CH716950A2 - Method for measuring at least one property of a balance of a sprung balance oscillator. - Google Patents

Abstract

The invention relates to a method for measuring at least one property of a balance, of an oscillator (11), balance (11a) - hairspring (11b), of a watch movement (5), comprising the following steps , carried out with the oscillator (11) in a state of oscillation: a) carrying out a plurality of measurements of at least one surface of said balance (11a) by means of at least one chromatic confocal sensor (13, 13 ' ); b) processing said measurements in order to determine said at least one property of said balance (11a). The invention also relates to a system adapted to carry out this method.

Description

Technical area

The present invention relates to the field of watchmaking. It relates more particularly to a method for measuring at least one static or dynamic property of a balance of a balance-spring oscillator.

State of the art

It is generally known to use optical measurements to determine certain properties or characteristics of watch components.

[0003] For example, the shape of components can be controlled and verified by means of an optical comparator, such as a contour projector or by a laser micrometer. Since the measurements are made with the part in question statically mounted on a support, they are laborious and thus impossible to do quickly on a production line.

It is also known to use optical measurements in order to determine the dynamic properties of a sprung balance oscillator. For example, document CH712940 describes a non-acoustic system for determining the rate of a watch, the amplitude of the balance, the moment in time (tMAX) when the amplitude of the balance is maximum, and the reference of the balance, for example the intermediary of a simple proximity sensor, which can in particular be optical. This sensor can, among other things, detect the passage of the arms of a balance or another change in geometry, size, type of surface or the like.

The SMEV Laser Velocimeter product, from the company Witschi Electronic SA, uses a laser to determine the passages through the rest position, the stopping points and the direction of each oscillation of the balance, the deviation of the oscillation of the ideal behavior (due to disturbances of the escapement) and to determine the course of the movement, among others.

[0006] Other examples of horological optical measurements are disclosed by documents CH706642, FR2780169, EP2881809, CH691992, CH521627 and others.

[0007] However, all of these dynamic optical measurements require that the balance wheel is completely exposed to the light source, whether it is a laser or other, which requires removing the movement from its cap in the production line and placing the movement on a support. Indeed, the material of a cap generates light reflections which interfere with the measurements. Moreover, these methods are relatively limited in the properties which they are capable of measuring.

[0008] The aim of the invention is therefore to provide a measurement method in which the aforementioned defects are at least partially overcome.

Disclosure of the invention

[0009] More precisely, the invention relates to a method for measuring at least one property of a balance of a sprung balance oscillator of a watch movement, as defined by claim 1. This method comprises the following steps, carried out with the oscillator in a state of oscillation:

A) performing a plurality of measurements of at least one surface of said balance by means of at least one chromatic confocal sensor; and

B) processing said measurements in order to determine said at least one property of said balance.

The use of a confocal chromatic sensor instead of a laser system or a conventional optical sensor makes it possible to perform various static and / or dynamic measurements of the balance and even makes it possible to carry out these measurements through a transparent cap, or through a crystal if the movement is nested and the balance is visible on the dial side or the back side. An integration of a system carrying out the process on a production line is thus favored, since it is not necessary to remove the movements of their caps, or that the movements can be nested during their control, for reasons which will appear later. In addition, a confocal chromatic sensor can take measurements at a relatively high frequency, thus conferring overall speed on the process.

The at least one confocal chromatic sensor can be arranged such that its measurement axis is substantially parallel to the axis of rotation of said balance. Alternatively, the at least one chromatic confocal sensor can be arranged such that its measurement axis is substantially perpendicular to the axis of rotation of said balance. It is possible to have a first and a second confocal chromatic sensor, oriented respectively parallel and perpendicular to the axis of rotation of said balance. Other angles are also possible.

The properties of the balance to be determined can be static, such as for example one or more of its dimensions, its eccentricity, the roughness of its surface or the like. Alternatively, the properties can be dynamic, such as, for example, its beating, its out-of-roundness, its precession in its pivots, its oscillation frequency and / or its amplitude.

In the case of determining the frequency of the balance, steps a) and b) can advantageously be repeated by orienting said movement according to several positions, in particular according to the standardized positions for chronometry measurements.

