High-speed portable infrared gas analyzer based on DSP

The infrared gas analyzer is a physical analytical instrument based on the principle that different gas molecules selectively absorb infrared light of a specific wavelength. It is a typical optical, mechanical, and electrical integrated intelligent sensor system. Compared with other gas sensor systems, it has the characteristics of high sensitivity, fast response, multiple types of analyzed gases, wide range, and continuous measurement. It has been increasingly widely used in traditional industries such as earthquake prediction, mine safety, oil exploration, atmospheric physics, medical and health care, pollution source monitoring, high-voltage equipment fault diagnosis, chemical process control, metallurgy, and even all the leading disciplines of the new technological revolution such as biological sciences, microelectronics, and new materials.

From the existing infrared analyzers in China, most online monitors need the cooperation of the host computer to complete the post-processing and storage of data, and the front end only completes the functions of signal detection and acquisition. Such a design is not applicable in some occasions where installation is limited or there is no supervision for a long time (such as pollution source monitoring). In view of this situation, this paper studies and develops a non-spectral infrared gas analyzer based on TI's DSP, which can independently complete the monitoring work and store the data in a large-capacity flash memory or transmit it remotely via GPRS. The simple human-machine interface of the instrument makes it easy to complete both measurement and instrument calibration. At the same time, the addition of the USB interface gives the instrument a larger expansion space, and the instrument has a variety of signal output methods, which can be easily connected to various systems.

1 System structure of infrared analyzer

The system structure block diagram of the infrared analyzer is shown in Figure 1. The TMS320F2812DSP sends a modulated signal of a certain frequency to control the infrared light source. The infrared light emitted by the infrared light source is absorbed by the specific gas after passing through the gas chamber filled with the gas to be measured, and selectively transmits through the filter, and finally reaches the corresponding infrared detector. The detector measures the intensity of the absorbed light energy, which reflects the intensity of the gas absorbing infrared light, and also reflects the concentration of the gas. The weak signal output by the infrared detector is passed through a precise preamplifier and a secondary amplification filter circuit to obtain a stable signal. The signal is converted by A/D and sent to the DSP for analysis and processing. After filtering and nonlinear correction, the final measurement data will be transmitted, saved, or displayed and refreshed according to the system settings and the current instrument status. Note that there are at least two measurement channels for the detector here, one for the measurement channel and one for the reference channel, so that the effect of differential measurement can be achieved, forming a suppression of system noise and interference. The other sensors and A/D channels are used to measure temperature, humidity and gas pressure, and participate in the concentration compensation calculation.



2 System hardware design

In the hardware design of the instrument, the selection of infrared light source, detector, DSP system and peripheral circuit, and signal amplification circuit is the key. The infrared light source of this instrument selects the IRL715 infrared light source, the wavelength range of which is from the visible light wavelength to 5μm, and the working life can reach 40,000 h under 5 V voltage drive. The detector selects the TPS2534G2 with two measurement channels. DSP uses the TMS320F2812 digital signal processor from TI of the United States. It is a 32-bit fixed-point DSP. The core provides a computing bandwidth of up to 150 MIPS, which greatly improves the control accuracy and data processing capability of the control system. Its peripheral circuits mainly include: data acquisition, switch output, human-machine interface. Storage system and USB communication interface. The signal amplification circuit uses two-stage AD8552 op amp circuits in series. The AD855X series has the function of automatic bias adjustment and is the first choice for low-frequency weak signal detection amplifier systems. Since this instrument focuses on the application of high-speed DSP systems in infrared analysis instruments, the following is a special description of the peripheral circuits of DSP.

2.1 Data acquisition

Data acquisition is mainly responsible for analog-to-digital conversion and signal acquisition, and sends various analog quantities that need to be measured to DSP after A/D conversion. The signals of the measurement channel and the reference channel are generated by the infrared detector and converted by an A/D converter with a conversion accuracy of 16 bits. A serial 16-bit A/D converter that can directly interface with TMS320F2812 is selected according to the requirements; other signals involved in the gas concentration compensation calculation, including humidity, temperature and atmospheric pressure signals, are converted by the A/D time-sharing of TMS320F2812 itself, with a conversion accuracy of 12 bits. After the above signals are converted, they are sent to the DSP processor cache. After a series of complex calculations, the concentration value of the measured gas is finally obtained. Here is the circuit design for signal acquisition of the measurement channel and the reference channel, as shown in Figure 2.



