Extraction of equivalent circuit parameters from S-parameter measurement data

[arjunshetty]

I have measured the S-parameters (S11) for a Schottky varactor upto 40GHz. I have a desired equivalent circuit model for my device. I would like to extract the equivalent circuit parameters to fit the measured S-parameters. Is there any software (preferably open source) that can help me do this extraction of parameters?

 

[BigBoss]

It's a modelling problem and only model engineers can extract component values for a predefined model of a element.
Because measured s-parameters should fit into predefined diode (or another component) model.For instance, a varicap diode model components can be defined by voltage vs capacitnace measurements and the model primitive components are so defined by these measured datas.You can not use arbitary model as you wish too.

 

[biff44]

it is critically important to de-embed the diode itself from the test fixture, especially at 40 GHz. What type of network analyzer calibration, and what test fixture, did u use. It might be necessary to resonate the diode with a known inductance to actually model the capacitance with any accuracy. It is easy to measure a resonance peak, it is hard to measure a small capacitance in the presence of other parasitic capacitances.

 

[arjunshetty]

I have attached the image of my device. This Schottky diode has been fabricated directly on a GaN film using lithography. I am not using any other text fixture. I am directly probing the device with a GSG probe (signal on the centre semi-circle and the 2 grounds on either side touching the ends of the bigger semi-circle). I am measuring the S-parameers by directly probing the device like this.

In this case, can I assume that there are no other parasitic capacitances involved? Do I still have to de-embed the diode or can I assume that the entire response is directly from my device?



Regards,
Arjun