FR2940424A1 - DEVICE FOR MEASURING A PURELY RESISTIVE MATERIAL THICKNESS THICKNESS THRESHOLD, MEASURING METHOD, METHOD OF DIMENSIONING SUCH A DEVICE AND USE OF SUCH A DEVICE IN AN EXHAUST POT. - Google Patents

Abstract

The invention proposes a device for detecting a reliable and accurate deposit threshold thickness, making it possible to reduce the influence of the connections on the accuracy of the measurements of detection of a threshold thickness of purely resistive material, while being durably resistant to exhaust conditions, simple and economical. To this end, the subject of the invention is a device for detecting a threshold thickness of a layer of purely resistive material, comprising a sensor provided with at least three electrodes (100a, 100b, 200a, 200b) for defining at at least two pairs of electrodes (100, 200), each pair being arranged in a main longitudinal direction (D) on one side of a support (1). Each electrode comprises a sub-electrode (100a-100b, 200a-200b) connected to a voltage source (2) via connections (10), the voltage source (2) being set to generate a current (I , I) between the sub-electrodes, and means for measuring the resistances or intensities between the pairs of electrodes. The connections (10) are arranged on the same face of the support as the electrodes, and laterally with respect to the electrodes, in a transverse direction (D), substantially perpendicular to the longitudinal direction (D) of the electrodes.

Description

DEVICE FOR MEASURING THICKNESS THICKNESS OF PURELY RESISTIVE MATERIAL, MEASURING METHOD, METHOD FOR SIZING SUCH A DEVICE, AND USE OF SUCH DEVICE IN EXHAUST POT.

The invention relates to a device for measuring a threshold thickness of a layer of purely resistive material, which can be used in particular for detecting a deposit of soot in an automobile exhaust. The invention also relates to a method for measuring a threshold thickness, a method of dimensioning such a device and a use of such a device in an exhaust pipe. The permanent reduction of pollutant emission limits from automobile exhausts imposed by national and international standards is a major constraint for car manufacturers. The Euro 4 European Regulation of 2005 allows a maximum pollutant level of 0.025 mg / km, but the future European Euro 5 regulation, planned for 2010, will lower this threshold to 0.005 mg / km, which will require a systematic use of particulate filters (DPF) for all vehicles and in all EU countries. This regulation will therefore significantly increase current needs. Improving FAP control systems for increased performance is therefore one of the current priorities for engine manufacturers. The patent application FR 07/04221, in the name of the Applicant, describes a method and a device for measuring a threshold thickness layer of purely resistive material applicable, in particular, to the soot deposits in automobile exhaust. This application describes, moreover, two methods of sizing the two pairs of electrodes constituting the differential sensor so that the resistances measured between these two pairs are equal only when the thickness of the deposit reaches a predefined threshold value: an analytical method and a numerical finite element simulation method.

However, the soot is deposited on the entire sensor including between the longitudinal connections of the electrodes with the processing circuit, which disturbs the detection. To solve this problem, the invention described in the document FR 07/04221 proposes two solutions. The first is to make a connection through holes made through the substrate, these holes being called vias. But this method introduces additional complexity and cost because the connections are not in the same plane as the electrodes, but must pass through the vias and then be connected to the processing circuit under the substrate. This extra cost and complexity are incompatible with mass production in the automotive industry. The second proposed solution consists in separating the measurement zones from the connection zones by depositing insulating covers on the connections. These covers must be made of electrically insulating material and chemically inert with respect to connections, soot and exhaust gases, at temperatures at least equal to that of the regeneration sequences of the particle filter, ie approximately 800 ° vs. If this solution allows the realization of a sensor in which the connections and the electrodes are substantially in the same plane, the deposition of the covers is a complex and expensive step of the manufacture of these sensors. In addition, after long-term tests, the Applicant has found that even caches made from the most suitable materials deteriorate prematurely. It is therefore imperative that the deposition of purely resistive material on the connections does not significantly disturb the equality of the resistances to the threshold thickness, equality conditioned by the sizing of the electrodes. The present invention therefore aims at providing a reliable and precise device, making it possible to reduce the influence of the connections on the accuracy

measurements of detection of a threshold thickness of purely resistive material, while being simple, economical and resistant to the conditions of use of the exhaust pipes. In order to overcome the drawbacks of the preceding solutions, the present invention proposes a device for detecting a deposition threshold thickness by differential resistance measurements between at least three electrodes forming at least two pairs of electrodes, the length and / or the width and / or the supply voltage and / or the spacing are adjusted according to the thickness of deposit to be detected, and whose connections have a conformation to minimize the influence of the deposit of purely resistive material on these connections . To this end, the subject of the invention is a device for detecting a threshold thickness of a layer of purely resistive material, comprising a sensor provided with at least three electrodes for defining at least two pairs of electrodes, each pair being disposed in a main longitudinal direction on a face of a support. Each electrode comprises a sub-electrode connected to a voltage source via connections, the voltage source being set to generate a current between the sub-electrodes. The device also comprises means for measuring the resistances or intensities between the pairs of electrodes. The connections are arranged on the same side of the support as the electrodes and laterally with respect to the electrodes, in a transverse direction substantially perpendicular to the longitudinal direction of the electrodes. In other words, the sub-electrode is the functional part of the electrode, the geometric dimensions of which can be adjusted to ensure equal resistance to the threshold thickness. The connection is the co-planar portion, contiguous and substantially perpendicular to the sub-electrode, whose geometric dimensions are pre-defined, and which connects the sub-electrode to the processing circuit. An electrode is therefore the assembly formed by a sub-electrode and a connection.

The notions of longitudinal and transverse direction are valid only for each pair of electrodes. The two pairs of electrodes may have different directions, which may eventually save space. On the other hand, in the case of a system with three electrodes, all the sub-electrodes must be arranged in substantially parallel main longitudinal directions. In the remainder of the description, the designation of the dimensions is given with respect to the sub-electrodes. Thus, the length of the connections and sub-electrodes is the dimension taken along the longitudinal direction. The width of the connections and sub-electrodes is the dimension taken in the transverse direction, substantially perpendicular to the longitudinal direction. In other words, the length of the sub-electrodes corresponds to their largest dimension, while the width of the sub-electrodes corresponds to their smallest dimension. Conversely, the length of the connections corresponds to their smallest dimension, while the width of the connections corresponds to their largest dimension. Due to the fact that the connections extend substantially perpendicular to the sub-electrodes, the manufacture and implementation of the device are easy and economical. In addition, the device remains accurate despite the large width of the connections. According to other embodiments: the connections of the electrodes may have a length and a width such that the ratio between the length of the connections and the length of the sub-electrodes is less than 1: 5, preferably between 1: 10 and 1: 100, and that the ratio between the width of the connections and the width of the sub-electrodes is greater than 5: 1 preferably between 10: 1 and 50: 1; The device may furthermore comprise means for comparing the resistances or intensities with one another and means for generating a signal when the measured resistances or intensities are equal;

The pairs of electrodes may differ by at least one first parameter taken from the width and the spacing of the sub-electrodes, and a second parameter taken from the spacing, the width, the length and the adjustment of the voltage source supplying the sub-electrodes, the second parameter being such that, in use, the equality of resistance or of intensity is obtained when the threshold thickness to be detected is deposited on the electrodes; A width and / or a first spacing of the first pair of sub-electrodes may be such that, in use, the derivative of the current between the electrodes of said first pair with respect to the thickness of the layer tends to zero when the Threshold thickness achieved, and a width and / or a second spacing of the second pair of sub-electrodes may be such that, in use, the current between the electrodes of said second pair increases substantially linearly with the thickness of the layer. when the threshold thickness is reached, the widths and / or the lengths and / or the spacings of the sub-electrodes and / or the adjustment of the supply voltage source being adapted so that, in use, the equality of resistance or intensity is obtained when the threshold thickness is reached; The sub-electrodes may be arranged in parallel, interdigitated or mixed manner; The sub-electrodes of the first pair may have a width of between 100 nm and 1 cm, preferably between 10 μm and 1 mm, typically from 30 μm to 250 μm, and the sub-electrodes of the second pair may exhibit a width between 500 nm and 5 cm, preferably between 50 pm and 5 mm, typically from 250 pm to 1 mm; The ratio between the width of the sub-electrodes of the first pair and the width of the sub-electrodes of the second pair can be between 1: 1000 and 10: 1, preferably between 1: 100 and 1: 1, typically from 1: 10 to 1: 2;

The sub-electrodes of the first pair may have a spacing between 100 nm and 1 cm, preferably between 10 μm and 1 mm, typically from 30 μm to 250 μm, and the sub-electrodes of the second pair have a spacing between 500 nm and 5 cm, preferably between 50 pm and 5 mm, typically from 250 pm to 1 mm; The ratio between the first and the second spacing may be between 1: 1000 and 1: 1, preferably between 1: 100 and 1: 2, typically from 1: 10 to 1: 3; and / or the device may further comprise a heating resistor disposed on the support at least partially around the pairs of electrodes, so as to ensure a substantially complete combustion of the resistive material deposit on the sub-electrodes and the connections. .

