USRP2 phase drift

I still get phase drift even with atomic external clock connected and
configured in software with config_mimo(MC_WE_LOCK_TO_SMA).

I tried with GPS signals and a sinusoid from the function generator as
the inputs. There is phase drift in both cases. The attachment shows a
picture of the phase error fluctuation. Does this problem cause by the
100 MHz main oscillator on the main board?

On 07/26/2010 11:22 AM, senlin peng wrote:

I still get phase drift even with atomic external clock connected and
configured in software with config_mimo(MC_WE_LOCK_TO_SMA).

I tried with GPS signals and a sinusoid from the function generator as
the inputs. There is phase drift in both cases. The attachment shows a
picture of the phase error fluctuation. Does this problem cause by the
100 MHz main oscillator on the main board?

Even if you lock the USRP2 to an external reference, there will still be
drift relative to other things not connected to the same reference. The
function generator has its own clock, as does GPS.

Matt

Hi Matt

Even if you lock the USRP2 to an external reference, there will still be
drift relative to other things not connected to the same reference.

Can you give me an example? The PLL on the daughter board is also
connected with the reference clock. I hope to know which part is not
connected to the reference clock.

The error has the same trend for all satellites. So I think the error
is not caused by the clock difference between the local clock and the
GPS clock.
Is there any method that I can decrease the phase drift in the USRP2
output?

The
function generator has its own clock, as does GPS.

Matt


Best Regards!

On 07/26/2010 03:26 PM, senlin peng wrote:

Hi Matt

Even if you lock the USRP2 to an external reference, there will still be
drift relative to other things not connected to the same reference.

Can you give me an example? The PLL on the daughter board is also
connected with the reference clock. I hope to know which part is not
connected to the reference clock.

ALL of the parts are locked to the reference clock.

The error has the same trend for all satellites. So I think the error
is not caused by the clock difference between the local clock and the
GPS clock.
Is there any method that I can decrease the phase drift in the USRP2 output?

You are seeing REAL phase shift between your reference and the signals
you are looking at. Atomic clocks are good but nothing is perfect.

Matt

On 07/26/2010 04:00 PM, Matt E. wrote:

Even if you lock the USRP2 to an external reference, there will still
be drift relative to other things not connected to the same
reference. The function generator has its own clock, as does GPS.

Matt

Somewhat disturbingly, I’ve also observed in the lab phase-hits between
two GPS receivers, being fed from
the same antenna! I suspect differences in the phase locking
algorithms or something. But you can’t
even rely on GPS to be consistent with itself (at least viewed by two
different GPS receivers!!).


Marcus L.
Principal Investigator
Shirleys Bay Radio Astronomy Consortium
http://www.sbrac.org

The transmission line can act as a phase shifter depending upon how
long it is and how the GPS signal is being split.

~Jeff

On 7/26/2010 5:54 PM, Marcus D. Leech wrote:

even rely on GPS to be consistent with itself (at least viewed by two
different GPS receivers!!).


~Jeffrey L., K1VZX

This forum is not affiliated to the Ruby language, Ruby on Rails framework, nor any Ruby applications discussed here.

| Privacy Policy | Terms of Service | Remote Ruby Jobs