and downspouts. Watch values change as
you add them.
----- Original Message -----
From: "Jim Miller" <JimMiller@STL-OnLine.Net>
To: "AntennaWare Reflector" <AntennaWare@Contesting.com>
Sent: Wednesday, April 16, 2008 9:44 PM
Subject: [Antennaware] modeling feedlines- coils - traps
My current thought obsession :
Here is your assignment for next week : L.B., where is this in your modeling
I guess I am looking for an example of a simple model with the feedline
modeled. Is there one in the examples that come with eznec? Name?
Can somebody send my a simple dipole with :
1) a coax direct fed,
2) with a 1x1 balun,
3) with window line
And a model with a coil :
4) at the feedpoint
5) somewhere out toward the middle
6) Also something like a trap dipole for 40 and 80 for instance (showing the
Thanks es 73, de Jim KG0KP
Antennaware mailing list
Antennaware mailing list