Re: NEC-LIST: Current source in NEC-2

Adam Turowski <aturowsk_at_elka.pw.edu.pl> wrote:
> ... I have tried to model antenna phase array [but] I couldn't get the
>right radiation pattern (a kardioid). I do the same modelling in ELNEC
>demo version and the pattern was right if the sources was the current
>sources. This result is obvious, but the question is : How could I do
>the same in NEC-2 ? (manual says that this is directly imposible, but
>maybe there is a trick )

One way to make a current source in NEC-4 or (I suppose) in NEC-2, is
to put a voltage source in a very short segment (so short that it
doesn't radiate significantly) located out in left field, and to
connect a one-quarter wavelength long transmission line from this
segment to the feedpoint of your antenna where you need a current
source. A quarter-wave line inverts impedance, so a voltage source V
becomes a current source V/Zo. This method is frequency-specific.

A frequency-independent method is to use a nonradiating two-port
network in place of a transmission line, setting the transadmittance
y21 = -1 Siemens, to turn a voltage source V into a current source
V/(1 ohm). Set y11=0 and y22=0. I believe that this is the trick
that ELNEC uses.

Either method can transform a single voltage source into as many
current sources (of different values) as you need. Just set the
char. Z's of the several transmission lines, or the transadmittances
of the nonradiating two-ports, in (inverse) proportion to the desired
feedpoint currents.

A simpler trick that works but is less elegant because it dissipates
power, is to put a sufficiently large resistance (e.g., 10**6 ohms) in
series with a voltage source.

-Chuck W1HIS
Received on Thu Sep 24 1998 - 09:58:25 EDT