FR2790313A1 - Non contact inductive sensor for measuring rectilinear or angular displacements - Google Patents

Abstract

The inductive sensor has a magnetic circuit (2) with an airgap and excitation winding (1). A movable part (3) slides in the airgap and a measuring circuit (4,5,v), see Fig. 2a,2b, measures the reluctance of the magnetic circuit so as to deduce a position of an object mechanically linked to the moving part (3). the part (3) has a section, measured perpendicular to the sense of sliding which is variable along the slide direction. The excitation winding can include a series connected primary winding (7) see Fig. 5, and a secondary winding (8) mounted respectively on the first and second pole either side of the airgap. The measuring circuit includes an a.c. and d.c. current source connected in parallel with the winding. The circuit also includes: - a low and high pass circuits and a multiplier circuit; - an analogue - digital converter which has a reference input connected to the low pass circuit.

Description

I The subject of the present invention is a contactless inductive sensor for

  the measurement of rectilinear or angular displacements. The measurement of rectilinear or angular displacements already calls upon many technologies which each answer a particular need in terms of cost, precision or environment. Among the known devices, mention may be made of potentiometers, optical encoders, capacitive sensors, magnetostrictive sensors and inductive winding sensors. There is also a variant of an inductive sensor in which the windings are reduced to simple printed circuit loops. In a patent document FR-A-2 682 760, a printed circuit displacement sensor is described. This sensor sometimes encounters difficulties of use either because the accessibility to the moving parts is difficult and the disassembly for repair is very complicated, or because it is often necessary to install electronic components as close as possible to the sensor. that the environment presents very strong disturbances

  radioelectric or intense ionizing radiation.

  The subject of the present invention is a displacement sensor which has substantially the same advantages as such a

  printed circuit device by providing, moreover, great ease of interchangeability and the absence of active electronic components in the immediate vicinity.

  The object of the invention is in fact essentially to solve assembly and disassembly problems which are difficult to dissolve with the above-mentioned printed circuit sensor, to remove electronic components which must be installed on the circuit of this sensor, and to improve its insensitivity to the climatic and radioelectric environment. The displacement sensor according to the invention essentially consists of one or more windings wound on a magnetic circuit in the gap of which moves a mobile element, ferromagnetic or not, of variable section. The mobile element is inert and therefore is not sensitive to temperature influences or electromagnetic influences. The invention therefore relates to an inductive sensor characterized in that it comprises a magnetic circuit with an air gap and an excitation winding, a movable element sliding in this air gap, and a circuit for measuring the reluctance of this magnetic circuit to deduce a position from an object mechanically connected to this mobile element, the mobile element having a section, measured perpendicular to the direction of sliding, which is variable

with this shift.

  The invention will be better understood on reading the

  description which follows and upon examination of the figures which

  accompany him. These are presented for information only and in no way limit the invention. The figures show: Figure 1: a schematic representation of the sensor of the invention; Figures 2a to 2b: a schematic representation of the principles of electrical supply of the winding of the invention; Figures 3a, 3b and 4: exemplary embodiments of mobile elements usable in the sensor of the invention; Figure 5: an improvement of the winding of the invention; Figures 6 to 8: an improved power supply of the winding of the invention and a corresponding measurement circuit; Figures 9 to 12: a representation of the effects of a splitting of the poles of the sensor of the invention to achieve drift compensation; Figure 13a to 13e: mobile elements usable with split air gaps according to Figures 9 to 12; Figure 14: a measurement circuit usable with the embodiments of Figures 9 to 12; Figure 15: a representation of the movable element allowing the measured position to be an angular position; Figure 16: a schematic representation of the sensor of the invention with a second magnetic circuit and a second air gap in which slides a mobile element allowing a measurement in rotation with a drift compensation; Figures 17 to 19: mobile elements usable with a sensor with two air gaps and allowing drift compensation; Figure 20: a mobile element usable in the context of Figure 16; Figure 21: an example of use of the sensor of the invention; Figure 1 shows how a device according to the invention can be arranged in its simplest design. A winding 1 forming an excitation winding is wound around a C-shaped magnetic circuit 2

  The opening of the C forms the air gap of the magnetic circuit. A mobile element 3 can move in the air gap of the magnetic circuit, preferably free of any material contact.

  Powered by alternating current, the winding 1 has an impedance, the value of which depends on the reluctance of the magnetic circuit. This reluctance is itself a function of the section of the movable element interposed in the air gap. As a result, the impedance of the coil is a function of the position of the movable element 3 in the air gap of the magnetic circuit. To measure displacements of an object, we link element 3 to this object. Element 3 is inert and easily supports the above-mentioned environmental constraints. Furthermore, the element 3 being devoid of material contact with the poles of the air gap, its disassembly from the object does not pose

of difficulty.

