WO2006089789A1

WO2006089789A1 - Method and device for determining gas concentrations or vapor concentrations contained in the atmosphere

Info

Publication number: WO2006089789A1
Application number: PCT/EP2006/001785
Authority: WO
Inventors: Hanns Rump; Reinhard Patzer; Carsten Supply; Jessica Gerhart; Rosemarie Brand-Gerhart
Original assignee: Automotive Ag
Priority date: 2005-02-25
Filing date: 2006-02-27
Publication date: 2006-08-31
Also published as: EP1910811A1

Abstract

Heated metal oxide sensors are widely used for detecting oxidizable and reducible gases and vapors. In prior art, the gas-dependent resistance is evaluated in a measuring manner in the effective layer by means of the physical unit called ohm , represented by the symbol R. According to the invention, the electric charge is measured and evaluated using the physical unit "Coulomb", said electric charge changing in the effective layer of the metal oxide sensor in the presence of gas. Disclosed are advantageous and suitable technical electrical methods.

Description

Method and device for the determination of gas or vapor concentrations contained in the atmosphere

The invention relates to a method for the determination of reducible or oxidizable gas or vapor concentrations contained in the atmosphere, using a arranged in this atmosphere, electrically heatable metal-semiconductor sensor, sensor, with sensitive active layer, and with an electrical connected to the sensor Evaluation circuit for

Generation of a switching signal, according to the preamble of claim 1, and a device for carrying out the method according to the preamble of

Claim 17.

State of the art:

Since 1962 gas sensors so-called TAGUCHI sensors are known in which heated layers of certain metal oxides such as tin dioxide are exposed to the air to be monitored for the purpose of detection of airborne gases or vapors Figure 1, prior art. The reference numeral 1.1 designates. the substrate carrying the sensor layers, which can be made of ceramic or of silicon. 1.2. is a heater, which can be made as an electrical resistor, for example made of platinum or in the case of silicon sensors can also consist of polysilicon. The heater 1.2 tempered the active layer to a certain temperature, which is typically between 300-450 ⁰ C. The electrical resistance of the active layer is tapped off via a finger-like contact structure 1.3. The active layer 1.4 consists of metal oxide, and may have different particle sizes. As gas-sensitive metal oxide substances are known, such as tin dioxide, zinc oxide, tungsten oxide, indium oxide, gallium oxide and some other metal oxides.

When gassed with light oxidizable gases, such as hydrogen, the gas penetrates into the metal oxide structure and reacts with the oxide, which is partially reduced. As a result, the electrical resistance of the metal oxide decreases. Other gases or vapors first adsorb to the surface of the active layer and then react with the metal oxides. Depending on the affinity of the gases or vapors the metal oxide used as the active layer and its working temperature and also according to its grain size and distance of the contact electrodes to each other, a different response of the active layer is detected on the gas present at different gases.

The change in the electrical resistance of the active layer as a function of the gas concentration generally follows a very strongly curved curve Figure 2, prior art. At low gas concentrations, the change quotient Rs / Ro, see 2.1, is very high. At higher gas concentrations, the curve is increasingly flatter and virtually asymptotic. Changes in the gas concentration have virtually no effect on the then very low electrical resistance of the gas sensor in the saturation case. The saturation occurs relatively early. In alcohol vapor saturation is already at about 2000ppm vol at industrially available sensors. reached. Higher gas concentrations are no longer measurable, which adversely limits possible applications.

It is known to add catalytically active substances to the metal oxides. Frequently used are, for example, platinum Pt or palladium Pd. The spectral sensitivity of the sensor to gas mixtures of different oxidizable gases changes with the presence of catalytic admixtures because they burn efficiently at the heated and air exposed surface of the active layer at high concentration of the catalyst with the atmospheric oxygen and therefore the deeper portion of the active layer, where Help the electrode structure of conductance is measured, no more or only slightly affect.

It is known that the characteristic R = resistance of the active layer; R Gas in Figure 2.2, with catalytic admixtures doped, prior art sensors is increasingly steep and that the asymptotic, very flat part of the curve begins at even lower gas concentrations.

It is also known that the heated active layer is electrically complex, and not only the generally evaluated ohmic resistance of the active layer is a function of the gas supply. As an equivalent circuit diagram in the literature, a circuit according to Figure 3, prior art, indicated. Where:

- 3.1 Resistance in the mass of metal oxide particles

- 3.2 Capacitance at the Transitions between the Metal Oxide Particles - 3.3 Capacitance at the Transitions between Metal Oxide Particles and

contact

3.4 Transition resistance between the metal oxide particles

- 3.5 Contact resistance between the metal oxide particles and the

Contacting. C.Diehl, 2000; J. Gutierrez, 1991; J.R. MacDonald, 1987; Weimar, 1992.

As a consequence, it becomes clear that when gassing the complex circuit structure changes electrically both in its resistance values and in the electrical capacitance present in the sensor. The resulting complex impedance of the sensor is additionally dependent on the type of gas offered. It has therefore been considered to take advantage of this fact, in order to be able to conclude from changes in the sensor impedance to the gas type, or to be able to preferentially detect desired types of gas, see also P196 17 297.7; WO 97/41423; US 200 20 11851.

The technical problem to be solved:

The ohmic resistance of an oxidic semiconducting gas sensor changes in normal air even at low gas offers compared to no gas supply by about one order of magnitude, ie by one order of magnitude. Influences from humidity, absolute humidity, and the temperature of the active layer also have a considerable influence on the detection result in this steep part of the characteristic curve. However, a direct inference from the resistance of the active layer to the gas supply is subject to great uncertainties. With gas or steam offers in the order of about 2000-5000ppm, depending on the gas / steam, the resistance of the active layer is shifted in the range of very small slope. If the gas supply is increased even further, the active layer is saturated and allows no further evaluations of the electrical resistance of the active layer more. The time it takes for the active layer to return to its normal air level after fumigation with high gas or vapor concentrations is very high. To offer some vapors - such as gasoline vapor in a concentration> 1% vol. - The electrical resistance of the sensor remains for a very long time several days lower than before fumigation.

If the sensor is stored for a long time, sorbates are deposited on the surface of the active layer. After switching on the sensor, the active-layer resistance is generally significantly reduced and reaches its resistance to normal air only after a longer break-in period, for example after days. In applications in which the sensor could be stored under gas / steam and in which a start of the measurements under gas might be necessary, this behavior is a considerable disadvantage.