The invention also relates to a method for measuring a threshold thickness of a layer of purely resistive material deposited on a sensor comprising at least three electrodes for defining at least two pairs of adjacent electrodes, each pair being arranged in a longitudinal direction. principal on one side of a support. Each electrode comprises a sub-electrode fed, via a connection disposed on the same face of the support as the sub-electrode and in a transverse direction, substantially perpendicular to the longitudinal direction of the sub-electrode, with a voltage defined generating a current between the sub-electrodes, the pairs of sub-electrodes differing by at least a first parameter taken from the width and spacing of each pair, at least a second parameter taken from the spacing, the width, the length and the supply voltage of the sub-electrodes being adjusted so that a first resistance or a first intensity between the electrodes of the first pair and a second resistance or a second intensity between the electrodes of the second pair are equal when the threshold thickness is reached.

According to a particular embodiment, a width and / or a spacing of the first pair of sub-electrodes are such that the derivative of the current between the electrodes of said first pair with respect to the thickness of the layer tends to zero when the Threshold thickness is reached, and a width and / or a spacing of the second pair of sub-electrodes are such that the current, between the electrodes of said second pair, increases substantially linearly with the thickness of the layer, when the threshold thickness is reached, the method further comprising the steps of: a) feeding the pairs of sub-electrodes with a defined voltage via connections disposed on the same side of the support as the sub-electrodes and according to a transverse direction, substantially perpendicular to the longitudinal direction of the sub-electrodes, b) measuring a first resistance or a first intensity between the electrodes of the first pair, c) measuring a second resistance or a second intensity between the electrodes of the second pair, d) comparing the second and first resistances or the first and second intensities, e) generating a signal when said resistors or said intensities are equal, the widths, and / or the lengths, and / or the supply voltages, and / or the spacings of the sub-electrodes being adapted so that this equality is obtained when the threshold thickness is reached. The invention also relates to an analytical method of dimensioning a preceding detection device, from pairs of theoretical electrodes, in which each theoretical electrode is decomposed, for calculation, into a sub-electrode SE and a C juxtaposed connection, the method comprising the steps of: a) calculating: 30 a1) ksE = tanh [7rs / 4e] / tanh [n (s + 2-esE) / 4e]

a2) kc = tanh [irs / 4e] / tanh [ir (s + 2 'c) / 4e], where s is the electrode spacing of the electrode pair, fSE is the width of the sub electrode, .ec represents the width of the connection and e represents the thickness of the deposit of purely resistive material, expressed in the same arbitrary unit ua; F31) calculating, for each pair of spacing sub-electrodes s, and of width fsE, the intensity ISE per unit length as a function of the thickness e by the following equations: IsE = (7r / 2) {ln [2 (1 + ksE ° '5) / (1-ksE °, $) J} - / for ksE 2 0.5 and IsE = (1 / 27r) ln [2 {1 + (] - ksE 2) 0 '25} / {I - (1-ksE2) ° '25} J for ksE2 <0.5; R2) calculate, for each pair of spacing connections s, and width ec, the intensity per unit length as a function of the thickness e by the following equations: IC = (n / 2) (ln [20 + kc ° '5) / (1-kc °' 5) J '-' for kc2? 0.5 and 1c = (1/270 ln [2 (1 + 0-kc9 + 25) / (] - (1-kc2) 0.25} J for kc2 <0.5;

X33) calculate, for each pair of theoretical electrodes, the total intensity Icot as a function of the thickness e by the following equation: Itot = LSE * ISE + Lc * Ic where LSE represents the length of the sub-electrode, Lc the length of the connection, expressed in arbitrary unit ua ', may differ from the arbitrary unit ua used to characterize the widths and spacings of the sub-electrodes and their connections as well as the thickness of the purely resistive material. y) plot the curve Itot = f (e) by varying the thickness e of the deposit of purely resistive material, b) among the plotted curves, choose two curves representative of a first pair and a second pair of electrodes theoretical

differing by at least one first parameter taken from the width and spacing of each pair of sub-electrodes; E) measuring, for each pair of theoretical electrodes, the intensities or resistances at a chosen threshold thickness and calculating the ratio of the intensities or the ratio of the resistances between the electrodes of each pair; C) manufacturing a sensor comprising two pairs of electrodes differing by said same at least a first parameter as the theoretical sub-electrodes, chosen in step S), and by a second parameter taken from the spacing, the width, the length and the adjustment of the supply voltage source of the sub-electrodes, the second parameter being such that, in use, the equality of resistance or of intensity is obtained when the threshold thickness to be detected is deposited on the electrodes , the sensor comprising connections disposed on the same side of the support as the sub-electrodes and laterally relative to the sub-electrodes, in a transverse direction substantially perpendicular to the longitudinal direction of the sub-electrodes. According to other embodiments: during step S), the representative curves of a first pair and of a second pair of electrodes differing by at least one first parameter taken from the width and the spacing of each pair of sub-electrodes may be chosen such that the derivative of the current between the electrodes of said first pair with respect to the thickness of the deposition layer tends to zero to the threshold thickness, and that the current, between the electrodes of said second pair increases substantially linearly with the thickness of the layer at the threshold thickness; In step S), the first and second pairs of theoretical sub-electrodes may respectively have a first and second width, but are of lengths, spacings and

identical supply voltages, and during the step), a sensor can be manufactured, comprising two pairs of sub-electrodes having respectively the same first and second widths as those of the theoretical sub-electrodes, lengths and spacings identical to those of the theoretical sub-electrodes, but in which the ratio between the applied voltage, in use, across the electrodes of the second and of the first pair, is equal to the ratio of the intensities or the resistances, measured at the threshold thickness chosen in the previous step E), between the pairs of theoretical electrodes; During step S), the first and second pairs of theoretical sub-electrodes may respectively have a first and a second width, but they are of identical lengths, spacings and supply voltages, and during step ), a sensor may be manufactured, comprising two pairs of sub-electrodes respectively having the same first and second widths as those of the theoretical sub-electrodes, supply voltages and spacings identical to those of the theoretical sub-electrodes, but in which the length SE2 of the sub-electrodes of the second pair satisfies the following equation: The sE2 * IsE2 (es) + Lc2 * Ic2 (es) = LSE1 * IsEl (es) + Lei * Here (es) with IsE2 (es), the intensity calculated between the sub-electrodes of the second pair at the threshold thickness, Lc2 the length of the connections of the second pair, lc2 (es) the calculated intensity between the connections of the second pair to threshold thickness, LsEl the lon of the first pair of electrodes, IsEl (es) the intensity calculated between the sub-electrodes of the first pair at the threshold thickness, Lc1 the length of the connections of the first pair, and Here (es) the intensity calculated between the connections of the first pair to the threshold thickness; In step 8), the first and second pairs of theoretical sub-electrodes may respectively have first and second spacings, but are of lengths, widths and voltages.

same step, and during the step), a sensor can be manufactured, comprising two pairs of sub-electrodes respectively having the same first and second spacings as those of the theoretical sub-electrodes, widths and lengths identical to those theoretical sub-electrodes, but in which the ratio between the applied voltage, in use, across the electrodes of the second and the first pair, is equal to the ratio of the intensities or resistances, measured at the threshold thickness chosen at step g) before, between the pairs of theoretical electrodes; During step b), the first and second pairs of theoretical sub-electrodes may respectively have first and second spacings, but are of identical lengths, widths and supply voltages, and during the step) a sensor may be manufactured, comprising two pairs of sub-electrodes respectively having the same first and second spacings as those of the theoretical sub-electrodes, widths and supply voltages identical to those of the theoretical sub-electrodes, but in which the length The sE2 of the sub-electrodes of the second pair satisfies the following equation: sE2 * IsE2 (es) + Lc2 * Ic2 (es) = LSE1 * ISE1 (es) + Lei * lci (es) with IsE2 (es), the intensity calculated between the sub-electrodes of the second pair at the threshold thickness, Lc2 the length of the connections of the second pair, IC2 (es) the intensity calculated between the connections of the second pair to the threshold thickness, LsEl the length r of the sub-electrodes of the first pair, IsE, (es) the intensity calculated between the sub-electrodes of the first pair at the threshold thickness, Lc1 the length of the connections of the first pair, and Here (es) l intensity calculated between the connections of the first pair to the threshold thickness. In step S), the first and second pairs of theoretical sub-electrodes may differ in width and spacing, but are of identical supply lengths and voltages, and

step C), a sensor may be manufactured, comprising two pairs of sub-electrodes respectively having the same first and second spacings and the same first and second widths as the theoretical sub-electrodes, lengths identical to those of the sub-electrodes theoretical, but in which the ratio between the voltage applied, in use, across the electrodes of the second and the first pair, is equal to the ratio of the intensities or resistances, measured at the threshold thickness chosen in step E ) before, between the pairs of theoretical electrodes; In step 8), the first and second pairs of theoretical sub-electrodes may differ in their width and their spacing, but are of identical lengths and supply voltages, and in step A sensor may be manufactured comprising two pairs of sub-electrodes having respectively the same first and second spacings and the same first and second widths as the theoretical sub-electrodes, identical supply voltages to those of the theoretical sub-electrodes, but in which length The sE2 of the sub-electrodes of the second pair satisfies the following equation: sE2 * IsE2 (es) + Lc2 * Ic2 (es) = LsE1 * IsE1 (es) + Lc1 * Ic1 (es) with IsE2 (es), the intensity calculated between the sub-electrodes of the second pair at the threshold thickness, Lc2 the length of the connections of the second pair, lc2 (es) the calculated intensity between the connections of the second pair to the threshold thickness, LSE1 the length sub-electrodes of the first pair, ISE1 (es) the intensity calculated between the sub-electrodes of the first pair at the threshold thickness, Lc1 the length of the connections of the first pair, and 1C1 (es) the intensity calculated between the connections of the first pair to the threshold thickness. The invention also relates to a use of the preceding measurement method, for detecting the deposition of a threshold thickness of soot layer in an exhaust pipe, in which the sub-electrodes are

fed laterally via connections disposed on the same side of the support as the sub-electrodes, in a transverse direction, substantially perpendicular to the longitudinal direction of the sub-electrodes.

