Waveguides - Science topic

The feed to a rod or wire dipole antenna can be well approximated by a discrete port if the fields at the feed are concentrated in a region that is small compared to a wavelength. If the feed region becomes significant compared to a wavelength, such as with microstrip feeding a planar dipole, then the impedance seen by a waveguide port feeding the microstrip will be different to the impedance seen by a discrete port. The fields around a discrete port contain a lot of short wavelength modes that die out very quickly at small distances from the port but contribute to the reactive impedance. The waveguide port does not generate these modes, but only the ones that propagate in the transmission line it is launching into. In CST it is necessary that this transmission line has a short distance from the port over which the transmission line is uniform, which prevents a waveguide port being used directly at the terminals of a dipole antenna and other structures not starting with uniform transmission lines. Presumably this is needed because of the way the code calculates or launches the waveguide modes.

As I understood, you are trying to define a refractive index varying with space which is feasible in principle. You can learn from https://optics.ansys.com/hc/en-us/articles/360034901993-Spatial-n-k-data-Simulation-object

Malcolm White and Shrivastav a.k.

First thank you for your response. I still try to measure the polarizer correctly. In my research I found a blog post. You may see it via link --> https://www.ainfoinc.com/blog/blog-how-to-measure-the-axial-ratio-of-the-circular-polarizer.html

In there the method is similar as Malcolm White said. And it looks more simpler and quicker to do for me. You both may see the pictures of my measurement setup. I put two 4 port OMT's back to back. Each OMT has 2 TX and 2 RX ports for vertical (V) and horizantal (H) polarizations. In this first setup I did the measurements S21-VV and S21HH. Then I put the polarizer between the OMT's as in picture at 45degree alinged and measure S21-VV and S21HH again. So it's like in blog post ILV1 ILH1 and ILV2 ILH2 I assume. Then I took amplitude division in linear scale;

Delta-V=ILV2/ILV1 and Delta-H=ILH2/ILH1

So if it's true so far Delta-V and Delta-H would be my polarizer's vertical and horizantal results in linear scale. Then I also took their amplitudes division to find Axial ratio --> AR= Delta-V/Delta-H

Then using this formula I calculated cross polar discrimination --> XPD=20*log10((AR+1)/(AR-1))

The results I had from those calculations in Matlab seems so perfect but also unrealistic beacause I've never seen a polarizer that has these kind of perfect XPD results. So I beleive I had made a mistake because I also couldn't see 90 degree phase difference between measurements. But I can't find where. If you have any idea please tell.

You can put a choke around the end of the waveguide - it only really needs it on the top and bottom walls, but it is easy to model all the way round. A 2 mm high air gap 1/4 wavelength deep ending with a shorting wall. Like a shorted length of rectanguar coax back along the outside of the guide wall. The 2 mm gap surrounding the open end will look like a good open and stop currents continuing along the outside of the guide, so making the end of the guide a low current point. I think this will also hamper the radiation.

Dielectric with a high permittivity will look more like a s/c. High permeability material (magnetic) looks more like o/c.

Thank you for answering, but there must be some way to efficiently calculate different polarizations along the longitudinal direction of the fiber

If you use a short, most of the power goes back as S11 and it is easier to see if some of it is missing, rather than using a good load and the answer being two values close to s11=0.

At a short, magnetic field is across the guide and the currents on the end wall are up and down, and on the top and bottom of the guide are mainly in the direction of propagation and don't couple to a longitudinal slot, likewise half a wavelength from this. A quarter wavelength from the short, at the open, the magnetic fields are longitudinal and the currents run up the sidewalls and sideways across the top and bottom of the guide, and couple more strongly with the slot, if it isn't on the centre-line where the currents meet and stop. I would be surprised if the slot resonance is insensitive to the slot length at this position.

Use a much shorter sample so it is not resonant. There are four peaks in one and seven in the other, so it needs to be 1/5 the length and 1/8 the length, probably, or shorter.

Hello,

Find attachment.

Thanks,

I think that -175 degrees is equal to 185 degrees. This is about 28 degrees out from 157 degrees. There are about 22 cycles along the waveguide, so you only need the guide wavelength to be about half a percent out and that would make the difference. Do you know the guide wavelength accurately enough?