The measurement of the sensor capacitance is not possible according to the state of the art without effort. Are known circuits according to Figure 4; where is:

- 4.1 The heated sensor heater is not drawn

- 4.2 frequency generator preferably sine-wave generator - 4.3. Base point resistance or electrical inductance

- 4.4 Comparison and evaluation circuit Phase, vector, imaginary part

The sensor capacitances changing under the influence of gas cause a phase rotation and additionally act like a frequency-dependent AC voltage divider. The ohmic components which change under the influence of gas act as voltage dividers and act together with the capacitances of the sensor on the phase position.

The evaluation of the obtained measured values and the inference of the capacitance components in the equivalent circuit diagram of the active layer is only possible with considerable technical and mathematical effort and with costs. In practical applications - beyond the scientific discussion - the sensor capacity is not considered a gas concentration characteristic and practical. table usable size has been used. In practical applications and in numerous patents which deal with the use of oxide semiconductive gas sensors, the electrical ohmic resistance of the active layer is usually evaluated as a gas-dependent variable.

The capacitances measured with the aforementioned complex measurement methods on industrially manufactured oxide semiconductor sensors are dependent on the operating temperature and are given in the literature as follows see FIG. 3 - Capacity 3.2. Normal air 30-5OpF, saturation below gas100pF - capacity 3.3 Normal air 100OpF, saturation below gas 1.50OpF

Technical task:

The invention is based on the object of demonstrating with inexpensive methods a new possibility with which the metal oxide semiconductor gas sensor offered gas or vapor or aerosol concentration in the atmosphere surrounding the sensor, to be determined with great accuracy.

Technical solution of the problem and disclosure of the advantages: The solution of the problem consists in a method of the type mentioned that the sensor located in the gas and vapor-dependent changing electrical charge [(Q) in Coulomb)] as a measure of the concentration used in the surrounding atmosphere of the sensor oxidizable or reducible gases or vapors and evaluated by the evaluation circuit.

In a further embodiment of the invention, the sensor is charged by means of a voltage source to its voltage level, so that build up charges in the active layer of the sensor, and after completion of the charging process, the sensor is de-energized, after which the electrical charge now stored in the sensor by means of the evaluation is determined before it can take place within the sensor, a self-discharge on the internal ohmic portions of the same. Preferably, the voltage source is a DC voltage source. The fundamental advantage of the method according to the invention is that the ohmic resistance of the metal-oxide-semiconductor gas sensor is not used as a measure of the gas or vapor concentration, but instead the electrical charge Q stored in the active layer of the sensor is measured. or vapor-dependent quantity, so that in this way the offered gas or vapor concentration in the atmosphere can be determined with great accuracy. Gas or vapor concentration is also understood as meaning a concentration of an aerosol.

According to the prior art, when the ohmic resistance of the metal-oxide-semiconductor gas sensor is used for measurement, charges flow through the sensor through one line and out through another line; the measurement takes place during the flow of the charges. In the subject invention, however, the sensor is not flowed through by charges, but it flow only shift currents. The sensor acts alternately as a sink and as a source of charges without being flown through by them. Instead, the charges flow through a line into the device and out of the device through the same line; the measurement takes place after the charges have flowed.

In a further embodiment of the method according to the invention, the electrical charge stored in the sensor after charging by the DC voltage source is determined by discharging the sensor by means of a charge carrier transfer method, in which the electrical charge is transferred very quickly in small charge quantities into a reference capacitor, before self-discharge can take place within the sensor via the internal ohmic portions thereof. This means that by means of the DC voltage source, the sensor is charged to a certain charge, which depends on the voltage of the DC voltage source and the already existing state of charge of the active layer of the sensor, said charge state of the sensor from the reducible or oxidizable gas contained in the atmosphere Vapor concentrations depends. Subsequently, by means of the charge carrier transfer method, the additional Voltage source charged to the sensor again dissipated and transferred to the reference capacitor, which can be measured by the sensor emitted charge.

In a further embodiment of the method according to the invention, the sensor is charged via defined discharge pulses into the reference capacitor to determine the charge stored after charging by the DC voltage source in the reference capacitor, depending on the size of the reference capacitor and a selected threshold value dependent on the charge of the sensor Number of discharge pulses, the electrical charge stored in the sensor is determined, wherein the size of the electrical charge of the sensor depends on the atmosphere in which the sensor is operated.

In a further embodiment of the method according to the invention, the discharge pulses for discharging the sensor into the reference capacitor take time between 1 and 10 microseconds, preferably 1 microsecond. Advantageously, the time between the individual discharge pulses can correspond approximately to the duration of the individual discharge pulse, namely between 1 to 10 microseconds, preferably also 1 microsecond.

In a further embodiment of the method according to the invention, the electrical evaluation circuit has a controlled switching bridge, with the aid of which the sensor is reversed during the measurement process with respect to two successive measurement cycles, wherein the electrical charge in the sensor is measured by the charge carrier transfer method, wherein a measuring cycle composed of the charge and the discharge of the sensor.

In a further embodiment of the method according to the invention, the sensor is connected to the DC voltage source via a switch actuated by the evaluation circuit, so that charges build up in the active layer of the sensor and then the switch is opened again and de-energized, the switch for a short time Time, for example in the range of 0.5 to 5 milliseconds, preferably 1 millisecond, closed and is then opened again and immediately thereafter the discharge pulses for discharging the sensor are initiated.

In a further embodiment of the method according to the invention, a further fast switch is closed and opened again for generating the discharge pulses via the evaluation circuit for a very short time, so that a charge transfer takes place in the reference capacitor, this process being repeated at short intervals, so that the Charge voltage in the reference capacitor increases until it corresponds to a predetermined reference voltage, which detects an electrical comparator, such as comparator.

In a further inventive embodiment of the method, the number of required until reaching the switching threshold of the electrical comparator discharge pulses from the evaluation circuit is counted, from this number and the size of the reference capacitor, its charging voltage and from the predetermined value of the comparison voltage of the electrical comparator in the sensor total stored electrical charge is determined, from which the concentration of the gases or vapors in the atmosphere of the sensor is derived.

In a further embodiment of the method according to the invention, the closing time of the fast switch of the evaluation circuit for transmitting the charge from the sensor to the reference capacitor is between 1 microsecond to 10 microseconds, the opening time of the switch being approximately of the same order of magnitude between two closing times.

