[600MRG] Polarization reception testing at W5EST: 630m WSPR SNR update

[James Hollander]

All--  The WSPR database shows some polarization-null reception runs here in the last two weeks.  Last night's run is typical.  WD2XSH/15 WSPR SNR departs from low SNR for a couple of hours (0630-0840Z and around 1230-1430Z, an hour or so past local sunrise) and then returns to low SNR other times during the night and day.  I can't explain these results that have been consistent a number of nights.

You can identify my polarization runs on the database because they feature nearby WD2XSH/15 mostly in the negative double digits. (Without nulling XSH/15 is +5dB WSPR SNR.)  

On the database, enter Call such as WD2XSH/15, Reporter W5EST, and sort by timestamp. http://wsprnet.org/drupal/wsprnet/spots    

I'm trying to avoid unfounded conclusions about receiving with simple equipment what seem to be polarization variations on 630m.  Any questions or words of wisdom are welcome. Send to mrsocion at aol.com.

Such cancellation nulling works the best on the strongest signals such as XSH/15 a few miles away.  Less frequently, I can null WG2XXM in Okla and WD2XSH/7 in Louisiana. XSH/7 nulls at a different canceller setting than XXM and XSH/15. At your locations elsewhere, a nearby 630m station that has ground wave to aid testing any time of day is the most convenient type of station to study this way.      

73,  Jim H   W5EST

 
DETAILS AND TESTING 



A homebrew noise canceller circuit with input coupling transformers takes the common mode signal coming down from the attic dipole on both twinleads (++) and separates it from the differential mode signal across the twinleads (+-).  
 
This dipole is fed directly by twinlead, and not by the often-used earth-grounded coax to a balun-fed dipole center.  The shack is on the 2nd floor.  The twinlead comes vertically down from the attic, and is coupled to the canceller circuit and SDR receiver. The "antenna" path continues in a non-recommended way via the home AC grounding 3rd-wire conductor from the shack down to earth to complete a "vertical." 
 
So I believe each conductor of the twinlead (and 3rd-wire path as well) acts like a vertical antenna to pick up vertically polarized signal and deliver it as a common mode signal (++) with respect to ground. Meanwhile, the dipole legs provide a capacitive hat for the twinlead vertical like a Marconi-T. But the dipole legs aren't shorted like a Marconi-T.   (Some ops think of common mode as a synonym for "noisy" because in their stations, common mode on coax generally involves an undesired noise or interference. But here I'm trying to get the twinlead to do double duty and pick up a vertically polarized RF desired signal and deliver it as common mode.)
 
The dipole also acts like a dipole to pick up horizontally polarized signal and deliver it as differential mode signal (+-) between the conductors of the twinlead relative to each other. At 630m the horizontally polarized component isn't very strong, especially with the dipole so close to the ground. But here the low-conductivity-earth, high-hillside location does permit some horizontal signal, I conclude.  
 
The coupling circuit of the canceller separates the common mode from the differential mode and phases the respective signals 180 degrees out of phase and equalizes their amplitudes and sums them together for the SDR receiver.  Then I look at the spectrum analyzer and waterfall windows of the SDR and adjust the canceller for the best null. The receiver audio goes to the software WSPR decoder on the PC. The WSPR decoder records the WSPR SNR as the SNR of the self-cancelled signal varies away from null and back to null during the night.  
 
I believe what the SNR represents is small variations in polarization away from null.  I'm also trying to perfect a way to measure or estimate the polarization angle itself, the angle away from the vertical polarization that 630m stations generally transmit.  But the variations in polarization are themselves important because they can provide meaningful indication of changes in the way the signal is propagating.  Propagation professionals and radio astronomers use a device called a polarimeter to do this type of thing, but usually at HF/VHF/UHF instead.
 
On the skeptic's side of things, the presumed differential mode signal versus common mode signal from my imperfect attic dipole/twinlead/grounding system might be just some confused artifact. I'm trying to make sure the signal null departures are indeed saying something about polarization itself.  Also, polarization and directionality are probably intertwined things. And I need to make sure the "differential mode signal" (horizontal) is actually there in the first place.  The "differential mode" mustn't be just an unseparated portion of the common mode (vertical) signal. So the differential mode has to be well-separated from common mode. 
 
So the polarization reception here is a work in progress. One good sign is that the dial adjustments on the canceller to null WD2XSH/7 in Louisiana differ from the dial adjustments that null XSH/15 in west Little Rock, Arkansas a few miles from here, and that null WG2XXM in Oklahoma.   WD2XSH/7 delivers a different ratio of differential mode current to common mode current in the antenna than XSH/15.
 
Another good sign involves the null being at or below noise level.  The resulting cancelled signal varies by about 10dB up and down and into the noise over about half a minute--while uncancelled XSH/15 signal varies 1-2dB on the same display. If the canceller were merely nulling part of a steady-angled linear polarized signal against another steady-angled part of itself, the variations from null ought to still be 1-2dB, and the cancelled signal would merely be lower in average strength level. 
 
A third good sign involves departure from null of the cancelled signal on a 2-minute WSPR time-slot SNR basis over periods of hours.  XSH/15 WSPR SNR departs from low SNR for a couple of hours (12:30-2:40am and around 5am to sunrise) and then returns to low SNR other times during the night and the day.  In the night, when it's departed from null, I can re-null the signal below the noise level.  So that says the increase in SNR isn't just due to an increase in signal strength.  It's due to a changing ratio of differential mode to common mode, which I interpret as a change in polarization.     
 

--END--