Embedded Files
One of the challenges using SNR data from receiver stations is the inherent variability in the data over time. This is shown in the figure below. Here a 1/4 wave telescoping ground-mounted vertical is used to test the variation of SNR data at receiver stations over time. The plot below shows the time-varying SNR data at a single receiver station, N2YCH over a time period of 7 hours on 20M. Note the large variability in the recorded SNR values. A simple method to characterize the SNR for this station for a given antenna is to compute a time-average value of the SNR.
Utilizing a single time-averaged SNR value computed for a test and reference antenna allows for a simple calculation of the relative gain;
Relative gain (dB) = SNR(test-avg) - SNR(Reference-avg)
The figure below shows the WSPR spots in time for a test antenna (Inverted L) and a reference antenna (1/4 wave Vertical) at station WA2TP on 40M. The time-average SNR values are shown as solid lines. The relative game is the difference between the time-averaged SNR values. (Please note that decibel values of SNR are not averaged. The values are first converted to a linear power term and then averaged. This average value is then converted back to decibel form to compute relative gain.)
The WSPR methodology therefore is to run a test over a long time period (5-7 hours) and transmit/collect SNR data over multiple receiver stations. Since I use a single transceiver with an antenna switch to the test and reference antenna, I alternate transmissions over the length of the test switching back and forth between antennas. In the figure above you can see an alternating pattern of transmissions at 30 minute intervals. By cycling back and forth between antennas over a long period of time, short term QSB (fading) and diurnal propagation shifts are averaged out. Gathering a high statistical density of spots per station allows you to extract a clean and repeatable relative gain from an inhernetly noisy WSPR data pool.
I have done extensive studies on the time interval and on the number of spots per antenna needed to get a converged solution. The details of this can be found in my paper published in QEX March/April 2026 edition entitled "Using WSPR to Predict Relative Antenna Gain of Directional Antennas". In short the guidelines I recommend are
Use a 30 minute time interval between antenna transmissions
Collect a minimum of 40 spots per antenna over the course of the test
I typically run at 1 watt and I use a 70% duty cycle. This gives the rig some time to cool down. I typically run an external fan behind the rig to keep it cool. I advise monitoring your rig temperature. Also, turn off sleep mode on your computer so that it runs the WSJTX software continuously. I typically run 5-8 hours depending on the propagation in order to collect 5000-10000 spots. Typically out of all those spots I find a couple dozen stations that offer up 40 or more spots per antenna.
Please take accurate readings of the time in which you switch antenna transmissions. You will need to know these times later on when determining which spots come from which antenna. You can monitor the progress of your WSPR spots on wsprnet.org
On poor propagation days, it may be a struggle even to get a couple thousand spots let alone 40 spots over antenna. My recommendation is to be cautious in interpreting results for less than 40 spots per antenna.
In the next section I describe how to download the WSPR results into Excel and extract relative gain.
Page updated
Google Sites
Report abuse