Antardipan Pal did you get your answer ?

Dalga kılavuzunun oluşması için gerekli ortamı sağlar. Hız, dalga boyu, kırıcılık indisi vb.

Thank you so much for your answer. I have found some similar answers in the literature. Thank you again

This can be done with Floquet Ports by observing the responses at oblique incidence.

Hello,

I suppose, use vacuum or air first for the inner space of waveguide and also port can be chosen by selecting edges not the whole face of the waveguide. For that one more outer waveguide can be selected.

Thanks,

I wouldn't expect CST to be that accurate for S21. If you don't run it for very long you have power still bouncing around inside the problem, and if you run it for a long time then numerical errors start to build up. How much ripple was there left on the S11? If you watch the S11 as the solution progresses you can see the ripple reduce every time a lump of energy arrives at the port (see the energy monitor). If the finishing point is -40 dB energy left there is still 0.01% of the energy left in the problem which has the potential to make +/-1% difference in the voltage at the output port which is */-2% of the power. The absence of this from the answer leaves ripple in the S parameter. This ripple is bigger than the differences you are seeing. I'm not sure that the impedance is calculated as you would hope, but I think you can't tell from these results if it's right or not.

Yeah sure

Absolutely, my friend Mostafa Shabani! Growing graphene on vertical surfaces like the walls of a SOI (Silicon-On-Insulator) waveguide is indeed feasible. This process, known as vertical graphene growth or VGG, involves techniques such as chemical vapor deposition (CVD) or plasma-enhanced chemical vapor deposition (PECVD). These methods allow for the controlled deposition of graphene onto various substrates, including vertical walls.

In the case of SOI waveguides, which are often used in photonics and integrated optics, growing graphene on their vertical walls can enhance their performance by introducing additional functionalities like electro-optic modulation or ultrafast photodetection. Plus, the exceptional properties of graphene, such as its high carrier mobility and optical transparency, make it an ideal candidate for such applications.

An interesting article to read:

However, it's essential to optimize the growth parameters and substrate conditions to ensure uniform and high-quality graphene deposition across the vertical walls. With careful engineering and precise control, achieving this goal is well within reach. So, in short, yes, growing graphene on the vertical walls of structures like SOI waveguides is not only possible but also holds great promise for advancing various technological applications.

Graphene into waveguides opens up a dynamic and versatile realm of possibilities. Graphene's extraordinary conductivity, coupled with its electrically tunable ability, is revolutionizing waveguide functionality. By applying an external electric field, waveguide properties such as propagation constant and impedance can be dynamically adjusted in real time, allowing unprecedented control of signal transmission. Moreover, graphene's inherent low loss at high frequencies ensures efficient long-distance propagation, which is essential for communication and sensing applications. Beyond mere conductivity, the interaction of graphene with light induces surface plasmons, promoting enhanced light-matter interactions and enabling nanophotonics devices with superior performance. Moreover, its nonlinear optical properties pave the way for pure optical modulation and frequency conversion within waveguide structures. Ultimately, the integration of graphene into waveguides not only expands their capabilities but also paves the way for next-generation photonics, promising advances in communications, sensing, and quantum technologies.

Marina Qasim Yes, the propagation constant of the CPW mode will vary slightly depending on the various conductor/spacing geometries.

If you are seeing this difference in simulation, for lines that should be identical, it could be an indication that the simulation is not sufficiently converged.

Hi Haneen.

I am not familiar with Lumerical but, usually, what you want requires 2 different simulation schemes. However, if you want to use the electrooptic effect to change the refractive index of a given region and evaluate the impact on a propagation mode, you can do so by calculating (most simulation platforms have a multi-physics module) the refractive index change for a given DC field and then use this new RI for an FDTD simulation.

Hope it helps...

I don't know what is happening with your simulation. Have you looked at the mode patterns that are giving you these results? They are in 2D/3D results and you can see the mode patterns at one frequency.

If the modes aren't what you expect you can enforce some modes using symmetry planes. I used a horizontal E wall and a vertical H wall to split the problem up into quarters, and the results were what I expected, except below cutoff, which was recently asked about here, and I don't understand the reason for that.

