Edge (fringing) capacitance in Sonnet and Momentum

[alexneverhurts]

I'm simulating the capacitance of a HEMT in off-state, and the capacitance values I got from Maxwell 2D (only simulated the cross-section), Sonnet, and Momentum are different from each other, and Maxwell 2D's result matches the experimental value very well. I wonder if 2.5D EM simulators like Sonnet and Momentum can calculate the fringing capacitance (capacitance from the edge of an electrode on top layer to the edge of an electrode on the layer below, while the to electrodes do not overlap) very accurately. Will it be helpful to shrink mesh size on the edge?

Thanks!

 

[Manjunatha_hv]

Hello alexneverhurts,

You should get the accurate fringing capacitance values from Sonnet, have you subsectioned the structure properly & entered correct values of Metal, Dielectric layer properties...
upload your sonnet project file (*.son), we can suggest you by looking into the model...


---manju---

 

[Joe_User]

There is also a published technique for quantifying the error in an EM simulator for a capacitor. See this paper if interested:

Erik H. Lenzing and James C. Rautio, "A Model for Discretization Error in Electromagnetic Analysis of Capacitors," IEEE Transactions on Microwave Theory and Techniques, Vol. 46, No. 2, February 1998, pp. 162-166.

 

[alexneverhurts]

Thank you! Here is my Sonnet model. In fact Sonnet gave me the lowest capacitance of all three methods. Perhaps my subsection was incorrect.

 

[Joe_User]

I think you will find that if you decrease your cell size (just in the Y direction), the capacitance will increase. I tried this with a simplified version (removed CPW feeds) and the capacitance increased from ~0.04 pF to ~0.05 pF. I used a cell size of 5 um by 0.25 um.

 

[Manjunatha_hv]

Hello alexneverhurts,

I modified your sonnet project file & I got better results compared your previous file...As I suspected it is due to less no. of subsections in Y direction...
See the attached Sonnet project file & the results picture....

Also I have observed you are not set the exact conductor material properties such as conductivity & thickness & also the Dielectric material each layer Loss tanget,
this may have some more effects on your Capacitance calculations...I recommend you to enter these values & see the results in Sonnet & measurement it should agree!!!


---manju---

 

[alexneverhurts]

Thank you Manjunatha_hv, and all other kind helps! Just one last question for Manjunatha_hv: I notice that you got capacitance value directly from the simulation result. Can I know how you did it? I always export the S21 and used another math program to fit it, and it was quite tedious. Thanks again!

 

[Manjunatha_hv]

Hello alexneverhurts,

Yes see the attached Ploting Equations using Sonnet response Viewer.pdf document for simple steps to plot the capacitance in Sonnet response viewer...

Also you can do lots of postprocessing in Sonnet Response Viewer (emgraph) itself,
see the below document (you can write your own equations to calculate different parameters from sonnet S/Y-parameter results without exporting files...
Application Note - Equations in the Sonnet V9 Response Viewer
**broken link removed**

---manju---

 

[alexneverhurts]

This thing rocks!