coupling to ring resonator cannot be seen in the movie monitor


#1

Hi all,

I’m simulating a ring resonator coupled to a waveguide in 3D. Although the transmission spectrum of the coupled waveguide shows the resonance features at frequencies I expect I am not able to see this coupling feature in a movie monitor. Basically the real time movie shows no coupling to the ring. Any help would be appreciated.

Thank you!


Problem with All pass ring resonator simulation
spatial distribution of the eigenmodes from the dipole cloud analysis
#2

Hi @hadiseh.alaeian

Can you please attach your simulation file for a review?

Thanks


#3

ring_res_Lumerical.fsp (672.5 KB)
Hi,
Attached please find the file.
Thank you!


#4

Hi,

I was wondering if you noticed sth problematic there which could resolve the issue.

Thank you!


#5

Hello,
The reason is the coupled light intensity is much smaller than the input excitation and the light traveling to the through port. Unfortunately, I can’t find any options for the start time of the recording. @bkhanaliloo Wasn’t this an option in earlier versions?

However, I reduced the dimension of the movie monitor and increased the resolution to 500 in x and y. Now, I can now see the coupled field.
Start at 00:33

Hope this is helpful!


#6

Hi @hadiseh.alaeian and @aya_zaki

In my best knowledge, we don’t have apodization for the movie monitor. If the filed intensity is weak inside the microring, I found out that adjusting the scale factor in the Edit movie monitor window is quite useful. You can choose this value to be for example 0.5 or 0.1. Please see the screenshot below:


Modeling a Ring Resonator used in an Experiment
#7

Thanks for the response!
I’m afraid I don’t understand what you meant by the “coupled light intensity is much smaller than the input excitation”. The resonator supports couple of resonating modes within that range as you can see from the through monitor. So I guess there should be proper coupling to the cavity from the feed waveguide.


#8

yes, there’s proper coupling. But a proper coupling coefficient can be 0.1. So the coupled power is 10% the input so it shows in the video as dark blue. That is what I meant.


#9

Hi,

Thanks for the response! As I mentioned I’m afraid I cannot see why the field intensity should be low in the ring while having the proper coupling to it. I’m attaching the spectrum of the through signal as well as the field profile at resonance case.

It seems that at resonance there is proper field inside the ring consistent with a transmission of ~23%. Am I missing sth here?


Ring resonator add-drop filter
#10

Thank you!
I just uploaded the through signal as the field profile at resonance showing more than 70% coupling at resonance (i.e. ~795nm)


#11

Yes, but here is what the time domain field intensity looks like:

The blue is the coupled |E|^2 . You can see how small it is compared to the Green (Through one)


#12

You’re absolutely right but I’m afraid that’s almost always the case when the input is broadband with very dispersive coupling. For example I’m attaching the spectrum of Lumerical Example for an add/drop filter for it’s through and drop signals. You can see the same problem there as well but the real time movie of their example seems to be much clearer compared to mine. That’s what confuses me.


#13

I guess I see what you’re saying though. So do you think if I reduced the pulse bandwidth just to be so narrow around the resonance that should be OK resolving the real time demonstration?

Thank you for all the help!


#14

I guess, yes, reducing the bandwidth should help. In addition, I am trying something else now; changing the component to be recorded to Ex. This is the one captured in the example’s video. There are positive and negative values in that video.


#15

Here is the new video:

I have tried changing the bandwidth and it has no great effect actually. Indeed, if we think thoroughly about dispersion here, it seems irrelevant since all wavelengths get coupled at the start of the simulation like the case when you are simulating a mere directional coupler. Then only some specific wavelengths get trapped in the ring due to constructive interference.


#20

Hi @aya_zaki
Thanks for the response!
I want to try some arrangements of dipole sources within that frequency range to see if I can properly see the coupling to the specific resonant mode.
My whole purpose is to see if a proper real time coupling to this mode can be obtained.
Also I’m wondering if the dipole cloud could be a proper way to find the spatial distribution of the eigenmode from a time domain analysis. Do you have any feedback on that?

Thank you!


#21

Hi @hadiseh.alaeian

I just checked the simulation file of the KB example again. The pulse used in the example is fairly broadband (1.48-1.63 um) however the movie monitor scale factor is set to 0.1 (rather than default value of 1):

I hope this answered your question why you can the field profile in the KB example and how you can modify your movie monitor settings to obtain similar resutls.

Regarding the rest of the conversation:

  • As @aya_zaki and yourself mentioned, the movie monitor captures the whole pulse profile. So, while you have around 80% coupling at a specific wavelength, a big portion of the pulse is not coupled. Thus, you need to modify monitor scale to capture this week fields.

  • FDTD accounts for both material loss and radiation loss. Material loss is included in material properties and radiation loss is calculated during the simulation. FDTD propagates the light in time domain by solving Maxwell’s eqautions and demonstrated exact behaviour of light field.

I am not recommending using dipole sources. First, since your light is routed with a waveguide before coupling to the ring cavity, you need to use mode source. Second, dipole source also emit a pulse which will have a bandwidth.

If you want to study the coupled field at a specific wavelength, a good approach would be to use time monitors. You can see how the light field evolves in time and by Fourier transforming of the data, you can obtain the frequency domain data. Alternatively, you can visualize spectrum from time monitor.

Please let me know if you have further questions and I am happy to help.

Thanks


spatial distribution of the eigenmodes from the dipole cloud analysis
#22

Hi @bkhanaliloo,

Thank you for the response!
I guess now it works as good as it could be considering the very dispersive coupling of the pulse.
I have a follow up question about the eigen modes which I believe I should create a new topic and your insight would be really appreciated there.

Thank you again,
Hadiseh


#24

continuing discussion in a new post