Hi all, I see an interesting trend and I'm not sure how to explain it ... I've done a y-factor measurement, sweeping the frequency, sampling rate, and rx-gain with an x310 with the SBX-120 daughterboard. The results seem similar to the published performance specs. The part I'm not sure quite how to explain is why the NF would vary with the sampling rate. Doesn't the X310 use a static sample rate and downsample in the FPGA? Why would this affect the NF? Why does the effect seem to be more exaggerated at higher gain settings? Why is the effect on NF very small at higher sampling rates but more pronounced at lower ones instead of being a linear change? Dustin

On 03/03/2025 22:23, Dustin Widmann via USRP-users wrote: > Hi all, > > I see an interesting trend and I'm not sure how to explain it ... > > I've done a y-factor measurement, sweeping the frequency, sampling > rate, and rx-gain with an x310 with the SBX-120 daughterboard. The > results seem similar to the published performance specs. The part I'm > not sure quite how to explain is why the NF would vary with the > sampling rate. Doesn't the X310 use a static sample rate and > downsample in the FPGA? Why would this affect the NF? Why does the > effect seem to be more exaggerated at higher gain settings? Why is the > effect on NF very small at higher sampling rates but more pronounced > at lower ones instead of being a linear change? > > Dustin > What noise inputs are you using for the two levels in your Y-factor? The amount of power represented at each sample-rate is different, and shooting from the hip here, the amount of energy   represented in the transition regions at different sample rates will be different.  For example, odd/even/factor-of-4   sample-rates have different pass-band shapes.

For an ADC, I believe the noise power is fairly constant. As sample rate increases, noise power density should decrease and SNR should improve. Thus a measured improvement in noise figure. Source: https://www.analog.com/en/resources/technical-articles/noise-spectral-density.html x310 ADC (ADS62P48): http://www.ti.com/lit/ds/slas635b/slas635b.pdf Mark On Mon, Mar 3, 2025, 7:48 PM Marcus D. Leech <patchvonbraun@gmail.com> wrote: > On 03/03/2025 22:23, Dustin Widmann via USRP-users wrote: > > Hi all, > > > > I see an interesting trend and I'm not sure how to explain it ... > > > > I've done a y-factor measurement, sweeping the frequency, sampling > > rate, and rx-gain with an x310 with the SBX-120 daughterboard. The > > results seem similar to the published performance specs. The part I'm > > not sure quite how to explain is why the NF would vary with the > > sampling rate. Doesn't the X310 use a static sample rate and > > downsample in the FPGA? Why would this affect the NF? Why does the > > effect seem to be more exaggerated at higher gain settings? Why is the > > effect on NF very small at higher sampling rates but more pronounced > > at lower ones instead of being a linear change? > > > > Dustin > > > What noise inputs are you using for the two levels in your Y-factor? > > The amount of power represented at each sample-rate is different, and > shooting from the hip here, the amount of energy > represented in the transition regions at different sample rates will > be different. For example, odd/even/factor-of-4 > sample-rates have different pass-band shapes. > > > _______________________________________________ > USRP-users mailing list -- usrp-users@lists.ettus.com > To unsubscribe send an email to usrp-users-leave@lists.ettus.com >

 On 03/03/2025 22:23, Dustin Widmann via USRP-users wrote:

 part I'm

 is the

 What noise inputs are you using for the two levels in your Y-factor?

 The amount of power represented at each sample-rate is different, and
 shooting from the hip here, the amount of energy
    represented in the transition regions at different sample rates
 will
 be different.  For example, odd/even/factor-of-4
    sample-rates have different pass-band shapes.


 _______________________________________________
 USRP-users mailing list -- usrp-users@lists.ettus.com
 To unsubscribe send an email to usrp-users-leave@lists.ettus.com

On 03/03/2025 22:59, Mark Gannet wrote: > For an ADC, I believe the noise power is fairly constant. As sample > rate increases, noise power density should decrease and SNR should > improve. Thus a measured improvement in noise figure. > > Source: > https://www.analog.com/en/resources/technical-articles/noise-spectral-density.html > > x310 ADC (ADS62P48): http://www.ti.com/lit/ds/slas635b/slas635b.pdf > > Mark Thanks.  Trying to get my head around this a bit.  The X310 runs the ADC at a fixed rate, and host-directed sample rates are   based on the actions of a DDC filter/sample-rate converter. > > > On Mon, Mar 3, 2025, 7:48 PM Marcus D. Leech <patchvonbraun@gmail.com> > wrote: > > On 03/03/2025 22:23, Dustin Widmann via USRP-users wrote: > > Hi all, > > > > I see an interesting trend and I'm not sure how to explain it ... > > > > I've done a y-factor measurement, sweeping the frequency, sampling > > rate, and rx-gain with an x310 with the SBX-120 daughterboard. The > > results seem similar to the published performance specs. The > part I'm > > not sure quite how to explain is why the NF would vary with the > > sampling rate. Doesn't the X310 use a static sample rate and > > downsample in the FPGA? Why would this affect the NF? Why does the > > effect seem to be more exaggerated at higher gain settings? Why > is the > > effect on NF very small at higher sampling rates but more pronounced > > at lower ones instead of being a linear change? > > > > Dustin > > > What noise inputs are you using for the two levels in your Y-factor? > > The amount of power represented at each sample-rate is different, and > shooting from the hip here, the amount of energy >    represented in the transition regions at different sample rates > will > be different.  For example, odd/even/factor-of-4 >    sample-rates have different pass-band shapes. > > > _______________________________________________ > USRP-users mailing list -- usrp-users@lists.ettus.com > To unsubscribe send an email to usrp-users-leave@lists.ettus.com >

50-ohm termination for cold ~15.5 dB ENR Noise Source for hot I've attached the scripts and the data, in case they should somehow prove userful. Dustin On Mon, Mar 3, 2025 at 10:49 PM Marcus D. Leech <patchvonbraun@gmail.com> wrote: > > On 03/03/2025 22:23, Dustin Widmann via USRP-users wrote: > > Hi all, > > > > I see an interesting trend and I'm not sure how to explain it ... > > > > I've done a y-factor measurement, sweeping the frequency, sampling > > rate, and rx-gain with an x310 with the SBX-120 daughterboard. The > > results seem similar to the published performance specs. The part I'm > > not sure quite how to explain is why the NF would vary with the > > sampling rate. Doesn't the X310 use a static sample rate and > > downsample in the FPGA? Why would this affect the NF? Why does the > > effect seem to be more exaggerated at higher gain settings? Why is the > > effect on NF very small at higher sampling rates but more pronounced > > at lower ones instead of being a linear change? > > > > Dustin > > > What noise inputs are you using for the two levels in your Y-factor? > > The amount of power represented at each sample-rate is different, and > shooting from the hip here, the amount of energy > represented in the transition regions at different sample rates will > be different. For example, odd/even/factor-of-4 > sample-rates have different pass-band shapes. > > > _______________________________________________ > USRP-users mailing list -- usrp-users@lists.ettus.com > To unsubscribe send an email to usrp-users-leave@lists.ettus.com