add dispersion model for comparison and use a few new API improvements

[dmarek-flex]

Big change is adding the dispersion model from scikit-rf. I also changed to a 3D PEC for the strip, since our conformal meshing is much improved.

Also added section describing the different definitions of impedance commonly used.

• Need to rerun with the next version of the solver (2.8.4 has some bugs which cause the notebook to error)

[github-actions]

Spell Check Report

CharacteristicImpedanceCalculator.ipynb:

880: "[4]   R. H. Jansen and M. Kirschning, “Arguments and an accurate Model for the Power-Current Formulation of Microstrip Characteristic Impedance”, Archiv für Elektronik und Übertragungstechnik (AEÜ), vol. 37, pp. 108-112, 1983.\n",
	Archiv ==> Archive

---

Checked 1 notebook(s). Found spelling errors in 1 file(s).
Generated by GitHub Action run: https://github.com/flexcompute/tidy3d-notebooks/actions/runs/15742913183

[daquinteroflex]

Looks good! Would it be possible to use the skrf prefix instead of rf because that way it's easier to work out what's external to tidy3d since rf and our microwave would appear to be from the same package.

[weiliangjin2021]

Looks great!

• "Internally, the ImpedanceCalculator computes impedance using ..." maybe change it to mention that it is one of the 3 approaches
• For Gaussian pulse, how is the result if it's constructed from from_frequency?
• Could you comment how the 3 methods compared to skrf result?

[dmarek-flex]

For Gaussian pulse, how is the result if it's constructed from from_frequency?

Basically no change

[dmarek-flex]

Could you comment how the 3 methods compared to skrf result?

Done, I think I found a better way to compare the methods and explain the differences when compared to scikit-rf. I want to still finalize the last cells to show a calculation of a lossy transmission line.

[weiliangjin2021]

For Gaussian pulse, how is the result if it's constructed from from_frequency?

Basically no change

In that case, better to use from_frequency to promote this approach?

[dmarek-flex]

For Gaussian pulse, how is the result if it's constructed from from_frequency?

Basically no change

In that case, better to use from_frequency to promote this approach?

Ah, ok! It did change the FieldTimeMonitorData, so I'll need to edit those plots.

[tylerflex]

should we merge this?

[dmarek-flex]

should we merge this?

I was waiting for the next version of the solver to rerun, so I can remove the temporary solver_version. I think I can do that with the current pre release.