Background of the Invention Chemical sensors are used to monitor the concentration of specific chemical species which may be in liquid or gaseous form. Generally, chemical sensors comprise a sensitive layer for detecting a specific chemical species and a heater for heating the sensitive layer. A problem with such chemical sensors is that humidity can interfere with the response of the sensor which can lead to inaccurate sensor readings. For example, in a carbon monoxide fire alarm comprising a carbon monoxide chemical sensor having a metal oxide sensitive layer, the effects of humidity can cause a false alarm or no alarm in error.

Humidity is always present in the atmosphere and can vary from one place to another and from day to day. Thus, in order to avoid false alarms due to humidity, it is desirable that the sensor response be corrected to account for humidity.

One possible solution would be to use a dedicated humidity sensor in addition to the chemical sensor. An algorithm must then be used to subtract the humidity signal from the chemical sensor response. Such a solution has a number of disadvantages. Since two separate sensors are used, their output signals must be synchronised and stable before a correction can be made. This increases the complexity of the system and increases the chance for errors and also requires that the effect of the humidity level be stored on a regular basis. In addition, this solution requires two sensors which requires more space than one sensor.

Another solution, which is used in US Patent No. 5,517,182, uses one sensor which comprises a sensitive layer for sensing carbon monoxide and a heater for heating the sensitive layer sequentially to two different temperatures by way of a pulsed signal applied to the heater. The response of the sensor is measured during the period when the sensitive layer is at

the lower temperature and the dynamic response is then de-convoluted to account for humidity.

Since the de-convolution of the sensor's response cannot discriminate properly the signal due to humidity, this solution does not sufficiently compensate for humidity.

There is therefore a need for an improved method and sensor system for determining a concentration of a chemical species with improved humidity compensation.

Summary of the Invention In accordance with a first aspect of the present invention there is provided a method for determining a concentration of a chemical species in the presence of a contaminant as recited in claim 1 of the accompanying claims.

In accordance with a second aspect of the invention there is provided a sensor system for determining a concentration of a chemical species in the presence of a contaminant as recited in claim 7 of the accompanying claims.

Brief Description of the Drawings A method and a sensor system for determining the concentration of a chemical species in accordance with the present invention will now be described, by way of example only, with reference to the accompanying drawings in which: FIG. 1 is a block schematic diagram of a sensor system in accordance with the present invention; FIG. 2 is a graphical representation of a preferred pulse voltage signal for heating the chemical sensor of the sensor system of FIG. 1; FIG. 3 is a graphical representation of the response of the chemical sensor of the sensor system of FIG. 1 over a heating cycle; FIG. 4 is a graphical representation of the time derivative of the chemical sensor response during a first period of the heating cycle; FIG. 5 is a graphical representation of the time derivative of the chemical sensor response during a second period of the heating cycle; and FIG. 6 is a graphical representation of the time derivative of the chemical sensor response during a second period for different concentrations of carbon monoxide.

Detailed Description of the Drawings Referring firstly to FIG. 1, a sensor system 2 in accordance with a preferred embodiment of the present invention comprises a chemical sensor 4 and a processing unit 6. The chemical sensor 4 comprises a sensitive layer, represented in FIG. 1 by resistance 8, and a heater, represented by resistance 10, for heating the sensitive layer. Contacts 12,14 to the sensitive layer are coupled to inputs 16,18 of the processing unit 6.

Contacts 20,22 to the heater are also coupled to the processing unit 6 for receiving a voltage signal generated by a signal supply circuitry 24 in the processing unit 6.

In the preferred embodiment described below, the chemical sensor 4 is a carbon monoxide chemical sensor having a metal oxide sensitive layer.

The present invention may however be applied to different types of chemical sensors having sensitive layers formed of material other than a metal oxide material.

The value of the resistance 8 of the sensitive layer varies according to the concentration of the species to be detected by the chemical sensor 4 and with temperature. Thus, the processing unit 6 monitors the resistance of the sensitive layer across contacts 12 and 14 in order to determine the response of the sensitive layer and hence the concentration of the chemical species.

