On 08/02/2021 12:37 PM, Marcin Wachowiak wrote:
Hello,
I am working on a phase coherent application using USRP B210. As the
phase difference between RX channels of B210 is relatively stable I
wanted to see how it behaves across the whole frequency range. I
performed some measurements and unfortunately for some frequencies I
observed strange distortions. The setup consists of a TX USRP, RF
splitter, matched cables and a second USRP with both RX channels
connected to the splitter.
Screams documenting the distortions:
https://drive.google.com/drive/folders/1DkjrVKz3ywv3ZE0eS1UYeCMTtQwL3Du8?usp=sharing
The received sine wave should have quite similar shape and spectral
properties across both RX channels of USRP.
What I observe instead is a symmetrical harmonic tone at one of the
RX, when the other does not record it.
There are also random distortions in one of the RX channels looking as
if the sine signal was split and shifted at some point in time (looks
like a some sort of buffer issue)
This kind of distortion for some settings persists for some is not
present or occurs regularly.
To provide some illustiation I attach recordings and screens of the
registered distortions.
usrpb210_sine_rx_distoriton2.mp4
<https://drive.google.com/file/d/1Hif0dZY2Ah7ipC13nfudhoK5HD0JeM9M/view?usp=drive_web>
usrpb210_sine_rx_distoriton.mp4
<https://drive.google.com/file/d/13niyBdl6H4w0Akdf28KeontnZKUsCUaW/view?usp=drive_web>
I tried to debug this problem but it appears to be related to sample
rate, TX tone frequency, carrier frequency and gains. (I checked
dynamic IQ imbalance or gain controls and changing of the parameters
did not affect anything)
There are now underflow/overflow flags present while I run the
flowchart. The signal that is provided to both RX is identical ( RF
mini-circuits RF splitter), and any outside interferences are limited
as the whole setup is connected via concentric cables. I tried to set
the number of receive and sent frames to 1024 to ensure continuity,
but that didn't help.
It is strange that for some frequencies this phenomenon occurs while
for others the phase difference variance is very low and there are no
such distortions.
Could You please explain the cause or propose some kind of solution to
this issue?
Kind regards,
Marcin Wachowiak
_______________________________________________
USRP-users mailing list -- [email protected]
To unsubscribe send an email to [email protected]
What are your signal levels going into the RX USRP?
They need to be well below -20dBm or so to protect against
non-linearity, and even then, that's awfully loud.
_______________________________________________
USRP-users mailing list -- [email protected]
To unsubscribe send an email to [email protected]
