DE102009029690A1 - Method for determining component concentration of gas in measuring gas chamber of motor vehicle, involves completing evaluation of parameter of gas from measured current and measured voltage change that occurs due to current change - Google Patents

Abstract

The method involves applying voltage (Up) to a pump cell (118) comprising two electrodes (114, 116) and a solid electrolyte (112) in a measuring phase, where one of the electrodes (114) is arranged in a hollow chamber (120). Current (Ip) i.e. limiting current is measured by the pump cell in the measuring phase. The current is changed around a predetermined value in another measuring phase. Voltage change occurring due to the current change is measured. Evaluation of a parameter of a gas is completed from the current and voltage change. An independent claim is also included for a sensor arrangement for determining a parameter of an exhaust gas in a measuring gas chamber.

Description

State of the art

The The invention is based on known sensor elements, which are based on electrolytic Properties of certain solids based, so the ability this solid to conduct certain ions. such Sensor elements are used in particular in motor vehicles, to measure air-fuel gas mixture compositions in which Case these sensor elements also known as "lambda probe" are and have an essential role in reducing pollutants in exhaust gases, both in gasoline engines and in diesel technology, play.

With the so-called air ratio "lambda" (λ), the relationship between an air mass actually offered and a theoretically required (ie stoichiometric) air mass for combustion is generally referred to in combustion technology. The air ratio is measured by means of one or more sensor elements usually at one or more locations in the exhaust system of an internal combustion engine. Accordingly, "rich" gas mixtures (ie gas mixtures with a fuel surplus) have an air ratio λ <1, whereas "lean" gas mixtures (ie gas mixtures with a fuel deficiency) have an air ratio λ> 1. In addition to automotive technology, such and similar sensor elements are also used in other areas of technology (in particular combustion technology), for example in aviation technology or in the control of burners, for. B. in heating systems or power plants. Embodiments and functional principles of such sensor elements are, for example, in Robert Bosch GmbH: "Sensors in the motor vehicle", 1st edition 2001, pp. 112-117 described.

Such sensor elements are now known in numerous different embodiments. One embodiment variant is the so-called "jump probe" whose measuring principle is based on the measurement of an electrochemical potential difference between a reference electrode exposed to a reference gas and a measuring electrode exposed to the gas mixture to be measured. Reference electrode and measuring electrode are connected to one another via the solid electrolyte, wherein doped zirconium dioxide (eg yttrium-stabilized ZrO ₂ , YSZ) or similar ceramics are generally used as the solid electrolyte due to its oxygen-ion-conducting properties.

alternative or in addition to jump probes also come so-called Pump cells are used in which an electrical pumping voltage applied to two electrodes connected across the solid electrolyte is, with a pumping current is measured by the pumping cell. The Sensor elements are usually in the so-called limit current operation operated, that is, in an operation in which the pumping voltage is chosen such that a through a diffusion barrier incoming oxygen completely to the counter electrode is pumped. In this mode, the pumping current is approximately proportional to the partial pressure of the oxygen in the exhaust gas mixture, see above that such sensor elements often as proportional sensors be designated. In contrast to jump sensors can be such Proportionalsensoren as so-called broadband sensors about a comparatively wide range for the air ratio lambda deploy. In many broadband sensors, the sensor principles described above become also combined so that the sensor elements one or more after sensors operating on the jump sensor principle ("cells") and one or more proportional sensors. So let yourself For example, the principle described above, according to the pump cell principle working "single cell" by adding expand a jump cell (Nernst cell) into a "double cell".

characteristic Feature of the broadband lambda probes is that they at least theoretically a unique, monotone characteristic over the lambda value which is ideally linear. To this Way, these sensor elements allow a determination the oxygen content over a wide range. In the In practice, however, it has been shown that broadband lambda probes are around the λ = 1-point does not work very well around. Another disadvantage is in that multicellular broadband lambda probes use a variety of Need electrode leads. So need For example, broadband lambda probes with a pumping cell and a Reference electrode and a heating element usually five or four leads. Since electrode leads a considerable Represent cost factor in the production of the sensor elements, this represents a significant disadvantage.

Basically, therefore, from the point of view of the reduced electrodes - and thus the number of leads - unicellular sensor elements are preferable to multi-cell sensor elements. For example, unicellular sensor elements can be constructed in such a way that both electrodes of a pumping cell are each arranged in separate cavities, which are each separated from the exhaust gas via a separate diffusion barrier. A connection to a reference air duct is thus eliminated in such a sensor element. A sensor element constructed in this way allows a measurement of the oxygen content in both the lean and the rich exhaust gas, ie both at a residual oxygen content or at an oxygen deficit. However, the characteristic of such an oxygen sensor is not unique in itself, For example, when the sensor element is acted upon by a constant voltage or a constant current, it has a characteristic curve with a V-shaped profile, that is, ambiguity. This is for example in DE 10 2005 054 144 A1 , to which reference can be made with respect to this problem explained. This ambiguity or ambiguity means that, for example, from a simple current-voltage measurement can not be clearly concluded whether the gas mixture is in the rich or lean air range.

Disclosure of the invention

The The present invention therefore provides a sensor element and a method and a sensor arrangement which overcomes the disadvantages of known methods and avoid devices at least largely. In particular the sensor element constructed comparatively simple and preferably has only a single cell structure on.

The The present invention is based on the recognition that a sensor element structurally designed with two diffusion barriers and two electrodes can be and electrically controlled so that overall generates a unique, monotone in the lambda value overall characteristic can be and so the function, for example, a broadband lambda probe can be fulfilled. It can both the described V characteristic can be measured, as well as an additional clear statement about whether the exhaust fat or lean, d. H. Lambda bigger or bigger smaller than 1. In other words, the present invention is based on the idea that it is determined on which "branch" the V-shaped characteristic of the current operating point lies. there can be, since this additional statement let win very quickly according to the invention, a Measuring speed, especially a sample rate, compared significantly increase conventional processes and one Measuring dead time to obtain this additional statement clearly to reduce. The total time required for the actual measurement and for obtaining the additional statement about The air range is needed, is thus very low. Furthermore, the resolution of the ambiguity can be by obtaining the additional statement even at a Disturbed background at any time and clearly perform. Furthermore, the inventive leads additional statement that also the resolution in the range around λ = 1, d. H. in an area with an oxygen content in the exhaust of about 0 percent, in addition to improve.

The proposed methods, the proposed sensor element and the proposed sensor arrangement with at least one of the invention Sensor elements can generally be used to determine at least a physical and / or chemical parameter of a gas be used in a sample gas space. Below a parameter is basically any measurable physical and / or to understand chemical property. In particular, you can however, the methods and apparatus for determining a concentration or a partial pressure of a gas component of the gas in the sample gas space are used, in particular for the determination of an oxygen partial pressure in an exhaust. The proposed method can be an inventive Use sensor element, and the sensor arrangement is set up, to perform a method according to the invention. In this respect, for possible embodiments of the Each case referred to the devices, and for possible embodiments of the devices on the process.