The signal supply circuitry 24 provides a pulsed voltage signal to the heater of the chemical sensor 4 so that the heater heats the sensitive layer to a first temperature during a first period and to a second temperature during a second consecutive period, the second temperature being greater than the first.

The inventors have recognised that during the first period, the chemical sensor response is a function of the concentration of the chemical species to be detected by the chemical sensor, which in the preferred embodiment is carbon monoxide, and the humidity level in the surrounding environment, whilst during the second period by arranging for the second temperature to reach an optimum value, the chemical sensor response is a function of the humidity level only. In other words, by selecting appropriate values for the heating cycle during the second period, the chemical sensor response is substantially constant with respect to changes in the concentration of carbon monoxide.

In the preferred embodiment, the optimum value is in the range of 380°C to 450°C. For a N02 sensor, the optimum value will be in the range of 250°C to 320°C. The temperature is determined by the level of the voltage

signal applied to the heater and its duration (i. e. the duration of the second period). The level of the voltage signal for the second period, and hence the value of the second temperature, is preferably determined by analysing the sensor's response over various voltage levels for a given concentration of carbon monoxide and by selecting the voltage level at which the sensor response is a maximum, that is, the voltage level at which the sensor is the most sensitive.

In the preferred embodiment, each heating cycle of the pulse voltage signal 26 (shown in FIG. 2) comprises a 1 volt pulse for a period of 10 seconds followed by a 5 volt pulse for a period of 5 seconds. Thus, in the preferred embodiment the first period is 10 seconds and the second period is 5 seconds. It will of course be appreciated that other heating cycles may also be used.

The value of the resistance 8 of the sensitive layer, and hence the response of the chemical sensor 4, varies over the first period and second periods of the heating cycle as shown by the curve 28 in FIG. 3. Since during the first period or cold period, the heater heats the sensitive layer to the first temperature which is lower than the second temperature during the second period or hot period, the resistance of the sensitive layer is higher in the first period than during the second period.

Thus, the method and sensor system for determining a concentration of a chemical species in accordance with the present invention measures the sensor response during both the first cold period and the second hot period and then processes both the measured responses to determine the actual concentration of the carbon monoxide for which humidity has been compensated.

Although the absolute values of the sensor response during the first and second periods could be used in order to determine the actual carbon monoxide concentration, in the preferred embodiment, the time derivative of the measured sensor response during both the first period and the second period is determined and then processed in order to determine the actual concentration. An advantage of processing time derivatives rather than absolute values is that variations in the resistance of the sensitive layer, which can occur for example due to the natural ageing of the metal oxide, do not affect the determined actual concentration of carbon monoxide. With the continuous drift of the sensitive layer's resistance, if the absolute humidity is compensated for as opposed to relative humidity, the sensor system may induce false alarms.

The processing of the measured sensor responses and the calculation of the time derivatives in the first and second periods and their subsequent processing may be carried out by the processing unit 6 or dedicated units (not specifically shown in FIG. 1) in the processing unit or discrete units (not shown). The processing unit 6 may for example comprise a microcontroller such as a HC05 microcontroller supplied by Motorola, Inc.

FIG. 4 shows a smoothed curve 29 representing the time derivative of the sensor response (dR/dt) during the first period after filtering. The time for the derivative of the sensor response to stabilise is the time constant To of the derivative during the first period. The time constant To is dependent on the carbon monoxide concentration and the humidity level and can be determined by monitoring the derivative during the first period.

In the preferred embodiment, the time constant is measured by sampling the time derivative of the sensor's response and by determining (via for example a counter) when x sampled points lie within a predetermined threshold window. x is chosen to allow the time derivative to stabilise.