In a further embodiment of the method according to the invention for obtaining further information from the complex circuit system of the sensor for evaluating the electrical charge stored in the sensor, both the electrical charge of the sensor and the electrical resistance of the sensor are determined separately by measurement and these two measured values are evaluated linked together. In a further embodiment of the method according to the invention, after the end of the measurement process, the reference capacitor is discharged via a further controlled switch and thus prepared for the next measurement cycle. In addition, the sensor can be completely discharged after each measurement via a further, controlled switch due to a short circuit.

In an advantageous application of the invention, the gas concentration dependent on the electrical charge of the sensor is used as a sensor signal for the purpose of alarm, or the Steuem or Regge. i Furthermore, the object is achieved by a device for measuring in the

Atmosphere contained reducible or oxidizable gas or vapor concentrations, with an arranged in this atmosphere, electrically heatable metal-semiconductor sensor, sensor, with sensitive active layer, and connected to the sensor electrical evaluation circuit for generating a switching signal, achieved in that the electrical charge [(Qj in Coulomb)] located in the sensor and changing as a function of gas or vapor is used as a measure of the concentration of the oxidizable or reducible gases or vapors present in the surrounding atmosphere of the sensor and can be evaluated by means of the evaluation circuit.

According to the invention, the charge stored in the sensor can be determined by electrically charging the sensor first at a DC voltage source so that charges build up in the active layer of the sensor and the sensor is deenergized after completion of the charging process, and the charge is then discharged via defined discharge pulses a reference capacitor is discharged, carrier transfer method, before it can take place within the sensor, a self-discharge on the internal ohmic portions of the same, depending on the size of the reference capacitor, its charge voltage and a selected value of a comparison voltage to one of the charge of the Sensor dependent number of discharge pulses, the electrical charge stored in the sensor is determined, the size of the electrical Charge of the sensor depends on the atmosphere in which the sensor is operated.

In a further embodiment of the device according to the invention, the electrical evaluation circuit has a controlled switching bridge, with the aid of which the sensor is reversed during the measurement process after each measurement cycle, wherein the sensor in each case in alternating polarity by means of a program-controlled circuit of the electrical evaluation circuit electrically to the voltage potential the DC voltage source is charged, the electrical charge of the sensor via two switchable electrical switch polarity and pulse is transmitted in the reference capacitor until the charging voltage of the reference capacitor reaches an electrical comparator predetermined reference voltage, and that of the sizes of frequency of the recharging pulses, capacitance and reached charging voltage of the reference capacitor, the electrical charge of the sensor is determinable, from which the concentration of the gases located in the ambient atmosphere of the sensor or vapors is derivable.

In a further inventive embodiment of the device, the charging time of the sensor via the first switch is a short period of time in the range of a few milliseconds, preferably between 0.5 to 5 milliseconds, wherein the transmission time of the charge from the sensor to the reference capacitor by means of another, fast switch only by means of very short-term closing intervals, preferably for a few microseconds per closing interval, in particular between 1 to 10 microseconds, preferably 1 microsecond.

In a further embodiment of the device according to the invention, the bridge circuit consists of four branches and a middle path in which the sensor is arranged, wherein two branches are connected to a feed line in which the first switch for connecting the DC voltage source, and the two other branches with Ground are connected, wherein in each branch depending on a switch for the polarity change of the voltage at the sensor is located each measuring cycle, and both ends of the center path and thus both ends of the sensor via a fast switch together on the reference capacitor and placed together on the one input of a comparator, at the other input a comparison voltage is applied, the comparator then a switching signal outputs as soon as the charging voltage of the reference capacitor has reached the comparison voltage of the comparator.

Brief description of the drawing, in which:

Figure 1 shows the known structure of a TAGUCHI gas sensor Figure 2 shows the course of the resistance of the active layer of the gas sensor at normal air as a function of gas concentration

3 shows an electrical equivalent circuit diagram of the heated active layer of a gas sensor according to the prior art

Figure 4 is a known from the prior art, complex circuit for measuring the sensor capacitance

FIG. 5 shows the characteristics of an oxide semiconductor gas sensor when exposed to gas / steam

FIG. 6 shows a patented principle circuit for determining the charge or the capacitance of an oxide semiconductor gas sensor. FIG. 8 shows a representation using the electrical circuit of FIG. 4, as in the short pauses between the individual discharging processes, the charges are equalized via the resistors in a plurality of current paths

Figure 9 is a representation of the course of the electric charge in the invention, which varies in proportion to the gas supply, even in such

Offered if the electrical resistance of the sensor is already in the asymptotic, saturated part of the characteristic curve

11 shows an electrical schematic circuit diagram for achieving the object of the invention, namely an electrical switch, for detecting the charge or energy stored in a semiconducting metal oxide gas sensor as a measure of the invention Concentration of offered gases 12 shows the dependence of the charge in the active layer of a semiconductor gas sensor on an oxidizable calibration gas, for example carbon monoxide (CO) for calibrating the measuring method according to the invention, and FIG. 13 shows the dependence of the charge in the active layer of a semiconductor gas sensor on a reducible calibration gas; For example, nitrogen oxide also for calibration of the measuring method according to the invention

Ways to carry out the invention:

Before describing the circuitry approaches, the physical principle underlying the invention will be described below, which is a prerequisite for the understanding of the invention.

The mode of action of the gas-dependent dielectric and thus gas-dependent, the different capacity of oxide semiconductor sensors based on a change in the permittivity of insulators by the adsorption of molecules from the gas or liquid phase. The relationship between adsorbed gas quantity and permittivity is generally not linear in the case of dielectric sensors, so that the gas concentration can be concluded via a calibration curve. With higher occupation density, saturation effects may occur, depending on the sensor material. In polarization processes caused by adsorption processes, a distinction is made between volume effects and interfacial effects at the electrodes. The change in volume permittivity is caused by the change in polarizability and conductivity of the sensitive material.

The polarization and ion conduction mechanisms are not in phase because of their finite relaxation times η with the electric field. The dielectric function E * w is therefore frequency-dependent and complex-valued. At the interfaces of the sensor material to the electrodes polarization effects occur, which depend on the specific properties of the interface.

For the investigation of the sensor properties as a function of the gas concentration, it was proposed to evaluate the sensor response for a given Measure frequency. The sensor response defines impedance or admittance and the quantities derived therefrom. For practical use, the capacitive component of the sensor response, which is measured by an evaluation circuit, would be important. The optimum frequency for such measurements can be found in the spectroscopic investigations. It is normally between 10 kHz and 1 MHz for oxide semiconducting gas sensors.

