RF simulation of MZ modulator

mode
device

#1

Dear all,

I’m using the example on knowledge base:
https://kb.lumerical.com/en/pic_modulators_travelling_wave_modulator.html

There are few points that confused me.

  1. the clearance between rib edge and the boundary of high dopant regions does not agree with the reference [1] (actually, a big difference: 0.95um, 1.75um). How can the web claim that the simulation results well match the experimental results?
  2. the calculated series resistance from the slab in the electrical simulation was not used in RF simulation. In the RF simulation, a new assumed value 8 ohm-cm is used for both n- and p-doped slab instead. Why?
  3. In RF simulation, the PN junction was replaced by some lumped 2D sheet elements. But I don’t understand this statement:"The resistivity is set to 8 ohm-cm for both the n-type and p-type silicon, corresponding to a resistance of 1.6 ohm-cm for the TWM. " (from https://kb.lumerical.com/en/modulators_rf-simulation.html). In the example, the value for resistor was set as 8000 ohm and capacitance was set as 2e-16 F in the simulation file. How to convert these values for 2D sheet elements?

And, how to set the dimension of 2D element? I found that the dimensions of a 2D capacity and resistor also affect the propagation constant of RF signal.

Thank you.


Traveling wave modulator的RF分析
#2

Hi @lih076, Thank you posting you question in KX. I have take a look at the example you have mentioned and unfortunately I am not sure why there is a difference between the resistance values between the DEVICE and MODE files. I will discuss this a colleague who is an expert in the field and get back to you. Unfortunately this weekend is a long weekend for us so the earliest I will be able to get back to you is next Tuesday. My apologies for the wait and thanks a lot for you patience in advance.


#3

Hi @lih076, apologies for the long wait. I have discussed this example with a colleague and it looks like the RF simulation was done separately from the electrical simulation and the resistance values were taken directly from the resistance reported in the paper. The resistance most likely does not match with the electrical simulation because as you’ve mentioned the dimension of the slabs is different from the paper in the electrical simulation. I am not sure why the structure is different though and will take a look at it/update it appropriately.

The values are defined in therms of sheet resistance. Since the thickness of the silicon slab is about ~0.1 um or 1e-3 cm, the sheet resistance comes out to be 8000 Ohm/square. Note that the width of the 2D sheets in the third dimension are made equal to their length. This was done because the RLC object itself takes the total resistance of the sheet as input so the 2D objects were created as a square so that the sheets resistance also denotes the total resistance of the 2D objects. I hope this also answers your last questions but please let me know if you have any more questions.