Home Forums SYSTEM Testing RSOA modulation bandwidth?

Viewing 3 reply threads
  • Author
    Posts
    • #16160
      zsliu
      Participant

      Dear all,

      I am going to test the modulation bandwidth of RSOA. I upload the setup. the issue is that I cannot get 1GHz signal after PD when I use 1GHz sinewave to modulate RSOA.

      Thanks

    • #16422
      Damian Marek
      Participant

      If you bias the Sine generator to 0.5 A.U. you will get the peak at 1 GHz.

      • #17064
        zsliu
        Participant

        Hi, Damian,

        as you suggested, i can detect the peak around frequency. now the problem is the modulation bandwidth unlike real RSOA which show us low pass property. any suggestion on that? thanks

        Zhansheng

        • #17111
          Damian Marek
          Participant

          In your particular example, I think the cutoff frequency is much higher than the values you swept. Try increasing the frequency to >10 GHz, I achieved a low pass filter shape. Although it was not perfectly flat.

          Regards

        • #17125
          zsliu
          Participant

          Thank you so much for your reply. The modulation bandwidth is generally less than 2GHz in a real RSOA component. the shape is going high from 100 MHz, may be increased by more than 5dB, then going to drop, even if I changed some parameters.

        • #17137
          Damian Marek
          Participant

          OK then the problem is something else! Do you have a paper with the results you are expecting? Showing the cutoff frequency and containing information about the RSOA parameters.

        • #17150
          zsliu
          Participant

          yes. please find the two attached documents. the curve in first file page14. the parameters are in the second file at page92.

        • #17183
          Damian Marek
          Participant

          Hi Zsliu,

          The OptiSystem team is determining whether the gain is a problem with the default parameters not corresponding to a peak gain at the default wavelength. For now can you please try the Linear or Lorentzian gain approximations for the material gain in the Enhanced tab.

          Regards,
          Damian

        • #17248
          zsliu
          Participant

          Hi,Damian

          I tried the linear gain approximations, which is generally used for bulk RSOAs. However, the results are not correct. I even cannot get sin waves . I haven’t figured out where the problem is. could you please help me?

          Thanks

        • #17256
          Damian Marek
          Participant

          I replaced some of the default parameters in the RSOA with those from the reference paper. (Active volume mainly)

          I got a good sine wave response, let me know if this is closer!

        • #17263
          zsliu
          Participant

          Damian, thank you so much for helping me lots on the rsoa module. The modulation bandwidth is still so high, more than 6 GHz when I used the linear gain approximations. I tried to change some parameters. however I havent figured out. any suggestion?

        • #17266
          zsliu
          Participant

          finally, I got the more reasonable modulation response. however the curve starts rising around 2.4GHz as shown in fig. some parameters may not be proper.

        • #17272
          Damian Marek
          Participant

          I think I used most of the parameters in the document you posted, you could also try the Lorentzian qpproximation. It is hard to say though since I only have one page to work from and it doesn’t say what type of model it is using. You can look at our help file which is accessed by the component properties window and see the background theory and references we use. You should be able to match the parameters with those from your reference.

          Regards

    • #17151
      zsliu
      Participant

      please find the files

    • #17153
      zsliu
      Participant

      sorry the second file is too large. I just extracted the corresponding page

Viewing 3 reply threads
  • You must be logged in to reply to this topic.