Method for testing of temperature sensors and a testing device Field of the invention

The present invention relates to a method and a device according to the preambles of the independent claims, and in particular to a method and a test device for a temperature sensor with negative temperature coefficient.

Background to the invention

The present invention is particularly intended for resistive temperature sensors, so-called thermistors, which are used in vehicles, but it may of course also be applied generally with resistive temperature sensors.

A thermistor is generally an electronic component, a resistor, whose resistance depends on temperature. There are many areas of application, e.g. in electronic thermometers, thermostats and overheating protection. There are thermistors with a very large number of resistance values and different sensitivities. There are ones with both positive temperature coefficient PTC and negative temperature coefficient NTC. The resistance of the positive type increases with rising temperature, and the resistance of the negative type drops when the temperature increases.

Power trains of trucks are currently provided with various resistive temperature sensors, thermistors, e.g. NTC, which has a negative temperature coefficient, and PT200, which has a positive temperature coefficient.

Figure 1 is a typical graph for an NTC thermistor, in this case of the 2 kQ/25°C type, which at the resistance 2 kQ therefore indicates a temperature of 25°C. With an electronic control unit it is possible to read off the resistance and thereby determine the temperature.

To measure the resistance, so-called voltage division is used in the control unit, as illustrated in Figure 2. Figure 2 is a simple schematic diagram of a temperature sensor with a temperature- sensitive resistor R2 and a resistor Rl with known resistance. The temperature sensor is connected to the voltage V _s and the current I flows through the resistors Rl and R2. The voltage U is measured across R2 and according to Ohm's law the following relationships then apply:

Vs=(I ⁺ R2)xI

I=U/R2

The following expression for the resistance R2 is thus arrived at: R2 = ^ The voltage U across the NTC resistor (R2) is thus measured when it is connected in series with a known resistor. This makes it possible to calculate the resistance and hence the temperature.

There are a number of legal requirements for heavy vehicles within the EU (e.g. Euro5 and Euro6), e.g. with regard to emission limits for exhaust gases, noise levels etc. Some of these requirements are related to certain temperature levels, e.g. there is no need to comply with emission requirements if for example the temperature is below +5°C. This in conjunction with the vehicle manufacturers' own needs results in requirements with regard to accuracy and reliability. Sometimes even redundancy (extra sensors) is required to meet legal requirements. There is also risk that less scrupulous

hauliers/haulage contractors might try to cheat by tampering with the temperature sensors so that they show errors, i.e. manipulation of the sensors. Here again there are legal requirements to protect the equipment against such practices. The patent specifications cited below refer to calibration and testing of thermistors.

US-2006/0104330 refers to a method and a device for calibrating a thermistor, particularly for a printer head in an inkjet printer. In simplified terms, this involves warming the thermistor by applying a constant voltage or a constant power and then measuring the resistance before and during the warming of the thermistor. Data for use in the calibration may thus be gathered. WO-98/55977 refers to an electronic heat detector consisting of a thermistor. By applying a voltage in the form of a test pulse for 10 seconds the function of the thermistor can be tested by seeing whether the thermistor becomes warm.

The object of the present invention is to improve the supervision of temperature sensors, particularly those used on vehicles, in order inter alia to prevent manipulation of such sensors and check their function. Summary of the invention

The above objects are achieved with the invention as it is defined by the independent claims.

Preferred embodiments are defined by the dependent claims.

The fact that according to the present invention the temperature sensor exhibits a resistance makes it possible to develop power in the sensor by applying across it a voltage pulse which has for example a magnitude of the order of 5 volts and a duration of not more than one second. When energy is supplied, the temperature of the sensor will also rise. The invention thus makes it possible to diagnose the sensor, i.e. to verify that it really exhibits a

temperature dependency. It is preferably also possible to calibrate the accuracy of the temperature measurement, and the offset voltage, and ensure that it shows correct temperatures.

In brief, the method according to the invention comprises

1. measuring the resistance,

2. applying a specified voltage for a specified time,

3. measuring the resistance and analysing how it changes after the voltage pulse applied. In one embodiment the difference between the resistances measured before and

after the voltage pulse applied to the temperature-dependent resistor is arrived

at, followed by analysing how the resistance changes after the voltage pulse by

comparing the measured resistance with a predetermined level related to the

difference. This predetermined level may for example be an increase in the

resistance of 63% relative to the resistance just at the end of the voltage pulse.

Measuring the time taken to reach this level makes it possible to determine

definitely the resistor's function.

