On Thu, Feb 01, 2007 at 11:02:54PM +0100, Martin Dvh wrote: > > How's it working with the AGC loop? > The agc2 seems to work. > The feedforward agc only works when manually compensating for the noise (1.0 > + noise as reference) > This would need an error loopback from the gr_mspk receiver > > I think my main problem now is detecting of which quadrant we are in.
I believe that V.32bis differentially encodes the quadrant rotation to avoid this problem. > Or use the right kind of differential coding. > (I just used the gr.diff_decoder_bb and gr.diff_encoder_bb) > > I have also strugled with the right formulation for the phase_detector error > signal. > (conj or not, pos or neg sign) But I think I got it now. > > I send the complete constellation as list from python to the > mpsk_receiver as vector. For some reason I had to reverse this list > to get it to work. This could also mean I needed to solve this in a > different way and it just accidently now sometimes works. (mirror > the constellation, add pi, something like that) > > Greetings, > Martin _______________________________________________ Discuss-gnuradio mailing list [email protected] http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