Hello,

Yes, you can model any waveguide structure in terahertz range as CST has option to simulate the structure for very very small parameter like dimension of waveguide structure.

Thanks,

Measuring the characteristic impedance of a rectangular waveguide requires specialized equipment and techniques. Unfortunately, I cannot provide a complete guide for measuring it because it's not a simple task for someone without the necessary instruments and expertise.

However, I can offer some general information and resources that might be helpful:

If you need to measure the characteristic impedance of your waveguide, I strongly recommend consulting with an expert in microwave engineering or using a calibrated network analyzer with proper training. Attempting the measurement without proper knowledge and equipment could damage your instruments or provide inaccurate results.

Dikdörtgenin merkezden y-eksenine göre simetrik olarak düşünülürse, ikinci grafiğin orta noktaya göre simetrik olduğu, birinci grafik ise cisim aynı şekilde x- y-eksenine göre simetrik olduğu için belli bir değerden sonra sabit kalır.

Dalga kılavuzunda yansıma olunca süreklilik sona erer. Artık başka bir dalga kılavuzu oluşur. Nedeni yansıma ve yüzey

Oluşabilecek en iyi dalga kılavuzu özelliği sınırlama ve kalite faktörünün ayrı ayrı düşünülmesi.

Photonic crystals have regular periodic structures, and the periodic change of refractive index makes it produce photonic bandgap, which can prohibit the propagation of electromagnetic waves due to the regulation of Bragg's diffraction in all directions.

If R=1-T-A this is using conservation of energy. All the numbers should be in the range 0 to 1. They are not logs, they are proportions.

The reflected power added to the absorbed power added to the transmitted power should equal the incident power.

You need to measure the absorption. You cannot calculate it if you don't know S11 (R). If You use the formula A=1-T this assumes R = 0.

Don't use logs unless you want ratios. The formula R=1-T-A doesn't work for logs. If 10% is reflected and 10% absorbed then 80% will be transmitted, 10% =100%-80%-10%. If you log this it turns into -1=0-(-.1)-(-1) which is totally wrong.

You have calculated/simulated/predicted that S11=R=10^-8. If that is right then very nearly A=1-T. You only know A to the accuracy with which you know S11 and T. If T is 0.1 then A is 0.9-10^-8

Kristal kafesin geometrik şekli dalga kılavuzunun oluşumunu etkiler.

Kristal kafesin geometrik şekli onun kararlılığını belirler. Bu yüzden dalga kılavuzunun verimliliğini etkiler.

Bağlantı noktasında pmc etkisi yaratmak için dalga kılavuzu bağlantı noktasında manyetik koruma tercihe bağlı.Eğer manyetik koruma tercih edilirse dalga hem manyetik dalga hem de normal dalga özelliği taşır. Manyetik koruma yoksa sadece normal dalga özelliği gösterir. GW cinsinden S-parametrelerinin doğru hesaplanması için doğru port boyutu ve boyutları oluşacak dalganın dalga boyunda olması gerekiyor.

70 ve 90 derecelik dalga kılavuzunu üst üste gelme şeklinde bir dalga kılavuzu oluşturulabilir. 70'in ve 90'nın katları olan bölgelerde dalga kılavuzu üst üste biner ve ortak noktalar oluşturur. Bu üst üste gelen noktalar birleştirilerek sırt dalga kılavuzu oluşturulabilir.

Simetrik dalga kılavuzunda p-tipi ve n-tipi malzemenin optik değişimi , eşik seviyesi, eğim verimliliği, direnc değişimi aynı şekilde artar veya azalır. Bu artış ve azalış birbirine göre simetriktir. Asimetrik olayda ise bu artma ve azalma n-tipi ve p-tipi malzemede farklı şekilde olur.

I think you need to ask for the number of modes you want in the simulation setup box.

You probably need to set up conditions at the port where the te and tm modes can exist, by putting PEC boundaries around it, for instance. The port needs to be big enough to support the modes you want.

The type of semiconductor can be determined by the 4-probe method or by using the hall effect. The width of the forbidden zone can be determined using a spectrometer, spectrophotometer.

