Nonlinear Simulations in ADS, How to

I am trying to conduct some nonlinear simulations in ADS to find IP2, IP3, etc ... I know that one of the main ways to do this is by using a nonlinear model of the amp and doing some harmonic balance and such, but my problem is that I do not have a model.

All I have is measured s-parameters for a MMIC and measured Power-in-Power-out data for this MMIC. So, I want to do a nonlinear analysis on this MMIC using a modified .s2p file. I've heard of this .S2D file format that can apparantly achieve this, but I can't seem to find many references on how to use it in ADS and how to accurately define it.

Any help or ideas? Thanks!

You cannot do this with S-Parameter data. S2P data is small signal so it cannot even simulate compression.

S2D data file would work the wrong way for you. Yes, it is based on s-parameter data but then you have to specify the gain compression. You set the IP3 performance for example as a method of deriving the gain compression rather than being able to predict the IP3 value. Also S2D simulation only models odd order products so IP2 is not possible.

There is another format P2D which is power dependant s-parameter data, a set of large signal s-parameter style measurements at a series of different drive levels. This will still not facilitate IP2 or IP3 simulations though as it only contains information about S21 at the fundamental frequency. There is no data to model the harmonics and therefore the intermodulation products.

To do these simulations you do need to have an appropriate non-linear model. At the MMIC level the only measurement based option for this would be X-Parameters but unfortunately this is so new that most MMIC manufacturers do not have this data available.

X-parameters on Wikipedia

X-parameters by Agilent Technologies

I too am looking for a way of extracting parameters from our InP FETs to form a non-linear model in ADS. I'm designing a dual-gate mixer, and I plan to use two InP FETs.

We have ADS 2004 and a lab with a VNA plus other equipment. We just don't have the budget to purchase the proper VNA and version of ADS to extract and utilize the new X-Parameter format.
(I'm pretty excited about this new format, but I've pitched this idea to my superiors without much luck)

Is there a parameter extraction procedure out there that we can use to fit an existing ADS FET model? A pdf or book reference?
(I've been using Microwave Mixers by S. Maas, but I'm looking for a more detailed reference with procedures )

Thanks