Given that samples conceptually are anywhere i the range {-1.0,1.0} why would
you not expect some zero-valued samples?
Consider a normalized sine wave. Does it ever cross the zero line?
Also if you’re getting a lot of maximum valued samples, your input power is too
high and/or your gain is too high.
Sent from my iPhone
> On Apr 22, 2021, at 2:33 PM, Glenn Hazelwood <[email protected]> wrote:
>
>
> I am using UHD 3.15 on a Jetson Xavier with Ubuntu 18.04.4 LTS with a n Ettus
> B205mini SDR. I am developing my application in C++. My cpu_format is 'fc32'
> (complex 32-bit floats) and my over-the-wire-format is sc16 (complex 16-bit
> ints).
>
> My application is receiving complex samples with one channel. I tune to a
> frequency, wait for the LO to settle and then recv() to get the samples like
> in the UHD example, rx_multi_samples.cpp. But then I increment the tune
> frequency, tune to the new frequency, wait for the LO again and then recv()
> again until I have received samples from all of the tuning frequencies of
> interest.
>
> Is it normal to get zero-valued samples even after waiting for the LO to
> settle?
>
>
> P.S. Bonus issue:
> Also when I plot the samples, I see some of the samples (real and complex
> parts) go to 1.0 and -1.0, which is the max value for several samples. I set
> the gain to 50.
>
> --
> Diftor heh smusma
> -Famous Vulcan Phrase ;)
> _______________________________________________
> USRP-users mailing list -- [email protected]
> To unsubscribe send an email to [email protected]
_______________________________________________
USRP-users mailing list -- [email protected]
To unsubscribe send an email to [email protected]
