Connecting models that don't have the same z0

[Nita E]

Hello everyone,

Myself and a few others are currently trying to simulate a power integrity model using skrf. The model is similar to what's shown in this image.

We recently hit a roadblock in development. Being a PI model, there are frequently mismatches and such since the characteristic impedance of power traces/planes tends to be very low.

If we wanted to for example connect a capacitor model (imported from a touchstone file) z0=50 with a board model (also from a touchstone file) z0=0.1, what would be the "correct" impedance settings? Additionally, what happens inside the code when there is an impedance mismatch?

We currently use Spice to verify skrf outputs and have not been able to find a solution for this yet.

Help is appreciated.
Thanks.

[Julien Hillairet]

Hi,

When connecting two Networks of different port impedances, the mismatch is handled by inserting a 2-port ‘mismatch’ network between the two connected ports. This mismatch Network is calculated with the impedance_mismatch() function.

So after, it depends on your desired behaviour. You may have to renormalize some of your networks before connecting them (cf. `renormalize()` method).

Hope it is more clear,

Best regards,

Julien

--
You received this message because you are subscribed to the Google Groups "scikit-rf" group.
To unsubscribe from this group and stop receiving emails from it, send an email to scikit-rf+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/scikit-rf/6039d301-3e15-465b-97e5-431003346820n%40googlegroups.com.

[Nita E]

Hi Julien,

Turns out that renormalization is just part of the issue, the main issue comes from failure to convert s parameters to z parameters. Problem mostly comes up when simulating touchstone files provided by capacitor manufacturers. The s2z method ends up dividing by numbers very close to zero so we end up with very large and meaningless values for the impedance matrix.

I have no idea if there are any alternative approaches to the conversion, but a warning or something similar that indicates some converted values might be off would be very useful in the future.

Thank you very much for your help.

Regards,

Nita