Abstract
Low Temperature Co-fired Ceramic (LTCC) technology has already been widely applied to facilitate circuits package for RF and microwave communication product miniaturization and modularization. This study presents an integrated method in which neural networks, back-propagation algorithm, and design of experiment are used to optimize the LTCC process parameter control. Following the design of experiment method, dominant factors were found for LTCC formulation, which was optimized for production.If this model functions pretty well, then, the experimental work and processing cost can be substantially reduced and the stringent fast time-to-market requirement will be also fulfilled.
