Hello,
I want to measure the phase difference between three synchronised USRP
channels (Common Ref & PPS), does software exist to do this, and to what
accuracy, please?
Sorry, this is a bit of a newbie question, and probably already covered!
Dave
On 17.06.2014 19:06, David M. wrote:
Hello,
I want to measure the phase difference between three synchronised USRP
channels (Common Ref & PPS), does software exist to do this, and to what
accuracy, please?
Sorry, this is a bit of a newbie question, and probably already covered!
Sure, the easiest way would simply be to put in a scope, and check by
looking – of course, that’s not very accurate.
There’s enough math operators in GNU Radio to see the actual phase
difference, such as conjugate and multiply, then convert to phase.
Cheers,
Martin
Hi David,
for starters, you could just divide one channel by the other and
calculate the result’s argument. There are blocks for both operations 
Short explanation:
Just consider the complex signals $s_1, s_2$ in polar coordinates
(instead of the usual real/imag representation)
$s_1(t) = a_1(t) * e^{j \phi_1(t)}$.
Dividing that by $s_2(t) = a_2(t) * e^{j \phi_2(t)}$ gives you
$p(t) = a_1(t)/a_2(t) * e^{j \phi_1 - \phi_2(t)}$, of which you can take
the argument to get the phase difference.
Greetings,
Marcus
I did an experiment similar to this a year or so ago. I synced all the
USRPs to the same 10MHz and PPS. I had pretty good results using libgps
to
get the GPS time and then using UHD to start streaming on a particular
PPS
edge on all units. With LFRX boards this will ensure that the DDCs start
at
the same time and start from zero degrees. I seem to remember there was
some small random variation from run to run. If you are just looking for
relative changes in phase difference I found this approach works fairly
well.
On Tue, Jun 17, 2014 at 10:36 AM, Martin B. [email protected]