Rotation Polarization Fades

Home Up

 

 

 

Circularly Polarized Waves

Circular polarization is traditionally generated from an antenna system that launches a V and H wave with phase lead or lag in the polarizations. This is traditionally considered to be from a V and H antenna, because we are conditioned to think in those terms for electric field polarization. The wave actually does not have to come from two sources, as referenced to the earth at 90 degrees (vertical) or 180 degrees (parallel to earth or horizontal). Two antennas, one at 45 and one at 35 degrees tilt, would work just as well. It , or it could also be any other polarization angle 90 degrees apart, like 20 degrees and 110 degrees.

If we stood at one point and looked into a circularly polarized wavefront, we would see an electric field rotating as it moved toward us. The arriving wave would be continually rotating with a rotational period of the reciprocal of the frequency. A 1.830 MHz signal would be rotating in a 1/1,830,000 period for 360-degrees rotation. This is about 0.546448 ÁS for a full 360-degree rotation!

A circularly polarized wave rotates so fast the only net effect is a 3dB loss of level into a fixed polarization antenna. Unless the wave becomes elliptical from propagation effects, this 3dB loss occurs for any receiving antenna tilt angle. The fact the wave rotates cannot cause deeps slow fades, because by definition circularly polarized signal rotate 360 degrees at the time period of one cycle of the operating frequency.

As a matter of fact, many FM Broadcast stations transmit circular polarized signals. This is so any receiving antenna tilt-angle still hears about the same signal signal level. You will hear the station fine on a dipole, a horizontal TV band Yagi, or a vertical car antenna. The signal does not fade up and down because of circular polarization.

Here are the rules for receiving.

Circular incoming wave:

Antenna Fading Loss  
Vertical None 3dB  
Horizontal None 3dB  
Tilted but single polarization None 3dB  
Circular same rotational sense complete very high  
Circular opposite rotational sense None 0dB  

 

Linear incoming wave:

Antenna Fading Loss Incoming
Vertical None 0dB V
Vertical complete very high H
Horizontal None 0dB H
Horizontal complete very high V
Tilted linear polarization variable cos^2 Φ Angled Φ
Circular same rotational sense very small 3dB linear any Φ
Circular opposite rotational sense very small 3dB linear any Φ

 

Polarization error loss is cos^2 Φ

45-degrees of cross polarization produces 0.5 signal level, or 3 dB loss.  90 degrees theoretically is infinite polarization coupling loss.

Since the circularly polarized wave rotates at the period of one cycle,  50% of the time it has a polarization loss that complements polarization increase during the other 50% of the RF cycle. The resulting waveform is a perfect signal, but at half-power, as the circularly polarized wave excites the linear antenna.

 

Signal Samples

Please excuse the channel level difference in the recordings and pictures below. My left ear is less sensitive than my right ear, so I have a habit of running a few dB more gain on my left ear. Since these are recorded off a master audio buss, the balance adjustment carries over to  recordings.

I cannot check for circular polarization. I do not have two antennas at the same point that are RF combined, or any way to calibrate to that point from the house. I can check for wave polarization rotation and phase instability. If the wave is circular, the very maximum level difference would be 3dB signal reduction. If the wave is elliptical, loss can be greater. Eventually it might be elliptical enough to be considered a linear wave. Any deep fading has to either be from very slow rotation of an elliptical wave, or a linear wave, or a complete signal loss for any polarization. Circularly polarized waves cannot cause deep fades in linear antennas. Circularly polarized waves can only cause deep fades in circularly polarized antennas.

While it may seem strange, if two receiving systems share phase locked oscillators for all conversion and detection functions, RF phase differences translate directly into audio phase differences. RF test equipment, such as vector voltmeters, use this technique to measure phase differences between GHz frequency band signals.

Here is a receiver phase verification. This is with a single antenna on both receiver channels. In order to do any useful phase tests, the receiver system has to be phase stable between channels in this test:

 

Start time = 0.4 seconds

 

Mid time = 22.23 seconds

 

 

End time = 56.9 seconds

 

 

Careful examination of phase with a common antenna used for both receivers proves phase stability. There is no phase drift between channels throughout the entire recording time. This test proved my RX channels are phase stable, and any drift in phase is from the antennas.

 

This picture below is from KH6AT at 48.74 seconds in recording time. Left (top) channel "rear" Beverage pair 880-ft long x  350-ft spacing wide, right (bottom) channel "front" Beverage pair 880-ft long x 330-ft spacing wide. Guessing about 1000-1200 feet SW/NE (Echelon) stagger in these antenna pairs. 

 Notice two things occur below:

1.) One antenna does not always fade at the same time the other antenna fades

2.) The relative phase between antennas slowly rotates

 

KH6AT signals in-phase at this time:

 

 

KH6AT at 1 min 5.35 seconds into the contact is 180 degrees out-of-phase on channels:

 

 

180-degrees phase drift occurred over a time period of ~45 seconds. The drifting or rotating phase is why I cannot directly combine my large antennas into one giant phased antenna. This is also at least partly why really long antennas just don't work well. If I make antennas really long, they have more fading. I believe this is because the wave at different areas of the antenna is slowly rotating in-and-out of phase. Clearly we can see that spatial effect on KH6AT's signal.

KH6AT sample signal

 

V and H Sample

Lack of directivity hampers S/N ratio, but here is a sample of a true horizontally polarized antenna and a reference vertical at the same location. 

Note the in-and-out fading is much faster in this comparison between a small horizontal loop at 280-feet, and a vertical wire up along the same supporting tower as a vertical. This signal is a steady carrier on 1820 kHz, unknown skywave source (but almost certainly a AM BC station harmonic to the northwest of me).

 

This is probably caused by a slow rotation of polarization, since the antennas are essentially at the same physical point.

I need to get stronger recordings of this.

 

 

Recorded sample of carrier