resonator

[Poonam Jindal]

Hi,

I have just created a resonator between two line defects and simulate the structure at 1550nm. The Ey plot is attached here for reference. can u please suggest that why am I getting this type of response. I mean why is there so much scattering at the input port and why the propagation inside the cavity and waveguide not smooth?

[Scott Newman]

Hello Poonam,

It is a bit tricky to diagnose the issue from just that one view. Are you able to share the fdt file?

Scott

[Scott Newman]

Hello Poonam,

It is a bit tricky to diagnose the issue from just that one view. Are you able to share the fdt file?

Scott

[Poonam Jindal]

Hi Scott

Thanks for the response.
I am attaching the fdt file for your reference.

[Poonam Jindal]

Hi Scott

Thanks for the response.
I am attaching the fdt file for your reference.

[Scott Newman]

Poonam,

The design you have included has no observation areas, are you certain that is the design you are using to get those results? Because I am unable to duplicate your field data. I do note a couple of things:

1. Why is your injected field 1.414 um when that waveguide is only about 0.6 um? This gives very poor overlap between the waveguide mode and your field which means much of the field will be dumped into the crystal where it cannot propagate.
2. Why is your source inside the crystal? If you inject from outside the crystal you either get power in the waveguide or it does not enter at all.
3. You seem to have both 2D and 3D simulation parameters set, which simulation are those results from?
4. You are using the default resolution (lambda/10) which is in my opinion too coarse for your structure with a source of 1.55 um. Have you done any convergence testing?
5. Finally I moved the input plane outside the crystal and changed the DFT specifications to be from 0.8 um to 3 um. What I observed from the observation area that there is couple into the waveguide from about 0.8 um up to about 1.4 um. It is my assessment that the waveguide in question does not support modes with wavelengths beyond about 1.4 um as it is too narrow.

Scott

[Scott Newman]

Poonam,

The design you have included has no observation areas, are you certain that is the design you are using to get those results? Because I am unable to duplicate your field data. I do note a couple of things:

1. Why is your injected field 1.414 um when that waveguide is only about 0.6 um? This gives very poor overlap between the waveguide mode and your field which means much of the field will be dumped into the crystal where it cannot propagate.
2. Why is your source inside the crystal? If you inject from outside the crystal you either get power in the waveguide or it does not enter at all.
3. You seem to have both 2D and 3D simulation parameters set, which simulation are those results from?
4. You are using the default resolution (lambda/10) which is in my opinion too coarse for your structure with a source of 1.55 um. Have you done any convergence testing?
5. Finally I moved the input plane outside the crystal and changed the DFT specifications to be from 0.8 um to 3 um. What I observed from the observation area that there is couple into the waveguide from about 0.8 um up to about 1.4 um. It is my assessment that the waveguide in question does not support modes with wavelengths beyond about 1.4 um as it is too narrow.

Scott

[Poonam Jindal]

Hi Scott
Thanks a lot for your response.
I have seen all the suggestions and will work upon it. Further, I have some doubts as follows-
1. Regarding the relation between injected field and waveguide dimensions.
2. How to adjust the mesh size (resolution) in relation to the source wavelength.
3. What changes in the waveguide parameters I need to do to get the result beyond 1.4um (else suggest the relation between them).

I will be very thankful to you as this will be really helpful in my research.
Hoping for a favorable reply.
Poonam

[Poonam Jindal]

Hi Scott
Thanks a lot for your response.
I have seen all the suggestions and will work upon it. Further, I have some doubts as follows-
1. Regarding the relation between injected field and waveguide dimensions.
2. How to adjust the mesh size (resolution) in relation to the source wavelength.
3. What changes in the waveguide parameters I need to do to get the result beyond 1.4um (else suggest the relation between them).

I will be very thankful to you as this will be really helpful in my research.
Hoping for a favorable reply.
Poonam

[Scott Newman]

1. I am not sure what your concern is with the injected field and waveguide dimensions. If your field is much larger than the waveguide and you inject inside the crystal and some of those wavelengths are in the bandgap then you are dumping power into the crystal that cannot go anywhere which will be problematic. If you want to use the larger field inject from outside.
2. When you open the 2D or 3D Simulation dialogs (seen in the tree on the left side of the product) you have options to change the “Mesh Delta” for X/Z (2D) as well as Y (3D). You have the auto check on these so the mesh will be automatically set to lambda / 10n where n is the highest index material in use. For your design this corresponds to only 4-5 points across one of your rods and only 2-3 points between rods. You will want to run a series of simulations with smaller mesh sizes to see how the mesh affects the results. There will come a point where the results stop changing.
3. This is not a simple question as there are a broad number of variables here such as lattice spacing, rod radius that also impact your band structure. Additionally you could look at removing more than one row or using different radii for the rows that are bordering the waveguide. There are many papers out there on design options for PhC waveguides.

