Accurate high speed empirically based predictive modeling of deeply embedded gridded parallel plate capacitors fabricated in a multilayer LTCC process

A novel technique is presented for the accurate, rapid, high frequency, predictive modeling of parallel plate capacitors with gridded plates manufactured in a multilayer low temperature cofired ceramic (LTCC) process. The method is empirical in nature and is based on the concept of incrementally constructing the model for a structure from well characterized individual building blocks. Building blocks are characterized by the use of test structures and measurements, and are modeled using passive lumped circuit elements. This method is applied to the predictive modeling of deeply embedded gridded parallel plate capacitor structures. The procedure has been experimentally verified, with accurate predictions of behavior obtained up to the second self resonance for large area gridded parallel plate capacitors. Since lumped element circuits are generated by this method, structure prediction speed is determined by circuit size and simulator small signal analysis time. The method is versatile and is well suited for circuit design applications.