CH715295A2 - Timepiece including a mechanical oscillator whose average frequency is synchronized with that of a reference electronic oscillator. - Google Patents

Abstract

The timepiece comprises a mechanical oscillator, formed by a mechanical resonator (6) and an oscillation maintenance device, and an auxiliary oscillator forming a reference time base (22). It includes a synchronization device (20) arranged to slave the average frequency of the mechanical oscillator to that of the auxiliary oscillator. This synchronization device comprises an electromagnetic braking device (26) which is formed of a coil and at least one permanent magnet and arranged so that an induced voltage is generated between the terminals of the coil in each alternation of the oscillation of the mechanical resonator. The synchronization device is arranged to be able to momentarily decrease the impedance between the terminals of the coil during distinct time intervals, any two successive time intervals having between their respective beginnings a time distance substantially equal to a positive integer multiplied by half of a set period for the mechanical oscillator.

Description

Technical area

The present invention relates to a timepiece comprising a mechanical movement whose walking is improved by a device for correcting a possible time drift in the operation of the mechanical oscillator which cadences the walking of the mechanical movement. The timepiece includes a mechanical oscillator whose average frequency is synchronized with a set frequency determined by an auxiliary electronic oscillator.

In particular, the timepiece is formed, on the one hand, by a mechanical movement comprising: - a mechanism indicating at least one time datum, - a mechanical resonator capable of oscillating along a general axis of oscillation around a neutral position corresponding to its state of minimum potential energy, - a device for servicing the mechanical resonator, forming with it a mechanical oscillator which is arranged to clock the progress of the indicator mechanism, and, on the other hand, by a synchronization device arranged to slave the average frequency of the mechanical oscillator to a set frequency determined by a reference time base.

Technological background

Timepieces as defined in the field of the invention have been proposed in some prior documents. Patent CH 597,636, published in 1977, proposes such a timepiece with reference to its FIG. 3. The movement is equipped with a resonator formed by a balance-spring and a conventional maintenance device comprising an anchor and an escapement wheel in kinematic connection with a barrel provided with a spring. This watch movement also includes a device for regulating the frequency of its mechanical oscillator. This regulating device comprises an electronic circuit and an electromagnetic braking device formed by a flat coil, arranged on a support under the lever of the balance, and two magnets mounted on the balance and arranged close to each other. both pass over the coil when the oscillator is activated.

The electronic circuit comprises a time base comprising a quartz resonator and used to generate a reference frequency signal FR, this reference frequency being compared with the frequency FG of the mechanical oscillator. The oscillator FG frequency is detected by the electrical signals generated in the coil by the pair of magnets. The comparison between the two frequencies FG and FR is carried out by a bidirectional counter receiving these two frequencies at its two inputs and supplying at the output a signal determining a difference in periods counted for the two frequencies. The electronic circuit further comprises a logic circuit which analyzes the output signal from the counter to control a circuit for applying braking pulses as a function of this output signal, so as to brake the balance wheel when the logic circuit has detected a time drift corresponding to a value of the frequency FG of the oscillator greater than the reference frequency FR. The braking pulse application circuit is designed to be able to generate a momentary braking torque on the balance wheel via an electromagnetic magnet-coil interaction and a switchable load connected to the coil.

Summary of the invention

An object of the present invention is to simplify as much as possible the electronic circuit of a synchronization device arranged to control the average frequency of the mechanical oscillator with a mechanical movement on a set frequency determined by an auxiliary electronic oscillator , without losing precision in the running of the timepiece equipped with such a synchronization device.

In the context of the present invention, it is generally sought to improve the precision of the operation of a mechanical watch movement, that is to say to reduce the maximum daily error of this mechanical movement and more overall, to significantly reduce a possible time drift over a longer period (for example a year). In particular, the present invention seeks to achieve such an aim for a mechanical watch movement whose gait is initially regulated at best. Indeed, a general object of the invention is to find a device for correcting the progress of a mechanical movement in the case where the natural functioning of this mechanical movement would lead to a certain daily error and therefore to an increasing temporal drift. (increasing cumulative error), without giving up that it can operate autonomously with the best precision it is possible to have thanks to its own characteristics, that is to say in the absence of the correction device or when the latter is inactive.

To this end, the present invention relates to a timepiece as defined in the field of the invention and in which the synchronization device comprises an electromagnetic braking device of the mechanical resonator, this electromagnetic braking device being formed at least one coil and at least one permanent magnet which are arranged so that an induced voltage is generated between the two terminals of the coil in each alternation of the oscillation of the mechanical resonator for a useful operating range of the mechanical oscillator, the synchronization device being arranged to be able to temporarily decrease the impedance between the two terminals of the coil. The timepiece is remarkable because the synchronization device is arranged so as to reduce the impedance between the two terminals of the coil during distinct time intervals TPet so that the beginnings of any two successive time intervals, among the distinct time intervals, have a time distance DT equal to a positive integer N multiplied by half of a set period T0c for the mechanical oscillator, ie DT = N⋅T0c / 2. synchronization device is arranged to determine by means of the reference time base the start of each of the distinct time intervals so as to satisfy the aforementioned mathematical relationship between the time distance DT and the setpoint period T0c.

Thanks to the characteristics of the invention, surprisingly, the mechanical oscillator of the watch movement is slaved to the auxiliary oscillator in an efficient and rapid manner, as will become clear from the detailed description of the invention which will follow. The oscillation frequency of the mechanical oscillator (slave mechanical oscillator) is synchronized with the set frequency determined by the auxiliary oscillator (master oscillator), and this without closed loop control and without the need for a motion measurement sensor d oscillation of the mechanical oscillator. The synchronization device therefore operates with an open loop and it makes it possible to correct both an advance and a delay in the natural progress of the mechanical movement, as will be explained below. This result is quite remarkable.

It is understood by "synchronization on a master oscillator" a servo (open loop, therefore without feedback) from the slave mechanical oscillator to the master oscillator. The operation of the synchronization device is such that the frequency at which the time intervals occur, where the impedance of the circuit connected to the two terminals of the coil is reduced, is imposed on the slave mechanical oscillator which cadences the march of the indicator mechanism of temporal data. More generally, it is not even necessary that the succession of such distinct time intervals occur periodically at a given frequency, since it is sufficient for the beginnings (or, equivalently, the middle instants) of any two successive time intervals among these distinct time intervals have a time distance DT as defined above, with a positive integer N which can vary over time. We are not here in the standard case of a forced oscillator, nor even in the situation of coupled oscillators.

In the present invention, the possible temporal distances DT, for a predefined set period T0c, determine the average frequency of the mechanical oscillator and therefore the timing of the operation of the mechanism. As the temporal distances are determined by a precise auxiliary oscillator, the average frequency is determined by this auxiliary oscillator so that the precision of the operation of the mechanism is in direct relation with that of the auxiliary oscillator. It is understood by 'timing the progress of a mechanism' the fact of rhythmizing the movement of the mobile elements of this mechanism when it is operating, in particular of determining the rotational speeds of its wheels and thus of at least one indicator of temporal data.

In a main embodiment, the mechanical resonator is formed by a pendulum oscillating around an axis of oscillation, and the synchronization device is arranged so as to periodically trigger the separate time intervals TP, which have a same value, and in such a way that the tripping frequency FD of these distinct time intervals is twice a set frequency F0c, equal by definition to the inverse of the set period T0c, divided by a positive integer M, ie FD = 2-F0c / M, the value of the separate time intervals TP being less than the set half-period, ie TP <T0c / 2. In a preferred variant, the value of the separate time intervals TP is expected to be less than a quarter of the setpoint period T0c, i.e. TP <T0c / 4.