Scott

[Scott Newman]

1. I am not sure what your concern is with the injected field and waveguide dimensions. If your field is much larger than the waveguide and you inject inside the crystal and some of those wavelengths are in the bandgap then you are dumping power into the crystal that cannot go anywhere which will be problematic. If you want to use the larger field inject from outside.
2. When you open the 2D or 3D Simulation dialogs (seen in the tree on the left side of the product) you have options to change the “Mesh Delta” for X/Z (2D) as well as Y (3D). You have the auto check on these so the mesh will be automatically set to lambda / 10n where n is the highest index material in use. For your design this corresponds to only 4-5 points across one of your rods and only 2-3 points between rods. You will want to run a series of simulations with smaller mesh sizes to see how the mesh affects the results. There will come a point where the results stop changing.
3. This is not a simple question as there are a broad number of variables here such as lattice spacing, rod radius that also impact your band structure. Additionally you could look at removing more than one row or using different radii for the rows that are bordering the waveguide. There are many papers out there on design options for PhC waveguides.

Scott

[Poonam Jindal]

Hi Scott,

Thanks for the response.

1. I have done the 2D simulation for the designed structure.
2. I have also placed the input source outside the structure.
3. As per your suggestion, I have broaden the input waveguide by removing a row of rods. But again a lot of power is leaking out from the waveguide as shown in the attached file.
4. I have also simulated the structure at different mesh sizes like 0.04, 0.038, 0.01 etc. but getting almost same results each time.

I am getting very confused. Please guide me on what am I doing wrong?

[Poonam Jindal]

Hi Scott,

Thanks for the response.

1. I have done the 2D simulation for the designed structure.
2. I have also placed the input source outside the structure.
3. As per your suggestion, I have broaden the input waveguide by removing a row of rods. But again a lot of power is leaking out from the waveguide as shown in the attached file.
4. I have also simulated the structure at different mesh sizes like 0.04, 0.038, 0.01 etc. but getting almost same results each time.

I am getting very confused. Please guide me on what am I doing wrong?

[Scott Newman]

I am just looking at your design and will get back to you.

[Scott Newman]

I am just looking at your design and will get back to you.

[Poonam Jindal]

Ok Sir,

I am also attaching the jpeg file of analyzer.

[Poonam Jindal]

Ok Sir,

I am also attaching the jpeg file of analyzer.

[Scott Newman]

There would appear to be some discrepancy between what you have sent me and where you got those results from. Either that or you are using a much older version of the product. I ran your design and at 1550 nm um while it does not propagate across the resonator as seen in yours it is certainly well confined (see 1550nm.PNG). The worst results I could get from your simulation were for 1682 nm which is outside the bandgap you are using so it is not surprising (see 1682nm.PNG).

Is the design you sent in fact the same one you used to get those results? Note that the one you sent did not have an observation area covering the whole XZ space as shown in your results so some changes were made.

It also looks like the colour scale is saturated. Can you include a version of that result with the colour bar so I can see the scale?

Scott

[Scott Newman]

There would appear to be some discrepancy between what you have sent me and where you got those results from. Either that or you are using a much older version of the product. I ran your design and at 1550 nm um while it does not propagate across the resonator as seen in yours it is certainly well confined (see 1550nm.PNG). The worst results I could get from your simulation were for 1682 nm which is outside the bandgap you are using so it is not surprising (see 1682nm.PNG).

Is the design you sent in fact the same one you used to get those results? Note that the one you sent did not have an observation area covering the whole XZ space as shown in your results so some changes were made.

It also looks like the colour scale is saturated. Can you include a version of that result with the colour bar so I can see the scale?

Scott

[Poonam Jindal]

Thank u so much for the response.

Yes, the design is same as I send you and that of the results. The only difference is that I have inserted the observation areas in the design as per your previous comments.
The colour scale was saturated. I have adjusted the scale and got better propagation.

[Poonam Jindal]

Thank u so much for the response.

Yes, the design is same as I send you and that of the results. The only difference is that I have inserted the observation areas in the design as per your previous comments.
The colour scale was saturated. I have adjusted the scale and got better propagation.

[Author]

Posts