DE102022210329A1 - Method for operating a sensor for detecting at least a portion of a measuring gas component with bound oxygen in a measuring gas - Google Patents

Abstract

A method for operating a sensor (100) for detecting at least a portion of a measuring gas component with bound oxygen in a measuring gas is proposed. In the method, an electronic control device (122) which has at least a first separate connection (P1) for a first pump cell (112) of a sensor element (110) of the sensor (100) and a second separate connection (P2) for a second pump cell (140) of the sensor element (110) of the sensor (100) is connected to the sensor element (110), wherein the first pump cell (112) is connected to the first separate connection (P1) by means of an electrically conductive connection (120), wherein the second pump cell (140) is connected to the second separate connection (P2) by means of an electrically conductive connection (146), wherein a measuring resistor (160) is provided in the electrically conductive connection (146) which connects the second pump cell (140) to the second separate connection (P2), wherein a current and/or voltage excitation of the second pump cell (140) for Generating a measurement signal (UP2p-P2n) at the measuring resistor (160), wherein the electrically conductive connection (146) connecting the second pump cell (140) to the second separate terminal (P2) is identified as intact if an amount of a temporal change in the measurement signal (UP2p-P2n) during the current and/or voltage excitation exceeds a predetermined threshold value (212), and is identified as defective if the amount of the temporal change in the measurement signal (UP2p-P2n) during the current and/or voltage excitation falls below the predetermined threshold value (212).

Description

State of the art

A large number of methods and sensors are known from the prior art for detecting at least a portion of the measuring gas component with bound oxygen in a gas mixture, in particular in an exhaust gas of an internal combustion engine, by detecting a portion of oxygen which is produced by a reduction of the measuring gas component with the bound oxygen.

Sensors for detecting at least a portion of the measuring gas component with bound oxygen in a gas mixture, which are also abbreviated or simplified to NO _x sensors or nitrogen oxide sensors, are used, for example, in Reif, K., Deitsche, KH. et al., Automotive Engineering Handbook, Springer Vieweg, Wiesbaden, 2014, pages 1338-1347 described.

Nitrogen oxide sensors (= NO _x sensors) that are used in automotive technology today work according to the limiting current principle, analogous to oxygen sensors such as lambda sensors. Such a nitrogen oxide sensor comprises a Nernst concentration cell, which is also called a reference cell, a modified oxygen pump cell and another modified oxygen pump cell, the so-called NO _x cell. An outer pump electrode, which is alternatively either exposed to the exhaust gas or has contact with the ambient air via an air duct, and an inner pump electrode in a first cavity that is separated from the exhaust gas by a diffusion barrier, form the oxygen pump cell. The Nernst electrode is also located in the first cavity and the reference electrode is located in a reference gas chamber, which together form the Nernst cell or reference cell. The NO _x cell comprises a NO _x pump electrode and a counter electrode. The NO _x pump electrode is located in a second cavity that is connected to the first inner cavity and separated from it by a diffusion barrier. The counter electrode is located in the reference gas chamber. All electrodes in the first and second cavities have a common return conductor.

When the nitrogen oxide sensor is in operation, the oxygen is removed from the first cavity of the so-called O2 cell, which is connected to the exhaust gas via a diffusion barrier. The resulting pumping current is then proportional to the oxygen content of the ambient air in the measuring gas or exhaust gas stream. The nitrogen oxides are pumped out in the NO _x cell. The nitrogen oxide NO _x in the atmosphere in the second cavity is reduced or broken down by applying a constant pumping voltage. The oxygen generated by the reduction or breakdown of the measuring gas component in the second cavity, which preferably comes from the reduction of the nitrogen oxide NO _x , is pumped out into a reference gas chamber. The applied pumping voltage against the resistance of the NO _x cell and the concentration of the nitrogen oxide NO _x or oxygen results in a pumping current that is proportional to the content of nitrogen oxide NO _x or oxygen and represents the NO _x measurement signal.

The resulting pump current I _P2 is therefore a measure of the NO _x concentration of the ambient air in the measuring gas or exhaust gas flow. With this arrangement, it is important that the nitrogen oxides are not pumped out at the oxygen cell, as otherwise no signal could be measured at the NO _x cell. This is achieved by gold doping the O2 cell. In addition, the O2 cell may only be operated at low pump voltages, as otherwise NO _x molecules would dissociate again.

Despite the advantages of the sensors and methods for operating them known from the prior art, they still contain potential for improvement. The temperature of the sensor element is controlled by pulse width modulation (PWM) of the heater supply (voltage, current). The PWM voltage is taken directly from the supply voltage (typically 13.5 V) of the sensor control unit (SCU) via a field effect transistor (FET). As a result, in the on phase of the PWM, the SCU supply voltage of the system is applied to the heating meander of the sensor element in the sensor sample. The current of NOx measuring signal is very small, for example. 4.5 µA at 1500 ppm NOx and therefore extremely sensitive to interference and coupling. Due to the structural proximity of the heating meanders to the NOx cell, a current is applied to the line between the P2 pin on the control unit (NOx measurement and control circuit) and the P2 pin during the on phase of the PWM signal due to capacitive coupling and leakage currents the NOx cell or the counter electrode of the NOx cell. This disturbance makes an offset to the actual NOx value measurable.

The requirements of the environmental authorities require a continuous and reliable diagnosis of line breaks in the line between the P2 pin on the control unit (NOx measuring and control circuit) and the P2 pin of the NOx cell or the counter electrode of the NOx cell (hereinafter also P2 Line). Monitoring of the lines of exhaust gas-relevant functions for interruptions (open circuit) must be carried out continuously, i.e. with no less than two sample values per second. So have to stick Due to these regulatory requirements, oxide sensors (NOx sensors) have a diagnostic function that can detect a line interruption (open circuit) in the conductive connections between the control unit and sensor element as well as the heater lines. This diagnosis must be carried out continuously during measurement operation.

Disclosure of the invention

A method for operating a sensor for detecting at least a portion of a measurement gas component with bound oxygen in a measurement gas is therefore proposed, which at least largely avoids the disadvantages of known methods for operating these sensors and which allows reliable and continuous diagnosis at intervals of at least 500 ms . In particular, a robust and continuous diagnosis of the lines and electrical wiring of the NOx cell of the sensor should be able to be carried out during measuring operation without influencing or disturbing the NOx measured values. If a changing NOx signal is present, it can be assumed in the system that the electrical connections to the NOx cell are intact. However, it is particularly important that a line interruption can be detected even in operating conditions in which the NOx measurement signal is (almost) equal to zero (e.g. at 0 ppm NOx after exhaust gas aftertreatment).

In a method according to the invention for operating a sensor for detecting at least a portion of a measurement gas component with bound oxygen in a measurement gas, in particular in an exhaust gas of an internal combustion engine, wherein the sensor comprises a sensor element, wherein the sensor element has a first pump cell, which has an outer pump electrode and an inner pump electrode and which is adjacent to a first cavity which is connected to the measurement gas, a reference cell, which has a Nernst electrode and a reference electrode and which is adjacent to a reference gas space, and a second pump cell, which has a pump electrode and a counter electrode and which is adjacent to a second cavity, an electronic control device, which has at least a first separate connection for the first pump cell and a second separate connection for the second pump cell, is connected to the sensor element, wherein the first pump cell is connected to the first separate connection by means of an electrically conductive connection, wherein the second pump cell is connected to the second separate connection by means of an electrically conductive connection, wherein a measuring resistor is provided in the electrically conductive connection which connects the second pump cell to the second separate connection, wherein by means of the Control unit, a current and/or voltage excitation of the second pump cell is carried out to generate a measurement signal at the measuring resistor. The electrically conductive connection that connects the second pump cell or its counter electrode to the second separate connection (P2 pin of the control unit) is identified as intact if the amount of a temporal change in the measurement signal while an excitation signal is applied to the second pump cell exceeds a predetermined threshold value, and is identified as defective if the amount of the temporal change in the measurement signal while an excitation signal is applied to the second pump cell falls below the predetermined threshold value.

The current and/or voltage excitation of the second pump cell generates an evaluable measurement signal at the measuring resistor that allows the distinction between an open or closed circuit. In other words, an IP2 signal is generated by an external excitation. It is not the amplitude of the voltage or current response measured (differentially at the measuring resistor) that serves as the detection criterion for a line break, but rather its change over time. Since the measured values of the amplitude of the response signal to a voltage jump are strongly influenced by component variations and interference in the NOx signal, they are not so suitable as a detection criterion. The temporal change of the differentially measured response signal to a voltage jump or voltage pulse is much more robust against component variations or interference in the NOx signal as a detection criterion for a line break. This avoids false positive detections of a line break. This means that an open circuit on the P2 line of the NOx cell can be reliably detected even during measurement operation. Even in operating conditions where the current IP2 is (almost) zero, such as at 0 ppm NOx, an open P2 line can be detected.

In a further development, a predetermined electrical voltage is applied to the second pump cell, with a voltage excitation of the second pump cell being carried out, the voltage excitation comprising a change in the predetermined electrical voltage for a predetermined time. By stimulating the voltage compared to the normal potential, a current flow occurs when the line is closed, which adds up to the NOx measuring current. This means that a measurable current flow occurs even with a gas mixture in which the IP2 signal is zero, for example with 0% O ₂ , H ₂ O ≥1% and NOx = 0 ppm. If the P2 line is disconnected, no more power can flow flow, even when stimulated over time by a pulse. This is an indication of an open circuit.