A sensor element is used which has at least one first electrode, at least one second electrode and at least one solid electrolyte connecting the first electrode and the second electrode. This first electrode, the second electrode and the solid electrolyte form components of a pumping cell. Preferably, as described above, the sensor element is constructed as a single cell, ie it has no further contacted electrodes except this pump cell. In principle, however, a single cell also includes sensor elements in which further, non-contacted electrodes are present and / or in which existing electrodes are designed in two or more parts with a plurality of interconnected partial electrodes. The first electrode and the second electrode are in each case arranged in at least one first cavity or in at least one second cavity, wherein the cavities are to be configured separated from one another. Via a first diffusion barrier, the first cavity or the first electrode should be acted upon with gas from the measurement gas space and the second cavity or the second electrode via a second diffusion barrier. The cavities may be wholly or partially unfilled, but may also be completely or partially filled with a gas-permeable medium, such as a highly porous ceramic medium, such as highly porous Al ₂ O ₃ . According to the invention, the two cavities have different volumes, which, as will be explained in greater detail below, enables a resolution of an ambiguity of the characteristic curves. In particular, the volume of the first cavity is smaller than that of the second cavity.

Alternatively, but in particular additionally, the resolution of the ambiguity of the characteristic curves can be made possible or improved by the fact that the first diffusion barrier is made denser than the second diffusion barrier, wherein below the density the diffusion barrier is to be understood in particular the measure with which an influx and / or diffusion of gases, in particular oxygen and / or hydrocarbons and / or CO and / or CO ₂ and / or H ₂ and / or H ₂ O, is affected in the first and second chamber.

The Electrodes may also be designed for this purpose that they have different areas, in particular that the electrode is in a larger of the two Cavities and / or in the cavity that is behind the more open of the two diffusion barriers, a larger one Surface than the electrode in the other cavity or on her own and / or together with her assigned Hollow chamber has a larger capacity as the other electrode, or the other electrode together with the other hollow chamber.

The Method according to the invention has a bipartition the operating mode, with at least two measuring phases provided are. The measuring phases described below must however, not necessarily performed in the order shown and it can be additional, not listed measuring phases are performed. Also Individual measuring phases can be completely or partially repeated be performed.

In a first measurement phase, a voltage U _{p is applied} to the pump cell and a current I _p measured by the pump cell. This measurement is preferably carried out in the limiting current operation of the pumping cell.

In a second measurement phase, the current I _p is then briefly changed by a predetermined value ΔI _p . In this case, a voltage change ΔU _p occurring in the case of this current change is measured. This second measuring phase serves to obtain an additional statement which makes it possible to produce a uniqueness of the first measuring phase. In an evaluation step, it is then concluded from the current I _p and from the voltage change ΔU _p to the parameter of the gas to be determined.

Alternatively, it is also possible not to change the current in the short term by a predetermined value ΔI _p , but to a predetermined value, in particular to the value whose amount is not less than 500 uA and / or not more than 50 mA, or its value 0 mA is.

As described above, the parameter may in particular comprise a concentration and / or a partial pressure of a gas component in the gas, in particular an oxygen partial pressure. Frequently, a unambiguous relationship between the current I _p and the concentration or the partial pressure often occurs in unicellular broadband sensor elements, for example in the form of an at least approximately V-shaped I _p (λ) characteristic curve and / or one likewise at least approximately V-shaped I _p (pO 2) characteristic curve, wherein pO 2 is, for example, the partial pressure of oxygen in the measurement gas. In this case, the measured voltage change .DELTA.U _p can be used to select from two possible concentrations, for example, in a V-shaped characteristic two possible concentrations or partial pressures, a certain concentration and / or a certain partial pressure, which then corresponds to the parameter to be measured. In other words, it can be determined, for example with the aid of the additional information about ΔU _p obtained in the second measuring phase, whether the gas is in a rich air range or in a lean air range.

To resolve the ambiguity, for example, to determine a current air range (fat / lean), for example, a threshold value method can be used. In this threshold value method, for example, the measured voltage change ΔU _p can be compared with one or more threshold values. According to this comparison, a specific partial pressure can be selected from two or more suitable concentrations and / or partial pressures or ranges of the concentrations and / or the partial pressure. For example, a simple threshold can be set, and it can, for example, when ΔU _p falls below or exceeds this threshold, it can be concluded that there is a certain air range. Accordingly, the correct value can then be selected, for example, from a V-shaped characteristic curve, that is, the correct air ratio λ can be assigned to the measured pumping current I _p .

A basic idea of the present invention therefore lies in the division of the operating mode into the at least two measuring phases. These can, for example, as described in more detail below, be carried out alternating in time, or intermittently, in the sense that the second measuring phase is carried out only at given times or in a case of need. In the first measurement phase, for example in a constant voltage or constant current mode, the possibly ambiguous parameter, for example a still ambiguous lambda value corresponding to a V-shaped characteristic curve, is determined. Alternately or additionally, in the second measuring phase, for example in the context of a very short-term measurement, the additional information ΔU _p resolves the ambiguity with additional information. For example, it can be stated Whether the present exhaust gas is rich or lean, ie, whether the V-shaped characteristic I _p -Lambda is on the "lean branch" or "fat branch".

This ambiguity is resolved by artificially inducing an asymmetry between the first electrode and the second electrode, by the different volumes of the cavities and / or by the measures described above. This will be described below using the example of oxygen detection, but the explanation has to be correspondingly transferred to other parameters or gas components to be detected. Since, with a constant current direction or constant direction of the applied pump voltage U _p , in the rich and in the lean air range once the one side (diffusion barrier) and another times the other side (diffusion barrier) limiting (though both diffusion barriers, especially in the case of different sizes Cavities, can be configured the same), the asymmetry of the arrangement causes the short-term change in the pumping current by ΔI _p in the rich or lean air-fuel range has different effects in changes in concentration in the cavities. Due to the strongly non-linear characteristic of the Nernst voltage between the two electrodes, this asymmetry, however, results in a marked change in the pump voltage, ie the pump voltage changes .DELTA.U _p differ greatly in the rich air range and in the lean air range. In this way it can be determined metrologically, in which of the two cavities currently present a diffusion limitation. If the diffusion restriction in the chamber, from which the oxygen ions are removed by the pumping current, the exhaust gas is lean. On the other hand, if the diffusion limitation exists in the chamber to which oxygen is supplied by the pumping process, then the exhaust gas is rich. The geometry of the sensor element can be optimized so that this distinction can be made in a comparatively short time, for example well below 50 ms, by means of a current and / or voltage signal.

Furthermore, the activation can then be optimized so that the obtained signal as a function of the oxygen content of the exhaust gas a characteristic curve by λ = 1 (0% oxygen in the exhaust gas) receives. The characteristic of this signal is usually similar to the signal of a jump probe. Thus, the proposed method combines in itself the characteristics of a simplified broadband probe and a jump probe. Thus, with an exhaust gas around λ = 1 targeted only the signal of the second measurement phase can be evaluated, since this is around λ = 1 is more accurate than the broadband signal I _p . Overall, this allows a simplified broadband probe and a jump probe in a sensor element combine, at a complexity level, which is only slightly above that of a simple jump probe.

