Learning Forum

[jasper]

Hello,

I simulated the following bragg grating:

(Sketch:)

(Setup in Lumerical:)

Since the sections are both single mode, the mode convergence sweep works out fine. I get the following results:

The spectrum has a typical bragg grating form, however the FWHM and Reflectance are too high compared to measurements and simulations with other programs. For example, for a corrugation depth of 0.075µm, the FWHM is about 2.3nm when it should be about 0.7nm and the reflectivity is approximately 1 when it should be significantly lower (0.7). I have used the same geometry setup and the same # of periods. The grating period was calculated using the approximation via the average neffs of the fundamental TE modes of each cell: (neff=sqrt((neff_l1^2)*(1-duty_cycle)+(neff_l2^2)*duty_cycle). These are my EME settings:

Could this perhaps be caused by the integrated wavelength sweep feature in EME? What is going on here?

I'd appreciate some help.

Thanks,

Jasper

[Guilin Sun]

It might be one reason. The built-in wavelength sweep does not count the dispersion, eg, it uses the neff of the initially calculated modes at the specified wavelength.

Other reason might be:

1: the geometrically averaged effective index is a approximation

2: need proper number of modes

In addition, the group span is set to 0.718143um. It may not matter much but I usually suggest to use fewer digits so to be easier to compare with experiment. 1nm length might be difficult to control in manufacturing.