Brief description of the figures

The invention will be described below in detail using the accompanying drawings, given by way of non-limiting examples, in which: <tb> Fig. 1 <SEP> shows a first embodiment of a timepiece according to the invention, <tb> fig. 2 <SEP> is a partial view of the first embodiment according to FIG. 1, <tb> fig. 3 <SEP> shows the electronic diagram of a first variant of the control circuit of the electromagnetic braking device according to the invention, <tb> fig. 4 <SEP> shows the electronic diagram of a second variant of the control circuit of the electromagnetic braking device according to the invention, <tb> fig. 5A, 5B and 5C <SEP> are graphs giving the temporal evolution of various physical parameters of the mechanical oscillator and of the synchronization device of the first embodiment for various relationships between the set frequency F0c and the natural frequency F0 of the oscillator mechanical, respectively F0> F0c, F0 <F0c, F0 = F0c, <tb> fig. 6 <SEP> shows the application of a first braking pulse to a mechanical resonator in a certain alternation of its oscillation before it passes through its neutral position, as well as the angular speed of the balance of this mechanical resonator and its position angular in a time interval during which the first braking pulse occurs, <tb> fig. 7 <SEP> is a figure similar to fig. 6 but for the application of a second braking pulse in a certain alternation of the oscillation of a mechanical oscillator after it has passed through its neutral position, <tb> fig. 8A, 8B and 8C <SEP> respectively show the angular position of a balance-spring during a period of oscillation, the variation of the movement of the watch movement obtained for a braking pulse of fixed duration, for three values of a constant braking torque, as a function of the angular position of the balance spring, and the corresponding braking power, <tb> fig. 9, 10 and 11 <SEP> respectively show three different situations which can occur in an initial phase following the engagement of the correction device in a timepiece according to the invention, <tb> fig. 12 <SEP> is an explanatory graph of the physical process occurring following the engagement of the correction device in the timepiece according to the invention and leading to the desired synchronization for the case where the natural frequency of the slave mechanical oscillator is greater than the set frequency, <tb> fig. 13 <SEP> represents, in the case of fig. 12, an oscillation of the slave mechanical oscillator and the braking pulses in a stable synchronous phase for a variant where a braking pulse intervenes in each alternation, <tb> fig. 14 <SEP> is an explanatory graph of the physical process occurring following the engagement of the correction device in the timepiece according to the invention and leading to the desired synchronization for the case where the natural frequency of the slave mechanical oscillator is less than the set frequency, <tb> fig. 15 <SEP> represents, in the case of FIG. 14, an oscillation of the slave mechanical oscillator and the braking pulses in a stable synchronous phase for a variant where a braking pulse intervenes in each alternation, <tb> fig. 16 and 17 <SEP> give, respectively for the two cases of FIGS. 12 and 14, the graph of the angular position of a mechanical oscillator and the corresponding oscillation periods for an operating mode of the correction device where a braking pulse occurs every four oscillation periods, <tb> fig. 18 and 19 <SEP> are respectively partial enlargements of figs. 16 and 17, <tb> Fig. 20 <SEP> represents, in a similar manner to the two preceding figures, a specific situation in which the frequency of a mechanical oscillator is equal to the braking frequency, <tb> fig. 21 <SEP> schematically shows the mechanical oscillator and the electromagnetic device of a second embodiment, <tb> fig. 22 <SEP> gives, in the context of the second embodiment, graphs of the time evolution of the angular position of the mechanical oscillator, of the voltage induced in a coil of the electromagnetic device as a function of a control signal of this stationary electromagnetic device, <tb> fig. 23 <SEP> schematically shows the mechanical oscillator and the electromagnetic device of a third embodiment, <tb> fig. 24 <SEP> gives, in the context of the third embodiment, graphs of the time evolution of the angular position of the mechanical oscillator, of the voltage induced in a coil of the electromagnetic device as a function of a control signal of this stationary electromagnetic device, <tb> fig. 25 <SEP> is similar to fig. 24 for a variant control of the electromagnetic device in the context of the third embodiment, <tb> fig. 26 <SEP> is a sectional view of the mechanical oscillator and the electromagnetic device of a fourth embodiment, <tb> fig. 27 <SEP> is cross section, along line A-A of the mechanical oscillator and the electromagnetic device of fig. 26, and <tb> fig. 28A, 28B and 28C <SEP> are graphs giving the time evolution of various physical parameters of the mechanical oscillator and of the synchronization device of the fourth embodiment for various relationships between the set frequency F0c and the natural frequency F0 of the mechanical oscillator, respectively F0> F0c, F0 <F0c, F0 = F0c.

Detailed description of the invention

A first embodiment of a timepiece according to the invention will be described with reference to Figs. 1 to 4 and 5A to 5C. In fig. 1 is shown, partly schematically, a timepiece 2 comprising a mechanical movement 4 which comprises at least one mechanism 12 indicating a time datum. The mechanism 12 comprises a gear train 16 driven by a barrel 14 (the mechanism is partially shown in FIG. 1). The mechanical movement also comprises a mechanical resonator 6, formed by a balance 8 and a hairspring 10, which is arranged on a plate 5 defining a support for the mechanical resonator, and a device for servicing this mechanical resonator which is formed by an exhaust 18, this maintenance device forming with the mechanical resonator a mechanical oscillator which cadences the progress of the indicator mechanism. The escapement 18 conventionally comprises an anchor and an escape wheel, the latter being kinematically connected to the barrel via the gear train 16. The mechanical resonator is capable of oscillating, around a neutral position (rest position / angular position zero) corresponding to its minimum potential energy state, along a circular axis (the radius of this axis is immaterial since the position of the pendulum along this axis is given by an angle). The circular axis defines a general axis of oscillation which indicates the nature of the movement of the mechanical resonator, which may for example be linear in another embodiment.

Each oscillation of the mechanical resonator defines an oscillation period which is formed by two half-waves, each between two extreme angular positions of the oscillation and with a rotation in opposite directions to the other. When the mechanical resonator reaches an extreme angular position, defining the amplitude of oscillation, its speed of rotation is zero and the direction of rotation is reversed. Each half-wave has two half-waves (the duration of which may be different depending on disturbing events), either a first half-wave occurring before the mechanical resonator passes by its neutral position and a second half-wave intervening after this passage by the neutral position.

Timepiece 2 comprises a synchronization device 20 of the mechanical oscillator, formed by the mechanical resonator 6 and the escapement 18, on a reference time base 22 constituted by an auxiliary oscillator which comprises a resonator crystal 35 and a clock circuit 36 maintaining the crystal resonator and delivering a reference frequency signal SR. The crystal oscillator defines a master oscillator. The reference time base is associated with the control device 24 of the synchronization device to which it supplies the signal SR. It will be noted that other types of auxiliary oscillators can be provided, in particular an oscillator integrated entirely in an electronic circuit with the control circuit. Generally, the auxiliary oscillator is by nature or by construction more precise than the mechanical oscillator arranged in the watch movement, this mechanical oscillator defining a slave oscillator within the framework of the invention. Generally, as will be understood below, the synchronization device 20 is arranged to slave the average frequency of the mechanical oscillator to a set frequency determined by the auxiliary oscillator.

Then, the synchronization device 20 comprises an electromagnetic braking device 26 of the mechanical resonator 6. By “electromagnetic braking” is understood a braking of the mechanical resonator generated via an electromagnetic interaction between at least one permanent magnet, carried by the resonator mechanical or a support of this mechanical resonator, and at least one coil carried respectively by the support or the mechanical resonator and associated with an electronic circuit in which a current induced in the coil by the magnet can be generated. In general, the electromagnetic braking device is thus formed of at least one coil 28 and at least one permanent magnet which are arranged so that an induced voltage is generated between the two terminals 28A, 28B of the coil 28 in each alternation of the oscillation of the mechanical resonator for a useful operating range of the mechanical oscillator. The coil 28 is of the wafer type (disc having a height less than its diameter), without ferromagnetic core. In the first embodiment, a plurality of bipolar magnets 30, 32 is provided which are arranged in a juxtaposed manner on the clamp 9 of the balance wheel with alternating magnetic polarities in the direction of the axis of oscillation 34. In an equivalent variant, an annular magnet is provided having an axial magnetization with successive sectors corresponding to the bipolar magnets 30, 32, these successive sectors having alternating polarities and each defining an angle at the center (an angular 'opening') having substantially same value. In the variant shown, the bipolar magnets 30, 32 define eight magnetized annular sectors each having an angular distance of 45 ° with alternating magnetic polarities. In the case of the first embodiment, there is an even number 2N of magnetized annular sectors, N being a positive integer, these sectors being arranged in a circular manner, in particular on the clamp 9 of the balance 8 forming the mechanical resonator 6.

The coil 28 is arranged on the plate 5 so as to be crossed by the magnetic flux of the bipolar magnets / annular sectors magnetized when the balance oscillates. Advantageously, the diameter of the coil 28 is provided so that it is substantially included in an angular opening, relative to the axis of oscillation, which is substantially equal to that defined by each bipolar magnet / annular magnetized sector. However, in other variants, the diameter of the coil 28 can be provided for larger and for example have an angular opening corresponding substantially to twice that of a magnetic annular sector. In addition, in another variant, a plurality of wafer coils are provided, having between them, taken in pairs, an angular offset corresponding to an integer number of magnetic periods (a magnetic period being given by the angular distance of two annular sectors adjacent magnets). These coils thus not presenting an electromagnetic phase shift (that is to say that the phase shifts are integral multiples of 360 °), the voltages induced in these coils each have a time variation identical and simultaneous to the others, so that these induced voltages add up. The plurality of coils can be arranged in series or in parallel. The number of magnetized annular sectors, the number of coils and their characteristic dimensions are selected according to the strength of the desired electromagnetic interaction to allow the desired servo control of the mechanical oscillator.

According to the invention, the synchronization device is arranged to be able to momentarily reduce the impedance between the two terminals of the coil. According to a general synchronization mode implemented in the synchronization device of the invention, the latter is arranged so as to reduce the impedance between the two terminals of the coil during distinct time intervals TPet so that the respective beginnings of any two successive time intervals, among these distinct time intervals, have a time distance DT equal to a positive integer N multiplied by half of a setpoint period T0c (i.e. by half - setpoint period) for the mechanical oscillator, i.e. DT = N⋅T0c / 2. The synchronization device is arranged to determine by means of the reference time base 22 the start of each of the distinct time intervals so as to satisfy the aforementioned mathematical relationship between the time distance DT and the setpoint period T0c.