According to a particular mode of use of the preceding measurement method, the signal generated in step e), when the first and the second resistance or intensity are equal, consists in controlling a particle filter regeneration step. Finally, the invention relates to an exhaust pipe provided with a particulate filter, comprising at least one preceding detection device, arranged upstream and / or downstream of the particulate filter for carrying out the process of the invention. previous measure. Other characteristics of the invention will be set forth in the following detailed description with reference to the appended figures which represent, respectively: FIG. 1, a diagrammatic plan view of a detection device according to the invention; FIG. 2 is a diagram illustrating an embodiment of the dimensioning method according to the invention, of adapting, for pairs of sub-electrodes of different widths and spacings, the length of one of the pairs as a function of the threshold thickness to be determined; FIG. 3 is a diagrammatic plan view of a pair of theoretical electrodes, each electrode being decomposed into a sub-electrode and a connection for the sizing calculation; FIG. 4, a diagram illustrating an embodiment of the sizing method according to the invention, of a sensor comprising electrodes provided with connections according to the invention; and FIGS. 5 and 6 are schematic plan views of sensors according to the invention, having different electrode geometries and a cleaning heating resistor.

The device for measuring a threshold thickness of purely resistive material according to the invention allows a differential resistance measurement or a differential intensity measurement of the layer of purely resistive material between two pairs of electrodes arranged on an insulating support. The expression two pairs of electrodes must be understood in a functional and non-structural sense, that is to say that the measurements can be made between two pairs of electrodes obtained with three physical electrodes, one of the electrodes belonging to to the two pairs of electrodes. The method essentially applies to electronically conductive or semiconductor materials, as opposed to dielectrics. Each pair of electrodes is characterized by the height h and the spacing s of the electrodes, by the width SE and the LSE length of the sub-electrodes, and by the width and the length Lc of the connections.

To obtain a differential resistance or intensity measurement, the widths and / or lengths and / or voltages and / or spacings are optimized with respect to the threshold thickness of the layer to be determined. An embodiment of a detection device according to the invention, illustrated in FIG. 1, comprises a sensor provided with two pairs of sub-electrodes 100 and 200, of different widths e1o1, e20, and spacings s102, s202. . The sub-electrodes 100a-100b, 200a-200b are arranged adjacently and in a main longitudinal direction DL on a face of a support 1. The device also comprises a voltage source 2 connected by wire bonds 5 to the sub-electrodes via electrical contacting connections 10 between the wires 5 and the sub-electrodes. Thus, a connection of a sub-electrode 100a, 200a of each pair 100, 200 is connected to the generator 2 set to deliver a defined voltage U, and a connection of the other sub-electrode 100b, 200b of each pair 100, 200 is connected to a measuring means 103, 203 of the resistors R1 and R2. This measuring means can be an ohmmeter or a

ammeter measuring the intensity of the current passing between the electrodes (the resistance is then obtained by Ohm's law). The device further comprises means for comparing (not shown) resistances or intensities with each other and means for generating (not shown) a signal when the measured resistances or currents are equal. The connections 10 are contact lines disposed on the same face of the support 1 as the sub-electrodes and which connect the latter to the rest of the electrical circuit by electrical connection son. These connections extend laterally with respect to the sub-electrodes, in a transverse direction DT, substantially perpendicular to the longitudinal direction DL of the sub-electrodes. The connections of the electrodes preferably have a length L and a width e such that the ratio between the length of the connections and the length of the sub-electrodes is less than 1: 5, preferably between 1:10 and 1. : 100, and that the ratio between the width of the connections and the width of the sub-electrodes is greater than 5: 1, preferably between 10: 1 and 50: 1. Due to this lateral arrangement and substantially perpendicular to the sub-electrodes, the connections 10 have very little influence on the accuracy of the detection. The voltage source is set to deliver a supply voltage between each pair of electrodes with the same voltage or with different voltages. The voltage generates a current between the electrodes of each pair when resistive material is deposited on the electrodes. According to the invention, the electrode pairs differ by at least a first parameter taken from the width and the spacing of the electrodes, and at least one second parameter taken from the spacing, the width, the length and the adjustment of the electrode. supply voltage source of the electrodes. The second parameter is such that, in use, the equality of

resistance or intensity is obtained when the threshold thickness to be detected is deposited on the electrodes. The widths 10 ,, 1201 and / or spacings 5102, S202 are chosen so that, on the one hand, the electric field lines resulting from the application of a voltage between the two electrodes 100a, 100b of the first pair of electrodes 100 remain mainly confined within the layer of purely resistive material and, secondly, for at least a portion of the electric field lines resulting from the application of a voltage between the electrodes 200a, 200b the second pair 200 remain outside the layer of material when the thickness of this layer has reached the threshold thickness es to be detected. Thus, as long as part of the electric field lines remain outside the layer of material, the current increases substantially linearly with the thickness of the layer. This evolution is illustrated on the curves C1 and C2 of FIG. 2 by the substantially linear portion C11, n-C211n of said curves. It is also said that the electrodes are not saturated. As the deposition thickness increases so that the electric field lines between two electrodes remain mostly confined inside the layer of material, the derivative of the current with respect to the thickness dl / de tends to zero. This evolution is illustrated on the curves C1 and C2 of FIG. 2 by the asymptotic part C1asy - C2asy of said curves. It is also said that the electrodes are saturated.

It suffices to vary only one of the parameters taken from the width of the electrodes or their spacing in order to obtain curves having distinct shapes, reflecting the fact that the first pair of electrodes develops the asymptotic character before the second pair. Preferably, to obtain accurate results while limiting the size of the electrodes and the amount of material used, the variation of said parameter is chosen for, on the one hand, that the first pair

of electrodes arrives at saturation before the deposition of the threshold thickness, that is to say that the derivative of the current between the electrodes of said first pair 100 with respect to the thickness of the deposition layer tends to zero to the threshold thickness es and, secondly, that the second pair of electrodes reaches saturation after the deposition of the threshold thickness, that is to say that the current, between the electrodes of said second pair 200 , increases substantially linearly with the thickness of the layer when the thickness of this layer has reached the threshold thickness es to be detected. To size the device according to the invention, it is necessary to determine the torque tto1-S102 such that for the first pair of electrodes 100 and the threshold thickness es, the derivative of the current with respect to the thickness dl / de tends towards zero, that is to say that the curve C1 has an asymptotic shape C1asy at the threshold thickness. It is also necessary to determine the torque t201-s202 such that for the second pair of electrodes 200 and the threshold thickness es, the current increases substantially linearly with the thickness of the layer, that is to say that the Curve C2 has a substantially linear shape C21, n at the threshold thickness. It is possible to determine these pairs E101-s102 and 1201-s202 by an experimental method or by an analytical method. This latter method is disclosed in the patent application FR 07/04221, in the name of the Applicant, for a detection device comprising a sensor provided with at least three electrodes defining at least two pairs of electrodes.

However, in this prior Application, the electrodes do not include lateral connections, but longitudinal connections. The mathematical equations given in this earlier application can not be applied as such to a device with side connections.

Indeed, in the previous method, when the length or the voltage were chosen as the second parameter, any one of these two parameters could be multiplied by the ratio of the intensities or the resistances between each pair of electrodes to a threshold thickness ( es) chosen.