The second measurement phase, which can be used for example only to resolve the ambiguity (which, as described above, but also can be additionally evaluated), preferably can only be switched on briefly, for example as a short-term interruption of the first measurement phase. Thus, the sensor element can be operated in the first measurement phase in particular during a predominant first portion of an entire measurement time, the sensor element being operated in the second measurement phase for a second portion of the total measurement time, which is for example a maximum of one third of the first component. For example, the second measurement phase can also only be used to carry out control measurements in order to ensure that the gas mixture is still in the concentration range and / or partial pressure range in question, for example a rich or lean air range. Thus, for example, the method can be carried out in such a way that an assumption about a candidate concentration range and / or partial pressure range, for example over an air-fuel ratio range (rich / lean), is stored in accordance with a respective last performed second measurement phase. The stored assumption can then be used in the at least one evaluation step. At regular or irregular intervals, for example in a case of need (for example, after a load change of the internal combustion engine) then a second measurement phase can be performed as a control step, which can be checked and optionally updated due to the voltage change ΔU _p measured in the second measurement phase. For example, the assumption in the form of a "flag" or another type of variable assignment in a data memory, such as a volatile or non-volatile memory, stored, wherein in the evaluation of the pumping current I _p the first measurement phase to this stored value with the assumption resorted to over the air range.

Of the The second portion may, for example, have a duration which 30 ms, preferably 10 ms and more preferably 5 ms or 1 ms does not exceed. Alternatively or additionally the first measurement phase and the second measurement phase also performed clocked be, d. H. in one at predetermined intervals performed Change between the first measurement phase and the second measurement phase.

The measurement of the voltage change .DELTA.U _p can be done, for example, that after ei After a switchover to the second measuring phase, the pumping voltage U _{p is} measured at a predetermined time and this is compared with the pumping voltage U _p previously applied in the first measuring phase. This measurement of the pumping voltage after a predetermined time t * after switching to the second measuring phase de facto means a measurement of the pumping voltage change .DELTA.U _p , which is at least approximately proportional to an additional amount of charge flowed .DELTA.Q = .DELTA.U _p * t *.

All in all can the sensor element in the first measurement phase, for example in operated in a constant voltage mode. In the second measurement phase For example, the sensor element can be operated in a constant current mode. However, other embodiments are conceivable, as below is explained in more detail.

As described above, in an air number range around λ = 1, the characteristic of the pump voltage change, that is, the ΔU _p lambda characteristic, can be used as additional information for accurately determining the concentration and / or the partial pressure. In this area, in which ΔU _p , similar to the Nernst voltage characteristic, shows a steep course, can at λ Lamb 1 small lambda values or lambda values in which the V-shaped I _p characteristic can be faulty , provide more exact values than this I _p characteristic.

The condition in which λ = 1, even with aged sensor element can still be reliably detected by a minimum of the pumping current, without the accurate knowledge of the absolute value of the pumping current (near 0 mA) would be crucial. It is also possible, during operation of the sensor element, to know that a state in which λ = 1 exists is to be used to adapt the characteristic of the pump voltage change, that is to say the ΔU _p lambda characteristic, to an aging state of the sensor element. For example, set thresholds can be adjusted, from the undershoot or overshoot is closed to the air range, be adjusted.

It is also possible, following a passage of the pumping current through a minimum, ie following a transition of the exhaust gas between "rich" and "lean", from the time derivative of ΔU _p determine in which direction the gas exchange took place, ie which exhaust gas is currently present or what its oxygen partial pressure is. With this information, the ΔU _p lambda characteristic can also be updated.

As described above, the measuring principle is based on the fact that, in particular by the different volumes of the cavities, and / or by other measures described above, an asymmetry in the pump cell is created because the supply or discharge of certain gas components from the cavities in precipitate different concentration changes. In this way, the currently limiting diffusion barrier can be determined. Due to the strongly non-linear Nernst voltage characteristic curve, even small changes in the ratio of the volumes make themselves noticeable in a significant change in the steepness of the ΔU _{p characteristic} curve.

Especially it is preferred if the second cavity is a factor of 1.1 to 50, preferably by a factor of 1.2 to 5 and more preferably by a factor of 1.5 to 3 larger volume as the first cavity. The smaller of the two cavities, So according to the nomenclature used here, the first cavity can basically also have a very small volume, up to the extreme case, in which this volume is at least almost completely disappears, so it becomes zero. The concept of the cavity is Thus, although preferred to understand that this one Space of finite volume, but is to be construed broadly, so that also the described borderline case, in which the volume of a Cavity at least approximately disappears includes should be. In the latter case, for example, the first diffusion barrier immediately adjacent to the first electrode. Again, this case in which the first cavity has a zero volume, so should to be covered by the subject invention. Overall, the conditions the void volumes and these volumes themselves over the Life of the sensor elements away, in contrast, for example on the properties of the diffusion barriers, comparatively minor changes subjected, so that the measuring principle of the invention has a high long-term stability.

Especially It is further preferred if the tightness of the first diffusion barrier clearly different from the tightness of the second diffusion barrier, especially if at a given difference in oxygen partial pressure which abuts along both diffusion barriers, at an operating temperature of the sensor, especially at 750 ° C, the rate of oxygen transport through the first diffusion barrier by at least 15%, in particular to at least 30%, more preferably at least 50%, is smaller as rate of oxygen transport through the second diffusion barrier.

The sensor element may in particular comprise a layer structure. In this case, the first cavity and the second cavity may be arranged in different layer planes. For example, the first cavity may be located on a surface of the sensor element facing the measurement gas space be arranged and separated from the measuring gas space, for example, by a gas-tight cover and the first diffusion barrier, whereas the second cavity may be disposed in a lower layer of the sensor element and may be acted upon, for example via a gas inlet hole with gas. Alternatively, however, the two cavities can also be arranged in deeper layer planes of the sensor element, for example in the same or in different deeper layer planes, and for example, both via a gas inlet hole, over an end face of the sensor element or on different, opposing surfaces or sides of the Sensor element are exposed to gas. Again alternatively, the two sensor elements can also be arranged in the same layer plane, for example once again in a layer plane on the surface of the sensor element and / or in deeper layer planes of the layer structure. In the latter case, for example, it is again possible to apply gas from the measuring gas space via gas inlet holes.

The proposed sensor arrangement has in addition to at least one inventive Sensor element further at least one controller. This control For example, all or part of a motor control be integrated into an internal combustion engine or may also be wholly or partially be designed as a separate control. The controller can be a or more electronic components as well as, alternatively or additionally, one or more microprocessors. The control is particularly adapted for carrying out the inventive Process. For this purpose, the controller can, for example programmatically be set up to carry out the method, for example by equipping with a corresponding computer program.

The Control may in particular comprise an electrical energy source. This at least one electrical energy source can, for example, a Voltage source, for example, a constant voltage source and / or an adjustable constant-voltage source, and / or a Power source, such as a constant current source and / or a comprise adjustable or adjustable constant current source.

As As stated above, the measurement phases 1 and 2 are preferably in different Measurement modes performed. Thus, in particular, the first measurement phase be performed in a constant voltage mode, and the second measurement phase in a constant current mode. Metrologically So this would be a switch, for example from a constant voltage source to a constant current source required. Since this would mean a relatively high outlay in terms of apparatus, it is particularly preferred to use only either a voltage source or to provide a power source. Nevertheless, in each other Measuring phase to provide the required operating mode can then a current source or a voltage source can be simulated.

For example, the energy source may comprise only one voltage source, by means of which the voltage U _{p is} generated in the first measurement phase. The controller can then be set up to simulate a current source in the second measuring phase by means of the voltage source and in this way to generate the current change ΔI _p . Alternatively, the energy source may comprise a current source by means of which in the first measurement phase a voltage source can be simulated in order to generate the voltage U _p , wherein in the second measurement phase the current change ΔI _p can be generated by means of this current source.

