GPC based adaptive crowning growth control in CZ-Si process

In conventional control system, crowning growth process has little consideration on controlling the microscopic quality of silicon crystal ingot. Phenomena of diameter fluctuation is observed and analyzed. Based on the cause of diameter fluctuation, a model based predictive control approach is proposed. Generalized Predictive Control (GPC) algorithm is used in the temperature feedforward path of crystal growth in a 300mm crystal growth test run, the controller shows very good improvement on suppress the diameter fluctuation after crowning growth finished. This ensures microscopic quality and increased the usable body length of crystal.