Modeling of LTCC filter using space-mapping neural network method

Modeling and analysis, optimization, and synthesis are the issues in design of low temperature co-fired ceramic (LTCC) filters. A multilayer filter is proposed based on the space mapped neuromodeling method. With optimizing and training the mapped relations between 3D structure and equivalent circuit in filter, physical structural characteristics of the filter can be got accurately and rapidly. This proposed model is validated through actual LTCC process. The filter exhibits a 300 MHz passband with the center frequency at 2.45 GHz and an in-band insertion loss of 2.2 dB. Measured results agree well with the high frequency simulation. The modeling method can be used for design on LTCC microwave passive components.