AnsweredAssumed Answered

MCS with 2xAD9361 (with different crystals)

Question asked by add.jky on Feb 11, 2018
Latest reply on Feb 12, 2018 by Vinod



I am trying to figure out a way to synchronize 2xAD9361 with MCS, but i have a restriction that binds me to use two different borads.


So say i have 2 different boards (different but identical), a master and a slave board, that each has an AD9361, a ZYNQ US+ and say 1ppm XO as a reference clock.


I have read that there are strict demands to use the same reference clock and to keep trace lengths the same.

say, i can make the trace lengths be the same so there won't be any skew (obviously there will be a phase offset and a tiny frequency offset because there are two XO's ).


The sync_in of the slave, will be coming from the master zynq, and i will make sure both the sync_in of the master and the slave will arrive to each AD9361 at the exact same time.


1. In this constellation, will the MCS process won't work at all? or maybe will it work with a constant error?

2. what if i use the same reference clock to provide clock to the zynqs, and then using MMCM in each fpga to distribute a clock to each AD9361, but still the master zynq will send the sync_in pulse to both AD9361 - will that work?


Any help will be much appreciated.