Advantageously, and again in the case where said property comprises the frequency of oscillation of said balance, steps a) and b) can be repeated several times. In doing so, the isochronism of the movement can be determined by measuring the frequency of oscillation of the balance along the unwinding of the mainspring of the watch movement.

Advantageously, the movement is fully wound before performing steps a) and b) for the first time, and then, by detecting the fact that the balance no longer oscillates, it is possible to calculate the power reserve of said movement, in particular on the basis of the time elapsed between the winding of the movement and the detection of the moment when the balance has ceased to oscillate.

Advantageously, steps a) and b) can be performed simultaneously for a plurality of balances belonging to a plurality of movements stacked in front of said confocal chromatic sensor so that the latter can measure all of the balances. The production flow can thus be increased.

The invention also relates to a system for measuring at least one balance of a sprung balance oscillator of a watch movement, comprising:

[0020] a support adapted to receive a watch movement comprising a sprung balance regulating member;

At least one confocal chromatic sensor arranged to measure at least one surface of said balance;

[0022] a controller suitable for receiving signals from said confocal chromatic sensor and for processing said measurements in order to determine said at least one property of said balance.

This system is thus suitable for carrying out the aforementioned method.

Brief description of the drawings

Other details of the invention will emerge more clearly on reading the following description, made with reference to the accompanying drawings in which: Figure 1 schematically represents a system according to the invention, suitable for carrying out the method of the invention; FIG. 2 diagrammatically represents the measurement of the passage of the arms of a balance by a confocal chromatic sensor; Figure 3 schematically represents a graph illustrating a first approach for determining the oscillation frequency of the balance; Figure 4 schematically represents a graph illustrating a second approach for determining the oscillation frequency of the balance; FIG. 5 schematically represents the simultaneous measurement of the balances of a plurality of movements housed in a stack of caps; Figure 6 schematically shows the use of the system according to the invention in a production line; and Figure 7 schematically represents an approach to determine the amplitude of oscillation of the balance.

Modes of Carrying Out the Invention

Figure 1 schematically illustrates a system 1 suitable for performing the method of the invention.

The system 1 comprises a support 3, such as a template, adapted to receive a clockwork movement 5, which can be contained in a cap 4 or placed directly on the support. The movement 5 can be placed on the support 3 either manually or automatically by a manipulator or an ad hoc loading device. Even if the movement 5 is mounted flat in Figure 1, it is also possible to mount it in a vertical position or in any tilted position, and it is also possible that the support 3 can be movable in order to position the movement in a plurality of spatial orientations.

The timepiece movement comprises a frame 7 consisting of a plate and bridges (not shown). This frame 7 carries a driving source 9 such as a driving spring housed in a barrel, which is in kinematic connection with a regulating member 11 of the balance 11a - hairspring 11b type. The kinematic link 12 between the driving source and the regulating member 11 typically comprises a finishing gear train driving an escapement of any type and is illustrated schematically by an arrow. Other forms of kinematic connections are also possible.

The system further comprises at least one chromatic confocal sensor 13, arranged to capture the surface of the balance 11a when it oscillates. In the illustrated embodiment, a first confocal chromatic sensor 13 is arranged to capture the upper face (according to the orientation of the figure) of the balance 11a, while a second confocal chromatic sensor 13 'is arranged to capture its outer wall. , that is to say its surface which typically extends parallel to the axis of rotation of the balance 11a.

To this end, the measurement axis of the first sensor 13 is substantially parallel to the axis of rotation of the balance 11a, and that of the second sensor 13 'is substantially perpendicular to the axis of rotation of the balance 11a. These measuring devices 13 are in communication with a processor 15, suitable for receiving and processing the signals emitted by said sensor (s) 13, 13 '.

A confocal chromatic sensor, which can be of the point, multipoint or linear type, comprises a panchromatic light source, arranged to illuminate the balance 11a, as well as a receiver 13a having an extended axial chromatism. This receiver 13a comprises at least one lens, the focal length of which strongly depends on the wavelength of the incident light (before focusing on the measured object) and of the collected light (after reflection on the measured object). Therefore, the wavelength (and therefore the color) of the detected light varies depending on the distance between the receiver 13a and the measured object. By measuring the received wavelength, the distance between the receiver 13a and the object can be determined with an accuracy which can reach 10 nm, with a measuring frequency which can go up to 60 kHz or even 100 kHz, but is typically between 10 kHz and 50 kHz. A detailed explanation of the operation of such sensors, which are commercially available, is contained in the publication Blateyron F. (2011) Chromatic Confocal Microscopy, Leach R. (eds) Optical Measurement of Surface Topography. Springer, Berlin, Heidelberg, ISBN 978-3-642-12011-4.

