Can the Optical Time Variant S-Parameter (SPAR) also use touchstone files?

JonathJonath Member Posts: 1

Dear Lumerical Support Team,

I want to model a EO modulator including wavelength-dependence.

The modulator elements, even the scripted modulator, can't handle wavelength dependence, but it seems the Optical Time Variant S-Parameter (SPAR) element can be used for this.

Is the s-parameter file format type then strictly limited to

(<output port name>,<mode label >,<modeID>,<input port name>,<modeID>,"modulation",<group delay>)

or can it have any other file format as listed on your support website about S-parameter file formats?

https://support.lumerical.com/hc/en-us/articles/360036618513-S-parameter-file-formats,

That website, btw, may be broken, the png images don't appear for me.

Could the touchstone file format also be used for this element?

Thanks in Advance!

Answers

  • JonathJonath Posts: 7Member
    edited November 19

    I apologize for double-posting, but I can't seem to edit my question:

    PS: Can you suggest any other way of modelling a wavelength-dependent modulator (the result should be some CM-compilable custom element with a specifiable "length" property, CW behaviour is the focus)?

    So far, all the approaches I have come up have certain drawbacks and limitations. an overview:

    1. Said Optical Time Variant SPAR element, but here the length of the modulator is fixed. To include Length, one would have to use a compound element (2 opt, 1 elec port) and then use the setup script to add such SPARs iteratively to the compound, say 500 * 10um SPARs for a 5mm long modulator.
    2. Compound element (2 opt, 1 elec port) with rectangular bands Wavelength Demultiplexer, then an OM element for each wavelength band in parallel, then wavelength multiplexer. (obviously impractical)
    3. Compound element (2 opt, 1 elec port) with a single OM element whose wavelength dependent complex effective index is created via setup script from a property "wavelength" of the compound element (the user has to specify this wavelength manually, since Interconnect compound element scripts cannot use popportframe("port") to obtain the input signal wavelength)
    4. Compound element (2 opt, 1 elec port) with a Waveguide element (where wavelength dependence is easy) whose exact spectral effective index (or even MODE data file) is determined via script by an electrical input to the compound element. This would actually be really nice and would be my prefered approach, but there seems no simple way for the setup script of the compound element to get the data from the electrical input port...
    5. an entirely scripted element ... I would really like to avoid this one...
  • JonathJonath Posts: 7Member

    I apologize for double-posting, but I can't seem to edit my question:

    Can you suggest any other way of modelling a wavelength-dependent modulator (the result should be some CM-compilable custom element with a specifiable "length" property, CW behaviour is the focus)?

    So far, all the approaches I have come up have certain drawbacks and limitations. an overview:

    1. Said Optical Time Variant SPAR element, but here the length of the modulator is fixed. To include Length, one would have to use a compound element (2 opt, 1 elec port) and then use the setup script to add such SPARs iteratively to the compound, say 500 * 10um SPARs for a 5mm long modulator, and connect them all to an electrical fork.
    2. Compound element (2 opt, 1 elec port) with rectangular bands Wavelength Demultiplexer, then an OM element for each wavelength band in parallel, then wavelength multiplexer. (obviously impractical)
    3. Compound element (2 opt, 1 elec port) with a single OM element whose wavelength dependent complex effective index is created via setup script from a property "wavelength" of the compound element (the user has to specify this wavelength manually, since Interconnect compound element scripts cannot use popportframe("port") to obtain the input signal wavelength)
    4. Compound element (2 opt, 1 elec port) with a Waveguide element (where wavelength dependence is easy) whose exact spectral effective index is determined by an electrical input to the compound element. This would actually be really nice, but there seems no simple way for the setup script of the compound to get the electrical input port data.
    5. an entirely scripted element (2 opt, 1 elec port), or a variation of option 1 with a really heavy setup script that generates the SPAR txt data, both of which I would like to avoid
  • GWANGGWANG Posts: 112Ansys Employee

    Hi Jonath ,


    Thank you for the detailed post. Waht is the wavelength dependency effect you want in the modulator? It seems like you are talking about somethinglike using the OM in conjuction with the waveguide element to tune the waveguide effective index based on the electrical input to the OM (and this is the way the OM element should be used, it should be used with a piece of waveguide element to model the waveguide propagation property). Please let me know some more details on this and we can help you from there.

  • JonathJonath Posts: 7Member

    Hi @GWANG ,

    Thank you for the reply!

    It is indeed as you state, with the important difference that the effective index modulation itself is also wavelength dependent within the operation wavelength band of this EO modulator.

    Thus it does not suffice to tune the effective index based purely on electrical input .

  • GWANGGWANG Posts: 112Ansys Employee

    Hi Jonath ,


    Thank you for the explanation. In this case, does the ring modulator sounds like what you want to model?

  • JonathJonath Posts: 7Member

    Hi @GWANG ,

    Thank you for your reply! It is not a ring modulator, just an EO modulator whose modulation efficiency, or the shift in complex effective index, is strongly dependent on wavelength.

    I looked into the ring modulator Interconnect element (not useful, it's basically the same as the other modulator elements) as well as the application example found in

    https://support.lumerical.com/hc/en-us/articles/360042322794-Ring-modulator

    This example does, in fact, show a similar behaviour as the modulator I am modelling: the applied electrical signal not only shifts the spectral Transmission by a scalar, but also changes its spectral shape.

    This, however, is not achieved by the modulator element placed in the ring circuit itself but arises from the ring circuit as a whole.

    For me, I know what the Transmission behaviour of the modulator looks like (sorry for forgetting to mention this before) since I know its complex effective index as a function of voltage & wavelength (ie I don't want to simulate it on a circuit level).

    I want to include this neff(V,λ) = neff0(λ) + DeltaNeff(V, λ) in an interconnect custom element much like the basic modulator element, but with the important difference that DeltaNeff is also wavelength dependent. In the application example:https://support.lumerical.com/hc/en-us/articles/1500003780782-GaAs-AlGaAs-Electro-Absorption-Modulator-

    this is handled via the Optical Time Variant S-Parameter (SPAR) element, but this only works only for a single Device length. I want the user to be able to specify the length in the custom element and the Transmission to change accordingly.

    Do you have any ideas? Thanks in advance.

Sign In or Register to comment.