Abstract: Pressure measurement and control systems are common application modules in embedded applications and are widely used in various measurement and control systems in modern industry. This paper introduces a pressure measurement and control system designed based on low-cost devices such as the MPxM2010 pressure sensor and 68HC908QT4 microcontroller produced by Freescale. The system's measurement accuracy is increased by combining simple analog circuits and software programming without significantly increasing hardware costs. Numerical calculations compare the system before and after the accuracy improvement, demonstrating a significant improvement in system accuracy. The method described in this paper can also be applied to other A/D conversion applications, greatly benefiting the reduction of system costs. Keywords: 68HC908QT4, MPxM2010, silicon piezoresistive sensor, A/D conversion, improved accuracy Introduction The MPxM2010 device produced by Freescale is a silicon piezoresistive pressure sensor. The MPxM2010 has high accuracy, and the output voltage has a good linear relationship with the input pressure. This sensor is a monolithic integrated circuit that integrates a pressure strain gauge and a membrane resistance network, and includes a laser-based fine-tuning module for temperature compensation and offset correction. The 68HC908QT4 microcontroller is a low-end 8-bit microcontroller with four 8-bit A/D conversion channels and a 16-bit PWM module, capable of A/D and D/A conversion. Combining these two chips creates a practical, low-cost pressure measurement and control system. The drawback is its relatively low accuracy; increasing the A/D bit depth would significantly increase the cost. This shortcoming can be mitigated through hardware integration and software programming, improving product performance without increasing hardware overhead. 1. Pressure Sensor Module Design The Freescale MPXM2010 is a silicon piezoresistive pressure sensor, its internal principle shown in Figure 1. The MPXM2010 has high accuracy, with a good linear relationship between the output voltage and the input pressure. This sensor is a monolithic integrated circuit that integrates a pressure strain gauge and a membrane resistance network, and includes a laser-based fine-tuning module for temperature compensation and offset correction. The MPXM2010 features the following: ◇ Pressure measurement range of 0–10 kPa, with an accuracy of ±0.01 kPa; ◇ Temperature compensation function between 0 and 85℃; ◇ Good linearity between output signal and pressure; ◇ Sensor contact surface can be selected with or without lead-out ports; ◇ Easy-to-use Tape & Reel package available, as shown in Figure 2. The MPXM2010's output signal is relatively weak, requiring the addition of an MOC2A60 chip to amplify the small signal and convert DC to AC. This allows direct control of the motor's power supply. During module debugging, the various components are separated for easier adjustments. An MC33179 operational amplifier is used, along with some resistors, to output the pressure sensor signal. The output signal magnitude can be adjusted by changing the resistance values. Figures 3 and 4 show the schematic diagram and PCB layout of the pressure sensor module design. 2. Pressure Measurement and Control System Design and Accuracy Improvement 2.1 Direct-Connected Pressure Measurement and Control System Generally, the design can be completed using the 68HC908QT4 A/D module. Simply connect the output of the pressure sensor module to the input of the 68HC908QT4 A/D module. Figure 5 shows the block diagram of the pressure measurement and control system. The 68HC908QT4 microcontroller has the following features: ◇ 4 KB Flash memory, 128 B RAM memory; ◇ 4-channel 8-bit A/D converter, 16-bit PWM module; ◇ Low price; the price per unit can drop below $1 for batches of 1000 or more. The MPXM2010 has a measurement range of 0–10 kPa. Limiting its output voltage signal to 0–5 V, its accuracy is: S = 5 V / 10 kPa = 500 mV/kPa. The 68HC908QT4's A/D converter is 8 bits, with a voltage limit of 5 V. Its accuracy is: R = 5 V / (2⁸–1) bit ≈ 19.61 mV/bit. The overall system pressure accuracy is: R/S = 19.61 / 500 kPa/bit = 0.03922 kPa/bit. To improve accuracy, the A/D converter is upgraded to 10 bits. The accuracy is: R/S = 0.03922 × (2⁸–1) / (2¹⁰–1) kPa/bit = 0.009776 kPa/bit. After upgrading the A/D converter to 12 bits, the accuracy is: R/S = 0.03922 kPa/bit. 22 × (28 - 1) / (212 - 1) kPa/bit = 0.002 442 kPa/bit. This approach does improve accuracy, but it increases hardware overhead. Using the PWM module of the 68HC908QT4 as a D/A converter can cleverly improve the accuracy of the A/D conversion. 2.2 The error in the improved pressure measurement and control system arises at the A/D level, where the decimal part is discarded; for example, 176.51 bits are treated as 176 bits. Solving this problem should start here, reducing the error. The error can be addressed using D/A conversion. The data read from the A/D converter is processed again by D/A conversion and then subtracted from the original data to obtain the result. The error cannot be directly fed back to the A/D converter, but it can be amplified before being fed back. Then, another A/D converter in the 68HC908QT4 is used to perform an A/D conversion on the amplified error. The MCU obtains the result, reduces it by the same factor, and adds it to the original A/D conversion result for a more accurate result. In Figure 6, the entire system can be divided into a pressure sensor module, an analog section, a microcontroller section, and an output circuit section. The key to improving accuracy lies in the design of the analog section, as shown in Figure 7. Assume the amplifier G has a magnification of 10. The A/D performance itself is not improved; the accuracy remains R = 19.61 mV/bit, which is the limit. After amplification by 10, the original maximum error of 19.61 mV/bit is expanded to 196.1 mV/bit. The A/D converter processes the amplified data, thus its capability is amplified by 10 times. During data processing, it is divided by 10 to restore the original value. Overall, it appears as if the accuracy R has been divided by 10, becoming 1.961 mV/bit. For example: the initial A/D conversion error is 10 mV, which becomes 100 mV after amplification. After another A/D conversion, the remaining error is 100 mV - 19.61 mV/bit × 5 bits = 1.95 mV, which becomes 0.195 mV after dividing by 10. The error is greatly reduced, with its limit being one-tenth of the original accuracy. The amplification factor G can be adjusted, but it must match the performance of the selected microprocessor and the accuracy of the circuit itself; choosing too high a factor is meaningless. In the circuit shown in Figure 7, Vm, D, and Vc are the same as in Figure 6. The calculated value of D is: D = (Vm - Vc) × (R14/R13)[1 + (R17/R16)]. The amplification factor of G is (R14/R13)[1 + (R17/R16)]. Conclusion In the product design and development process, cost is a very important factor. By cleverly utilizing the modules within a microcontroller and supplementing them with corresponding simple analog circuits, the chip's utilization efficiency can be greatly improved, and system performance can be enhanced. Making the most of what you have on hand for improvement and reinvention often yields twice the results with half the effort.