The measured current flow depends on the magnitude of the voltage change of the excitation pulse and, in particular, on the duration of the excitation pulse. This means that the current flow increases with the duration of the excitation.

In a further development, the predetermined electrical voltage is increased for the predetermined time. This also increases the measurable current flow.

In a further development, the predetermined electrical voltage is applied to the second pump cell in the form of at least one voltage pulse for the predetermined time, the temporal change in the measurement signal being determined by calculating a difference between a first measured value of the measurement signal at the beginning of the voltage pulse and a second measured value of the measurement signal End of the voltage pulse is determined. The expression “beginning of the voltage pulse” does not refer to a point in time when the voltage pulse is excited, but rather refers to a point in time from which the measurement signal has settled. The temporal change in the response signal to a voltage pulse is thus determined. The voltage value is measured at predetermined times at the beginning of the pulse response and shortly before its end. A difference signal is formed between these two voltage values. This difference signal serves as a detection criterion.

In a further development, the predetermined time for the voltage pulse has a duration of 1 ms to 10 ms and preferably 1.5 ms to 5 ms. The current response within the first milliseconds of a voltage pulse can thus provide a clear criterion for whether the conductive connection to the sensor element is interrupted. For this reason, a short voltage pulse of, for example, 2 ms to 5 ms in length is sufficient to detect a line interruption. A short duration of these diagnostic signals is also important because monitoring the lines for a malfunction should not interrupt the measurement of the NOx signal (current IP2) for too long or interfere too much.

In a further development, the predetermined electrical voltage is applied to the second pump cell in the form of a first voltage pulse for a first predetermined time and a second voltage pulse for a second predetermined time, the first voltage pulse and the second voltage pulse having an opposite polarity, an integral of the first voltage pulse and the second voltage pulse has a value of zero. Since individual voltage pulses disrupt and falsify the sensitive NOx measurement signal due to the charge shifts on the NOx electrode, pairs of voltage pulses are applied to the line. The first pulse and the subsequent counterpulse have the same duration and the same amount of amplitude, but a polarity with opposite signs. The integral of the two positive and negative pulses should be equal to zero.

In a further development, several pulse pairs with the first voltage pulse and the second voltage pulse are applied to the second pump cell, with the polarities of successive voltage pulse pairs alternating regularly or irregularly. If pulses are applied periodically to a regulated electrical system, there is a possibility that the system will be disturbed or periodically excited. For the electrical circuit of the NOx cell, this means that the pulse pairs for detecting the conductive connection to the NOx cell can possibly cause a shift in the NOx measurement signal. If the resonance frequency of the voltage control circuit on the NOx cell coincides with an excitation frequency, oscillating disturbances in the measurement signal can occur. In order to compensate for charge shifts on the NOx electrodes, the first measurement pulse is followed by a counter pulse of the same size with reversed voltage polarity. Due to the electrical properties of the sensor ceramic and the electrical circuit, small charge shifts can occur after the measurement pulses, even though the pulse and counter pulse have the same duration and the same amplitude. Depending on whether the first pulse has positive or negative polarity, there may be a slight positive or negative offset in the measurement signal. To avoid a shift in the measurement signal, the polarity of the pulse sequence can be changed so that the charge shifts balance out again.

In a further development, the number of positive polarities of the first voltage pulse of the voltage pulse pairs and the number of negative polarities of the first voltage pulse of the voltage pulse pairs are identical. This reliably prevents a shift in the measurement signal. The irregular sequence of pulse changes prevents the detection signal from developing a frequency that coincides with the resonance frequency of the NOx cell. In addition to a sequence of pulse pairs, individual pulses can also be omitted in the sequence to further improve the spectral characteristics.

Alternatively or additionally, the method may further include checking the plausibility of the measurement signal if the electrically conductive connection that connects the second pump cell to the second separate connection is identified as defective. To avoid a false-positive detection of a line interruption error, a two-stage detection procedure can be used. The monitoring pulses continuously monitor the state of the line at 500 ms intervals. If the monitoring pulses detect a possible interruption of the electrically conductive connection to the sensor element, this is checked again with a determination pulse.

In a further development, to check the plausibility, the predetermined electrical voltage is applied to the second pump cell in the form of at least one detection voltage pulse for a predetermined detection time, the temporal change in the measurement signal being determined by calculating a difference between a first measured value of the measurement signal at the beginning of the voltage pulse and a second Measured value of the measurement signal is determined at the end of the detection voltage pulse, the detection time for the detection voltage pulse being longer than the predetermined time for the voltage pulse. An amplitude of the detection voltage pulse is preferably the same size as an amplitude of the voltage pulse. Detection with this detection pulse has a minimal probability of error. A long-lasting pulse (typically 20 ms - 40 ms) compared to the monitoring pulse (typically 2 ms - 5 ms) is given to the line. The voltage change measured at the shunt, the difference signal, shows voltage values of significantly more than 100 mV when the line is closed, while values close to zero are measured when the line is interrupted. Due to this large detection signal compared to the monitoring pulse, incorrect detection is very unlikely.

In a further development, the control device is designed to regulate a voltage on the second pump cell, the control device further having an excitation signal source, a voltage setpoint being controlled as a reference variable, the control device controlling a time sequence of changes in the voltage setpoint, using measurements of a voltage drop on the Measuring resistor, an interruption in the line that connects the second pump cell to the second separate connection can be detected.

In a further development, the control device also has an excitation signal source and an operational amplifier, which is used as a non-inverting stage. The excitation signal source is connected as an input voltage to the non-inverting input (+) of the operational amplifier. The voltage present at the NOx cell is fed back to the inverting input (-) of the operational amplifier. The differential input signals of the operational amplifier and its gain factor determine the voltage at the output of the operational amplifier to which the NOx shunt is applied and which therefore also drives the current through the NOx shunt. The operational amplifier of the control unit electronics, which is connected as a voltage follower, regulates the voltage on the NOx cell of the sensor element to the setpoint of 450 mV. A voltage jump is applied to the NOx cell by increasing the setpoint of the operational amplifier connected as a voltage follower. The voltage drop across the NOx shunt cannot be measured directly by the ASIC. Instead, the voltage at the input and output of the operational amplifier is measured differentially. These are also the measuring points of the analog-to-digital converter in the NOx measuring ASIC. If there is a voltage jump at the input of the operational amplifier, these signals differ from those of the voltage drop at the measurement shunt. When the line is closed, the voltage increases steadily. When the line is open, this voltage drops steadily after a short increase.

In a further development, a predetermined electrical current is impressed into the second pump cell, with a current excitation of the second pump cell being carried out, the predetermined electrical current being increased for a first predetermined time and the predetermined electrical current being reduced for a second predetermined time, the first predetermined time and the second predetermined time are identical in length. An alternative to the voltage pulse is a current pulse on the P2 line. In the measuring state of the NOx sensor, a controlled current source generates a current pulse for a short time, increases it by a certain value and then generates a current pulse with the opposite polarity for the same time. It doesn't matter whether the positive current pulse or the negative current pulse is carried out first. By impressing a current pulse (pump current), a voltage swing at the NOx shunt will be measurable. This behaves differently when the circuit is closed than when it is open. This can be used as a distinguishing feature for the detection of an open circuit on the P2 line. The temporal change in the measurement signal is also taken into account for the diagnosis.

In a further development, the predetermined electrical voltage is applied to the second pump cell in the form of a first voltage pulse for a first predetermined time and a second voltage pulse for a second predetermined time. The integral over the time of both pulses must be zero in order to avoid an imbalance in the NOx Cell by pumping up or pumping down on one side. By impressing a current pulse (pumping current), a voltage swing on the P2 line can be measured. This behaves differently in a closed circuit than in an open circuit. The voltage swing is higher in a closed circuit than in an open circuit. This can be used as a distinguishing feature for the detection of an open circuit on the P2 line.

Furthermore, a computer program is proposed which is configured to carry out each step of the method according to the invention.

Furthermore, an electronic storage medium is proposed on which a computer program for carrying out the method according to the invention is stored.

The invention also includes an electronic control device which contains the electronic storage medium according to the invention with said computer program for carrying out the method according to the invention.

Finally, the invention also relates to a sensor for detecting at least a proportion of a measurement gas component with bound oxygen in a measurement gas, in particular in an exhaust gas of an internal combustion engine, comprising a sensor element, wherein the sensor element has a first pump cell, which has an outer pump electrode and an inner pump electrode and the is applied to a first cavity, which is connected to the measurement gas, a reference cell, which has a Nernst electrode and a reference electrode and which is applied to a reference gas space, and a second pump cell, which has a pump electrode and a counter electrode and which is applied to a second Cavity is present, the sensor further having an electronic control device according to the invention.

In the context of the present invention, a solid electrolyte is understood to mean a body or object with electrolytic properties, i.e. with ion-conducting properties. In particular, it can be a ceramic solid electrolyte. This also includes the raw material of a solid electrolyte and therefore the formation as a so-called green or brown compact, which only becomes a solid electrolyte after sintering. In particular, the solid electrolyte can be formed as a solid electrolyte layer or from several solid electrolyte layers. In the context of the present invention, a layer is understood to mean a uniform mass in a flat extension of a certain height, which lies above, below or between other elements.

