EP1565735A1

EP1565735A1 - Method and device for determining the water content and conductivity in the ground and in bulk materials

Info

Publication number: EP1565735A1
Application number: EP03772275A
Authority: EP
Inventors: Ruth Bernhard
Original assignee: Helmholtz Zentrum Muenchen Deutsches Forschungszentrum fuer Gesundheit und Umwelt GmbH
Current assignee: Helmholtz Zentrum Muenchen Deutsches F
Priority date: 2002-11-30
Filing date: 2003-10-30
Publication date: 2005-08-24
Also published as: US20050212532A1; DE10256064A1; US7126352B2; AU2003279330A1; DE10256064B4; WO2004051248A1

Abstract

The invention relates to a method and device for determining the water content and conductivity in the ground and in bulk materials. The aim is to disclose a method for the above, characterised by little technical complexity. Said aim is achieved with a method comprising an introduction of a capacitive sensor (4) into the ground or the bulk material, charging the sensor with a constant external voltage supply with a given initial current from a given first voltage threshold value to a given second voltage value and either determination of a voltage/time diagram, or measurement of the first charging time between both voltage threshold values necessary for the above and charging the sensor with the constant external voltage supply for determination of a second charging time, whereby either the initial current and/or at least one of the both voltage threshold values are altered and a determination of the water content and the conductivity by verification of the both charging times or of the voltage/time diagram.

Description

Method and device for determining the water content and conductivity in soils and bulk materials

The invention relates to a method and a device for determining the water content and conductivity in soils and bulk materials according to the first and sixth claims.

The dielectric constant ε _m between two electrodes serves as a reliable measure of the water content in a bed or a soil sample. Water itself has a very high dielectric constant compared to many other materials, so that even a low level of moisture in the measurement volume, ie in the soil sample or in the bed, leads to a significant change of ε _m . Typical areas of application for such moisture determinations can be found particularly in the monitoring of soils and in the quality control of agricultural cereal products (e.g. for flour, legumes, soybeans), but also in general in the production and conveying technology of moisture-sensitive goods (e.g. sugar ).

In addition to a determination of the water content in a bed or a soil via the dielectric constant ε _m , a parallel determination of the conductivity is also possible. The electrical conductivity of a moist material can be used as a measure of contamination, which has a significant influence on the electrical resistance of the moist material.

In this respect, a parallel measurement of the dielectric constant ε _m and the conductivity of a bed or a soil segment is suitable not only for the determination of the water content but also of the contamination dissolved in the water. The following methods for the simultaneous determination of a dielectric constant ε _m and a conductivity σ _m in materials are known for the characterization of soils and bulk materials:

In the so-called TDR method [1], differences in transit time along a waveguide (TDR probe) are used which generally change with the moisture in a material around this waveguide. The pulse runs along in a waveguide, a second waveguide serves as a reference mass. The dielectric constant is calculated from a measured pulse speed in the TDR probe and then, for example, the water content θ of a material is determined. An increase in the dielectric constant leads to a decrease in the pulse speed. The relationship between ε _m and θ must be assumed to be known. To operate and register the TDR signal, a cable tester is required, which must have a bandwidth of a few 4 GHz to over 100 GHz depending on the time resolution required.

[2] also describes the determination of the conductivity σ _{m of} a material using the TDR method from the damping of the impulse.

An evaluation of a TDR signal is complex, the required cable tester and an evaluation unit connected to it with the required resolution for runtime measurements are very expensive.

Alternatively, [3] proposes a capacitance method using an electromagnetic resonant circuit to determine the dielectric constant ε _{m of} the material, which can be implemented much more economically than the TDR method _. is. A sensor is designed as a capacitor with two capacitor electrodes connected to an oscillator. An oscillator, an analog electronic circuit, generates a sinusoidal output signal, the frequency f _{OSz of which is} determined by the capacitance of the sensor C _s , but also by other components. elements, in particular the necessary inductance L for the oscillator, is determined. The connection is

A higher dielectric constant is shown by a lowering of f _OS z

A frequency can be recorded precisely, reliably and inexpensively using simple means. As a result, exact determinations of dielectric constants are also possible with this method.

The frequency f _osz generally has a dependency on the conductivity of the material examined, an increasing conductivity generally leading to lower frequencies. A higher dielectric constant is exchanged in this way. However, this influence of the conductivity decreases with increasing frequency, whereby one preferably measures with high frequencies to determine the dielectric constant. Even if the influence of conductivity is reduced, it is still retained.

