[linrad] Re: Challenges!!

[Leif Asbrink <sm5bsz.com; leif@xxxxxxxxxxxxxxxx>]

Hello Chuck,

> I'd like to see you expand the scope of #4 to include splatter from any
> source.
Sure, if someone would be kind enough to send a recording:-)

> For those of us that mostly DX the AM broadcasters, splatter is the
> huge enemy.
Really? Are there bad AM transmitters out there or do you want
to listen near a strong local AM station?

> A solution that I'd like to see is a sort of Costas loop where:
> (1) the signal is split into two paths where one is demodulated with an
> in-phase carrier and the second with a 90 degree shifted carrier,
Linrad already has this feature:
http://www.sm5bsz.com/linuxdsp/run/am.htm

> (2) the "Q" output is shifted an additional 90 degrees, and (3) the two demodulated
> outputs are selectably added or substracted.
This will be equivalent to receiving one sideband or the other, 
Linrad sends both I and Q to stereo head-phones to allow
your brains to do the processing. I think that will often
be a better solution. If one wants one sideband only, I think
it will be a better idea to just set the appropriate filter and
use the BFO to receive the SSB signal that remains.

BUT, if the problem is splatter, much better solutions are possible.
The splatter is a distortion on the modulation. It should be possible
to detect the modulation in the offending channel and to compute
the expected splatter which could then be removed. In principle
Linrad could learn exactly what non-linearities the offending
station suffers from and then build exactly the emitted signal
with an accuracy detemined by the S/N of the main components of
the offending station.

To try to implement something like this, I need a wideband recording.
Wide enough to include all of the desired station as well as the full
spectrum of the offending station (splatter on both sides included.)

73

Leif / SM5BSZ




> 
> Chuck
> WD4ELO
> 
> -----Original Message-----
> From: Linrad mailinglist [