Since the dielectric function can not be measured directly, its conversion into another representation, e.g. as impedance, necessary. In dielectric sensors with insulating sensitive materials, essentially the polarizability of the sensor layer is changed by the adsorption of gases. A change in the ionic conductivity by the dissociation of adsorbed molecules or their reaction products is also possible. Both processes influence the permittivity of the material.

Not always does the adsorption of gases lead to a measurable change in the dielectric material properties. In molecules with no permanent dipole moment, such as N2, _O2 or CO2, which are physisorbed in larger amounts, there is no change in the electrical properties is observed of oxide semiconductors. For amorphous materials, anisotropy of the dielectric function is generally not expected. Likewise, it can be assumed for the applied measuring voltages that non-linear field effects are negligible. A scalar description of the polarizability α of the individual molecules is therefore sufficient.

The polarizability α of a molecule is due to a number of different processes whose summation gives the total polarizability α _ges . These are the orientation polarization α _Δ , the electronic polarization α _e ι (displacement of the outer electrical _{shell relative} to the atomic nucleus) and the ionic polarization α j _0n (displacements of the atoms of a molecule _{relative to} one another). The two latter polarization mechanisms show energetic reasons (resonance vibrations) only from the microwave range on a frequency dependence. in the Low-frequency limiting case up to 1 MHz, the contributions of the electronic and ionic polarizability are constant, namely: α ∞ = α _e ι + a _{\ on} .

The orientation polarization α _{Δ of} the polar groups becomes, according to Debye, assuming a free movement without mutual interaction

P ² εo 3 k T calculated from its electric dipole moment p.

Not only molecular dipoles show orientation polarization, but also ions that can jump locally between several neighboring defect sites. The contribution of this hopping mechanism for orientation polarization is described mathematically analogous to that of molecular dipoles and is therefore included in α _Δ . These localized charges do not contribute to electrical conductivity because their mobility is too low.

The static polarizability α _{ges of} a molecule at low frequency is the sum of the Teilpolarisierbarkeiten, namely: α _ges = α _Δ + α °° = α _Δ + α _e ι + αi _0π .

In many cases, the electronic and ionic fraction α °° Gesamtpolarisierbarkeit ability α _{ges is} considerably lower than that of the orientation polarization α _Δ . The electrical susceptibility x is calculated neglecting the mutual interaction of the different polarization mechanisms from a simple summation of the polarizabilities a _\ of the individual molecules multiplied by their number Ni. This gives the static polarization of a solid and, with the definition of the dielectric displacement, the relative dielectric constant. Since the ionic and the electronic polarizability have only a small proportion of the dielectric constant, their change will also be slight in the case of pure physisorption of gases with a low dipole moment.

According to the ideas of adsorption the adsorbed molecules form on the inner surface of the adsorbent a mono- or multimolecular layer, which has the character of a liquid. To estimate the static Dielectric constant, therefore, the adsorbate adsorbent system can be treated as a heterogeneous dielectric with separate phases.

According to the invention, approaches for calculating the dielectric constant of heterogeneous dielectrics were sought, which add up the dielectric constant of the individual phases according to their volume fractions. The calculated capacities are subject to large loss factors due to the parallel electrical resistances.

Furthermore, these are not conventional surface capacitors, but charge carriers which occur at the contact junctions between the metal oxide grains of the active layer as well as between the grains of the active layer and the contacting structures.

If a DC voltage is applied to the sensor, current flows through the resistors 8.5 R3, 8.3 R1 and 8.4 R2 shown in FIG. The electrical charges at the contact junctions do not affect the current flow.

If an alternating voltage is applied to the sensor, current flows through the resistors, as above, and through the capacitors, which form a capacitance according to the following expression: 1

Xc = 2πf * C

The capacitive resistance goes to infinity at DC, and to zero at infinity. If no parallel resistors exist, two capacitors are connected in series. The capacity of both capacitors is then calculated as follows:

C ₁ + C ₂

Cges ⁼

Ci * C ₂

The resulting capacitor C _tot must be smaller than the smallest capacitor in the series in any case. The phase shift of an applied sinusoidal voltage is calculated on each R / C element as a function of, inter alia, the frequency. If a very large frequency is applied, for example 1,000 KHz, the capacitance of the relatively large capacitor C1,11 of about 100 oPF is relatively small and the phase shift is also small. The capacitor C2 8.2 is relatively small with 30-10OpF and the capacitive resistance is relatively high and in relation to the resistor R2 connected in parallel so that a well evaluable phase shift can be measured.

In order to determine the capacitor C1, one chooses a relatively low frequency, e.g. 5KHz, so that the capacitive resistance of C1 in relation to the ohmic resistance R1 is about 1: 1, to cause a maximum phase shift of the alternating current. The current path then passes through the resistors R1, R2, R3 and over the capacitive resistance of C1.

When fumigated, the capacitors mentioned change in the sense that the capacitance increases, whereas the resistances change in the sense that they become smaller under gas.

Because both processes are inextricably linked, the direct evaluation of the located in a sensor electrical capacitances or electrical charges is very complicated and difficult, because usually only the way of determining the frequency-dependent impedance is possible. If only the electrical charge were determined, which is stored in the abovementioned electrical capacitors of an oxidic semiconductor gas sensor after application of a DC voltage, the electrical charge would be a measure of the concentration of oxidizable or reducible gases conducted to the sensor.

With regard to the solution according to the invention, a circuit whose principle is shown in FIG. 6 can be used to detect and measure the electrical charge present in the active layer of an oxide semiconductor gas sensor. The sensor 6.6 is connected via the switch S1, 6.1, which is actuated by a central evaluation circuit, to a voltage source, preferably a DC voltage source, wherein charges build up in the active layer of the sensor. Thereafter, the switch is opened again and de-energized.

Via the central evaluation circuit, a second, fast switch S2, 6.2 is now closed and opened again for a very short time - in the range of a few microseconds. There is a charge transfer into the reference capacitor C, 6.4. This process is repeated at short intervals, with the charging voltage in the reference capacitor increasing until it equals a given reference voltage, which an electrical comparator, for example a comparator, Compi, 6.5 detects.

The number of required until reaching the switching threshold, the comparison voltage, recharging pulses is counted by the evaluation circuit. From the magnitudes of the frequency of the charging pulses, the capacitance of the reference capacitor and the charging voltage of the same, the electrical charge of the sensor is determined, from which values the concentration of the gases or vapors in the ambient atmosphere of the sensor is derived.

