In the gnuradio/gr-noaa/README file under usrp_rx_hrpt.py, there is a
comment that
reads as follows:
“The present HRPT demodulator is only tested at decimation 16. The
only other
valid decimation rates are 24 and 32, which may work, but with more bit
errors. No
other decimation rates will work.”
My (perhaps naive) question is whether the important parameter is
the decimation
rate, or the sampling rate. For example, with 64 MHz USRP, a
decimation of 16 would
give 4 MHz sampling, whereas with a higher rate USRP, the same
decimation would
yield a higher sampling rate.
I would appreciate it if someone could clarify this point for me.
Hi Daniel,
that README refers to things of the past (when you specified a
decimation rather than the sampling rate you want), and “USRP” refered
to what we nowadays call the USRP1.
In the version of the usrp_rx_hrpt.grc, from which the .py gets
generated, the USRP source is just configured to 4MS/s.
The question is whether 4MS/s is really the optimum rate; I haven’t
studied the HRPT signal enough to answer this. In principle, for the
100MHz devices (N210/N200), 4MS/s is a suboptimal rate, since the
decimation (25) is odd, and I’d try with something less strange, like
5MS/s.