Hi Dave,
You wrote:
That was done by measuring the "strength" of the tone when analysing
the wav files with Spectrum Lab. The receiver AGC made the peak noise
level the same I could read off the difference in tone level (in a
tiny bandwidth) .
Did you turn your receiver's AGC off (or to "manual") when recording the
files ? If not, the increase of the audio signal will not tell you the
true story.
Besides this, I played with some recent recordings which Jim sent me.
Trying to subtract the spectrum of the Loran signal from the "wanted"
signal, but this didn't work too well. And the automatic multi-notch
function (which also acts upon the FFT'ed audio stream) also kills
CW-type signals.. too bad. So the removal of the Loran noise must be
done in the time domain, maybe similar to a classic "QRM eliminator" on
the rcvr's antenna input. Here is how such a circuit could work:
1.) Build a broadband receiver for 100 kHz, which recovers the waveform
of the Loran pulses
2.) Regenerate the pulses (constant amplitude) by means of a
Schmitt-Trigger.
3.) Use these pulses to drive an analog switch to "blank" the 137 kHz
receiver input.
Or (a bit more complicated due to the need to demodulate "positive" and
"negative" pulses):
Add the harmonics of the regenerated Loran signals, with amplitude and
phase carefully adjusted, to the 137 kHz signal (no blanking but
subtraction). This may produce less unwanted sidebands, but requires an
SSB-type receiver for the Loran signal with controlled 100 kHz L.O. .
Well these are just a few thoughts... not verified the principle because
Loran is rather weak here (or other kinds of QRM too strong ! ).
Cheers,
Wolf DL4YHF .
|
