Richard Fry wrote:
"Roy Lewallen" wrote
If you do a bit of analysis with a modeling program, I think you'll
find that if you generate a circularly polarized signal, it'll become
nearly linearly polarized once it reflects from the ground. ... There
might
be a way to generate a signal that's circularly polarized after
reflection, but I don't know how to do it.
________

Roy, won't a c-pol signal remain c-pol after a low-angle terrain
reflection, except that its rotation sense is reversed? (The magnitudes
of the v-pol and h-pol reflection components are nearly the same, but
there is a 180-degree phase reversal in the v-pol reflection with
respect to the h-pol reflection.)

This has been applied with good results in analog TV broadcasts using
c-pol, because a c-pol receiving antenna rejects reflections of the
transmitted signal -- which effectively reduces the multipath "ghosts"
seen on a TV set when linearly polarized receive antennas are used.

A head-on reflection to a flat surface results in the polarization sense
reversal you mention. But reflection at a shallow angle doesn't. To see
why, get the EZNEC demo program and run separate elevation plots of a
dipole and a vertical, save the first plot, then superimpose it on the
second. Begin with a perfect ground. You'll see that the field from the
horizontal antenna is zero at zero elevation angle, while the field from
the vertical is maximum. So if you reflect a circularly polarized signal
from a perfect ground at a very low angle, you'll end up with a
vertically (linearly) polarized field. The horizontal component
disappears (that is, the field disappears each time it rotates to
horizontal) because the reflection is equal in magnitude to and out of
phase with the direct signal, so the two sum to zero at a distant point.
If you look at a higher angle where the horizontal and vertical fields
are equal, that's an angle at which you'd maintain circular polarization
if those same two antennas were spaced and phased for circular
polarization. At the elevation angle of the next null in the horizontal
field, you'll again get a purely vertically polarized field. The
conditions change when the ground has a finite conductivity, but you can
use the same general process to understand what happens.

EZNEC+ or NEC-2 users can model particular antennas and directly see
what happens to the polarization circularity. In EZNEC+, click on the
Desc Options line, Plot and Fields tabs, and choose one of the circular
polarization choices for Fields To Plot. Then run a calculation and look
at the plot. You get the polarization circularity by clicking the FF Tab
button to show the pattern data in tabular form. The AxR column shows
the axial ratio - the ratio of major to minor axis in dB. 0 dB
represents perfect circularity, 99.99 dB means the field is completely
linear, and in between represents various degrees of elliptical
polarization. Begin with an antenna in free space and verify its
polarization circularity in some direction. Then elevate the antenna,
add a ground, and note the effect. A simple test antenna is a pair of
crossed dipoles, very close but not touching, fed in quadrature, which
will generate fairly good circular polarization broadside to the plane
containing the antennas.

Roy Lewallen, W7EL