Abstract: To meet the process requirements of MOCVD systems, a PLC-based MOCVD computer control system was designed. This paper introduces the system's composition, implementation principle, and software design. The use of a programmable logic controller (PLC) as the core control system improves the system's automation level and ensures its reliability and flexibility. Keywords : PLC; MOCVD; computer control system [b][align=center]Design of MOCVD control system based on PLC GUO Run-qiu, CHEN Xian-Neng[/align][/b] (School of Electromechanical Engineering, Xidian University, Xi'an 710071, China) Abstract : According to the technical requirements of MOCVD system, we designed a MOCVD computer control system based on PLC. This paper introduces the system composition, realization principle and software design. Applying PLC to serve as the control system increases the level of automation and ensures the reliability and flexibility of the system. KEY WORDS : PLC; MOCVD; computer control system 1 Introduction MOCVD (Metal Organic Chemical Vapor Deposition) is a new technology for preparing high-quality semiconductor crystals. The advantages of this technology are [1]: it can be made into materials with various thin film structures; it can be made into epitaxial films with large area and high uniformity; it can accurately control the thickness, composition and doping concentration of the film; flexible gas source control technology, rapid switching technology of gas source, and fully automatic control of the growth process minimize the influence of random human factors and have good repeatability. To make these characteristics of MOCVD work smoothly, the process parameters must be strictly controlled. However, the process parameters of MOCVD are particularly numerous and complex, which puts forward increasingly higher requirements for the control method. Therefore, it is necessary to adopt computer automatic control. At present, most of the MOCVD control systems rely on foreign imports, which are costly. Developing MOCVD equipment with independent intellectual property rights will be a key link in the development of my country's optoelectronic industry and is of great significance. In particular, with the launch of the "National Semiconductor Lighting Project", the localization of MOCVD has become very urgent. According to the specific process requirements of the MOCVD control system, we independently developed and designed a PLC-based MOCVD control system. The system uses a host computer and a programmable controller to realize the control and management of the entire system. The field test operation shows that the system has stable performance and fast response. 2. System Composition and Implementation Principle This system mainly consists of a computer, a Siemens PLC S7-300 (the core of the control unit), a temperature control system, a gas handling system, and a reaction chamber. The basic structure of the control system is shown in Figure 1. 2.1 The host computer is an industrial control computer. It uses WINCC industrial control configuration software to communicate with the PLC via MPI, receiving information from the PLC and simultaneously sending commands to it. It is responsible for system monitoring, data recording, alarm recording, data analysis, and parameter configuration. 2.2 The PLC is chosen as the controller because of its high reliability, strong anti-interference capability, complete hardware support, convenient maintenance, and suitability for harsh industrial application environments. As the core controller of the system, the PLC is responsible for the operation of the entire system, including the acquisition of various signals, data processing, and the control of various output signals. Input signal acquisition includes flow, pressure, and alarm signals from various instruments and sensors. Output signals involve control quantities from solenoid valves, contactors, motors, pressure controllers, flow controllers, RF induction heaters, etc. 2.3 Temperature Control System The temperature control system comprises a temperature controller, induction heater, host computer, and PLC. This is a closed-loop control system. The temperature controller collects the reaction chamber temperature in real time via thermocouples and feeds it back to the host computer via RS232 serial port. After processing by the host computer's control algorithm, a suitable control quantity is calculated and transmitted to the PLC. The PLC then runs a program to control the induction heater to regulate the reaction chamber temperature. 2.4 Gas Handling System The gas handling system mainly consists of chemically polished stainless steel pipes, a gas purifier, a flow controller, a pressure controller, solenoid valves, and pneumatic valves. The main function of the gas control system is to control the opening of various valves in the gas path by controlling the pressure and flow controllers, thereby controlling the proportion of various gas sources and delivering the reactant to the reaction chamber through the pipes. To ensure the purity of the reactant, the pipes must be well-sealed. The pressure and flow control in the gas path are both accomplished by the pressure and flow controllers. Sensors transmit the collected actual measurement values ​​to the control system, which compares the collected actual values ​​with the set values ​​in real time. If the user is not satisfied with the control effect, closed-loop control can be used to modify the transmitted setpoints in real time. 3 System Software Design The system control mainly refers to the automatic and manual control of signals through PLC, thereby realizing the control of the heating system, gas flow and gas pressure, pneumatic valves, etc. Our designed MOCVD control system has two control modes: automatic control program and manual control program, which can be switched between each other. Its subroutines mainly include step sequence control, analog output control, analog input control, digital output control, and digital input control. 3.1 Step Sequence Control In the MOCVD control system, the controlled step running time and the required cycle position are different depending on the different recipes. One of the design difficulties of this system is that when writing the program, the start and stop positions of the cycle cannot be known in advance. How to write a program that can be used for multiple different recipes? In view of the requirements of the MOCVD system process and combined with the operation flow of this system, the sequential control design method is adopted to control the actions and commands between different times and execute the different sequence cycle control strategy. This method flexibly and accurately employs a loop control program to achieve different functions at different loop positions according to different formulas. Its basic idea is to divide the system's working cycle into 50 sequentially connected stages, called steps, and then use programming elements (memory bits M) to represent each step. Each step sets a start flag, an end flag, a loop entry flag, and a loop end flag. The transition condition between steps can be an external interrupt input "pre-jump" signal or a signal provided by the timer for each step. For handling uncertain loop positions, at the end of each step, it checks if the loop end flag is 1. If not, it jumps directly to the next step. If it is 1, it checks if the remaining loop count is 0. If it is 0, it jumps to the next step; otherwise, it decrements the remaining loop count by 1 and jumps to the step sequence where the loop began. The algorithm flow is shown in Figure 2. 