[USRP-users] Signal harmonics in WBX output

Daniele Nicolodi daniele at grinta.net
Thu Oct 30 13:15:32 EDT 2014


Hello,

I'm experimenting with a WBX daughterboard connected to a N210 and
controlled via GNURadio. I'm sorry if my questions below are naive, but
I'm new to the field and I didn't find answers in the documentation.

What I obaserve, programming the system to output a single 80 MHz tone,
letting the UHD driver choose about the tuning, are quite strong signals
at the odd harmonics, where the 240 MHz harmonic is only 10 dB below the
80 MHz carrier. The even harmonics are much smaller, but their amplitude
increases quickly with gains above 20 dB: at 20 dB gain the 160 MHz
harmonic is at -60 dB from the carrier, at 30 dB gain it increases to be
-20 dB from the carrier.

Are those harmonics expected? There are ways to have the WBX generate
less undesired harmonics? What's the maximum recommended gain to
minimize those signal artifacts?

I've read on the documentation about the "mode_n=integer" parameter that
can be specified to the tune request, but it did not change what I
observe. When is this setting useful?

Things get more interesting when I step away from 80 MHz!

At 80 MHz + x Hz (with x small such that the frequency is obtained
tuning the WBX LO at 80 MHz and bridging the difference through DSP
synthesis) the even harmonic start to be amplitude modulated at
frequency x. Probably two close frequency start to beat with each other
producing the amplitude modulation.

The real trouble comes when in this configuration I increase the gain
above 20 dB: also the 80 MHz carrier starts to be amplitude modulated ad
frequency x. Of course this is not what I would like to see :)

Obviously I can keep the gain below 20 dB, add a low pass filter and be
happy with it.  However, I'm wondering if those symptoms are there also
at lower gain settings, and are simply too small do be observed on my
spectrum analyzer. For my application I require low phase noise, but
also good amplitude stability, therefore I would like to know what are
the best practices to avoid the observed artifacts.

Thank you a lot!

Cheers,
Daniele




More information about the USRP-users mailing list