In the context of the present invention, an electrode is generally understood to mean an element which is able to contact the solid electrolyte in such a way that a current can be maintained through the solid electrolyte and the electrode. Accordingly, the electrode can comprise an element on which the ions can be incorporated into the solid electrolyte and/or removed from the solid electrolyte. Typically, the electrodes comprise a noble metal electrode, which can be applied to the solid electrolyte, for example as a metal-ceramic electrode, or can be connected to the solid electrolyte in some other way. Typical electrode materials are platinum cermet electrodes. However, other precious metals, such as gold or palladium, can also generally be used.

In the context of the present invention, a heating element is to be understood as meaning an element which serves to heat the solid electrolyte and the electrodes to at least their functional temperature and preferably to their operating temperature. The functional temperature is the temperature at which the solid electrolyte becomes conductive to ions and is approximately 350 °C. A distinction must be made between this and the operating temperature, which is the temperature at which the sensor element is usually operated and which is higher than the functional temperature. The operating temperature can be, for example, from 700 °C to 950 °C. The heating element can include a heating area and at least one supply path. In the context of the present invention, a heating area is to be understood as meaning the area of the heating element which overlaps with an electrode in the layer structure along a direction perpendicular to the surface of the sensor element. The heating area usually heats up more than the supply line during operation, so that they can be distinguished. The different heating can be achieved, for example, by the heating area having a higher electrical resistance than the supply line. The heating area and/or the supply line are designed, for example, as an electrical resistance track and heat up when an electrical voltage is applied. The heating element can be made, for example, from a platinum cermet. The invention can be demonstrated directly by a shortened waiting time until the sensor is ready for operation after starting. The respective potentials can be measured on the supply lines.

In the context of the present invention, voltage excitation of the second pump cell involves the application of an electrical voltage not equal to 0 V to the second pump cell or the electrical line to the NOx counter electrode. Changing the voltage across the electrodes changes the amount of charge they can carry. A current flows into or out of the electrodes. This current flow causes a detectable measurement signal at a measuring resistor in this line.

In the context of the present invention, current excitation of the second pump cell means the imprinting of an electrical current not equal to 0 A into the second pump cell or the electrical line to the NOx counterelectrode. By impressing a current pulse (pump current), a voltage swing is caused on the electrical line to the NOx counter electrode, which can be measured at a measuring resistor in this line. This behaves differently when the circuit is closed than when it is open.

The invention can be easily and simply demonstrated by monitoring the electrical signals on the NOx cell line. If, during normal measuring operation, specific voltage pulse/current pulse sequences are measured on the line between the sensor and the control unit using an oscilloscope, the circuits and methods described in this invention are used.

Brief description of the drawings

Further optional details and features of the invention result from the following description of preferred exemplary embodiments, which are shown schematically in the figures.

• 1 einen prinzipiellen Aufbau eines erfindungsgemäßen Sensors,
• 2 einen Teil des Sensors mit einem Teil eines daran angeschlossenen Steuergeräts,
• 3 ein vereinfachtes Blockschaltbild der erfindungsgemäßen Schaltung,
• 4 ein erweitertes Blockschaltbild der erfindungsgemäßen Schaltung,
• 5 ein vereinfachtes elektrisches Modell der zweiten Pumpzelle 140 und der Schaltung im Steuergerät,
• 6 einen Spannungssprung und Spannung am NOx-Mess-Shunt für eine geschlossene und unterbrochene Verbindung zur NOx-Zelle,
• 7 das Kurzzeitverhalten für den Spannungssprung und Spannung am NOx-Mess-Shunt für eine geschlossene und unterbrochene Verbindung zur NOx-Zelle,
• 8 einen Spannungssprung und eine differentiell gemessene Spannung an den Eingängen des NOx-Mess-ASIC für eine geschlossene und unterbrochene Verbindung zur NOx-Zelle,
• 9 das Kurzzeitverhalten für den Spannungssprung und eine differentiell gemessene Spannung an den Eingängen des NOx-Mess-ASIC für eine geschlossene und unterbrochene Verbindung zur NOx-Zelle,
• 10 einen Spannungssprung und eine differentiell gemessene Spannung an den Eingängen des NOx-Mess-ASIC für eine geschlossene und unterbrochene Verbindung zur NOx-Zelle,
• 11 das Kurzzeitverhalten für den Spannungssprung und eine differentiell gemessene Spannung an den Eingängen des NOx-Mess-ASIC für eine geschlossene und unterbrochene Verbindung zur NOx-Zelle,
• 12 verschiedene Spannungs-Messsignale gemessen mit einem Oszilloskop,
• 13 verschiedene Spannungs-Messsignale,
• 14 verschiedene Spannungs-Messsignale gemessen mit einem Oszilloskop,
• 15 interne Signale der Auswerteinheit im Steuergerät 122 innerhalb des Monitoring-Pulses,
• 16 das Differenzsignal der Auswerteinheit im Steuergerät 122 innerhalb des Monitoring-Pulses,
• 17 verschiedene Spannungs-Messsignale gemessen mit einem Oszilloskop,
• 18 eine beispielhafte Definition der Abfolge der wechselnden Polarität der Pulspaare,
• 19 eine beispielhafte die spektrale Verteilung des NOx Mess-Signals,
• 20 an Umgebungsluft Häufigkeitsverteilungen eines NOx-Signals,
• 21 die Änderung der Steigung der Pulsantwort unter Annahme der größten spezifizierten Drift dieses Bauteils,
• 22 das Differenzsignal für typische Bauteilewerte und für den Fall der Kombination von Worst-Case-Werten,
• 23 interne Signale der Auswerteinheit im Steuergerät innerhalb des Monitoring-Pulses,
• 24 Verteilungsdichtefunktionen des Differenzsignals und des NOx Mess-Signals,
• 25 ein Flussdiagramm des kontinuierlichen Monitorings der leitenden Verbindung zur NOx-Zelle,
• 26 die Abfolge der Anregungspulse und der Pulsantwort des differentiell gemessenen Spannungsabfalls am Shunt bei einer falsch positiven Leitungsunterbrechungsdetektion mit Monitoring-Pulsen,
• 27 die Abfolge der Anregungspulse und der Pulsantwort des differentiell gemessenen Spannungsabfalls am Shunt bei einer falsch positiven Leitungsunterbrechungsdetektion mit Monitoring-Pulsen,
• 28 verschiedene Spannungs-Messsignale des Feststellungs-Pulses gemessen mit einem Oszilloskop, und
• 29 das Differenzsignal der Auswerteinheit im Steuergerät innerhalb des Feststellung-Pulses.

Show it:

• 1 a basic structure of a sensor according to the invention,
• 2 a part of the sensor with a part of a control device connected to it,
• 3 a simplified block diagram of the circuit according to the invention,
• 4 an extended block diagram of the circuit according to the invention,
• 5 a simplified electrical model of the second pump cell 140 and the circuit in the control unit,
• 6 a voltage jump and voltage at the NOx measuring shunt for a closed and interrupted connection to the NOx cell,
• 7 the short-term behavior for the voltage jump and voltage at the NOx measuring shunt for a closed and interrupted connection to the NOx cell,
• 8th a voltage jump and a differentially measured voltage at the inputs of the NOx measurement ASIC for a closed and interrupted connection to the NOx cell,
• 9 the short-term behavior for the voltage jump and a differentially measured voltage at the inputs of the NOx measuring ASIC for a closed and interrupted connection to the NOx cell,
• 10 a voltage jump and a differentially measured voltage at the inputs of the NOx measurement ASIC for a closed and interrupted connection to the NOx cell,
• 11 the short-term behavior for the voltage jump and a differentially measured voltage at the inputs of the NOx measuring ASIC for a closed and interrupted connection to the NOx cell,
• 12 various voltage measurement signals measured with an oscilloscope,
• 13 various voltage measurement signals,
• 14 various voltage measurement signals measured with an oscilloscope,
• 15 internal signals from the evaluation unit in the control unit 122 within the monitoring pulse,
• 16 the difference signal from the evaluation unit in the control device 122 within the monitoring pulse,
• 17 various voltage measurement signals measured with an oscilloscope,
• 18 an exemplary definition of the sequence of alternating polarity of the pulse pairs,
• 19 an example of the spectral distribution of the NOx measurement signal,
• 20 frequency distributions of a NOx signal in ambient air,
• 21 the change in the slope of the pulse response assuming the largest specified drift of this component,
• 22 the difference signal for typical component values and for the case of a combination of worst-case values,
• 23 internal signals from the evaluation unit in the control unit within the monitoring pulse,
• 24 Distribution density functions of the difference signal and the NOx measurement signal,
• 25 a flowchart of the continuous monitoring of the conductive connection to the NOx cell,
• 26 the sequence of excitation pulses and the pulse response of the differentially measured voltage drop at the shunt in the event of a false positive line break detection with monitoring pulses,
• 27 the sequence of excitation pulses and the pulse response of the differentially measured voltage drop at the shunt in the event of a false positive line break detection with monitoring pulses,
• 28 various voltage measurement signals of the detection pulse measured with an oscilloscope, and
• 29 the difference signal from the evaluation unit in the control unit within the detection pulse.

Embodiments of the invention

1 shows a basic structure of a sensor 100 according to the invention, which is particularly suitable for carrying out the method according to the invention. In the following explanations, identical or comparable components or features are provided with the same reference numerals.

