Hello Everyone: I am sweeping between 50MHz - 6GHz with UBX40 + N200. Whenever I cross fc = 1500MHz with this transceiver, however, I see a significant (>10dB) change in the noise floor. What is the cause, and what could be done to suppress (or characterize) such behavior? I understand that UBX40 switches its internal RF circuitries when it crosses 500MHz and 1.5GHz, and I understand that the filtering of the circuit can affect the noise floor, but I do not know the RF hardware design well enough to understand the exact issue. Is it being overloaded by radio stations and BTS? If so, why it is not overloaded anymore when it crosses 1500MHz boundary (I don't see any additional filters on the block diagram of the UBX40)? Regards, Kyeong Su Shin

Hello Kyeong Su Shin, The transition is caused by a change in the signal path. Above the 1.5 GHz frequency the low noise amplifier is switched from an MGA-62563 to a VMMK-3603. The Low Band LNA is followed by a low pass filter with a cutoff frequency of 1.7 GHz. It may be that you have a strong signal below 1.7 GHz which is causing an increased noise floor which is then removed. Here is the schematics of the UBX. http://files.ettus.com/schematics/ubx/ubx.pdf Is the change a decrease (improvement) in the noise floor above 1.7 GHz? Do you observe the same change if you remove your antennas and, if possible, attach 50 ohm terminators? That will isolate the cause to internal vs external. The noise figure for the UBX is almost level over frequency, and has no change at 1.5 GHz based on our test results. Page 64 has a good plot of Noise Figure and Gain over frequency. http://files.ettus.com/performance_data/ubx/UBX-without-UHD-corrections.pdf Regards, Derek On Wed, Jan 25, 2017 at 12:41 AM, Kyeong Su Shin via USRP-users < usrp-users@lists.ettus.com> wrote: > Hello Everyone: > > I am sweeping between 50MHz - 6GHz with UBX40 + N200. Whenever I cross fc > = 1500MHz with this transceiver, however, I see a significant (>10dB) > change in the noise floor. What is the cause, and what could be done to > suppress (or characterize) such behavior? > > I understand that UBX40 switches its internal RF circuitries when it > crosses 500MHz and 1.5GHz, and I understand that the filtering of the > circuit can affect the noise floor, but I do not know the RF hardware > design well enough to understand the exact issue. Is it being overloaded by > radio stations and BTS? If so, why it is not overloaded anymore when it > crosses 1500MHz boundary (I don't see any additional filters on the block > diagram of the UBX40)? > > Regards, > Kyeong Su Shin > > _______________________________________________ > USRP-users mailing list > USRP-users@lists.ettus.com > http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com > >

Hello Derek, Thank you for the quick response! Yes, this happens when and only when the USRP is connected to an antenna (outdoor discone, so very likely to be receiving strong broadcast or cellular station signals). Change occurs at 1.5GHz, and yes, the noise floor decreases after that frequency (GNURadio PSD plot screenshots attached). It is probably caused by external signal, but I was confused because that 1.7GHz low pass filter is only present for the sub-1.5GHz signal path and there was no such thing (high-pass filter) for above 1.5GHz (according to the block diagram). Could it be the LNA? Regards, Kyeong Su Shin On Wed, Jan 25, 2017 at 1:07 AM, Derek Kozel <derek.kozel@ettus.com> wrote: > Hello Kyeong Su Shin, > > The transition is caused by a change in the signal path. Above the 1.5 GHz > frequency the low noise amplifier is switched from an MGA-62563 to a > VMMK-3603. The Low Band LNA is followed by a low pass filter with a cutoff > frequency of 1.7 GHz. It may be that you have a strong signal below 1.7 GHz > which is causing an increased noise floor which is then removed. > > Here is the schematics of the UBX. > http://files.ettus.com/schematics/ubx/ubx.pdf > > Is the change a decrease (improvement) in the noise floor above 1.7 GHz? > Do you observe the same change if you remove your antennas and, if > possible, attach 50 ohm terminators? That will isolate the cause to > internal vs external. > > The noise figure for the UBX is almost level over frequency, and has no > change at 1.5 GHz based on our test results. Page 64 has a good plot of > Noise Figure and Gain over frequency. > http://files.ettus.com/performance_data/ubx/UBX- > without-UHD-corrections.pdf > > Regards, > Derek > > > On Wed, Jan 25, 2017 at 12:41 AM, Kyeong Su Shin via USRP-users < > usrp-users@lists.ettus.com> wrote: > >> Hello Everyone: >> >> I am sweeping between 50MHz - 6GHz with UBX40 + N200. Whenever I cross fc >> = 1500MHz with this transceiver, however, I see a significant (>10dB) >> change in the noise floor. What is the cause, and what could be done to >> suppress (or characterize) such behavior? >> >> I understand that UBX40 switches its internal RF circuitries when it >> crosses 500MHz and 1.5GHz, and I understand that the filtering of the >> circuit can affect the noise floor, but I do not know the RF hardware >> design well enough to understand the exact issue. Is it being overloaded by >> radio stations and BTS? If so, why it is not overloaded anymore when it >> crosses 1500MHz boundary (I don't see any additional filters on the block >> diagram of the UBX40)? >> >> Regards, >> Kyeong Su Shin >> >> _______________________________________________ >> USRP-users mailing list >> USRP-users@lists.ettus.com >> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com >> >> >