RU2523089C2 - Calibration of semiconductor gas sensors and device to this end - Google Patents

Abstract

FIELD: instrumentation.

SUBSTANCE: invention relates to analysis of gases. Proposed method is implemented with the help of hardware-software measuring complex. It consists in that sensor sensitive element is heated cyclically in clean air preset (PGC-1) number of time (K times) to temperature T1 and cooled to temperature T2. Then, during the next K cycles of heating and cooling test gas mix PGC-2 is fed to sensitive element area. Then, during the next K cycles of heating and cooling test gas mix PGC-3 is fed to sensitive element area. Then, during the next K cycles of heating and cooling test gas mix PGC-N is fed to sensitive element area. Family is constructed from N=4 time dependencies of conductivity of gas-sensitive ply σ(t) for every gas mix and for fixed time moment in the cycle ti to define calibration characteristic. Calibration curve thus constructed is approximated and loaded in the processor of intelligent gas module sensor. Said processor is mounted in gas analyser for sensor processor is polled by central processor of gas analyser to output readings of measured gas concentration at its display.

EFFECT: higher accuracy and validity of calibration.

5 cl, 14 dwg

Description

The invention relates to the field of gas analysis and the study of physical and chemical processes on the surface of semiconductor materials and is aimed at improving the quality of gas control devices manufactured by domestic manufacturers.

The invention can be used in the field of gas analytical instrumentation in terms of its metrological support - calibration by means of calibration gas mixtures in the mass production of intelligent semiconductor gas sensors to control impurities of methane, propane, hydrogen, carbon monoxide, ammonia in the air of the working zone of industrial enterprises.

A known method for determining the health of a semiconductor gas sensor, consisting in the analysis of the Fourier spectrum of the time dependence of the conductivity of the gas-sensitive layer of the semiconductor sensor during its cyclic heating and cooling in the temperature range of gas adsorption in the atmosphere of which the sensor is located (see patent RU No. 2396554, G01N 27/00 , 2010).

The disadvantage of using this method for calibrating gas sensors of a semiconductor type is the low accuracy and selectivity of the measurement, since the optimization of the composition and operating temperature of the semiconductor material of the sensor does not completely suppress the possibility of interaction with gaseous impurities and additional influencing factors.

Closest to the proposed invention, the method of possible calibration of the gas sensor of the semiconductor type is the method according to Patent RU No. 2371709, G0127 / 14, 2009.

The method consists of determining the concentration of hydrogen in the presence of gaseous impurities and consists in measuring the electrical signal at the output of a semiconductor sensor with a sensitive layer of metal oxide when it is heated to a given temperature, the value of the conductivity of the sensitive layer of the semiconductor sensor is determined by the value of this signal, compare it with the previously obtained calibration value and determine the concentration of hydrogen. In this case, the signal at the output of the semiconductor sensor is measured continuously, cyclically heating it to a temperature T1 and cooling it to a temperature T2. The time derivative of the conductivity of the sensor sensitive layer is determined over the time interval between the end of heating to temperature T1 and the end of cooling to temperature T2. The conductivity value, which is a function of gas concentration, is determined. Then, the presence and number of local minima of the dependence of the conductivity of the sensitive layer on time in the interval between the end of heating and the end of cooling is determined, and if there are two such minima, the electrical signal at the output of the semiconductor sensor is measured at the time between the first and last local minimum, in where the absolute value of the time derivative of the conductivity reaches a minimum, if the local minimum was one, then the electrical signal at the output of the semiconductor sensor and measured at the time between the end of heating and the last local minimum, in which the absolute value of the derivative of conductivity in time reaches a minimum, and the value of the measured electrical signal judges the conductivity of the sensitive layer of the semiconductor sensor, which determines the concentration of hydrogen.

The disadvantage of this method is the lack of a practical, applicable in the mass production of intelligent semiconductor sensors sensor calibration technique that requires more functionality and capabilities of a structured database of calibration results.

The objective of the invention is a practical, applicable in industrial mass production calibration of gas sensors of the semiconductor type to control the concentration of gas components of CH4, C3H8, H2, CO, NH3 in the air of the working area of industrial enterprises.

The task is carried out as follows.

