Simulating s-parameters for a directional coupler

[rizamox]

Hello
I’m running the attached simulation in order to extract S-parameters using FDTD for the shown structure. The simulation is using a combination of input source (pointing down), field monitors, and mode expansion monitors to simulate the S-parameters. The results obtained from mode expansion for frequency monitor at port1 shows that Port1.a and Port1.b only contain 0+0i, and hence S11(=Port1.a/Port1.b), S21(=Port2.b/Port1.b), S31(=Port3.a/Port1.b), and S41 have Indefinite values because a division by 0 occurs. I have checked all the design parameters to make sure there is nothing wrong. Everything looks fine. This problem only happens at port1. What do you think? What can cause this problem?

Through power at center wavelength (Input1): -1.#IND%.
Cross power at center wavelength (Input1): -1.#IND%.
Loss (Input1): -1.#IND%.
Through power at center wavelength (Input2): 72.1603%.
Cross power at center wavelength (Input2): 7.6635%.
Loss (Input2): 20.1762%.

DC_Sparam.lsf (21.3 KB)
materials.lsf (1.2 KB)
port1ain1.txt (2.9 KB)

[gbaethge]

Hi @rizamox,

You are using the same mode expansion monitor for Port1 and Port3. it is correct since both correspond to identical waveguides. However, the way the mode expansion monitor is set, it does the mode expansions of the fields from the monitor using their absolute coordinates. As Port1 is largely shifted from Port3, there is 0 coupling so a and b are equal to 0, hence the undefined results.

You can either add another mode expansion monitor for Port1, or check the box “align to frequency monitor center”, in the mode expansion tab of the mode expansion monitor. This will shift the center of the mode expansion monitor to align with the center of the frequency monitor. You just need to make sure the monitors’ and mode expansion monitors’ center are aligned with the center of the waveguide they cover. This way, the mode expansion should work correctly.