[TowerTalk] Re: Horizontal + Vertical Polarization Question

[Jim Lux]

At 09:39 AM 11/21/2003 -0500, Martin Ewing AA6E wrote:
>K9OM asks:
>
>Has anyone experimented with feeding both a horizontally polarized and
>vertically polarized antenna simultaneously on HF?  Such as, feeding a 
>horizontally polarized yagi and a  vertical on 20, 15, or 10 meters?
>------
>
>Well, in my time, we called this circular polarization.  That's what you 
>have if you feed H and V 90 degrees out of phase, anyway, and if the phase 
>centers of the two antennas coincide.  If the antennas are not symmetric, 
>you'll get elliptical polarization, i.e., circular plus linear.  If the 
>antennas are physically offset, you'll get "interesting" interference 
>fringes on top of the normal radiation pattern.
>
>Circular polarized antennas are insensitive to the (linear) polarization 
>angle of the incoming wave, so that eliminates one source of 
>QSB.  (Propagation can twist the polarization angle, but typical HF 
>propagation favors H polarization.  It's the same reason your Polaroid 
>sunglasses work.) On the other hand, a right-hand circular antenna rejects 
>left-hand circular waves, so if both sides are playing this game, you'd 
>better agree which sense to use.  The optimum antenna (on receive, but 
>same argument for transmit) for a linearly polarized wave is linearly 
>polarized at the same orientation.  If you use circular receive for a 
>linear transmit, you are going to be down 3 db against an optimum linear 
>antenna.  So there's no free lunch.

I think, though, that given the random polarization of the incident wave 
when it actually hits the antenna(especially given all the stuff around the 
typical ham antenna.. relatively few of us have antennas on tall 
non-conductive towers, fed with optical fiber, over perfectly flat salt 
water marshes, etc.), the fact that most CP (or, more properly elliptical 
polarized) antennas have no deep nulls when receiving a LP signal is 
significant. Sure, you lose 3dB on the peaks, but you also don't get the 
20-30 dB fade when the signal happens to be cross polarized. (one reason 
why TV stations broadcast CP/EP.. it keeps multipath fades from being so deep)

It would be very interesting(!) if someone were to actually measure the 
incident polarization of typical skywave signals (actually, someone 
probably has.. the radio astronomy folks working at HF have to deal with 
nulling out interferers, etc., and I'm sure some one in the SW broadcast 
business has looked at this), particularly in terms of the short time 
statistics.

If someone could suggest a simple way to do it, I might do it 
myself.  Perhaps a couple of short dipoles, fed through a FET switch, to a 
receiver like the PCR1000 hooked up to a PC for data logging? Flip the 
switch 10 or 100 times a second, etc.  I could even look at the NCDXF 
beacons or, perhaps something like WWV/WWVH.  Shortwave broadcast might be 
a good source of a transmitted signal at a known frequency and time.