GB2314634A

GB2314634A - method of operating a solid electrolyte probe and determining the internal resistance thereof

Info

Publication number: GB2314634A
Application number: GB9713673A
Authority: GB
Inventors: Hong Zhang; Alexander Ketterer
Original assignee: Siemens AG
Current assignee: Siemens AG
Priority date: 1996-06-27
Filing date: 1997-06-27
Publication date: 1998-01-07
Anticipated expiration: 2017-06-27
Also published as: DE19625899C2; DE19625899A1; GB2314634B; GB9713673D0; FR2750501B1; FR2750501A1

Abstract

A solid electrolyte oxygen probe of the voltaic type is supplied directly with a clocked pumping current which generates reference gas, and the supply of pumping current is synchronised with the lambda controller frequency. The internal resistance of the probe is determined from the probe output while the pumping current is being applied. Pumping is started when the probe output reaches a threshold value (point B) at which the probe signal is stable. Application of the pumping current causes the sensor output to follow the curve BCC' instead of the usual dotted curve. From the probe output before (point B) and after (point C) the pumping phase, the internal resistance of the sensor can be determined and used to control heating of the probe.

Description

METHOD FOR OPERATING AN OXYGEN PROBE The invention relates to a method for

operating an oxygen probe, and in particular to a method of operating an oxygen probe to provide an input to a lambda control system.

An oxygen probes operating according to the principle of the oxygen concentration cell is known from DE 44 10 016 Al. Such a probe is used in order to detect the oxygen concentration in an exhaust gas and the manner in which the exhaust gas is expelled from an internal combustion engine so that the air/fuel ratio which is supplied to the internal combustion engine can be adjusted. An oxygen probe operating according to the principles of an oxygen concentration cell has an electrochemical cell comprising a solid electrolyte body, which conducts oxygen ions, a reference electrode and a measuring electrode. A reference gas is admitted to the reference electrode and an exhaust gas to be measured is admitted to the measuring electrode. An electromotive force is generated between the reference electrode and the measuring electrode, the magnitude of the electrodemotive force being dependent on the difference in oxygen concentration between the reference gas and the exhaust gas.

Generally, the atmosphere is used as a reference gas in this method. A relatively small current (pumping current) is applied between the reference electrode and the measuring electrode, and generates in the electrochemical cell an oxygen pumping action to effect the entry of gas from the atmosphere, as the reference gas, into the reference cell.

The output signal of the lambda probe is strongly influenced by its temperature, and this influence has a disruptive effect on the precision of the lambda control system to which the signal produced by the probe is input. As a result, measures are known for compensating for the effect of the temperature of the probe on the output signal. In this connection, it should be noted that the exhaust-gas probe is not ready for operation until above a temperature threshold which is generally above the exhaust-gas temperature in no load operation. For the purpose of temperature compensation, it is known from DE-OS 29 28 496 for example, additionally to heat the probe electrically and to control its temperature in order to achieve a lambda signal which is as precise as possible. In this connection, the measured internal resistance of the ceramic probe body is used as an actual value for the temperature control.

The invention seeks to provide a method with which is the pumping current of an oxygen reference cell of an oxygen probe can be provided in a simple way, and reliable information about the internal resistance of the probe can be obtained at the same time.

According to the present invention, there is provided a method for operating an oxygen probe having:

a concentration cell, to which a gas to be analysed is admitted; an oxygen pumping cell, to which a reference gas is admitted; and a power source; the method comprising the steps of:

supplying a discontinuous pumping current from the power source to the oxygen pumping cell in dependence on the output of the oxygen probe, to make the reference gas available by means of a pumping action; and evaluating the output signal of the probe during the supply of the pumping current in order to determine the internal resistance of the probe.

By supplying the reference cell with a discontiuous, or clocked, pumping current, in which the pumping current on-phases are synchronised with the output of the oxygen sensor, then in a simple way, by measuring the probe voltage before and during the pumping current phase, the internal resistance of the probe can be established with a high level of accuracy and then used as an input variable for the heating control of the lambda probe.

As a trigger event for the pumping current phase, during which time the pumping current is applied to the oxygen pumping cell, evaluations are made regarding whether threshold values at which a stable probe signal is present are exceeded or whether there is a fall below these threshold values. If the pumping current is then applied for a short time, the total resistance of the circuit changes and an alteration of the probe signal is obtained, which alteration is caused solely by the application of the pumping current and is not influenced by lambda changes. After a minimum pumping time, which is determined by the system, the internal resistance of the oxygen probe can be determined.

