Cecil Moore wrote:

Funny, I don't see "fully cancel" anywhere in my posting.
I probably should have said "tend to cancel".

A free space vertical with horizontal radials in EZNEC
has horizontal radiation more than 40 dB down from the
vertical radiation. That's a high degree of cancellation.


The issue is the horizontal opposing radials only have that degree of
cancellation for perfectly horizontal directions.

You will be able to see your statement isn't true if you place the
antenna in freespace and look at pattern distortion at various
elevation angles. For example, the 2-d plot is skewed 2.11 dB from
being circular at - 45 and +45 degrees elevation. The skewing gets
worse at larger angles from the plane of the radials.

If the radials were REALLY radiating -40dB in all directions as you
wrongly assume, there would NOT be significant FS change in the azimuth
pattern at various elevations.

You looked at horizontal radiation, but the horizontal radials peak
radiation is vertically polarized and nearly off the radial's ends.
(Just like in a dipole pattern.)

The radials do indeed radiate enough to change the pattern a
significant amount (but not at zero degrees), but the largest problem
is decoupling the feedline shield. The fewer radials are used, the
bigger the problem becomes.

There are VERY good reasons everyone settled on four radials, and it
isn't the old wive's tale about making the antenna look good. Four
radials is a reasonable compromise between excessive common mode
problems and tolerable common mode feedline current problems, pattern,
and cost.

Don't feel bad though Cecil. Many people miss this point, even card
carrying Mensa members.

73 Tom