In this paper we describe a method of accurately modeling newpassive devices by deembedding the many building blocks, from which theyare built, using just a few test structures. We use a nonlinearoptimizer to find the optimal equivalent circuit models for the buildingblocks by fitting extensive high frequency measurements of the teststructures behavior. We also demonstrate that the variation in thecomplete equivalent circuit models, based only on the circuit buildingblocks, can be used to predict such variations in actual fabricateddevices. Three sets of gridded parallel plate capacitor structures arefabricated on a Low Temperature Co-fired Ceramic (LTCC) process, using2-layers of a 12-layer process. S-parameter measurements are taken andelement values from Partial Element Equivalent Circuits (PEEC)deembedded using an HSPICE Optimization algorithm. Statisticalvariations in the deembedded element values of the structures buildingblocks are calculated and used in an HSPICE Monte Carlo Simulation tool.S-parameters are converted to y-parameters for convenience, and bothmeasured and predicted y-parameters compared. The comparison between theMonte Carlo simulation results and the measured data, determines thatthe statistical variation of the component values provides an accuraterepresentation of the overall capacitor performance. The key results inthis research are: the number of test structures needed is much lessthan the number of parameters and building blocks to be extracted; thismethodology is much faster than other finite element like methods; andthe equivalent circuit models and the process used to create them isaccurate enough to also model the variations in the actual fabricateddevices
展开▼