With a device according to the invention, comprising lateral connections, only the voltage, as the second parameter, can be multiplied or divided by the ratio of the intensities or resistances to accurately detect the threshold thickness. When one chooses to adapt the length as the second parameter, it is not possible to multiply it by the intensity ratio calculated in step), without risking a certain approximation of detection of the threshold thickness. However, in some cases, this approximation is sufficient. For accurate dimensioning, the present invention provides a method in which the calculation is performed by decomposing each subelectrode of its lateral connection. This decomposition is shown diagrammatically in FIG. 3, with reference to the pair of theoretical electrodes 100. The sub-electrode thus has the width tSE and a length LSE equal to the total length Ltot of the electrode minus the length L, The connection has the width 1, and the length Lc, so that for the first pair of electrodes 100 "spacing s equal to 1u in the example illustrated in Figure 4a, each electrode is decomposed, for the calculation, in a subelectrode SE of length LSE equal to 0.95 ua 'and width tSE of 1 ua, and in a connection c of length L, equal to 0.05 ua' and of width f 40 ua . The curve I = f (e) of each pair of theoretical electrodes is then obtained as follows: a) calculate: a1) ksE = tanh [ns / Ie] / tanh [n (s + 2E) / 4e]

a2) kc = tanh [irs / 4eJ / tanh [7r (s + 2 'c) / 4e], In these equations, s represents the spacing of the electrodes of the electrode pair, tSE represents the width of the sub- electrode, tc represents the width of the connection and e represents the thickness of the deposit of purely resistive material, expressed in the same arbitrary unit ua This step is performed for each electrode of each pair. Then, in a step (3i), for each pair of spacing sub-electrodes s and of width fSE, the intensity per unit of length is calculated as a function of the thickness e, by the following equations:

PI) IsE = (n / 2) {ln [2 (1 + ksE ° 5) / (1-ksE ° 's) J} -' for ksE 2? 0.5 and 15 IsE = (1 / 27r) 1n [2 {1 + (1-ksE 0'251 / 0 -0 -ksE 2) ° '25} J for ksE 2 <0.5 Then, in a step P2, for each pair of spacing connections s and width fo, the intensity I is calculated per unit length as a function of the thickness e by the following equations: I ~ = (7r / 2) (ln [2 (1 + k ~ Ô)) / (1-k, ° .5) J '-' for k, 2? 0.5

And = (1/270 ln [2 {1 + (1-k, ')' L5} / {1- (1-k, 0.25) J for kc 2 <0.5 Then, in a step [33 for each pair of theoretical electrodes, the total intensity Itot is calculated as a function of the thickness e by the following equation: ## EQU1 ## where LSE represents the length of the sub-electrode, Lc the length of the connection, expressed in arbitrary unit ua ', may differ from the arbitrary unit ua used to characterize the widths and spacings of the sub-electrodes and their connections as well as the thickness of the material purely resistive.

Then, in a step y, the curve Itot = f (e) is plotted by varying the thickness e of the deposit of purely resistive material. In practice, a multitude of curves corresponding to a multitude of theoretical electrodes is drawn, by varying the spacing and the width of the electrodes. Among the curves drawn, in a step 8), two curves representative of a first pair and a second pair of theoretical electrodes differing by at least one first parameter taken from the width and spacing of each pair are chosen. . Then, in a step e), for each pair of theoretical electrodes, the intensities or the resistances are measured, at the chosen threshold thickness, and the ratio I "1 / I" 2 of the intensities 1 "1 is calculated, 1 "2 or the ratio R" 2 / R "1 of the resistors R" 2, R "I between the electrodes of each pair. Finally, in a step), a sensor comprising two pairs of electrodes 100 "and 200" differing by said same at least a first parameter as the electrodes chosen in step 8) (in the example, the width and the spacing) and by a second parameter taken from the spacing, width, length and supply voltage of the electrodes. The second parameter is such that, in use, the equality of resistance or intensity is obtained when the threshold thickness to be detected is deposited on the electrodes. In the manufactured sensor, the connections are arranged on the same face of the support as the sub-electrodes and laterally with respect to the sub-electrodes, in a transverse direction DT, substantially perpendicular to the longitudinal direction DL of the sub-electrodes. It is preferable, to limit their influence on the detection accuracy, that the connections have the same widths f and lengths Lo as those chosen for the theoretical electrodes.

When it is chosen to vary, as second parameter, the supply voltage, it will differ according to a ratio equal to the ratio of the intensities or the ratio of the resistances, measured at the threshold thickness in the previous step E) between the pairs of theoretical electrodes.

When it is chosen to vary the length of the sub-electrodes as a second parameter, the ratio of the lengths of the sub-electrodes will differ from that of the intensities or resistances measured at the threshold thickness in the previous step E). To calculate the length of the second pair of electrodes making it possible to determine with precision, when the threshold thickness es is reached, one looks for the sE2, the length of the sub-electrodes of the second pair, so that the ratio between the intensities between the two pairs of electrodes is equal to 1 to the threshold thickness. In other words, we look for the sE2 such that: sE2 * ISE2 (es) + Lc2 * Ic2 (es) = LSE1 * ISE1 (es) + Lc1 * lc1 (es) With IsE2 (es), the calculated intensity between the sub-electrodes of the second pair at the threshold thickness, Lc2 the length of the connections of the second pair, IC2 (es) the intensity calculated between the connections of the second pair at the threshold thickness, LSE1 the length of the sub -electrodes of the first pair, IsE1 (es) the intensity calculated between the sub-electrodes of the first pair at the threshold thickness, Lc1 the length of the connections of the first pair, and Ic1 (es) the calculated intensity between the connections of the first pair to the threshold thickness. The longer the length Lc of the connections is negligible compared with that of the sub-electrodes LSE, the more the ratio of lengths between the two pairs of electrodes of the sensor will be close to that of the intensities measured at the threshold thickness between the theoretical electrodes r (es) . Then, to multiply the length by the ratio of the intensities or the resistances will be able to give a sufficient precision.

Four embodiments of the sizing process according to the invention are described below.

According to a first exemplary embodiment, during step 8), the first and second pairs of theoretical electrodes respectively have a first and second width, but are of identical length, spacing and supply voltages.

According to a first variant of the first exemplary embodiment, during step c), a sensor is produced comprising two pairs of electrodes 100 "and 200" respectively having the same first and second widths as those of the theoretical electrodes, a length and a spacing identical to that of the theoretical electrodes, and in which the ratio between the applied voltage, in use, across the electrodes of the second 200 "and the first 100" pair, is equal to the ratio of the intensities I "1 / 1 "or resistors R" 2 / R "1, measured at the threshold thickness chosen in step y) before, between the pairs of theoretical electrodes. According to a second variant of the first exemplary embodiment, during the step), a sensor is made comprising two pairs of electrodes 100 "and 200" respectively having the same first and second widths as those of the theoretical electrodes, voltages d supply and spacings identical to those of the theoretical electrodes, and wherein the length of the second pair is calculated as follows: - the intensities of the first and second pairs of theoretical electrodes, designated respectively by the indices 1 and 2, are given depending on the thickness e of the layer, by: Itoti = LsEl * ISEI + Lcl * Here and Itot2 ù LSE2 * ISE2 + Lc2 * Ic2

the ratio of the intensities measured between the two pairs of theoretical electrodes to the threshold thickness of the layer is given by: r (es) = Itotl (es) / Itot2 (es) For the intensities or the resistances measured between the two pairs of electrodes of the sensor are equal when the threshold thickness of the resistive material layer is reached, it will be necessary to have: tot2 (es) = Iton (es)

where tot2 (es) is the intensity measured between the second pair of sensor electrodes at the threshold thickness. SEE 2 is sought such that the different lengths satisfy the following equation: SE2 * IsE2 (es) + Lc2 * Ic2 (es) = LSE1 * ISE1 (es) + Lc1 * lc1 (es) with IsE2 (es) ), the intensity calculated between the sub-electrodes of the second pair at the threshold thickness, Lc2 the length of the connections of the second pair, lc2 (es) the intensity calculated between the connections of the second pair to the threshold thickness, LSE1 the length of the sub-electrodes of the first pair, ISE1 (es) the intensity calculated between the sub-electrodes of the first pair at the threshold thickness, Lc1 the length of the connections of the first pair, and IC1 (es) the intensity calculated between the connections of the first pair to the threshold thickness. The solution of this equation makes it possible to determine the length of the second pair of sub-electrodes of the sensor to ensure the crossing of intensities or resistances to the threshold thickness. According to a second exemplary embodiment of the sizing method, in step a), the first and second pairs of theoretical electrodes respectively have first and second spacings, but are of length, width and supply voltages. identical. According to a first variant of the second exemplary embodiment, a sensor is produced during the step) comprising two pairs of electrodes 100 "and 200" respectively presenting the same first and second

second spacings than those of the theoretical electrodes, widths and lengths identical to those of the theoretical electrodes, and in which the ratio between the voltage applied, in use, across the electrodes of the second 200 "and the first 100" pair, is equal to the ratio of intensities I "1/1" 2 or resistors R "2 / R" 1, measured at the threshold thickness chosen in step y) before, between the pairs of theoretical electrodes. According to a second variant of the second exemplary embodiment, during the step), a sensor is made comprising two pairs of electrodes 100 "and 200" respectively having the same first and second spacings as those of the theoretical sub-electrodes, widths and supply voltages identical to those of the theoretical electrodes, and in which the length of the second pair is calculated as follows: the intensities of the first and second pairs of theoretical electrodes are given as a function of the thickness e of the layer, by: Itotl = LSE1 * ISE1 + Lc1 * lc1 and Itot2 = LSE2 * ISE2 + Lc2 * Ic2

