#2 The "Tmet" parameter is ignored for coupled microstrips

[Anonymous]

When feeding the same parameters into a similar program (transcalc.sourceforge.net) I get the exact same results if and only if I specify Tmet=0 in both calculators.

Thus I draw the conclusion that Tmet is ignored and assumed to be zero.

[Dan McMahill]

it appears you are correct. Thanks for the report I will fix this. I guess wcalc-1.1 will be somewhat short lived :(

[Dan McMahill]

https://github.com/dmcmahill/wcalc/pull/22 has added some correction for Tmet. I also think it remains to be fully verified. I was looking at the original references and also at what transcalc did. I think they may have a factor of 2 bug in one of their equations for the correction (one of the references uses space between traces as "2s" as opposed to just "s"). Unfortunately the primary reference for coupled microstrip does not give full details on how the thickness correction should be applied. Transcalc just applies the correction to the calculation of static effective dielectric constant and characteristic impedances but not to the dispersion equations. I ended up doing the same but am not convinced one way or the other as to which is more accurate.

I have some work in progress in figuring out how to use Free FEM for doing some model validation although I'm only using that to solve a 2-D electrostatic problem and so I won't be able to get dispersion from that. However, I should be able to at least look at metal thickness effects in the static case. I'm still sorting out how to best make use of that tool but this all can be part of a different discussion over on github.