Optiwave Systems Inc.
Product Manager – OptiBPM, OptiFiber and OptiGrating
Forum Replies Created
You don’t need to apologize for the need to learn something. I had one customer who remarked – “I’ve been working with gratings all my career and never seen that effect before – well, you never stop learning until you are dead”. How True!!
There is no direct connection between temperature and wavelength. In the sensors section of OptiGrating, the temperature influences the spectrum though material properties thermo-optic and thermal expansion coefficients. The thermo-optic effect is the temperature changing the refractive index of the material. The thermal expansion coefficient is the strain (expansion, displacement) caused by heat. The expansion changes the grating period and optical gratings are very sensitive to period. The author of OptiGrating thought that sometimes the heat might not be applied uniformly over the whole length of the grating. It might be applied in some distribution, hotter in some places than others, and therefore it would be a non-uniform grating. One would have to calculate the spectrum in steps, as done for apodized, chirped or otherwise nonuniform gratings.
I can recommend putting the sensing segment inside a Bragg cavity. i.e. two gratings in the stop band in sequence. The spectral properties of a Bragg cavity are very sensitive to changes in the cavity.
It looks like I’m not getting through. OK, let’s take it more slowly, a little bit at a time, line-by-line:
1)OptiGrating cannot calculate the spectrum of a non-uniform grating. It knows the spectra of uniform gratings only.
2) OptiGrating can also calculate the spectrum of two or more gratings connected in sequence.
3) Therefore, when asked to calculate a nonuniform grating, it finds an approximation by dividing the non uniform grating into a specified number of uniform gratings connected in sequence. each of those gratings takes its parameters from the corresponding location on the nonuniform grating.
4) OptiGrating uses scattering matrix theory to combine the spectra of uniform gratings connected in sequence.
5) The number of uniform gratings is specified in the field labelled “Number of Segments” in the dialog box “Grating Definition”. It is user defined because there is a trade-off between simulation time and accuacy. The user makes the most practical choice for each case.
If you follow all of this, I believe I’ve answered your question. If not, can you tell me at what point you are no longer following me?
Sensor applications often lead to nonuniform gratings. If the sensed environmental parameter varies continuously over the length of the grating, the grating parameters also vary, resulting in a nonuniform grating.
Optigrating uses the solution to the coupled mode equations for uniform gratings. It approximates nonuniform gratings by dividing the length of the grating into steps uniform segments. You noticed the sensor response changes with steps. Try a nonuniform grating and look at the profile, you should see it changes too. For a nonuniform grating you should see the grade resolved into a staircase where the number of steps over the whole length is steps. To calculate the spectrum of a nonuniform grating, Optigrating will use the scattering matrix of each uniform segment to calculate the spectrum of the whole grating.
Correction: The speed of light is 300 µm/ps.
In fact, a change in the electric field will propagate from the electrode at the speed of light. (More precisely, at the speed of propagation of radio waves. Material constants have different values in the radio range than at optical frequencies.) The speed of light is 300 mm/ps, and your dimensions are on the order of µm. Therefore the transit time for propagating the change in the electric field is much shorter than ps. I’m sure the time constant of your electrodes will be longer than that!
In OptiBPM, one can define waveguides created by diffusion of specific metals into specific crystals. These features follow the technical process for each diffused waveguide technology. Examples can be found in OptiBPM_Tech_Background_and_Tutorials.pdf,Lesson 7: applying predefined diffusion processes. For example the substrate can be lithium niobate, the required process parameters are already defined in OptiBPM. One cannot have a substrate of dielectric material, since the diffusion process is not defined. I recommend following lesson 7 and modifying it to suit your application as far as possible.
There does not appear to be an attachment to this post. Can you try to attach the project file? (The one with extension .bpd)
Lithium niobate is an insulator. It can become a waveguide if Ti is diffused into the surface. What do you mean by lithium niobate on insulator(LNoI)?
Switching speed will be limited to the speed of the slowest part of the system. In MZI, the slowest process will probably be the time it takes to charge the electrodes to create the electric field. The electrodes have both resistance and capacitance, and the time constant of the electric circuit will probably be the longest response time.
No, apodization is not the same as fringe visibility. Apodization is used in defining the refractive index distribution in the grating, as shown in equation (15) of OptiGrating Users Reference 42.pdf. I don’t believe OptiGrating has a fringe visibility parameter.
Optiwave software does not model the electrical properties of semiconductors and will not model electrical contact problems, however, if the p or n doping level is high enough, the doped Si has negligible resistance. One can model the highly doped regions as an electric conductor, as shown in OptiBPM_Tech_Background_and_Tutorials.pdf, Lesson 10: Electro-Optic Modulator. In Lesson 10, the highly doped regions are modelled as perfect electrical conductors and the static electric field external to the conductors is found and allowed to modify the semiconductor optical properties by the EO effect.
Thank you for sending the sample project file. Now I can clearly see the problem. These waveguides have been formed with a Mg stripe evaporated to lithium niobate. Magnesium diffusion causes the refractive index of LiNbO to become less, not greater. Therefore these stripes have refractive index less than the substrate. Therefore there is no mode. You ask for modal input. The mode solver fails to find it because it doesn’t exist. There will be an error message indicating a problem with mode solving. try using Titanium diffustion, instead of Magnesium!
We have been unable to reproduce this issue at Optiwave. Can you send us a project that creates this error message?
Optiwave OptiBPM and OptiFDTD do not have semiconductor models, as is needed for analysis of PN junctions. These softwares are implementations of the BP method and FDTD method respectively. These methods are for modelling of optics only. OptiBPM, on the other hand, does have a feature that allows the definition of electric conductors of arbitrary shape. It will calculate the static electric field if potential is applied to the electrodes. It will calculate the modified optical field from the EO effect. The details are in OptiBPM_Tech_Background_and_Tutorials.pdf, Lesson 10: Electro-Optic Modulator.
There is a great deal of finformation in the OptiBPM manual for the creation of multilayer waveguides. Tutorial 10 describes how to make a very complicated electro optic modulator design. You might not need the electrodes, but Tutorial 10 shows how multilayers are created.
There is not likely to be a problem with the creation of the design, the issue is more likely to be related to the limitations of the BP method itself. SiO often has refractive index contrast too large for the Beam Propagation Method. Your simulation might require FDTD. OptiBPM uses much the same interface as OptiFDTD, so if you can specify it in OptiBPM, you can probably do the same in OptiFDTD.