The ratio of the intensities measured between the two pairs of theoretical electrodes to the threshold thickness of the layer is given by:

r (es) = Itot1 (es) / ltot2 (es)

In order for the intensities or the resistances measured between the two pairs of electrodes of the sensor to be equal when the threshold thickness of the layer of resistive material is reached, it will be necessary to have:

I'tot2 (es) = 'tom (es) 30

where tot2 (es) is the intensity measured between the second pair of sensor electrodes at the threshold thickness. we look for the sE2 such that the different lengths satisfy the following equation: sE2 * IsE2 (es) + Lc2 * Ic2 (es) = LSE1 * IsEl (es) + Lei * lci (es) with IsE2 (es) , the intensity calculated between the sub-electrodes of the second pair at the threshold thickness, Lc2 the length of the connections of the second pair, IC2 (es) the intensity calculated between the connections of the second pair at the threshold thickness , LSE1 the length of the sub-electrodes of the first pair, IsE1 (es) the intensity calculated between the sub-electrodes of the first pair at the threshold thickness, Lci the length of the connections of the first pair, and 1.1 (es) the intensity calculated between the connections of the first pair to the threshold thickness. The solution of this equation makes it possible to determine the length of the second pair of sub-electrodes of the sensor to ensure the crossing of intensities or resistances to the threshold thickness. If the length L, connections is negligible compared to that of the sub-electrodes LSE, then the ratio of lengths between the two pairs of electrodes of the sensor will be close to that of the intensities measured at the threshold thickness between the theoretical electrodes r (es ). In addition, if the currents measured between the electrodes of the second pair are in the linear range of the curves I (e) at the threshold thickness, then the ratio of the total length of the second pair of electrodes of the sensor (L 'sE2 + LC2) to that of the second pair of theoretical electrodes (LSE2 + Lc2) will be practically the same as the ratio of the intensities 1 "1/1" 2 or the resistances R "2 / R" t, measured, to the threshold thickness chosen in step y) of the preceding sizing process, between the pairs of theoretical electrodes. A third embodiment of the sizing process is illustrated in FIG. 4.

A device with lateral connections according to the invention is illustrated in FIG. 4a, and the diagram representing the evolution of the intensity between the pairs of electrodes 100 "and 200" of this sensor as a function of the thickness of deposit is represented. in Figure 4b.

Steps a) to b) are implemented with pairs of theoretical electrodes differing by two first parameters: the width and the spacing of the sub-electrodes. Thus, as an example for the pair of electrodes 100 ", Lot is calculated, in step R), with: sl = 1ua? SE1 = 1 ua LSE1 = 0.95 ua ', ic, = 40 ua Lc, = 0.05 ua '.

Then, in a step y, the curve ltot = f (e) is plotted by varying the thickness e of the deposit of purely resistive material. In the example illustrated in FIG. 4b, only the curves C4 and C, 00 'have been represented. For example, for the electrode pair 100 ", this curve is illustrated by the curve C100 ,, of Figure 4b.Also, the curve C4 illustrates, in this example of Figure 4b, the evolution of the intensity as a function of the deposition thickness for a pair of electrodes of the same length as those of the pair 100 ", but with sub-electrodes of different width and spacing. In this example, curve C4 was obtained by calculating ksE2 with s2 = 10 ua, ESE2 = 10 ua, LSE2 = 0.95 ua ', and k, 2 with s = 10 ua, Lc2 = 40 ua and Lc2 = 0 , 05 ua '. The electrodes corresponding to the curves C1oo., And C4 are therefore the theoretical electrodes. To manufacture the sensor for detecting the deposition of the threshold thickness when the intensities or the resistances between the electrodes of the sensor are equal, the SE2 length of the second pair of sub-electrodes is sought. SE2 must be such that the different lengths satisfy the following equation: sE2 * IsE2 (es) + Lc2 * Ic2 (es) = LSE1 * IsEl (es) + Lci * lci (es) with IsE2 (es) , the intensity calculated between the sub-electrodes of the second pair at the threshold thickness, Lc2 the length of the connections of the second pair, IC2 (es) the intensity calculated between the connections of the second pair at the threshold thickness , LsEl the length of the sub-electrodes of the first pair, IsEl (es) the intensity calculated between the sub-electrodes of the first pair at the threshold thickness, Lc, the length of the connections of the first pair, and lci ( es) the intensity calculated between the connections of the first pair to the threshold thickness.

Solving this equation yields a length of sE2 equal to 2.18. The sensor thus dimensioned thus comprises a first pair of electrodes 100 "of following dimensions: s1 = 1 ua, eSE1 = 1 ua, LsE1 = 0.95 ua ', fc1 = 40 ua and Lc1 = 0.05 ua', and a second pair of electrodes 200 "with the following dimensions: s2 = 10 ua, tsE2 = 10 ua, sE2 = 2.18 ua ', i2 = 40 ua and L2 = 0.05 ua'. By way of comparison, for a device without lateral connections, the length of the electrodes of the second pair would be equal to 2.19 units of length (ua '), while the total length of the electrodes of the second pair of electrodes 200 " with lateral connections is 2.23 units in length (ua '), ie 2.18 units in length for the sub-electrodes, plus 0.05 units of length of the connections. , this surplus length has almost no influence on the overall size of the sensor.

Indeed, when one compares the length LE (2.19 ua ') of the electrodes of the second pair of a device without lateral connections, to the length of the sE2 of the sub-electrodes of the pair 200 "(2, 18 ua '), one notices that the influence of the connections on the dimensioning of the electrodes is weak.So the juxtaposition, with the sub-electrodes, of lateral and substantially perpendicular connections of great width and short length, does not

changes only very slightly the dimensioning of the functional part of the electrodes, that is to say the sub-electrodes. In addition, for lateral connections of large electrodes it is useless to maximize their width (ie their largest dimension) or to minimize their length (i.e. their small dimension), since in the linear region of the curve, the current per unit length is practically the same for the connections and for the sub-electrodes. One could, for example, make connections in the form of a lateral square, side equal to the width of the sub-electrodes. This saves space. The increase in the size relative to the sensor without connection is therefore essentially related to the lateral connections of the small pair of electrodes, the size of the large electrodes can be very limited as explained above.

Thus, by a simple comparison of the intensities or resistances between the electrodes of the two pairs, it is possible to accurately detect that the deposition of purely resistive material on the electrodes has reached the threshold thickness. Equivalently, it would also have been possible to calculate the length SE of the sub-electrodes of the first pair of electrodes (the one whose electrodes are the least spaced apart), so that the different lengths satisfy the equation next :

LsE1 * IsEl (es) + Lc, * Here (es) = LsE2 * IsE2 (es) + Lc2 * Ic2 (es) with ISE, (es) the intensity calculated between the sub-electrodes of the first pair at threshold thickness, LcI the length of the connections of the first pair, and Ic, (es) the calculated intensity between the connections of the first pair to the threshold thickness, LSE2 the length of the sub-electrodes of the second pair, IsE2 (es), the intensity calculated between the sub-electrodes of the second pair at the threshold thickness, Lc2 the length of the connections of the

second pair, lc2 (es) the calculated intensity between the connections of the second pair to the threshold thickness. Equivalently, it would also have been possible to divide the voltage across the first pair of electrodes according to the ratio equal to the ratio of the intensities I ", / 1" 2 or the resistors R "2 / R" 1, measured at the threshold thickness chosen in step y) before, between the pairs of theoretical electrodes. Thus, according to the invention, it is necessary to vary at least two parameters, during the sizing phase of the sensor, so that the curves representing the change in intensity between the electrodes as a function of the thickness deposition intersect at the threshold thickness to be detected. In other words, it is appropriate to choose these parameters (length, width, spacing, supply voltage) so that the intensities or resistances measured between each pair of electrodes are equal when the thickness of the deposition layer has reaches the threshold thickness to be detected. According to a fourth embodiment of the sizing process, during step b), the first and second pairs of theoretical electrodes respectively have first and second spacings, but are of identical length, width and supply voltages. . During the step), instead of adapting the length or the voltage (as in the first and second exemplary embodiments), the measurement method according to the fourth embodiment consists in adapting the width or the spacing of the electrodes. depending on the thickness threshold to be determined. In other words, when the pairs of theoretical electrodes have electrodes of identical widths and differ only in their spacing, the method according to the fourth embodiment consists in adapting only the width of the sub-electrodes of the sensor so that the curves intersect at the threshold thickness. Similarly, when the pairs of theoretical electrodes are of identical spacings and differ only in the width of

electrodes, the method according to the fourth embodiment consists in adapting only the spacing of the electrodes of the sensor so that the curves intersect at the threshold thickness. Thus, the electrode pairs of the sensor differ only in the width and spacing of the electrodes. However, in this case, it is necessary that the two pairs of electrodes moreover have a different s / f ratio, otherwise the two curves representative of the evolution of the intensity I as a function of the thickness e deposition would have the same asymptote and consequently would never cross each other. The variation of two parameters to cross the curves of each pair of theoretical electrodes to the chosen threshold thickness may be sufficient for certain values of threshold thickness, but difficult or impossible to implement for other values of threshold thickness.

