What to do with S-parameter causality/passivity problem

Discussion created by weiming.qian on May 25, 2012
Latest reply on May 28, 2012 by weiming.qian



I use HyperLynx v8.2 to extract the S-parameter model of a differential channel on Xilinx Virtex-7 Dev board. However, the Touchstone Viewer tells me that the model is non causal. What should I do with this non causal model? Is there a way to fix it? Or can I ignore it as non-causal behaviour only happens on return loss curve, which I do not care too much.


The non-causal model and its mixed-mode model are attached.


Cheers, Weiming