MIxed-mode conversion calculations

[Gabriel Cardona]

I'm plotting an s8p file generated from Ansys HFSS, consisting of 2 differential pairs (8 ports total). When I plot the mixed-mode response for SDD21, for example, in VNA tools, the results are slightly different than what HFSS shows. I believe the VNA tools, when calculating the mixed mode response for each differential port is using the full 8x8 s-parameter matrix, thus the SDD21 response includes the response from ports 5-8, rather than just ports 1-4, which is what HFSS is calculating. I believe this is incorrect, because the second differential pair (ports 5-8), while there may be some coupling to the first (ports 1-4), this coupling should not be included in the differential response because the second pair is not necessarily correlated to the first pair.

Has anyone seen this behavior?

Thanks

 

[Michael Wollensack METAS]

> I believe the VNA tools, when calculating the mixed mode response for each differential port is using the full 8x8 s-parameter matrix

Yes that's right.

See Appendix C Mixed-Mode S-Parameters of Math Reference of VNA Tools.

If you want to see the mixed mode S-parameters only of port 1-4 or 5-8 take a subset before.