If the electrical charge stored in the sensor is small, many reloading operations are required to reach the switching threshold. If the electrical charge stored in the sensor is large, few transhipment operations are required to reach the switching threshold.

The discharge of the electrical charge stored in the sensor takes place at very short intervals - in the microsecond range. A complete discharge of the charge stored in the sensor is not possible via the resistors connected in parallel because they are in each case Schottky junctions, which have a certain threshold voltage, and thus do not behave like ohmic resistors. In the short pauses between the individual discharge processes, the charges are equalized via the resistors in several current paths, as shown in Figure 8, 8.6, 8.7 and 8.8. The charge stored in the sensor can thus be determined according to the invention by first electrically charging the sensor to a DC voltage source and then discharging it via defined discharge pulses into a reference capacitor, depending on the size of the reference capacitor and the selected threshold value depending on the charge of the sensor Number of electrical charges stored in the sensor. The electrical charge of the sensor depends on the atmosphere in which it is operated.

In normal air, which is free of oxidizable gases or vapors, the charge of the oxide semiconductor sensor corresponds to an electrical capacitor in the order of about 20-10OpF, depending on the design of the sensor.

When the sensor is oxidized with oxidizable gas, this displaces the oxygen stored in the sensor, with the result that the electric field between the individual Schottky barriers decreases, which results in an increase in the electrical charge. For very large gas or steam offers, the electrical charge in the sensor increases, so that it corresponds to a capacitor in the order of up to 1000OpF.

When the sensor is gassed with reducible gas, such as ozone or nitrogen oxides, the reverse effect occurs because the charge decreases with the presence of these gases in relation to the charge at normal air.

It is advantageous that the electrical charge of the sensor is approximately proportional to the gas supply, Figure 9, and the characteristic in contrast to

Resistance characteristic of Figure 9.2 has an advantageously much smaller curvature. It is also advantageous that the electric charge 9.1 varies proportionally to the gas supply even with gas offers, if the electric

Resistance of the sensor is already in the asymptotic, saturated part of the characteristic and no more evaluation allows. Therefore, it is highly advantageous to use the energy or charge stored in a semiconducting metal oxide sensor as a measure of the concentration of gases offered.

In the simplest way, a basic electrical circuit according to FIG. 11 is suitable for this purpose. The capacitance of the sensor is connected to a voltage via a switch. When the switch is opened, the charge of the capacitor discharges according to a time function, the time being a measure of the capacity or charge. A disadvantage of this simple solution that the discharge does not take place via constant parallel resistances to the internal capacitance, but that the resistive components of the sensor resistance in their value are dependent on the offered gas concentration.

In order to eliminate these disadvantages, it is proposed according to the invention to control the gas sensor over a controlled switch and a defined time, e.g. less than 1 msec, to apply to a DC voltage and then to discharge the sensor via a controlled, fast switch in a plurality of very short discharge pulses, preferably in the microsecond range, into a reference capacitor until the voltage at the reference capacitor defines a defined Threshold reached. From the number of required discharge pulses and the size of the reference capacitor and from the predetermined threshold value, the electrical capacitance, ie the total electrical charge stored in the sensor can be determined.

After the end of the measurement process, the reference capacitor can be discharged via another controlled switch and thus prepared for the next measurement cycle. It makes sense to completely discharge the gas sensor via a controlled switch after the measurement due to a short circuit.

It is thus highly advantageous that the sensor is neither flowed through by direct current nor alternating current during the actual measurement process, but that there is a static and only by the gas supply-dependent electrical charge (in Coulomb), which in the carrier transfer process very is quickly transferred in small amounts of charge in the reference capacitor before self-discharge can take place on the inner ohmic portions of the sensor.

5 Oxide gas sensors are not independent in their characteristics of the polarity of the connected measuring voltage, which subsequently leads to a further, preferred solution of the invention.

In practice, the "switches" described below are implemented by transistor switches or similar semiconductor switches which are precisely controlled by a program controlled microcontroller in their timing.

In the preferred solution of the invention, a circuit according to FIG. 10 is used. The sensor 10.6 to be measured is located in a switching bridge which is formed from the four switches 10.2, 10.3, 10.4 and 10.5 which are actuated electronically by a central evaluation circuit. The switches of the switching bridge are set after each measuring cycle in such a way that the sensor to be measured is always reversed so that advantageously no electrolysis processes can take place in the sensor. 0

The further switch 10.1 charges a certain time, e.g. <1msec, the capacitance of the sensor from a positive voltage source U to the electrical level of this voltage. Depending on the position of the switching bridge in each case the positive pole of the charged sensor via two electrical lines to two switches 10.7 and 5 10.7 'out. Leading to the positive pole of the sensor switch 10.7 or 10.7 'transmits by multiple and very short-term closing of the switches 10.7 and 10.7' successively stored in the sensor electrical charge in a reference capacitor 10.8, while the negative pole - practically as third away from the sensor to be measured leading line - flows through the switching bridge as o "ground" in the circuit.

Only the positive pole of the sensor respectively facing switch 10.7 or 10.7 'is immediately after completion of the charging of the sensor via a series very short discharge pulses charge the reference capacitor 10.8 until the downstream electrical comparator 10.10 detects the achievement of a comparison voltage 10.9 and generates an output signal. The number of discharge pulses until reaching this threshold is a measure of the electrical charge stored in the sensor, which alone can serve to determine the charge of the sensor. It is thus in a measurement cycle of a switch 10.7 and in the following measurement cycle of the other switch 10.7 ¹ briefly pressed several times. After each measurement cycle, the reference capacitor 10.8 is completely discharged by briefly closing another switch 10.11.

A natural discharge of the sensor capacitance of the sensor is carried out by the capacitors connected in parallel sensor resistance, Figure 8. In addition, the electrical charge in the sensor is electrically discharged after each measurement cycle by short-circuiting the sensor short, which is not shown in Figure 10.

In addition to the evaluation of the electrical charge stored in the sensor, it is proposed to determine both the electrical charge of the sensor and the electrical resistance of the sensor separately from each other in a suitable manner, and to link these two measured values in the evaluation for further information to gain the complex circuit system of the sensor.

