Saturday, June 20, 2020

A Ham Radio Series 5 - Antenna Polarization

One of those questions that comes up from time to time is from someone who has heard antennas are polarized and wants to know if that matters.  The answer, as always, is that it depends. 

The polarization of the radio waves from an antenna is ordinarily given as the polarization of the electric fields.  As a recap, a radio wave is an electromagnetic wave, like ordinary light but lower in energy.  In the 1800s, you may recall from reading or a class at some time, it was thought that light propagated in a medium called the Luminiferous Aether because there was ample evidence light traveled in waves and how can there be waves that aren't waves of something.  Without spending too much time here, the presence of the Aether was disproved by the Michelson-Morley experiment and then finally buried by early 20th century advances.  Long before that, Maxwell described how an electromagnetic field can propagate forward as a changing electric field created a changing magnetic field which created a changing electric field ... repeating forever. 


Animation of EM wave propagation; by convention the electric field is called E while the magnetic field is B.  The important part to notice here is that both fields maintain the same orientation as they propagate.  

The general rule is that horizontal antennas produce horizontally polarized radio waves while vertical antennas produce vertically polarized antennas.

Why do we care?  Anytime radio propagation is local or line of sight, the polarization will be maintained and various radio services have established preferences for polarization.  An antenna set for one polarization rejects the other.  The theoretical attenuation on a cross polarized signal, like a horizontally polarized signal on a vertical antenna, is infinite.  You never get truly infinite rejection because that depends on the antennas being perfectly perpendicular to each other, and very small angular differences matter.  A cheap and dirty workaround, although you lose signal compared to antennas in alignment, is to rotate one antenna 45 degrees to the vertical.  This loses 3 dB compared to perfect alignment, but is an easy workaround if you were to be unsure of how signals would arrive at your antenna, vertically or horizontally polarized. 

Virtually any service that uses vehicle mounted antennas or handheld radios is based on vertically polarized signals - taxis, police, fire, emergency medical, and ham radio FM services (most popularly the 2m or 440 MHz handhelds or car mounted radios) come to mind.  I know of no exceptions.  While vertical polarization is the default on the 2m repeaters, if you drop down to 144.200 and switch over single sideband, the default there is horizontal polarization.  A few times over the years, I've heard guys just getting started on sideband using their vertical 2m FM antennas and being disappointed until they rotate the antenna. 

If the propagation is ionospheric, the polarization gets "scrambled "in the ionosphere by a phenomenon called Faraday rotation.  This goes for your attempts at working the other side of the world or the guy behind the next hill by sending your signal straight up.  If you're using a vertical to work some remote island and they're using a horizontally polarized beam (or dipoles) polarization doesn't matter. 

Horizontal and vertical polarization are just the basics, there are other kinds.  Circular polarization, (the general case is elliptical polarization) is a natural phenomenon - see the previous link on Faraday rotation - and is commonly used in some satellite downlinks.  Quadrafilar helix or turnstile antennas are common amateur approaches to receiving weather satellites.  The tricky part here is that there are two types of circular polarization, too: Right Hand and Left Hand Circular Polarization (RHCP and LHCP).  Those antennas reject the other polarization and I don't know of a trick like mounting the vertical or horizontal antenna at 45 degrees. 

If you're interested in HF, polarization is less important than the antenna pattern you get from either a horizontal or vertical antenna.  The effect of antenna height is talked about in this post from February of 2018.  

The same 40 meter dipole mounted at 66' above ground (left) and one foot above ground (right).  On the left it's a respectable pattern for long distance contacts.  On the right, it's an NVIS (Near Vertical Incidence Skywave) for local contacts.  Vertical antennas tend to have low angles of radiation, like the dipole at 66' but with even lower take off angles. 


  1. Good primer, SiG! Regarding the Cross Polarization Loss, when I was in the Satellite TV biz, I had to set up ground station receivers for specific transponders and specific satellites. Part of that was setting up the antenna controllers for both Azimuth/Elevation, and "Cross Pole". Since it's easier to find the "Null" instead of the "Peak", we'd set the Cross Polarization by peaking on the transponder in the correct polarization, and then look at the feed that had the other polarization, and and rotate the feed to null it out. The most I ever saw at C-Band was about 35~36dB difference between the Horizontal and Vertical Polarizations. I've heard of it being as high as 40dB+ in certain cases, but our feeds weren't as "good" as the ones on the satellite.

