Research on low power consumption of wireless MEMS methane sensor

Aiming at the problem that the high power consumption of traditional methane sensors does not realize the comprehensive detection of coal mines underground, based on MEMS methane detection technology, wireless methane sensing low-power research is proposed. First of all, based on the analysis of power correlation, we select the modularized design with low power, which includes power supply, sensing element driver, acquisition, and processing modules. We study the change of sensitive resistance, response time, and power consumption of the MEMS methane sensing element and sensor in different operating modes. Meanwhile, experimental tests are carried out. The experimental result shows that under the premise of ensuring the timeliness and stability of methane gas detection, the power consumption of the wireless MEMS methane sensor in one minute is only 2.4% of the traditional catalytic methane sensing element.


Introduction
As an important traditional energy industry in China, to enhance the level of coal mine intelligence and promote the high-quality development of China's coal industry, the National Energy Administration clarifies the goals of coal mine intelligence development and forms intelligent devices for comprehensive perception, real-time interconnection, dynamic prediction, and other collaborative controllers [1] .Comprehensive detection in coal mines cannot be separated from the data support of front-end sensors, and methane monitoring is the top priority of the environmental management of coal mines [2] .The commonly used methane sensors for detection include catalytic, thermal conductivity, infrared, and laser spectral absorption principles.The catalytic principle is used for low-concentration detection at a low price.Thermal conductivity is only used for high-concentration detection of over 4%, while infrared and laser can be used for full-range detection, but it is expensive [3] .The catalytic combustion methane sensor has the highest cost-effectiveness, the highest usage, and the highest power consumption.It is wired to transmit at a fixed electricity level, with fewer sensors on load, short transmission distance, and portable and wireless devices that require more battery energy.Its working time is short, making it impossible to achieve comprehensive underground detection in coal mines, and seriously restricting the intelligent development of the coal mining industry [4] .
With the development of MEMS sensing technology, compared with traditional methods, MEMS sensing has the characteristics of low power consumption, easy integration, and intelligent implementation [5] .More and more scholars are investing in the research of MEMS sensing technology.Su et al. [6] are researching MEMS methane catalyst materials, Zhou et al. [7] are researching ultra-low power gas detection MEMS microheaters, and Dominik et al. [8] are researching MEMS hydrogen sensors for automotive detection.This article uses MEMS methane sensing elements for power consumption correlation analysis, conducts low-power research on wireless MEMS sensors, provides reliable technical solutions for product development, and promotes intelligent development.

Power consumption correlation analysis
The sensor power consumption means the loss of power during its normal operation.Sensor power consumption can be mainly divided into dynamic power consumption and static power consumption.
total P is the total power consumption, and static P is the static power consumption when powered on but not working, which is related to cc V .As cc V increases, the static current of static I increases, too [9] .
Dynamic power consumption ( dynamic P ) is the power consumption caused by state changes during normal operation, consisting of the charging and discharging power consumption C P of the load capacitor and the instantaneous conduction power consumption TC P .The power consumption of the product during the state transition process of the TC P signal is low.dynamic P mainly comes from C P , accounting for over 80% of the total power consumption of the circuit [10] .
 is the activity factor which means the probability of changes in circuit nodes; C is the load capacitor; cc V is the power supply for the circuit; f is the clock frequency at voltage cc V ; TC I is the time average of the pulse current.According to the above analysis, to reduce power consumption, the first step is to reduce dynamic power consumption.When the power supply voltage meets the application requirements, we try to reduce the type and voltage value of the power supply voltage as much as possible.When there are no special requirements for clock frequency, it can be reduced to the lowest value.Static power consumption mainly aims to reduce the voltage value of the power supply.At the same time, hardware modular design and power management can be adopted to turn off the power supply of nonworking modules to reduce the static current value.This article mainly conducts low-power research on wireless MEMS methane sensors through hardware modular design, reducing power supply voltage, working mode and frequency, and other methods.

