Hello 

I did write a piece of software under Linux for use with the elektor kit (may 
2007). Being a programmer rather than a radio-man, I had (and have) to learn a 
lot on signals and processing. Currently my software works well, written in 
C++, using qt4 for the gui, using hamlib for the rig interface and using 
portaudio for interfacing the soundcard. The basic configuration handles analog 
modes (am, fm, usb, lsb), rtty, psk, navtech, mfsk fairly well. An attempt to 
decode cw is implemented, letting the user set relevant parameters, it is 
working, however should be made more intelligent.
I am working to understand olivia and included the open source implementation, 
but it needs quite some work before releasing.

The front end reads samples (standard 96K), the band filter is selectable in 
terms of bandwidth, positioning in the positive half of
the field, as well as in terms of kind. Currently IIR filters as butterworth 
etc are implemented, as well as FIR filters (with 255, 511 filterelements, so 
fairly sharp) and FFT convolution implementations of the FIR filters. The 
inputrate can be selected.

The backend runs with a lower rate (also selectable), and does the
decoding.

The user interface has spectrum and waterfall displays (front end for the whole 
96K, backend depending on the selected decoding mode).

The implementation is straightforward and was a very nice programming exercise 
(first program in about 8 years). It could benefit from more
advanced techniques for handling and improving the signals at various stages 
during processing and that is where I need input from this group. Another field 
where I could use some input is in descriptions for varieties of digital modes.

The elektor kit does work (not very sensitive though), I am thinking on another 
board (but preferably cheap), but the main concern is
the software.

Obviously, the software is under the GPL available for interested parties.

all the best, waiting for reactions
jan





Reply via email to