How to simulate a solt line as a transmission line in HFSS?

EDAboard.com Forum Index -> Software Problems, Hints and Reviews -> How to simulate a solt line as a transmission line in HFSS?

Author

Message

Yanyan Zhang

Joined: 01 Jul 2006
Posts: 7

29 Mar 2007 4:25 slotline calculator

I want to use HFSS to simulate a slot line structure with the FR4 material as a transmission line to get the characteristic impedance and effective dielectric constant. The first question is about the port, should I use the wave port or the lumped port? Well, lumped port is better for the microstrip line structure, for the field is constrained around the strip; yet for the slot line, the field extends beyond the gap especially for a low dielectric constant material. The second question is how to simulate the free space enviroment. Is PML boundary better than the radiation boundary, considering to remove the wave reflection? Thanks for your advice!

Google
AdSense
Google Adsense

29 Mar 2007 4:25 Ads

Azulykit

Joined: 18 Feb 2007
Posts: 851
Helped: 103
Location: All over the Map.

08 May 2007 5:49 hfss transmission line

I have seen variations of this question a number of times now.

I have several recommendations. Look into the on-line help for HFSS for "Calculating Characteristic Impedance". That will suggest that you use a waveport adjusted to calculate Zpv which is the most appropriate form for a slot line transmission line. The section also points to background references if you want to look into the question further.

You will need to assign an impedance line for the calculation.

I usually like to examine the port field display, particularly if there is the possibility of higher order modes in your model.

I would surround the model with an airbox and a radiation boundary (assuming that is appropriate e.g. the line is not enclosed inside a box).

If you are concerned about the fields at the edge of the waveport not terminating correctly you can force it to a radiation boundary or you can increase the size of the port. Ypu should see signs of a problem in the port field display.

You might want to examine a number of modes to make sure you have the excitation you want.

I also like to sanity check the results with another tool. Polar Instruments is one choice and there are free calculators all over the web although finding one with a slotline may be a little challenging. I am not implying that Polar is a better tool than HFSS. Wadell and Simon each have books, each with collections of equations for transmission line impedance calculations. If you need the exact titles let me know and I will give you the details. The books are on a bookshelf 15 miles from where I am right now.

Yanyan Zhang

Joined: 01 Jul 2006
Posts: 7

08 May 2007 17:54 slotline transmission line

Hello, Azulykit!

Thank you so much for your advice! I basically set up the slotline model in HFSS according to your ideas. I use two ways to get and compare the characteristic impedance and the effective dielectric constant. One way is as you suggest, to use the wave port and the impdeance line to find the Zpv and the propagation constant Gamma. And the second way, I use the S parameter to calculate both quantities. And I find a book entitled " Microstrip lines and slotlines" by K.C.Gupta. etc. This book provides some closed-form equations to calculate both quantities.

I feel that my simulation results are somewhat not stable. For the port set up, the port size has a great impact on the simulation. The length and width and the distance away from the airbox which is added the radiation boundary seem to influence the results greatly. I find if I set the wave port entirely cover the cross section of the soltline, and set the suitable distances away from the airbox, the simulation results agree with the theoretical ones to a large degree. Yet by checking the field distribution on the port , I notice that such large wave port leads to the appearing of higher order modes. If I reduce the size of the port, the result is not good. I need to adjust the airbox again.

Cohn's paper "Slot line on a dielectric substrate" says the slot line mode is non-TEM mode in nature, so I feel really puzzled about the slot line mode . Maybe there is some reflection from the radiation boundary, and the airbox sizes thereby influence very much the simulation. I tried to use the PML boundary (one layer), but not much improvement. Besides, the width of the metal strip has a big affect. The width can not be too small, otherwise it becomes the coupled strips and not the slotline structure. When the width gets larger, the simulation result changes greatly. I think once the width is big enough, the simulation should be stable.

I do think the simulation of the slot line in HFSS is more challenging. Yes, as you sugest. I need to use another tool to have a try. Thanks again for your time and help! I really appreciate it very much!