Advantageous developments of the invention are set out in the subclaims.

For a better understanding of the present invention, and to show how it may be brought into effect, reference will now be made, by way of example, to the accompanying drawings, in which:

Figure 1 shows a basic representation of an electrical equivalent circuit of a lambda probe with pumped oxygen reference; and Figure 2 shows the output signal of such a lambda probe for an active lambda control.

Figure I shows the equivalent circuit diagram of a heated, planar lambda probe, which works according to the principle of the galvanic oxygen-concentration cell with ZrO2-solid electrolyte. This known lambda probe has a measuring cell and an oxygen pumping cell (reference cell). A so-called Nernst voltage develops between an electrode which is allocated to the measuring cell and an 02-reference electrode which is allocated to the pumping cell. The magnitude of this voltage is dependent on the partial oxygen pressures in the measuring cell and in the reference cell. In the circuit diagram, this voltage is termed voltage source UN. At the series circuit arrangement of this voltage source and the internal resistance RLS of the probe, the probe signal VLS can be picked up, which signal contains information about the oxygen concentration in the exhaust gas of an internal combustion engine. This information is used in a known way to control the air/fuel ratio of the internal combustion engine.

The 02-reference cell of such a lambda probe has no connection to the atmosphere, and a relatively small current is generated, with the aid of a voltage source and a series resistance, in order to fill the reference cell with oxygen. This current results in the entry of oxygen into the reference cell by means of the so called oxygen pumping action. In Figure 1, VCC denotes a stabilised supply voltage, which generates such a pumping current PI by way of a series resistance RVOR.

The magnitude of the pumping current PI must be limited, because an excessively high current causes the so-called blackening or darkening phenomenon, by which oxygen ions are removed from the solid electrolyte elements and as a result, a very rapid worsening of the action of the oxygen pumping element, and consequently a worsening of the operation of the whole lambda control device, takes place.

The pumping current PI is applied by actuating an electronic switchgear SCH1, which is controlled by means of signals from the control device STG in a manner explained later with the aid of Figure 2.

The lambda probe is only ready for operation above a minimum operating temperature, and consequently regulation of the air/fuel mixture is not possible until the probe has reached its operating temperature.

As a result, the heating of the probe is accelerated by the application of electrical probe heating. Apart from this, the probe heating ensures that in operating ranges of the internal combustion engine (e.g. no-load running) in which the heat output of the exhaust gas is not sufficient, the probe temperature is kept constantly at a predetermined value.

A heating control is necessary to keep the probe temperature constantly at a predetermined value because only a defined temperature of the probe delivers with high accuracy a signal which represents the oxygen content in the exhaust gas. If the temperature of the sensor varies greatly, the probe signal is not only is dependent on the air ratio X, but is also dependent on the temperature in an undesired manner.

In Figure 1, this probe heating is represented by means of a heating resistance RH, to which can be applied a heating voltage UH by way of a further switchgear SCH2 by means of trigger signals from the control device STG. In this connection, the heating of the lambda probe is regulated to a constant internal resistance of 200 0, for example. For this purpose, it is necessary to know the value of the internal resistance RLS of the probe.

With the aid of Figure 2, it is explained both how the pumping current PI can be made available for the reference cell and how the internal resistance RLS of the probe can be determined in a very precise manner.

The graph shows the change with respect to time of the probe signal VLS for an active lambda control and a warmed-up probe during a complete controller oscillation. Apart from this, various threshold values for the probe signal are shown. In this connection, VM_MMV_MAX denotes an average maximum probe voltage and V1S-MV-MIN denotes an average minimum probe 6voltage. A "rich" recognition threshold is denoted with VM_RICH, and a "lean" recognition threshold is denoted with VLS_LEAN. The time span between the point A, at which the probe signal VLS exceeds the "rich" recognition threshold VLS - RICH, up to the point D at which the probe signal VLS falls below the "lean" recognition threshold VLS - LEAN, is defined as a rich range VLS-CYC-AFR. Furthermore, the time span between the point D and a point F, at which the probe signal VLS again exceeds the "rich" recognition threshold VLS-RICH, is defined as a lean range VLS-CYC-AFL. The sum of these two times VLS - CYC - AFR + VLS-CYC-AFL corresponds.to the period of the controller oscillation.

is As already mentioned in the introduction, the pumping current must not exceed a maximum value which is essentially predetermined by the probe construction.