Hardware design
The MEMS methane sensor adopts a modular design approach, mainly consisting of a power processing module, sensing elements and control driver module, signal processing module, microprocessor, and output module.The module diagram is shown in Figure 1.
The power processing module is used to supply power to each module, the power supply voltage should be minimized as much as possible, and the selection of components should minimize the inconsistency of power supply voltage among various modules.The power supply voltage of the MEMS methane sensing element needs to be 3.3 V, the power supply range of the microprocessor and output module is 1.7 V~5.5 V, and the power supply range of the signal processing module is 2.0 V~5.5 V.This article adopts a unified 3.3 V power supply.The control driver module controls the sensing components and signal processing circuits to power on or off through a microprocessor.The MOS transistor driving switch circuit used in this article has low conduction resistance, low static power consumption, fast switching speed, and low driving current.To improve sampling accuracy and anti-interference ability, the MEMS methane sensing element is shaped and amplified by an operational amplifier and then passed through a low-pass filter.It is directly converted to a digital signal by a 16-bit analog-to-digital conversion chip, and the sensor supply voltage is collected to detect power supply stability, as shown in Figure 2. The output and microprocessor units of this article use low-power Bluetooth with built-in microprocessors for data exchange.At the same time, the Bluetooth temperature detection function can be used to obtain the temperature of the tested environment, compensate for the temperature impact of the sensing components, and improve environmental adaptability.In application, only a simple peripheral circuit is needed.

Software design
The main function of MEMS methane sensor software is to control the power supply voltage of the sensing element through the microprocessor port.The microprocessor collects the sensing element signal and detects environmental temperature parameters.Then it compensates and processes them, converts them into concentration values, and outputs wirelessly.The entire software framework of the device is shown in Figure 3.The microprocessor needs to control the sensing element to turn on and cannot be completely dormant.The sensing element has switching losses during conduction and cut-off, which are usually much greater than conduction losses.Moreover, the faster the switching frequency is, the greater the loss is.We try to shorten the switching time to reduce the loss during each conduction and the switching frequency, thus reducing the switching frequency per unit of time.
The low-power design is achieved by adjusting key parameters through low-power management units, setting extended wireless connection intervals, broadcast intervals, and their durations, opening or closing corresponding peripheral resources, reducing the operating frequency of peripherals, and enabling the processor to operate in a low current state.Protocol optimization is used to avoid duplicate data transmission.

Experimental testing
When MEMS methane sensing elements react with the target gas, the heating resistor uses a direct current power supply to heat the sensitive material to a specific temperature to react with the target gas.The power consumption of the sensing element mainly comes from the heating resistor.Based on not affecting the performance of the sensing element, to reduce power consumption, the performance of the sensing element is studied by comparing the changes in component power consumption, response time, and sensitive resistance between continuous heating and pulse heating.
The experiment used four methane-sensing elements to test the power consumption under a continuous power supply.We applied a heating voltage of 3.30 V at both ends of the heating resistor.By recording the resistance values in air and a 0.50% methane environment, the sensitivity of the element is shown in Table 1 The response time test is under the condition of continuous power supply and heating, with the sensing element operating stably in a clean-air environment.0.50% methane gas is introduced, and the change in resistance over time is recorded, as shown in Figure 4. From the results, it can be seen that under continuous heating, the response time of MEMS sensing elements is very fast, reaching a response value of 90% within about 3 seconds.To further reduce power consumption, the performance of methane sensing elements was tested in a low-power mode by using pulse power heating.According to the "AQ6203-2006 Low Concentration Catalytic Methane Sensor for Mining", the response time shall not exceed 20 seconds.Based on not affecting the response time, its pulse working cycle shall be set as 20 seconds.When studying the pulse heating of sensing elements, the pulse is tested under different duty ratios of 15%, 25%, and 50%, and the impact on the performance of the sensing element is analyzed.According to the material-sensitive characteristics of the metal oxide methane sensing element, the sensing element is unheated and the sensitive resistance value is at the MΩ level.In the heating state, the resistance value of the sensitive resistor is below 20 kΩ.Prototype 1 was tested and the distribution of sensitive resistance at high and low levels during pulse heating in air is shown in Figure 5. From Figure 5, it can be seen that as the pulse high-level heating time increases, the resistance value of the sensitive resistor continues to increase and then tends to stabilize.At a duty cycle of 25%, the sensitive resistance of the sensing element reaches a stable state of 50% duty cycle, and the average power consumption of pulse heating is reduced by 75% compared to continuous heating.
The power consumption of MEMS methane wireless sensors varies under different working modes.The power consumption of the sensor is tested under relative idle mode, continuous heating mode, and pulse heating mode.The average power consumption for testing is shown in Table 2.As the mode of the pulse cycle during 20 s is heating for 5 s, the average power consumption per minute is only 8.40 mW.The power consumption of traditional catalytic elements is 360 mW, which is only 2.4% of the traditional catalytic methane sensing elements.The practical application of MEMS methane sensors generally adopts a multi-point layout and wireless networking method.Low power design can change the connection interval, slave latency, and protocol optimization to reduce data retransmission based on actual application scenarios, further reducing power consumption.

