On Monday 23 March 2009 16:35:45 you wrote: > We've got the patch in the queue for testing. It looks correct, > though I wonder how the block even worked at all given this issue. > > Tom--could you comment?
Although I am not Tom ... Obviously, if there is no carrier offset, freq (IMHO misnamed, should be something like phase_increment ...) is 0, and adding 0 twice is no problem. As long as freq is smaller than pi/M, it gives some additional noise, but it should still work. The costas loop sets freq so that the average value is correct. Example: 3 Samples per symbol carrier offset of 0.2 (radians) freq should be 0.2/3 As freq is added 4 times instead of instead of 3 times (once per sample), freq is set to 0.05 this gives an phase error of -0.016 for the first sample, 0 for the second and +0.016 for the 3rd. Stefan -- Stefan Brüns / Bergstraße 21 / 52062 Aachen mailto:lurch at gmx.li http://www.kawo1.rwth-aachen.de/~lurchi/ phone: +49 241 53809034 mobile: +49 151 50412019 _______________________________________________ Discuss-gnuradio mailing list [email protected] http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
