RE: [Discuss-gnuradio] Flipping problem

[Bob McGwier]

I agree with this analysis unless Isaac tells us he has a carrier and baud
synchronizer in his system.  If he is looking at raw output from the USRP,
indeed the two oscillators (as they would in any real system) shift
frequency over time and drift in and out of any phase relationship.

Bob


ARRL SDR Working Group Chair
Member: ARRL, AMSAT, AMSAT-DL, TAPR, Packrats,
NJQRP, QRP ARCI, QCWA, FRC.
"Trample the slow ....  Hurdle the dead"


-----Original Message-----
From: address@hidden
[mailto:address@hidden On Behalf Of
Matt Ettus
Sent: Tuesday, July 29, 2008 2:40 AM
To: isaacgerg
Cc: address@hidden
Subject: Re: [Discuss-gnuradio] Flipping problem

isaacgerg wrote:
> I am sending a known sequence of samples from one USRP to another using
the
> Basix RX/TX d'boards setting the frequency to 24e6.
>
> When I rx the sequence, the correlation of it keeps flipping, but not in a
> way that suggests residual carrier.  It seems as if I am experiencing an
> instantaneous flip.
>
> My gain on both the RX and TX is set to 500. The signal amplitude I rx is
> ~1200. When I run the GRC model, the Rx says I have no residual carrier,
> invert is set to false, the dxc frequency is set to -24e6 and the baseband
> to zero.  On the TX side, I have these same values with the exception of
the
> dxc freq, it is not set to 24e6. Why the flip??   What do all these things
> mean???
>
> Is 24e6 a bad freq to use with the basic TX/Rx?
>
> Thanks in advance!
>   

I believe what you are seeing here is that the two USRPs are on two 
different frequencies.  This is not a fault of the USRP.  No two 
oscillators will be on exactly the same frequency, and so in any 
practical receiver you must synchronize. 

Matt


_______________________________________________
Discuss-gnuradio mailing list
address@hidden
http://lists.gnu.org/mailman/listinfo/discuss-gnuradio