In the embodiments described, the mechanical resonator is formed by a pendulum rotating around an axis of oscillation. In the synchronization modes implemented in the synchronization devices which are represented in FIGS. 5A to 5C and 28A to 28C, provision is made for periodically triggering the separate time intervals TP during which the impedance between the terminals of the coil is reduced, that is to say that these time intervals are provided with a time distance DT between them which is constant. The tripping frequency FD of these separate time intervals is twice the set frequency F0c, which is by definition the inverse of the set period T0c, divided by a positive integer M, i.e. FD = 2⋅F0c / M. Then, preferably, the separate time intervals TP have the same value which is expected to be less than the set half-period, ie TP <T0c / 2. Finally, the synchronization device is arranged so as to generate a short circuit between the two terminals 28A and 28B of the coil 28 during the separate time intervals TP to reduce the impedance between the two terminals of this coil.

In the variant of the first embodiment described with the aid of FIGS. 5A to 5C, the integer M is two (M = 2), so that the tripping frequency FD is equal to the set frequency F0c and the successive time distances Td are equal to the set period T0c. Then, the value of the separate time intervals TP is advantageously less than a quarter of the setpoint period T0c, ie TP <T0c / 4. In this first embodiment, as can be seen in FIGS. 5A to 5C, the electromagnetic braking device 26 is arranged so that an induced voltage is generated in the coil 28 substantially without interruption for any oscillation of the mechanical resonator 6 in the useful operating range of the mechanical oscillator formed by this resonator mechanical.

Before considering in more detail Figs. 5A to 5C, the behavior of a mechanical oscillator subjected to short duration braking pulses will first be summarized here, although a more detailed discussion of this subject is given below. It has been observed that, when the braking pulse is generated between the start of an alternation and the passage of the resonator by its neutral position in this alternation, such a braking pulse induces a negative time shift in the oscillation of the resonator. Thus, the duration of the alternation concerned is increased relative to the duration T0 / 2 of an alternation during the natural oscillation of the mechanical oscillator. This therefore generates a punctual decrease in the frequency of the mechanical oscillator and makes it possible to generate a certain delay in the running of the timepiece to correct, if necessary, an advance taken by this mechanical oscillator. On the other hand, when the braking pulse is generated between the passage of the resonator through its neutral position in an alternation and the end of this alternation, such a braking pulse induces a positive time phase shift in the oscillation of the resonator. Thus, the duration of the alternation concerned is reduced relative to the duration T0 / 2 of an alternation during the natural oscillation of the mechanical oscillator. This therefore generates a point increase in the frequency of the mechanical oscillator and makes it possible to generate a certain advance in the running of the timepiece to correct, if necessary, a delay taken by this mechanical oscillator.

In Figs. 5A to 5C are shown, in a stable phase of synchronization obtained by the synchronization device according to the invention, the curves of the angular position and the angular speed of the balance spring 6 as well as a digital control signal Se generated in the control circuit 24 and supplied to a switch 40 arranged to short-circuit the two terminals 28A, 28B of the coil 28 (see fig. 3 and 4) during pulses 58 which define the distinct time intervals TP. In addition, in these figures are shown a signal of the voltage induced in the coil 28, resulting from the oscillation of the mechanical resonator 6 and the short-circuit pulses 58, and a signal of the braking torque applied to the mechanical resonator during the short circuit pulses. It will be noted that the stable phase represented here occurs following a transient phase (initial phase) which will be described below. Remarkably, during the stable phase, also called synchronous phase, the oscillation frequency of the mechanical resonator is slaved to the set frequency F0c and the first and second parts TB and T Short-circuit pulses 58 have a substantially constant and defined. In this stable phase, the synchronization device automatically stabilizes, without a sensor measuring a parameter of the oscillation of the mechanical resonator 6 and without feedback loop, the oscillation frequency of this mechanical resonator at the set frequency F0c.

[0023] FIG. 5A corresponds to a situation where the natural frequency F0 of the mechanical oscillator of the timepiece is greater than the set frequency F0c, so that this timepiece without the synchronization device would exhibit a positive time drift corresponding to an advance in the progress of the timepiece. It is observed that the short-circuit pulses 58 occur around an extreme angular position, that is to say that the distinct time intervals TP include an inversion of the direction of the oscillation movement which occurs between an alternation A2 and a alternation A1 of the oscillation when the speed of rotation (angular speed) is zero. The oscillation periods are equal to the setpoint period T0c, but it is noted that the two alternations A1 and A2 which constitute each oscillation period are not equal. Indeed, the alternation A1 lasts here longer than the alternation A2, because greater braking takes place in the alternation A1, before the mechanical resonator passes through its neutral position (angle 0 °), than in the alternation A2 after the mechanical resonator has passed through its neutral position. It will be noted that no braking torque is applied to the mechanical resonator neither after the passage of the mechanical resonator through its neutral position in the alternation A1, nor before the passage of the mechanical resonator through its neutral position in the alternation A2.

The braking pulse is formed by two small lobes 50 located respectively on each side of the instant of passage of the mechanical resonator by the extreme angular position, having a central symmetry relative to this instant (the opposite mathematical signs of two lobes 50 comes from the change of direction in the oscillation movement), and from a lobe 52 of greater amplitude which intervenes in the alternation A1 of each oscillation period, in the first half-alternation before the passage of the mechanical resonator by its neutral position. The effects of the two lobes 50 compensate each other and therefore generally do not cause any phase shift in the oscillation of the mechanical resonator, while the braking torque caused by the lobe 52 in each half-wave A1 generates an increase in the duration of the latter, so that the duration of the oscillation period concerned is equal to that of the setpoint period T0c. The instantaneous oscillation frequency is thus equal to the setpoint frequency F0c which is, as indicated, less than the natural frequency F0 of the mechanical oscillator. The appearance of the lobe 52 only in the alternations A1 results from the fact that the middle instants of the short-circuit pulses 58 occur with a certain delay relative to the passages of the mechanical resonator by an extreme angular position, this resulting from the fact that the natural frequency F0 of the mechanical oscillator is greater than the set frequency F0c. Indeed, the part TB of the pulses 58 occurring before the passage of the mechanical resonator through an extreme position is less than the part T Ades pulses 58 occurring after this passage.

[0025] FIG. 5B corresponds to a situation where the natural frequency F0 of the mechanical oscillator of the timepiece is lower than the set frequency F0c, so that this timepiece without the synchronization device would have a negative time drift corresponding to a delay in the progress of the timepiece. We also observe that the short-circuit pulses 58 occur around an extreme angular position and that the alternation A1 lasts longer than the alternation A2, because greater braking occurs in the alternation A2, here after the passage of the mechanical resonator through its neutral position (angle 0 °), only in the alternation A1 before the mechanical resonator passes through its neutral position. As in the previous situation, no braking torque is applied to the mechanical resonator either after the mechanical resonator has passed through its neutral position in alternation A1, or before the mechanical resonator has passed through its neutral position in alternation A2 . The braking pulse is here formed of two small lobes 50 located respectively on each side of the extreme angular position and of a lobe 54 of greater amplitude which intervenes in the alternation A2 of each period of oscillation, in the second half-wave after the mechanical resonator has passed through its neutral position.

The effects of the two lobes 50 always compensate each other, while the braking torque caused by the lobe 54 in each alternation A2 generates a reduction in the duration of the latter, so that the duration of the period of oscillation concerned is equal to that of the setpoint period T0c. The instantaneous oscillation frequency is thus equal to the set frequency F0c which is, as indicated, greater than the natural frequency F0 of the mechanical oscillator. The appearance of the lobe 54 only in the alternations A2 results from the fact that the middle instants of the short-circuit pulses 58 occur here with a certain advance relative to the passages of the mechanical resonator by an extreme angular position, this resulting from the fact that the frequency natural F0 of the mechanical oscillator is less than the set frequency F0c. Indeed, the part TAdes pulses 58 intervening after the passage of the mechanical resonator by an extreme position is less than the part TBdes pulses 58 intervening before this passage.

To be complete, there is shown in FIG. 5C a situation where the natural frequency F0 of the mechanical oscillator of the timepiece is equal to the set frequency F0c. It follows from this situation that the part TAdes pulses 58 occurring after the mechanical resonator has passed through an extreme angular position is equal to the part TB of the pulses 58 occurring before this passage, so that the parts 50A of the braking pulses intervening in the half-waves A2 just before the passage of the mechanical resonator through an extreme position have the same profile, with an opposite mathematical sign, as the parts 50B of the braking pulses intervening in the alternations A1 just after this passage and thus have a central symmetry relative to the instant of passage through the extreme angular position concerned. Consequently, the effects of parts 50A and 50B of the braking pulses occurring during each short-circuit pulse 58, and therefore of each separate time interval TP, compensate each other so that, in this particular case, the device synchronization does not influence the progress of the timepiece, which is precise insofar as it is naturally synchronous with the reference time base 22.

