On Fri, Jun 19, 2020 at 4:57 AM Fernando <ferna...@samara.com.es> wrote:
> In the GRC Frequency mod block the doc says > > > analog_frequency_modulator_fc > make(float sensitivity) -> frequency_modulator_fc_sptr > Frequency modulator block. > float input; complex baseband output > Takes a real, baseband signal (x_m[n]) and output a frequency > modulated signal (y[n]) according to: > > Where x[n] is the input sample at time n and is the frequency > deviation. Common values for are 5 kHz for narrowband FM channels such > as for voice systems and 75 KHz for wideband FM, like audio broadcast FM > stations. > In this block, the input argument is , not the frequency > deviation. The sensitivity specifies how much the phase changes based on > the new input sample. Given a maximum deviation, , and sample rate , the > sensitivity is defined as: > Constructor Specific Documentation: > Build a frequency modulator block. > Args: > sensitivity : radians/sample = amplitude * sensitivity > > > There is somejhing missing at: " Where x[n] is the input sample at > time n and ____________ is the frequency deviation" > > > Here https://wiki.gnuradio.org/index.php/Frequency_Mod the same is > missing too. > > I guess it refers to sensitivity = K = > (2*pi*freq_deviation)/samp_rate correct? > > > Second question related to this one. > > If I the use Quadrature_demod for demodulation of the signal, the gain > parameter will be the inverso of this one isnt it? > > > gain=samp_rate/((2*pi*freq_deviation) but when I insert the quadrature > demod block it is autofilled with > > samp_rate/(2*math.pi*fsk_deviation_hz/8.0) where does the 8.0 come from? > > This would give full scale deviation with 4x oversampling, I think. So, this might be set by default for a digital mode. > > best regards > > > > > > > > > > >
