Trying to design a CPW in HFSS

First let me say I'm very inexperienced as far as electronics goes (I’m a physics student) so be nice

I'm having some issues trying to design a simple CPW in HFSS, i searched the site up and down and found many many threads on this but none were able to help me I even downloaded the PowerPoint that is floating around here and didn't find much help in that either.

My issue is the Z0 is calculated to be 51 ohms but in hfss it’s around 39 ohms and then after about 14 GHz it shoots off to 1000ohms!

I have made the wave ports 3 times the width of the waveguide and the exact same issue.

What would be a great help is if someone knows of an example CPW HFSS file (or maybe someone quickly makes one?) that I can use to compare and figure out where I have went wrong

Other than the size of the wave port is there any other considerations i have to take into account with the CPW that perhaps I have overlooked?

As I said I’m very inexperienced in this so please explain what you mean if your using HFSS terms

Thanks and any help will be greatly appreciated.

Do you really need EM analysis for your CPW work?
And if you do, why do you want to use HFSS instead of a planar EM solver?
A planar solver like Momentum, Sonnet, IE3D is optimized for planar problems and much easier to handle.

I see HFSS used all over the place here, for the most trivial things, but most users seem to have no idea what they are doing and get all sort of crazy results.

Well the uni has HFSS so we kinda just used that, are any of those free?

We do need an EM sim for the problems though we plan to compare simulation and physical results.

Might stick them on my laptop.

Johnt447 said:

are any of those free?

Click to expand...

No, these are all commercial products, similar to HFSS. They are very popular for PCB work because they are specialized = more efficient for this type of circuits.

I asked about the need for EM because with HFSS, you can do PCB work if planar EM solvers tools are not available, but it will take an expert user and a lot of time. Maybe you can get away with circuit models instead?