[0028] FIG. 3 is a diagram showing a first alternative embodiment 24A of the control circuit 24 of the synchronization device 20. The control circuit 24A is connected on the one hand to the clock circuit 36 and, on the other hand, to the coil 28. The clock circuit maintains the quartz resonator 35 and generates in return a clock signal SR at a reference frequency, in particular equal to 2 <15> Hz. The clock signal SR is successively supplied to two dividers DIV1 and DIV2 (these two dividers can form two stages of the same divider). Divider DIV2 provides a periodic signal SDdirectly to a timer 38 ("Timer"). On each detection of a characteristic transition in the periodic signal SD, the timer turns on the switch 40 during a time interval TP, to short-circuit the coil 28, by providing it with a control signal Se, having a frequency of triggering FD identical to that of the periodic signal SD, which periodically triggers the timer 38. As the duration of the braking pulses (corresponding to that of the short-circuit pulses) is here provided preferably less than T0c / 4 (for example T0c = 250 ms) and even much lower than this value in the case under consideration, in particular between 10 ms and 30 ms, the timer 38 receives a timing signal from the divider DIV1.

For example, in the case of a set frequency F0c = 4 Hz and a trigger frequency FD equal to this set frequency, as in the example given in figs. 5A to 5C, the divider DIV2 directly supplies trigger pulses to the timer at the frequency FD = 4 Hz. If it is planned to supply a short-circuit pulse every second, that is to say every four periods d oscillation, and thus to have a time distance DT = 1 s between the separate time intervals TP where the impedance between the terminals 28A and 28B of the coil is reduced, then we can use the terminal output of a clock divider circuit classic which provides, at the output of the terminal stage of its division chain by two, a periodic signal at the frequency of 1 Hz. For a trigger frequency FD = 4 Hz mentioned above, it is also possible to use a circuit classic watch divider, but taking the signal supplied two stages before the terminal output in the division chain. It will be noted that the control circuit 24A of the synchronization device is very simple. It can be easily miniaturized and its electrical consumption is very low. No microcontroller is required.

It will be noted that in a particular synchronization mode, provision may be made to generate the short-circuit pulses in groups, for example a succession of sequences with four pulses in four successive oscillation periods and then no pulse for ten seconds, or for forty periods for a frequency F0c = 4 Hz. In another synchronization mode, provision may be made to vary the time intervals TP (therefore the duration of the short-circuit pulses), for example by providing a longer duration in an initial phase, to generate a greater braking torque, than in a nominal speed which follows it. Note that the synchronization process is robust. For example, it is not necessary for the time intervals TP to be measured precisely, that is to say with as much precision as the time distances DT between the beginnings of these time intervals. Thus, a timer could be provided with its own timing circuit, less precise than the reference time base 22.

In a second alternative embodiment 24B, shown in FIG. 4, of the control circuit 24 of the synchronization device 20, the dividers DIV1 and DIV2 together form a conventional watch divider circuit which therefore supplies as output a periodic signal SD having a frequency equal to 1Hz. This signal SD is supplied to a counter at N which defines an additional divider, which generates the periodic signal SP which it supplies to the timer 38. The control signal Se supplied by the timer to the switch 40 has a trigger frequency FD equal to that of the periodic signal SP. Thus, in an example where the setpoint frequency F0c of the mechanical oscillator is equal to 4 Hz (F0c = 4 Hz) and the number N is equal to eight, the triggering frequency FD of the periodic signals SPet Se is then 1 / 8 Hz, which means that a braking pulse (short-circuit pulse) is provided per 32 setpoint periods T0c, i.e. approximately one pulse after 32 periods of the mechanical oscillator insofar as its natural frequency F0 is expected close to the set frequency F0c.

In fig. 4, the synchronization device further comprises a supply device 44 formed by a rectifier circuit 46 (of the single or double alternation type) and by a storage capacity Cal connected to ground (reference potential of the synchronization device). The rectifier circuit is constantly connected at the input to a terminal of the coil so that apart from the short-circuit pulses it can rectify a voltage induced in the coil 28 by the permanent magnets 30,32. This rectified induced voltage accumulated in the storage capacity is used to power the synchronization device in the useful operating range of the mechanical oscillator. The control circuit 24B of the synchronization device is very simple and autonomous. It consumes little and takes a minimum of energy from the mechanical oscillator to effectively perform the synchronization according to the invention.

We will describe below, with reference to Figs. 6 and 7, a remarkable physical phenomenon highlighted in the context of developments leading to the present invention and involved in the synchronization process implemented in the timepiece according to the invention. Understanding this phenomenon will make it possible to better understand the synchronization obtained by the synchronization device regulating the progress of the mechanical movement.

In Figs. 6 and 7, the first graph indicates the instant tp1 at which a braking pulse P1, respectively P2 is applied to the mechanical resonator considered in order to correct the progress of the mechanism which is clocked by the mechanical oscillator formed by this resonator. The last two graphs respectively show the angular speed (values in radian per second: [rad / s]) and the angular position (values in radian: [rad]) of the oscillating member (hereinafter also 'the pendulum') of the mechanical resonator over time. Curves 90 and 92 correspond respectively to the angular speed and the angular position of the free-swinging balance (oscillation at its natural frequency) before the intervention of a braking pulse. After the braking pulse are shown the speed curves 90a and 90b corresponding to the behavior of the resonator respectively in the case disturbed by the braking pulse and the case undisturbed. Likewise, the position curves 92a and 92b correspond to the behavior of the resonator respectively in the case disturbed by the braking pulse and the case undisturbed. In the figures, the instants tp1 and tp2 at which the braking pulses P1 and P2 occur correspond to the time positions of the middle of these pulses. However, the start of the braking pulse and its duration are considered as the two parameters which temporally define a braking pulse.

By braking pulse, we understand the momentary application of a torque to the mechanical resonator which brakes its oscillating member (balance), that is to say which opposes the oscillating movement of this oscillating organ. In the case of a non-zero torque which is variable, the duration of the pulse is generally defined as the part of this pulse which has a significant force torque to brake the mechanical resonator. It should be noted that a braking pulse can have a large variation. It can even be chopped and form a succession of shorter pulses.

Each free oscillation period T0 of the mechanical oscillator defines a first half-wave A0 <1> followed by a second half-wave A0 <2> each occurring between two extreme positions defining the amplitude of oscillation of this mechanical oscillator , each alternation having an identical duration T0 / 2 and having a passage of the mechanical resonator through its zero position at a median instant. The two successive alternations of an oscillation define two half-periods during which the pendulum undergoes respectively an oscillating movement in one direction and then an oscillating movement in the other direction. In other words, an alternation here corresponds to a rocking of the pendulum in one direction or the other direction between its two extreme positions defining the amplitude of oscillation. In general, there is a variation in the oscillation period during which a braking pulse occurs and therefore a point variation in the frequency of the mechanical oscillator. In fact, the temporal variation concerns the only alternation during which the braking pulse intervenes. By "median instant", we understand an instant occurring substantially in the middle of the half-waves. This is precisely the case when the mechanical oscillator oscillates freely. On the other hand, for the alternations in which regulation pulses intervene, this median instant no longer corresponds exactly to the middle of the duration of each of these alternations due to the disturbance of the mechanical oscillator caused by the regulation device.

We will first describe the behavior of the mechanical oscillator in a first case of correction of its oscillation frequency, which corresponds to that shown in fig. 6. After a first period T0 then begins a new period T1, respectively a new alternation A1 during which a braking pulse P1 occurs. At the initial instant tD1 begins the alternation A1, the resonator 14 occupying a maximum positive angular position corresponding to an extreme position. Then comes the braking pulse P1 at the instant tp1 which is located before the median instant tN1 at which the resonator passes through its neutral position and therefore also before the corresponding median instant tNo of the undisturbed oscillation. Finally, the alternation A1 ends at the final time tF1. The braking pulse is triggered after a time interval TA <1> following the instant tD1 marking the start of the alternation A1. The duration TA1 is less than a half-wave T0 / 4 minus the duration of the braking pulse P1. In the example given, the duration of this braking pulse is much less than a half-wave T0 / 4.

In this first case, the braking pulse is therefore generated between the start of an alternation and the passage of the resonator through its neutral position in this alternation. The angular speed in absolute value decreases at the time of the braking pulse P1. Such a braking pulse induces a negative time shift TC1 in the oscillation of the resonator, as shown in fig. 6 the two curves 90a and 90b of the angular speed and also the two curves 92a and 92b of the angular position, that is to say a delay relative to the theoretical undisturbed signal (shown in dashed lines). Thus, the duration of alternation A1 is increased by a time interval TC1. The oscillation period T1, including the alternation A1, is therefore extended relative to the value T0. This generates a punctual decrease in the frequency of the mechanical oscillator and a momentary slowing down of the associated mechanism whose operation is clocked by this mechanical oscillator.