In one aspect of the present invention, the testing of a temperature sensor takes

place at the same time as beginning to use the sensor, in order thereby to obtain

measured values which definitely identify the sensor and can be used in

calibration. These measured values may be stored for subsequent use in

comparison with values measured subsequently. Any subsequent measurement which shows that the values differ too much means that the sensor is no longer

working as intended.

The method and the test device according to the present invention afford inter

alia the following advantages compared with prior art:

• no need for redundant temperature sensors

• sensor can be diagnosed without using other sensors

• manipulation can be prevented

• sensor can be calibrated for more accurate measurement

The fact that the present invention involves analysis of how the resistance changes after the voltage pulse applied also affords substantial advantages compared with prior art. First measuring the resistance, followed by applying a specified voltage for a specified time and thereby raising the temperature of the sensor, thereafter measuring the resistance again and finally analysing how the resistance changes after the voltage pulse applied provides a substantially improved appraisal of the characteristics and function of the sensor. The fact that the present invention analyses how the resistance changes after the voltage pulse applied/the warming results in an appraisal which better correlates with how the sensor and its output signal are affected by the surroundings. As part of the present invention, the inventors have for example found that the location and configuration of the sensor, e.g. with respect to surrounding protective material of various kinds, may substantially affect the resulting characteristics and working of the sensor. The present invention takes this into account in an effective and robust way. As part of the present invention, the inventors have found that the prior art procedure whereby measurements, e.g. with respect to time constants, take place during the time when the sensor is being actively warmed does not provide a true picture of its resulting characteristics.

Brief description of drawings

Figure 1 is a schematic graph illustrating how temperature varies with resistance in the case of an NTC resistor.

Figure 2 is a schematic diagram of a known temperature sensor.

Figure 3 is a block diagram illustrating the test device according to the present invention when testing a temperature sensor.

Figure 4 is a schematic graph illustrating the present invention.

Figure 5 is a schematic graph illustrating the present invention.

Figure 6 is a flowchart illustrating the method according to the invention.

Detailed description of preferred embodiments of the invention

The invention will now be described in detail with reference to the attached drawings. Figure 3 is a block diagram illustrating the test device according to the present invention when testing a temperature sensor.

Figure 3 thus depicts a test device 2 for testing a temperature sensor 4, which sensor comprises a temperature-dependent resistor R2 with a resistance R and a negative temperature coefficient. The test device comprises a measuring unit 6 adapted to determining the resistance R of R2 at a first time tl and calling it R _tl . This is for example achieved by measuring the voltage across R2 and then calculating the resistance R of R2 as described above in relation to Figure 2. As the temperature depends on the resistance of R2 an appraisal of the temperature may also be obtained. The test device 2 further comprises a voltage supply circuit 8, a control unit 10, a switch 12 and a calculation unit 14. The switch 12 is adapted to being controlled by the control unit 10 so that the voltage circuit 8 is connected and applies to the resistor R2 a voltage pulse 16 (see Figure 4) with a predetermined voltage amplitude U _T and a predetermined duration T. Immediately after the voltage pulse, the resistance R of R2 is determined by the measuring unit 6 at a second time t2 which it calls R^. The measuring unit is then adapted to determining the resistance R of R2 after time t2 in order to identify how the resistance changes from level R _t2 to R _t i.

The resistance R is preferably determined substantially continuously after time t2. This may be done by sampling the resistance after t2 at a high enough sampling rate to ensure that changes in the resistance are identified.

The measuring unit 6 delivers the resistance values determined (R _t i, Rt2 and R) to said calculation unit 14 which is adapted to determining at least one measured value of a parameter which represents the change in the resistance after time t2, and to generating on the basis of said measured value an information signal 18 which is representative of the sensor's function.

In one embodiment, the measuring parameter is an amount of time t after time t2. The change in the resistance after time t2 is determined by calculating the absolute value of the resistance difference AR = | R _tl - Re | , calculating R _a = R _t2 + a x AR, where a is a predetermined constant between 0 and 1 , and determining a measured value t _a when the measured resistance R is equal to R _a . The value of a in one embodiment is 0.63. In the case of certain types of resistors R2 it may be more advantageous to adopt other values which more clearly reflect their function.