A preferred embodiment of the invention, which is easier to implement, consists in varying at least three parameters, as is the case for the example illustrated in FIG. 2. The best results are obtained for ranges of spacing and / or width in which the current flowing between the first pair of electrodes, when the threshold thickness es is reached, is in the asymptotic range of the curve C ,, whereas the current flowing between the second pair of electrodes, when the threshold thickness is reached, lies in the substantially linear domain of the curve C2. However, when the spacing is equal to the width for each pair of electrodes, the choice of s = of other than the threshold thickness is not imperative, although it is preferable. In general, it is appropriate to choose dimensions of the relatively distant electrode pairs in order to obtain a sufficiently large angle between the curves representative of the two currents (more exactly the tangents to each curve) to obtain a precise cross.

In addition, the advantages of the lack of insulating layer and the electrical contact on the same side of the support as the electrodes compensate for the surplus amount of conductive material to be used to make the electrodes and connections. Indeed, the device according to the invention is economical since it is necessary neither to make vias, nor to deposit insulating material. In addition, the device has improved resistance to aging and is technically easy to implement because the connections are practical.

In the different embodiments described, the width of the first pair of sub-electrodes 100 "can be chosen between 100 nm (nanometers) and 1 cm (centimeter), preferably between 10 pm (micrometers) and 1 mm (millimeter). typically from 30 μm to 250 μm The width of the sub-electrodes of the second pair 200 "may be chosen between 500 nm and 5 cm, preferably between 50 μm and 5 mm, typically from 250 μm to 1 mm. More generally, the width of the first pair 100 "may be chosen to be less than or equal to es / 2, preferably between es / 10 and es / 4. sub-electrodes 20 of the first pair 100 "and the width of the sub-electrodes of the second pair 200" can be between 1: 1000 and 10: 1, preferably between 1: 100 and 1: 1, typically 1: 10-1: 2. In addition, the electrode spacing of the first pair 100 "is selected between 100 nm and 1 cm, preferably between 10 pm and 1 mm, typically from 30 pm to 250 pm, and the spacing of the electrodes of the second pair 200 "is between 500 nm and 5 cm, preferably between 50 μm and 5 mm, typically from 250 to 1 mm, and preferably the ratio between the first and the second spacing is between 1 1000 and 1: 1, preferably between 1: 100 and 1: 2, typically from 1: 10 to 1: 3.

In the preceding description, the mathematical formulas are complex and a number of approximations have been made, in particular by neglecting the thickness h of the electrodes. These approximations have led to a certain inaccuracy when the threshold thickness to be detected is low, that is to say for threshold thickness values of the order of a few heights, in particular between 2 and 20, and when the spacing of the electrodes is small, especially for spacings smaller than the threshold thickness. When it is necessary to take into account the thickness h of the electrodes, the calculation of the curve I = f (e) is done differently depending on whether the thickness of the deposit is less than or greater than the thickness h of the electrodes . Thus, between a zero deposit thickness and a deposit thickness equal to the thickness h of the electrodes, the computation of the curve I = f (e) is done by calculating the current I by the usual law I = L * e / s, with L, the total length of the electrodes (sub-electrode + connection), and s their spacing. The applied voltage and resistivity of the layer are assumed to be unitary (as in all previously processed current equations). When the thickness of deposition is greater than the thickness h, the computation of the curve I = f (e) is done using the equations given previously.

In other words, when it is necessary to take into account the height h of the electrodes, the computation of the curves I = f (e) consists of translating the curves obtained in step (3): along the abscissa axis, the height of the electrodes, expressed in the same arbitrary unit ua as the spacings and widths, along the ordinate axis, of the intensity of current throughout the height of the electrodes, calculated by the classical law, is l = Lh / s, that is to say the ratio of the height of the electrodes by their spacing multiplied by their length.

The connection with the origin of the translated curves is done with a line of slope Lis. This slope is identical to the slope at the origin of the curves before translation.

Thus, by taking the curve of FIG. 4b, an electrode height of 0.2 ua would induce a translation of 0.2 ua along the abscissa axis, a translation of 0.2 intensity units along the axis ordinates for the solid line curve C1oo (s1 = sEi = 1u.a.; = = 40u.a.; Ltot = ..a. '= 0.95 ua' (LsE1) + 0,05 ua '(Lc1 )) and a translation of 0.02 intensity units along the y-axis for the dashed curve C4 (s = fsE = 10u.a; f = 40u.a.; Ltot = 1 ua '= 0, 95 ua '(SE2) + 0.05 ua' (Lc2)). The ratio of the intensities of the two pairs of theoretical electrodes, equal to 2.23 to the threshold thickness for a zero electrode height, would then have a value of 2.72 for an electrode height of 0.2 μ.a.

The material of the electrodes also influences the accuracy of the resistivity measurement. Thus, the electrodes consist preferentially of doped silicon, platinum, gold, silver-palladium or metal oxides when the atmosphere of use is corrosive, as for example in a muffler in use. . If the atmosphere permits, the usable materials may also be aluminum, copper, tin, etc. The electrodes can thus be made of various materials and varied, as long as their electrical resistivity remains negligible compared to that of the layer whose thickness is to be measured. In addition, it must be ensured that the electrical resistance of the electrodes is negligible compared to that of the layer whose thickness is to be measured. The accuracy of the measurement is all the more important that the manufacturing processes of the sensor, for example to deposit the electrodes on the insulating support, will be accurate. For example, screen printing allows only 5-10% sizing, especially for the most electrodes.

short and closest. Better accuracy could be obtained by depositing the electrodes by photolithography on alumina substrate. Figures 5 and 6 illustrate different shapes and arrangement of the pairs of electrodes. The two pairs of electrodes may, as illustrated in Figure 5, be rectangular. To solve the problems due to a large difference in lengths between the two pairs of electrodes, it can be envisaged that the electrodes of the first pair are rectangular and that the electrodes of the second pair are interdigitated (Figure 6). For highly resistive layers, two pairs of interdigitated electrodes are possible. A three-electrode design can also be considered, provided that the central electrode has the same length and / or width as that calculated during the design phase. The measurement method and the detection device can be used to detect the deposition of a threshold thickness of a layer of soot particles in an exhaust pipe comprising a particulate filter (DPF), and to control the monitoring of the regeneration of the FAP. For this purpose, the signal generated when the first and second resistors or intensities are equal is to control the particle filter regeneration. For this, at least one detection device according to the invention is disposed upstream of the particle filter for implementing the measuring method according to the invention. This arrangement therefore makes it possible to determine that a threshold quantity of soot has been supplied to the particulate filter. This regeneration, which is conducted at about 800 ° C., is not always sufficient to eliminate all the soot deposition on the sensor of the detection device according to the invention. To clean the sensor and completely eliminate soot deposition on the electrodes and, optionally, to ensure a sufficient operating temperature of the sensor, a heating resistor 500 is disposed on the carrier, at least partially around the electrode pairs

and connections, so as to provide substantially complete combustion of the resistive material deposit on the electrodes and the connections. The heating resistor is preferably platinum, but as for the electrodes, other conductive materials are conceivable. The arrangement of the heating resistor 500 should provide as homogeneous a distribution of temperature as possible at the soot deposit covering the electrodes (FIGS. 5-6). The use of the differential assembly assumes indeed the same resistivity value of the deposit at all locations on the surface covering the electrodes and therefore the same temperature for the two pairs of electrodes. The heating resistor may be deposited on the same side of the support as the electrodes, around them (Figures 5-6), or on the other side of the support, under the electrodes. The deposition technologies of the heating resistor are the same as those described for the electrodes.