Under a "simulated energy source" is a mode to be understood by an appropriate regulation of existing ones Energy source another type of energy source is "simulated". This can be done by a suitable controller, for. B. a continuous or quasi-continuous, or a value and / or time discrete Controller done. For example, a voltage source can thereby be simulated that a current is impressed such that to form a set target voltage. The simulation a power source by means of a voltage source can be at appropriate Done way. The existing energy source can be operated modulated be, for. B. by a time or amplitude modulated pulse width modulation (PWM). In such a mode of operation can be different Measured variables are regulated, for. B. on the voltage in the off phase of an amplitude modulated PWM, for example perform a current-resistance compensation (IR compensation). By measuring further voltages of the PWM can be in such an operation z. B. also the affected ohmic Measure the internal resistance from which the sensor temperature is determined leaves.

Thus, a "simulation" can be understood, for example, as a measurement mode in which, in fact, the respective other of the variables U, I is determined as the actual quantity and / or variable to be actually controlled, but at regular or irregular intervals, for example, by control measurements it is determined whether the actually controlled variable corresponds to the setpoint. Thus, for example, a voltage source can be simulated by adjusting a current such that the setpoint voltage to be set is formed. In regular or irregular moderate intervals, for example, then be checked by a voltage measurement, if the voltage is still the setpoint, and the current can be readjusted accordingly. The simulation of a current source by means of a voltage source can be carried out in a corresponding manner. In this way, the electrical control of the sensor arrangement can be made comparatively simple, so that also the method according to the invention, which is based on the mentioned switching between the operating phases, can be easily realized metrologically.

With in other words, the effect of the present invention is based on by targeted transport of oxygen ions to one of the two electrodes to set a reference gas environment (for example oxygen rich, as explained below), similar to a pumped one Reference. This is done by first targeted in the second measurement phase Oxygen ions are transported to this electrode. Furthermore will be in the second measurement phase after getting a certain state has set a measurement of the Nernst voltage of the other electrode made against this generated reference. If the exhaust gas is fat, will a magnitude large Nernst tension measured, it is lean, so is an amount measured small Nernst voltage. Similarly, a low-oxygen reference be generated at the current directions and relative potentials turning back.

In the present invention, however, another effect for distinguishing between rich and lean exhaust gas comes into play: in the lean case, molecular oxygen is decomposed at the first electrode and pumped against a gas with a high oxygen content (λ> 1). In the case of the variation of the pumping current, this changes at most slightly (compared with the following case) and the stroke of ΔU _p is small. In the rich case (λ <1), however, CO ₂ and H ₂ O are decomposed at the first electrode, which is only possible with the applied pump voltages (for example at 450 mV) because the gas composition at the second electrode (λ = 1 ) gives a sufficient effective pumping voltage (low reverse voltage). Since, in the case of a brief change in the pumping current, the gas composition changes rapidly due to the denser diffusion barrier and / or the smaller cavity (λ << 1 or λ >> 1), the Nernst countervoltage also changes considerably, and the changed pumping current is maintained a significant stroke of ΔU _p .

Brief description of the drawings

embodiments The invention is illustrated in the drawings and in the following Description explained in more detail.

It demonstrate

1 a characteristic of a typical jumping probe;

2 a characteristic of a typical multicell broad-band probe;

3 and 4 different embodiments of the prior art, single-cell sensor elements;

5 a typical characteristic of the sensor elements in the 3 and 4 ;

6 an embodiment of a sensor element according to the invention;

7 Characteristics determined by the method according to the invention; and

8th a schematic flowchart of an embodiment of a method according to the invention.

In the 1 and 2 typical measuring curves of known sensor elements are shown. So shows 1 a waveform of a jump probe, in which the voltage between two electrodes is measured. For example, one of these electrodes may be exposed to a reference air space with a known gas mixture composition. The trace follows the equation:

und

die Partialdrücke des Sauerstoffs im Abgas (Messgasraum) bzw. dem Referenzluftraum. Die Nernstspannung ist hier als Funktion der Konzentration des Sauerstoffs c (O₂) aufgetragen und zeigt einen charakteristischen Sprungverlauf bei 0% Sauerstoff (λ = 1).Here R denotes the universal gas constant, T the temperature, F the Faraday constant and

and

the partial pressures of the oxygen in the exhaust gas (sample gas space) or the reference air space. The Nernst voltage is plotted here as a function of the concentration of the oxygen c (O ₂ ) and shows a characteristic jump course at 0% oxygen (λ = 1).

In 2 In contrast, a typical signal curve (referred to here as "I") is shown schematically as a function of the oxygen concentration or the air ratio for a prior art multi-cell broadband lambda probe. For the measuring principle of these broadband lambda probes, for example, on Robert Bosch GmbH: Sensors in the motor vehicle, 1st edition 2001, pages 116 to 117 to get expelled. It can be seen that the signal has at least approximately a clear, linear course.

In contrast to the signal measured with a jump probe according to 1 and that with egg According to a multi-cell broadband probe structure, the measured signal 2 are in the 3 and 4 simple, the prior art and designed as a pump cell sensor elements 110 shown schematically in section. The sensor elements each comprise a solid electrolyte 112 which is a first electrode 114 and a second electrode 116 combines. The solid electrolyte 112 and the electrodes 114 . 116 together form a pump cell 118 , The electrodes 114 and 116 are essentially equivalent. The first electrode 114 is in a first cavity 120 arranged, which via a first diffusion barrier 122 with a sample gas chamber 124 communicates, and the second electrode 116 is in a second cavity 126 arranged, which via a second diffusion barrier 128 communicates with the sample gas space. The electrodes 114 . 116 are via electrode leads 130 . 132 contactable.

The solid electrolyte 112 may include, for example, yttria-stabilized zirconia (YSZ). The electrodes 114 . 116 For example, they may comprise noble metal electrodes, for example platinum electrodes, in particular in the form of ceramic-metal composites (cermet). The diffusion barriers 122 . 128 For example, they may comprise a porous ceramic material, such as an alumina or a zirconia.

The embodiments according to the 3 and 4 differ essentially by the arrangement of the pumping cell 118 , While in the embodiment according to 3 the electrodes 114 . 116 on opposite sides of the layered solid electrolyte 112 are arranged, and thus in different layer planes of the sensor element 110 are arranged, are the electrodes 114 . 116 at the in 4 shown embodiment arranged in the same layer plane. In this case, the electrodes are 114 . 116 for example, on the surface of the sensor element 118 arranged and through a cover layer 134 from the sample gas space 124 separated. Furthermore, what is only in 4 is shown, a heating element 136 be provided by means of which, for example, the properties of the solid electrolyte 112 and / or the diffusion barriers 122 . 128 , in particular the limit currents, can be adjusted. Such a heating element 136 can for example in the embodiment according to 3 be provided.

Furthermore, the cavities 120 . 126 be designed either as completely unfilled cavities. Alternatively or additionally, these cavities 120 . 126 also be completely or partially filled with a highly porous material, such as a gas-permeable, highly porous alumina.

