WSPR transmit frequency distribution - 20m

4 posts / 0 new
Last post
n6gn's picture
Last seen: 10 hours 18 min ago
Joined: 2010/07/27 - 01:38
WSPR transmit frequency distribution - 20m
Image icon 20mdist.png29.87 KB

I've been on WSPR for over five years, all of that time on 2m and 70cm, but I also have had a WSPR station on HF much of it. On VHF/UHF WSPR mandates good frequency control to function at all and my station is GPS referenced, almost always with accuracy and stability better than 1e-10. Much of my HF operation has also been this precise, with the exception of times I've used UltimateQRSS U2 and U3 beacons which don't yet allow both center frequency and tone accuracy to this degree.

This precision is quite a bit more than the average WSPR HF station uses or needs. But as I've reported before, having it has allowed the few of us on HF WSPR that are so referenced to observe the Doppler shift/smearing due to the motion of the ionosphere as the bands are opening/closing. This generally amounts to no more than 1 Hz at 20m but it is easily identifiable in the database. Locally we've been to identify aircraft scatter on 10m and 20m signals, sometimes even yielding dual spots.

For fun I decided to have a look to see what the 20m WSPR segment looked like over time. Since the spots of others' stations I made were known to be within 1 Hz of their transmitted frequency, as reported by WSPR, WSPR-X or WSJT-X, I thought that a histogram of a significant number of spots might be revealing.

The attached histogram of 458 unique-station spots during November 2015 gives an idea of several things. One is where the most WSPR QRM is likely to be. This could be useful for choosing and setting transmit frequency. The big spike at band center probably represents those stations who either are GPS referenced or who have taken the care to use the frequency adjustment tools within WSPR to set their radios accurately. The spread around this center is some kind of statement about the average accuracy of 20m WSPRers radios combined with the "preferred frequency" for those that have deliberately chosen to transmit somewhere other than band center.

I find it interesting that this distribution is not symmetrical about band center. Whether it is because our radios are, on average, high in frequency due to some subtle bias in manufacturer setting, crystal ageing etc or whether stations choosing to QSY from band center to "avoid the QRM" tend to go up rather than down in frequency, I have no idea.

My initial action as a result of seeing this data is to QSY from my long-held spot at band center down to a "quieter spot" in the band. On a busy weekend afternoon, well over 20 stations are sometimes seen at the same time so maybe doing this will decrease QRM and increase spotting of my signal slightly.

Whatever you may make of this, I offer it here for your amusement.

Glenn n6gn