Is there a way/trick to force the results of an AMC component to be fully symmetrical? I guess not but I just want to ask to make sure we are not missing out on something. The thing is that I created a cross model but I noticed that some S-parameters are not identical (very close though) when they should. For example, for a cross with all widths the same, the phase of S13 is different from the phase of S24. Any input is welcome.