Re: [Discuss-gnuradio] question in source code of gr_sig_source_f

[Eric Blossom]

On Thu, Aug 06, 2009 at 12:39:15PM -0700, Milo Wong wrote:
> Hi,
> 
> I was browsing the GNU radio C++ API. Here I have a couple of questions
> about the class reference of gr_sig_source_f. (Signal generator)
> 
> 1) For the type of public member function waveform(), i.e. gr_waveform_t
> waveform () const. Where can I refer to the definition of "gr_waveform_t"?

Hi Milo,

You can find it in gnuradio-core/src/lib/gengen/gr_sig_source_waveform.h

(Given that there were only 3 includes in gr_sig_source_f, it didn't
 take long to find it...  grep would have worked too)

  http://gnuradio.org/doc/doxygen/gr__sig__source__f_8h-source.html
  http://gnuradio.org/doc/doxygen/gr__sig__source__waveform_8h.html

> Is there any description of how waveforms(sine wave) are generated?

Use the source Luke!

  gnuradio-core/src/lib/gengen/gr_sig_source_f.cc
  

> 2)  I am going to generate a set of sine waves at different frequencies. If
> these sine waves are obtained from the iteration equations of harmonious
> oscillators:   v(t+dt)=v(t) - omega^2 * x(t)dt ; x(t+dt)=x(t) + v(t)dt, is
> that different from the way of sine wave generation in gnu radio. Do I need
> to create a new sine wave generation block instead of the defaut one? (x(t)
> - displacement, v(t) - velocity)
> 
> Thank you,
> Milo

No, there's no reason to reimplement the sine generator unless
you enjoy that kind of thing.  First thing to remember is that the
block does what it says: it generates a sine wave of the specified
frequency and amplitude.

We also have blocks to generate sinusoids with frequency a function
of its input stream.  See e.g, gr_frequency_modulator_fc or gr_vco_f

Eric