The sensor 100, which is designed to detect at least a portion of a measurement gas component with bound oxygen, hereinafter referred to as nitrogen oxide NOx, in a gas mixture, for example an exhaust gas from an internal combustion engine, comprises a sensor element 110 and a first pump cell 112, which is formed between an outer pump electrode 114 and an inner pump electrode 116. The outer pump electrode 114, which is separated from the environment of the sensor 100 by means of a porous aluminum oxide layer 118, has a first electrically conductive connection 120, via which a first pump current I _P1 can be generated in the first pump cell 112. The first electrically conductive connection 120 is connected to a first connection P1 of an external electronic control unit 122. In order to obtain a complete circuit, the inner pump electrode 116 also has a second electrically conductive connection 124, which leads to a common connection COM of the external electronic control unit 122. The first pump cell 112 is located at a first cavity 126, which is located inside the sensor element 110 and is connected to the measuring gas. By generating the first pump current I _P1 in the first pump cell 112, a first portion of oxygen ions, which are formed from molecular oxygen from the gas mixture, can be transported between the first cavity 126 and the environment of the sensor 100. Two diffusion barriers 128 are present in the entry path from the environment to the first cavity 126.

The sensor element 110 also has an electrical reference cell 130, which has a Nernst electrode 132 and a reference electrode 134. While the Nernst electrode 132 has the second electrically conductive connection 124 together with the inner pump electrode 116 to the common connection COM, the reference electrode 134 has a separate third electrically conductive connection 136 to a supply voltage U _Vs , which has a connection Vs des external electronic control device 122 provides the required supply voltage Vs. The reference cell 130 lies against a reference gas space 138. A second portion of the oxygen ions from the first cavity 126 and/or from the environment of the sensor 100 are transported into the reference gas space 138 by applying a reference pumping current between the connection Vs and the common connection COM. Here, the value for the reference pump current is set such that a fixed proportion of the oxygen ions are formed in the reference gas space 138. In this context, the value for the first pump current I _P1 is preferably also set such that a fixed ratio results between the first proportion of oxygen ions in the first cavity 126 and the second proportion of oxygen ions in the reference gas space 138.

The measurement gas component nitrogen oxide NO _x with the bound oxygen still contained in the gas mixture reaches a second pump cell 140 of the sensor element 110, which can also be referred to as a “NO _x pump cell,” largely unaffected, in particular by diffusion. The second pump cell 140 has a NO _x pump electrode 142 and a NO _x counter electrode 144 and is located on a second cavity 145 inside the sensor element 110. The second cavity 145 is separated from the first cavity 126 by one of the diffusion barriers 128. At least one of the two electrodes, NO _x pump electrode 142 and/or NO _x counter electrode 144, are designed such that when a voltage is applied, additional molecular oxygen can be generated from the measurement gas component NO _x by means of catalysis, which is formed in the second pump cell 140.

While the NO _x pump electrode 142 has an electrically conductive connection which leads to the common terminal COM, the NO _x counter electrode 144 has a fourth electrically conductive The fourth electrically conductive connection 146 has a second pumping current I _P2 that can be applied to the second pumping cell 140. The fourth electrically conductive connection 146 is connected to a second connection P2 of the external electronic control unit 122. When a second pumping current I _P2 is applied to the second pumping cell 140, a portion of additional oxygen ions that were formed from the additional molecular oxygen is transported into the reference gas space 138.

The sensor element 110 also has a heating element 148, which has a heating line 150 with the lines HTR+ and HTR-, via which a heating current can be introduced into the heating element 148, which can bring the sensor element 110 to the desired temperature by generating a heating power.

The reference electrode 134 is connected to the connection Vs on the control unit 122 via the third electrically conductive connection 136. The Nernst voltage Uvs between Vs and COM can be measured via an analog-digital converter 152. The control unit 122 has an analog-digital converter 152 which is connected to the reference electrode 134 via the connection Vs for the supply voltage U _Vs and the third electrically conductive connection 136.

The control device 122 has a COM voltage reference 154, which is connected to a common return conductor 155 of the sensor element 110. The common return conductor 124 connects the NOx pump electrode 142, the Nernst electrode 132 and the inner pump electrode 116 to COM. The control device 122 also has an excitation signal source 156, which is connected to the setpoint input of the voltage regulator 166. The controlled variable input of the voltage regulator 166 is connected to the connection P2 on the NOx counter electrode 144 of the sensor element 110. The controlled variable is the voltage between P2 and COM. The manipulated variable of the voltage regulator 166 controls the input of a voltage driver 168. The measuring resistor 160 is arranged between the output of the voltage driver and the pin or connection P2.

In the method according to the invention for operating the sensor 100, a current and/or voltage excitation of the second pump cell 140 is carried out by means of the control unit 122 in order to generate a measurement signal at the measuring resistor 160. The electrically conductive connection 146, which connects the second pump cell 140 to the second separate connection P2, is identified as intact if a temporal change in the measurement signal exceeds a predetermined threshold value, and is identified as defective if the temporal change in the measurement signal falls below the predetermined threshold value. A predetermined electrical voltage U _P2 is thus applied to the second pump cell 140, wherein a voltage excitation of the second pump cell 140 is carried out. The voltage excitation comprises a change in the predetermined electrical voltage U _P2 for a predetermined time. The predetermined electrical voltage U _P2 can be increased for the predetermined time.

The predetermined electrical voltage U _P2 is preferably applied to the second pump cell 140 in the form of at least one voltage pulse for the predetermined time. The temporal change in the measurement signal is determined by calculating a difference between a first measured value of the measurement signal at the beginning of the voltage pulse and a second measured value of the measurement signal at the end of the voltage pulse. The predetermined time for the voltage pulse has a duration of 1 ms to 10 ms and preferably 1.5 ms to 5 ms, for example 2 ms to 5 ms.

The measuring signal comprises an electrical voltage U _P2P-P2N in the form of a difference between an electrical voltage U _P2P at the output P2P of the voltage driver 168 and an electrical voltage UP2 at pin P2 on the other side of the measuring shunt 160. This voltage is also present as a controlled variable at the input of the voltage regulator 166.

3 shows a simplified block diagram of the circuit according to the invention for detecting an interruption of the conductive connection to the NOx cell or second pump cell 140. Shown is an example of an interruption 161 of the fourth electrically conductive connection 146 between the NOx control unit 122 and the NOx counter electrode 144. The fourth electrically conductive connection 146 is also referred to as the P2 line because it connects the P2 pin of the control unit 122 to the NOx cell or the counter electrode 144. The block 162 represents the control of the setpoint or the command variable. The block 164 represents the voltage setpoint, which is fed to a block 166 for the voltage regulation, the control variable of which is in turn fed to a block 168 for the voltage driver. A measuring shunt 160 is connected between the output of the voltage driver 168 and the NOx cell 140. The voltage UP2 at the NOx cell 140 serves as a control variable for the controller 166. The signals P2N at the input of the controller 166 and P2P at the output of the operational amplifier 158 are fed to a differentially measuring analog-digital converter 170. The output of the analog-digital converter 170 represents the block 172 for evaluating the measured voltages. The evaluation The test results are fed to a line break detection logic block 174, which in turn communicates with block 162.

4 shows an extended block diagram of the circuit according to the invention for detecting an interruption in the conductive connection to the NOx cell or second pump cell 140. In contrast to 3 is in 4 a compensation circuit 176 is shown, which effects a direct coupling, for example a feedback capacitor 178, between the output P2P of the operational amplifier 158 and the input of the controlled variable on the controller 166. A feedback resistor 180 is located between the input P2N of the controller 166 and the point P2 on the NOx cell 140. In addition, there are ESD capacitors 181 on the control devices in the fourth line 146 between the measuring resistor 160 and the NOx counter electrode 144 of the sensor element 110 -page indicated. In the line between pin P2 on the control unit side and the counter electrode 144 on the side of the sensor element 110, line interruptions 161 can occur that should be monitored.

The electrical circuit for regulating the voltage at the second pump cell 140 is represented by the blocks of the voltage regulation 166, the voltage driver 168, the voltage setpoint 164, the measuring resistor 160 and the second pump cell 140. There is also block 162, in which the voltage setpoint is controlled as a reference variable. In this way, jumps or pulses of the reference variable can be impressed. The measurement of the NOx current is determined by the voltage drop between the input and the output of the voltage control. In the simplified representation, this is the voltage at the measuring resistor 160. The differentially measured voltages are converted into digital measured values using an analog-digital converter 170 and evaluated. The logic controls the timing of the changes in the voltage setpoint and the measurements of the voltage drop across the measuring resistor 160. With the help of these values, a line interruption can be detected.

5 shows a simplified electrical model of the second pump cell 140 and the circuit in the control unit 122 for investigating the electrical effects of line interruptions. The same components as in the 1 to 4 provided with the same reference numerals and shown as electrical circuit symbols. The functions of the blocks of the voltage regulator 166 and the voltage driver 168 are implemented by a non-inverting operational amplifier. The signal at the input P2N of the operational amplifier 158 is the controlled variable. Signals are tapped at the input P2N and output P2P of the operational amplifier 158 and fed to an analog-digital converter 170.