In the proposed method, in accordance with GOST 13320-81 GAS ANALYZERS INDUSTRIAL AUTOMATIC calibration of the semiconductor sensor is carried out using a hardware-software measuring complex and consists in the fact that the sensor element is cyclically set (K times) in clean air (PGS-1) to a temperature of T1 and cooled to a temperature of T2, then during the next K heating and cooling cycles a calibration gas mixture PGS-2 is supplied to the area of the sensor, then for the next K cycle OVS serves calibration gas mixture PGS-3 in the region of the sensitive element, then during the next K cycles serves the calibration gas mixture PGS-N in the region of the sensitive element, build a family (in figure 2, 4) of N = 4 time dependences of the conductivity of the gas-sensitive layer σ (t) for each gas mixture and for a fixed time ti in the cycle, determine the calibration characteristic

$${\sigma }_i^k,ti,T\mathrm{one}i,T2i,\Delta {12}^i=F\left(C{i}^k\right),\left(\text{one}\right)$$

where k = 1,2,3, ..., N; i = 1-CH4, i = 2-C3H8, i = 3-H2, i = 4CO, i = 5-NH3; ti is the point in time measured from the start of heating of the sensor at which N measurements of the conductivity take place in the 6, 12, 18, ... 6 × N cycle of heating and cooling the sensor; T1i is the heating temperature of the sensor in the ith gas environment, T2i is the cooling temperature of the sensor in the ith gas environment, Δ12 ⁱ is the cooling time of the gas sensor element from T1 to T2; Ci ^k is the concentration of the ith gas component in the PGS-k calibration gas mixture supplied to the sensor.

The functions σ (t) are continuous functions with the first and second derivatives dσ / dt and dσ / dt (dσ / dt) characteristic of each gas.

по семейству из N реализации σ(t) нагрева и охлаждения чувствительного элемента сенсора с длительностью цикла Δtc, получаемых обработкой электрического сигнала от интеллектуального газового сенсора, осуществляемой пакетом программ SEMSENSOR и MATLABSEM в интегрированных операционных системах Silicon Laboratories IDE и MATLAB 7.01 персонального компьютера с предварительным вводом данных калибровки в меню КАЛИБРОВКА/ВВОД ДАННЫХ и последующей ЗАПИСЬЮ ДАННЫХ калибровки на экране монитора, выбором временного сечения, вычислением и сглаживанием кусочно-аппроксимируемой градуировочной характеристики (1), сохранением семейства из N реализаций и градуировочной характеристики в базе данных персонального компьютера.Calibration characteristics for each i-th gas are determined using the time section algorithm - determining at a fixed point in time ti N values $${\sigma }_i^k$$

for a family of N implementations of σ (t) heating and cooling a sensor element with a cycle time Δtc obtained by processing an electric signal from an intelligent gas sensor implemented by the software package SEMSENSOR and MATLABSEM in the integrated operating systems Silicon Laboratories IDE and MATLAB 7.01 of a personal computer with preliminary input calibration data in the CALIBRATION / DATA ENTRY menu and then RECORDING calibration data on the monitor screen, selecting a temporary section, calculating and smoothing piecewise approximations uemoy calibration characteristics (1), preserving the family of implementations and N calibration characteristics in a PC-based data.

The calibration characteristic obtained in this way is approximated and loaded into the processor of an intelligent semiconductor sensor, which is installed in the gas analyzer, where in the operational mode the measurements are interrogated by the central processor of the gas analyzer and the readings of the measured concentration of the gas component are displayed on its display. The invention allows the calibration of semiconductor smart sensors with increased accuracy and reliability in terms of their mass production.

Such a construction of a method for calibrating semiconductor gas sensors provides a solution to the problem - high-quality calibration of semiconductor sensors in mass production and additionally allows you to determine the physicochemical properties of semiconductor materials with catalytic additives in order to optimize the composition and manufacturing technology of the sensor.

Figure 1, 2, 3, 4, 5, 6 shows a structural diagram and a device for implementing the method of calibration of semiconductor sensors.

