You can find the spot size converter example on our Knowledge Base. In this subsection, we
will be creating the same simulation file from scratch.
First, open the material database and confirm that the silicon material is present.
This will be used for the waveguide and taper. The materials for the rest of the structures
are dielectric and we can directly specify the refractive index in the structure object.
Next, add the substrate using a rectangle primitive. Edit the rectangle, and in the
Geometry tab, set the position and spans of the structure. In the Material tab, set the
index to 1.465. Add the input waveguide which is also a rectangle.
The waveguide will have a have a y span of 0.4 um and a z span of 0.2 um. Set the material
to silicon from the drop down menu. For the taper, you can use the isosceles trapezoid
from the object library. Edit the structure group and apply a rotation around the z axis
so that the orientation of the taper is along the x-axis. Set the position and spans of
the trapezoid. The lx top and lx base properties are the width of the taper at the output and
input. Set the material to silicon. Add a rectangle for the SiON waveguide. Set
the position and spans. Set the index to 1.5 and the mesh order to 3. The taper has a mesh
order of 2 so the region where the two objects overlap will be represented by the taper,
since it has a smaller mesh order. Now, add the EME solver object and set the
background index to 1.465. This will form a cladding above the waveguides. Then, set
the wavelength to 1.5 um. In the EME setup tab, set the starting x position
which will correspond to the left side of the solver region, and set the y, and z geometry of
the solver region. The x span of the simulation region will be decided by the group spans
you use. Set the number of cell groups to 3, and set the group spans in the cell group
definition table so that cell group 1 covers the uniform input waveguide, cell group 2
covers the taper, and cell group 3 covers the uniform output waveguide. Set the number
of cells to 1 in uniform regions and 19 in taper. You need to use CVCS subcell method
for taper region where the cross-section is smoothly varying. See the Convergence Testing section
for more details about how to determine the appropriate number of cells and modes to use.
Select the "display cells" option so the cells will be outlined in orange in the view ports.
The settings in the other tabs are same as the FDE solver, and we will use the default settings.
Once you set up the EME solver object, you can see that all the cells and ports are automatically
generated under the EME object. Since the Si waveguide and taper are relatively
small, it is a good idea to add a mesh override over them to apply a finer mesh in the transverse
directions. Set the mesh multiplier in the y and z directions to 5. Using a higher value here will result
in a finer mesh. You can define the mesh region directly or based on the structure.
Finally, in the ports, select "Fundamental TE mode" for each port as these are the input
and output modes of interest. With everything set up, click the Run button
in the main menu to find the modes in each cell. The Job Manager shows the simulation
progress. Actions in this step include: calculating modes for each cell, normalizing the modes
and calculating mode overlaps. Further details will be provided in the Solver Physics section.