Varactor modeling using HFSS+ADS

[yuip]

[SOLVED] Varactor modeling using HFSS+ADS

Hi everyone!

I'm an undergraduate student and as a part of my final project I need to model a varactor diode using a waveguide. I've design a simple model in the HFSS, drawing the waveguide and the terminals of the diode, with 3 ports: 2 wave ports for the waveguide and 1 lumped port linking both terminals.

The idea is to export the 3x3 S Matrix to ADS and using that program, along with the measurements, to find a good model of de varactor. But I have a few questions regarding this process:

1) About the lumped port: should it have the dimensions of the actual diode or is it enough to draw a sheet that fits the diameter of the terminals?

2) If the waveguide is drawn in the way that its biggest dimension is along x axis, I understand that, depending on how I draw the lumped port it can affect the results. So, how should I draw the 2D sheet that represents the lumped port? I've drawn it in the XZ plane so it minimize the perturbation over the E field which propagates inside the waveguide, but I'm not sure if that's okey.

3) Once I export the 3x3 S Matrix and put it in a S3P box, a couple of questions arise:
3a) In two of the three ports I have to put Terms to obtain the S Matrix from the simulations. Those would be the wave ports of the waveguide. So I only have left the lumped port to put my varactor model. Does the model have to be connected between that port and ground?
3b) In HFSS there's no ground. How does the fact that in ADS you have a ground affect the results from simulations? Because the varactor would be connected between two terminals, not the lumped port and ground as it would be in ADS...

Thanks in advance for your answers.

Regards,

Daniel.

 

[yuip]

Well, I've got the answer for some of the questions, in case someone is interested. Though they might not be too accurate, it's what I found out asking some people and through simulations.

1) They should fit the diameter of the pin, not the real dimesions of the diode. It's a lumped port to get the 3x3 S Matrix, not the diode.

2) The diferences between both representations are too small, so doesn't matter in which plane the port is drawn.

3a) Yep, that's the way of connect the diode.
3b) The ground is to refer all the S parameters to the same point.

So I think all the problems are solved, unless someone has a better approach.

BTW: I'm doing some simulations to check that the results between HFSS and HFSS+ADS are consistent and they are good.