Referring to FIG. 7, the behavior of the mechanical oscillator will be described below in a second case of correction of its oscillation frequency. After a first period T0 then begins a new oscillation period T2, respectively an alternation A2 during which a braking pulse P2 occurs. At the initial instant to2 begins the alternation A2, the mechanical resonator then being in an extreme position (maximum negative angular position). After a quarter of period T0 / 4 corresponding to half-wave, the resonator reaches its neutral position at the median instant tN2. Then comes the braking pulse P2 at time tp2 which is located in alternation A2 after the median time tN2 at which the resonator passes through its neutral position. Finally, after the braking pulse P2, this alternation A2 ends at the final instant tF2 in which the resonator again occupies an extreme position (maximum positive angular position in the period T2) and therefore also before the corresponding final instant tF0 of the undisturbed oscillation. The braking pulse is triggered after a time interval TA2 following the initial instant tD2 of alternation A2. The duration TA2 is greater than a half-wave T0 / 4 and less than a half-wave T0 / 2 reduced by the duration of the braking pulse P2. In the example given, the duration of this braking pulse is much less than half-wave.

In the second case, considered, the braking pulse is therefore generated, in alternation, between the median instant at which the resonator passes through its neutral position (zero position) and the final instant at which this alternation ends. . The angular speed in absolute value decreases at the time of the braking pulse P2. Remarkably, the braking pulse here induces a positive time phase shift TC2 in the oscillation of the resonator, as shown in fig. 4 the two curves 90b and 90c of the angular speed and also the curves 92b and 92c of the angular position, ie an advance relative to the theoretical undisturbed signal (shown in broken lines). Thus, the duration of the alternation A2 is reduced by the time interval TC2. The oscillation period T2 including the alternation A2 is therefore shorter than the value T0. This consequently generates a punctual increase in the frequency of the mechanical oscillator and a momentary acceleration of the associated mechanism whose operation is clocked by this mechanical oscillator. This phenomenon is surprising and not intuitive, which is why the skilled person has ignored it in the past. Indeed, obtaining acceleration of the mechanism by a braking pulse is a priori surprising, but this is indeed the case when this step is clocked by a mechanical oscillator and when the braking pulse is applied to its resonator.

The aforementioned physical phenomenon for mechanical oscillators is involved in the synchronization process implemented in a timepiece according to the invention. Unlike general teaching in the watchmaking field, it is not only possible to decrease the frequency of a mechanical oscillator by braking pulses, but it is also possible to increase the frequency of such a mechanical oscillator also by braking pulses. Those skilled in the art expect to be able to practically only reduce the frequency of a mechanical oscillator by braking pulses and, as a corollary, to be able only to increase the frequency of such a mechanical oscillator by the application of driving pulses. during an energy supply to this oscillator. Such an intuition, which has imposed itself in the watchmaking field and therefore comes premium on board to the mind of a person skilled in the art, turns out to be false for a mechanical oscillator. Thus, as will be explained below in detail, it is possible to synchronize, via an auxiliary oscillator defining a master oscillator, a mechanical oscillator which is otherwise very precise, whether it temporarily has a frequency that is slightly too high or too low. It is therefore possible to correct a too high frequency or a too low frequency only by means of braking pulses. In summary, the application of a braking torque during an alternation of the oscillation of a balance-spring causes a negative or positive phase shift in the oscillation of this balance-spring depending on whether this braking torque is applied respectively before or after the balance-spring has passed through its neutral position.

The synchronization process resulting from the correction device incorporated in a timepiece according to the invention is described below. In fig. 8A is shown the angular position (in degrees) of a mechanical clockwork resonator oscillating with an amplitude of 300 ° during an oscillation period of 250 ms. In fig. 8B is shown the daily error generated by braking pulses of a millisecond (1 ms) applied in successive periods of oscillation of the mechanical resonator as a function of the instant of their application within these periods and therefore depending on the angular position of the mechanical resonator. Here, we start from the fact that the mechanical oscillator operates freely at a natural frequency of 4 Hz (undisturbed case). Three curves are given respectively for three pairs of force (100 nNm, 300 nNm and 500 nNm) applied by each braking pulse. The result confirms the physical phenomenon described above, namely that a braking pulse occurring in the first quarter of the period or the third quarter of the period generates a delay originating from a decrease in the frequency of the mechanical oscillator, whereas a braking pulse occurring in the second quarter of the period or the fourth quarter of the period generates an advance originating from an increase in the frequency of the mechanical oscillator. Then, it is observed that, for a given force torque, the daily error is equal to zero for a braking pulse occurring at the neutral position of the resonator, this daily error increasing (in absolute value) as one s' approaching an extreme position of the oscillation. At this extreme position where the speed of the resonator goes through zero and where the direction of movement changes, there is a sudden reversal of the sign of the daily error. Finally, in fig. 8C is given the braking power consumed for the three values of force torque mentioned above as a function of the instant of application of the braking pulse during a period of oscillation. As the speed decreases as it approaches the extreme positions of the resonator, the braking power decreases. Thus, as the daily error generated increases as it approaches the extreme positions, the braking power required (and therefore the energy lost by the oscillator) decreases significantly.

The error generated in fig. 8B can correspond in fact to a correction for the case where the mechanical oscillator has a natural frequency which does not correspond to a set frequency. Thus, if the oscillator has a too low natural frequency, braking pulses occurring in the second or fourth quarter of the oscillation period can allow correction of the delay taken by the free oscillation (undisturbed), this correction being more or less strong depending on the instant of the braking pulses within the oscillation period. On the other hand, if the oscillator has an excessively high natural frequency, braking pulses occurring in the first or third quarter of the oscillation period can allow correction of the advance taken by the free oscillation, this correction being more or less strong as a function of the instant of the braking pulses in the period of oscillation.

The teaching given above makes it possible to understand the remarkable phenomenon of the synchronization of a main mechanical oscillator (slave oscillator) on an auxiliary oscillator, forming a master oscillator, by the only periodic application of braking pulses on the resonator. mechanical slave at a braking frequency FFR advantageously corresponding to twice the set frequency F0c divided by a positive integer N, i.e. FFR = 2⋅F0c / N. The braking frequency is thus proportional to the set frequency for the oscillator master and only depends on this setpoint frequency as soon as the positive integer N is given. As the setpoint frequency is expected to be equal to a fractional number multiplied by the reference frequency, the braking frequency is therefore proportional to the reference frequency and determined by this reference frequency, which is supplied by the auxiliary oscillator which is by nature or by construction more precise than the main mechanical oscillator.

The aforementioned synchronization obtained by the correction device incorporated in the timepiece of the invention will now be described in more detail with the aid of figs. 9 to 22.

In fig. 9 is shown on the upper graph the angular position of the mechanical slave resonator, in particular of the balance-spring of a timepiece resonator, oscillating freely (curve 100) and oscillating with braking (curve 102). The frequency of the free oscillation is higher than the set frequency F0c = 4 Hz. The first braking pulses 104 (hereinafter also called 'pulses') intervene here once per oscillation period in half-cycle between passing through an extreme position and passing through zero. This choice is arbitrary because the system provided does not detect the angular position of the mechanical resonator; it is therefore just one possible hypothesis among others which will be analyzed later. So here we are in the case of a slowdown of the mechanical oscillator. The braking torque for the first braking pulse is here provided greater than a minimum braking torque to compensate for the advance that the free oscillator takes over a period of oscillation. This has the consequence that the second braking pulse takes place a little before the first within the quarter of the period in which these pulses occur. Curve 106, which gives the instantaneous frequency of the mechanical oscillator, indicates that the instantaneous frequency decreases below the set frequency from the first pulse. Thus, the second braking pulse is closer to the preceding extreme position, so that the braking effect increases and so on with the following pulses. In a transient phase, the instantaneous frequency of the oscillator therefore gradually decreases and the pulses gradually approach an extreme position of the oscillation. After a certain time, the braking pulses include passing through the extreme position where the speed of the mechanical resonator changes direction and the instantaneous frequency then begins to increase.

Braking has the particularity that it opposes the movement of the resonator regardless of the direction of its movement. Thus, when the resonator goes through an inversion of the direction of its oscillation during a braking pulse, the braking torque automatically changes sign at the instant of this inversion. There are then braking pulses 104a which have, for the braking torque, a first part with a first sign and a second part with a second sign opposite to the first sign. In this situation, we have the first part of the signal which intervenes before the extreme position and which opposes the effect of the second part which intervenes after this extreme position. If the second part decreases the instantaneous frequency of the mechanical oscillator, the first part increases it. The correction then decreases to stabilize finally and relatively quickly at a value for which the instantaneous frequency of the oscillator is equal to the set frequency (here corresponding to the braking frequency). Thus, the transient phase is followed by a stable phase, also called synchronous phase, where the oscillation frequency is substantially equal to the set frequency and where the first and second parts of the braking pulses have a substantially constant and defined ratio.

