Empathy List Archives

usrp-users@lists.ettus.com

Discussion and technical support related to USRP, UHD, RFNoC

View all threads

Re: [USRP-users] spectral leakage ?!

Marcus D. Leech

Mon, Oct 5, 2015 4:58 PM

On 10/05/2015 11:39 AM, Amina M. Fellan wrote:

Thanks a lot Marcus and James, that fixed it :)

I have one more question, when I transmit a BPSK signal with the
benchmark_tx.py script I get the spectrum in the attached photo. As
can be seen, there is about 5 dB dip of the received signal's power at
the center frequency compared to the center_freq +/- 25 KHz, is that
normal ?

Cheers,
Amina

It is expected with a BPSK signal for there to be somewhat more energy
in the sidelobes than closer to the notional carrier
frequency.

Date: Fri, 2 Oct 2015 10:24:07 -0400
From: mleech@ripnet.com
To: james.humphries@ettus.com
CC: afellan@hotmail.com; usrp-users@lists.ettus.com
Subject: Re: [USRP-users] spectral leakage ?!

You can use offset tuning to move the DC anomaly out-of-band. Both in
the TX and RX direction:
http://files.ettus.com/manual/page_general.html
All superhet TX/RX have some amount of LO leakage. With
direct-conversion (Zero IF), that leakage happens to be in-band. So,
you can make this more like a non-zero-IF design using offset tuning.
There will still be leakage, but it will be out-of-band.

On 2015-10-02 09:55, James Humphries via USRP-users wrote:

 Some LO leakage is typical with the system. When you say sometimes
 not necessarily at the center frequency, how far away from center
 is it?
 -Trip

 On Thu, Oct 1, 2015 at 7:53 AM, Amina M. Fellan via USRP-users
 <usrp-users@lists.ettus.com <mailto:usrp-users@lists.ettus.com>>
 wrote:

     Hello everybody !
     I'm using 2 USRPs N210 equipped with SBX daughterboards with
     GNU Radio to establish BPSK communication over ISM band. For
     the latter, I started with the benchmark_tx/_rx.py examples.
     currently I'm trying to measure the SNR at the receiver. I
     tried using the SNR estimation blocks in the GRC but they seem
     to be giving erroneous readings, so I decided to try with a
     spectrum analyzer to estimate the SNR using (S+N)/N.
     I noticed that even when there's no transmission going on, so
     trying to measure the noise floor, a signal appears on the
     spectrum and sometimes even not necessarily at the center
     frequency. This also can be seen using the uhd_fft script from
     GNU Radio. Attached is a photo of how the spectrum looks if
     the USRP is on without transmitting/receiving !!
     Any clues what is causing this problem ??
     Thanks for your time !

     Cheers,
     Amina
     *
     *

     _______________________________________________
     USRP-users mailing list
     USRP-users@lists.ettus.com <mailto:USRP-users@lists.ettus.com>
     http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com


 _______________________________________________
 USRP-users mailing list
 USRP-users@lists.ettus.com  <mailto:USRP-users@lists.ettus.com>
 http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com

On 10/05/2015 11:39 AM, Amina M. Fellan wrote: > Thanks a lot Marcus and James, that fixed it :) > > I have one more question, when I transmit a BPSK signal with the > benchmark_tx.py script I get the spectrum in the attached photo. As > can be seen, there is about 5 dB dip of the received signal's power at > the center frequency compared to the center_freq +/- 25 KHz, is that > normal ? > > Cheers, > Amina > It is expected with a BPSK signal for there to be somewhat more energy in the sidelobes than closer to the notional carrier frequency. > * > * > > > ------------------------------------------------------------------------ > Date: Fri, 2 Oct 2015 10:24:07 -0400 > From: mleech@ripnet.com > To: james.humphries@ettus.com > CC: afellan@hotmail.com; usrp-users@lists.ettus.com > Subject: Re: [USRP-users] spectral leakage ?! > > You can use offset tuning to move the DC anomaly out-of-band. Both in > the TX and RX direction: > http://files.ettus.com/manual/page_general.html > All superhet TX/RX have some amount of LO leakage. With > direct-conversion (Zero IF), that leakage happens to be in-band. So, > you can make this more like a non-zero-IF design using offset tuning. > There will still be leakage, but it will be out-of-band. > > On 2015-10-02 09:55, James Humphries via USRP-users wrote: > > Some LO leakage is typical with the system. When you say sometimes > not necessarily at the center frequency, how far away from center > is it? > -Trip > > On Thu, Oct 1, 2015 at 7:53 AM, Amina M. Fellan via USRP-users > <usrp-users@lists.ettus.com <mailto:usrp-users@lists.ettus.com>> > wrote: > > Hello everybody ! > I'm using 2 USRPs N210 equipped with SBX daughterboards with > GNU Radio to establish BPSK communication over ISM band. For > the latter, I started with the benchmark_tx/_rx.py examples. > currently I'm trying to measure the SNR at the receiver. I > tried using the SNR estimation blocks in the GRC but they seem > to be giving erroneous readings, so I decided to try with a > spectrum analyzer to estimate the SNR using (S+N)/N. > I noticed that even when there's no transmission going on, so > trying to measure the noise floor, a signal appears on the > spectrum and sometimes even not necessarily at the center > frequency. This also can be seen using the uhd_fft script from > GNU Radio. Attached is a photo of how the spectrum looks if > the USRP is on without transmitting/receiving !! > Any clues what is causing this problem ?? > Thanks for your time ! > > Cheers, > Amina > * > * > > _______________________________________________ > USRP-users mailing list > USRP-users@lists.ettus.com <mailto:USRP-users@lists.ettus.com> > http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com > > > _______________________________________________ > USRP-users mailing list > USRP-users@lists.ettus.com <mailto:USRP-users@lists.ettus.com> > http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com >