Re: NEC-LIST: Question about Average Gain calculation

Hi Scott,

I think there might be an additional small error in your RP card. I believe
you have to add one more theta step and one more phi step in order to "close
the circle" of the pattern.

I can't claim credit for knowing this beforehand, but your posting got me
curious so I looked at the way that the commercial program NEC-Win Plus
(Nittany Scientific) builds the NEC deck when the user requests an average
gain calculation. Sure enough, if the model is over perfect ground the RP
card is (for example)
  RP 0 31 121 1002 0 0 3 3
and in Free Space the card is
  RP 0 61 121 1002 0 0 3 3
The actual step sizes can be partially controlled by the user but in all
cases there is one extra theta step and one extra phi step such that the
theta range is 0 to 90 inclusive (or 0 to 180 inclusive for Free Space) and
the phi range is 0 to 360 inclusive.

This is also shown in Example 3 of the NEC User's Guide (Part III). That
example is for a vertical dipole over perfect ground and the RP card is
  RP 0 10 2 1301 0. 0. 10. 90.
This shows a theta range of 0 to 90 inclusive (0, 10, 20, ... 90). Phi only
goes from 0 to 90 inclusive but I think that's okay in this case since the
pattern is symmetrical in the azimuth plane. I tried running this model
with more phi steps to cover a complete circle and the average gain was the
same.

Except in special cases like this where you can take advantage of known
symmetry in the pattern, I think the trick is to make sure that the solid
angle used in the averaging is either 2*pi steradians (Perfect ground) or
4*pi steradians (Free Space). That is, a full hemi-sphere or a full sphere.
Using your (latest) RP card of
  RP 0 90 180 1002 0 0 2 2
results in a solid angle used in the averaging of only 3.9766*pi steradians,
just shy of a full sphere (4*pi) and hence possibly giving a slightly
inaccurate average gain (depending on the actual shape of the pattern and
the size of the steps).

Dan AC6LA
www.qsl.net/ac6la/

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