On 09/13/2019 01:01 PM, Lapointe, Benjamin - 0333 - MITLL via USRP-users
wrote:
To provide more details:
-The two USRP devices each have their own GPSDOs.
-I have separate TX and RX programs. Each one synchronizes to their
GPSDO and then synchronizes phase by creating streamers with a
time_spec with a specified gps_second in the future, following the
instructions in the manual.
The issue is: I am seeing a random phase offset in the received data
each time I run my programs. I don’t see a phase offset if I specify
a zero Hz center frequency, so there might be a clue there. For my
application I need to specify a non-zero center frequency.
I’m just trying to find out whether this result is expected with the
HW and/or SW setup that I have, or whether it is a bug in my current
version of UHD, or I am doing something wrong? Has anyone ever tried
to synchronize two USRPs each with their own GPSDO using non-zero
center frequencies?
Someone mentioned off-list that there was a bug in earlier versions of
UHD that caused the CORDIC not to get reset but that has been fixed in
newer versions. I’m running UHD 3.15.0.
Someone else asked off-list whether I am using the same or separate
GPSDOs. The answer was separate GPSDOs, but does that make a difference?
Thanks,
-ben
Fine-scale phase-noise will absolutely be different between two GPSDOs,
even on the same antenna. They'll also sometimes experience phase-slips
with respect to each other. I've seen this in the lab many times.
The difference here is that at zero frequency, the CORDIC doesn't spin.
So this likely has to do with the initial state of the CORDIC. I'm not sure
that the CORDIC fixes have been back-ported to N210--not a lot of
"new" work on the N210 device in the last couple of years.
_______________________________________________
USRP-users mailing list
USRP-users@lists.ettus.com
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
