Learning Forum

[Songyu Yao]

Happy new year!
I was trying to optimize my 2x2 mmi in EME to be lossless and 50:50 splitting ratio. However, the propagation profile looks uncontinuous. I was first thinking the input tapers were not long enough which lead to changing modes. After considerably increasing the taper length,  the results didn't become better. I think the mesh are small enough and for EME as well. So what might be the reason of this problem?

[Amrita Pati]

Hi Songyu,

Happy New Year! Would you be able to share a screenshot of the "EME setup" tab? And a screenshot of the XY and XZ view of the actual structure would also help. The discontinuity could be due to reasons such as the nonavailablity of enough modes for the expansions of fields. But I just want to make sure the geometry and the EME object has been setup correctly before we move on to troubleshooting.

Regards,
Amrita

[Songyu Yao]

Hi Amrita,

 

Thanks for your kind reply! I attached these screenshots here, but I'm afraid that the XY and XZ views are not that clear. Anyway, the MMI is Si3N4 material, which consists of two pairs of input and output waveguides and tapers, and a MMI body, on the top of a SiO2 layer and a Si substrate. The number of modes for EME solver is 20.

 

Kind regards,

Songyu

[Amrita Pati]

Hi Songyu,

Thanks for sharing the screenshot. I believe the EME setup looks correct. As you previously menioned, the field distribution in the y-direction looks pixelated. You seem to be using a mesh override region, so I will ask you to reduce the mesh size. Also, what do the other results, like the total transmission look like? Do they look reasonable. I will also recommend you to perform convergence testing with regard to parameters such as the number of modes for expansion, the number of cells in the taper sections etc. I will suggest you to go through the following articles for more details on the convergence testing:

1.EME Convergence Testing: An Intuitive Approach

2. Convergence testing process for EME

Regards,
Amrita