  A simple means of measurement, but not very stable in temperature, consists, as shown in FIGS. 2a and 2b, of supplying the winding either by a generator 4 of alternating current i and of measuring a voltage variation v directly at the terminals of the coil , or to supply the winding with a DC voltage generator 5 and to measure a voltage variation v across a resistor R. These two methods, if they make it possible to understand in a simple way how the displacement of the mobile element can be translated into a measurable voltage can be perfected for their practical implementation. Below are described

  other means more suitable for obtaining a stable measurement.

  The mobile element 3 can be made of non-magnetic conductive material in which case it develops, in a small thickness, eddy currents which oppose the

  passage of the magnetic flux which circulate in the air gap.

  This has the effect of obtaining a reduction in the impedance of the coil when the section of the movable element, has

in the air gap, increases.

  The movable element 3 can also be made of ferromagnetic material permeable to flow. In which case the impedance of the coil increases when the section of the movable element increases during its movement in the air gap of the magnetic circuit. The eddy currents essentially developing on the surface of the conductor, the thickness of the movable element has little effect. It just needs to be thick enough for currents to flow through it. This is called the skin effect. On the other hand, in the operating mode with variable winding permeability, the flux in the air gap depends as much on the width as on the thickness of the ferromagnetic mobile element. There are therefore at least two possibilities for producing the movable element, either the thickness is constant and the width varies (FIG. 3a) or the width is constant and it is the thickness which

  is variable (Figure 3b). Alternatively, the two vary.

  Other solutions can also be implemented to vary the cross section of the magnetic flux in the air gap when the movable element is moving. Thus the movable element while retaining a constant thickness and width 5 can be hollowed out with an opening which more or less allows the flow to pass depending on the position of this movable element. FIG. 4 represents a cursor of this type which can also be made either of electrically conductive material or of material

ferromagnetic.

  As a variant, a mobile element in FIGS. 3a, 3b or 4

contains both materials.

  As the coil 1 is arranged in the magnetic circuit 2 shown in Figure 1, the device has a large leakage flow which disturbs the measurement of the movement of the movable element. A better arrangement is to separate the winding into two parts and place each half-coil 7 and 8, of Figure 5, directly around

  poles of the magnetic circuit 2.

  In addition to reducing the leakage flow, the arrangement of the coils as shown in FIG. 5 allows the sensor to be used in two different operating modes. The first mode, identical to that described above, consists in measuring the variation of the total impedance of the two coils connected in series when, under the action of the displacement of the mobile element, the reluctance of the magnetic circuit varies. This operating mode is said to

variable reluctance.

  The second mode consists in using one of the coils as the primary of a transformer and the other coil as the secondary. By moving the mobile element changes the coupling between the two coils. This has the effect of varying the voltage induced in the secondary and thus obtaining a measurement proportional to the displacement of the element.

mobile.

  In both operating modes, the sensor remains sensitive to temperature variations. One means of reducing this sensitivity may consist in placing a temperature probe as close as possible to one of the two coils, and in making a correction to the measurement voltage so that the variations attributable to the temperature are minimized. Aside from the fact that this process complicates the sensor somewhat by adding a probe and by the presence of additional wires, the compensation obtained is only imperfect since the probe never exactly measures the temperature of the coils. The essential interposition of insulating materials always causes a temperature gradient between the core of the coils and

the sensitive element of the probe.

  A better method according to the invention consists in using the fact that any winding made of electrically conductive material (excluding superconductive materials) consists of two main elements: a choke and a resistor. This resistance is none other than that of the wire used to make the coil. In current embodiments, the windings are executed by means of copper wires whose resistance is practically proportional to the temperature with a coefficient of variation equal to 0.0039 / C. The measurement of the resistance of the (or) coil of the sensor makes it possible to correct the influence of the temperature without adding any additional measuring means. In addition, this correction is all the more effective as the information gathering takes place at the very heart

of the active element.

  The main part of the process therefore consists in separating, on the one hand, the variation of the self-function as a function of the displacement of the mobile element and, on the other hand, the variation of the resistance proportional to the temperature. This is obtained by running the winding (s) through two superimposed currents. One of the currents is alternating, and its amplitude is a function of the variation of the inductor and therefore of the displacement of the mobile element. The other current is continuous,

  and its value is a function of the temperature variation.

