JPH01316652A - Gas detector - Google Patents

Abstract

PURPOSE:To reduce the power consumption of a gas sensor by using the gas sensor by intermittently heating the sensor by applying heating pulses having a width which is smaller than the thermal time constant of the heater of the sensor to the heater. CONSTITUTION:A semiconductor substance 6 of a metal oxide, the resistance value of which is changed by a gas, is heated by applying heating pulses to a heater 4 and the gas is detected based on the resistance value of the substance 6. By constituting a pulse power source of oscillation circuits Osc 1 and Osc 2, the circuit Osc 2 is actuated by the output of the circuit Osc 1 and the heater 4 is heated through a transistor Tr. The width of the heating pulses applied to the heater 4 is made shorter than the thermal time constant of the heater 4, for example, the 1/2 of the time constant and, at the same time, intervals of the pulses are made longer than the time constant, for example, desirably >=2 times as long as the time constant. In addition, the duty ratio of the heating pulses is set to <=1/20.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to gas detection using a change in the resistance value of a metal oxide semiconductor. This invention particularly relates to reducing the power consumption of the gas sensor used.

[Prior Art] Reducing the power consumption of gas sensors is one of the fundamental challenges in the field of gas detection. Research to date has focused on miniaturizing sensors.

However, it is true that there is a limit to the miniaturization of sensors.

As for the invention, we considered operating the gas sensor by intermittently heating the gas sensor for a very short period of time.

We have discovered that it is possible to detect gas by setting the heating time of the sensor to less than the thermal time constant. It was also found that the interval between heating can be quite long. For example, for a gas sensor with a thermal time constant of 100 m5ec, in the specification of 2003, the time required for the sensor to rise in temperature from the start of heating to 90% of the steady heating temperature is defined as the thermal time constant. ), 1 per second
The sensor worked after applying heating pulses every 5 m5ec (see Figure 4).

This procedure reduces the power consumption of the gas sensor to I/l OOa degrees compared to continuous heating.

[Problem of the Invention 1 An object of the present invention is to reduce the power consumption of a gas sensor.

[Structure and operation of the invention] In the present invention, a heating pulse having a width equal to or less than the thermal time constant of the gas sensor is applied to the heater of the gas sensor, for example, a pulse having a width of 1/2 of the thermal time constant, and the sensor is intermittently heated for use. . The interval between pulses is longer than the thermal time constant of the sensor, preferably at least twice, more preferably at least 5 times longer than the thermal time constant of the sensor. Further, the duty ratio of the heating pulse is set to I/10 or less, preferably 1/20 or less.

The gas sensor can operate and detect gas even under such conditions. The power consumption of the sensor is determined by the power during heating, the width of the heating pulse, and the duty ratio. Therefore, by making the heating time for one time less than or equal to the thermal time constant and reducing the duty ratio, power consumption can be reduced.

The width of the heating pulse is less than the thermal time constant and the sensor does not reach the steady-state value of the heating temperature. However, if the pulse width is constant, the temperature reached by the sensor will also be constant, and no problem will occur. And it is easy to form a pulse with a constant width.

[Example] Gas sensor tank construction example In this example, a power-saving gas sensor developed by Nonda et al. (described in Japanese Patent Application No. 174,420/1982) was used. The feature of this gas sensor is that a heat-resistant insulating coating such as Al t O* is applied to the surface of a metal heating element such as a Fe-Cr-A1 alloy wire, and the metal oxide semiconductor is used as a carrier. That is, a metal heating element is used as a heater, and a metal oxide semiconductor whose resistance value changes depending on gas and its ff114 are provided on an insulating coating. The heater and the metal oxide semiconductor are insulated with an insulating coating.

In the example used, a Fe-Cr-Al alloy wire with a wire diameter of 20 μm was used as the metal heating element, a single Sn0w layer was used as the metal oxide semiconductor, and an approximately 1 μm thick A
It was made into a l t Oa film. This sensor consumes 70 mWatt when continuously heated to 300°C.

