Re: LF: RE: Re: WSPR -2 SNR

["Roelof Bakker" <roelof@ndb.demon.nl>]

Hello Clemens,

Thank you for your help.

This evening I did a rerun from my PERSEUS recording.
This is a bit tricky in regard to timing, however it can be done.

Instead of the original bandwidth of 540 Hz, I used a filter bandwidth of  
2400 Hz.
This time WG2XJM produced a SNR of -28, whilst with a bandwidth of 540 Hz  
the SNR was -33.
The difference is 5 dB.

The noise in a bandwidth of 540 Hz is 6.5 dB lower than in a bandwidth of  
2400 Hz.
Joe mentions a bandwidth of 2000 Hz and then the difference is 5.6 dB,  
which agrees a little better.
Note that decoding was not compromised by using the wider filter.

The reason that SNR's below -30 dB are rare, might be that in conventional  
receivers/transceivers it often is not possible to select a 500 Hz filter  
in USB mode and shift the passband center 1700 Hz in regard to the  
(suppressed) carrier.
73,
Roelof, pa0rdt