Is at the sensor elements 110 according to the embodiments in the 3 and 4 at temperatures of typically 500 to 1100 ° C applied an external voltage (in 3 exemplified by "500 mV"), so can between the electrodes 114 . 116 an oxygen ion current can be measured. The applied voltage is also referred to as pumping voltage U _p and the measured current as pumping current I _p . Such a pumping current I _p shows the in 5 illustrated typical course as a function of the oxygen concentration or the oxygen partial pressure in the gas of the sample gas space 128 , For example, an exhaust gas of an internal combustion engine. Characteristic of the in 5 characteristic curve shown is the V-shape, ie a monotone decrease in the oxygen flow in the rich exhaust gas (λ <1, in 5 With 138 referred to) with increasing oxygen deficiency, and a monotone increase in the lean exhaust gas (in 5 with the reference number 140 designated) with increasing oxygen content. A similar course applies to a wiring of the sensor elements 110 with a constant current and measuring the voltage. Due to the illustrated ambiguity, ie that a given pumping current two lambda values are assigned, no clear conclusion can be made on the oxygen content of the exhaust gas.

In order to explain the V-shaped profile, it is assumed below by way of example that the pump voltage U _{p is selected} as a fixed voltage in its sign such that oxygen ions in the ceramic solid electrolyte 112 from the second electrode 116 to the first electrode 114 be pulled like this in 3 is shown symbolically. However, another pumping direction is possible in principle. The reason for the monotonous increase of the sensor signal according to 5 in the lean area 140 is that the wake of oxygen molecules through the second diffusion barrier 128 is limited. It can with a constant voltage of typically 300 to 600 mV only as much oxygen through the ceramic diffusion barrier 128 be transported and thus measured as a current, such as oxygen in the same material extent through the diffusion barrier 128 nachströmt. This is the principle of a lean-flow limit current probe 140 , The fact that in front of the first electrode 114 another diffusion barrier 122 is present, this does not change anything fundamental. The expansion of the oxygen ions at the first electrode 114 takes place unbraked, since there is already lean exhaust gas, which by additional, on the diffusion barrier 122 only slowly draining oxygen is made even leaner. On the Nernstgegenspannung this, however, has only a very small influence, since the signal, as from 1 it can be seen, in the lean area with a further addition of oxygen changes only insignificantly.

In the rich exhaust gas (area 138 in 5 ), however, is virtually no oxygen in the second cavity 126 available. Therefore, by applying the pumping voltage U _p to the second electrode 116 virtually no molecular oxygen decomposed. Rather, other oxygen-containing gases, in particular H ₂ O and / or CO ₂ , are converted with the elimination of oxygen into H ₂ or CO, ie not completely oxidized gases. In principle, such an operation would be limited by the backflow of the oxidized gases H ₂ O and CO ₂ . Practically dominated here, however, another mechanism. So takes place at the first electrode 114 In the rich exhaust gas, a reverse reaction takes place, in which not fully oxidized gases, such as H ₂ and CO, and other unsaturated molecules are saturated with oxygen, that are further or completely oxidized. But these are in their wake by the first diffusion barrier 122 limited. More oxygen is being built up than non-saturated molecules are present or beyond the first diffusion barrier 122 can be replenished, then the Nernst counter voltage increases, and the current is suppressed. Determining in the lean air range 140 So is the Gasgemischzusam composition in the second cavity 126 whereas in the fat air range 138 the gas composition in the first cavity 120 dominated. At a reversal of the pump voltage, which in 3 is given as an example with 500 mV, the conditions are reversed because the electrodes 114 . 116 are essentially equivalent.

The in the 6 illustrated embodiment of a sensor arrangement 142 According to the invention provides that the cavities 120 . 126 are dimensioned differently. In particular, the volume of the first cavity 120 smaller than that of the second cavity 126 , Furthermore, the first diffusion barrier 122 denser than the second diffusion barrier 128 , The sensor arrangement 142 includes a sensor element 110 and a controller 144 , what a 6 is shown only symbolically and which is explained in more detail below. The control 144 can via the electrode leads 130 . 132 with the electrodes 114 . 116 be connected, this can apply a voltage and / or current (in 6 is exemplified by a pump voltage of 450 mV) and can, as in 6 over the interface 146 be hinted at, be charged with additional information from the outside or submit information to the outside.

In 6 is symbolically a sensor element 110 similar to the one in 3 shown construction, ie in a structure with a vertical arrangement of the pumping cell 118 with opposing electrodes 114 . 116 , Analogously, however, a structure would be similar to 4 conceivable or another, not shown construction of the sensor element 110 ,

With the in 6 illustrated sensor arrangement 142 can basically a characteristic similar to 5 be recorded. This characteristic is in 7 represented and there with the reference number 148 characterized. It shows the typical, from 5 known V-shaped course with the said ambiguity. The inclusion of a pumping current I _p at a predetermined pumping voltage U _p is also referred to as the first measuring phase. As from the ambiguity according to 7 It can be seen, however, for example, from a measured pump current of 100 uA can not say for sure whether an oxygen concentration of -1.75% or an oxygen concentration of about 1% is present.

To dissolve this ambiguity a second measuring phase is proposed in which by a short-term selective increase or decrease of the pump current by an amount .DELTA.I _p is determined, in which chamber a boundary just present. If the response signal, ie the voltage change between the electrodes, falls 114 . 116 (or alternatively, the current change between these electrodes) is large, so is currently the smaller of the two cavities 120 . 126 determining or limiting. If the answer turns out to be small, then the larger of the two cavities 120 . 126 determining, since a larger gas change is required for a correspondingly large signal. For example, what is referred to as a "big answer" or "small answer" may be determined by means of one or more thresholds. An example of such a threshold is in 7 with the reference number 150 designated. The characteristic of the second measuring phase, ie the measured at a current change of .DELTA.I _p changing the pumping voltage .DELTA.U _p, which refers to the left scale, is in 7 with the reference number 152 designated. If the measured value measured in the second measuring phase is correspondingly above the threshold value in the illustrated example 150 , so dominates the smaller first cavity 120 , and the exhaust gas in the sample gas space 124 is in the rich air range 138 whereas at a reading below the threshold 150 the larger second cavity 126 dominates and the exhaust gas in the sample gas space 124 in the lean air range 140 located. With the help of this additional information, the ambiguity of the previously measured pumping current can be resolved. Thus, for example, in the case of a measuring signal in the second measuring phase, it may be smaller than the predetermined threshold value 150 It can be seen that the pump current of 100 μA in this illustrated example corresponds to a composition of 1% oxygen.

The threshold 150 , from which a signal is considered large or small depends in particular on the relative size of the volumes of the cavities 120 . 126 , or from the tightness of the diffusion barriers 122 . 128 from. Furthermore, the threshold depends 150 As a rule, from the additional or less pumped charge amount or the current change .DELTA.I _p . Because the cavity sizes 120 . 126 are subject only to a small extent aging, in particular lower aging effects than those usually over the life of the sensor element 110 slowly adding diffusion barriers 120 . 128 , the value of the volumes of the cavities 120 . 126 , in particular the value of the ratio of these volumes, usually very accurately over the life of the sensor elements 110 be specified. Alternatively, this ratio or the threshold value 150 also in operation, for example, at regular or irregular intervals, be re-determined at λ = 1, for example by the minimum of the characteristic 148 is detected and the associated value of the characteristic 152 is determined.

To explain the principle of measurement of the two measuring phases, let us again refer to the Nernst voltage curve in 1 referred to the voltage at the pump cell 118 clarified. In this case, a measuring method with the first and the second measurement phase initially in the lean air range 140 and then in the rich air range 138 explained. In the lean air range, for example, at an oxygen concentration of 3% in the sample gas space 124 For example, in the first measurement phase, a pump voltage of 450 mV is applied between the electrodes 114 . 116 placed. In the following, again assume the polarity described above, in which the second electrode 116 acts as a mounting electrode for oxygen, so it is acted upon by a negative electrode potential. However, the explanation for a reverse pumping voltage is analogous. In this lean air range, the limiting factor is the night-time flow of oxygen through the second diffusion barrier 128 , Without pumping current lies in both cavities 120 . 126 an atmosphere of about 3% oxygen in front of what is in 1 symbolic with the lean point 154 is indicated. It should be noted, however, that the percentages are to be understood as exemplary only.