The first sensor 13 can be mounted so that it is movable in at least one direction parallel to the plane of the balance 11a, in order to be able to take measurements of several parts of its surface, and the second sensor 13 'can be movable in at least one direction perpendicular to the plane of the balance for the same purpose. Alternatively, the sensor (s) 13, 13 ′ may be fixed relative to the support 3. For the rest, if the first sensor 13 is arranged to measure at two locations offset angularly with respect to the axis of rotation of the balance 11a, the direction of rotation of the balance 11a can be determined. This can be done by using two sensors arranged in parallel, a multipoint chromatic confocal sensor arranged to perform measurements at at least two discrete points, a chromatic confocal sensor which can be moved or whose beam can be deflected to be able to cut the balance 11a at two positions. different angles, or a confocal chromatic sensor in combination with a conventional optical sensor (for example laser).

In addition, still other means, which are capable of being measured by the sensor 13 or 13a (as the case may be), can be used to determine the direction of rotation of the balance 11a. For example, two keys with different depths can be engraved on two adjacent arms or in two different places of the rim, the analysis of the measured height giving rise to a spectral shift of the peak relative to the arms. This offset can be processed digitally to determine the direction of rotation.

The polarizer can alternatively be formed by a simple code (for example a point followed by a dash) engraved on an arm or on the rim of the balance 11a, the reading direction of this code (for example „point-dash “Or“ dash-dot ”) indicating the direction of rotation. Of course, this code can be repeated several times on the arms or on the rim of the balance.

Instead of engraving such keys, they can also be formed by the use of materials of different reflectivity, different surface finishes (polished or matte finish) or similar, the amplitude of the spectral peak varying and can be detected in order to determine the presence of the polarizer.

Again alternatively, the focal point of the confocal chromatic sensor can be positioned so that it intercepts the arms of the balance wheel 11a as well as the outer turn of the spiral spring 11b when the latter is the most extended. The last turn thus gives rise to a signal on a dedicated spectral peak (thanks to its different position of the balance 11a) for all the vibrations during which the balance spring 11b extends, which indicates the direction of rotation of the balance 11a. This approach also applies to point sensors of the laser type. The same principle also applies if the focal point of the confocal chromatic sensor is positioned such that it intercepts the arms as well as the anchor when the latter is in its outermost position, which again makes it possible to determine the direction of rotation of the balance 11a in a similar manner.

A major advantage of a confocal chromatic sensor over a conventional optical sensor is that the presence of transparent reflecting surfaces between the sensor 13 and the balance 11a does not introduce any interference into the signal. Since we know the approximate position of the balance 11a relative to the sensor 13, we can simply ignore any signal which corresponds to a shorter distance, or we can consider exclusively the signal which varies over time. Indeed, in the case of the use of a transparent cap 4 in which the movement 5 is located, or if the movement is nested so that the balance 11a is visible through a glass located on the dial side or side. bottom, sensor 13 will emit three signals. Two of these correspond to the reflection of the two surfaces of the cap or of the glass, the third, stronger (in terms of collected luminous flux), corresponds to the balance 11a (or else to the plate 7 in the case where the sensor 13 is arranged to detect the passage of an arm or an adjustment screw of the latter and that this element is not located directly in the axis of the sensor 13 at the precise moment of the measurement). The first two signals corresponding to the cap or to the glass can thus be ignored, for example, by applying an ad hoc digital filter to the data received.

For the rest, the quality of the measurements taken is independent of the state of the detected surface, which can therefore be polished, rough, irregular or the like, without any influence on the precision of the measurements, which is not the case for most conventional optical sensors.

In doing so, the sensor (s) 13, 13 'can perform various measurements of at least one property of the balance 11a and / or of the balance 11a - spring 11b when the oscillator is in a state of oscillation. These oscillations ideally (but not necessarily) have an amplitude of at least 180 ° so that the balance 11a travels 360 ° of arc in front of the sensor 13, 13 'during two successive alternations.

