Having used some of the WSJT suite during the Norfolk Island VK9NA expedition in January 2011, and following some postings to the VK1 mailing list in recent weeks, I was curious to know how to use WSPR.  The best reference is at the website of its author, Joe Taylor W1JT.

WSPR is part of a suite of software tools that use digital signal processing (DSP) to detect and decode very weak signals, much weaker than can be even detected by the human ear, let alone understood.  Possiblly the best known modes are the FSK441 mode used for meteor scatter contacts on vhf bands, and the JT65 series used for terrestrial dx contacts and Earth-Moon-Earth (EME) contacts on various vhf and uhf bands.  The JT65 modes were used at VK9NA.

Having downloaded and installed the software I then had to see what audio levels were suitable.  I first tried connecting a cable from the speaker/headphone socket of the radio to the mike input on the computer.  Levels were very sensitive and I had to cut everything down, but even then it didn’t work well.  I needed to cut down the audio level output from the radio, an FT817.  It was overdriving the mike input of the sound chip in the laptop.  If there had been a “line in” option I think that would have worked almost without any further change. The microphone input is more sensitive as it is designed for the much lower level of a microphone.  I wired up a potentiometer to enable the sound level going to the computer to be set as a fraction of the output from the radio and that worked very well.  I finished up setting that to about 10% of full scale.

The website http://wsprnet.org/drupal/ is the next resource I found very useful.  It lists the beacon frequencies commonly used world wide, for each band.  I have tried the 7, 10 and 14 MHz bands and found it worked very well.  A good read of the user manual is advisable.  Adjusting the input level on noise to be around 0 db was quite important.  After that it was just a matter of tuning the radio to the correct dial frequency, using USB mode.

Before long I found the screen was gradually building up a list of callsigns received and their signal levels, frequency offsets and the stated power level of their transmitters.  Most seem to be 5w but some are less than 1w and one notable station indicated 1000 watts, but was no stronger than others running 5w, so I think he specified his power incorrectly.

The WSPR screen looked like this at one stage today.

Thanks to Ian VK1HOW for sparking my curiousity about this mode. This must be providing a wealth of data for propagation students.