The applied voltage was about 0.6V, and the resistance value of the heating element was 5Ω. Further, the thermal time constant is 1oOnsec. 90% here
Since the thermal time constant of the response is 00 m5 ec, the thermal time constant for the response up to l/e is about 40 m5 ec. Here, e represents the base of the natural logarithm.

This gas sensor is used as an example of a sensor with low power consumption, and any other sensor may be used. For example, a structure in which a heater, a metal oxide semiconductor layer, and an electrode are provided on a thin glass film proposed by the inventor may be used (Japanese Patent Application No. 5596/1982).

6 to 8 show the gas sensor 2 used in the example. In Fig. 6, 4 is Pe-Cr-A with a wire diameter of 20 μm.
It is a metal heating element made of L alloy wire (Kantal manufactured by Gabrius of Sweden, Kanthal is a trade name). Reference numeral 8 denotes a heat-resistant insulating coating provided on almost the entire surface of the metal heating element 4. Application of alumina sol and thermal decomposition at 800° C. were repeated 10 times to obtain an alumina coating 8 with a thickness of about 17 zm. 6 is 5n
Metal oxide semiconductors such as Ot, IntO3, ZnO, etc.
Here, S n(OCH3)3(0(CHJPN Ht
) was added dropwise and thermally decomposed at 500°C to give 5nOt.

Reference numeral 10 and 12 are electrodes on which Au is vacuum-deposited, and reference numeral 14 is a heat-resistant insulating substrate made of alumina. Also, 16 is hollow,
This is to prevent the raw material solution from touching the substrate 14 during thermal decomposition of SnO.

18. 20. 22 are gold printed electrodes. Both ends of the metal heating element 4 are welded to the electrodes 18 and 22, and the electrode IO is fixed to the printed electrode 18 and the detection electrode 12 is fixed to the printed electrode 20 with gold paste 24. did.

In the enlarged sectional view of FIG. 7 and the overall view of FIG. 8, numeral 30 denotes a lead frame which also serves as an external pin, and each lead frame 30 and the printed electrodes +8, 20, and 22 are connected by lead wires. Further, 26 is a welded portion between the metal heating element 4 and the printed electrode 18.22, and 28 is a bonding layer between the lead frame 30 and the substrate 14. Furthermore, 32 is a heath made of synthetic resin or the like, and 3 and 1 are covers made of synthetic resin or the like.

Characteristics of the Gas Sensor FIG. 4 shows the relationship between the width of the heating pulse and the resistance value of the sensor 2. The sensor used was the one mentioned above. The horizontal axis represents the voltage applied to the metal heating element 4, and the vertical axis represents the resistance value of the metal oxide semiconductor 6 immediately before the end of the heating pulse. The atmosphere is 20
℃, humidity 65%, temperature during continuous heating is applied voltage 0.6■
It is about 3006C. Heating is 5 m5ec once per second
The heating pulse was performed with a width of −100 m5 ec. 5 m
The sensor is operating even with heating of 5 ec (1/20 of the thermal time constant) or 15 m5 ec (15% of the thermal time constant). Then the thermal time constant or its l / 2 n degree heating pulse (
100 msec or 50 m5ec), the characteristics are almost the same as in the case of continuous heating. Also, if the heating time is sufficiently shorter than the thermal time constant (5 m5 ec or 15 m5 ec
), it is preferable to make the voltage applied to the sensor slightly higher than in the case of continuous heating.

FIG. 5 shows the response characteristics to + o o ppm ethanol. The heating conditions were such that a pulse with a voltage of 0.7 V was applied once every second with a width of 5 m5 ec. In this example, a load resistor of IOKΩ and a power supply of 5 V were connected to the sensor 2, and the voltage across the load resistor was sampled immediately after the heating pulse ended. Sensor 2 has a fast response and can be used for practical purposes.

The table below shows the relationship between six types of heating conditions and characteristics.