The graphs of FIG. 10 are similar to those of FIG. 9. The major difference is the value of the natural frequency of the free mechanical oscillator which is less than the set frequency F0c = 4 Hz. The first pulses 104 occur in the same half-wave as in fig. 9. As expected, a decrease in the instantaneous frequency given by the curve 110 is observed. The oscillation with braking 108 therefore momentarily takes even more delay in the transient phase, this until the pulses 104b begin to encompass the passage of the resonator. by an extreme position. From this moment, the instantaneous frequency begins to increase until reaching the set frequency, because the first part of the pulses occurring before the extreme position increases the instantaneous frequency. This phenomenon is automatic. Indeed, as long as the duration of the oscillation periods is greater than the duration of the setpoint period T0c, the first part of the pulse increases while the second part decreases and consequently the instantaneous frequency continues to increase until a stable situation where the setpoint period is substantially equal to the oscillation period. So we have the desired synchronization.

The graphs of FIG. 11 are similar to those of FIG. 10. The major difference comes from the fact that the first braking pulses 114 occur in another half-cycle than in FIG. 10, namely in a half-cycle between the passage through zero and the passage through an extreme position. According to what has been explained above, here we observe in a transient phase an increase in the instantaneous frequency given by the curve 112. The braking torque for the first braking pulse is provided here greater than a minimum braking torque to compensate for the delay that the free mechanical oscillator takes over a period of oscillation. This has the consequence that the second braking pulse takes place a little after the first within the quarter of the period in which these pulses occur. Curve 112 in fact indicates that the instantaneous frequency of the oscillator increases above the set frequency from the first pulse. Thus, the second braking pulse is closer to the following extreme position, so that the braking effect increases and so on with the following pulses. In the transient phase, the instantaneous frequency of the oscillation with braking 114 therefore increases and the braking pulses progressively approach an extreme position of the oscillation. After a certain time, the braking pulses include passing through the extreme position where the speed of the mechanical resonator changes direction. From this moment, we have a phenomenon similar to that exposed above. The braking pulses 114a then have two parts and the second part decreases the instantaneous frequency. This decrease in the instantaneous frequency continues until it has a value equal to the set value for the same reasons as given with reference to FIGS. 9 and 10. The frequency decrease stops automatically when the instantaneous frequency is substantially equal to the set frequency. We then obtain a stabilization of the frequency of the mechanical oscillator at the set frequency in a synchronous phase.

Using FIGS. 12 to 15, the behavior of the mechanical oscillator in the transition phase will be exposed for any instant when a first braking pulse occurs during an oscillation period, as well as the final situation corresponding to the phase synchronous where the oscillation frequency is stabilized on the set frequency. Fig. 12 represents an oscillation period with the curve S1 of the positions of a mechanical resonator. In the case considered here, the natural oscillation frequency F0 of the free mechanical oscillator (without braking pulses) is greater than the setpoint frequency F0c (F0> F0c). The oscillation period conventionally comprises a first alternation A1 followed by a second alternation A2, each between two extreme positions (tm-1, Am-1; tm, Am; tm + 1, Am + 1) corresponding to the amplitude oscillation. Next, there is shown, in the first half-cycle, a braking pulse "Imp1" whose mid-time position occurs at an instant ti and, in the second half-cycle, another braking pulse "Imp2" whose mid-time position occurs at a moment t2. Impulses Imp1 and Imp2 have a phase shift of T0 / 2, and they are special because they correspond, for a given profile of the braking torque, to corrections generating two unstable equilibria of the system. As these pulses occur respectively in the first and the third quarter of the oscillation period, they therefore brake the mechanical oscillator to an extent which exactly corrects the excessively high natural frequency of the free mechanical oscillator (with the frequency of braking selected for the application of the braking pulses). It will be noted that the impulses Imp1 and Imp2 are both first impulses, each one being considered for itself in the absence of the other. It will be noted that the effects of the impulses Imp1 and Imp2 are identical.

If a first pulse occurs at time t1 or t2, there will therefore theoretically be a repetition of this situation during the next oscillation periods and an oscillation frequency equal to the set frequency. Two things should be noted for such a case. First, the probability that a first impulse occurs exactly at time t1 or t2 is relatively small although possible. Second, if such a particular situation arises, it cannot last long. In fact, the instantaneous frequency of a balance spring in a timepiece varies a little over time for various reasons (amplitude of oscillation, temperature, change in spatial orientation, etc.). Although these reasons constitute disturbances that we generally seek to minimize in fine watchmaking, the fact remains that in practice such an unstable balance will not last very long. It will be noted that the higher the braking torque, the closer the times t1 and t2 are to the two passage times of the mechanical resonator by its neutral position which follow them respectively. It will also be noted that the smaller the difference between the natural oscillation frequency F0 and the setpoint frequency F0c, the more the times ti and t.2 also approach the two passage times of the mechanical resonator by its neutral position which follow them respectively.

Let us now consider what happens as soon as we deviate a little from the time positions t1 or t2 when applying the pulses. According to the teaching given with reference to fig. 8B, if an impulse occurs to the left (anterior time position) of the impulse lmp1 in the zone Z1a, the correction increases so that during the following periods, the previous extreme position Am-1 will gradually approach the impulse braking. On the other hand, if a pulse occurs on the right (posterior time position) of the pulse Imp1, on the left of the zero position, the correction decreases so that during the following periods the pulses drift towards this zero position where the correction becomes nothing. Then the effect of the pulse changes and an increase in the instantaneous frequency occurs. As the natural frequency is already too high, the pulse will quickly drift towards the extreme position Am. Thus, if a pulse takes place to the right of the pulse Imp1 in the zone Z1 b, the following pulses will gradually approach the next extreme position Am. The same behavior is observed in the second half-wave A2. If an impulse takes place to the left of the impulse Imp2 in the zone Z2a, the following impulses will gradually approach the previous extreme position Am. On the other hand, if a impulse takes place to the right of the impulse Imp2 in the zone Z2b , the following pulses will gradually approach the next extreme position Am + 1. It will be noted that this formulation is relative because in reality the frequency of application of the braking pulses is imposed by the master oscillator (given braking frequency), so that it is the periods of oscillation which vary and in fact c ' is the extreme position in question which approaches the instant of application of a braking pulse. In conclusion, if a pulse occurs in the first alternation A1 at a time other than ti, the instantaneous oscillation frequency evolves in a transient phase during the following oscillation periods so that one of the two extreme positions of this first alternation (positions of reversal of the direction of movement of the mechanical resonator) gradually approaches the braking pulses. The same is true for the second alternation A2.

FIG. 13 shows the corresponding synchronous phase a final stable situation occurring after the transient phase described above. As already explained, as soon as the passage through an extreme position occurs during a braking pulse, this extreme position will lock onto the braking pulses provided that these braking pulses are configured (the force torque and the duration) to be able to sufficiently correct the time drift of the free mechanical oscillator at least by a braking pulse intervening entirely, as the case may be, just before or just after an extreme position. Thus, in the synchronous phase, if a first pulse occurs in the first half-wave A1, either the extreme position Am-1 of the oscillation is set on the impulses Imp1a, or the extreme position Amde the oscillation is set on the pulses Imp1b . In the case of a substantially constant torque, the pulses Imp1a and Imp1 b each have a first part whose duration is shorter than that of their second part, so as to exactly correct the difference between the too high natural frequency of the slave main oscillator and the set frequency imposed by the master auxiliary oscillator. Similarly, in the synchronous phase, if a first pulse occurs in the second half-wave A2, either the extreme position Amde the oscillation is set on the impulses Imp2a, or the extreme position Am + 1 of the oscillation is set on the pulses Imp2b .

Note that the impulses Imp1a, respectively Imp1b, Imp2a and Imp2b occupy stable relative time positions. Indeed, a slight deviation to the left or to the right of one of these pulses, due to an external disturbance, will have the effect of bringing a next pulse towards the initial relative time position. Then, if the temporal drift of the mechanical oscillator varies during the synchronous phase, the oscillation will automatically undergo a slight phase shift so that the ratio between the first part and the second part of the impulses Imp1a, respectively Imp1b, Imp2a and Imp2b varies to an extent which adapts the correction generated by the braking pulses to the new frequency difference. Such behavior of the timepiece according to the present invention is truly remarkable.

Figs. 14 and 15 are similar to Figs. 12 and 13, but for a situation where the natural frequency of the oscillator is lower than the set frequency. Consequently, the pulses Imp3 and Imp4, corresponding to an unstable equilibrium situation in the correction made by the braking pulses, are respectively located in the second and the fourth quarter of period (instants t3 and t4) where the pulses generate an increase of the oscillation frequency. The explanations in detail will not be repeated here because the behavior of the system follows from the previous considerations. In the transient phase (fig. 14), if a pulse takes place in the alternation A3 to the left of the pulse Imp3 in the zone Z3a, the previous extreme position (tm-1, Am-1) will gradually approach the following pulses. On the other hand, if a pulse takes place to the right of the Imp3 pulse in the zone Z3b, the next extreme position (tm, Am) will gradually approach the following pulses. Similarly, if a pulse takes place in the alternation A4 to the left of the pulse Imp4 in the zone Z4a, the previous extreme position (tm, Am) will gradually approach the following pulses. Finally, if an impulse takes place to the right of impulse Imp4 in zone Z4b, the next extreme position (tm + 1, Am + 1) will gradually approach the following impulses during the transition phase.