The support must be chosen in order to: - withstand the drastic conditions of the automobile exhaust, in particular the strong temperature variations (100 to 900 ° C), the corrosive character of the exhaust gases, - be compatible at the electrical and mechanical level with the deposition of electrodes and with that of a heating resistor making it possible to ensure the combustion of the soot on the sensor, during the regeneration sequences of the FAP. The flat substrates made of 96-99.9% alumina, conventionally used for the deposition of thick or thin layers in hybrid microelectronics, seem well suited. Nevertheless, other insulating substrates are conceivable, such as ceramics, glasses, silicon oxide, magnesium oxide, zirconium oxide, aluminum nitride, silicon nitride, boron nitride etc. The support may consist of either an electrical insulator or a conductor or semiconductor 30 coated with an electrical insulator. It may be a metal coated with an electrical insulator, zirconia, silicon carbide, etc. For applications in

medium less hot and corrosive than that of the exhaust pipes, the support can also be plastic or monocrystalline silicon. The shape of the support is preferably flat, but it can be adapted according to the use as, for example, a curvilinear or cylindrical shape to be used in a muffler, or in gloves fingers. The sensors usable in automobile exhaust described above, can also be used in boilers, especially fuel oil, or in fireplaces. The interest is an optimization of the cleaning frequencies of the boilers and chimney sweeping. The minimum thickness that can be measured is of the order of magnitude of the width of the thinnest electrode. By using silicon technologies, it is therefore possible to consider measuring layer thicknesses starting from 0.1 μm. For some applications, it is possible to replace the quartz balance advantageously to control in situ thickness during deposition. The measurement method and the differential resistive sensor described in the present invention allow the detection of a threshold thickness of a purely resistive material, regardless of its resistivity. They are applicable to the determination of the thickness of any purely resistive material with finite electrical resistivity. Thus, the sensor is suitable for measuring the same threshold thickness in a wide range of resistivity without requiring recalibration as a function of the resistivity of the material and / or the deposition and measurement conditions (flow, temperature and pressure of the gases exhaust). The measuring method and the detection device according to the invention are particularly suitable for determining a threshold thickness of a particulate deposit, in particular a deposition of soot particles whose resistivity may be between 10 'and 10 + 8 ohm. depending on the composition of the soot (particularly its hydrocarbon content) and the temperature of the

deposition (the temperature of the engine and exhaust gases influences the composition of the deposit and its resistivity). In addition, its low cost of manufacture with materials proven in automobile exhaust, the compatibility of its dimensions with that of the pot, the simplicity and reliability of its operation in such hostile conditions, make the sensor perfectly meets the specifications car manufacturers. Many variants and alternatives can be made without departing from the invention and in particular: at least one detection device according to the invention can also be arranged downstream of the particulate filter. • The connections do not necessarily have the same lengths and widths between them.

Claims (24)