Table-7* Continuous heating 250 6.5 5.0 3
．． 0100msec/sec 800 13
7.0 3.550m5ec/see 600
8.0 8.0 4.015m5ec/see
350 6.0 5.0 3.55m5ec
/see 300 4.0 4.5 4.5
15m5ec/1Osec 300 4.5 5
．． 0 4.015m5ec/ll1in 400
3.0 6.0 4.010 m5ec/mi
n* cxs I OMΩ IOMΩ IMΩ
*The gas concentration is 100 ppm each, and the sensitivity is the ratio of the resistance value in air and gas.In the last example, heating is performed once every minute with a pulse of 10+n5ea, and the resistance value immediately before the next heating is In this example, the resistance value in air cannot be measured, so the resistance value in gas is displayed.The heating voltage is 0.7V in each case, and the resistance value is displayed in ohms.

The following is clear from these data:

One heating time is 50% of the thermal time constant (50m5ec/5
ec), 15% (l 5 m5ec/5ee), 5%
(5m5ec/5ec), gas can be detected. No problem will arise if the interval between heating is set from 1 second (10 times the thermal time constant) to 1 minute (600 times the thermal time constant). Then, by applying a heating pulse that is equal to or less than the thermal time constant and increasing the interval between heating pulses, the power consumption of the sensor 2 is rapidly reduced. Furthermore, the response speed of sensor 2 is under the harshest conditions (5m5ec/sec
heating pulse) without any problem. Since all data so far have shown satisfactory results, there is no clear lower limit on the width of the heating pulse or upper limit on the pulse interval. According to estimates and 12, the lower limit of the width of the heating pulse would be 7j<, and the upper limit of the pulse interval would be about IO~30 minutes.

In Fig. 9, a voltage of 3.2 V is applied to the heater 4 once every second, and
The characteristics when adding 2 m5ec each are shown. 0.21se
Gas can also be detected with c/see heating pulses. Further, a delay with respect to the heating pulse is seen in the response of the sensor output Vout. Then, after approximately 10 m5 ec has passed after the end of the heating pulse, the sensor output returns to the value before the pulse.

This result suggests the following.

If the width of the heating pulse is extremely short, the response of the sensor output will be delayed from the pulse because it will take time for heat to be conducted from the heater 4 to the metal oxide semiconductor 6. In this region, the energy per pulse is more important than the pulse width or power. In other words, the lower limit of the pulse width has no particular value, and may be within a range that is possible in terms of the circuit and allowed by the withstand voltage of the heater 4. This lower limit is Iμse
It would be around c.

If the width of the heating pulse is set to 4 degrees shorter than the thermal time constant (100 m5ec), the heating pulse will end when the temperature of the sensor 2 becomes stable.However, in reality, this is not a particular problem. The pattern of temperature change of sensor 2 is
It is determined by the power supply voltage and the oscillation conditions of the oscillation circuits 0scl and 0sc2. Therefore, if these things are stable, the heating pattern will be constant and no problems will occur. It is extremely easy to obtain a stable iEb and a stable oscillation circuit 0scl or 0sc2.

Furthermore, as is evident from the last data in Table 1, the heating pulse is used for heat cleaning of sensor 2;
Detection may be performed after cooling.

A specific example of the circuit studied by the inventor is shown below along with numerical constants.

Example of K-way circuit FIG. 1 shows a circuit diagram of the embodiment. In the figure, r>t
+ is an appropriate power supply, here 1.5VX2 3V7[i
source. This is intended for battery use of the device. 0scl is an oscillation circuit that emits a pulse with a width of 10m5ec once per second, and 0sc2 is an oscillation circuit that generates a pulse with a width of 25μ at 500μ5ecfi.
This is an oscillation circuit that emits pulses of sec.

Then, using the output pulse of the oscillation circuit 0scl as a strobe signal, the oscillation circuit 0sc2 is caused to oscillate. 0sc1.0
The sc2 constitutes a pulse power source. However, when a logic circuit such as a microcomputer is used as the auxiliary circuit, the pulse power supply may be built into the logic circuit such as the microcomputer.

2 is the gas sensor listed in 11η, Tr is a transistor connected to the metal heating element 4, and you can change it to any switch you like.

It I is the load resistance of the sensor 2, and the voltage across it is the sensor output Vout.

