Hi all,Yesterday a sizable group of us conducted on-the-air tests of SuperFox (SF) mode using a pre-release version of WSJT-X 2.7.0-RC7. We included some operation using old-style Fox-and-Hound (FH), so that direct comparisons could be made.
W7YED acted as SuperFox (SF) for about half an hour, from 0130 0200 UTC. Then N7QT ran as SF for 20 minutes and as FH for 7 minutes. Both stations are in locator CN87. The two attached plots summarize results related to received SNRs and achievable QSO rates, based on signals received at K9AN (EN50) and W2ZQ (FN20). Propagation was excellent throughout the test session. More that 60 different hound calls were worked during the session.
As SuperFoxes W7YED made 165 QSOs in 29 minutes (rate 341 QSO/hour) and N7QT made 109 QSOs in 18 minutes (363 Q/h). N7QT then made 30 QSOs in 7 minutes using FH, rate 257 Q/h. These QSO rates are all very good and within the available statistics they are roughly equal.
The attached plots show received SNRs at K9AN and W2ZQ for both fox stations over the full test hour, roughly 0130 to 0230 UTC. Signals near the end were several dB stronger at K9AN than at W2ZQ, but otherwise the plots are nearly identical. Both show very clearly the fundamental advantage of SF over FH.
The SF signals were strong, around +17 dB most of the time at both K9AN and W2ZQ. N7QT had one extended fade near the middle of his SF time, but was still many dB above the SF decoding threshold. After N7QT switched to FH his initial CQ (a single-stream transmission) was strong, +18 dB at both K9AN and W2ZQ. But then, as N7QT made QSOs mostly using 3 or 4 streams, his "per stream" SNR dropped by as many as 9-14 dB. This is exactly as expected for the multi-streamed FH waveform.
If propagation were such that single-stream SNRs are around -10 dB instead of +17 dB, SuperFox could still make QSOs at rates around 350/hour. But in those conditions an old-style Fox would need to throttle back to 1 or 2 streams and his QSO rate would be more like 50/hour, at best.
Stated another way: propagation could have been 20 dB worse than that experienced at K9AN and W2ZQ, and all but 3 or 4 of the 274 SF QSOs by W7YED and N7QT could still have been completed. In the same conditions hardly any (maybe 2 or 3?) of the FH QSOs would have succeeded.
Look for a public release of WSJT-X 2.7.0-RC7 in the near future, with many more details.
-- Joe, K1JT
SNRs_at_K9AN.pdf
Description: Adobe PDF document
SNRs_at_W2ZQ.pdf
Description: Adobe PDF document
_______________________________________________ wsjt-devel mailing list wsjt-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/wsjt-devel
