Learning Forum

[kosala.dhanawansha]

Hello,

As the "spectrum vs wavelength" graph shows, the source has a relatively high bandwidth even if we select a single wavelength of 1.55um. We also use an (n,k) material with the particular n,k values ​​only corresponding to the 1.55um wavelength in the simulation. Doesn't it make any inaccuracy when we get the Pabs (normalized value wrt the source power) at 1.55um wavelength as the source injects energy other than 1.55um in to the simulation volume.

[Guilin Sun]

no, it works fine. Since FDTD is a time domain method, using a pulse is the best way to simulate fast.

However if at this wavelength there is strong resonance you may not know it. I would suggest that you use two monitors. One is for this desired wavelength, the mother uses braoder say1.54~1.56um with 20 points to know if there is resonace (you will need to override the global setting). In case there is, you will see if the broader result is smooth arouf 1.55um. If not, you will need to increase simulation time and/or decrease autoshutoff min. The simulation autoshutoff level is averaged for the whole pulse which is broadband.