Comsol de çoklu simülasyon aynı anda farklı şekillerde gerçekleşebilir. Birden fazla simülasyon varsa, birden fazla bağlantı noktası ve sınır noktası vardır. Bu simülasyon elektrik devreleri ve comsol yazılımı içinde geçerlidir. Elektrik devrelerinin bağlantı noktalarında ve sınır noktalarında oluşan simülasyon denilen düğümler comsol yazılıma yansıtılır ve bunlar comsol yazılımda simülasyon örneği oluşturur.

A search for cutoff of circular waveguide found this

If you feed it using an H-plane T then both arms will be in phase. If you feed it using an E-plane T the arms will be out of phase with each other, which will give a null on boresight.

Hi Kumar Neupane

Usually, the height of cladding and buried oxide (BOX) in silicon photonics is determined by the foundry and typically varies between 2 to 3 microns. To avoid any mode scattering, you can set the cladding and BOX height to 3 μm. Besides, you can use perfectly matched layers (PML) to reduce the reflection from boundaries.

I noticed that in your mph file, you haven't considered the silicon substrate. If your intention is to only study the waveguide modes, you can extend the BOX and cladding heights to any value larger than 5 times the maximum wavelength.

Please refer to the attached mph file where I have added a PML layer and made modifications to the geometrical parameters.

Thanks,

Aydin

Hello,

In microwave characterizations, you must make sure that you followed the right procedures for characterization such as calibrations and sample preparations. Many factors could have affected your results.

I have the following advice for you:

1- Re-calibrate the VNA more than one time and characterize the sample each time to check the difference between results.

2- If you have any air gap in the sample, the results will never be accepted or OK. You must make sure to TOTALLY avoid the air gap. In my opinion, the best way to avoid air-gap is to fabricate your sample inside the sample holder (the one which you use during characterization) you can refer to my article in IET

@ One more thing, are you sure that there are not any air bubbles inside your sample? Having an air bubble inside the sample will also affect your results.

@ Have you checked the homogeneity of your samples? It is crucial to ensure that your sample is homogeneous.

3- You can also check the air gap by simply directing a strong light (for example, a mobile torch) to one side of the sample and observing the light on the other side of the sample (this method could help but it is not always effective).

4- Both sides of the sample should be smooth as possible.

5- If possible, try to characterize any other samples apart from your samples and check the results.

6- If you followed all mentioned advice and the results are still the same, then, I advise you to increase the thickness of your sample and also fabricate more than 5 samples and characterize all of them.

Any tiny or very small air gap will destroy your results. The air gap can never be accepted in such characterizations.

All the best!

fdtd is really a finite volume method so there are many ways to implement boundaries/interfaces that cut through the rectangular grid. The useful ways out of the many ways to do this are the ones that maintain accuracy and do not alter the stability condition. Additionally, many ways that have been published are not stable even if the publication's results appear stable.

Stair-stepping is not bad, once you realize that your discretization now has to deal with another length scale in the problem ( via the slope of the stair stepped curve). A curved boundary, say PEC, that is too stair-stepped behaves like a diffraction grating and if the wavelengths in the domain are larger than the period of this grating then the stair-stepped boundary "homogenizes" to an impedance boundary. But as dxdx,dz--->0 this artifact goes away and you recover the boundary condition intended.

At the end of the day, if your interest is to actually solve a problem, just use a very fine discretization so stair-stepping is not a problem. There are fast clusters with huge caches and RAM to make this possible.

Grafiğin eğimi nasıl bir dalga kılavuzu olduğunu belirtir.

Dikdörtgen bir dalga kılavuzu oluşturmak veya simüle etmek kaynağın açık veya kapalı olmasına bağlı değil önemli olan dalga kılavuzunun tam oluşması ve ölçüiebilmesidir.

The difference in the results obtained using the waveguide port and discrete port could be due to several reasons:

Dalga kılavuzuna yandan bakıldığında şeklinin eksik olup olmaması önemli değil, önemli olan dalga kılavuzunun devamının olmasıdır. Dalga kılavuzu farklı bir ortama maruz bırakıldığında oluşan yeni dalga kılavuzuda farklı olur.

