Error encountered while performing simulation of MZI- switch

[Dr.Sudip Roy]

Kindly let me know about the error encountered while I was performing the simulation of the MZI-Switch. I am not knowing how to rectify this error and the cause behind it. Please suggest me an appropriate way to rectify it and perform the simulation smoothly.

[Scott Newman]

Hello Dr. Roy,

It would appear from the error message that there is an issue with the mode solver used in the source injection. Would it be possible for you to attache the design file so I can take a look?

Scott

[Dr.Sudip Roy]

I have enclosed the design file, Mr.Scott Newman. Pls let me know about it as soon as possible

[Scott Newman]

I have reviewed your design and and the error and I have confirmed that my original statement was correct. The error you are encountering is because the mode solver has determined that the structure you have does not support a mode at 1.3um.

Looking into your design I see that your waveguide has a refractive index of 1.5 while the surrounding material has a refractive index of 2.1 which means that this structure will not support a mode and therefore why the simulation is terminating with the error.

Your waveguide should have a higher refractive index than the surrounding material.

Scott

[Dr.Sudip Roy]

Thank you for your prompt reply, Mr.Scott Neuman. I was able to resolve this issue. I am also facing another problem. While simulating the NOR-Gate, my output is getting distorted. Earlier I have faced the same problem while simulating MZI too. I am enclosing the screenshots and design files for your kind perusal. Please let know about it as soon as possible.

[Steve Dods]

Hello Dr. Roy,

I have observed NOR-gate and mzi-output.png. It is clear these designs will not function as NOR gate. These structures both have two couplers. To function as a NOR gate, both couplers must be lossless 3dB couplers. However, after BPM simulation it is clear these couplers are very far from the 3dB lossless spec. It is clear from nor-trial.png that the first coupler encountered transfers almost all optical power to the lower waveguide of the MZI. In order to acheive the NOR function, the power of the lower wavguide must cancel the power in the other one. This is not possible in the given design, almost all power transfers to the lower waveguide. Since there is no power in the upper waveguide, no cancellation is possible in the second coupler. function cannot be achieved. One can also see a great deal of radiation loss from the S Bend waveguides. The corners are too sharp for a diffused waveguide. Diffused waveguides are very weakly guiding and cannot tolerate much curvature
Regards, Steve

[Dr.Sudip Roy]

I am not knowing what is wrong with my design. Kindly let me know in detail what should I do to get the accurate output. I simulated it by seeing the layout of it, from a research paper in order to get familiar with the software. I have enclosed the research paper for your kind reference.

[Steve Dods]

The MZI and NOR gate require lossless 3dB couplers to function properly. The MZI consists of two such couplers connected in series. NOR-gate-output1.png and mzi-output1.png seem to have 2 couplers in the layout, as required.

The optical circuits shown in
Vol. 54, No. 28 / October 1 2015 / Applied Optics
each have two or three lossless 3dB couplers connected in series or some combination of series and parallel.

The first coupler in NOR-gate-output.png does not appear to be a ossless 3dB coupler. The power does seem to be approximately evenly divided on the output waveguides, but there is a great deal of radiation loss, evident from all the light observed outside the waveguides.

The first coupler in mzi-output1.png seems to have similar trouble. MZI and NOR rely on total cancellation from interference between different waveguides. this will not be possible if there is so much radiation loss. I noticed the waveguides in this project are diffused. Diffused waveguides cannot tolerate much bending. Even a small curvature in the waveguide arc causes significant loss.

I recommend experiments with arc waveguides untill the minimum radius of curvature is determined. All designs for couplers should use either straight wqaveguides, or, if S bends or arcs are used, the radius of curvature should be equal or greater than as determined above to avoid radiation loss. Once that is achieved, adjustment should be made to couple 3dB to each output waveguide, to get suitable design for a lossless 3 dB coupler. With such devices, he MZI and NOR illustrated in Vol. 54, No. 28 / October 1 2015 Applied Optics will be possible by connecting the couplers as shown in that paper.

[Author]

Posts