I'm conducting a frequency observation experiment for the next several days looking at the close-in spectrum of the 7.335MHz CHU carrier as received here, looking to observe doppler shifting and multipath effects as the ionosphere changes. CHU is a time and frequency standard broadcast from near Ottawa, Ontario (450km NW of my location).
The spectrogram image will evolve very slowly (several hours for each row in the strip chart). Since the transmitted carrier frequency of CHU is maintained to a high degree of accuracy (5 parts in 1012), if the ionosphere were completely stable, the spectrogram plot of the received signal would be a thin straight line right at 7.335MHz. Instead, the spectrogram reveals some of the complex and highly dynamic changes various layers of the ionosphere undergo throughout the day/night cycle. A great description of this technique (and beautiful spectrograms) can be found at the site of Murray Greenman, ZL1BPU.
I am using a Flex SDR-1000 receiver with an HPSDR Ozy/Janus audio interface. Both radio and HPSDR hardware are disciplined to a Trimble Thunderbolt GPSDO for frequency stability and accuracy to about one part in 1011. Each FFT bin (pixel) in the spectrogram is about .023 Hz. Setup should be stable starting Tue, 23 Dec 08, 0200Z (local sunrise is 1211Z, sunset 2116Z).
40m CHU has been battling QRM since the broadcast band has been adjusted to include its frequency. Voice of Russia is swamping the signal here until 0600Z, and you can clearly see when other stations come and go during the day (some of whose carriers are off by a few Hz). Starting 27 Dec, I will switch to 90m 3.33Mhz, which should be a clear frequency, then go back to 7MHz after CHU moves on 1 Jan.