  2. Virtually any service that uses vehicle mounted antennas or handheld radios is based on vertically polarized signals - taxis, police, fire, emergency medical, and ham radio FM services (most popularly the 2m or 440 MHz handhelds or car mounted radios) come to mind. I know of no exceptions.

    Except for the one Americans might need to be using soon:

    1. I'm gonna butt-in here for Sig and answer that that's not very practical for mobile use due to the long wavelengths required. You'd need a 33' long antenna at 14 MHz, and 66' at 7MHz. If you use "loaded" antennas for the elements your radiation efficiency isn't very good.

      NVIS gets a lot of talk, but isn't very practical for some applications.

  3. Sorta off topic but about antennas. SiG, I am trying to build a fixed satellite antenna based on a design that Cebik published in QST about 19 years ago. It is a pair of crossed Moxons pointed vertically. I have built the 2 meter out of 1/4" aluminum rod successfully threading the ends of the driven elements 1/4-20 thread. I am trying to build the 70 cm antenna using 3/32" aluminum welding rod. Research showed that I should probably thread that at #2-56; so I bought a die and tried threading it. About 1/4" in I sheared the rod. This was doing the threading dry. With your machining mastery, would using oil and backing off, clearing the die and continuing threading, repeating in small increments allow me to get the 1" or so of thread I need on that small rod?

    1. My disclaimer is that I have no experience with aluminum welding rod and how hard or tough or brittle it is. Your 1/4-20 on 1/4" rod is easy to thread because 1/4" is the major diameter - the peak of the threads - which the welding rod ought to be very close to.

      I think the problem with the second one is that 3/32" is .0938" and that's bigger than the threads are. Major diameter for a #2 screw is 0.086" so you're cutting almost 8 thousandths of metal from the peak of each thread and it has to go somewhere. I'd try to thread it 4-40. The major diameter of 4-40 threads is .1120 and the bottom of the threads is .0813. It might feel loose when you're done, but it should work. The closest screw size is #3 at 0.0990 but good luck finding a #3 die and nuts. (A quick check shows Bolt Depot has #3-48 stainless nuts!)

      If you have a way of turning that 3/32" rod down just a little bit, down to .086", that's got to help if you really want to do #2-56. If not, maybe some threading oil, turn the die 1/4 to 1/2 turn and then break it back. Clean it out often. It's really something where I'd be surprised if it worked. It's better to go really slow than tempt that nasty "click of doom!"

      I'll see if I have some sort of rod lying around that's 3/32 that I can try to thread.

    2. Thanks for the infor. I did a quick search o the internet and a reference I found related 3/32 to the diameter of a #2 screw thread and it didn't get into the major and minor diameters. So that is the way I went. To try to have this antenna up for next weekend, I think I am going to have to go with using #12 AWG copper wire the way Cebik built the antenna. I was trying to be fancy and make a nice looking antenna.

      All I have is a bench vice and hand threading equipment; no fancy machining equipment (yet), so I don't think I can turn the material down. The material may be too soft as it is 4043 fill rod and I don't know how one hardens aluminum (more research needed on my part). May go in the morning and see if the local hardware store or auto parts shop has a 4-40 die I can afford as here in Bandera, TX we are in the boonies. After that, I may go searching for #3 dies and hardware

    3. #12 wire gauge is .0808 which is a little undersized for #2. I bet you could thread that. I don't know how the mechanical properties would work out, but it's a band-aid.

      I assume either the USPS or the Big Brown Truck comes out where you are? I order from Travers all the time:

      McMaster has a couple, of course,
      a bit more expensive and no mention of brand.

    4. Tired threading the #12 wire and it works. Probably no worse than using a 4-40 on the welding rod. The copper is hard enough that it didn't shear off. The nuts do not feel too lose so I am going with that for my 70 cm antenna.