Hi Samuel, Is there any reason why you do the quadrature demodulation before the resampler? Usually the CPU limitations require to decimate the signal as soon as possible to a manageable rate for the quadrature demod block. Also, if the FSK symbols are not RC filtered and represent a square wave, I remember reading that the M&M clock recovery algorithm does not deal well with a square wave. Perhaps your resampler deals with that?
Hope this helps, Adrian On 1/22/19, samuel verdon <sam.ver...@gmail.com> wrote: > hello Everybody, > > I am trying to recover my fsk signal with the Muller&muller block. > Here is my configuraiton: > Sample rate of USRP = 240K > bitrate of signal = 9600bps > > receiving the signal from the USRP, > apply a low pass filter -> quadrature demod ->rational resampler -> m&m > sync block > > > My signal contain a preamble of 100ms of 1010 (0x55). after my > Muller&Muller block, I can see the recovering working for my signal but not > for my Preamble. I have pulse on the preamble which sometime are visible on > the data part as well. > https://ibb.co/hfLt62K <https://ibb.co/hfLt62K> > https://ibb.co/xqRkRP7 > > I have tried also with Polyhpase clock sync block but this one took to > much time to be syncronised with my preamble. > > Can someone help me by explaining this pulse? I porbably do not use > correctly the sync block ( I have left the default value except for the > omega=sps ). Can someone redirect me to a documentation to use it > correctly? > let me know if I have to give more information. > > Thank you in advance for your help > > -- > Samuel Verdon > _______________________________________________ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
