I am having problems with critical coupling in an SOI ring resonator with waveguides 220 nm high by 500 nm wide.
I have tried this with both Palik silicon and setting a custom (n,k) material with a given propagation loss of 5 dB/cm (corresponds to a round trip transmission of 0.995 for this 83 um cavity).
My approach was to shrink the FDTD to just the bus and the linear coupler of the racetrack and use the output monitor to get the power transmission T, and then extract t = sqrt (T), which is the coupling coefficient of the ring. I varied the separation between the bus and the linear racetrack waveguides until I got t close to 0.995 - corresponding to a separation of 350 nm and a coupling length of 10 um. Then I made the varFDTD cover the entire ring and simulated.
I end up getting an odd spectrum, with oscillations in the pass band and a dip that drops to 0.75 transmission, which is certainly not critically coupled. Can someone look through and see if I have made any obvious mistakes, or if there is a bigger problem?
MackCavity 80um.lms (345.7 KB)