2.2 Switch output

The switch output mainly includes three channels: detector temperature control signal, optical lens window temperature control signal, element alarm trigger signal (concentration, temperature, humidity) and one time trigger signal. The general I/O port of the DSP processor can be used to realize the output of digital quantity, and 74HC244 realizes the drive output of each control signal. Figure 3 is a digital quantity output circuit diagram. It should be noted that the power supply voltage of the TMS320F2812 processor is 3.3 V, while the power supply voltage of the 74HC244 is 5 V. The two devices have a level conversion problem in the interface circuit and cannot be directly connected. LVCl*5 is used as a level converter to achieve level matching between TMS320F2812 and 74HC244.



2.3 Human-machine interface

The human-machine interface uses the LCMl68651 liquid crystal module as the display and an interrupt-mode matrix keyboard. The task of the human-machine interface is mainly to receive keyboard instructions, complete instrument settings, calibration, measurement and other operations, provide real-time concentration change curve drawing, and provide query and display functions for historical data.

2.4 USB interface circuit

The USB interface is used for communication between the lower computer and the upper computer. The data transmission rate of USB is very high, so it can not only be used to transmit commands, but also can transmit data in real time, including original measurement values, current concentration values, or historical records. Figure 4 shows the USB interface circuit schematic.



2.5 Data storage

The data that the infrared analyzer needs to store include: gas measurement record number, date, time, concentration value, temperature, humidity, atmospheric pressure, etc. The amount of recorded data depends on the size of the saved items, the type of storage, the frequency of storage and the duration of storage. Considering the requirements of online measurement and unmanned measurement, a large-capacity flash memory K9F1G08UOA is used as the data storage medium of the instrument, and the total data storage capacity is greater than 128 MB.

3 System software design

The quality of software processing directly determines the processing speed of the system and the accuracy of the calculation results. Figure 5 shows the main process of the instrument software. The basic principle is to use the ring buffer to cache data and other information as necessary, so as to achieve higher-speed measurement and data transmission without affecting the human-computer interaction experience of the system. The most time-consuming and performance-impacting part of the entire software processing process is the processing of raw data, which involves filtering, environmental and detector compensation correction, elimination of gas absorption cross interference and other algorithms. In this instrument, these algorithms are specially optimized for DSP to ensure the fast operation of the system.



The system first completes initialization, including initialization of DSP and its peripheral circuits, creation and setting of ring buffers for raw data, target data, keyboard and other information, etc. Set the modulation frequency of infrared light and start the instrument, the instrument will enter the waiting state, the user can set the parameters of the instrument at this time, the key or remote start command will make the instrument enter the measurement state. In the measurement state, the instrument will cycle through to determine whether there is data or command to be processed, and then perform corresponding processing, such as remote command execution, raw data calculation, key command execution, target data transmission and storage, LCD interface refresh, etc.

Various raw data will be read by the interrupt program according to the set sampling rate and stored in the raw data ring buffer. The raw data includes the readings of the measurement channel and the reference channel, various compensation signals, measurement time, etc. After the raw data is processed. The calculated target value is placed in the target data buffer. If the system has settings for transmission, display or saving, the data in the target data buffer will be used in sequence. All ring buffers will maintain their own read and write pointers and modify them after the corresponding operations are completed. Some commands with higher priority and special cases are not restricted by the above process. For example, operations such as stop commands will be directly processed in the interrupt.

4 Experimental results

The specific gas types measured by this instrument are not mentioned in the previous article. The reason is that this instrument can be used to measure the concentrations of various gases, such as CO2, CO and HC, by replacing filters of different wavelengths and making corresponding parameter modifications. In the experiment, CO2 was selected as the test gas type. At an ambient temperature of 25°C and a standard atmospheric pressure, this instrument was used to measure various standard concentrations of CO2 gases certified by the national metrology department. The actual results show that the absolute error is within 0.3% and the relative error is within 2%, which has good measurement accuracy.

5 Conclusion

Infrared gas analyzers involve multiple disciplines such as optics, mechanics, electronics, computers, communications, and information fusion, and require relatively high software and hardware design and integration capabilities. The infrared gas analyzer described in this article, with the help of the powerful computing power of SP, not only meets the requirements of high-speed measurement, but also greatly enhances the portability and installability of the instrument because it is separated from the host computer. Through on-site operation and debugging, many advantages of the instrument have been demonstrated, such as simple and reliable structure, convenient installation and maintenance, convenient operation, long-term data storage, USB expansion, etc. I believe that with the upgrading of online gas analyzers, the application prospects of this instrument will become wider and wider. (Electronic Design Engineering Author: Ren Yumiao Bi Xueqin)