REVENDICATIONS1. Device for detecting a threshold thickness of a layer of purely resistive material, comprising a sensor provided with at least three electrodes (100a, 100b, 200a, 200b) for defining at least two pairs of electrodes (100, 200, 100 ", 200"), each pair being arranged in a main longitudinal direction (DL) on a face of a support (1), each electrode comprises a sub-electrode (100a-100b, 200a-200b, SE, SE1, SE2) connected to a voltage source (2) via connections (10, C, Cl, C2), the voltage source (2) being set to generate a current (II, 12) between the sub-electrodes , and means for measuring the resistances or intensities between the pairs of electrodes, characterized in that the connections are arranged on the same face of the support as the electrodes, and laterally with respect to the electrodes, in a transverse direction (DT) substantially perpendicular to the longitudinal direction (DL) of the electrodes. 2. Detection device according to claim 1, wherein the connections of the electrodes have a length (Lc) and a width (Cc) such as the ratio (Lc / LSE) between the length (Lc) of the connections and the length (LSE). ) of the sub-electrodes is less than 1: 5, preferably between 1:10 and 1: 100, and that the ratio (ic / iSE) between the width (tc) of the connections and the width (iSE) of the sub-electrodes is electrodes is greater than 5: 1 preferably between 10: 1 and 50: 1. 3. Detection device according to one of claims 1 or 2, further comprising means for comparing the resistances or intensities between them and means for generating a signal when the measured resistances or currents are equal. . 4. A sensing device according to any one of claims 1 to 3, wherein the electrode pairs differ in at least one of the sub-electrode width and spacing, and a second one of the spacing, the width, the length and the adjustment of the supply voltage source of the sub-electrodes, the second parameter being such that, in use, the equality of resistance or intensity is obtained when the thickness threshold to be detected is deposited on the electrodes. 5. Detection device according to claim 4, wherein a width (fSE1) and / or a first spacing (Si) of the first pair (100 ") of sub-electrodes are such that, in use, the current derivative between the electrodes of said first pair (100 ") with respect to the thickness (e) of the layer tend to zero when the threshold thickness (es) is reached, and a width (fSE2) and / or a second spacing (s2 ) of the second pair of sub-electrodes may be such that, in use, the current between the electrodes of said second pair increases substantially linearly with the thickness of the layer, when the threshold thickness (es) is reached, the widths and / or lengths and / or spacings of the sub-electrodes and / or adjustment of the source of supply voltage being adapted so that, in use, the equality of resistance or intensity is obtained when the threshold thickness (es) is reached. 6. Detection device according to any one of claims 1 to 5, wherein the sub-electrodes are arranged in parallel, interdigitated or mixed. 7. Device according to any one of claims 1 to 6, wherein the sub-electrodes of the first pair (100, 100 ") have a width (101, fSE1) of between 100 nm and 1 cm, preferably between 10 and 1 mm, typically from 30 μm to 250 μm, and the sub-electrodes of the second pair (200, 200 ") have a width (201, fSE2) of between 500 nm and 5 cm, preferably between 50 μm and 5 mm, typically 250 μm to 1 mm. 8. Device according to claim 7, wherein the ratio between the width (101, fSE1) of the sub-electrodes of the first pair 30 and the width (201, fSE2) of the sub-electrodes of the second pair is between 1: 1000 and 10: 1, preferably between 1: 100 and 1: 1, typically from 1: 10 to 1: 2. 9. Device according to any one of claims 1 to 8, wherein the sub-electrodes of the first pair have a spacing (102, si) between 100 nm and 1 cm, preferably between 10 pm and 1 mm, typically from 30 μm to 250 μm, and the sub-electrodes of the second pair have a spacing (202, s2) between 500 nm and 5 cm, preferably between 50 μm and 5 mm, typically from 250 μm to 1 mm. The device according to claim 9, wherein the ratio between the first (102, si) and the second (202, s2) spacing is between 1: 1000 and 1: 1, preferably between 1: 100 and 1: 2. typically 1: 10 to 1: 3. 11. Device according to any one of claims 1 to 10, further comprising a heating resistor (500) disposed on the support (1) at least partially around the pairs of sub-electrodes and connections, so as to ensure a substantially complete combustion of the deposit of resistive material on the electrodes and the connections. 12. A method for measuring a threshold thickness (e5) of a layer (3) of purely resistive material deposited on a sensor comprising at least three electrodes (100a, 100b, 200a, 200b) for defining at least two pairs (100, 200). , 100 ", 200") of adjacent electrodes, each pair being arranged in a main longitudinal direction (DL) on a face of a support (1), characterized in that each electrode comprises a sub-electrode (SE, SE1 , SE2) fed, via a connection (10, C, Cl, C2) disposed on the same face of the support (1) as the sub-electrode and in a transverse direction (DT), substantially perpendicular to the longitudinal direction (DL) of the sub-electrode, with a defined voltage (U, U1oo, U200), generating a current between the sub-electrodes, the pairs of sub-electrodes differing by at least a first parameter taken from the width and the spacing of each pair, at least one second parameter taken from the spacing, the width, the length and the supply voltage of the sub-electrodes being adjusted so that a first resistance (Ri) or a first intensity (II) between the electrodes of the first pair (100, 100 ") and a second resistance (R2) or a second intensity (12) between the electrodes of the second pair (200, 200 ") are equal when the threshold thickness (es) is reached. The measuring method according to claim 12, wherein a width ('ESE1) and / or a spacing (si) of the first pair (100' ') of sub-electrodes are such that the derivative of the current between the electrodes of said first pair with respect to the thickness (e) of the layer (3) tends to zero when the threshold thickness (es) is reached, and a width (isE2) and / or a spacing (s2) of the second pair of sub-electrodes are such that the current, between the electrodes of said second pair, increases substantially linearly with the thickness of the layer, when the threshold thickness (es) is reached, the method further comprising the steps of a) supplying the pairs of electrodes with a defined voltage (U, U, 00, U200) via connections (10, C, Cl, C2) arranged on the same face of the support as the electrodes and according to a transverse direction (DT), substantially perpendicular to the longitudinal direction (DL) of the electrodes, b) measuring a first resistance (RI) or a first intensity (II) between the electrodes of the first pair, c) measuring a second resistance (R2) or a second intensity (12) between the electrodes of the second pair , d) comparing the second and first resistances or the first and second intensities, e) generating a signal when said resistances or intensities are equal, the widths (fSE), and / or the lengths (LSE), and / or the voltages 30, and / or the spacings (s) of the sub-electrodes being adapted so that this equality is obtained when the threshold thickness (es) is reached. 14. An analytical method for dimensioning a detection device according to any one of claims 1 to 11, from pairs of theoretical electrodes, in which each theoretical electrode is decomposed, for calculation, into a sub-electrode SE. and a juxtaposed C connection, the method comprising the steps of: a) calculating: al) ksE = tanh [irs / 4eJ / tanhf (s + 2Esp) / 4eJ a2) kc = tanh [7rs / 4eJ / tanh [ir (s + 2ec) / 4e], where s represents the electrode spacing of the electrode pair, tsE represents the width of the sub-electrode, cc represents the width of the connection and e represents the thickness of the purely material deposit resistive, expressed in the same arbitrary unit ua; p,) calculate, for each pair of spacing sub-electrodes s, and of width ksE, the intensity ISE per unit length as a function of the thickness e by the following equations: ISE = (7r / 2) { ln [2 (1 + ksE ° '') / (1-ksE ')]} -' for ksE '? 0.5 and ISE = (1 / 2n) 1n [2 {I + (1-ksE2) + 25) / {1- (1-ksE ') D.2'}) for ksES 0.5; R2) calculating, for each pair of spacing connections s, and of width fc, the intensity per unit of length as a function of the thickness e by the following equations: le = (7r / 2) {ln [2 (1 + kc 0)) / (1-kc 0. ') J} -' for kc '? 0.5 and the = (1 / 27x) 171 [2 {1 + (1-kc ') 0 "' '} / {I (1-kc 2f 2'} J for kc '<0.5; R3) calculate, for each pair of theoretical electrodes, the total intensity Itot as a function of the thickness e by the following equation: Itot = LsE * Isr_ + Lc * lc where LSE represents the length of the sub-electrode, Lc la length of the connection, expressed in arbitrary unit ua ', which may differ from the arbitrary unit used to characterize the widths and spacings of the sub-electrodes and their connections as well as the thickness of the purely resistive material. Itot = f (e) by varying the thickness e of the deposition of purely resistive material, (0) among the curves drawn, choose two curves representative of a first pair and a second pair of theoretical electrodes 10 differing by minus a first parameter taken from the width and spacing of each pair of sub-electrodes; s) measuring, for each pair of theoretical electrodes ues, intensities or resistances at a chosen threshold thickness (es) and calculate the ratio (II / I2, I "t / I" 2) of the intensities (II, 12, ri, I "2) or the ratio (R2 / R1) resistors (R2, R4) between the electrodes of each pair; C) manufacturing a sensor comprising two pairs of electrodes differing by said same at least a first parameter as the theoretical sub-electrodes, chosen in step b), and by a second parameter taken from the spacing, the width, the the second parameter being such that, in use, the equality of resistance or of intensity is obtained when the threshold thickness to be detected is deposited on the electrodes, the second parameter being such that, in use, electrodes, the sensor comprising connections disposed on the same side of the support as the sub-electrodes and laterally with respect to the sub-electrodes, in a transverse direction (DT), substantially perpendicular to the longitudinal direction (DL) of the sub-electrodes. . 15. An analytical sizing method according to claim 14, wherein, in step b), the representative curves 30 of a first pair and a second pair of electrodes differ by at least a first parameter taken from the width and spacing of each pair are chosen so that the derivative of the current between the electrodes of said first pair (100 ") with respect to the thickness of the deposition layer (3) tends to zero to the threshold thickness (es), and that the current, between the electrodes of said second pair (200 ") increases substantially linearly with the thickness of the layer, the threshold thickness (es). 16. Sizing method according to one of claims 14 or 15, wherein: in step b), the first and second pairs of theoretical sub-electrodes may respectively have a first and second width, but are of length, identical spacings and supply voltages; and in step C), a sensor may be manufactured, comprising two pairs of sub-electrodes having respectively the same first and second widths as those of the theoretical sub-electrodes, lengths and spacings identical to those of the sub-electrodes theoretical, but in which the ratio (U200 / U, U / U100) between the voltage (U, U200, U100) applied, in use, across the electrodes of the second and the first pair, is equal to the ratio of the intensities (I1, 12, ri, I "2) or resistors (R1, R2), measured at the threshold thickness (es) chosen in the previous step s), between the pairs of theoretical electrodes. 17. Sizing method according to one of claims 14 or 15, wherein: in step a), the first and second pairs of theoretical sub-electrodes may respectively have a first and second width, but are of length, identical spacings and supply voltages; and in step c), a sensor may be manufactured, comprising two pairs of sub-electrodes respectively having the same first and second widths as those of the theoretical sub-electrodes, supply voltages and spacings identical to those of the theoretical sub-electrodes, but in which the length SE2 of the sub-electrodes of the second pair satisfies the following equation: The sE2 * IsE2 (es) + Lc2 * Ic2 (es) = LSE1 * IsEl (es) ) + Lc1 * lci (es) with IsE2 (es), the intensity calculated between the sub-electrodes of the second pair at the threshold thickness, Lc2 the length of the connections of the second pair, lc2 (es) the intensity calculated between the connections of the second pair at the threshold thickness, LSE1 the length of the sub-electrodes of the first pair, IsEl (es) the intensity calculated between the sub-electrodes of the first pair at the threshold thickness, Lc1 the length of the connections of the first pair, and Ic1 (es) the AC intensity lculated between the connections of the first pair to the threshold thickness. 18. Sizing method according to one of claims 14 or 15, wherein: in step a), the first and second pairs of theoretical sub-electrodes may respectively have a first and second spacing, but are of lengths; identical widths and supply voltages; and - during the step), a sensor can be manufactured, comprising two pairs of sub-electrodes respectively having the same first and second spacings as those of the theoretical sub-electrodes, widths and lengths identical to those of the sub-electrodes theoretical, but in which the ratio between the voltage applied, in use, across the electrodes of the second and the first pair, is equal to the ratio of the intensities or resistances, measured at the threshold thickness chosen in step s ) before, between the pairs of theoretical electrodes; 19. Sizing method according to one of claims 14 or 15, wherein: in step 8), the first and second pairs of theoretical sub-electrodes may respectively have a first and second spacing, but are of length, identical widths and supply voltages; and during the step), a sensor may be manufactured, comprising two pairs of sub-electrodes respectively having the same first and second spacings as those of the theoretical sub-electrodes, widths and supply voltages identical to those of the sub-electrodes. -oretical electrodes, but in which the length SE2 of the sub-electrodes of the second pair satisfies the following equation: sE2 * lsE2 (es) + Lc2 * Ic2 (es) = LsE1 * lsE1 (es) + Lc1 * lc1 (es) with IsE2 (es), the intensity calculated between the sub-electrodes of the second pair at the threshold thickness, Lc2 the length of the connections of the second pair, Ic2 (es) the calculated intensity between the connections of the second pair at the threshold thickness, LsE1 the length of the sub-electrodes of the first pair, IsE1 (es) the intensity calculated between the sub-electrodes of the first pair at the threshold thickness, LC1 the length of the connections of the first pair, and Ic1 (es) the calculated intensity e between the connections of the first pair and the threshold thickness. 20. Sizing method according to one of claims 14 or 15, wherein: in step 8), the first and second pairs of theoretical sub-electrodes may differ in their width and their spacing, but are of lengths and identical supply voltages; and in step C), a sensor may be manufactured, comprising two pairs of sub-electrodes respectively having the same first and second spacings and the same first and second widths as the theoretical sub-electrodes, lengths identical to those of theoretical sub-electrodes, but in which the ratio between the applied voltage, in use, across the electrodes of the second and the first pair, is equal to the ratio of the intensities or resistances, measured at the threshold thickness chosen at step s), between the pairs of theoretical electrodes. 21. The sizing method according to one of claims 14 or 15, wherein: in step 8), the first and second pairs of theoretical sub-electrodes may differ in width and spacing, but are of length and identical supply voltages; and in the step), a sensor may be fabricated comprising two pairs of sub-electrodes having the same first and second spacings and the same first and second widths respectively as the theoretical sub-electrodes, identical supply voltages to those of the theoretical sub-electrodes, but in which the length S e 2 of the sub-electrodes of the second pair 15 satisfies the following equation: sE 2 * IsE 2 (es) + Lc 2 * Ic 2 (es) = LSE 1 * IsE 1 (es) + Lcl * Ic1 (es) with IsE2 (es), the intensity calculated between the sub-electrodes of the second pair at the threshold thickness, Lc2 the length of the connections of the second pair, lc2 (es) l intensity calculated between the connections of the second pair to the threshold thickness, LsE1 the length of the sub-electrodes of the first pair, IsEl (es) the intensity calculated between the sub-electrodes of the first pair to the thickness threshold, Lc, the length of the connections of the first pair, and l 1 (es) the intensity calculated between the connections of the first pair to the threshold thickness. 25 22. Use of the measuring method according to one of claims 12 or 13, for detecting the deposition of a threshold thickness of soot layer in an exhaust pipe, wherein the sub-electrodes are fed laterally via of connections disposed on the same side of the support as the sub-electrodes, in a transverse direction (DT), substantially perpendicular to the longitudinal direction (DL) of the sub-electrodes. 23. Use of the measuring method according to claim 13, for detecting the deposition of a threshold thickness of soot layer in an exhaust pipe, during which the signal generated in step e), when the first and the second resistance or intensity are equal, consists in controlling a particle filter regeneration step. 24. exhaust pipe provided with a particulate filter, comprising at least one detection device according to any one of claims 1 to 11, arranged upstream and / or downstream of the particle filter for the implementation of the measuring method according to one of claims 13 or 14.