1、Thermo Tank Temperature Control System Based On STM32 Biao QIU( ) , Shi-guang LI( ), Zheng-zhong GAO( ), Xu ZHANG( ), Yu RUI( ) (School of Information and Electrical Engineering, Shandong University of Science and Technology, Qingdao 266510, China) Abstract-this paper introduced a thermo tank temper
2、ature control system based on STM32, Firstly, the temperature acquisition is realized by the high-precision electrical bridge based on constant current source. Then the augmented PID algorithm realized by software is adopted Butterworth filter is used to convert the output PWM of STM32 to current si
3、gnal which is used to control the semiconductor control rectifier to adjust the temperature. Calibration check and practical application both indicated that the system was reliable, high-precision, practicable and could meet reality needs. Key words-STM32； thermo tank； temperature acquisition； PID M
4、anuscriptNumber: 1674-8042(2011)01-0064-03 Dio: 10.3969/j.issn.1674-8042.2011.01.16 1 introduction Thermo tank can be divided into low temperature thermo tank and high temperature thermo tank according to temperature range. Heating control thermo tank is one kind of high temperature thermo tank and
5、has a wide range of applications in industrial, medical and scientific areas. As some special thermo tank control system require high precision in temperature acquisition and control, the system designed in this paper can measure temperatures from 16 to 80 and its precision is superior to 0.05 . As
6、ARM is gradually occupying the microelectronics market for its powerful function and low cost, it is of important practical significance and value to design a temperature control system based on ARM with high precision, simple structure and low cost. 2 Basic control principles of thermo tank In this
7、 system, temperature acquisition of the inner thermo tank is realized by using platinum resistance as temperature sensor and bridge circuit based on constant current source. Then compare the actual temperature with the temperature set by touch screen. By using augmented PID algorithm to adjust, STM3
8、2 outputs 16-bit PWM signals. Then convert PWM signal to voltage signal to control the conduction angle of Semiconductor Control Rectifier(SCR) which controls the heating tubes. System control principle is shown in Fig.1. Considering the system accuracy and stability requirements, features of this s
9、ystem include: powerful and high speed ARM STM32F103 as the controller, augmented PID algorithm, and full use of on-chip resources of microcomputer such as ADC, USART and 16-bit PWM output for great control accuracy. Fig 1 System control principle 3 hardware design This system includes temperature a
10、cquisition bridge circuit, STM32F103, color LCD touch screen control circuit, filtering circuit and SCR. In addition, the system has a good man-machine interaction function and can realize real-time monitoring and control by using 5.6 inches color LCD and touch screen. Temperature control system str
11、ucture is shown in Fig.2. Fig 2 System structure 3.1 temperature acquisition and A/D conversion Among the thermal resistance temperature sensors, platinum resistance, with advantage as high precision, stable performance, corrosion resistance and easy to use, is the ideal temperature acquisition comp
12、onent widely used in industrial environments and control systems. As the temperature acquisition range is 16 to 80 , Pt1000 is chosen as temperature sensor, which resistance changes with temperature according to certain rules and has good high precision and stable performance. Unbalanced bridge meas
13、urement is typical in detect circuits using platinum resistance as temperature sensors1. However, the nonlinearity between platinum resistance and temperature and nonlinearity of unbalanced bridge lead to acquisition error, thus we improved the temperature acquisition bridge circuit. Use constant cu
14、rrent source to power the bridge, connect the two bridge arms with precise operational amplifier that is low noise and low temperature drift, use 4DH2 to constitute constant current source circuit which outputs 0.5 A current, thus the current in platinum resistance is equal to constant current sourc
15、e. The ADC of STM32F103 is used to convert analog voltage of temperature into digital signal. The 12-bit ADC is a successive approximation analog-to-digital converter and has the function of self-calibration. D/D conversion of each channel can be performed in single, continuous, scan or discontinuou
16、s mode, and in this system we use continuous mode. The result of ADC is stored in right-aligned 16-bit data register which improves the conversion speed. In addition, the analog watchdog feature allows the application to detect if the input voltage goes outside the user-defined high or low threshold
17、s. 3.2 TM32F103 on-chip resources TM32F103 can work in -40 105 and this meets the requirements of industrial environment. It incorporate the high performance ARM Cortex-M3 32-bits RISC core operating at a 72 MHz frequency, high speed embedded memories (Flash memory up to 128Kbytes and SRAM up to 20K
18、bytes) to store data and program, and an extensive range of enhanced I/Os, most of which have alternate functions and peripherals connected to two APB buses. It has three general purpose 16-bit timers plus two watchdogs, as well as standard and advanced communication interface USART used to communic
19、ate with LCD2. More importantly, it offers two 12-bit ADCs with 1 s conversion speed which make it suit for fast acquisition and fast processing. It is one of the important reasons for this system to choose TM32F103 as the core controller. 3.3 Filtering and conversion circuits In order to realize th
20、e convention from PWM signal to analog output, we use the second order low pass filter to filter out the high frequency components and keep DC component and changing duty cycle of PWM signal so that the analog voltage output is got then. Fig.3 shows the designed Butterworth filter. After filtering,
21、convert PWM signal to 02.5 V to control thyristor conduction angle3. Thus we realized the precise control of heating temperature. Fig 3 Butterworth filter 4 Software design 4.1PID control algorithm This system uses PID control algorithm which is a basic control method widely used in industrial proce
22、ss control method widely used in industrial process control. Augmented PID control algorithm4 is uk - uk- 1 = KP ( ek - ek- 1 ) + K 1 ek + K D ( ek - 2ek- 1 + ek- 2 ) . However, if this algorithm was used directly, it could generate a lare overshoot and cause integral saturation easily when starup,
23、stop or adjust substantially. In order to inhibit the emergence of this phenomenon, we use integral separation as an improvement. Integral separation wont work until actual temperature is approaching the settings. When it works, it can eliminate static error and improve precision5. Block diagram of
24、integral separation PID is shown in Fig.4. Fig 4 Integral separate PID algorithm block diagram 4.2 Touch screen software design It makes human-computer interface much more friendly, more convenient and faster by using touch screen. Use dedicated control chip ADS7843 to connect AMT9532, four-wire res
25、istive touch screen, withSTM32F103, process the touch screen signals6. Touch screens software design flow chart is shown in Fig.5. Fig 5 Touch screen flow chart Use standard thermometer with 0.001 precision as calibration to check the experimental results. Specific methods: set different temperature
26、s within the appropriate range though touch screen, wait until the temperatures shown in the LCD are stable, then calculate the errors based on the actual temperature of standard thermometer with formula: Error=|set-actual|/set. The check results are shown in Tab.1. Tab.1 Calibration results 6 concl
27、usion By using 16-bit PWM output, simple filtering circuit conversion circuit, software design and floating-point operations, this system realized 16-bit D/A. Conversion which is very hard for common MCU to realize. The system temperature range is 680 and the resolution of 16-bit control signal could reach to 10 . The experimental