For a common determination of dielectric constant ε _m and

Conductivity σ _{m of} a material, the sensor can therefore be regarded as a complex resistance in an equivalent circuit diagram. A network analyzer is again required to determine the complex resistance with a real and an imaginary component, which increases the effort for a simultaneous detection of the water content and the conductivity accordingly.

The object of the present invention is therefore to provide a further method and a device for carrying out a method for determining the water content and the conductivity in soil and debris, which is also characterized by less effort. The object is achieved by a method and a device according to the features of the first and sixth patent claims. Preferred embodiments of the invention are the subject of the dependent claims.

The invention relates to a method for determining the water content and conductivity in soils and bulk materials. The method initially involves placing a capacitive sensor in the ground or in the bulk material. The sensor is then charged with an external constant current source, an initial current being adjustable and thus predeterminable. The capacitive sensor is charged from a predeterminable first voltage threshold value to a predeterminable second voltage threshold value, with either the voltage curve over time being recorded as continuously as possible and recorded as a voltage-time diagram, or the charging time required for this being measured between the two voltage threshold values. In the event that only the charging time and not the voltage curve over time is recorded, a second charging process of the sensor with the constant external voltage supply is required to determine a second charging time, for which purpose either the initial current and / or at least one of the two voltage threshold values has to be changed , The determination of the water content and the conductivity by verifying the two charging times or the voltage-time diagram.

If a voltage-time diagram is recorded, the verification is carried out by comparing the recorded voltage curve profile with the solution of a corresponding linear system of differential equations.

In the case of recording only the charging times based on two different parameter sets ^ which differ from one another in the voltage threshold values and / or the initial current, it is advisable to record each of the two charging times redundantly by periodically charging and discharging the sensor and to eliminate any malfunctions by averaging the individual charging times. If one uses an upper and a lower voltage threshold value as initial trigger for a discharge or Charging process for the sensor, the period of the measurement signal is composed of the charging and discharging time of the sensor. The period of the measurement signal is generated from the charging and discharging time of the sensor, the charging time having known discharging behavior being determined in a particularly advantageous manner with the aid of a simple and exact frequency measurement of the periodic loading and unloading. The two frequencies of the two periodic measurement signals are also used to determine the water content and the conductivity.

For the recording of the measurement signals, i. H. A storage oscilloscope, a transient recorder or a fast measurement data acquisition card, which forwards the acquired data to a process computer or a PC for further evaluation, is ideal for voltage-time diagrams, the periodic measurement signals or the charging times.

The invention is explained on the basis of embodiments with the following figures. Show it

Fig. 1 is an equivalent circuit diagram of the sensor in its basic form

2 shows the basic structure of the device,

Fig. 3 shows the equivalent circuit diagram of the sensor with insulated electrodes in the presence of a high conductivity of the dielectric.

4 a and b the simplified circuit diagrams of the device of a first embodiment and the switchable resistor 13 in the sensor branch, 5 a and b the simplified circuit diagrams of the device of a second embodiment and of the resistor circuit 18 in the reference branch, ie in the voltage divider branch,

6 the frequency signals determined as a function of the sensor capacitance,

7 a and 7 b, the frequencies determined, plotted against the conductivity of the soil or the bed, and

Fig. 8 several charging curves and an upper voltage threshold T in a voltage-time diagram.

Fig. 1 shows the equivalent circuit diagram of the sensor in a floor segment or a bed.

In principle, the conductivity σ _{m of} a material can be represented by an ohmic resistance R _s , which is connected in parallel to C _s . R _s and σ _m have the relationship

R _s = G / σ _m (2),

where G is a constant geometry factor of the sensor, which can be determined by a calibration measurement.

With their large dipole moments, which align in an electrical field, water molecules significantly influence the dielectric constant of water ε _w . The electrical conductivity in water is also not a metallic conductivity with displaceable charges and an opposite charge on the stationary phase, but an ionic conductivity. This means that in this case the applied electric field separates the charges due to the existing conductivity and pulls them further apart, so that a dipole is induced. The dipoles drift apart. follows until the resulting electric field of the separated ions compensates for the external field.

