[Discuss-gnuradio] gnuradio filtering

[Eric Hill Matlis]

Hi-

I'm not sure how to implement a band-pass filter.  I have an am-modulated 
waveform with the carrier at about 2.858 MHz, and modulations between 1 
kHz and 5 kHz.  I wish to filter out anything above 6.5 kHz, and also 
remove 60-hz noise I believe I am seeing; preferably I would like to 
band-pass between say 300 Hz and 6500 Hz.

However, while the low-pass filter seems to work, the band-pass doesn't 
seem to be working as expected.  I seem to loose most of my signal.  I'm 
not sure how to describe what's going on, so I attached two figures 
showing each filter's response.  These figures are also available at:

www.nd.edu/~ematlis/z.gnuradio/am_lpfilter.jpg
www.nd.edu/~ematlis/z.gnuradio/am_band_pass_filter.jpg

My code is:

        adc_rate = self.u.adc_rate()                # 64 MS/s
        usrp_decim = 250
        self.u.set_decim_rate(usrp_decim)
        usrp_rate = adc_rate / usrp_decim           # 256 kS/s
        chanfilt_decim = 16
        demod_rate = usrp_rate / chanfilt_decim     # 16 kHz
        audio_decimation = 1
        audio_rate = demod_rate / audio_decimation  # 16 kHz


        if options.rx_subdev_spec is None:
            options.rx_subdev_spec = pick_subdevice(self.u)

        self.u.set_mux(usrp.determine_rx_mux_value(self.u, 
options.rx_subdev_spec))
        self.subdev = usrp.selected_subdev(self.u, options.rx_subdev_spec)


        #chan_filt_coeffs = gr.firdes.band_pass (1,           # gain
        chan_filt_coeffs = gr.firdes.low_pass (1,           # gain
                                            usrp_rate,   # sampling rate
                                            #3e2,
                                            6.e3,        # passband cutoff
                                            100,       # stopband cutoff
                                            gr.firdes.WIN_HANN)
        self.lpfilter =  gr.fir_filter_ccf 
(chanfilt_decim,chan_filt_coeffs)

        self.magblock = gr.complex_to_mag()
        self.volume_control = gr.multiply_const_ff(self.vol)



        # now wire it all together
        self.connect (self.u, self.lpfilter)
        #self.connect (self.hpfilter, self.lpfilter)
        self.connect (self.lpfilter, self.magblock)

        if plot2:
            self.post_filt = fftsink.fft_sink_f (self, self.panel, 
title="AM Demodulated FFT", fft_size=4096, sample_rate=audio_rate)
            self.connect (self.magblock,self.post_filt)
            vbox.Add (self.post_filt.win, 4, wx.EXPAND)

        if plot3:
            self.scope = scopesink.scope_sink_f(self, self.panel, 
title="AM Demodulated Time Series", sample_rate=audio_rate)
            self.connect(self.magblock, self.scope)
            vbox.Add (self.scope.win, 4, wx.EXPAND)

thanks!

eric

am_lpfilter.jpg
Description: JPEG image

am_band_pass_filter.jpg
Description: JPEG image