Re: [Discuss-gnuradio] GMSK MOD question

[Tom Rondeau]

On Tue, Jun 12, 2012 at 4:03 PM, Vivian Paola Triana Galeano
<address@hidden> wrote:
> Hi Everyone,
> My name is VIvian and I am working with GMSK modulation. I have a little
> doubt about it.
>
> I am transmitting a cosine signal using UHD sink and source blocks, USRP1,
> daughter board RFX 2400. The spectrum that I transmit is the one that you
> can see in the attached files.
>
> I use a network analyzer to see the real spectrum that I am sending and the
> spectrum that I obtain can be seen in the  other photo that I attach.
>
> Finally I look the spectrum received in GNU Radio, which is attached to this
> email too.
>
> I found that the spectrum that I send is the half of the size that I see in
> the network analyzer. I mean the spectrum that I send has a bandwidth of 200
> KHz ( I see this in GNU Radio)  but the one that I obtain looking at the
> network analyzer is approx. 400 KHz. When i see the received signal in GNU
> Radio is perfect, its bandwidth is the same that I send ( 200 KHz). I dont
> know why I see a different bandwidth in the network analyzer. Anyone knows
> what is happening?
>
> I would appreciate any help you can give me. Thanks in advance.
>
> Vivian.

Hi Vivian,

Just remember that the GNU Radio scope plots don't have to tell you
the truth. Everything in GNU Radio really just runs off a unitless
sample rate of 1, so samples/second are actually meaningless to it. We
tell the scope what the sample rate is and it sets the x-axis scale
accordingly, but we can easily lie to it. My guess is that there is a
factor of two in the sample rate of the system that you aren't
accounting for when creating the plot.

Tom