First, measurements of the shape of the balance 11a can be performed. The first sensor 13 can measure the height of the arms and of the rim very quickly, when the balance 11a is oscillating. The second sensor 13 'can measure the eccentricity of the balance 11a by measuring the distance between the sensor 13' and the outer wall of the balance 11a.

Each of the sensors 13, 13 'can also measure the roughness of the surface of the balance, of the patterns formed on the balance as a decoration, or the shape of areas ablated by laser during a correction of walking or of the balancing of oscillator 11. In the latter case, a preliminary measurement before said ablation can also be carried out.

If the support 3 and the sensors 13 are integral in displacement relative to each other and are capable of being pivoted in space using an ad hoc robot, the swing of the pendulum can be measured, by turning these elements over so that the axis of the balance 11a comes into contact with its other bearing, located on the balance bridge (not shown). During these measurements, the movement 5 can be manipulated in space, ideally in its cap 4, while keeping the spatial relationship between the movement 5 and the sensor (s) 13, 13 '. Likewise, the out-of-roundness of the balance 11a can be measured.

One can also perform other dynamic measurements. For example, the precession of the balance shaft in its bearings can be measured in real time, by measuring the upper surface of the balance 11a via the first sensor 13, as well as the outer wall of the balance 11a via the second sensor 13 '. In such a case, a preliminary (almost) static measurement by means of the two sensors 13, 13 'can also be carried out, in order to measure the surface and profile of the balance 11a.

The rate of oscillator 11 can also be measured in one or more positions, the system being able to be arranged to manipulate the movements in order to position them according to the standardized positions for chronometry measurements. This can be done by means of a pivotable support 3 as mentioned above.

Several mathematical approaches, some of which are similar to those already used in connection with conventional optical measurements as mentioned above, can be used to determine the rate of oscillator 11.

For example, by measuring the passage of the arms of the balance 11a through the first sensor 13, as illustrated schematically in Figure 2, or the passage of the adjustment screw through one or the Other of the first 13 or second 13 'sensors, conventional strategies can be applied. Subsequently, reference will be made exclusively to measurements of the passages of the arms in front of the sensors, but the same principles also apply to the other structures of the balance 11a mentioned.

An example of such a strategy is illustrated in FIG. 3, which schematically represents a digitized version of the signal emitted by the sensor 13 when it measures the passage of the arms of a four-arm balance. A signal "0" corresponds to the presence of an arm, the signal "1" corresponding to the absence of an arm and therefore a measurement of the relative position of the frame with respect to the sensor 13. Since the sensor 13 measures the distance to an object, and that the arm of the pendulum 11a is closer to the sensor 13 than the underlying frame is, the presence of an arm generates a measurement of "0" while the absence of 'one arm generates a measurement of „1“.

First, we measure the passages of the arms during several alternations, then we identify the fastest arm for the two vibrations composing the oscillation, that is to say the passage of the arm having a duration t0-> 1 - shortest t1-> 0, in which t0-> 1 represents the moment when the signal rises from 0 to 1 and in which t1-> 0 represents the moment when the signal falls from 1 to 0. In other words, this passage is the one where the signal remains at "0" for the shortest time for each alternation.

Then, the period T = 1 / f0 + ΔT is calculated by subtracting values of t0-> 1 and t1-> 0 for the fastest passage of an arm during the i + 2-th and i-th alternation, f0 being the nominal frequency, and ΔT being the time difference with respect to the nominal period T0 = 1 / f0. In other words, T = T0 + ΔT.

Finally, the walking oscillation by oscillation is extracted by the calculation On = -86400 * ΔT * f0 in which walking is expressed in seconds per day. Oscillation-by-oscillation calculations can be averaged over several oscillations to arrive at an average value.

On the basis of the same measurements, one can also extract the amplitude of the balance 11a, by the following method.

For each alternation, the moment of trebr cusp is identified. of the balance, which represents the local axis of symmetry in the time-stamping data, as illustrated in FIG. 4. To do this, we identify the number of arms passing under the sensor 13 during an alternation and we determine the associated angles θi. For example, for a balance 11a with four arms, θi = m × 90 °, m being the index of each arm.