The aforementioned effect of the water molecules as dipoles is shown in the equivalent circuit. Fig. 1 by the capacitor 1 with the capacitance

_w ⁼ ε _ma {)

shown (dielectric constant ε _m , capacitance in air C _a ). C _{a is} the capacitance of sensor 1 or the capacitance of the space between the insulation when insulated electrodes are used when there is air in the space. If there is material in the gap with a dielectric constant ε _m , the total capacitance from equation (2) increases. Parallel to this is a series circuit consisting of the ionic resistor 2 with an ohmic resistance R _s = G / σ _m (Eq. (2)) and a further capacitor 3 with a capacitance

on ⁼ Sion ^ a V ^.

exists, which reflects the effect of ion separation, where ει _{on is given} by a function of σ _m :

ε _ion = g (σ _m ) (5).

In the first approximation, the relationship applies

ε _ion = α σ _m (6)

where α must be determined as the material constant for a soil to be examined or for a bed. α links the two sizes eats R _s and Ci _on , so that in addition to C _s, only one additional quantity has to be determined, namely R _s .

2 shows the basic structure of the device with a multi-vibrator circuit for determining the water content and the conductivity in a simplified circuit diagram. The sensor 4 forms together with the load resistor R _a 5 an RC element which essentially determines the time constant and thus the frequency of the circuit. The multivibrator circuit essentially consists of an operational amplifier as an active element 9, which is connected on the one hand with its signal output 6 to the load resistor 5 and the part of the voltage divider (resistor 11), on the other hand the voltage 7 between sensor 4 and load resistor 5 and the voltage 8 between taps two further resistors 10 and 11 (voltage divider). The time dependence of the charging curve (voltage 7) at the sensor 4 is determined by the capacitance value of the sensor, the resistor R 5 and _a further specification of circuit and the sensor. If the voltage 7 at the sensor 4 reaches a certain threshold voltage U _s ι, then the active element 9 switches over and the sensor 4 is discharged. At this stage, the voltage 6 changes to another constant value U ₂ . When a lower, ie second, threshold voltage U _s2 is reached, the discharge is ended, the active element is switched over again and the sensor is recharged. The output signal 6 is thus a digital square wave signal. With a smaller sensor capacity, charging and discharging processes are faster and the output signal shows a higher frequency.

In this device, the dielectric constant ε _{m of} a material is first determined using the capacitance method with the aid of the multivibrator circuit, the sensor 4 having the capacitance C _{s being} charged via a resistor R _a 5 and discharged again at a specific voltage. A periodically alternating charging and discharging results in a signal sequence with a frequency f, the time constant being determined by the RC element with R _a and C _s . The occurring frequency f f = h (R _a , C _s ) (7)

the capacitance of the RC element is thus influenced by the resistance, an increasing dielectric constant causing a decreasing frequency f. A digital signal is generated directly, which is less susceptible to interference and in which the frequency can also be easily determined. In addition, the circuit does not require an inductance L, which can also be disturbed from the outside.

However, a precise analysis shows that the frequency f is also influenced in this case by the conductivity σ _{m of} the material. The frequency f decreases as σ _{m of} the material increases. The decisive point of this dependency is no longer the frequency as in the case of the oscillator, but the resistance R _s , which is given by the conductivity of the material as above and which is parallel to the sensor capacitor C _s 1 and which leads to C _{s is} charged less, but the capacitor C _10n 3 is connected in series (FIG. 1). Increasing the conductivity leads to a lower R _s and also to a lower frequency. It is therefore not clear whether a lower frequency is due to an increase in the dielectric constant or an increase in the conductivity. A higher conductivity can therefore simulate a higher dielectric constant. Only if

R _a «Rs ⁽ 8)

then the frequency can be viewed independently of the conductivity. However, a small load resistance R _a causes a high frequency for a given sensor capacitance C _s . Given the specifications of the sensor, this condition cannot be met with the currently available electronic components, since measurement solutions with uncertainties in the determination of C _s and thus ε _m .

For a simultaneous determination of dielectric constant ε _m and conductivity σ _m , the device according to FIG. 2 has to be modified. The modifications relate to additional electrical circuits on the digital multivibrator circuit, with which various electrical properties Ei can be varied in a targeted manner. The determination is carried out with measurements with different settings Ei on the same examination object, several frequencies fi being determined. The properties of the material ε _m and σ _m can then be determined from these frequencies by calculation.

The expenditure on equipment is comparatively low in comparison to the TDR method, which uses expensive cable testers or the capacity method with the expensive network analyzer. Exact frequency measurements can be reliably achieved with comparatively simple means. A measurement with two settings Ei and E ₂ , ie a determination of two frequencies fi and f ₂ is completely sufficient for the solution of the task.

