Re: NEC-LIST: Fw: Lambda/4 monopoles lamdba/4 apart

At 9:51 AM -0700 9/22/03, Phil Ede wrote:
>...I have... modelled a simple pair of Lambda/4 monopoles spaced
>lambda/4, and fed 90 degrees apart in HFSS using perfect ground, and
>get a good cardiod pattern with a small rear null.
>
>I have modelled an almost identical confifration in NEC, and no
>matter how I tweak the detail (frequency, number of segments, number
>of wires in each monopole, wire diameter, EK command etc) I get a
>much worse pattern than using HFSS..
>
>Can anyone tell me where I might be going wrong using NEC ?
>...

I don't know what HFSS is, so I can't comment on the validity what
you did with it or the results you got from it; but, with NEC, it
appears to me that your mistake was exciting the monopoles with
equal-magnitude, quadrature-phased, _voltage_ sources. Because the
driving-point impedances seen by these sources are radically
different, the currents in the two monopoles are not at all what you
wanted; i.e., their magnitudes are not equal and their phases do not
differ by 90 degrees.

Feed the monopoles with equal-magnitude, quadrature-phased, currents by

(1) deleting those voltage sources;

(2) connecting a transmission line (use the TL command) from the fed
segment (second from the bottom) of each monopole to the same, one,
new, very short (0.001 wl or less) segment, located anywhere that's
legal;

(3) giving these two transmission lines equal characteristic
impedances (Zo = 1 ohm will do), and making both their electrical
lengths equal to 90 degrees at the operating frequency; and

(4) putting a source (a voltage source will do) in the
aforementioned, new, very short segment (to which both TL's) connect.

You will get a more nearly perfect cardioid pattern if you make the
monopole wires thinner.

-Chuck, W1HIS

-- 
The NEC-List mailing list <nec-list_at_gweep.ca>
http://www.gweep.ca/mailman/listinfo.cgi/nec-list