  The variations of one and the other of the two currents are measured separately. We deduce from the measurement of this other current the value of the temperature with which we correct the measurement

of the first current.

  FIG. 6 shows by an equivalent diagram the manner in which a direct current ic coming from a generator 9 of direct voltage is superimposed on an alternating current, ia, coming from a generator 10 of alternating voltage. The two generators are connected to the terminals of the winding 11 of the sensor. A low-pass filter composed of a resistor R3 and a capacitor C1 connected in parallel with the winding 11 makes it possible to extract at its terminals a continuous component Vc proportional to the temperature. A high-pass filter comprising a capacitor C2 in series by a terminal with the winding 11 makes it possible to collect, by its other terminal, only

  the alternating voltage Va proportional to the displacement.

  Among the various correction methods that can be

  imagined, there are two whose simple implementation perfectly meets the desired goal.

  A first correction method consists in introducing an analog multiplier device into a signal conditioning stage proportional to the displacement measurement. As shown in FIG. 7, the product of a voltage proportional to the displacement by a determined fraction of the voltage proportional to the temperature makes it possible to correct the drift. Thus a multiplier circuit acts as a variable gain amplifier controlled by a

voltage.

  In this case, the two generators 9 and 10 are coupled to the winding 11 by a differential amplifier 12. The measurement signal taken at the terminals of this winding 11 is injected into an amplifier 13. The output of the amplifier 13 is connected by the high-pass filter C1-C2 to an envelope detector 14 which measures the rms value. It is also connected by the low-pass filter R3-Cl to an amplifier 15, the output of which is connected to a potentiometer R. This potentiometer R enables the measurement chain to be tared. The other terminals of the potentiometer R are connected to ground and to a first input of a multiplier 16. The other input of the multiplier 16 receives the output signal from the amplifier 14. The output of the multiplier 16 delivers the measurement signal corrected. Filter cutoff frequencies are lower than signal frequency

alternative power.

  The second correction method implements an analog-to-digital conversion of the voltage proportional to the displacement. The correction thus operates digitally. The assembly of FIG. 8 is similar to that of FIG. 7. It suffices that the reference voltage of an analog-digital converter 16.1 connected in place of the multiplier 16 is subject to the DC voltage proportional to the temperature. By this device a binary number obtained at the output of the converter 16.1 is the image of the displacement of the mobile element corrected for a possible

temperature drift.

  Another means of combating the temperature drift of a displacement sensor consists in giving it a geometry such that the measurement is the result of the difference of two voltages which vary in opposite directions. We then obtain a differential measurement. This process shows that to the decrease in drift due to temperature is added an improvement in linearity, and a lower sensitivity to electromagnetic disturbances

(parasites).

  This is achieved for example by dividing the pole in half

  bottom of the magnetic circuit 17 shown in FIG. 9.

  Each half-coil 19 and 20 receives half of the magnetic flux produced by the excitation coil 18. There are then two air gaps and two magnetic circuits, one passing through the pole 19, the other passing through the pole 20. By analogy 9 with the other inductive sensors with rectilinear or angular differential transformers, we can say that the winding 18 constitutes the primary and that each winding 19 and 20 is a secondary.5 Figure 10, voluntarily rid of its primary and secondary windings, shows that the single movable element 3 in FIG. 1 has been divided into two inverted sectors (otherwise mechanically linked) so that when one of the interposed surfaces10 moves in one of the air gaps increases while it decreases in l 'other. Figure 11, voltages V1 and V2 are induced across each secondary 19 and 20. By the simple difference Vm = V1-V2, we obtain a measurement voltage Vm, differential, proportional to the movement of the movable element. The two windings being in series, the voltage Vm is directly accessible across all of these windings. Another way of making a structure

  differential is shown in Figure 12. The two half

  secondary 19 and 20 are no longer, as in Figure 10 placed on the same side facing the primary 18, but arranged on either side of a primary which is split in two

  identical primary half-windings 18 and 18bis.

  Electrically, these two semi-primaries being connected in series

  are traversed by the same alternating and or direct current.

  In its simplest mode of operation, we find strictly the same electrical diagram as that which is

shown in Figure 11.

  A consequence of the different arrangement of the secondary with respect to the primary represented either in FIG. 10 or in FIG. 12 necessarily results in a different arrangement of the two constituent parts of the mobile element. While being integral with the same movement, these two parts are located either in the same

  plane (Figure 10) or in two parallel planes (Figure 12).