A1. A2 is an operational amplifier, D1 to D3 are diodes, C
I is a capacitor, R2-R4 is a 1° resistor, and these will form a peak hold circuit. That is, the peak of the output Vout is accumulated in the capacitor CI, and this is transferred to the operational amplifier A.
Take it out from 2. R5 is the output resistance of operational amplifier WA2, M
is a meter for displaying detection results. Load resistance R
The meters from I to M are used as gas detection means. The reason for using the peak hold circuit here is as follows. The sensor output VouL varies depending on the heating pulse. The signal for which the normal amount is also significant is the signal at the point near the end of the heating pulse where the sensor temperature is the highest.

At this time, due to the temperature dependence of the sensor, the output Vou
t indicates a peak. This peak is then captured and processed by gas detection means. Furthermore, the duration of the peak of Vout is generally short, and signal processing is facilitated by bolding the peak with a peak hold circuit.

FIG. 2 shows a circuit that specifies the sampling timing of the sensor output Vout. In this circuit, the oscillation circuit Oscl
At the end of the strobe signal, the A/D conversion circuit A/D
Trigger Enono. Note that it is preferable that sample 7 has a built-in lead circuit in Al1) conversion circuit A/D. Without a built-in sample and hold circuit, V out
Since the sampling time is short, a high-speed A/D conversion circuit is required. Such circuits are generally expensive. Then, the A/D converted output is sent to the display circuit D isp! ay and displays the output of the A/D conversion circuit as control output Control.
This circuit detects the gas based on the resistance value of the metal oxide semiconductor 6 at the end of the heating pulse. However, the sampling timing of the output is arbitrary, and may be, for example, immediately before the end of the heating pulse, or after the sensor 2 has cooled down.

FIG. 3 shows the operation of the circuit of FIG. 1. Oscillation circuit 0sc
The oscillation circuit 0sc2 is operated by the output pulse 1, and the metal heating element 4 is heated via the transistor Tr. The duty ratio of the oscillation circuit 0sc2 is 1/20, and the output of the power source Eb is 3■, so this means that the metal heating element 4 is heated with a voltage of 0.7V during the output pulse (width 10m5ec) of the oscillation circuit 0scl. equal.

In this way, from a 0.7V power supply to a 3V or 5V power supply,
This is because it is easier to obtain power sources such as Oscillation circuit 0scl
The output is a pulse with a width of 10 m5ec once per second, and the power consumption of the sensor 2 is reduced to about 1/100 compared to the case of continuous heating.

Then, gas is detected from the peak value of the output Vout (in the case of FIG. 1) or the output VouL at a specific sampling period (in the case of FIG. 2).

[Effects of the Invention] According to the present invention, it is possible to reduce the power consumption of the gas sensor.

[Brief explanation of the drawing]

Fig. 1 is a circuit diagram of the embodiment, Fig. 2 is a circuit diagram of a modified example, and Fig. 3 is an operation waveform diagram of the embodiment. FIG. 2) shows the waveform of the sensor temperature, and FIG. 3) shows the waveform of the sensor output. FIG. 4 and FIG. 5 are characteristic diagrams of the embodiment. Figure 6 is a front view of the main parts of the gas sensor used in the example, Figure 7
The figure is an enlarged sectional view of a main part of the gas sensor used in the example, and FIG. 8 is a sectional view of the gas sensor used in the example. FIG. 9 is a characteristic diagram of the example. In the figure, 2 gas sensor, 1 fully mature metal body, 6 metal oxide semiconductor.

Claims (2)

[Claims] (1) In a gas detection device that detects gas by heating a metal oxide semiconductor whose resistance value changes depending on the gas using a heater, a heating pulse is applied to the heater to intermittently heat the gas sensor. A pulse power source is provided, the width of the heating pulse is made shorter than the thermal time constant of the gas sensor, the interval between heating pulses is made longer than the thermal time constant, and the duty ratio of the heating pulse is set to 1/10 or less. A gas detection device featuring: (2) The gas detection device according to claim 1, characterized in that the width of the heating pulse is 1/2 or less of a thermal time constant, and the duty ratio of the heating pulse is 1/20 or less. gas detection device.