[Antennaware] feedline modeling
n6ry at arrl.net
Thu Sep 24 10:26:15 PDT 2009
It's usually much easier and often more accurate to use the built-in
transmission line capability in EZNEC for modeling feed lines. The
exception is for cases, such as stubs in collinears or Sterba curtains,
where the physical transmission line carries common mode current - in
other words, if the line currents are not balanced (and often there is
coupling to other parts of an antenna). For coax fed antennas, you can
use the technique outlined by W7EL in his diptl.ez sample model, to
assess the impact of using an ideal balun, a practical balun, or no balun.
The built-in transmission line capability takes care of those "special
cases", such as multiples of 1/4 wl or 1/2 wl, and all lengths in
between or beyond. It also calculates the effect of line loss, which
can be very important for shorted or open lines, used for matching, or
other lines which operate at high SWR.
Your approach to modeling transmission lines with parallel wires can
work fine, but it may be necessary to use a high number of segments to
get reasonable accuracy. And as W7EL notes (thanks K9AY), the short
wire segments for the source and load can exceed the minimum
recommendations. Be sure and check how well your model works by
checking it at the frequencies where it is 1/4 wl or 1/2 wl long with
loads equal, higher, and lower in Z than the line. Then check that the
impedance at the source end is what you expect.
73, Terry N6RY
On 2009-09-24 9:15 AM, Paul Decker wrote:
> Using EZ-NEC, is it reasonable to model an open wire feed line, for example a 600 ohm ladder line using two parallel wire segments shorted at each end by two other wire segments, one end being the source, the other being the load?
> I know I can add a transmission line and change the impedance, but what about if I'm modeling the "special" transformation properties of specific lengths of transmission lines?
> If not, how is this done?
> Paul (KG7HF)
More information about the Antennaware