In the synchronous phase (fig. 15), if a first pulse occurs in the first half-wave A3, either the extreme position Am-1 of the oscillation is set on the pulses Imp3a, or the extreme position Amde the oscillation is set on Imp3b pulses. In the case of a substantially constant torque, the pulses Imp3a and Imp3b each have a first part whose duration is longer than that of their second part, so as to exactly correct the difference between the too low natural frequency of the oscillator main slave and the setpoint frequency imposed by the master auxiliary oscillator. Similarly, in the synchronous phase, if a first pulse occurs in the second alternation A4, either the extreme position Amde the oscillation is set on the impulses Imp4a, or the extreme position Am + 1 of the oscillation is set on the pulses Imp4b . The other considerations made in the context of the case described above with reference to FIGS. 12 and 13 apply by analogy to the case of Figs. 14 and 15. In conclusion, whether the natural frequency of the free mechanical oscillator is too high or too low and whatever the instant of the application of a first braking pulse within an oscillation period, the correction device of the invention is effective and quickly synchronizes the frequency of the mechanical oscillator, clocking the movement of the mechanical movement, on the set frequency which is determined by the reference frequency of the master auxiliary oscillator, which controls the braking frequency at which the braking pulses are applied to the resonator of the mechanical oscillator. This remains true if the natural frequency of the mechanical oscillator varies and even if it is, in certain periods of time, greater than the set frequency, while in other periods of time it is lower than this set frequency.

The teaching given above and the synchronization obtained thanks to the characteristics of the timepiece according to the invention also apply to the case where the braking frequency for the application of the braking pulses is not equal to the set frequency. When applying a pulse per oscillation period, the pulses taking place at unstable positions (t1, Imp1; t2, Imp2; t3, Imp3; t4, Imp4) correspond to corrections to compensate for time drift in a single oscillation period. On the other hand, if the planned braking pulses have a sufficient effect to correct a time drift during several oscillation periods, it is then possible to apply a single pulse per time interval equal to these several oscillation periods. We will then observe the same behavior as for the case where a pulse is generated by period of oscillation. By considering the periods of oscillation where the pulses intervene, we have the same transient phases and the same synchronous phases as in the case described above. In addition, these considerations are also correct if there is a whole number of alternations between each braking pulse. In the case of an odd number of alternations, one alternates, as the case may be, from alternation A1 or A3 to alternation A2 or A4 in FIGS. 12 to 15. As the effect of two pulses offset by an alternation is identical, it is understood that synchronization is carried out as for an even number of alternations between two successive braking pulses. In conclusion, as already indicated, the behavior of the system described with reference to FIGS. 12 to 15 is observed as soon as the braking frequency FFR is equal to 2F0c / N, F0c being the set frequency for the oscillation frequency and N a positive integer.

Although not very interesting, it will be noted that synchronization is also obtained for a braking frequency FFRsuperior to twice the set frequency (2F0), namely for a value equal to N times F0 with N> 2. In a variant with FFR = 4F0, there is just a loss of energy in the system with no effect in the synchronous phase, because one pulse out of two occurs at the neutral point of the mechanical resonator. For a higher FFR braking frequency, the pulses in the synchronous phase which do not occur at the extreme positions cancel their effects two by two. We therefore understand that these are theoretical cases without much practical sense.

Figs. 16 and 17 show the synchronous phase for a variant with a braking frequency FFR equal to a quarter of the set frequency, a braking pulse therefore occurring every four periods of oscillation. Figs. 18 and 19 are partial enlargements respectively of FIGS. 16 and 17. Fig. 16 relates to a case where the natural frequency of the main oscillator is greater than the set frequency F0c = 4 Hz, while fig. 17 relates to a case where the natural frequency of the main oscillator is greater than this set frequency. It is observed that only the oscillation periods T1 * and 12 *, in which braking pulses Imp1b or Imp2a, respectively Imp3b or Imp4a, intervene, exhibit a variation relative to the natural period T0 *. The braking pulses generate a phase shift only in the corresponding periods. Thus, the instantaneous periods here oscillate around an average value which is equal to that of the setpoint period. Note that, in Figs. 16 to 19, the instantaneous periods are measured from a zero crossing on a rising edge of the oscillation signal to such a following crossing. Thus, the synchronous pulses which occur at the extreme positions are entirely included in periods of oscillation. To be complete, fig. 20 shows the specific case where the natural frequency is equal to the set frequency. In this case, the oscillation periods T0 * all remain equal, the braking pulses Imp5 occurring exactly at extreme positions of the free oscillation with first and second parts of these pulses which have identical durations (case of a constant braking torque), so that the effect of the first part is canceled out by the opposite effect of the second part.

In an improved variant, the synchronization device is arranged so that the braking frequency can take several values, preferably a first value in an initial phase of the operation of the synchronization device and a second value, less than the first value, in a normal operating phase following the initial phase. In particular, the duration of the initial phase will be selected so that the normal operating phase occurs while the synchronous phase has probably already started. More generally, the initial phase includes at least the first braking pulses, following the engagement of the synchronization device, and preferably most of the transient phase. By increasing the frequency of the braking pulses, the duration of the transient phase is reduced. In addition, this variant makes it possible, on the one hand, to optimize the braking efficiency during the initial phase to ensure the physical process leading to synchronization and, on the other hand, to minimize the braking energy and therefore the energy losses for the main oscillator during the synchronous phase which lasts as long as the synchronization device is not deactivated and the mechanical movement works. The first braking pulses can occur near the neutral position of the resonator where the effect of the braking is less on the time difference generated for the oscillation of the main oscillator. On the other hand, once synchronization has been established, the braking pulses take place near the extreme positions of this oscillation where the effect of the braking is greatest.

Referring to Figs. 21 and 22, a first variant of a second embodiment of the invention will be described which is surprising by the simplicity of its electromagnetic braking device. This second embodiment differs from the first mode essentially by the magnetic system of the electromagnetic braking device which consists, in the first variant, of a single bipolar magnet 60 carried by the balance 8A of the mechanical resonator 6A and, in a second variant, by a single pair of bipolar magnets. In the first variant, when the resonator 6A is in its neutral position (situation shown in FIG. 21), a reference half-axis 62 starting from the axis of oscillation 34 and passing through the center of the magnet 60 defines a zero angular position ("0") in a polar coordinate system centered on the axis of oscillation and fixed relative to the stage of the watch movement. The coil 28, which completes the electromagnetic braking device in addition to the magnetic system, is integral with the plate and has an angular offset relative to the zero angular position. Preferably, the angular offset of the coil is substantially 180 °, as shown in FIG. 21.

In fig. 22 shows the curve 70 of the angular position of the pendulum 8A as a function of time, in the useful operating range of the mechanical oscillator considered which has in this range an amplitude greater than 180 ° and preferably greater than 200 ° (case shown), and the curve 72 of the voltage induced in a synchronous phase of the operation of the synchronization device. Thus, in each alternation of the oscillation of the mechanical resonator 6A, two pulses of induced voltage 74A and 74B are observed having substantially the shape of a period of a sinus. It is noted that the pulses 74A and 74B are separated two by two by time zones without induced voltage in the coil 28. In a variant ensuring great stability in the running of the timepiece, the distinct time intervals TP, defined by the short-circuit pulses 58A generated at the set frequency F0c and thus occurring in each oscillation period, are substantially equal to or greater (case shown) than the time zones with no induced voltage in the coil around the two extreme positions of the mechanical resonator in the useful operating range. However, as will be seen later, this condition is not necessary, the time intervals TP can be less than the duration of these time zones without induced voltage.

It is observed that, provided that the natural time drift of the timepiece remains within a nominal range for which the synchronization device has been dimensioned and generally after a transient phase which follows the activation of the synchronization device, this timepiece enters a stable and synchronous phase and where the mechanical oscillator has the set frequency F0c at which the short-circuit pulses 58A are generated here, and this regardless of the angular position of the balance 8A during 'a first short circuit pulse. Fig. 22 corresponds to a situation where the natural oscillation frequency F0 of the mechanical oscillator is slightly lower than the set frequency F0c. It follows from this situation that, in each oscillation period T0c, a first separate braking pulse, which is generated in the initial zone of each short-circuit pulse by an induced voltage pulse 74A and which occurs in the second half alternation A22of the second alternation A2 (at the start of the separate time intervals TP), is stronger than a second distinct braking pulse which is generated in the final zone of each short-circuit pulse by an induced voltage pulse 74Bet which intervenes in the first half-wave A11of the first half-wave A1 (at the end of the separate time intervals TP). Two braking pulses are distinct when they are separated by a time interval of non-zero duration.

