Re: [Discuss-gnuradio] Questions on the usrp_fft and usrp_oscope

[Matt Ettus]

Quoting Sachi <address@hidden>:

> Hi, guys
>
> I met some confusions when I tried the usrp_fft and
> usrp_oscope.
>
> I generate a 2MHz, peak to peak 1V, sin wave using the
> Agilent signal generator, and connect it to usrp.
> However, the result of usrp_fft always shows a very
> strong component at DC, besides the 2MHz. Why is that?

There is DC offset.  I will explain this in a separate email.

> Then I tried usrp_oscope, it works pretty well. But
> when I zoom in, I found the wave form doesn't seem to
> be sine, but something like square or triangle. Do I
> miss any parameters here?

At 8 MS/s, a 2 MHz sine wave only has 4 samples per cycle, so it will look like
a triangle wave when you "connect the dots".  Set the DDC frequency t something
closer to your signal frequency, so it will do downconversion and you can see
more points per cycle.  For example,

sudo ./usrp_oscope.py -c 1.9M


Matt