The method according to the invention makes use of the fact that the electrodes 142, 144 of the second pump cell 140 carry large amounts of charge and, in simplified terms, act like a large capacitor with a capacity of, for example, approximately 20 μP to 50 μF. The ESD capacitors 181 for electrical protection of the circuit on the control unit side, however, only have very small capacities. An important difference in the circuit is that there is the electrical compensation circuit 176 in the control loop, which consists of the feedback resistor 180 and the feedback capacitor 178. This means that there is a direct connection between the manipulated variable and the controlled variable in the control loop. The differential voltage measurement to determine the NOx current is therefore not carried out directly on the measuring resistor 160, but between the connection of the shunt 160 to the voltage driver output of the operational amplifier 158 and the connection of the feedback resistor 180 and the feedback capacitor 178 at the inverting input of the operational amplifier.

6 shows a voltage jump and voltage at the NOx measurement shunt 160 for a closed and broken connection to the NOx cell 140 according to a hardware simulation. 7 shows the short-term behavior for the voltage jump and voltage at the NOx measuring shunt 160 for a closed and interrupted connection to the NOx cell 140 according to a hardware simulation. The electrical circuit of the control unit 122 was simulated using a hardware simulation tool. The NOx cell 140 of the sensor element 11 is simplified by a large capacity, for example 30 µF. On the X axis 182 is in 6 the time is plotted in s and in 7 the time is plotted in ms. The voltage in mV is plotted on the Y axis 184. Curve 186 indicates the voltage U _P2n at the input of operational amplifier 158 when the electrical connection to NOx cell 140 is closed. Curve 188 indicates the voltage U _P2n at the input of operational amplifier 158 when the electrical connection to NOx cell 140 is open. The curve 190 indicates the voltage U _IP2 at the measuring resistor 160 when the electrical connection to the NOx cell 140 is closed. The curve 192 indicates the voltage U _IP2 at the measuring resistor 160 when the electrical connection to the NOx cell 140 is open.

The operational amplifier 158 of the control unit electronics, which is connected as a voltage follower, regulates the voltage at the NOx cell 140 of the sensor element 110 to the setpoint value of 450 mV. A voltage jump is applied to the NOx cell 140. by increasing the setpoint of the operational amplifier 158, which is connected as a voltage follower. The simulation data in 6 and 7 show the voltage jump U _P2n measured against COM at input P2N of operational amplifier 158, and the voltage drop U _IP2 measured differentially at measuring shunt 160. With a closed electrical connection to NOx cell 140, current I _P2 increases steadily over a longer period of time. With an open electrical connection to NOx cell 140, a charging current flows briefly, which quickly decays exponentially. With an open connection, the ESD capacitors 181 with a capacitance of, for example, 10 nF and 3.3 n F on the side of control unit 122 charge quickly until the new target voltage is applied to the capacitor. With a closed connection, larger currents must flow over a longer period of time, since with the large capacitance of NOx cell 140 (approx. 30 µP), significantly larger amounts of charge must flow until the voltage at NOx cell 140 is regulated.

8th shows a voltage jump and a differentially measured voltage at the inputs P2N and P2P of the NOx measuring ASIC, which differs from the differential voltage at the measuring shunt 160, for a closed and broken connection to the NOx cell 140 according to a hardware simulation. 9 shows the short-term behavior for the voltage jump and a differentially measured voltage at the inputs of the NOx measuring ASIC for a closed and interrupted connection to the NOx cell 140 according to a hardware simulation. On the X-axis 182, 8th the time in s and 9 the time is plotted in ms. The voltage is plotted in mV on the Y-axis 184. The curve 186 indicates the voltage U _P2n at the input of the operational amplifier 158 with a closed electrical connection to the NOx cell 140. The curve 188 indicates the voltage U _P2n at the input of the operational amplifier 158 with an open electrical connection to the NOx cell 140. The curve 194 indicates the differentially measured voltage U _P2P-P2N at the inputs of the operational amplifier 158 with a closed electrical connection to the NOx cell 140. The curve 196 indicates the differentially measured voltage U _P2P-P2N at the inputs of the operational amplifier 158 with an open electrical connection to the NOx cell 140.

The voltage drop across the NOx shunt 160 can, as shown in the circuit model in 5 shows, cannot be measured directly by the ASIC. Instead, the voltage U _P2P-P2N is measured differentially at the input P2N and output P2P of the operational amplifier 158. These are also the measuring points of the analog-digital converter 170 in the NOx measuring ASIC. Like the simulation data in 8th and 9 show, these signals differ from those of the voltage drop at the measuring shunt 160. When the line is closed, the voltage U _P2P-P2N increases steadily. When the line is open, this voltage U _P2P-P2N drops steadily after a short increase.

10 shows a voltage jump and a differentially measured voltage at the inputs of the NOx measurement ASIC for a closed and interrupted connection to the NOx cell 140, measured experimentally with the hardware circuit of the sensor control unit 122. 11 shows the short-term behavior for the voltage jump and a differentially measured voltage at the inputs of the NOx measuring ASIC for a closed and interrupted connection to the NOx cell 140, measured experimentally with the hardware circuit of the sensor control unit. The measurements were carried out using an oscilloscope. On the X axis 182 is in 10 the time is plotted in ms and in 11 the time is plotted in s. The voltage in mV is plotted on the Y axis 184. The oscilloscope measurement curves in the 10 and 11 show the voltage jump measured at the ASIC pin UP2N against COM as well as the voltage measured differentially between the ASIC pins P2P and P2N. Curve 186 indicates the voltage U _P2n at the input of operational amplifier 158 when the electrical connection to NOx cell 140 is closed. Curve 188 indicates the voltage U _P2n at the input of operational amplifier 158 when the electrical connection to NOx cell 140 is open. Curve 194 indicates the differentially measured voltage U _P2P-P2N at the inputs of operational amplifier 158 when the electrical connection to NOx cell 140 is closed. Curve 196 indicates differentially measured voltage U _P2P-P2N at the inputs of operational amplifier 158 when the electrical connection to NOx cell 140 is open. In the first milliseconds after the start of the voltage jump, the same signal behavior is observed that was also calculated with the hardware simulation tool and in the 8th and 9 is shown.

The signal behavior examined and shown here is used in the context of the present invention to derive a criterion for detecting an interruption of the conductive connection to the NOx cell 140. In this case, it is not the amplitude of the differentially measured voltage U _P2p-P2n that serves as the detection criterion for a line interruption, but rather its change over time. Since the measured values of the amplitude of the response signal U _P2p-P2n to a voltage jump are very strongly influenced by component variations and disturbances in the NOx signal, they are not suitable as a detection criterion in every case. The change over time of the differentially measured response signal U _P2p-P2n to a voltage jump or voltage pulse has As a detection criterion for a line break, it offers significantly better robustness against component variations or against interference with the NOx signal. This way, false positive detections of a line break can be avoided.

In particular, it is shown above that the current response within the first milliseconds of a voltage pulse can provide a clear criterion as to whether the conductive connection to the sensor element 110 is interrupted. For this reason, a short voltage pulse of approx. 2 ms to 5 ms in length is sufficient to detect a line interruption. A short duration of these diagnostic signals is also important because the monitoring of the lines for a malfunction should not interrupt the measurement of the NOx signal (current I _P2 ) for too long or disrupt it too much.

Since individual voltage pulses can disturb and distort the sensitive NOx measurement signal U _IP2 due to charge shifts on the NOx electrode, voltage pulse pairs are sent to line 146. The first pulse and the following counterpulse have the same duration and the same amplitude, but a polarity with opposite signs. The integral of the two positive and negative pulses should be zero.

12 shows various voltage measurement signals measured with an oscilloscope. On the X-axis 182 is 12 the time in ms. On the Y-axis 184 the voltage is plotted in mV. The upper part of 12 shows a voltage double pulse with opposite polarity measured at the ASIC pin P2N against COM. The measurement curves in the lower part of 12 show the voltage U _IP2 measured differentially at the shunt 160 with an oscilloscope and the voltage U _P2P-P2n measured differentially between the ASIC pins P2P and P2N. Curve 186 gives the voltage U _P2n at the inverting input of the operational amplifier 158 with a closed electrical connection to the NOx cell 140. Curve 188 gives the voltage U _P2n at the inverting input of the operational amplifier 158 with an open electrical connection to the NOx cell 140. The signals of curves 188 and 186 are almost identical. Curve 190 gives the voltage U _IP2 at the measuring resistor 160 with a closed electrical connection to the NOx cell 140. Curve 192 gives the voltage U _IP2 at the measuring resistor 160 with an open electrical connection to the NOx cell 140. Curve 194 indicates the differentially measured voltage U _P2p-P2n between the output of the operational amplifier and the inverting input of the operational amplifier 158 when the electrical connection to the NOx cell 140 is closed. Curve 196 indicates the differentially measured voltage U _P2p-P2n at the inputs of the operational amplifier 158 when the electrical connection to the NOx cell 140 is open. Both signals were measured both for the case of a conductive connection to the NOx cell 140 and for the case of a line break. When comparing the two differentially measured signals U _P2p-P2n, the influence of the feedback resistor 180 and the feedback capacitor 178 can be clearly seen. A direct measurement of the signal at the shunt 160 would be very complex for circuitry reasons. The amplitude of the differential signal U _P2p-P2n at the ASIC measuring input is significantly smaller than that of the signal at the measuring shunt 160. This measurement shows why the temporal change of the response signal provides a more robust criterion than the amplitude.

