[USRP-users] TwinRx transfert function
Marcus D. Leech
patchvonbraun at gmail.com
Wed Nov 6 12:49:33 EST 2019
On 11/06/2019 12:28 PM, Olivier Ravard wrote:
> 1. The tune frequency is 1200 MHz in this case but we obtain same order of magnitude
> for fluctuations if another frequency is tuned (but transfert function could be different).
> Our tests includes L band acquisitions (950 - 2150 MHz).
> 2. The gain in this case is set in order to have 15 dB backoff (which is margin relative
> to the
> maximum ADC amplitude). In this case, we use 85dB gain because the noise generator gives
> a level around -135 dBm/Hz. But we observed the same things when we sample satellite
> signals using 65dB gain.
> 3. what do you mean by "overruns".
Also, How was this spectrum actually capture (in the software sense)?
I'll note that 80Msps is not a valid delivered sample-rate for the
X310 with any of the standard master clocks, so to capture 80Mhz of
(analog) bandwidth, you'd need a 100Msps sample rate, which IS
> Le 06/11/2019 à 17:59, Marcus D Leech a écrit :
>> That is more passband ripple than I would expect.
>> What is your tunes frequency. What gain setting? Are you getting overruns during the
>> Sent from my iPhone
>>> On Nov 6, 2019, at 6:32 AM, Olivier Ravard via USRP-users <usrp-users at lists.ettus.com>
>>> We use X300 with 2 TwinRx boards for acquisitions in L band (UHD 3.10.3).
>>> It seems that the transfert function of the couple (X300 - TwinRx) shows strong
>>> To measure these fluctuations, we put a white noise generator at the input (this
>>> generator was
>>> measured before with a spectrum analyser and it show a very flat transfert function
>>> over 1 GHz)
>>> and we measure the PSD of the sampled 80MHz bw signal for the 4 inputs.
>>> The result is shown on figures below:
>>> inputs 1 and 2 (first TwinRx):
>>> inputs 3 and 4 (second TwinRx):
>>> Are these fluctuations normal ?
>>> USRP-users mailing list
>>> USRP-users at lists.ettus.com
More information about the USRP-users