Hi Yeo,
On Tue, 2018-05-08 at 08:06 +, Yeo Jin Kuang Alvin (IA) via USRP-
users wrote:
> Hi all,
>
> Theoretically, in frequency domain, the spectrum for SQUARE wave
> should be a SINC
No. That's wrong. The spectrum of a periodic function, such as the
square wave, must be discrete.
In fact, the square wave has a known decomposition into cosines of odd
multiple frequencies of the square wave frequency – wikipedia has the
correct formula.
> and for the spectrum of RAMP wave should be decreasing with every
> odd harmonics.
I think a ramp is a sawtooth wave, or do you mean triangular wave?
In general, all these square-wave-convolved-with *something* waves have
decreasing energy in higher-frequency spectral lines.
> A question that I want to ask is, will I get to see these frequency
> spectrum in the Spectrum Analyzer after being transmitted out from
> the USRP B210?
How would you ever see them perfectly? The frequency domain of a
continuous signal isn't sampling-frequency periodic, whereas everything
in DSP is. So, that's not how the math works.
> As I have tried to send a Square and a Ramp wave, I don’t see any
> nice expected waveforms.
Phew! I thought you might have broken the math! But if you only see
things that look like they've been sampled and windowed, then things
are alright.
> Is it because of the IQ Modulation in the AD9361 of the USRP B210
> that causes me not seeing the expected results? Or is it because of
> some configurations not done properly? Hope to get a reply soon!
It's because you're not observing the continuous-time Fourier transform
– you really can't. Sampling a continuous signal and observing it for a
finite amount of time doesn't allow for that, no matter what device you
use or how you configure it. Only through applying a-priori knowledge
of your signal (for example, knowledge of your periodicity, to make the
sampling period an integer factor of that) would allow you to get the
correct line spectra a periodic function has. But you weren't even
expecting those.
Best regards,
Marcus
___
USRP-users mailing list
USRP-users@lists.ettus.com
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com