It is of course possible to indicate a number of different threshold values for the resistance after time t2 which may then be calculated by the same formula for R _a as above and which for example allow a to have ten values between 0 and 1. Corresponding measured values are then determined for t. The accuracy of the measurement is thus increased. The calculation unit 14 is adapted to comparing the information signal 18 generated, or parameters in the information signal, with a set of information signals, or parameters for information signals, which represent characteristics of different temperature sensors, and to generating on the basis of the comparison an identity signal 20 which represents the characteristic which best corresponds to the information signals of the sensor measured. Examples of parameters in the information signal may be single measured values, two or more measured values, specific curve shapes and the rate of change (i.e. the derivative) of the resistance. Figure 4 shows a curve representing the voltage pulse 16 and a curve representing the temperature change of R2 during the voltage pulse. The voltage pulse is the lower curve, which has a clearly square shape. Beginning and end times (tl and t2) are marked, and the pulse length T. The upper curve shows how the temperature changes in the resistor during the voltage pulse. It rises quickly initially but the rate of rise decreases and the curve levels out towards a stable value. At t2 the temperature starts to decrease and then drops back towards the previous level.

Figure 5 is a graph schematically illustrating the underlying principle of the present invention.

The upper graph in Figure 5 shows how the temperature changes when the voltage pulse is applied to the resistor R2. From a TO level the temperature rises to Tl at the end of the pulse before gradually returning to TO after the pulse. The lower graph in Figure 5 shows how the resistance through the resistor R2 varies when the voltage pulse is applied. Before the pulse, the resistance is at Rtl but at the end of the pulse at time t2 it has dropped to Rt2. The present invention analyses the pattern after the voltage pulse. To illustrate the principle, three different patterns designated A, B and C in the diagram are shown after the voltage pulse for three different resistors. These resistors exhibit different temperature behaviours reflected in different lengths of time for the resistance to increase from Rt2 to a predetermined resistance indicated in the diagram as Rta. Curve A reaches the resistance value Rta after ta seconds, curve B after tb seconds and curve C after tc seconds. As the time taken for the resistance to change from Rt2 to Rta is unique to each type of temperature-sensitive resistor, the function can be definitely determined by means of the present invention by comparing the time measured with a set of times for different types of temperature-sensitive resistors.

If curve A represents "approved" working of the resistor, the identity signal 20 will for example contain the information that the resistor is approved and what type of resistor it is.

The length (duration) adopted for the voltage pulse will depend on what function the resistor R2 has. Long pulse length risks destroying the resistor. One embodiment adopts a voltage pulse duration T of less than 1 second, e.g. within the range 5-50 ms, preferably about 10 ms.

The amplitude UT adopted for the voltage pulse is preferably less than 36 volts, which in vehicle applications is normally the maximum available voltage. One embodiment adopts voltage pulse amplitude U _T within the range 3-36 volts, more preferably 3-7 volts and preferably 5 volts.

The invention comprises also a method for testing a temperature sensor which comprises a temperature-dependent resistor (R2) with negative temperature coefficient and the resistance R. The method will now be described primarily with respect to the flowchart in Figure 6.

The method comprises the steps of

A - determining the resistance R of the resistor R2 at a first time tl and calling it R _t i, B - applying to the resistor R2 a voltage pulse with a predetermined amplitude U _T and a predetermined duration T,

C - determining the resistance R of the resistor R2 after the voltage pulse at a second time t2 and calling it R _t2 ,

D - measuring the resistance R of R2 after time t2,

E - determining at least one measured value of a parameter which represents the change in the resistance after time t2, and F - generating on the basis of said measured value an information signal which is representative of the temperature sensor's function.

Said measuring parameter is preferably the length of time t after time t2, and step E comprises

- calculating the absolute value of the resistance difference AR = | Ru - R__ | ,

- calculating R _a = R _t2 + a ^χ AR, where a is a predetermined constant between 0 and 1 , and

- determining a measured value t _a when the measured resistance R is equal to R _a .

In a preferred embodiment, a = 0.63.

It is of course possible to indicate a number of different threshold values for the resistance after time t2, which may then be calculated by the same formula for R _a as above and allows a to have for example ten values between 0 and 1. Corresponding measured values of t are then determined. The accuracy of the measurement is thus increased.

In one embodiment, the method comprises comparing the information signal generated at step F, or parameters in the information signal, with a set of information signals, or parameters for information signals, which represent characteristics of different temperature sensors, and generating on the basis of the comparison an identity signal which represents the characteristic which best corresponds to the information signal of the temperature sensor measured. Examples of parameters in the information signal may be single measured values, two or more measured values, specific curve shapes and the rate of change (i.e. the derivative) of the resistance. The above discussion in connection with the description of the test device applies also to the method with regard to the duration and amplitude of the voltage pulse.

The present invention is not restricted to the preferred embodiments described above. Sundry alternatives, modifications and equivalents may be used. The above embodiments are therefore not to be regarded as limiting the invention's protective scope which is defined by the attached claims.