Figure 7 shows a family of temporary realizations of σ (t) - the dependence of the conductivity of the gas-sensitive layer of the semiconductor sensor for different concentrations of methane when it is fed into the sensor region from high-pressure cylinders with calibration gas mixtures PGS-1, PGS-2, PGS-3, PGS -4, ПГС-5 through a supply tube, dressed on the fitting of a 5-cell chamber, on top of which 5 intelligent gas modules are fixed.

On Fig depicts the calibration characteristic and the inverse smoothed calibration characteristic for methane.

Figure 9 shows a family of 5 temporary realizations of σ (t) - the dependence of the conductivity of the gas-sensitive layer of the semiconductor sensor for different concentrations of hydrogen.

Figure 10 shows the calibration characteristic and the inverse smoothed calibration characteristic for hydrogen — calibration gas mixtures PGS-1, PGS-6, PGS-7, PGS-8, PGS-9.

Figure 11 shows a "window" of the Silicon Laboratories IDE operating system with the Semsensor software package for setting the heating mode and measuring the concentrations of CH4 and C3H8, broadcasting and loading the calibration curve into the C8051F410 processor of an intelligent gas semiconductor sensor.

12 shows a “window” of the MATLAB 7.01 operating system with the Matlabsem software package for calibrating semiconductor sensors.

On Fig, 14 shows the electrical and wiring diagram of the printed circuit board of the intelligent semiconductor sensor.

A device for implementing a method for calibrating semiconductor gas sensors in particular SG21XX-A contains a gas electronic module consisting of a primary gas-sensitive transducer - sensor (in particular, SG21XX-A sensor in a typical TO-2 package with a metal grid in the lid) with four output electrodes inserted into the four holes of the PLSS type connector soldered into the printed circuit board of the electronic module. The heating mode of the gas-sensitive layer of the sensor is programmed during calibration and is performed by the C8051F410 processor by supplying from the internal digital-to-analog converter an IDACO electric signal - voltage U (t) of a given shape to the positive input of the operational amplifier (ОУ) AD8542AR, and the output voltage of the ОУ is then supplied to the ceramic-metal heater of the sensitive element heating it by supplying power from 50 to 500 degrees Celsius with measuring the steady-state temperature of the heater T1 (t1) by voltage on the heater STUDIO Un (t) and the output signal from the op amp AIN0 AD623AR (DA5) and heater current Ir (t) on the output signal AIN1 DU AD623AR (DA1: 1) according to the formulas

$$\begin{array}{l}T\mathrm{one}=\mathrm{twenty}+\left(\mathrm{one}/\alpha \right)\left(\left(Rn\left(ti\right)/R\mathrm{twenty}\right)-\mathrm{one}\right)\\ \mathrm{and}s\left\{Rn\left(ti\right)=R\mathrm{twenty}(\mathrm{one}+\alpha \left(T\mathrm{twenty}\right)Rn=Rn\left(ti\right)/In\left(ti\right)\right\}\left(\text{3}\right)\end{array}$$

where 20 is room temperature, α - TCS is the temperature coefficient of resistance, R20 is the heater resistance at room temperature in clean air at Un = 0, In = 0, t1 is the time section in 5 calibration implementations σ (t);

with measurement of σ (ti) according to the electrical signal voltage AIN2 according to the formulas

$$\begin{array}{l}\sigma \left(ti\right)=\mathrm{one}/Rs\left(ti\right),Rs\left(ti\right)=Us/Is,\\ Us=+3.3B-AIN2Is=AI\mathrm{<figure-callout\; id="2"\; label="N"\; filenames="00000005.png,00000017.png"\; state="\{\{state\}\}">N</figure-callout>}2/R16\left(\text{four}\right)\end{array}$$