The previously described pulse pairs are also referred to as monitoring pulses in the context of the present invention, since they are used to monitor the conductive connection 146 to the NOx cell 140 at 500 ms intervals and to detect interruptions in the line 146. The temporal change in the response signal U _P2p-P2n to a voltage pulse is determined while the first pulse of the pulse pair is applied to the line 146. The voltage value U _P2p-P2n is measured at predetermined times at the beginning of the pulse response and shortly before its end.

13 shows various voltage measurement signals. On the X axis 182 is in 13 the time is plotted in ms. The voltage in mV is plotted on the Y axis 184. The upper part of 13 shows a voltage double pulse with opposite polarity measured at the ASIC pin P2N against COM. The measurement curve in the lower part of 13 shows the voltage U _P2p-P2n measured differentially between the ASIC pins P2P and P2N. Curve 186 indicates the voltage U _P2n at the input of operational amplifier 158 when the electrical connection to NOx cell 140 is closed. Curve 194 indicates the differentially measured voltage U _P2p-P2n at the inputs of operational amplifier 158 when the electrical connection to NOx cell 140 is closed.

13 shows the times of the two voltage measurements within the pulse response. A difference signal is formed between these two voltage values DU _P2P-P2n = U _P2p-P2n (2nd sample) - U _P2p-P2n (1st sample). This difference signal serves as a detection criterion. 2nd sample refers to the measurement shortly before the end of the pulse and 1st sample refers to the measurement at the beginning of the pulse.

14 shows various voltage measurement signals measured with an oscilloscope. On the X-axis 182 is 14 the time in ms. On the Y-axis 184 the voltage in mV. The upper part of 14 shows a voltage double pulse with opposite polarity measured at the ASIC pin P2N against COM. The measurement curves in the lower part of 14 show the voltage U _P2P-P2n measured differentially between the ASIC pins P2P and P2N with an oscilloscope. The curve 186 indicates the voltage U _P2n at the input of the operational amplifier 158 with a closed electrical connection to the NOx cell 140. The curve 188 indicates the voltage U _P2n at the inverting input of the operational amplifier 158 with an open electrical connection to the NOx cell 140. The curve 194 indicates the differentially measured voltage U _P2p-P2n between the output of the operational amplifier and the inverting input of the operational amplifier 158 with a closed electrical connection to the NOx cell 140. The curve 196 indicates the differentially measured voltage U _P2p-P2n between the output of the operational amplifier and the inverting input of the operational amplifier 158 with an open electrical connection to the NOx cell 140.

The oscilloscope measurement signals in 14 show the different U _P2p-P2n response signals when the connection to the NOx cell 140 is closed and when the line is interrupted. It is easy to see that when the connection is closed (conductive), the difference signal DU _P2p-P2n has a significantly larger value than when the line is interrupted.

15 shows internal signals from the evaluation unit in the control unit 122 within the monitoring pulse. On the X axis 182 is in 15 the time is plotted in s. The voltage in mV is plotted on the Y axis 184. The curve 198 in the upper part of 15 indicates sample values at the beginning of the monitoring pulse for U _P2p-P2n (1st sample). The curve 200 in the upper part of 15 indicates sample values towards the end of the monitoring pulse for U _P2p-P2n (2nd sample). Curve 202 indicates the NOx measurement signal U _IP2 between P2P and P2N, which is measured when no detection pulse is active. The curve 204 in the lower part of 15 indicates the difference signal DU _P2P-P2n = U _P2p-P2n (2nd sample) - U _P2p-P2n (1st sample). The difference signal DU _P2p-P2n is formed from the two sample values within the monitoring pulse, U _P2p-P2n (2nd sample) and U _P2p-P2n (1st sample). In the example in 15 If the connection to the NOx cell 140 is initially conductive, as indicated by the time area 206, then it is briefly interrupted, as indicated by the time area 208, and after a few seconds it is conductive (closed) again, as indicated by the time area 210 stated.

16 shows the difference signal DU _P2p-P2n of the evaluation unit in the control unit 122 within the monitoring pulse. On the X-axis 182, 16 the time in s. The voltage in mV is plotted on the Y-axis 184. The curve 204 indicates the difference signal DU _P2P-P2n = U _P2p-P2n (2nd sample) - U _P2p-P2n (1st sample). The difference signal DU _P2p-P2n , which is used to detect a line break, is in 16 thus shown again. Within the scope of the present invention, a threshold value 212 is defined in the form of a voltage threshold. If the difference signal DU _P2p-P2n is below the threshold value 212, an interruption of the conductive connection to the NOx cell 140 is detected. If the difference signal DU _P2p-P2n is above this threshold value 212, the connection is considered to be intact.

However, if pulses are periodically applied to a controlled electrical system, there is the possibility that the system will be disturbed or periodically excited. With regard to the electrical circuit of the NOx cell 140 described above, this means that the pulse pairs for detecting the conductive connection to the NOx cell 140 may cause a shift in the NOx measurement signal. If the resonance frequency of the control circuit of the voltage on the NOx cell 140 coincides with an excitation frequency, oscillating disturbances in the measurement signal may occur.

In order to compensate for charge shifts on the NOx electrodes 142, 144, the first measurement pulse is followed by a counterpulse of the same size with reversed voltage polarity. Due to the electrical properties of the sensor ceramic and the electrical circuitry, although the pulse and counter-pulse have the same duration and the same amplitude, small charge shifts can occur after the measurement pulses. Depending on whether the first pulse has positive or negative polarity, there may possibly be a small positive or negative offset of the measurement signal.

To avoid a shift in the measurement signal, the polarity of the pulse sequence can be changed so that the charge shifts balance out again.

17 shows various voltage measurement signals measured with an oscilloscope. On the X-axis 182 is 17 the time in ms. On the Y-axis 184 the voltage is plotted in mV. The upper part of 17 shows voltage double pulses with opposite polarity measured at ASIC pin P2N against COM. The measurement curves in the middle part of 17 show the voltage U _P2p-P2n measured differentially between the ASIC pins P2P and P2N with an oscilloscope. The lower part of 17 gives for the diff ference signal DU _P2p-P2n the voltage levels when the line is closed or broken. Curve 214 gives the voltage U _P2n - COM at the ASIC pin P2N against COM measured for double voltage pulses with opposite polarity. Curve 216 gives the differentially measured voltage U _P2p-P2n between the output of the operational amplifier and the inverting input of the operational amplifier 158. Curve 218 gives the level of the voltage difference signal DU _P2p-P2n when the line is closed for a pulse with positive polarity. Curve 220 gives the threshold value for positive polarity. Curve 222 gives the level of the voltage difference signal DU _P2P-P2n when the line is broken for a pulse with positive polarity. Curve 228 gives the level of the voltage difference signal DU _P2p-P2n when the line is closed for a pulse with negative polarity. Curve 226 indicates the threshold value for negative polarity. Curve 224 indicates the level of the voltage difference signal DU _P2P-P2n when the line is interrupted for a pulse with negative polarity. To avoid a possible offset of the NOx measurement signal, the polarity of the pulse pairs is changed so that the first pulse has a positive or negative polarity equally often.

To avoid that the signal for detecting an interruption of the conductive connection to the NOx cell 140 has a characteristic frequency that may coincide with the resonance frequency of the system, a technique is used that distributes the frequencies of the excitation by the monitoring pulses over the spectrum.

18 shows an example definition of the sequence of alternating polarity of the pulse pairs in a sequence of 16 pulse pairs. On the X-axis 182, 18 the time in ms. The voltage U _P2n-COM measured at the ASIC pin P2N against COM is plotted on the Y-axis 184. The curve 230 indicates the voltage pulses. In the 18 In the implementation described, the sequence of pulse pairs, the first pulse of which has positive or negative polarity, can be defined. In this example, a repeating sequence of 16 pulse pairs is defined. The irregular sequence of pulse changes can prevent the detection signal from forming a frequency that coincides with the resonance frequency of the NOx cell 140. In addition to a sequence of pulse pairs, individual pulses in the sequence can also be omitted in order to further improve the spectral characteristics.

19 shows an example of the spectral distribution of the NOx measurement signal for the cases where no monitoring pulses are applied, where pulses with the same polarity as the first pulse are applied, or where pulses with alternating polarity of the first pulse are applied. On the X-axis 182, 19 the frequency f is plotted in Hz. On the Y-axis 184, the spectral power density is plotted as the magnitude P1(f).

Curve 232 indicates the spectral distribution of the NOx measurement signal when no monitoring pulses are applied. Curve 234 indicates the spectral distribution of the NOx measurement signal when pulses with the same polarity as the first pulse are applied. Curve 236 indicates the spectral distribution of the NOx measurement signal when pulses with alternating polarity of the first pulse are applied. The spectral lines caused by the monitoring pulses that repeat at a frequency of 2 Hz are clearly visible. Possible shifts in the NOx signal can be determined by analyzing the distribution of the relative frequency of the NOx measurement values.