As described with reference to FIG. 1, the sensor 4 corresponds to a parallel connection of the actual sensor capacitance C _s 1 and the series connection of resistor R _s 2 and the capacitor Ci _θn 3, which is determined by the conductivity σ _{m of} the material. These electrodes of sensor capacitance 1 must be insulated if the material has a high conductivity. This insulation acts in the equivalent circuit as a series connection of two further capacitors Ci 12 (see FIG. 3). The additional components of the actual sensor capacitance C _s 1, ie the resistor R _s 2, the capacitor Ci _0n 3 and the capacitors C 12, significantly influence the time course of the charging and discharging processes and thus also the frequency of the output signal 6. By modification of the circuit in Fig. 2 it is now possible to

_Determine sizes of C _s 1 and R _s 2 together with Ci _θn 3. This enables simultaneous determination according to equations (3) and (2) of ε _m and σ _{m of} a material.

In the following, two ^embodiments' of the device for determining the water content and the conductivity in soils and solids, but also described other materials which each contain a corresponding modification.

The first embodiment is shown in FIGS. 4 a and b as basic circuit diagrams. The resistor R _a 5 according to FIG. 2 has been replaced in FIG. 4 a by a switchable resistor 13 with two switch positions. This switchable resistor 13 consists of two series-connected resistors 15 and 16 and a switch 17, which can be controlled by a signal HL 14, with two switch positions, one switch position bridging the resistor 16 (cf. FIG. 4b). The switchable resistor 13 thus realizes two different series resistors depending on the switch position, namely either a resistor R _a ι corresponding to the resistor 15 or a resistor R _a2 corresponding to the sum of the resistors 15 and 16. Two load resistors can therefore be set. The influences of R _al and R _a on the charging and discharging processes differ significantly from one another and, in the case of an unchanged sensor arrangement with an unchanged material, cause two different frequencies fi and

The alternative second embodiment is shown in Fig. 5 a and b. It differs from the first embodiment in that the modification to the circuit does not change the charge current, but rather the lower and upper voltage threshold values on the sensor. The resistors 10 and 11, which determine the threshold voltages U _s ι and U _s2 for switching the active element in the basic circuit according to FIG. 2 as a voltage divider in the form of the signal 8, have been replaced by a ^' resistance circuit 18. The resistor 20 therefore corresponds to the resistor 10 from FIG. 2, while the resistor 11 in the resistor circuit is replaced by a series connection of the resistors 21 and 22, the resistor 22 being designed to be bridged by a switch 23, controlled by a switching signal 19.

A circuit according to FIG. 4 a or 5 a is calibrated with an exchange of components of sensor 4, ie with capacitors Cj (j = l ... n) with a known capacitance instead of the sensor capacitance C _s 1 and with resistors R _k ( k = l ... m) instead of the resistor R _s 2 and dependent capacitors C _n for C ± _on 3. When calibrating with known capacitances C _j and known resistors R _k , functions are obtained as a result of the calibration

These functions, plotted in a diagram in FIG. 6 as frequency f in MHz over the capacitance C _s in pF characterize the circuit properties. The curve family 24 indicates the function with a small load resistance R _a ι, the capacitance C _{s being} varied according to the horizontal x-axis up to 800 pF and the conductance of R _s between 0 and 470 μS. The family of curves 24 is reduced to a curve in FIG. 6, which indicates that the conductance of R _{s has} no major influence on the function. The family of curves 25, on the other hand, shows a comparable measurement, but with a different capacitance and a resistance R _a2 , which has about 6 times the value of R _ai . In this case the function scatters on the diagram, ie the influence of R _s on the frequency is clearly visible. The functions (9) are such that the values C _s and R _{s can be} calculated at measured frequencies fi and f ₂ and the material properties ε _m and σ _m can thus be determined.

7 a and b show the data of the family of curves 24 and 25, plotted as frequency f in MHz (vertical y-axis) over the conductivity in μS (horizontal x-axis). The family of curves shown 26 and 27 represent the measurement results with different capacitances C _s . The family of curves 26, determined with on the basis of the aforementioned small working resistance R _al , runs almost horizontally, ie there is almost no dependence on R _s or σ _m . However, if the above-mentioned larger working resistance corresponding to R _{a2 is} used, a significant dependency on R _Ξ can be observed in the measurement results, ie in the family of curves 27, which also depends on the capacitance C _s considered.