Will now in the first measurement phase, a constant pumping voltage U _p between the electrodes 114 . 116 applied, so a pumping current I _{p flows} through the pumping cell 118 , Will the pump cell 118 operated in the limiting current range, the oxygen is completely from the second cavity 126 pumped out so that there approximately the condition λ = 1 or C (O ₂ ) = 0 (ie oxygen concentration zero) prevails. This is in 1 symbolically with the reference number 156 (λ = 1 point) indicated. The one in the first cavity 120 pumped oxygen causes the gas mixture in this first cavity 120 becomes a little leaner, but at the electrode potential of the second electrode 114 practically nothing changes. Between the electrodes 114 . 116 So is the voltage of for example 450 mV. For example, at this voltage, a pump current I _p of 310 μA is measured. By means of the characteristic 148 in 7 Now, two potential oxygen concentrations can be considered as possible candidates for the actual oxygen concentration in the sample gas space 124 determine.

Now follows the second measuring phase in which a constant voltage measurement switches over to a constant current measurement. In this case, the pump current I _{p is} temporarily increased by a fixed ΔI _p or alternatively, since ΔI _{p can} also be selected negative, decreased. Alternatively, the pumping current I _p in this phase can also be set or adjusted to a fixed value.

Because the pump cell 118 was already operated before the increase (or decrease) by ΔI _p in the limiting current, now finds in the second cavity 126 a decomposition of oxygen-containing gases instead, for example, a water decomposition. The resulting oxygen ions become the first cavity 120 pumped. Now it is noticeable that in 1 the second electrode 116 which are in the larger cavity 126 is arranged, is arranged in the steep region of the characteristic. At a fixed amount .DELTA.I _p is a fixed predetermined number of oxygen molecules, if necessary. Water molecules in the large cavity 126 decomposed. This will cause the atmosphere in this second electrode cavity 126 something fatter, however, because this decomposition of a constant number of hydrogen molecules to a comparatively large volume of the cavity 126 distributed, or over the relatively open diffusion barrier 128 If water or O ₂ or CO ₂ can rapidly flow in, the concentration of oxygen in the electrode cavity as a whole decreases relatively negligibly (a negative oxygen concentration in this figure is considered an oxygen deficiency). While the lean point 154 that is, the atmosphere in the first electrode cavity 120 on the curve in 1 moved a little to the right, the atmosphere moves in the second cavity 126 from the λ = 1 point 156 only slightly to the left. This results in an increase in the pumping voltage by ΔU _p , which, however, is comparatively low. This is in 7 symbolic with the dot 158 which shows an example of a measuring point of a pumping current increase ΔU _p in the lean region. This pumping current increase 158 is below the threshold 150 , which in turn suggests that the air ratio in the lean area 140 lies. This information of the second measuring phase can be used to set the pumping current I _p measured in the first measuring phase assign a clear value.

Analogously, the following is the situation in the rich air range 138 explained. The starting point, for example, is an oxygen deficit of 3%, which is in 1 symbolic with a bold point 160 is designated. Again, it should be noted that this example is just one of many possible situations. In this fat point 160 If no pumping voltage is applied, first of all there prevails again in both cavities 120 . 126 an atmosphere with a negative oxygen content of 3%, ie an oxygen deficit of 3% compared to a stoichiometric equilibrium.

Now, in turn, first the first measurement phase is performed, in which, with the same polarity as in the case of the lean gas mixture described above, a pumping voltage U _p to the pumping cell 118 is created. It finds, as described above, at the first electrode 114 an inverse reaction takes place in which non-fully oxidized gas components, such as fuel gases (for example, hydrocarbons, hydrogen are oxidized by oxygen that has been removed.) A limiting factor is thus the first diffusion barrier 122 and the post-flow of fuel gases through this first diffusion barrier. In the limiting current operation, therefore, oxygen is through the solid electrolyte 112 from the second cavity 126 in the first cavity 120 pumped to react there with fat gases. By this removal of oxygen from the second cavity 126 the atmosphere is in the second cavity 126 slightly fatter, leaving the working point in this second cavity 126 from the fat point 160 in 1 slightly to the left. As from the Nernst curve in 1 can be seen, thereby changing the electrode potential of the second electrode 116 but only insignificantly.

In contrast, however, all react through the solid electrolyte 112 supplied oxygen molecules directly with fuel gases in the first cavity 120 , Through the first diffusion barrier 122 Just as many fat gases are in the first cavity 120 replenished that the subsequently supplied grease gases and by the solid electrolyte 112 transport oxygen molecules stoichiometrically balance each other. However, this means nothing else than that in the first cavity 120 the stoichiometric equilibrium point λ = 1 is present. In other words, the operating point travels in the first cavity 120 towards the λ = 1 point 156 , The mixed gas composition in the second cavity 120 Moves a little further into the area of the rich gas mixture composition, for example towards the point 160 , which does not change the voltage. Between the two electrodes 114 . 116 So again, for example, is the pumping voltage of 450 mV, corresponding, for example, since the boundary currents of the diffusion barriers 122 . 128 may differ, a pump current of I _p = 160 uA.

Now, in turn, the second measuring phase is switched over and the pumping current I _p is increased (or decreased) by a constant value ΔI _p by the same value ΔI _p as in the lean case described above. For example, this in turn may be a current increase by ΔI _p = 300 μA. In total, 460 μA are now flowing in the second measuring phase at short notice. Now it is noticeable that, in contrast to the lean case described above, in the rich air range, the operating point is the one with the smaller cavity 120 equipped first electrode 114 is in the range of λ = 1, ie in the steep region of the Nernst curve. So while the short-term pumping current increase in the second measurement phase, the atmosphere in the second cavity 126 slightly to the left of the point 160 in 1 wanders, wanders the atmosphere in the first cavity 120 from the λ = 1 point 156 out in 1 to the right, because the atmosphere in this cavity 120 is moved in the direction of the lean air range. Since the same number of molecules is decomposed by the constant current increase ΔI _p as in the lean case described above, this number of decomposed molecules now, however, on the smaller volume of the first cavity 120 In terms of percentage, the atmosphere changes in this first cavity 120 in the rich case stronger than the atmosphere in the second cavity 126 in the lean case. At the same time, through the denser diffusion barrier 122 only limited flow of fat molecules, which react with the pumped oxygen. Another effect is that it increases the necessary pumping voltage, because during the second measurement phase, the contents of the cavity 120 gets lean. Due to the strongly nonlinear course of the Nernst characteristic in 1 This affects a much larger change ΔU _{p of} the pumping voltage U _p than in the lean case. This is in 7 symbolic with the dot 162 illustrated, which terized the pump voltage increase in the fat area charak. This pump voltage increase 162 is significantly above the threshold 150 , so that out of the comparison with the threshold 150 can be concluded that a rich gas mixture composition in the sample gas space 124 is present. This information of the second measurement phase, together with the information about the pumping current I _p previously obtained in the first measurement phase, in turn allows an unambiguous assignment of the measurement signal of the first measurement phase to an oxygen concentration, an oxygen partial pressure and / or an air ratio λ. The described voltage measurements can advantageously take place in an unloaded state of the pumping cell.

