Gaussian Pulse Description [FDTD]

[Damian Marek]

Question:

I want to setup a Gaussian Modulated Continuous Wave input field for my structure. I have a rough idea what the time offset and half width mean, but is there any documentation on the exact meaning of these variables?

[Serena Zhang]

Hello, Damian

I went over the tutorial in the attachment and have some questions regarding the formulas on page 5. How did you come up with the formula to calculate the center wavelength if the start and end wavelengths are known? BTW, there is a typo in the formula.

Thanks

Serena

[Damian Marek]

Hi Serena,

The correct formula should be lambda0 = lambda_s + 1/e * (lambda_e – lambda_s)

Thanks for pointing this out.

[Haya AlSharif]

Thank you for the very helpful information.
What is e in the corrected equation of the center wavelength??
lambda0 = lambda_s + 1/e * (lambda_e – lambda_s)

[Damian Marek]

That’s the base of the natural logarithm or Euler’s number with the value: ~2.718

[shengsheng]

I use the formula in the attachment. And my start wavelength is 0.7 um and end wavelength is 2.5 um. I use these data to calculate the time offset and time domain half width. However, I cannot get the imagine as shown in the attachment. Can you give me some suggestions?

Attachments:

1. 

   .png
   

[Damian Marek]

You will notice one of the assumptions in the formula on the page containing IV is that the effective bandwidth (wavelength bandwidth divided by center wavelength) needs to be less than 100%. In your case (0.7-2.5)/1.55 is 116% which is greater. So that little helpful equation won’t help as it should.

Try increasing the time delay by an additional 0.5eE-14 s, from what I can tell that should look better.

[amin moazami]

hi , is there any exact formula for getting these two values (time offset & half width)?
for wave lenght = 500 um , i can not find the suitable value ;(

another Q is that , does the mesh size affect on time offset and half width ?

[asmita]

I’d not understand, how to choose the number of runs. As stated in “Deep understanding…”, run should make the GMCW pulse convergent to zero, but what should be the location of the observation point for this. For instance, if I have a waveguide coupled to reonator, then at different points the GMCW pulse will reach at different points. So, how one should decide the total number of time steps for the run?

[Damian Marek]

Answer:

Simply put the time offset is the time between t = 0 and the center of the Gaussian pulse. This has to be large enough, so that the entire pulse is in t > 0. The half width determines the width of the time signal and also the spectrum domain bandwidth of the pulse. For a complete overview of these variables and how to choose them refer to the attached PDF.

Attachments:

1. Deep_Understanding_the_GMCW.pdf

[Author]

Posts