[Discuss-gnuradio] Fwd: usage of firdes.root_raised_cosine (double gain, double sampling_freq, double symbol_rate, double alpha, int ntaps) function

[Linda20071]

Additional notes on this issue:

I have played with the taps in this Polyphase Clock Sync block using mpsk_stage3 file, and I found:

1.  firdes.root_raised_cosine (samp_rate/sps, samp_rate, 1.0, 0.35, 11*4*nfilts) makes the output before synchronization freeze. Besides, the constellation graph after synchronization couldn't display anything.

2.  firdes.root_raised_cosine (nfilts, nfilts, 1.0/float(sps), 0.35, 11*sps*nfilts),  firdes.root_raised_cosine (nfilts, nfilts*sps, 1.0, 0.35, 11*sps*nfilts), and firdes.root_raised_cosine (nfilts, nfilts, 1.0, 0.35, 11*sps*nfilts) all work well with mpsk_stage3.grc  (https://raw.githubusercontent.com/gnuradio/gr-tutorial/master/examples/tutorial7/mpsk_stage4.grc).

We all know "Polyphase Clock Sync" is the key block for synchronization in demodulation process. Could somebody give out some insight on how these RRC parameters (rrc_taps) should be chosen? 

Thank you.

________________________________________________________________________________

I guess I may have a misunderstanding on the usage of  firdes.root_raised_cosine (double gain, double sampling_freq, double symbol_rate, double alpha, int ntaps) function.

In the flowgraph, the parameters are set up as follows:

samp_rate = 32k;

upsampling_rate (sps) = 4;

# of pylyphase filter in Polyphase Clock Sync block nfilts = 32;

ntaps = 11*4*32;

If I set the gain as 1.0/float(sps) = 1.0/4.0, the rrc_taps for Polyphase Clock Sync should be (equation 1): 

firdes.root_raised_cosine (1.0/4.0, 32000, 32000/4, 0.35, 11*4*32).

However, in the wiki example mpsk_stage4 ( https://raw.githubusercontent.com/gnuradio/gr-tutorial/master/examples/tutorial7/mpsk_stage4.grc) ), it is set up as (equation 2):

firdes.root_raised_cosine(nfilts, nfilts, 1.0/float(sps), 0.35, 11*sps*nfilts), which works great for the mpsk_stage4.grc. Could somebody explain why this rrc_taps shouldn't be as in equation 1? 

 

Thank you.