FIG. 5 shows a smoothed curve 30 representing the time derivative of the sensor response (dR/dt) during the second period after filtering. As can be seen from the curve 30 in FIG. 5, the time derivative of the sensor response or resistance of the sensitive layer does not vary much during the second period. In fact, from experiments, the inventors found that the time derivative of the sensor response during the second period was substantially constant for different concentrations of carbon monoxide (see curves 32,34, 36 in FIG. 6). As can be seen from FIG. 6, the time derivative decreases as the humidity level increases.

The method for determining the concentration of carbon monoxide in accordance with the preferred embodiment of the present invention will now be described with reference to FIG. s 1-6.

Initially a calibration process is performed to determine the time constant To of the derivative of the sensor's response during the first period for different known concentrations of carbon monoxide and different known levels of humidity. From this calibration process, first data, such as the look-up shown in Table 1 below, can be produced which holds the determined time constants To for the different carbon monoxide concentrations and for the different humidity levels.

Table 1 PercentageofHumidity 20%50%80% CarbonOppm>10.Os10.0s 7. 0s Monoxide15ppm 10. 0s7.5s5.5s Concentration60ppm7.2s4.2s4.2s I00ppm6.Os3.2s2.2s

At the same time as table 1 is being produced, the value of the derivative of the sensor's response during the second period is determined for the different known concentrations of carbon monoxide and the different known levels of humidity so as to collect second data for the values of the derivative for different humidity levels. The curves 32-36 of FIG. 6 may be produced from this second data.

Preferably, for each concentration of carbon monoxide and for each level of humidity, the time constant To during the first period of a heating cycle and the value of the derivative in the second period of the heating cycle is measured over three heating cycles and the average values stored.

Measuring over three cycles improves data reliability. However, it is possible to use the values after one heating cycle only.

The calibration may be performed for each chemical sensor device.

However, this can be extremely time consuming. For chemical sensors integrated on a semiconductor substrate, the first and second data will be substantially the same for the same family of sensors which means that calibration need only be performed once for each family. Thus, sensor calibration is preferably performed once for each class of sensors having similar sensing resistances.

Once the calibration has been performed the first and second data is held in the processing unit 6.

In normal operation of the sensor system 2, the pulsed signal 26 is applied to the heater and the derivative of the sensor's response during a first period of the heating cycle is determined and the time constant of the determined derivative is measured. The derivative of the sensor's response during a second period of a heating cycle is determined and the value of the derivative is measured. From the measured value and using the second data held in the processing unit, the actual level of humidity can be determined.

The determined level of humidity and the measured time constant is then compared with the first data in order to determine the actual concentration of carbon monoxide.

Preferably, the measured value of the derivative in the second period of a previous heating cycle is used with the time constant of the derivative in the first period of the next heating cycle. In other words, the first measurements are taken in the second period (or hot period) and then in the first period (or cold period) of the next heating cycle.

In the preferred embodiment this comparison step comprises using the determined level of humidity to select one of the columns in the pre- stored look-up table, table 1, and then by reading off the concentration of carbon monoxide for the measured time constant in the selected column, the actual concentration of carbon monoxide can be determined.

Preferably, in normal operation the time constant To during the first period of a heating cycle and the value of the derivative in the second period of a heating cycle is measured over three heating cycles and the average values determined and used for determining the actual concentration of carbon monoxide.

The present invention therefore provides a method and sensor system which discriminates and measures properly the humidity and actual carbon monoxide (or other chemical species depending on the type of chemical sensor) concentration.

Furthermore, the present invention determines the actual concentration of the carbon monoxide which has been properly compensated to account for humidity without the need for an additional dedicated humidity sensor. Moreover, as the humidity signal (derivative in the second period) is synchronous with the carbon monoxide signal (derivative in the first period), it is not necessary to store the history of the sensor response in real-time in order to determine the actual concentration of the carbon monoxide.

The preferred embodiment utilises derivative curves to determine the actual concentration of carbon monoxide. This has an advantage in that no errors are introduced due to drift of the resistance of the sensitive layer.

Although the invention has been described with respect to compensating for humidity, the present invention may compensate for other contaminants, in addition or instead of humidity.