For this reason, a value for the continuous pumping current is specified by the probe manufacturer. It is, however, also possible to work with a higher pumping current for a short time, if subsequently a correspondingly long pause is maintained in which no pumping takes place. By selecting the pulse-pause ratio or the clock-pulse ratio, a current can then be supplied to the reference cell, the average value of which current corresponds to the predetermined value of the continuous pumping current.

The internal resistance RLS of the lambda probe is determined by evaluating the probe signal VLS before and after the pumping current phase. By switching on the pumping current (switchgear SCH1 closed), the total resistance of the circuit alters on the basis of the series resistance RVOR (Figure 1) which is now connected, and a signal superelevation occurs during pumping.

The pumping current, however, is not applied until the probe signal changes from rich to lean or vice versa and the probe signal reaches a stable state.

This is the case when the probe signal VLS either exceeds the average maximum probe voltage VLS - MMV - MAX (point B) or falls below the average probe voltage VLS-MM-MIN (point E).

When the probe signal VLS reaches the threshold value VLS - MMV - MAX (point B), then for a certain time T - PI, also termed pumping current phase, the pumping current PI is switched on. As a result of this, the probe signal VLS no longer follows the curve which is drawn in with a broken line, but instead the voltage rises further up to a point C' at which the pumping current is switched off again.

After a minimum pumping current time C - T - PI - MIN, which must be observed in order to ascertain the effects of the voltage alteration which has been caused and to be able to determine from this the internal resistance RLS, the voltage VLS - UP_PI,,,,. is measured at the instant C. If the measured voltage before the pumping current phase is denoted with VLS - UPn (point B) the internal resistance of the probe can be calculated according to the following relation:

-V -pl RLS=RVORO (VLS-UP' LSUP -1) (VLS- UP-PI, 1 - VCC - VORF) In this connection, VCC denotes the supply voltage, which generates the pumping current PI via the series resistance RVOR, and VOFF denotes an offset volt.age, which takes into account the electrical performance of the circuit according to Figure 1, e.g.

the saturation voltage of the transistor (switchgear SCH1), to which the series resistance RVOR is connected.

The internal resistance can also be calculated in an analogous way by subtraction of the corresponding voltage values if the probe signal VLS has fallen below the threshold VLS-MMV-MIN.

In order to determine the internal resistance of the lambda probe, it is therefore not necessary to evaluate the continuous pumping current, which has a small magnitude (typical values are in the range of 20 gA), which can result in measurement inaccuracies, but instead, by clocking the pumping current with higher currents (typically 1mA) whenever the probe signal has reached a stable state, the internal resistance can be determined very precisely. The voltage alterations to be evaluated are thus dependent exclusively on the influence of the pumping current and are not influenced by lambda fluctuations. If the application of the pumping current is not synchronised with respect to time with the output signal of the probe, but instead at an instant at which the sensor signal is strongly non-stationary, then measured values are obtained which result by superposition from the connection of the series resistance and the sensor performance and consequently falsify the result of the resistance definition.

In order not to destroy the probe by pumping currents which are too high and/or pumping current times which are too long, the clocked pumping current, the pumping current time and also the rest period which follows the pumping current time, in which rest period no pumping current flows, must be kept constant. In the following, the duty factor is denoted with the expression C-FAC-T_PI. The value of this duty factor resu I lts from the probe characteristic values, in particular from the recommended continuous pumping current and the maximum permissible pumping current.

The generation of the pumping current PI by the supply voltage VCC which is applied via the series resistance RVOR results in the above-mentioned voltage superelevation of the probe signal VLS. For this reason, the probe signal during the pumping current phase cannot be used for the evaluation of the lambda control. Therefore, as described in the case of determining the internal resistance of the probe, the pumping current phase does not begin until the probe signal has reached a stable state. If the probe signal VLS exceeds the threshold VLS - MMV-MAX, or if there is a fall below the threshold VLS - MMV - MIN, the pumping current PI is switched on for the duration T-PI.

In order to calculate this pumping current time T PI, it needs to be decided whether the lambda control is active and the probe internal resistance, which represents a gauge for the probe temperature, has fallen below a certain threshold value.

