Help about the simulation of Zo of CPW at low frequency

[jslee]

The simulatioin results is something wroing at low frequency.The Zo will increase rapidly as the frequencydecreases.I dont how to get correct answer.
May you help me? Thank you

 

[Supaswing]

well, I dont have your data but if your structure is something like Metal-SiO2-Si this behavior should be correct. There is some old work about this, check slow wave mode in google.
By the way, how did you do your simulation, which software/method?

 

[flyhigh]

Hi,

not able to view your structure right now. You talking about CPW and your file is named MicrostripLine?

As I saw that you are using HFSS, it might be the case that your ports are too small for the frequency you are analysing the structure. Try increasing the port area as microstrip (and CPW) EM field occupy more and more area as the frequency decreases. If your port is too small, you are getting much smaller equivalent capacitance of the structure thus your Zo increases.

The term low frequency is a bit etheric, numeric value in GHz might be more informative.

If the main aim of your simulation is to calculate Zo, there are much simpler methods, and even closed-form formulae are very accurate.

flyhgih

 

[rautio]

If you have a lossy line, it is entirely possible that your low frequency results are OK. When R per unit Length becomes about equal to ωL per unit length (as ω becomes small), the Zo becomes complex and very large at low frequency.

If your CPW is lossless, then the large Zo at low frequency is wrong. I have seen this happen with some kinds of interpolation.

If you are analyzing grounded CPW, you might have a microstrip mode, in addition to the CPW mode. This can really mess things up. Make sure your microstrip impedance is very high (i.e., the ground plane is a long ways away).

I assume you are using the planar solver of HFSS to do this problem. If you are using the volume meshing solver, you should switch immediately to the planar solver. If the planar solver gives you bad answers, you should switch to another planar solver. A good free one is SonnetLite (www.sonnetsoftware.com, I work for Sonnet). The interpolation works much better than any of the others out there (we do something very different from all the other guys) and you should be getting answers within one or two hours of downloading (this includes going through the Help->Tutorial). There are even CPW examples included.