Further advantages of the solution according to the invention: FIG. 5 shows in contrast the characteristic curves of an oxide semiconductor gas sensor when subjected to gas / steam. The characteristic 5.2 is that of the ohmic resistance of the active layer, the characteristic 5.1 is that of the course of the electric charge, measured in Coulomb, the active layer. It can be seen that the characteristic 5.2 in the beginning is very steep and then expires relatively quickly in an asymptote, namely in the saturation. The characteristic curve 5.1 is significantly less steep in the beginning and changes with the supply of gas up to the range of very high gas concentrations, in which the resistance of the active layer has long been saturated. The characteristic curve 5.1. is not ideal linear, but allows a computational linearization with much less effort. Because slight errors or environmental influences lead to large measured value deviations in the steep region of the characteristic curve 5.2, this is when the characteristic curve 5.1 is evaluated. however, not the case.

These properties are extremely advantageous for the practical application of semiconductor oxide sensors. While the resistance of the active layer regresses in its chemical structure after massive fumigation under reaction with the atmospheric oxygen, the gas molecules influencing the electrical charge diffuse out of the active layer very quickly. The sensor becomes "faster" when the charge in the sensor is measured.

This is a very great advantage in many applications beyond the range of concentrations that can be evaluated in the range of very high concentrations, especially in combination with the faster recovery time after massive fumigation. For example in the following commercial applications:

- a) Measurement of the concentration of fuel gases, such as methane, propane, butane, in the air: At concentrations above about 5000ppm addition, the gas-dependent sensor resistance of oxide semiconductor gas sensors is pushed into the region of saturation and the characteristic is asymptotic. Changes in the gas concentrations result in no significant, evaluable changes in the resistance of the active layer. The gas sensor can therefore only be used in the range of small concentrations of 1-100 ppm up to medium concentrations <10,000 ppm for measurement purposes. With higher supply of gas, the sensor becomes so low that sometimes it takes days to return to the range of normal values. The evaluation of the electrical charge of the sensor allows an extension of the measuring range up to approx. 50% vol. Gas content. The reset times of the measured sensor

Charge after fumigation is very short with a few minutes. - b) measurements in exhaust gases:

The concentrations of hydrocarbons and / or nitrogen oxides in the exhaust gas can reach very high levels. Measured with oxide semiconductor gas sensors, the sensors typically operate in the saturated, almost asymptotic region of the characteristic curve. Advantageously, semiconductor sensors in the exhaust gas give very good and stable results, if not the resistance of the active layer but the electrical charge of the active layer is evaluated.

- c) Control of ventilation systems:

The gas or vapor concentration can be very different outside or inside the object to be ventilated and can assume very high concentrations. It is important that the actual gas concentration is detected very quickly and safely even at high concentrations and that in normal air after the end of fumigation, the sensor reaches its normal operating point very quickly, which is advantageously faster than when evaluating the electrical resistance of the active layer. The inventive evaluation of the electrical charge of the sensor Q allows on the one hand the measurement of higher gas concentration, for example: industrial objects, hazardous and fuel storage, underground parking, tunnel, and ensures a very fast recovery after fumigation. The influence of air humidity on the measured charge quantities in the sensor is significantly lower than the influence of humidity on the electrical resistance of the active layer.

- d) Measurement of the gases that are produced in food production, baking, frying, cooking, roasting:

The composition and concentration of in o.g. Processes outgoing gases are dependent on the progress of the manufacturing process. It is desired to automate these processes, which requires a quick and cost-effective analysis of the outgoing gases / vapors. The concentrations can become very high. The evaluation of the electrical

Resistance of the active layer is largely impossible because the operating point of the sensor is in the asymptotic, saturated region. Advantageously, the teaching of the invention is used to make this application for oxide semiconductor sensors possible.

- e) Measurement of the oxygen content in exhaust gases, for example lambda probe: A lambda probe is a semiconducting metal oxide sensor, often consisting of zirconium oxide. The same gas-physical and chemical laws apply as for all other oxide semiconductor gas sensors. The measurement of the electric charge as a gas-dependent variable achieves a more linear characteristic than in the determination of the electrical conductance and is therefore advantageous.

Several technical possibilities are conceivable for measuring the electric charge stored in an oxide semiconductor gas sensor, measured in Q = Coulomb. Alternatively, the gas-dependent electrical capacitance of the sensor can be determined by the sensor is flowed through by an alternating current, which leads to evaluable effects when changing the sensor capacity.

The preferred solution for measuring the gas-dependent charge of the sensor is the so-called charge transfer method, as described above in its technical

Details are described. It is possible to represent the charge carrier method by an integrated circuit. The measurement of the electric charges present in oxidic sensors can be done with the patents / patent applications /

Utility models described in technical methods DE 2 99 2441U1; EP 1 153 404 A1; EP 1 131 641 A1, without these being the inventive

Impair thoughts.

Thus, the inventive method for determining the reducible or oxidizable gas concentration in a metal oxide semiconductor gas sensor, such as a Zinndioxid gas sensor, surrounding atmosphere, consisting of an electrically heated metal oxide semiconductor gas sensor and an electrical evaluation circuit, is that in a metal oxide semiconductor gas sensor located and changing depending on the gas electrical charge, measured in Coulomb, is measured by means of a charge carrier transfer method and used as a measure of the concentration of oxidizable or reducible gases or vapors present in the ambient atmosphere of the sensor. With the aid of a controlled switching bridge, the sensor is reversed in each case, the electrical charge in the sensor being measured by the charge carrier transfer method. The gas concentration-dependent electrical charge of a metal-oxide-semiconductor sensor can be used as a sensor signal for the purpose of alarming or control or regulation.

For calibration of the measuring method according to the invention and the device can proceed as follows. The electrical landing amount Q in the active layer of a semiconductor gas sensor is a measure of the sum of the concentrations of oxidizable or reducible gases. In order to conclude from the charge amount to the gas concentration, the characteristic curve, namely the dependence of the charge as a function of the gas supply is determined in a first step. As the calibration gas, for example, carbon monoxide (CO) or methane is used, as shown in FIG. 12 as an example. All other organic gases are also suitable in principle. Of course, the determined characteristic applies only to the associated gas, other gases show a different characteristic curve, which relates to the steepness of the curve and its curvature. The cause is the specific reactivity of the active layer to various gases or vapors or aerosols.

The thus determined characteristic curve of the charge quantity Q is stored as a function of the gas supply, for example in ppm, as a formula or table with several interpolation points, as a polynomial, in the electronic evaluation, which can be a microcontroller. Now as a result of a gas offer, for example

Methane, which changes the amount of charge Q stored in the active layer, can be directly deduced from the charge quantity Q on the gas supply, for example in ppm, via the aforementioned table.

