Re: FM Transmission with a vector of bits

[Jeff Long]

Use the "Frequency Mod" block as a frequency modulator to make things simpler.

The GUI Sink picture you show could be right. You're multiplying a 360 kHz square wave by another slower square wave and looking at the first harmonic.

On Mon, Aug 10, 2020 at 7:54 AM Anthony B. <anthonyld508@gmail.com> wrote:

Hello,

Thanks for your answer. I've taken a break on this issue and now going
back to it.

As you describe it, I created a Python class that generates my binary
sequence (which is a PCM Major Frame based on IRIG 106). Then I set the
data of the Vector Source to my binary sequence so that it moves it to
the downstream blocks. I don't really know if it's currently the right
way to go, as I am trying to get a feeling of the bitrate that goes over
the flowgraph and it just outputs as fast as it can. I therefore use a
Repeat Block to slow down the samples stream and knowing the sample rate
of the blocks downstream, I can get an idea of how much bits are
processed (that's what I am trying to apply).

In order to transmit it to a USRP device using FM modulation, I mostly
got inspiration from the RTTY transmitter that is based on the same
idea. It is described as an example of the VCO use here :

https://wiki.gnuradio.org/index.php/VCO

In my case, I settled the coefficients before the VCO in order to obtain
a 350 kHz deviation from the NRZ-L sequence I generated (-1 or +1). I
can visualize this deviation in the Spectrum plot after the VCO.
However, I am quite far from the theoretical spectrum from the IRIG 106
standard I'd get in a PCM/FM transmission. It is said that the optimal
deviation should be 0.35 * bitrate, however knowing the current bitrate
is kinda challenging in GNU Radio and in the theoretical spectrum it is
based on 1Mbps.

Currently I am trying to set the parameters in order to have a better
spectrum shape at the output, but I don't succeed to get something
workable. In my tests I am using a Vector Source feeding the downstream
blocks either with -1 or +1 as my binary sequence have.

Would you have suggestions on how to proceed ?

I've attached my current flowgraph, the Spectrum plot and the result I
am trying to converge to.

Thanks for your answers !

Anthony