In the present method, therefore, the strong non-linearity of the Nernst voltage curve in accordance with 1 caused enormous signal amplifications, which can cause significant voltage jumps even with a short-term current excitation. Thus, minimal changes in the lambda value of the tested chamber can shift the Nernst reverse voltage by up to several 100 mV. Thus clearly evaluable response signals .DELTA.U _p can be achieved by pumped circulation smallest operations to a current .DELTA.I _p here. An increase or decrease of the pump current by 100 μA for 10 ms or even only for 1 ms is sufficient to achieve a signal of, for example, 300 mV in the lean and only 50 mV in the rich region with suitably selected sensor element geometry. Alternatively, a fixed pump current I _p can be selected in this phase. The distinction between these two signals is technically easy to accomplish and could prove to be unaffected by typical electromagnetic interference (EMC) in the automotive environment.

In 8th is a possible flowchart of a method according to the invention shown very schematically. The reference number denotes 164 schematically a timing, so a timing, which between two measurement phases 166 . 168 switches. As described above, the first measuring phase includes, for example, a measurement of the pumping current at a predetermined pumping voltage. The second measurement phase 168 includes a short-term pump current boost, followed by a measurement of the associated pump voltage change. In addition, further measurement phases can be predetermined. As described above, for example, a clocked switching between the measuring phases 166 . 168 or, if necessary, switching on the second measuring phase 168 , For this purpose, the timing can be 164 for example, what in 8th with the reference number 170 is referred to, resort to additional information, such as information about a load change of an engine or any other change in an operating condition of an internal combustion engine, which is a renewed control measurement in the form of a second measurement phase 168 could make necessary.

As described above, the second measurement phase becomes 168 obtained additional information (step 172 ), which in an evaluation step 174 for evaluation of the pump current signal from the first measurement phase 166 can be used to resolve the above-described ambiguity of this signal.

Alternatively, as in 8th through the dashed arrow 176 is shown, also a direct evaluation of the information from the second measurement phase 168 respectively. For example, as above described, in an air speed range at λ = 1 around (ie C (O ₂₎ = 0% around) an immediate evaluation of the _p by the pumping current increase .DELTA.I pumping voltage variation .DELTA.U _p caused from the second measurement phase 168 done because the characteristic 152 In this area has a characteristic steep course and may be easier and more reliable to evaluate than the characteristic curve 148 , which is often subject to uncertainties in this area.

On the basis of in 6 symbolically represented control 144 An embodiment of the control will also be described below by way of example. The control 144 points in the in 6 illustrated embodiment, an electrical energy source 178 in the form of, for example, a power source 180 on. This electrical energy source 178 is preferably used for both measuring phases. Instead of a power source 180 may, alternatively or additionally, be provided a voltage source, as described above. Furthermore, the controller has 144 in the 6 illustrated embodiment, a voltage measuring device 182 on which of the power source 180 is connected in parallel, as well as a current measuring device 184 which are in series with the source of electrical energy 178 is switched.

The power source 180 is applied to the two sensor electrodes 114 . 116 via the electrode leads 130 . 132 connected. The current I _p is chosen in the first measurement phase so that between the two electrodes 114 . 116 sets the desired pumping voltage U _p of, for example, 450 mV. This can also be done in a regulated manner. The electrode voltage can affect the internal resistance of the pumping cell 118 or the sensor element 110 also be measured by modulation of the current. By this measure, let the time until a steady state is achieved, greatly shorten.

If the state of a constant pumping current I _{p is} reached, this current value is temporarily stored, for example. If this state of a constant current is reached, the first measurement phase can be terminated in principle. Now, for a fixed time, this current I _p plus (optionally minus) the fixed current value ΔI _{p is} impressed.

For example, if 310 μA was required beforehand to maintain the 450 mV, the current is now increased to 410 μA, for example. With a reduction, the current can also be negative overall. After a defined time (for example 10 ms or even 100 μs), the voltage reached is measured and this voltage or the difference .DELTA.U _p to a previously required Voltage, with a reference value (threshold 150 in 7 ) compared. Is the measured voltage difference or voltage greater than the threshold value 150 , so the exhaust gas is rich, is the measured voltage or voltage difference below the threshold 150 , so the exhaust is lean. In a reverse interpretation of the sensor element 110 in 6 , ie at one compared to the second cavity 126 larger first cavity 120 or at a pole reversal compared to the embodiment described above, the conditions would be reversed and the curve 152 for example, at the threshold value axis 150 or a symmetry axis mirrored. In 7 is the threshold 150 chosen such that it is about 150 mV, corresponding to the minimum of the I _p characteristic 148 , By means of this information, it is possible to switch over between the two characteristic curves (rich and lean) and output a clear measured value with regard to the oxygen concentration, the air ratio or the oxygen partial pressure from the measured value obtained in the first measuring phase, which is still ambiguous. Thereafter, the cycle described can begin with the measurement phases from the beginning.

As described above and as in 8th by the reference number 176 indicated, the response signal .DELTA.U _p , so the voltage increase or voltage decrease, either an increase in current or current reduction, can also be used directly as a measure of the gas mixture composition. The signal curve 152 is up to a scaling factor and an offset substantially the characteristic of a jump probe, as for example in 1 is shown. This is known to be particularly accurate in the range around λ = 1. The offset can be determined during operation, for example, as the value of the second measurement phase with a minimum signal of the first measurement phase. In order to increase the accuracy of the broadband probe thus obtained still, therefore, also the voltage signal directly .DELTA.U _p can be used (positive or negative, or alternatively, also the current signal) in the region around λ = 1, instead of the charged overall signal. In order to increase the accuracy even further and to enable a particularly simple type of evaluation, the voltage signal ΔU _p less the offset could be used directly in the range around λ = 1 instead of the total signal being calculated. This signal is in the range around λ = 1 except for a gain error, a very accurate lambda signal. Thus, the advantages of a broadband probe and a jump probe in a single sensor element 110 combined.

In an alternative embodiment of the invention, the current is not increased or decreased by a certain value ΔI _p , but for some time to a fixed value, e.g. B. set to 0. However, this also corresponds to a change of the current I _p by a value .DELTA.I _p (I _p ), but in this case the value .DELTA.I _p just corresponds to the previously measured in the measuring phase 1 current I _p and has a negative sign. As a result, the incoming downstream flow through the first diffusion barrier 122 or the second diffusion barrier 128 in relation to the size of the cavity 120 respectively. 126 measured. With a suitable choice of these parameters to each other can thus also a V-shaped curve of the voltage response can be achieved, instead of the characteristic 152 in 7 which resembles the characteristic of a jump probe. By comparing each voltage value thus obtained with an expected value can also be between a rich air range 138 and a lean air range 140 be differentiated. The same applies if it is not jumped to a current I _p = 0, but to another set value I _p , or if in constant voltage operation, the voltage is temporarily interrupted or switched to another value. However, the intrinsic disadvantage of these embodiments over the embodiment described above is that no such unique characteristic is achieved as in the method described above. Furthermore, the method described above allows the choice of the diffusion barrier properties of the diffusion barriers 122 . 128 regardless of the size of the cavities 120 . 126 , This allows greater freedom in the construction of the sensor element 110 ,

In various forms of expression, initial voltage jumps or current jumps can also be used directly after changing the current or the voltage as a measure of the internal resistance of the sensor element 110 and therefore used for its temperature.