If a reducible gas, such as ozone or nitrogen oxide, is offered, the sensor behaves as shown in FIG. The amount of charge Q increases with increasing the gas supply. The process is thus inverse to the process of providing oxidizable gases. However, the charge Q follows the gas supply, for example in ppm, very accurately and very reproducibly. Here, too, a calibration curve can be used to deduce the amount of gas offered from the measured charge quantity Q.

Since it is known to produce somewhat selective semiconductor gas sensors using special preparation methods and special active-substance formulations, which do not react broadly to all gases offered, it is possible with the method and apparatus according to the invention to selectively measure individual gases or gas groups, such as, for example Methane, propane, ozone, nitrogen oxide or nitrogen dioxide. The electrical charge present in the metal oxide semiconductor gas sensor is thus used and evaluated as a measure of the concentration of the oxidizable or reducible gases or vapors present in the ambient atmosphere of the sensor. For this purpose, the sensor is advantageously charged to the voltage level of a DC voltage and switched off after the end of the charging process, after which the now stored in the sensor electrical charge is determined in a suitable manner by measurement.

Furthermore, with the aid of a controlled switching bridge, the sensor can be charged electrically in alternating polarity, whereby the sensor is charged electrically to a specific voltage potential by means of a program-controlled circuit, and the electrical charge of the sensor is polarity-correct via a fixed and two switchable electrical conductors and in pulse manner is transferred to a reference capacitor until a voltage detected by an electrical comparator on the reference capacitor is reached, and that from the magnitudes frequency of the Umladeimpulse, capacitance of the reference capacitor and charging voltage of the reference capacitor, the electrical charge of the sensor is determined, resulting in the concentration of in the ambient atmosphere of the sensor located gases or vapors is derived. The gas concentration-dependent electrical charge of a metal-oxide-semiconductor sensor is advantageously used as a sensor signal for the purpose of alarming, or controlling or regein. In accordance with the method, the electrical charge of a metal oxide semiconductor sensor determined by the gas concentration is used in gas engineering applications, wherein the gas concentration dependent electrical resistance of the active layer of the metal oxide semiconductor sensor is also determined, and both the electrical charge and the electrical resistance be appropriately linked in the evaluation and taken into account for the evaluation of the gas or vapor-containing atmosphere surrounding the sensor.

It is thus advantageously the electrical capacitance of the sensor determined by the gas-dependent impedance of the sensor by the sensor is traversed by an alternating current of suitable frequency, and by the voltage across a series-connected electric capacitor or resistance or electrical inductance as a measure of the Gas concentration is used.

In a further advantageous embodiment, the capacitance of the sensor is determined by the gas-dependent impedance of the sensor by the sensor is traversed by an alternating current of suitable frequency, and by the resulting phase shift between the injected AC voltage and measured behind the sensor AC voltage is used as a measure of the gas concentration ,

In the apparatus for determining the proportions of oxidizable or reducible gases or vapors contained in the atmosphere surrounding a metal oxide semiconductor gas sensor, for example tin dioxide gas sensor, with an electrically heated metal oxide semiconductor gas sensor and a suitable electrical evaluation circuit, the in The electrical charge located in the sensor is used and evaluated as a measure of the concentration of the oxidizable or reducible gases or vapors present in the ambient atmosphere of the sensor. For this purpose, the sensor can be charged to the voltage level of a DC voltage and switched off after the end of the charging process, after which the now stored in the sensor electrical charge is determined in a suitable manner by measurement. With the aid of a controlled switching bridge, the sensor is electrically charged in alternating polarity, with the aid of a program-controlled circuit, the sensor is electrically charged to a certain voltage potential, and the electrical charge of the sensor via a fixed and two switchable electrical conductor with the correct polarity and in pulses in one Reference capacitor is transferred until a detected by an electrical comparator voltage is reached at the reference capacitor, and wherein the sizes of the charge, reference capacitor and charging voltage of the reference capacitor, the electrical charge of the sensor is determined, resulting in the concentration of the ambient in the Sensor located gases or vapors is derived. In this case, the charging time of the metal-oxide-semiconductor sensor via the switch is preferably a short period of time, preferably <1 msec, wherein the transfer of the charge from the metal-oxide semiconductor gas sensor into the reference capacitor by means of at least one further switch takes place only by means of very short-term closing intervals, preferably during less microseconds per closing interval.

Advantageously, the gas concentration dependent electrical charge of a metal oxide semiconductor sensor is now used as a sensor signal for the purpose of alarm, or the control or Regge.

Industrial Applicability:

The measuring method according to the invention and the electrical circuits are particularly suitable for use in oxidic semiconductor gas sensors for the purpose of measuring reducible or oxidizable gas concentrations contained in air. A common feature of all measuring methods is that the electrical charge, measured in Coulomb, of the sensor is determined in a suitable manner and used as a measure of the gas or vapor concentration surrounding the sensor, and thus the electrical charge of the sensor as a measured variable in measuring , Control and regulating circuits is used.

Claims

claims:

1. A method for the determination of reducible or oxidizable gas or vapor concentrations contained in the atmosphere, using an arranged in this atmosphere, electrically heatable metal-semiconductor sensor, sensor, with sensitive active layer, and with an electrical evaluation circuit connected to the sensor Generation of a switching signal, characterized in that the sensor located in the gas or vapor-dependent changing electrical charge [(QJ in Coulomb)] used as a measure of the concentration of existing in the surrounding atmosphere of the sensor oxidizable or reducible gases or vapors and means the evaluation circuit is evaluated.

2. The method according to claim 1, characterized in that the sensor is charged by means of a voltage source to its voltage level, so that build up charges in the active layer of the sensor, and after completion of the charging process, the sensor is de-energized, after which the now stored in the sensor electrical charge is determined metrologically by means of the evaluation circuit before it can take place within the sensor self-discharge via the inner ohmic portions of the same.

3. The method according to claim 1 or 2, characterized in that the determination of the stored after charging by the voltage source in the sensor electrical charge by means of a charge transfer method by a discharge of the sensor by the electric charge very quickly in small Charge quantities is transferred to a reference capacitor before it can take place within the sensor self-discharge via the internal ohmic portions of the same.