When the probe is cold, it has a very high internal resistance (typical values are in the MQ range) and there is no measurable voltage available as a probe signal which could be evaluated. A voltage, which results from the protective circuit of the probe, is adjusted at the analysis electronics. If the probe is heated by the exhaust gases and/or by the electric heating, the probe begins to work and gives off a voltage which is dependent on the air ratio Lambda and the temperature. The temperature, however, is approximately stable after a certain time. An interrogation is therefore carried out to see whether the probe internal resistance RLS has fallen below the threshold value C - RLS_PRE_MIN and there is therefore a warm probe which is ready for operation. The threshold value C_RLS_PRE_MIN is applied in a manner which is dependent on the probe type.

If the lambda control is active and the probe internal resistance RLS is less than or equal to this threshold value C - RLS-PRE - MIN, then a further interrogation is carried out to see whether the probe signal VLS lies above the average maximum value VLS-MMV-MAX or below the average minimum value VLS-MMV-MIN.

Consequently, it is possible to distinguish between the following cases:

a) lambda control active RLS 5 C - RLS-PU-MIN VLS > VLS_MMV_MAX in this case, the pumping current time T-PI is calculated by the formula T_PI = VLS_CYC_AFL C-FAC-T-PI where VLS - CYC - AFL denotes the lean range, i.e. the time during which the probe signal indicates a lean mixture and which time is measured by means of a time counter; and C FAC T PI denotes the duty factor which results from the average pumping current (e.g.: 20 AA continuous pumping current) and the choice of amplitude of the pumping current which is applied for a short time (e.g.: imA).

b) lambda control active RLS!-. C-RLS-PRE_MIN VLS < VLS-MMV-MIN in this case, the pumping current time T-PI is calculated by the formula T_PI = VLS_CYC_AFR C-FAC-T-PI.

where VLS CYC AFR denotes the rich range, i.e. the time' during which the probe signal indicates a rich mixture and which time is measured by means of a time counter; and C - FAC T - PI again denotes the duty factor already defined above.

In both cases, the pumping current time T-PI must lie within a limited range:

C-TP-I-MIN < T_PI < C_TP_I_MAX.

The minimum pumping current time C - T - PI - MIN must be observed, because the measuring process of the probe voltage requires a certain time span. The maximum pumping current time C - T - PI - MAX must not be exceeded, because otherwise the probe could be damaged. The control frequency of the lambda control has some influence on the formula for calculating the pumping current time T_PI via the times VLS-CYC-AFL and VLS - CYC - AFR, and so at certain working points with very low control frequencies and thus very long periods it is possible that the pumping current PI remains switched on for so long that the probe is damaged or even destroyed.

C) lambda control active RLS > C-RLS-PRE_MIN VLS > VLS-MMV-MAX or VLS < VLS-MMV-MIN in this case, the pumping current time T-PI is calculated according to the formula T_PI = C_T_PI_MIN.

In the case of a probe which is still cold, it is not possible to measure the times VLS - CYC - AFL and VLS-CYC-AFR as in the case of a warm probe (cases a) and b)). For this reason, the minimum pumping current time is used for safety reasons.

In the case of an active lambda control, the start and the duration of the pumping process are selected as a function of the control frequency of the lambda controller. If the lambda controller is not active, the pumping process is also not dependent on the probe signal and when calculating the pumping current time, it only needs to be decided whether the probe is warm or cold. The pumping process does not have to be triggered by the thresholds VLS - WV-MAX or VLS_WV_MIN, but can take place whenever necessary.

d) lambda controller not active RLS s C_RLS_PRE_MIN (warm probe) In this case, an applicable standard value C - T - PI - STND is selected as pumping current time, which is established by tests on the test stand:

T_PI = C T_PI_STND.

e) lambda controller not active RLS > C-RLS-PRE_MIN (cold probe) If the internal resistance, which reflects the temperature of the probe, has not yet fallen below the threshold value C - RLS - PRE-MIN, pumping is carried out only during the minimum pumping current time, which permits just one measurement of the probe voltage:

T_PI = C_T_PI_MIN.

The pumping current phase in the case of an inactive lambda control always starts with a pause, while in the case of an active lambda control, the exceeding of the thresholds VLS - MW-MAX or VLS-MW-MIN is used as a trigger event and the pumping current phase is begun immediately.

The pause between two successive pumping current phases turns out to be:

T-PI T-PI-OFF C-FAC-T-PI in all cases.

Claims

1. A method for operating an oxygen probe having: a concentration cell, to which a gas to be analysed is admitted; an oxygen pumping cell, to which a reference gas is admitted; and a power source; the method comprising the steps of:

is

2. A method according to claim 1, wherein the timing of the supply of the pumping current is determined by evaluating threshold values, at the attainment of which the output signal of the oxygen probe has a stable state.

