Hello Andy,
This is an interesting question.
The last few years I have increased the readibility of very weak non
directional beacons by decreasing the filter bandwidth. The results
have been quite spectacular. Using a SPM-30 selective level meter, I found
looking at the bouncing analogue meter that the minimum required S/N ratio
for aural copy was between 5 and 10 dB.
However a PERSEUS SDR is much better facilitated for this kind of
measurement.
Attached is a small part of my PERSEUS GUI and shows the signal of NDB
BK from Kazachstan, which is excellent to copy in a 5 Hz wide bandwidth.
Creating this picture was not trivial, though.
The signal level excursions of the band noise are around +/- 3 dB and the
static disrupted the measurement from time to time. Hence 15 second of not
by static interrupted recording was played back in a loop for 5 minutes.
The averaged band noise is around -143 dBm and the averaged signal (BK in
morse code, 7 w.p.m.) is -135 dBm.
Thus 8 dB over the band noise.
This may appear to be a wide margin, but both band noise and signal suffer
from quick changes in signal strength.
In practice the signal to noise ratio varies from 4 to 8 dB.
I will try to do some more measurements on weak daytime signals, but that
will have to wait until later this week.
On an other note, using software that takes advantage of the repititive
nature of beacons, it is possible to achieve a considerable improvement in
S/N. This could also be used instead of QRSS and enable the receiving
party to control the needed time for reliable identification, depending on
propagation.
Best regards,
Roelof Bakker, pa0rdt
Weak CW.jpg
Description: JPEG image
|
