Over the last few months I have been playing with WSPR and I wanted to be able to use it as a rough way to evaluate the relative performance of different antennas. My goal was to test how well two different antennas that I made for field use compare to my base station antenna.
My base station’s loop skywire was tested in its current configuration, 270 feet of wire strung between several trees. The two test antennas were put up in the same configurations that I intended to use them in the field.
- Loop Skywire (average height of about 35 feet)
- 80/40 Loaded Dipole (inverted V, center at 21 feet, ends at 5 feet)
- 80/40 Resonant End Fed (sloper, loaded end at 25 feet, feedpoint at 3 feet)
I did 24 hour WSPR runs using 5 watts of power with each antenna on successive days. The idea was to test each antenna in as equivalent band conditions as possible. By using the WSPRnet Database, I was able to collect signal reports from every station who heard me over the 24 hours. I then put all of this data into a spreadsheet. I calculated the average, median, maximum, and minimum dB signal reports for each antenna on every band I did the test. I also generated the same results for stations within a 500 mile radius as well as within a 300 mile radius. The purpose of these additional calculations was to evaluate each antenna’s performance for EMCOMM situations. Finally I calculated the same data for the distances from the receiving stations.
The first thing that I noticed when looking at the WSPR data was how closely the end fed and loaded dipole performed on both bands. The end fed seems to have a slight edge, but I would put this down to it being higher in the air rather than any inherent design advantage. On both bands the end fed is about 1dB better than the dipole, but given band fluctuations from day to day and the inaccuracy inherent to WSPR signal reports I’m going to call this a draw. Even the receiving station distance numbers were strikingly close to one another. This makes sense since both of these antennas are essentially identical, except one is fed in the center and the other is not.
On 80 meters the loop is clearly the best performer, besting both of the test antennas by at least 2dB overall and by several dB for regional contacts. It also reached much further out with almost double the average and more than double the maximum distance to a receiving station. Add in that it had the most spots from almost 40% more stations and it is clearly the most effective antenna.
On 40 meters the loop’s results are more complicated. Looking solely at the signal report data the loop is the worst performer of the three. I found this to be a ridiculous assertion because of how well this antenna performs in my personal experience for both regional and DX communications. One explanation for the relatively poor overall report is that the loop easily outperformed the other antennas in average and median distance to receiving station in addition to receiving 27% more signal reports from 33% more receiving stations. This could have skewed the results because more distant stations with additional spots would give weaker signal reports.
To test this idea I dug a little deeper into the data and looked at stations that heard all three antennas. What I found was that the loop generally had more spots from the same station than the other two antennas. These extra spots always came at the poorest times of day for propagation and consequently resulted in very low signal reports. When the receiving station had spots for all three antennas at the same time of day, the loop almost always had the highest signal report, usually by multiple dB. This combination of factors pulled down the average and median signal reports and masked how well the loop performs on 40 meters. I think this information also points to just how good of a performer the loop is on 80 meters because in spite of having a similar problem to overcome it still received the best signal reports by far.
Using WSPR to evaluate antennas is not an exact science and I am far from an expert statistician, so these results are by no means definitive. That said, I think this was a worthwhile exercise and resulted in some interesting data that generally correlates with my first hand experience using these antennas.