Rs, Us, Is - resistance, voltage, current of the gas-sensitive layer

The microprocessor C8051F410 was used to process the analogue signal AIN2 from the electric secondary converter of the concentration of the gas component, which provides the secondary conversion of the conductivity of the gas-sensitive layer to a digital signal transmitted via the interface interface on the basis of the SPI protocol to the gas analyzer central processor AT91SAM7X128 (into which the smart gas sensor is installed) and a normalized output analog electric signal U (t) from 0 to 3 V, proportional to the conductivity g the sensitive layer, which is analyzed by a hardware-software complex based on a personal computer using the SEMV1N and MATLABV_2 software packages implemented in the integrated operating systems Silicon Laboratories IDE and MATLAB 7.01, and when supplying calibration gas mixtures from cylinders with an open valve through an adjustable needle leak and so on through a silicone tube with a length of 50 cm and an inner diameter of 4 mm through a nozzle cap into the hole with a metal mesh cover of the housing of the semiconductor sensor SG21XX carry it calibration, as a result of which the calibration characteristic is written in the memory of the processor of the smart gas sensor according to the formula (1) - the dependence of the conductivity of the gas on the concentration of the gas component in the air.

Thus, the analysis of the prior art allows us to conclude that the proposed calibration method has a novelty, inventive level and has several advantages compared to the prototype.

Claims (5)

1. The method of calibration of semiconductor gas sensors, which consists in the fact that to determine the concentration of the gas component, in particular hydrogen, and calibrate the sensor in the presence of gaseous impurities, an electrical signal is measured at the output of a semiconductor sensor with a sensitive layer of metal oxide when it is heated to a given temperature, from the value of this signal, the conductivity of the sensitive layer of the semiconductor sensor is determined, stored, compared with the previously obtained calibration The hydrogen concentration is determined by this value, characterized in that the semiconductor sensor is calibrated using a hardware-software complex using the SEMSENSOR and MATLABSEM software package written in the assembler programming language implemented on a PC in the integrated operating environments Silicon Laboratories IDE and MATLAB 7.01, such so that a cyclically predetermined number of times (K times) is heated the sensitive element of the semiconductor sensor in clean air (PGS-1) to a temperature of T1 and cooled to a temperature of T2, then in t the next K cycles of heating and cooling feed the calibration gas mixture PGS-2 to the area of the sensor, then for the next K cycles feed the calibration gas mixture PGS-3 to the region of the sensor, then for the next K cycles feed the calibration gas mixture PGS-N into the region of the sensitive element, build a family of N = 4 time dependences of the conductivity of the gas-sensitive layer σ (t) for each gas mixture and for a fixed time ti in the cycle, determine the calibration characteristic, Rouge in intelligent processor semiconductor sensor 5 -

where k = 1, 2, 3, ..., N; i = 1-CH4, i = 2-C3H8, i = 3-H2, i = 4CO, i = 5-NH3; ti is the point in time measured from the start of heating of the sensor at which N measurements of the conductivity take place in the 6, 12, 18, ... 6 × N cycle of heating and cooling; T1i is the heating temperature of the HS sensitive element in the ith gas environment, T2i is the heating temperature of the HS in the ith gas environment, Δ12 ⁱ is the cooling time of the HS from T1 to T2; Ci ^k is the concentration of the ith gas component in the PGS-k calibration gas mixture supplied to the sensor. 2. The method for calibrating semiconductor gas sensors according to claim 1, characterized in that the calibration algorithm contains the initial (before calibration) testing of the operability of the smart module and the initial setting of the initial parameters of the heating mode and recording the identification data of the semiconductor sensor. 3. The method for calibrating semiconductor gas sensors according to claim 1, characterized in that there is an OscilloGRAPH calibration mode, which consists in an exponential law of increasing or decreasing the concentration of the gas component in the region of the sensing element, in which a family of up to 16 time can be built in one window realization of σ (t) with a heating and cooling cycle from 5 sec to 20 sec. 4. The method for calibrating semiconductor gas sensors according to claim 1, characterized in that the family of temporary implementations σ (t) in the CALIBRATION mode and calibration characteristics, any cycle in a single heating mode, in a cyclic training mode and stabilization of the sensor sensitivity, operating bias voltage graphs amplifiers of the electronic unit for normalizing the output electrical signal of the conductivity of the gas-sensitive layer can be recorded and archived into the database both in graphical form and in a text file. 5. The method of calibrating semiconductor gas sensors according to claim 1, characterized in that there is a graphical menu that allows you to carry out the transfer operations of curves with overlapping other graphs, insert explanatory texts anywhere on the graph, smooth out a piecewise-approximated curve, determine the equation of the curve, closest to the smoothed calibration characteristic (1).

Images