Conclusion
With the rapid development of Internet of Things technology, the research and development of lowpower methane wireless sensors is of great significance.This article is based on MEMS methane sensing technology for low-power device software and hardware design.While ensuring sensor performance, it reduces the power consumption of methane sensors by controlling the opening and closing of sensing elements and optimizing signal acquisition, processor sleep management, and data output.It also studies the changes in sensitive resistance, response time, and power consumption of MEMSmethane-sensing elements under different working modes.We conduct various performance comparison tests to reduce sensor power consumption in combination with practical applications.The paper provides reliable technical solutions for the development of wireless mine perception products and plays a huge promoting role in the development of underground IoT perception in coal mines.

Figure 2 .
Figure 2. Diagram of drive control and signal processing.

Figure 3 .
Figure 3. Software framework diagram.The microprocessor needs to control the sensing element to turn on and cannot be completely dormant.The sensing element has switching losses during conduction and cut-off, which are usually much greater than conduction losses.Moreover, the faster the switching frequency is, the greater the loss is.We try to shorten the switching time to reduce the loss during each conduction and the switching frequency, thus reducing the switching frequency per unit of time.The low-power design is achieved by adjusting key parameters through low-power management units, setting extended wireless connection intervals, broadcast intervals, and their durations, opening or closing corresponding peripheral resources, reducing the operating frequency of peripherals, and enabling the processor to operate in a low current state.Protocol optimization is used to avoid duplicate data transmission.

Figure 4 .
Figure 4. T90 test in the continuous heating state.According to the "AQ6203-2006 Low Concentration Catalytic Methane Sensor for Mining", the response time shall not exceed 20 seconds.Based on not affecting the response time, its pulse working cycle shall be set as 20 seconds.When studying the pulse heating of sensing elements, the pulse is tested under different duty ratios of 15%, 25%, and 50%, and the impact on the performance of the sensing element is analyzed.According to the material-sensitive characteristics of the metal oxide methane sensing element, the sensing element is unheated and the sensitive resistance value is at the MΩ level.In the heating state, the resistance value of the sensitive resistor is below 20 kΩ.Prototype 1 was tested and the distribution of sensitive resistance at high and low levels during pulse heating in air is shown in Figure5.

Figure 5 .
Figure 5. Resistance changes of sensing elements at high and low levels during pulse heating.From Figure5, it can be seen that as the pulse high-level heating time increases, the resistance value of the sensitive resistor continues to increase and then tends to stabilize.At a duty cycle of 25%, the sensitive resistance of the sensing element reaches a stable state of 50% duty cycle, and the average power consumption of pulse heating is reduced by 75% compared to continuous heating.The power consumption of MEMS methane wireless sensors varies under different working modes.The power consumption of the sensor is tested under relative idle mode, continuous heating mode, and pulse heating mode.The average power consumption for testing is shown in Table2.Table2.Average power consumption test for different working states.

Table 1 .
. Continuous heating power consumption and sensitivity testing of methane elements.

Table 2 .
Average power consumption test for different working states.