4. The method according to any one of the preceding claims, characterized in that the voltage source is a DC voltage source. 5. The method according to any one of the preceding claims, characterized in that is discharged to determine the charge stored by the DC voltage source in the sensor, the sensor via defined discharge pulses in the reference capacitor, wherein depending on the size of the

5 Reference capacitor and a selected threshold after a dependent of the charge of the sensor number of discharge pulses, the electrical charge stored in the sensor is determined and the size of the electrical charge of the sensor depends on the atmosphere in which the sensor is operated. 0

6. The method according to claim 5, characterized in that the discharge pulses for discharging the sensor into the reference capacitor take time between 1 microsecond to 10 microseconds.

5. The method according to any one of claims 5 or 6, characterized in that the time between the individual discharge pulses corresponds approximately to the duration of the individual discharge pulse, namely between 1 microsecond to 10 microseconds.

8. The method according to any one of the preceding claims, characterized in that the electrical evaluation circuit has a controlled switching bridge, with the aid of which the sensor is reversed during the measuring operation with respect to two consecutive measuring cycles, wherein the electrical charge in the sensor by the charge carrier transfer Method, wherein a measurement cycle is composed of the charge and the discharge of the sensor.

9. The method according to claim 1 or 2, characterized in that the sensor (6.6) is connected via an actuated by the evaluation switch (S1, 6.1) to the DC voltage source, so that build up charges in the active layer of the sensor and then the switch is opened again and de-energized, the switch for a short time, for example in the range of 0.5 to 5 milliseconds, preferably 1 Millisecond, closed and then reopened and immediately thereafter the discharge pulses for discharging the sensor are initiated.

10. The method according to any one of the preceding claims, characterized in that for generating the discharge pulses via the evaluation circuit for a very short time, a further switch (S2,6.2) is closed and opened again, so that a charge transfer into the reference capacitor (C1, 6.4 ), wherein this process is repeated at short intervals, so that the charging voltage in the reference capacitor increases until it corresponds to a predetermined reference voltage (10.8), as determined by an electrical comparator (Compi, 6.5), such as comparator.

11. The method according to claim 10, characterized in that the number of required until reaching the switching threshold discharge pulses from the evaluation circuit is counted, wherein from this number and the size of the reference capacitor and from the predetermined value of the comparison voltage of the electrical comparator in the Sensor total stored electrical charge is determined, from which the concentration of the located in the atmosphere of the sensor gases or vapors is derived.

12. The method according to any one of claims 10 to 11, characterized in that the closing time of the switch (S2, 6.2) of the evaluation circuit for transmitting the charge from the sensor to the reference capacitor (C1.6.4) is between 1 microsecond to 10 microseconds, wherein the Opening time of the switch between two closing times is approximately of the same order of magnitude.

13. The method according to claim 1, characterized in that for obtaining further information from the complex circuit system of the sensor for evaluating the electrical charge stored in the sensor both the electrical charge of the sensor and the electrical resistance of the sensor separately determined by measurement and these two Measured values are linked together during the evaluation.

14. The method according to any one of the preceding claims, characterized in that discharged after the end of the measuring operation, the reference capacitor via a further controlled switch and thus prepared for the next measurement cycle.

15. The method according to claim 14, characterized in that the sensor is completely discharged after each measurement via a further, controlled switch by short circuit.

16. The method according to any one of the preceding claims, characterized in that the dependent of the gas concentration of the electrical charge of the sensor is used as a sensor signal for the purpose of alarm, or the control or Regge.

17. An apparatus for measuring reducible or oxidizable gas or vapor concentrations contained in the atmosphere, with an arranged in this atmosphere, electrically heatable metal-semiconductor sensor, sensor, with sensitive active layer, and with an electrical evaluation circuit connected to the sensor for generating a switching signal, characterized in that the sensor located in the gas or vapor-dependent changing electrical charge [(Q) in Coulomb)] used as a measure of the concentration of existing in the surrounding atmosphere of the sensor oxidizable or reducible gases or vapors and means the evaluation circuit can be evaluated.

18. The device according to claim 17, characterized in that the charge stored in the sensor can be determined by the sensor is first electrically charged to a DC voltage source, so that build up charges in the active layer of the sensor and switched off after completion of the charging process, the sensor and the charge is then discharged via defined discharge pulses into a reference capacitor, carrier transfer method, before a self-inductance is applied within the sensor. Discharge over the internal ohmic portions of the same can take place, wherein depending on the size of the reference capacitor, the charging voltage and a selected threshold value of the comparison voltage after a dependent of the charge of the sensor number of discharge pulses, the electrical charge stored in the sensor is determined, the size of the electrical charge of the sensor depends on the atmosphere in which the sensor is operated.

19. The device according to claim 18, characterized in that the electrical evaluation circuit has a controlled switching bridge, with the aid of which the sensor is reversed during the measurement process after each measurement cycle, wherein the sensor in each case in alternating polarity by means of a program-controlled circuit of the electrical evaluation circuit is electrically charged to the voltage potential of the DC voltage source, and that the electrical charge of the sensor is transmitted via two switchable electrical switch polarity correct and in pulses in the reference capacitor until the charging voltage of the reference capacitor reaches the comparison voltage predetermined to an electrical comparator, and that from the sizes of the frequency Umladeimpulse, capacity and reached charging voltage of the reference capacitor, the electrical charge of the sensor can be determined, from which the concentration of the gases in the ambient atmosphere of the sensor o the vapors is derivable.

20. Device according to one of claims 17 to 19, characterized in that the charging time of the sensor via the first switch (10.1) is a short period of time, preferably between 0.5 to 5 milliseconds, and the transmission time of the charge from the sensor in the Reference capacitor by means of the further switch (10.7, 10.7 ') takes place only by means of very short-term closing intervals, preferably for a few microseconds per closing interval.

21. Device according to one of claims 17 to 20, characterized in that the bridge circuit consists of four branches and a middle path in which the sensor (10.6) is arranged, wherein two branches with a supply line, in which the first switch (10.1) for connecting the DC voltage source (U) is connected, and the other two branches are connected to ground, wherein in each branch depending on a switch (10.2,10.3,10.4,10.5) for the polarity change The voltage at the sensor is located after each measurement cycle, and both ends of the middle path and thus both ends of the sensor via a fast switch (10.7.10.7 ') together on the reference capacitor (10.8) and put together on one input of a comparator (10.10) are at the other input a comparison voltage (10.9) is set, the comparator (10.10) then emits a switching signal as soon as the charging voltage of the reference capacitor (10.8) has reached the comparison voltage (10.9) of the comparator (10.10).