Regarding IE3D EM Simulation of CPW

[cs.gautam]

Dear Friends,

I am designing Microstrip and Coplanar circuits using IE3D EM software.
I am using Lin-gauge in order to calculate the parameters like impedance, angle, width, Effective E.
When, i am calculating the same parameters through TLN parameters in IE3D then i am getting different results.
In case of Microstrip, results are same but in case of coplanar results are different.
I am using the differential ports (Advanced Extension/ Extension for MMIC) for coplanar.

I am giving you one example:

In order to design 50 ohm CPW on Rogers/Duroid (substrate thickness=.635, Er=10.8, Metal thickness=35um)
I calculated the W & S using lin gauge. The values are: W (width of the central Conductor) = 0.4mm and S (Spacing between two ground planes)=0.21.

Then i did the simulation through IE3D software, Then i calculated the value of line impedance through TLN Parameters and it came 67 ohm instead of 50 ohm.

What is the reason behind it? i do not know. I also calculated TLN Parameters at low frequencies (Less than 3 GHz).

 

[WimRFP]

In low Z CPW without ground plane, air gap between center conductor and ground plane is small, so you get current density concentration at the edges of the traces Did you use edge width option and did you view the meshing before simulation?

you may determine Z0 based on a piece of line (quarter wave) and two ports. In the smith chart display, you can change the termination impedance. When the termination impedance (both ports) equals Z0, you virtually have S11, S22 = 0.

 

[cs.gautam]

Thanks for your reply.
Are you talking about IE3D Software?
It seems that you are talking about ADS Software.

 

[WimRFP]

Yes I am talking about IE3D. In the simulation dialog you can select the width of the segments at the open edge of a structure. This enables the simulator to solve non-uniform current distribution that frequently appears at edges of structures without using a small mesh over the complete structure. In newer versions you can use AEC (automatic edge cells with more cells in parallel to get better accuracy).

 

[cs.gautam]

Dear Friend,

I am using IE3D v 15. So, AEC is available. i have gone upto 3 layer.
My question is related to impedance measurement difference between linegauge and TLN of IE3D.
Do you have any idea about Asymmetric coplanar waveguides?
I also want to measure the impedance of asymmetric coplanar waveguides. Due to unavailability of line-gauge
for asymmetric coplanar, i am not able to do.

 

[WimRFP]

I never used a TLN function in IE3D.

When I simulated impedance of arbitrary structures, I used the two port method as mentioned before, and some other methods, but these are more elaborate. Did you also also include the thickness of the traces in IE3D? When trace thickness is no longer very small with respect to the air gap between center trace and ground traces in same layer, trace thickness will reduce Z0.

Regarding asymmetric CPW, you can put them in the simulator as any other transmission line structure.

Regarding AEC, did you look at the meshing to make sure it is reasonable (that means fine mesh at edges and course mesh in the center?

I forgot to mention that you can use ATLC (http://atlc.sourceforge.net) to calculate Z0 of arbitrary structures (so also for your asymmetrical CPW). You need to spend some time, but it gives useful results.