[SOLVED] Lumped Port Slotline Simulation in HFSS

[Zikloa]

Hi all, new user here. Hoping that I could find some help with simulating a slotline in HFSS. I've read many of the previous posts about this topic but they haven't been successful in helping me get a sensible simulation.

I think the main issue I'm having is positioning the lumped port. Where should this be placed on the substrate? How wide/tall should it be? Does it just need to touch each of the conductors? In any case, here's what I have right now:

Also, is a lumped port on a driven terminal solution the best way to obtain the s-parameters for a slotline design, or should I be considering another simulation setup?

Any help would be greatly appreciated, TYIA.

 

[PlanarMetamaterials]

Welcome, Zikloa.

Does your simulation setup have any other objects in it, other than what you've shown? There are numerous setup issues if that is all there is.

I find I get the best results using wave ports, but a lumped port should be ok. It should be placed between the two conductors (i.e., in the gap).

 

[Zikloa]

Hey, thanks for getting back to me!

The only other object besides those pictured is a vacuum box containing the model. A radiation boundary has been applied to the faces of this box. Two other boundaries have been applied, those being PerfE boundaries on each of the conductors.

Regarding the port, when you say it should be in the gap, that means on top of the substrate? If so, is the only relevant dimension the width (to ensure that it touches both conductors), or does the length (along x axis) also play a role?

 

[PlanarMetamaterials]

Regarding the port, when you say it should be in the gap, that means on top of the substrate?

Yes.

If so, is the only relevant dimension the width (to ensure that it touches both conductors)

Essentially, yes.

or does the length (along x axis) also play a role?

It will introduce some minor parasitics; there is no right answer, but I've found empirically that the best response seems to be from making the surface approximately square (length = width).

 

[Zikloa]

Okay, makes sense. One last question: you said that you prefer to use a wave-port for this type of measurement, why?

 

[PlanarMetamaterials]

Wave ports re-create the appropriate modal fields of only the desired mode, which is usually what your realistic excitation should be doing. The lumped ports will excite a spectrum of modes, which while realistic, is probably to a larger degree than you would typically find in a good excitation. Really, it's about modelling your real-world excitation to the best possible degree.