This post is a bit old... but I'm having a similar problem. Created a modified gr_sig_source_c block such that attributes are assigned by pseudo-randomly generated numbers (restricted to ranges such as would be proper, as well as swapping waveform types for integers 0-5 and randomly assigning one). It compiles and runs. Amplitude and frequency appear to vary, but sine/cosine waves are the only ones which generate properly. Is there a known solution to this?
~ Francesco B. Syed Faisal Shah wrote: > > Hi fellows > > We are trying to transmit and receive train of square wave over > wireless using RFX 2400. We modified the usrp_siggen.py file > accordingly to generate square wave. We used > > ./usrp_siggen.py -f 2450M -w 100000 > > Is the value with -f option correct? We assume that -f option refers > to center frequency at antenna but in an http://lists.gnu.org/archive/html/discuss-gnuradio/2006-07/msg00097.html > earlier post it was mentioned that it is the DUC frequency. Which > one is correct? > > On the receiver side, we used usrp_rx_cfile.py script to write the > received signal samples on a file. The received signal does not look > like a square wave rather a sinusoid. In some cases, the sinusoid > appears to be as harmonic of the square wave, sometimes as a modulated > sinusoid. > > If anyone knows the cause and solution for this problem, please guide us. > > 1. Are we missing anything here? Without communicating with USRP, we > made sure that gr_sig_source_c.cc (underlying function that > usrp_siggen.py calls) generates a square wave. > > 2. Do we need to do anything with the USRP tune function? Can we > specify the IF frequency at the transmitter and receiver explicitly? > > 3. Or we cannot transmit/receive square wave through GNU radio. We > understand that may be some filters on the usrp board or daughter > board will not allow high frequency signals that will distort the > corners of the square wave but getting only a sinusoid does not seem > right.... > > 4. Our ultimate interest is not in square wave. For our purpose, we > need to generate a shortest possible pulse, say Gaussian pulse, > transmit it over the air and implement some ranging algorithms. We > started with square wave and got stuck. > > Any help/ideas will be highly appreciated. > > Thanks > > Faisal > > > _______________________________________________ > Discuss-gnuradio mailing list > [email protected] > http://lists.gnu.org/mailman/listinfo/discuss-gnuradio > > -- View this message in context: http://www.nabble.com/Problem-receiving-square-wave-tp10009840p20485774.html Sent from the GnuRadio mailing list archive at Nabble.com. _______________________________________________ Discuss-gnuradio mailing list [email protected] http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