  With this difference, we find all the possible uses either of ferromagnetic materials of variable section, or of materials both non-magnetic and good conductors of electricity. Figures 13a to 13e give some examples of embodiments of the mobile element, two of which implement an etching process used for the production of printed circuits (Figures 13a and 13b). This embodiment is particularly interesting because in addition to the good reproducibility obtained by chemical etching, it is possible to obtain mobile elements whose design can be developed either to improve the linearity of the sensor or to obtain any law of variation . Figures 13c and 13d are similar to Figures 3a and 3b respectively. FIG. 13e gives an example of arrangement of the mobile element so that it is adapted to the differential structure of the magnetic circuit shown in FIG. 12. In this case, the two mobile elements are in planes parallel to of

different altitudes.

  In Figure 11, a simple way to obtain a differential measurement voltage has been described. However, this means does not completely eliminate the effect of temperature on the sensor. Another slightly more complex implementation process achieves much better results. This other ratiometric method consists in realizing the ratio between the difference V1-V2 and the sum Vl + V2 of the output voltages of the two windings. In a more concise form, the measurement voltage Vm obtained by means of the ratiometric method is written: Vm = V1-V2 V1 + V2 In addition to the fact that the stability as a function of the temperature is considerably improved, it it appears that certain non-linearity defects are naturally compensated for. Furthermore, the fact that the sum V1 + V2 remains practically constant during satisfactory operation of the sensor can be used. Monitoring the sum V1 + V2 is therefore the means to check the correct functioning of the sensor and to alert in the event of a fault. Figure 14 shows schematically how the signals from the sensor are used both for measuring the

  movement only for monitoring a fault.

  In this case, two envelope detectors 23 and 24 measure the voltages across the windings 19 and 20 respectively. The signals V1 and V2 respectively measured are processed in an adder-divider 27 carrying out the quotient defined above. As regards monitoring the state of the sensor, an addition circuit 25 supplies a signal proportional to Vl + V2 which it suffices to compare in a comparator 26 to a threshold equal to an expected sum Vl + V2. Preferably, these treatments are performed digitally, the expected threshold being a binary state. In the event of too large a deviation from the threshold, a fault signal is produced. So far, the movable element consists of one or two sectors whose section is a linear function of the displacement. These mobile elements can themselves move either in translation or in rotation depending on whether one wishes to measure a rectilinear or angular displacement. FIG. 15 gives an example of a translation in circular form of the mobile element previously drawn in a rectilinear manner. The drawing has only one

  surface, the cross section of which is proportional to the angle of rotation but it is not difficult to apply to it all the configurations with two surfaces which are suitable for the magnetic circuits 30 of FIGS. 9 and 12.

  Since differential measurement is most often used, doubling the air gap makes it easier to

such measure.

  There is another way to obtain a measurement proportional to the displacement of the movable element without

  this uses a linear variation of the section.

  In this case, a so-called Resolver process implements a sinusoidal variation for one of the measurement voltages, and a cosine for the other. If, as before, the two measured voltages Vl and V2 are called, the sensor is arranged so that these two voltages are of the form: Vl = V sin ot. sin 0 V2 = V sin and. cos 0 oo is a pulse of measured alternating voltages, and o 0 is the angle by which the movable element moves as shown in Figure 16. In this case, the two air gaps are spaced at 90 one of the another along a circular track of the movable member. Commercial electronic components in common use make it possible to undergo a whole series of transformations at the voltages Vl and V2 so that these can be multiplied by cos 4 and sin Q. The multiplications are such that: Vl cos = V sin ot. sin 0. cos V2 sin 4 = V sin t. cos 0. sin The difference obtained is reduced to: V1 cos - V2 sin = V sin ot. sin (O -) After removal of the pulse carrier O from the generator 10, a detector delivers an error signal proportional to = sin (O -). By an internal servo loop, the device ensures that this error signal is zero. It follows that if sin (O - 4) = 0, the angle 4, the value of which is also known, is in

  permanence equal to the angle 0 that we want to measure.

  Generally reserved for precision measurements, the so-called Resolver process also makes it possible to carry out a

360 angular measurement.

  By a reverse path to that taken to pass from the rectilinear sensor to the angular sensor, we can develop a circular sinusoidal mobile element to transform it into a rectilinear mobile element and thus measure

  a linear displacement. This is shown in Figure 17.

  In this case the variation of the section follows the shape of a sinusoid. Whether the displacement is angular (figure 16) or rectilinear (figure 17) the spatial phase shift making it possible to obtain sin 0 and cos 0 can be achieved using two elementary sensors (such as that represented in the figure) offset by a quarter of a period. Figure 18 shows that there is another way to achieve the same result using a single movable element, with two surfaces offset by a quarter of a period which move in the air gap of a sensor whose structure is similar to that shown either at the

Figure 9 or Figure 12.

