Design And Research On The Control System Of 12-inch Single-spindle Dicing Saw

Dicing saw is one of the key equipment for semiconductor packaging process,mainly used for cutting and separating materials such as LED chips,IC chips,Ga As,silicon and ceramics.At present,the domestic dicing saw market is mainly occupied by foreign brands,the purchase price of equipment is expensive and the delivery time is long,which seriously restricts the development of China’s semiconductor industry.In order to improve the market share of domestic dicing saw and promote the development of domestic semiconductor industry,the control system of 12-inch singlespindle dicing saw is developed in this paper,and the main research contents are as follows:(1)The general structure,working principle and flow of the dicing saw were introduced.The solution design of grinding wheel blade wear detection system,vision recognition alignment system and motion control system were completed.(2)The principle and process of contact grinding wheel blade wear inspection were described.A scheme for contact grinding wheel blade wear detection based on switching signals was proposed.Hardware circuit design and programming were completed for the wear detection scheme,and experiments were conducted for grinding wheel blade wear detection.The experimental results show that the grinding wheel blade wear detection scheme based on switching signal has small detection error,good stability and high precision.Then,to improve the inspection efficiency,a non-contact grinding wheel blade wear inspection scheme was proposed.To reduce the impact of broken blades on wear detection performance,a broken grinding wheel blade detection method was designed.(3)The design of a visual recognition alignment system was carried out to achieve automatic alignment cutting.The selected image processing tool was Cog IPOne Image Tool,the template matching tool was Cog PMAlign Tool,and the template matching algorithm was Pat Max.The experimental results show that the algorithm has good stability and high accuracy.Then the design of the online template production and automatic identification alignment method was presented.Finally,the design of the autofocus method was carried out,and the commonly used sharpness evaluation function was quantitatively evaluated.The Variance function and Brenner function were selected as the optimal evaluation functions for the coarse focus and fine focus processes,respectively.After testing,the automatic focus can be used to find the focal position accurately that is in line with the focus requirements of the visual identification and alignment system.(4)Motion control system design and cutting test experiment.The composition and functions of the upper computer software were introduced,the design of the motion controller was completed,and the hardware circuit design and programming of the spindle drive system were carried out.In the servo system design,trap filters were used to suppress the resonance of the servo system and improve the dynamic response and smoothness;the servo system control method was designed to ensure the positioning accuracy of the motion axes.The test results show that the servo system has excellent performance.The feasibility of the control system was demonstrated by green silicon carbide dresser board cutting experiments.