Energy can only travel in the waveguide in the permitted waveguide modes. If there is only one mode permitted, the energy must travel in that mode, or not at all. If the incident energy does not match that mode in angle or distribution, then some will be reflected. There will be modes that can be added together to match the incident field, but they will be non-propagating (evanescent) and result in energy being stored near the generation point. They look like series and/or parallel inductance and/or capacitance. This reactance corresponds to the mismatch in impedance that results in the reflection. If many propagating modes are possible, then the energy is spread among the modes that can be combined to give the closest approximation to the incident wave. Look up mode conversion in waveguides to see how this can be calculated.

Miguel Ferrando-Rocher

Hello dear friend, thank you

According to the article:

Do you think it is easy to use the thesis of trapezoidal dielectric shape in reality?

Dear Fathima,

I think this video will help you,

Best regards,

Fotonlar engele çarptıktan sonra fotonda kimyasal bozunma gösterir ve foton eski özelliğini kaybeder. Engele yaklaştıkça ışık spektrumları engelle etkileşim halindedir. Bu yüzden hem engelde kimyasal değişim meydana gelir hem de fotonlarda kimyasal bozunma meydana gelir. Işık ışınları eski özelliğini kaybederek farklı şekilde kırılır.

please follow the inverter calculations procedure, the S-Matrix could be obtained using any of the EM simulation softwares.

Hello, I have rudimentary knowledge on CAD Software, but I don't know CST-Studio very well.

You can use an "extrusion" method. It might have another name in CST.

This turns the shape into a solid object by extending the 2D surface into a 3D object along one axis. You can input the thickness of the resulting object.

Make sure that there is a single surface with correct specifications. Some programs display surfaces but these are actually only a graphical guide.

If extrusion does not work, maybe the surface has to be created as an actual object first.

I'm sorry but I don't know how the materials work in CST.

Good Luck!

For waveguide fed horn antenna, Excitation method and radiation boundary implementation are shown in attached figure. pl. implement this in to your model. Hope it will meet your requirement.

I did a search for "how does single mode optical fibre maintain polarization"

The equation can be found here

@Nagarajan Nallusamy ... Thank for your reply. I know this formula but in the simulation design there is no stress in the SiN waveguide. Therefore no refractive index has changed. I want to know how to get stress in the waveguide structure as like the PZT structure.

I think that one have to take into consideration the effective index that depends greatly on the dimension of the waveguide.

Dear Sabah al-ithawi ,

Please follow this weblink to comprehend polarizations, i.e., TE and TM in the directional coupler. https://s3.amazonaws.com/fip-3/Full-Explorer/index.html

May you please give a reference to your formula of ER mentioned in the document?

Hi Andrei!

Yes, I checked it recently, here is the thing:

Basically, the fields are correct.

When I use it as a splitter the fields are correct, but when I use it as a combiner... well, they are correct, but I have to use "edit source" and correct the phases at the appropriate outputs. For e.g. I designed a 1to4 combiner for X-band, and all the 4 output ports were in-phase. Nonsense.

Malcolm White , yes. But in one of such video had been showed excited Ports with all the area of the extreme boundary.

Fundamentals of microwave and RF Design by Michael Steer, NC State.

I measure 201 points of S parameters from each side, total 402, see the picture attached. Now I found that the shift of S21 changes if I disassemble the scheme and then assemble again, though I had several times the same 3 % deviation. In my last measurements I had -0.21 % and +0.49 %. It looks like it depends on accuracy of assembling.

In any case I am very thankful to you for your interest.@

See

I believe that equation 6 may be wrong and the signs in both top brackets should be negative.

You can generalise the equations to use any three different calibration loads (I think), but it is best if they are significantly different from each other.

Hi @Azadeh Karimi,

recently, I met the same problem when I want to use out-of-phase signal as the light source. I don't know if you have already solve this problem?

Dear Dr Sewidan..I guess that you would like to couple from a conventional (monomode?) fiber to a dielectric ? waveguide on a chip (optical image line?)We may also discuss this via normal e-mail (Fritz Caspers@cern.ch