3.2 Analog Output Control Analog output mainly includes 8 pressure channels, 20 flow channels, and temperature. In analog output, preventing shock is a very important indicator. To prevent impact, a creeping incremental output control strategy is adopted during output, ensuring that the analog output gradually increases to the required output value within a specified time. Each increment should be as small as possible to reduce the possibility of impact and ensure the growth process proceeds smoothly. The basic idea is as follows: At the start of each step, the step number is read, and the target setpoint (IN2), the output value of the previous step (IN1), and the number of steps to climb (D) are called. The step size S = (IN2 - IN1) / D is calculated. Then, the actual value and the setpoint are compared to determine whether the actual value should be added to or subtracted from the step size. Finally, it is checked whether the actual value reaches the setpoint. If so, the current step ends. Two cases are considered: the step size is greater than or equal to 0 or negative. Figure 3 shows the program algorithm flowchart when the step size S is greater than or equal to 0. The analog output program mainly uses the Statement List (STL) programming method, which is a language similar to assembly language. It has a higher execution speed than ladder diagrams, occupies less memory space, and can handle complex loops and jumps. This solution effectively addresses the challenges of handling multiple analog signals, complex step sequences, and unpredictable loops in this system, while limiting CPU memory. 3.3 Analog Input Control: The MOCVD control system has 29 analog input channels. Direct input via analog input modules would require 29 input points, resulting in high costs. Considering the system's relatively low real-time requirements for analog signal acquisition, ADG408 decoding and channel selection are used for time-division multiplexing. Each ADG408 can connect to 8 analog signals; using 4 analog channels allows for 32 analog inputs, significantly reducing the cost of the analog input subsystem. This is a superior choice when real-time requirements are not high. The analog input subroutine employs a time-division multiplexing input scheme, where the decoder selects one channel at a specific time as the output, transmitting it to a channel on the analog input module. The timing of ADG408 chip decoding and PLC analog input reading processing must be strictly separated to avoid reading confusion. To ensure that at any given moment during decoding and PLC reading, only one analog input is in the strobe and input reading state. As shown in Figure 4, eight analog inputs AI1-AI8 are connected to the ADG408. A program is written to output digital signals to control the ADG408's enable terminal EN and signal control terminals A2, A1, and A0, thereby achieving time-division selection of one of the multiple analog inputs and inputting it into the PLC's analog input module for data storage and processing. 3.4 Digital Output Control The controlled objects of digital output mainly include solenoid valves, contactors, motors, pneumatic valves, etc. For digital output control, the programming concept is to call the corresponding digital data from the data area at the beginning of each step. Simultaneously, to achieve real-time control by the host computer, it first checks whether the host computer monitoring system has modified the output value of a certain digital quantity in real time. If the host computer has modified it, the effective output value of the digital quantity is based on the modified value; otherwise, it is output according to the recipe settings in the recipe table. 3.5 Digital Input Control Digital input control mainly refers to the alarm and fault handling program of the system. The alarm program design includes automatic and manual. The alarm signal is detected by the sensor and transmitted to the PLC. The program makes corresponding safety protection actions according to the alarm signal, gives the trigger signal to make the alarm signal light light up, the buzzer sound, the system operation is suspended, the induction heater is cut off, or the corresponding flow and pressure controller is turned off. 4 Conclusion The control system proposed in this paper is applied to the second generation MOCVD system of Xi'an University of Electronic Science and Technology. Compared with the first generation MOCVD control system, it has made great improvements, especially in the step sequence subroutine design and analog output control. In the step sequence control, the sequential control design method is used to control the actions and commands between different times. Compared with the first generation shift control method [2], the step sequence control method is more flexible and reliable for solving complex loop problems. In the analog output control, the PLC statement list (STL) programming method is used to write the analog incremental climbing subroutine, which solves the problem that in the first generation system, a large number of analog outputs are calculated and processed by the host computer and then controlled by the PLC, resulting in excessive load on the host computer, control delay, and slow response speed. Field testing of the system demonstrates that the control system is stable, fast, and safe, fully meeting the process requirements and possessing high application value. Furthermore, the successful development of this system will promote the development of the domestic microelectronics industry, placing it in a leading position domestically. The innovations of this paper lie in proposing the design and implementation of a PLC-based MOCVD control system. In particular, advanced control concepts are employed in the software programming, using sequential control to solve the control of complex step sequences in the MOCVD system. For analog output control, the PLC's Statement List (STL) programming method is used to write analog incremental ramp subroutines, which have a faster processing speed than ladder diagrams, lower memory usage, and prevent potential surges in analog output. The proposed control scheme fully meets the system's process requirements. References 1. Shang Wensheng, Liao Changjun, Fan Guanghan, et al. Development and prospect of modern MOCVD technology. Journal of South China Normal University, 1999(3) 2. Xie Baohui. Research and design of MOCVD control system. Xi'an University of Electronic Science and Technology Paper, 2005.3 3. Zheng Sheng, Gong Jianping, Zhang Xue. Principles and applications of modern programmable controllers. First edition Beijing: Science Press, 1999 4. Liao Changchu. Application technology of S7-300/400 PLC. Beijing: Machinery Industry Press, 2005.1 5. Wang Guicheng, Song Lin, Xu Xinhe. Design of a remote monitoring system based on PLC. Microcomputer Information, 2005 Author Biography: Guo Runqiu, female, 1960, Jiangsu, Professor, research direction: computer control technology Chen Xianneng, male, 1980, Zhejiang, Master, research direction: computer control technology Professor, research direction is computer control techniques. CHEN Xian-neng, male, born in 1980, Zhejiang, Master's degree, research direction is computer control techniques. Author contact information: Contact person: Chen Xian-neng; Address: P.O. Box 197, Xi'an University of Electronic Science and Technology, 710071; Email: [email protected]