Then, the instants of passage of the middle of each arm under the sensor 13 are determined, by means of the calculation tMIDi = (t0-> 1+ t1-> 0) / 2 - trebr.

This calculation is performed for each alternation.

Subsequently, a "fit" is carried out, that is to say an adjustment of a mathematical model, of these instants according to a sinusoidal law, under the assumption of a relatively low damping (which can therefore be ignored), by the calculation θi = θ0cos (2π × f0 × tMIDi) + θoffset, where

Θ0 is the oscillation amplitude for the alternation,

Θoffset represents the angular offset of the measurement axis of the sensor 13 relative to the position of the arm closest to the measurement axis when the balance 11a is at rest (for a balance with 4 arms, -90 ° <θoffset <90 °), and

F0 is the nominal frequency, which is fixed (4 Hz for example).

The value of the oscillation amplitude θ0est thus extracted alternately by alternation. The adjustment principle of this mathematical model is illustrated schematically in figure 7.

A second approach for calculating the rate on the basis of the same measurements is as follows.

First, we identify times of trebr cusp. of the balance from the instants of passage of all the arms under the sensor 13, for each alternation, as described above.

Then, the period T = 1 / f0 + ΔT is calculated by subtracting the values of trebr. for the i + 2-th and i-th alternation.

Finally, we extract the step oscillation by oscillation: On = -86400 * ΔT * f0

This approach can be more precise, since the determination of trebr. is more precise than t1-> 0 since there is already an averaging effect of two signal transitions for the same arm when determining trebr.

Of course, other mathematical approaches are possible, those described above not being in any way limiting. For example, the mathematical model can also incorporate terms and parameters representing the damping related to the tribology of the pendulum pivot and can also incorporate terms representing disturbances related to the interaction with the escapement and the impulse received from the pendulum. 'anchor.

It should also be noted that all of the calculations described above can also be carried out by sensing the passage of an adjusting screw, or of another characteristic of the balance 12a, by means of one or the other of the sensors 13, 13 ', if applicable.

Contrary to the principles of optical measurements of the prior art, an approach based on the use of a confocal chromatic sensor makes it possible to perform gait measurements by exploiting almost any non-uniform characteristic of the gait. balance wheel, that is to say not only the arms or screws, thanks to its distance measurement accuracy up to 10 nm. For example, since the outer wall of a balance is never perfectly cylindrical, it is possible to use its eccentricity or out-of-roundness measured by the second sensor 13 'to take measurements of the rate of oscillator 11. In such a case, modifications to the gait determination methods described above are within the abilities of those skilled in the art.

A great advantage of such a capture of the non-concentricity of the surface of the rim by one or the other (or all) of the sensors 13, 13 'is that one can identify the instants of cusp of the balance by identifying the axes of symmetry of the distance vs time signal. The precision over these instants can be extremely good since a much larger number of data is taken into account compared to the detection of the passages of the arms, which is in particular advantageous in the case where the balance has few arms.

FIG. 5 illustrates another possibility which is opened up by the use of a confocal chromatic sensor 13.

In this figure, a stack of movements 5, each being housed in a corresponding cap 4, is arranged so that the first sensor 13 can measure the passage of the arms of each movement 5 at the same time, the movements 5 being angularly indexed and positioned so that the measuring axis of the sensor 13 is cut by the arms of the rockers 11a. The other elements of system 1 have not been shown so as not to overload the figure.

Of course, the frames of the movements 5 are perforated so that the balance of the lower movement can be observed by the sensor 13, and the sensor 13 is arranged to be able to take measurements over a maximum distance suitable to reach the balance. 11a lowest in the stack.

Even if the signal processing is made more complex by this arrangement, since sometimes the arms of different balances will intersect the axis of the sensor 13 at the same time, it allows a determination of the rate of several movements 5, as well as the amplitude of oscillation of the corresponding balances 11a, in parallel.

FIG. 6 schematically illustrates the use of a system according to the invention on a production line, only the first sensor 13 having been shown. As mentioned above, since a confocal chromatic sensor is compatible with the use of transparent caps 4, the movements 5 can be controlled in series without removing the movements 5 and their caps 4, regardless of the type of measurements to be made. carry out. The control of nested movements can also be carried out in the case where the balance is visible through a glass, as mentioned above. In each of these cases, the speed of the measurements allows a high flow.