20 shows the frequency distribution of a NOx signal in ambient air. On the X-axis 182, 20 the NOx content is plotted in ppm. The relative frequency is plotted dimensionlessly on the Y-axis 184. Curve 238 indicates the frequency distribution of the NOx measurement signal for the case that no monitoring pulses are applied. Curve 240 indicates the frequency distribution of the NOx measurement signal for the case that pulses with the same polarity as the first pulse are applied. Curve 242 indicates the frequency distribution of the NOx measurement signal for the case that pulses with alternating polarity of the first pulse are applied. In particular, 20 the frequency distribution of a NOx signal where no detection pulses are applied. This distribution is compared with a NOx measurement signal where monitoring pulses are applied, where the first pulse of the pulse pair always has a positive polarity. Here, the mean value of the NOx measurement signal is shifted by approx. 0.5 ppm. However, if the monitoring pulses have a changing polarity of the first pulse (with an irregular sequence, see above), the mean value of the NOx measurement signal corresponds to that of a NOx measurement signal where no pulses are applied.

The detection method according to the invention, which uses the temporal change in the response signal U _P2p-P2n to a voltage pulse, is significantly more robust against changes and drifts in the electrical properties of the sensor element and the components on the control device side. The slope of the pulse response can be influenced in particular by changes in the feedback capacitor 178.

21 shows the change in the slope of the pulse response assuming the largest specified drift of this component (component scatter, temperature, aging). On the X-axis 182, 21 the time in ms. On the Y-axis 184 the voltage is plotted in mV. The upper part of 21 shows the voltage U _P2n measured at the ASIC pin P2N against COM. The measurement curves in the lower part of 21 show the voltage U _P2P-P2n measured differentially between the ASIC pins P2P and P2N. The curve 244 indicates the voltage U _P2n at the inverting input of the operational amplifier 158 with a closed electrical connection to the NOx cell 140 and typical components of the controller 122. The curve 246 indicates the voltage U _P2n at the inverting input of the operational amplifier 158 with an open electrical connection to the NOx cell 140 and typical components of the controller 122. The curve 248 indicates the voltage U _P2n at the inverting input of the operational amplifier 158 with a closed electrical connection to the NOx cell 140 and component drift. The curve 250 indicates the voltage U _P2n at the inverting input of the operational amplifier 158 with an open electrical connection to the NOx cell 140 and component drift. Curve 252 indicates the differentially measured voltage U _P2p-P2n that is measured between the output of the operational amplifier and the inverting input of the operational amplifier 158 with a closed electrical connection to the NOx cell 140 and typical components of the controller 122. Curve 254 indicates the differentially measured voltage U _P2p-P2n between the output of the operational amplifier and the inverting input of the operational amplifier 158 with an open electrical connection to the NOx cell 140 and typical components of the controller 122. Curve 256 indicates the differentially measured voltage U _P2p-P2n between the output of the operational amplifier and the inverting input of the operational amplifier 158 with a closed electrical connection to the NOx cell 140 and component drift. Curve 258 indicates differentially measured voltage U _P2p-P2n between the output of the operational amplifier and the inverting input of the operational amplifier 158 with an open electrical connection to the NOx cell 140 and component drift. If the worst-case component drifts and variations are combined, the distance between the voltage values of the difference signal DU _P2p-P2n , which is used to detect a line break, decreases.

22 shows the difference signal DU _P2P-P2n for typical component values and for the combination of worst-case values. On the X axis 182 is in 22 the time is plotted in ms. The voltage in mV is plotted on the Y axis 184. Curve 260 indicates the difference signal DU _P2P-P2n for typical component values. Curve 262 indicates the difference signal DU _P2P-P2n for the case of combining worst-case values. In particular shows 22 the difference signal DU _P2P-P2n for the case of a conductive connection, as indicated by a time range 264, and a broken connection, as indicated by a time range 266. However, the distance to the detection threshold 212 is still large enough to carry out robust monitoring. An important advantage of the detection method, which uses the temporal change in the response signal U _P2p-P2n to a voltage pulse, is its robustness against interference with the NOx signal.

23 shows internal signals from the evaluation unit in the control unit 122 within the monitoring pulse. On the X axis 182 is in 23 the time is plotted in s. The voltage in mV is plotted on the Y axis 184. The curve 268 in the upper part of 15 indicates sample values at the beginning of the monitoring pulse for U _P2p-P2r , (1st sample). The curve 270 in the upper part of 23 indicates sample values towards the end of the monitoring pulse for U _P2p-P2n (2nd sample). The curve 272 indicates the measurement signal U _IP2 at the measuring resistor 160. The curve 274 in the lower part of 23 indicates the difference signal DU _P2p-P2n = U _P2p-P2n (2nd sample) - U _P2p-P2n (1st sample). The difference signal DU _P2p-P2n is formed from the two sample values within the monitoring pulse, U _P2p-P2n (2nd sample) and U _P2p-P2n (1st sample). In the example in 23 the NOx signal 272 is severely disturbed. It can be clearly seen from the measurement curve 274 that the difference signal DU _P2p-P2n has significantly lower interference.

24 shows distribution density functions of the difference signal DU _P2p-P2n and the NOx measurement signal U _IP2 . On the X-axis 182 is 24 the voltage is plotted in mV. The relative frequency is plotted dimensionlessly on the Y-axis 184. Curve 276 indicates the difference signal DU _P2P-P2n . Curve 278 indicates the NOx measurement signal U _IP2 . The low disturbance of the difference signal DU _P2p-P2n compared to the NOx signal U _IP2 shows the probability densities of the signals in 24 .

An alternative or additional modification or addition to the method according to the invention is described below. In order to avoid a false-positive detection of a line interruption error, a two-stage detection method is used. The monitoring pulses are used to continuously monitor the state of the line at 500 ms intervals. If a possible interruption of the electrically conductive connection to the sensor element is detected with the monitoring pulses, this is checked again with a determination pulse.

25 shows a flowchart of the continuous monitoring of the conductive connection 146 to the NOx cell 140 (P2 Open Circuit Monitoring). In step S10, the above-described current and/or voltage excitation of the second pump cell takes place to generate a measurement signal. In particular, the voltage pulses described above are generated, for example with a frequency of 2 Hz. In step S12, the change over time in the measurement signal is evaluated. It is thus checked whether the temporal change in the measurement signal exceeds the predetermined threshold value 212 or not. This can be done for two consecutive voltage pulses. If the threshold value is exceeded during the determination in step S12, the method proceeds to step S14 and the electrically conductive connection 146 is identified as intact. The process then returns to step S10. If the threshold value is undershot during the determination in step S12, the method proceeds to step S16 and the electrically conductive connection 146 is provisionally identified as defective.

To check plausibility, at least one detection voltage pulse is applied to the second pump cell 140 for a predetermined detection time in step S16. The temporal change in the measurement signal is determined by calculating a difference between a first measured value of the measurement signal at the beginning of the voltage pulse and a second measured value of the measurement signal at the end of the detection voltage pulse. The detection time for the detection voltage pulse is longer than the predetermined time for the voltage pulse. In step S18, the temporal change in the measurement signal is evaluated. In this way, it is checked whether the temporal change in the measurement signal during the detection pulses exceeds the predetermined threshold value 212 or not. This can be done for any number of consecutive voltage pulses. If the threshold value is exceeded during the determination in step S20, the method proceeds to step S20 and the electrically conductive connection 146 is identified as intact. The process then returns to step S10. If the threshold value is undershot in the determination in step S16, the method proceeds to step S22 and the electrically conductive connection 146 is identified as defective. Optionally, debounce conditions can be set in step S24 and the method then returns to step S10.

The procedure with the optional plausibility check of the measurement signal is described in more detail below. Detection with this detection pulse has a minimal probability of error. A pulse of, for example, 20 ms - 40 ms, which is longer than the monitoring pulse with a duration of, for example, 2 ms - 5 ms, is sent to line 146. The voltage change measured at shunt 160, the difference signal DU _P2p-P2n , shows voltage values of significantly more than 100 mV when line 146 is closed, while values close to zero are measured when line 146 is interrupted. Due to this large detection signal compared to the monitoring pulse, false detection is very unlikely.

26 shows the sequence of the excitation pulses U _P2n and the pulse response of the differentially measured voltage drop U _P2p-P2n at the shunt 160 in the event that a false-positive line interruption is detected with the monitoring pulses. On the X axis 182 is in 26 the time is plotted in ms. The voltage in mV is plotted on the Y axis 184. Curve 280 indicates the excitation voltage U _P2n . The monitoring pulses are designated 282 and the detection pulse is designated 284. Curve 286 indicates the differentially measured voltage drop U _P2p-P2n at shunt 160. Periods of detection of a closed electrical connection to the NOx cell 140 are designated 288 and periods of an open electrical connection to the NOx cell 140 are designated 290. In the example in 26 the monitoring pulses 282 initially detect a false-positive line interruption. However, the response of the detection pulse 284, whose signal U _P2p-P2n increases steadily, shows that there is no line interruption. The temporal change in the measurement signal therefore exceeds the threshold value 212. The false-positive detection was detected and the NOx measurement operation can be continued.

27 shows the sequence of the excitation pulses U _P2n and the pulse response of the differentially measured voltage drop U _P2p-P2n at the shunt 160 in the event that a false-positive line interruption is detected with the monitoring pulses. On the X axis 182 is in 27 the time is plotted in ms. The voltage in mV is plotted on the Y axis 184. Curve 280 indicates the excitation voltage U _P2n . The monitoring pulses are designated 282 and the detection pulse is designated 284. Curve 286 indicates the differentially measured voltage drop U _P2p-P2n at shunt 160. Periods of detection of a closed electrical connection to the NOx cell 140 are designated 288 and periods of an open electrical connection to the NOx cell 140 are designated 290. In the example in 27 there is a line interruption. The detection pulse 284 shows a signal U _P2p-P2n near zero. The temporal change in the measurement signal falls below the threshold value 212. The error of the sensor 100 is confirmed and an interruption of the electrically conductive connection to the NOx sensor element 140 is diagnosed.

