[USRP-users] LO frequency offset

Francois Quitin fquitin at ulb.ac.be
Fri Jun 8 18:20:05 EDT 2012

Hi Nick, 


Thanks a lot for your answer ! It makes much more sense now. Does it also
work on the USRP N200 to select “external” reference without connecting a
reference (and not just the N210) ?


Is there another way to bypass the digital correction steps of the TCXO ? 


Thanks a lot, 




De : Nick Foster [mailto:nick at ettus.com] 
Envoyé : vendredi 8 juin 2012 14:46
À : Francois Quitin
Cc : usrp-users at lists.ettus.com
Objet : Re: [USRP-users] LO frequency offset


The behavior is due to the TCXO used in the N200 series. Small TCXOs
typically use digital correction steps; the step size in this particular
TCXO is relatively large. You can avoid the problem by using an external
reference. You can also configure the N210 to select the "external"
reference, but don't connect a reference. In this case, your absolute
frequency error will go up several orders of magnitude (comparable to the
USRP2), but you will not see the "jumps".



On Fri, Jun 8, 2012 at 2:41 PM, Francois Quitin <fquitin at ulb.ac.be> wrote:

Dear all, 


I’m designing a frequency synchronization algorithm on our USRP-N200 with a
WBX daughterboard, and I’ve been observing some strange behavior with it. So
I’ve run the following test: 


I use one USRP-N200 to transmit a constant pilot tone, and I use another
USRP-N200 as receiver. I store the received pilot tone to a file, and
post-process this data to extract the LO frequency offset  (Center frequency
of 964 MHz, sample rate of 200kHz). In some cases there is no problem (first
picture), the LO frequency offset changes over time, but nothing surprising.
However, sometimes (second picture), the LO frequency offset “jumps” several
dozen of Hz. 


I asked a colleague from another university to run the same test with their
USRPs-N200, and they have observed the same behavior: the LO frequency
offset sometimes “jumps”. Even stranger, when I use our older USRP-2, I
never observe this behavior. 


Has this happened to someone before? Have I broken something on our boards?
Or is there some flaw in the clock distribution on the USRP-N200? 


Thanks a lot, 


François Quitin, Ph.D.


BAEF Post-doctoral research fellow

Electrical & Computer Engineering Dept. 

University of California

Santa Barbara, CA 93106-9560

Phone: +1-805-883-8599 <tel:%2B1-805-883-8599> 

Email: fquitin at ece.ucsb.edu

Home: www.ece.ucsb.edu/~fquitin 


USRP-users mailing list
USRP-users at lists.ettus.com


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/attachments/20120608/6a4fe7bf/attachment-0002.html>

More information about the USRP-users mailing list