*Hi Josh, Thanks for replying. I have understood your answer and I wanted to refine my question which I previously asked. It is: *
> > On 10/10/2012 08:53 PM, Ashish Raste wrote: > > *Hi all, > > > > Recently, I tried testing how the data flows from one USRP to another in > a > > 2x2 setup. My configurations went as follows: > > > > 2 USRPs were connected via a MIMO. Each USRP was connected to an Ethernet > > cable and so both of them were in two different subnets. Now when I set > one > > of the USRPs clock source and time source to MIMO, I observed that both > of > > them were synchronized and believe that its acting as a slave for the > > Master's clock and time source. > > What I'm not able to understand is whether the slave is receiving its > data > > via the MIMO (from the Master) or via its own RF input i.e whether . I > > thought that modulating the RF input to both the USRPs would show some > > difference on the FFT plot but both were identical. Could any one please > > tell me how the data flow occurs by suggesting a good test case? > > If both usrps are attached to an ethernet cable, the data does not > travel over the mimo cable. > > *Revised Question: 2 USRPs were connected via a MIMO. Each USRP was connected to an Ethernet cable and so both of them were in two different subnets (This is the unshared Ethernet Mode) . I have a host PC which runs the GRC design. Now when I set one of the USRPs clock source and time source to MIMO, and the other USRP’s clock source and time source to default, I observed that both of them were synchronized and believe that one is acting as a slave for the Master's clock and time source.. What I'm not able to understand is the baseband data path between the USRPs and the host PC. That is, whether the slave’s baseband data within the GRC design is received/transmitted via its own Ethernet interface and the master’s basedband data within the GRC design is received/transmitted via its own Ethernet interface? So, this means that the MIMO cable is only used for sending the clock and time signals from the Master to the Slave ?* > If only one usrp is attached to an ethernet cable, the mimo cable acts > like a network switch. > > > > > My second question is: > > In a 4X4 USRP setup, how can we synchronize all the 4 USRPs without a > MIMO > > connection (between 2 USRPs within a pair), using a 10MHz reference given > > to each of the USRP and a PPS source? > > You will want to share a 10MHz ref and PPS signal to the two master > devices. The other two devices in the configuration can get reference > and time, each from a mimo cable. > > *Is there any way to synchronize these 4 USRPs without using MIMO, but only using the ref and PPS? Because I see that your suggestion on the setup is similar to the configuration depicted in the image attached with this mail.* *Also please note that my USRPs have GPSDO for referencing purpose, if needed.* > http://files.ettus.com/uhd_docs/manual/html/sync.html > > CHeers, > > -josh > > Looking forward to your reply. Thanks and Regards, -- Ashish
<<attachment: 4x4_MIMO_config.png>>
_______________________________________________ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio