Diffusion furnace control using a learning control method

Abstract

A learning control method is applied to a semiconductor diffusion furnace. In the proposed learning controller, an adjustment signal besides a virtual reference has been employed and it is shown that the signal improves the response in the high-frequency spectra. It is also found that the control method is effective for disturbances due to both loading a wafer boat and burning operations. Experimental results show that the proposed method is very effective for high-performance temperature control and gives errors five times smaller than associated with a conventional LQ controller. >