QUOTES INCLUDE IN THE DESCRIPTION

This list The documents listed by the applicant have been automated generated and is solely for better information recorded by the reader. The list is not part of the German Patent or utility model application. The DPMA takes over no liability for any errors or omissions.

Cited patent literature

• DE 102005054144 A1 [0006]

Cited non-patent literature

• - Robert Bosch GmbH: "Sensors in the motor vehicle", 1st edition 2001, pp. 112-117 [0002]
• - Robert Bosch GmbH: Sensors in the motor vehicle, 1st edition 2001, pages 116 to 117 [0052]

Claims (17)

Method for determining at least one parameter of a gas in a measuring gas space ( 124 ), in particular for determining a concentration of a gas component of the gas, in particular for determining an oxygen partial pressure in an exhaust gas, wherein a sensor element ( 110 ) is used, wherein the sensor element ( 110 ) at least one first electrode ( 114 ), at least one second electrode ( 116 ) and at least one the first electrode ( 114 ) and the second electrode ( 116 ) connecting solid electrolyte ( 112 ), wherein the first electrode ( 114 ) in a first cavity ( 120 ) and via a first diffusion barrier ( 122 ) with gas from the sample gas space ( 124 ), the second electrode ( 116 ) in one of the first cavity ( 120 ) separate second cavity ( 126 ) and via a second diffusion barrier ( 128 ) with gas from the sample gas space ( 124 ) is applied, wherein the first cavity ( 120 ) and the second cavity ( 126 ) have different volumes, the method having at least the following measurement phases: in a first measurement phase, a voltage U _{p is applied} to a first electrode ( 114 ), the second electrode ( 116 ) and the solid electrolyte ( 112 ) comprehensive pump cell ( 118 ) and a current I _p , preferably a limiting current, through the pumping cell ( 118 ) measured; and - in a second measurement phase, the current I _{p is changed} by a predetermined value .DELTA.I _p , wherein a change in voltage occurring during this current change .DELTA.U _{p is} measured, wherein in at least one evaluation step from the current I _p and from the voltage change .DELTA.U _p to the parameter the gas is closed. Method according to the preceding claim, wherein the parameter comprises a concentration and / or a partial pressure of a gas component in the gas, wherein an at least partially non-unique relationship between the current I _p and the concentration and / or the partial pressure, wherein the measured voltage change ΔU _p is used to select from a plurality of possible concentrations and / or partial pressures a certain concentration and / or a certain partial pressure. Method according to the preceding claim, wherein a threshold value method is used, wherein the voltage change ΔU _{p is} compared with one or more threshold values, according to this comparison of two or more candidate concentrations and / or partial pressures a certain concentration and / or a certain partial pressure is selected. Method according to one of the preceding claims, wherein the sensor element ( 110 ) is operated during a predominant first portion of a total measurement time in the first measurement phase, wherein the sensor element ( 110 ) is operated for a maximum of one third of the first portion lasting second portion of the total measurement time in the second measurement phase. Method according to one of the preceding claims, wherein according to a respective last performed second measurement phase an assumption about a concentration range and / or partial pressure range coming forth is stored, in particular a presence of a rich or lean gas mixture composition in the sample gas space ( 124 ), wherein in the at least one evaluation step, the stored assumption is used, wherein in regular or irregular circumstances, a second measurement phase is performed as a control step, wherein due to the voltage change ΔU _p measured in the second measurement phase, the assumption is checked and / or updated. Method according to one of the two preceding claims, wherein the second portion has a duration which is 10 milliseconds and preferably does not exceed 5 milliseconds. Method according to one of the preceding claims, wherein the voltage change .DELTA.U _{p is} measured after a predetermined time after switching to the second measuring phase. Method according to one of the preceding claims, wherein the first measurement phase and the second measurement phase performed clocked become. Method according to one of the preceding claims, wherein the sensor element ( 110 ) is operated in the first measuring phase in a constant voltage mode, wherein the sensor element ( 110 ) is operated in the second measuring phase in a constant current mode. Method according to one of the preceding claims, wherein a concentration and / or a partial pressure of oxygen in the gas is determined, wherein in a range around an air ratio λ = 1 around for the exact determination of the concentration and / or the partial pressure, a known relationship between ΔU _p and the concentration and / or the partial pressure is used. Method according to one of the preceding claims, wherein the pump cell ( 118 ) is unloaded at least one voltage measurement. Sensor element ( 110 ) for determination min at least one parameter of a gas in a sample gas space ( 124 ), in particular for determining a concentration of a gas component of the gas, in particular for determining an oxygen partial pressure in an exhaust gas, wherein the sensor element ( 110 ) at least one first electrode ( 114 ), at least one second electrode ( 116 ) and at least one the first electrode ( 114 ) and the second electrode ( 116 ) connecting solid electrolyte ( 112 ), wherein the first electrode ( 114 ) in a first cavity ( 120 ) and via a first diffusion barrier ( 122 ) with gas from the sample gas space ( 124 ), the second electrode ( 116 ) in one of the first cavity ( 120 ) separate second cavity ( 126 ) and via a second diffusion barrier ( 128 ) with gas from the sample gas space ( 124 ) is applied, wherein the first cavity ( 120 ) and the second cavity ( 126 ) have different volumes. Sensor element ( 110 ) according to the preceding claim, wherein the second cavity ( 126 ) has a volume larger by a factor of 1.1 to 50, preferably by a factor of 1.2 to 5 and particularly preferably by a factor of 1.5 to 3, than the first cavity ( 120 ). Sensor element ( 110 ) according to one of the two preceding claims, wherein the sensor element ( 110 ) comprises a layer structure, wherein the first cavity ( 120 ) and the second cavity ( 126 ) are arranged in the same layer plane or in different layer planes. Sensor element ( 110 ) according to one of claims 11 to 13, wherein the first diffusion barrier ( 122 ) is made denser than the second diffusion barrier ( 128 ), in particular significantly denser. Sensor arrangement ( 142 ) for determining at least one parameter of a gas in a measuring gas space ( 124 ), in particular for determining a concentration of a gas component of the gas, in particular for determining an oxygen partial pressure in an exhaust gas, comprising at least one sensor element ( 110 ) according to one of the preceding, to a sensor element ( 110 ), further comprising at least one controller ( 144 ) for controlling the sensor element ( 110 ), wherein the sensor arrangement ( 142 ) is arranged to perform a method according to one of the preceding method claims. Sensor arrangement ( 142 ) according to the preceding claim, wherein the controller ( 144 ) an energy source ( 178 ), wherein the controller ( 144 ) is set up to the energy source ( 178 ) in the first measuring phase and the second measuring phase in one of the following ways: - the energy source ( 178 ) comprises a voltage source, the controller ( 144 ) is arranged to generate the voltage U _p by means of the voltage source in the first measuring phase, wherein the controller ( 144 ) is arranged to simulate a current source in the second measuring phase by means of the voltage source and in this way to generate the current change ΔI _p ; - the energy source ( 178 ) comprises a power source ( 180 ), whereby the controller ( 144 ) is set up by means of the power source ( 180 ) in the first measuring phase to simulate a voltage source and to generate the voltage U _p , wherein the controller ( 144 ) is set up in the measuring phase by means of the power source ( 180 ) to produce the current change ΔI _p .