The influence of the quantities C _s , R _s and Ci _θn on the frequency of the circuit is obvious from the given relationships and is given by the above function gι, ₂ (Cj, R _k ).

For real measurements of soils or fillings, this relationship is to be used as an inverse function for the determination of the quantities C _s and R _s from fi and f ₂ . The dimensioning of the resistors R _al and R _a2 for a calibration depends on the ratios of these real measurements. In this respect, R _al should be chosen as small as possible. This variable must also be adapted to the maximum possible current that the active element 9 can generate. In a first approximation, an average conductivity value must be assumed for a given material for the real measurement. With the design of the sensor, the geometry factor G is then determined from equation (2), so that an average value for R _s results from the previously taken conductivity value. For a calibration, the value R _a2 should correspond to this average value R _s if possible.

Fig. 8 shows several charging curves as a time course of the signal 7 in V (vertical y-axis) over time in μs (horizontal x-axis). Furthermore, a threshold voltage is indicated as a horizontal line with the voltage value T. Curve 28 shows the time course at C _Ξ = 80 pF and a conductivity of 0 (resistance R _s is infinitely large). Curves 29 and 30 correspond to the same capacitance C _s , but with resistances R _s = 4 kΩ and 40 Ω, respectively. Each of these curves reaches after a time shown as t _τ in the diagram Voltage threshold value T. All three curves 28 to 30 show an almost identical initial course in the lower region, but rise in significantly different ways at higher voltages. As a result, in particular in the embodiment according to FIGS. 4 a and b, the frequency explained at the outset, which is composed of periodically successive charging and discharging processes, is influenced by the resistance R _s as the voltage T increases as the threshold voltage for an interruption of the charging process ,

C _s and R _s , and with known α of C ± _on , from the two frequencies f _x and f ₂ can therefore be determined using the following two methods:

In the first method, the sensor circuit is calibrated by integrating calibration capacitances C _s and Ci _0n and resistors R _s , _k into the circuit and determining the respective frequencies f 1, _k and f _{2 k} . The measured frequencies fi and f ₂ can then be assigned on the basis of table values and, if appropriate, linear interpolation, optimized variables for C _s and R _s .

In the second method, the curve shape of a charging process (see FIG. 8) is assumed. With the aid of a linear coupled differential equation, a curve profile according to FIG. 8 can be described arithmetically and the respective frequencies fi and f ₂ can be determined with known threshold voltages. The sizes C _s and R _s are then determined by a fit procedure

In a specific application for the second method, a sensor is immersed in deionized water with a conductivity of 1 μS / cm (case d) and in normal tap water with a conductivity of 540 μS / cm (case n), the conductivities being in a conventional device of the prior art at a measuring frequency of 400 Hz. The frequencies fi and f ₂ in case d are determined as f _id = 1.542 MHz and f _2d = 2.622 MHz, for case n as fi _n = 1.257 MHz and f _2n = 2.216 MHz. It is obvious that the different conductivity influences the frequencies. With the aid of the second method described _above , the following variables can be determined: C _wd = 47 pF, C _wn = 48 pF, L _d = 1072 Ω, L _n = 1709 Ω,

Cion, d = 129 pF and Ci _0n , n = 205 pF, where the coupling _{coefficient is} α = 0.12. This shows that the values for the effect of the water molecules, indicated by the capacitance C _w , are almost the same and the conductivities L _d change according to the measurement using a different method. However, it must be taken into account here that the conductivities measured by the two methods mentioned are not linearly related. This is caused by the fact that the conductivity was measured at 400 Hz and the method described works at 2 MHz.

The charging curves according to FIG. 8 are verified for the determination of the water content and the conductivity by recording them with a transient recorder or another data storage unit for evaluation.

This is done on the one hand by calculating a charge curve with a system of differential equations and comparing it with the measured curve. In the event of a deviation, both the input value C _s and the input value R _{s are} changed (fit procedure). The corresponding values for C _s and R _{s are} used only when the measurement curve and the calculated curve match, in order to calculate the values ε _m and σ using equations 3 and 2.