This figure also shows, by means of the dashed arrow, still other possibilities. By causing movements 5 to circulate in front of the sensor 13 so that their rate and / or amplitude of oscillation can be measured several times, the latter can be measured in several states of winding of their mainspring. These measurements can thus determine the isochronism of the movements and also their power reserve. If the mainsprings are initially fully wound, the winding level will decrease with each passage of each movement past the sensor 13. When the mainspring is no longer able to provide sufficient torque for the movement 5 to operate, the balance 11a will no longer be in oscillation, which can be detected by sensor 13. Subsequently, the power reserve can be determined with an accuracy corresponding to the interval between two measurements of the same movement, by comparing the time of winding with the moment of detection that the balance 11a no longer oscillates.

Of course, the movements can be identified by a bar code, QR code, RFID or the like, as generally known, in order to assign the data to the corresponding movements.

Finally, it is also possible to mount the sensor 13 on a motorized arm in order to perform measurements on movements mounted on acoustic devices of the Witschi type or the like. By correlating the optical and acoustic measurements, it is even possible to deduce the variation in the lifting angle of the movement during its power reserve.

Although the invention has been previously described in connection with specific embodiments, other additional variants can also be envisaged without departing from the scope of the invention as defined by the claims.

Claims (11)

1. Method for measuring at least one property of a balance of an oscillator (11) balance (11a) - hairspring (11b) of a watch movement (5), comprising the following steps, carried out with the oscillator (11) in a state of oscillation: a) performing a plurality of measurements of at least one surface of said balance (11a) by means of at least one confocal chromatic sensor (13, 13 '); b) processing said measurements in order to determine said at least one property of said balance (11a). 2. Method according to the preceding claim, wherein said at least one confocal chromatic sensor (13, 13 ') is at least one of: - a first confocal chromatic sensor (13) arranged so that its measurement axis is substantially parallel to the axis of rotation of said balance (11a), and - a second confocal chromatic sensor (13 ') arranged so that its measurement axis is substantially perpendicular to the axis of rotation of said balance (11a). 3. Method according to one of the preceding claims, wherein said at least one property is chosen from: - one or more dimensions of said balance (11a); - the eccentricity of said balance (11a); - the swing of said balance (11a); - the roughness of the surface of said balance (11a); - the out-of-roundness of said balance (11a); - the frequency of oscillation of said balance (11a); - the amplitude of oscillation of said balance (11a); - the precession of said balance (11a) in its pivots. 4. Method according to one of the preceding claims, wherein said movement (5) is in a transparent cap (4) when said measurements are carried out. 5. Method according to one of the preceding claims, wherein said at least one property comprises the frequency of oscillation of said balance (11a), steps a) and b) being repeated with said movement (5) positioned in several spatial orientations. . 6. Method according to one of the preceding claims, wherein said at least one property comprises the frequency of oscillation of said balance (11a), steps a) and b) being repeated several times. 7. Method according to the preceding claim, wherein said movement (5) is substantially fully wound before performing steps a) and b) for the first time. 8. Method according to the preceding claim, wherein, following detection of the fact that said balance (11a) no longer oscillates, the power reserve of said movement (5) is calculated. 9. Method according to one of the preceding claims, wherein steps a) and b) are performed simultaneously for a plurality of balances (11a) belonging to a plurality of movements (5) stacked in front of said confocal chromatic sensor of such. so that the latter can measure all of their balances (11a). 10. System (1) for measuring at least one property of a balance (11a) of a balance oscillator (11a) - hairspring (11b) of a watch movement (5), comprising: - a support (3) adapted to receive a watch movement (5) comprising a balance regulating member (11a) - hairspring (11b); - at least one confocal chromatic sensor (13, 13 ') arranged to measure at least one surface of said balance (11a); - a controller (15) adapted to receive signals from said confocal chromatic sensor and to process said measurements in order to determine said at least one property of said balance (11a). 11. System according to the preceding claim, wherein said at least one confocal chromatic sensor (13, 13 ') is at least one of: - a first chromatic confocal sensor (13) arranged such that its measurement axis is substantially parallel to the axis of rotation of said balance (11a), a second, a confocal chromatic sensor (13 ') arranged so that its measurement axis is substantially perpendicular to the axis of rotation of said balance (11a).