Alan:
Yes, it's true that if the signal is wider than your receiver and is
coherent (like a pulse), then the SNR will increase if you widen the
bandwidth, inspite of the fact that the noise increases in proportion to the
bandwidth. This is because, with a transmission like a pulse, all the
separate spectral lines add IN VOLTAGE, so the peak voltage increases faster
than the noise level as you make the receiver wider. This explains why you
can hear the galloping horses in a 6kHz wide AM receiver by ear (at least in
the south of the UK - I can't hear them up here) but not hear anything by
ear in a narrow CW filter. The latest version of my GPS-locked receiving
idea works in the same way - I can copy these LORAN spurii 25dB better by
adding-up the voltage of all 300 of them across the SSB passband than I can
by trying to see a single line in a narrow passband.
But that isn't really what I find surprising about the 42uW figure I have
just calculated. These LORAN lines, when seen on Argo or similar software,
are around the same sort of level as the weak TA signals we are exchanging
these days. We choose to do the TA tests on frequencies which avoid these
lines, so they must be around the same level if they annoy us to that
extent. If 42uW is a typical erp of a LORAN line, are we only transmitting
signals of that order ourselves? Surely most amateur stations would like to
think they are transmitting a few hundred mW. Something doesn't make sense.
I really don't mind if I have made a mistake here!
73
Peter G3PLX
|