3. A method according to claim 1 or 2, wherein the duty factor for the pumping current is determined by a continuous pumping current, which is dependent on the probe construction, and the maximum permissible pumping current which can be applied for a short time.

4. A method according to any preceding claim, wherein the determination of the internal resistance of the probe takes place by evaluating the probe signal before and during the supply of the pumping current.

5. A method according to claim 4, wherein the internal resistance is determined according to the following relationship:

-V RLS=RVORO (VLS-UP' LSUP-Pl") (VLS- UP-Pl. 1 - VCC - VORF) -14where:

RLS = internal resistance of oxygen probe RVOR = series resistance VLS-UPn = measured voltage before the pumping current phase VLS-UP-PIn+l = measured voltage during the pumping current phase vcc = supply voltage which generates the pumping current PI via the series resistance RVOR VOFF = offset voltage

6. A method according to one of the preceding claims, wherein is in the case of an active lambda control, the internal resistance of the probe is compared with a threshold value which characterises a probe which is ready for operation, if there is a fall below this threshold value for the internal resistance, a check is subsequently made to see whether the probe signal lies above the threshold value which represents an average maximum probe voltage or below a threshold value which represents an average minimum probe voltage, and as a function thereof, the period for which the pumping current is supplied is determined in different ways.

7. A method according to claim 6, wherein if the threshold value is exceeded, the pumping current period is determined as T_PI = VIS_=_AFL C-FAC-T-PI where:

T-PI = pumping current period VLS_=_AFL = time during which the probe signal indicates lean, C-FAC-T-PI = duty factor -is-

8. A method according to claim 6, wherein if there is a fall below the threshold value, the pumping current period is determined as T_PI = VW_=_APR C-FAC-T-PI where:

T-PI = pumping current period VW_=_AFR = time during which the probe signal indicates rich, C-FAC-T-PI = duty factor

9. A method according to any preceding claim, wherein the pumping current period is limited by a lower threshold value and an upper threshold value.

10. A method according to any preceding claim, wherein is in the case of an active lambda control, the internal resistance of the probe is compared with a threshold value which characterises a probe which is ready for operation, if this threshold value for the internal resistance is exceeded, it is inferred that the probe is not ready for operation, a check is subsequently made to see whether the probe signal lies above the threshold value which represents an average maximum probe voltage or below a threshold value which represents an average minimum probe voltage, and in both cases, the pumping current period, irrespective of the times during which the probe signal indicates lean, or rich,, is determined as:

T_PI = C_T_PI_MIN where:

T-PI = pumping current period C-T PI-MIN = minimum pumping current time

11. A method according to any preceding claim, wherein in the case of an inactive lambda control, the internal resistance of the probe is compared with a threshold value which characterises a probe which is ready for operation, if there is a fall below this threshold value, the pumping current phase is determined as:

T_PI = C_T_PI_STND where:

T-PI = pumping current period C-T_PI_STND = standard value which can be applied, if this threshold value is exceeded, the pumping current period is determined as:

T_PI = C_T_PI_MIN where:

T PI = pumping current period is C-T_PI_MIN = minimum pumping current time.

12. A method according any preceding claim, wherein the pause between two pumping current periods is determined as TPI TIPI-OFF= 1 -1 C-FAC-T-PI

13. A method of operating an oxygen probe substantially as herein described, with reference to the accompanying drawings.

14. A method for operating an oxygen probe which works according to the principle of the galvanic oxygen concentration cell with solid electrolyte, the outpu signal of which oxygen probe changes abruptly in the event of a theoretical air/fuel ratio and is used as an input variable for a lambda controller of a lambda control system, having a concentration cell and an oxygen pumping cell, with a gas to be analysed being admitted to the concentration cell and a reference gas which has a predetermined oxygen concentration being admitted to the oxygen pumping cell, and having a power source for generating a pumping current, so that in order to make the reference gas available, the pumping action is carried out, wherein the oxygen pumping cell is supplied with a clocked pumping current, with the pumping current phases being synchronised with the lambda controller frequency, and during the pumping current phase, the output signal of the probe is evaluated in order to determine the internal resistance of the probe.

15. An oxygen probe operated in accordance with the method as claimed in any preceding claim.

16. An internal combustion engine having an is oxygen probe as claimed in claim 15.