After trying QRSS with an SDR-1000 and noticing a frequency wavering it was suggested to try it at different frequencies to see if the waver changes with received frequency. Got around to doing that test and the results are below.
It does appear that the severity of the wandering is proportional to the frequency being received. I tested against WWV and CHU stations from 20 MHz down to 2.5 MHz and the images are below. For comparison I also did the same test using a Kenwood TS-2000. The setup used on the SDR-1000 was an SDR-1000 in enclosure with RFE board and PowerSDR v0.1.3 software. The soundcard used by the SDR 1000 for this test is Soundblaster Extigy. My previous QRSS images were done with a Turtle Beach Santa Cruz -- I changed sound cards to see if perhaps the sound card was the source of the wobble.
Above shows a Spectran trace of WWV at 15 MHz. A 3-4 Hz wobble is clearly seen in the SDR trace. The bold trace is the main WWV carrier and the lighter trace is the encoded time code WWV sends as a PWM modulated subcarrier 100 Hz above the main carrier.
WWV at 20 MHz (shown above) was considerably weaker, definitely not audible. Spectran was able to dig it out with both the SDR-1000 and the TS 2000. Again we see a bit of a waver. I'm not sure what causes the other harmonic traces showing for both radios. It's hard to tell with how weak the trace is, but the frequency wobble seems a little larger than at 15 Mhz.
Moving down to 10 MHz WWV (shown above) it looks like the magnitude of the wobble is not as severe. 5 MHz WWV (below) also shows a smaller wobble.
All of the screen shots to this point were done with Spectran with a resolution set to 0.67 Hz. Below is a trace of CHU (Canadian Time Beacon) at 7.335 MHz at 0.67 Hz resolution, followed by one of the same station at 0.17 Hz resolution. The time scale changes between the two captures as can be seen by the red ticks at the bottom of the trace. Each tick is 10 seconds.
At 2.5 MHz even with 0.17 Hz resolution the wobble seems to be less than 7.335 MHz as seen below.
The last capture I did was at 15 MHz with a 0.17 Hz resolution.
Comparing the 2.5 MHz and this 15 MHz capture would seem to show that the magnitude of the wobble is proportional for frequency.
I don't know the cause of the wobble. I've seen it with 2 different sound cards. I also did a quick test of changing the windowing the SDR software is using and that did not seem to make a difference. Might have to look at that a little more to be sure it does not have an effect. The fact that it seems to be proportional to frequency should give a clue as to the cause, but I've yet to puzzle it out! Thoughts anyone?
Update 20 September 2004: I've found the cause of the wobble. Read more here.
Comments to: Bill Tracey (kd5tfd@ewjt.com)
Last Updated: 20 September 2004
Copyright © Bill Tracey 2004