Alternatively, the dielectric constant ε _{m of} the material and the conductivity σ result from two settings Ei and E ₂ , ie by determining two frequencies fi and f _2, ie two charging times. Here, too, the charge curve is calculated using the differential equation system for both Ei and E ₂ , from which the corresponding frequencies, ie voltage value T, are determined and compared with fi and f ₂ . In the event of a deviation, the input values C _Ξ and R _{s are} changed again and a new employed immediately. This fit procedure is repeated until there is a match between the calculated and measured frequencies within a specified tolerance.

In a simple process, the measured frequencies are compared with the calibration values f _1/2 = gι, ₂ (Cj, R _k ) (equation (9)) and the appropriate values C _s and R _{s are determined,} if necessary, by interpolation. The values ε _m and σ are calculated according to equations (3) and (2).

The following designs are suitable as capacitive sensors for determining the water content and conductivity using the aforementioned devices and methods:

Sensors, each with an insulated wire as an electrode, are particularly suitable for the depth-resolved measurement of soil water content. Ideally, each of the two wires consists of several sections which are parallel to one another and which are arranged in a plane such that each section is adjacent to one electrode with two sections of the other electrode. The resulting flat surface and the sensitive thickness, which depends on the wire distance, determines the sensitive volume of the sensor.

Sensors consisting of two insulated plates, which are inserted into the ground as electrodes in an upright position and parallel to one another, are preferably used when the depth resolution is irrelevant. The vertical installation of the plates minimizes the interference of the sensors with the water movement in the floor.

Sensors consisting of two or more rigid, insulated electrodes which are pierced into the material are particularly suitable for use in bulk goods. This arrangement of the electrodes in the bulk material essentially corresponds to that of a TDR probe according to the prior art. literature

[1] Topp, G.C., J.L. Davis, and A.P. Annan. 1980. Electromagnetic determination of soil water content: measurement in coaxial transmission lines. Water Resour. Res. 16: 574-582.

[2] Dalton, F.N., W.N. Herkelrath, D.S. Rawlins, and J.D. Rhoades.

1984. Time domain reflectometry: simultaneous measurement of soil water content and electrical conductivity with single probe. Science 224: 989-990.

[3] Eller, H. and A. Denroth. 1996. A capacitive soil moisture sensor. J. Hydrology 185: 137-146.

LIST OF REFERENCE NUMBERS

1 capacitor, sensor capacity

2 ohmic resistance R _s

3 capacitor, capacitance Ci _θn

4 sensor

5 load resistance R _a

6 output signal

7 tension

8 signal

9 active element, operational amplifier

10 resistance

11 resistance

12 capacitor, capacitance Ci

13 switchable resistor

14 signal HL

15 resistance

16 resistance

17 switches

18 resistor circuit

19 switching signal

20 resistance

21 resistance

22 resistance

23 switches

24 family of curves

25 family of curves

26 family of curves

27 family of curves

28 curve

29 curve

30 curve

Claims

Claims:

1. A method for determining the water content and conductivity in soils and bulk solids, comprising the following process steps: a) introducing a capacitive sensor into the soil or the bulk solids, b) charging the sensor with a constant external voltage supply via a predeterminable initial current from a predefinable one first voltage threshold up to a predeterminable second voltage threshold and either determining a voltage-time diagram or measuring the required first charging time between the two voltage thresholds and charging the sensor with the constant external voltage supply to determine a second charging time, with either the initial current and / or at least one of the two voltage threshold values has changed, and c) determining the water content and the conductivity by verifying the two charging times, with a periodic measurement signal for each of the two charging times a frequency is generated, whereby a period of the measurement signal is composed of the charging and discharging time of the sensor and the frequencies of the two periodic measurement signals are also used to determine the water content and conductivity, or the voltage-time diagram by comparing one recorded voltage curve with a solution of a corresponding differential equation system.

2. The method according to claim 1, wherein a transient recorder is used to record the voltage-time diagram.

3. Device for determining the water content and conductivity in soils and bulk materials, comprising a) a capacitive sensor in series with a series resistor for setting an initial current, b) a voltage divider connected in parallel with at least two resistors and a branch arranged in between for setting a voltage threshold value, and c) an operational amplifier with constant voltage source, with a feedback Part of the voltage divider and the series resistor at the two inputs and the output of the operational amplifier is connected, and the sensor and the rest of the voltage divider is connected between the two inputs and ground.

4. The device according to claim 3, wherein the series resistor is switchable between two resistance values.

5. The device according to claim 3, wherein a resistance of the voltage divider can be switched back and forth between two resistance values.