I have looked back at some other results this evening. There was QSB at
this time. This was using my signal to Wootton Bassett 104km.
17:32 500 G3ZJO de G4MDH Op4 104 km -26 dB in Wootton Bassett with
0.0002w + 5m inv.L - rx
17:27 500 G3ZJO de G4MDH Op8 104 km -33 dB in Wootton Bassett with
0.0002w + 5m inv.L - rx
17:12 500 G3ZJO de G4MDH Op4 104 km -25 dB in Wootton Bassett with
0.0002w + 5m inv.L - rx
17:04 500 G3ZJO de G4MDH Op8 104 km -30 dB in Wootton Bassett with
0.0002w + 5m inv.L - rx
Eddie
On 09/02/2012 22:23, James Moritz wrote:
Dear Graham, Eddie, LF Group,
There is 3 dB gain in each X2 time step
One would expect so, but Eddie's results apparently contradict this.
But, if the transmitted signal power and noise power spectral density
remain the same, one would expect the indicated SNR to be the same
also, independent of the mode selected, since the SNR is referred to a
constant bandwidth (2500Hz), rather than the actual bandwidth occupied
by the signal.
I think without using path simulator or similar these are a
little subjective ..
This is certainly true; 3dB variation in either signal or noise level
can occur very quickly on 500kHz. Some time ago, watching G4JNT's WSPR
beacon on 500k, the signal level would sometimes vary by more than
30dB during a single 2 minute TX period. That was in darkness over
about 120km as I recall; a shorter path during daylight might have
much less variation, but 3dB could still easily occur.
What is needed is some independent way of determining the SNR during
each transmission. But a bit tricky when the signals are very weak.
Alternatively, transmit both signals simultaneously, from the same
transmitter, over the same path, and decode both simultaneously,
which, conveniently, Opera can do with some modes at least. This type
of test seems like a good application for the G0NBD dual-channel
beacon...
Cheers, Jim Moritz
73 de M0BMU
|