  The devices described above provide an absolute measurement of the rectilinear or angular displacement sometimes at the expense of the resolution. One known way of improving this resolution consists in using rectilinear or angular mobile elements whose surfaces are repeated periodically in space. Two examples are given in FIGS. 19 and 20 and it is easy to see that it is possible to further increase the periodicity of the patterns for

increase the resolution.

  One of the peculiarities of this sensor is that it has no material contact with the mobile element whose rectilinear or angular displacement is to be measured. This is illustrated by FIG. 21 where it can be seen that a large shaft 1, which is not accessible by any of its ends, is difficult to use for conventional sensors. A solution sometimes used, consists in receiving the shaft 1 in a toothed crown in its middle part and in transmitting the movement, by a series of gears, to an angular sensor. This means of setting

  expensive opener necessarily introduces games and non

  linearities that distort the measurement. By cons, if we have the movable element 2 (according to one of the configurations described above) in place of the toothed ring, the sensor

  3, proper, becomes an easily interchangeable insert during maintenance operations.

  Thus the mobile element can become an integral part of the mechanism whose displacement is to be measured. And this is all the more easily since it can be of multiple shape, adapt to the mechanisms and be produced in different ferromagnetic or non-magnetic materials and conductors of electricity. The magnetic circuit of the sensor thus becomes an external, removable and easily removable element.10 Furthermore, the various improvements presented, mixed continuous-alternative power supply, compensation for drifts by processing the signals delivered or by signals moving in opposite directions, presence of two air gaps , sensor fault detection, quotient of measurements and

  sine discrimination can be used together or in various combinations.

Claims (11)

  1 - Inductive sensor characterized in that it comprises a magnetic circuit (2) with an air gap and an excitation coil (1), a movable element (3) sliding in this air gap, and a circuit (4, 5, v ) for measuring the reluctance of this magnetic circuit in order to deduce therefrom a position of an object mechanically connected to this mobile element, the mobile element having a section, measured perpendicular to the   direction of sliding, which is variable with this sliding.   2 - Sensor according to claim 1, characterized in that the excitation winding comprises a primary coil (7) and a secondary coil (8) mounted respectively on a first and a second pole of the air gap and supplied with series.   3 - Sensor according to one of claims 1 to 2,   characterized in that the measuring circuit comprises a source (9) of alternating current and a source (10) of   direct current connected in parallel to the winding.   4 - Sensor according to claim 3, characterized in that   that the measurement circuit includes a low-pass circuit (R3-   Cl), a high-pass circuit (Cl-C2) and a multiplication circuit '(16). Sensor according to claim 4, characterized in that the measuring circuit comprises an analog-digital converter (16.1), one reference input of which is connected to the low pass circuit.   6 - Sensor according to one of claims 1 to 5,   characterized in that the movable element has a profile measured along the direction of the sliding with which the change in reluctance is linearly proportional to the sliding, for example a triangular profile.   7 - Sensor according to one of claims i to 6,   characterized in that the mobile element has a profile measured along the direction of the sliding with which the reluctance variation is alternately increasing then decreasing with respect to the sliding, for example a sinusoidal profile.   8 - Sensor according to one of claims 1 to 7,   characterized in that the movable element has a profile measured along the direction of sliding, the variation of which is spatially periodic.   9 - Sensor according to one of claims 1 to 8,   characterized in that it comprises two magnetic circuits (19, 20) and two air gaps, the movable element comprises two parts (21, 22) mechanically linked, the section of one of the parts adopting in an air gap a variation opposite to that of the other party in the other air gap with respect to the direction of sliding.   - Sensor according to one of claims 1 to 9,   characterized in that the magnetic circuit comprises two air gaps, the movable element being shaped so that its sections pass successively in each of these air gaps when it slides.   11 - Sensor according to claim 10, characterized in that the movable element comprises two sinusoidal patterns penetrating into the two air gaps with a shift of a quarter   period. 12 - Sensor according to one of claims 9 to 11,   characterized in that the measurement circuit includes a circuit (27) for measuring a ratio of the difference of the signals measured in the two air gaps to the sum of the   signals measured in these two air gaps.   13 - Sensor according to one of claims 1 to 12,   characterized in that the movable member has a section of a non-magnetic conductive material which is variable with this slip.   14 - Sensor according to one of claims 1 to 13,   characterized in that the movable element has a section of ferromagnetic material which is variable with this slip.