linear model vs s-parameter of ne68100 problem
biased it, and checked the S parameter, and compared it to an S parameter file.
(I veryfied the DCpoint and the DE embeded)
Any suggestions would be appreciated,
Best,
Elad
What is the problem?
sorry, the S parameter was diffrent
That's the problem with models: They're just models. I would rather rely on the S-parameters (which are measurements anyhow) and preferrably ones that you measured on the substrate of your choice. You can use spice models for quick and dirty analyses, like DC bias or harmonic content. Even phase noise prediction to a certain extent, but if you really want to know: build and measure the VCO and optimize by hand.
A few hints why the model might be different: is the substrate on which the S-parameters were measured taken into account? Where was the reference plane for the S-parameter measurement? Are the package parasitics included in the spice model of the transistor?
Simulate the circuit using the transistor's nonlinear model, including the bias networks for the desired bias point (the same as in the s-parameters)
Simulate the Large signal Gain. If you are using Microwave Office, it is in the Nonlinear simulations list,named LSSnm, and compare with the s-parameters of your file
Surely the responses simulated will differ a little, if the nonlinear model is good enough
tx. i worked with die. i added the wire bond to the nonlinaer model, and non. it is diffrent
I like to build an very low phase noise vco. so i need a good model.
can u jelp me ?