28 shows various voltage measurement signals of the detection pulse measured with an oscilloscope. On the X-axis 182 is 28 the time is plotted in ms. The voltage is plotted in mV on the Y-axis 184. The curve 292 indicates the voltage U _P2n at the input of the operational amplifier 158 when the electrical connection to the NOx cell 140 is closed. The curve 294 indicates the voltage U _P2n at the input of the operational amplifier 158 when the electrical connection to the NOx cell 140 is open. The curve 296 indicates the differentially measured voltage U _P2p-P2n between the output of the operational amplifier 158 and the inverting input of the operational amplifier 158 when the electrical connection to the NOx cell 140 is closed. The curve 298 indicates the differentially measured voltage U _P2p-P2n between the output of the operational amplifier and the inverting input of the operational amplifier 158 when the electrical connection to the NOx cell 140 is open. The oscilloscope measurement signals in 28 show a detection pulse with a 30 ms long excitation signal U _P2n . The excitation pulses U _P2n are designed as a pulse pair with opposite polarity, similar to the monitoring pulses. The response signals U _P2p-P2n show the voltage curve for the case of an electrically conductive connection (curve 296) and for the case of a line interruption (curve 298).

29 shows the difference signal DU _P2p-P2n of the evaluation unit in the control unit 122 within the detection pulse. On the X axis 182 is in 29 the time is plotted in s. The voltage in mV is plotted on the Y axis 184. Curve 300 indicates the difference signal DU _P2P-P2n = U _P2p-P2n (2nd sample) - U _P2p-P2n (1st sample). The difference signal DU _P2p-P2n , which is used to check the plausibility of a line interruption detection, is in 29 thus presented in more detail. Within the scope of the present invention, the threshold 212 is defined in the form of a voltage threshold. If the difference signal DU _P2p-P2n is below the threshold value 212, an interruption in the conductive connection to the NOx cell 140 is detected. If the difference signal DU _P2p-P2n is above this threshold value 212, the connection is considered intact. In the example in 29 If the connection to the NOx cell 140 is initially conductive, as indicated by the time area 302, then it is briefly interrupted, as indicated by the time area 304, and after a few seconds it is conductive (closed) again, as indicated by the time area 306 The large differences in the voltage signal DU _P2p-P2n between the conductive and interrupted connection to the sensor element 110 as well as the large distance between the voltages and the detection threshold 212 become clear here.

The procedure can be modified as follows. The application of the circuit and method described here for detecting an interruption in the line to the NOx cell 140 is not only limited to NOx sensors. In principle it is applicable to all electrochemical cells that have electrodes with a large capacity. In addition to using the electrical connection with voltage pulses, excitation with current pulses is also possible.

Proof of whether the invention is used in a third-party product can be easily and easily provided in an electronics laboratory. Pulse trains and pulse patterns can be measured with an oscilloscope at the electrical connections to the sensor element. By coupling an interference signal with a signal generator, it can be examined whether the temporal change in the pulse response may serve as a diagnostic criterion. An analysis of the opened product will then provide more detailed information about the possible use of the circuit and diagnostic procedure described here. Here the signals can be measured directly on the pins of the integrated circuits.

QUOTES INCLUDED IN THE DESCRIPTION

This list of documents listed by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Non-patent literature cited

• Reif, K., Deitsche, K-H. et al., Automotive Engineering Handbook, Springer Vieweg, Wiesbaden, 2014, pages 1338-1347 [0002]

Claims (15)

Method for operating a sensor (100) for detecting at least a portion of a measurement gas component with bound oxygen in a measurement gas, in particular in an exhaust gas of an internal combustion engine, comprising a sensor element (110), wherein the sensor element (110) comprises a first pump cell (112) which has an outer pump electrode (114) and an inner pump electrode (116) and which is adjacent to a first cavity (126) which is connected to the measurement gas, a reference cell (130) which has a Nernst electrode (132) and a reference electrode (134) and which is adjacent to a reference gas chamber (138), and a second pump cell (140) which has a pump electrode (142) and a counter electrode (144) and which is adjacent to a second cavity (145), wherein an electronic control unit (122) which has at least a first separate connection (P1) for the first pump cell (112) and a second separate connection (P2) for the second pump cell (140) is connected to the sensor element (110), wherein the first pump cell (112) is connected to the first separate connection (P1) by means of an electrically conductive connection (120), wherein the second pump cell (140) is connected to the second separate connection (P2) by means of an electrically conductive connection (146), wherein a measuring resistor (160) is provided in the electrically conductive connection (146) which connects the second pump cell (140) to the second separate connection (P2), wherein a current and/or voltage excitation of the second pump cell (140) for generating a measuring signal (U _P2p-P2n ) is carried out at the measuring resistor (160) by means of the control device (122), wherein the electrically conductive connection (146) which connects the second pump cell (140) to the second separate connection (P2) is identified as intact if an amount of a temporal change in the measuring signal (U _P2p-P2n ) exceeds a predetermined threshold value (212) during the current and/or voltage excitation, and is identified as defective if the amount of the temporal change of the measuring signal (U _P2p-P2n ) during the current and/or voltage excitation falls below the predetermined threshold value (212). Procedure according to Claim 1 , wherein a predetermined electrical voltage (U _P2 ) is applied to the second pump cell (140), wherein a voltage excitation of the second pump cell (140) is carried out, the voltage excitation comprising a change in the predetermined electrical voltage (U _P2 ) for a predetermined time . Procedure according to Claim 2 , wherein the predetermined electrical voltage (U _P2 ) is increased for the predetermined time. Procedure according to Claim 2 or 3 , wherein the predetermined electrical voltage (U _P2 ) is applied to the second pump cell (140) in the form of at least one voltage pulse for the predetermined time, wherein the temporal change of the measurement signal (U _P2p-P2n ) is determined by calculating a difference between a first measured value of the measurement signal at the beginning of the voltage pulse and a second measured value of the measurement signal (U _P2p-P2n ) at the end of the voltage pulse. Method according to the preceding claim, wherein the predetermined time for the voltage pulse has a duration of 1 ms to 10 ms and preferably 1.5 ms to 5 ms. Method according to the preceding claim, wherein the predetermined electrical voltage (U _P2 ) is applied to the second pump cell (140) in the form of a first voltage pulse for a first predetermined time and a second voltage coil for a second predetermined time, the first voltage pulse and the second voltage pulse have an opposite polarity, wherein an integral of the first voltage pulse and the second voltage pulse has a value of zero. Method according to the preceding claim, wherein a plurality of pulse pairs comprising the first voltage pulse and the second voltage pulse are applied to the second pump cell (140), wherein the polarities of successive voltage pulse pairs alternate regularly or irregularly. Method according to the preceding claim, wherein the number of positive polarities of the first voltage pulse of the voltage pulse pairs and the number of negative polarities of the first voltage pulse of the voltage pulse pairs are identical. Method according to one of the Claims 4 until 8th , wherein the method further comprises checking the plausibility of the measurement signal if the electrically conductive connection (146) connecting the second pump cell (140) to the second separate terminal (P2) is identified as defective. Method according to the preceding claim, wherein for plausibility the predetermined electrical voltage (U _P2 ) is applied to the second pump cell (140) in the form of at least one detection voltage pulse for a predetermined detection time, the temporal Change in the measurement signal is determined by calculating a difference between a first measured value of the measurement signal at the beginning of the voltage pulse and a second measured value of the measurement signal at the end of the detection voltage pulse, the detection time for the detection voltage pulse being longer than the predetermined time for the voltage pulse, wherein an amplitude of the detection voltage pulse is preferably the same size as an amplitude of the voltage pulse. Method according to one of the preceding claims, wherein the control device (122) is designed to regulate a voltage at the second pump cell (140), the control device (122) further having an excitation signal source (156), wherein a voltage setpoint is controlled as a reference variable, wherein the Control device (122) controls a time sequence of changes in the voltage setpoint, using measurements of a voltage drop across the measuring resistor (160) to interrupt the electrically conductive connection (146) that connects the second pump cell (140) to the second separate connection (P2), is detectable. Computer program arranged to carry out each step of the method according to one of the preceding claims. Electronic storage medium on which a computer program according to the preceding claim is stored. Electronic control device (122), which comprises an electronic storage medium according to the preceding claim. Sensor (100) for detecting at least a portion of a measurement gas component with bound oxygen in a measurement gas, in particular in an exhaust gas of an internal combustion engine, comprising a sensor element (110), wherein the sensor element (110) has a first pump cell (112) which has an outer pump electrode (114) and an inner pump electrode (116) and which is adjacent to a first cavity (126) which is in communication with the measurement gas, a reference cell (130) which has a Nernst electrode (132) and a reference electrode (134) and which is adjacent to a reference gas chamber (138), and a second pump cell (140) which has a pump electrode (142) and a counter electrode (144) and which is adjacent to a second cavity (145), wherein the sensor (100) further comprises an electronic control unit (122) according to the preceding claim.

Images