Thus, in the synchronous phase, during each time interval TP where there is a short circuit of the coil, the positive phase shift generated by the voltage pulse 74B in each half-wave A22 is greater than the negative phase shift generated by the voltage pulse 74 In each half-wave A11, so a correction of the running of the timepiece occurs here in each period of oscillation to ensure synchronization of the mechanical oscillator on the basis of reference time. As indicated previously, the generation of short-circuit pulses at the set frequency is a special case. In another variant, short-circuit pulses are generated with a lower frequency corresponding to a fraction of the set frequency. More generally, it is expected that the time distance DT, separating the same characteristic instant from any two successive short-circuit pulses, satisfies the mathematical relation DT = M⋅T0c / 2, M being any positive integer. Thus, in the case of a periodic generation of the braking pulses, the triggering frequency FD of these braking pulses is selected to satisfy the mathematical relation FD = 2⋅F0c / M (note that the two distinct braking pulses, generated in each time interval TP respectively when the two induced voltage pulses 74A and 74B appear, are considered together as the same braking pulse for the question of the temporal distances and the triggering frequency). Those skilled in the art will be able to select a frequency which is sufficiently high, and therefore a value of M which is not too high, to ensure the desired synchronization.

In a second variant of the second embodiment, the electromagnetic braking device comprises a magnetic system formed by a pair of permanent magnets with axial magnetization and opposite polarities, these two magnets being arranged symmetrically with respect to a half pendulum reference axis and close enough to each other to add two induced voltage lobes which they generate respectively when this pair of magnets pass opposite the coil. The reference half-axis defines a zero angular position when the mechanical resonator is in its neutral position. The coil has an angular offset relative to the zero angular position so that a voltage induced in this coil occurs, when the mechanical oscillator oscillates in the useful operating range, at least in an alternation of each period of oscillation substantially before or after the mechanical resonator has passed through its neutral position in this alternation. The angular offset of the coil is also preferably equal to 180 °. The extreme angular positions of the mechanical resonator in the useful operating range are, in absolute values, greater than the angular offset which is defined as the minimum angular distance between the zero angular position and the angular position of the center of the coil. This second variant corresponds to the electromagnetic device shown in FIG. 23, but without the second pair of magnets 66, 67 which relates to the third embodiment which will be described below.

In a third embodiment, shown in FIGS. 23 to 25, the magnetic system of the electromagnetic braking device consists of a first pair of bipolar magnets 64, 65 and a second pair of bipolar magnets 66, 67 both carried by the balance 8B of the mechanical resonator 6B , as well as a coil 28. Each pair of magnets has an axial magnetization of opposite polarities. The two magnets of the first pair are arranged symmetrically relative to a reference half-axis 62A of the balance 8B, this reference half-axis defining a zero angular position when the mechanical resonator is in its neutral position. In fig. 23, it will be noted that the balance is in an angular position θ equal to 90 ° (θ = 90 °). The coil 28, as in the second embodiment, has an angular offset relative to the zero angular position, this offset being preferably substantially equal to 180; but other angular offsets can be provided in other variants. The induced voltage curve 76 generated in the coil when the mechanical resonator oscillates is shown in FIG. 24, superimposed on the curve 70 giving the angular position of the balance 8B.

The positioning of the coil 28 at an angle of 180 ° (variant shown in Fig. 23) is a preferred variant, because the electromagnetic system that the coil forms with the first pair of magnets 64, 65 generates in each alternating two pulses of induced voltage 78A and 78B exhibiting symmetry relatively to the instant of passage of the resonator 6B through its neutral position. There is therefore a pulse 78Ad in each first half-wave A11, A21 and a pulse 78B in each second half-wave A12, A22. Thus, the induced voltage pulses 78A and 78B have substantially the same amplitude and are each located at the same time distance from a passage of the mechanical resonator 6B through an extreme angular position, so that they are all capable of generating, during '' a coil short-circuit, a braking torque of the same intensity and a phase shift, positive or negative depending on the case, of the same value in the oscillation of the mechanical resonator. Then, as has been explained above, it will be noted that an angular offset of 180 ° also has the advantage of great efficiency for the braking pulses generated. In addition, it will be noted that the amplitude of the balance in the useful operating range of the mechanical oscillator is conventionally greater than 180 °, which therefore allows the generation of pulses of induced voltage and thus being able to generate braking pulses. , by reducing the impedance between the two terminals of the coil 28 to correct the running of the timepiece.

In a first variant shown in FIG. 24, the value of the separate time intervals TP is substantially equal to or greater than the duration of a time zone without induced voltage in the coil 28 around each extreme angular position of the mechanical resonator in the useful operating range of the mechanical oscillator. However, this value of the separate time intervals TP is expected to be less than the set half-period, ie TP <T0c / 2. In the synchronous phase of the synchronization method according to this first variant, the short-circuit pulses 58B are called between two induced voltage pulses 78A, 78B surrounding an extreme angular position and two distinct braking pulses intervene respectively at the start and at the end of each time interval TP, these two distinct braking pulses corresponding to two quantities of energy taken from the mechanical resonator which are variable (the variation of one being opposite to the variation of the other, so that if one of the two amounts of energy increases or decreases the other respectively decreases or increases), according to a time drift positive or negative of the mechanical oscillator in question. Note that fig. 24 corresponds to the particular case where the natural frequency of the mechanical oscillator is equal to the set frequency, so that the two quantities of energy mentioned above are here identical.

In Figure 25, similar to FIG. 24, a second variant is shown in which the value of the separate time intervals TP is less than the duration of a time zone without induced voltage in the coil 28 around each extreme angular position of the mechanical resonator. The desired synchronization is also obtained. Indeed, in the synchronous phase, the short-circuit pulses 58C also remain in a time window which is framed by two pulses of induced voltage 78A, 78B surrounding an extreme angular position. The time position of the separate time intervals TP can vary within this time window during at least one terminal part of the transient phase (58C1 pulse) or in the synchronous phase if the natural frequency of the mechanical oscillator is very close to the frequency of the setpoint, especially if it varies very slightly around this value. Generally, we observe in the synchronous phase, depending on whether the time drift of the mechanical oscillator is negative or positive, short-circuit pulses 58C2 or 58C3 which occur respectively in the half-alternations A12 and A21 of periods of oscillation in a partially simultaneous manner respectively with the induced voltage pulses 78Bet 78A, so that they generate braking pulses in the respective half-vibrations. Only the aforementioned electromagnetic system, formed of the coil and the first pair of magnets, intervenes to ensure the desired synchronization in the synchronous phase of the synchronization process, the second pair of magnets then being without influence for this synchronization process.

The second pair of bipolar magnets 66, 67, which is momentarily coupled to the coil 28 in each alternation of the oscillation of the mechanical resonator, essentially serves for the electrical supply of the synchronization device, although it can intervene in a transient phase (initial phase after activation of the synchronization device) of the synchronization process. The timepiece includes a power supply circuit, formed by a rectifier circuit of an induced voltage in the coil and a storage capacity, and the second pair of bipolar magnets has a middle semi-axis 68 between its two magnets which is offset from the angular offset presented by the coil 28 relative to the reference half-axis 62A, so that this center axis is aligned with the center of the coil when the mechanical resonator is in its rest position. The supply circuit is connected, on the one hand, to a terminal of the coil and, on the other hand, to a reference potential of the synchronization device at least periodically when the mechanical resonator passes through its neutral position, but preferably constantly. The second pair of magnets generates pulses of induced voltage 80A and 80Blors of the balance 8B passing through the angular position zero, these pulses having a greater amplitude than the pulses generated by the first pair of magnets and used to power the the storage capacity, the voltage of which is represented by curve 82 in FIG. 24.

Referring to Figs. 26, 27 and 28A - 28C, a fourth embodiment of the invention will be described below. This fourth embodiment is distinguished from the other embodiments essentially by the arrangement of the magnetic system. The shaft 82 of the balance 8C is pivoted between the plate 5 and a balance bridge 7 around the axis of oscillation 34. A bipolar magnet 84 with radial magnetization is arranged on the shaft 82 and placed in an opening 87 d 'A plate 86 of material with high magnetic permeability, in particular of ferromagnetic material. The plate 86 defines a magnetic circuit with a core 89 around which a coil 28C is arranged, in the manner of a conventional timepiece motor. The plate 86 has two isthmus 88 at the opening 87 which partially prevent the magnetic flux of the magnet from closing in on itself without passing through the coil core. However, preferably, these isthmus are provided less thin than in the case of a timepiece motor to limit the variation of potential magnetic energy of the permanent magnet 84 as a function of its angle of rotation.

Figs. 28A to 28C are similar to figs. 5A to 5C, but for the fourth embodiment. The voltage curve induced in figs. 28A and 28B corresponds to a particular case where the amplitude of oscillation is substantially equal to 180 °. For a higher amplitude, the voltage curve induced in the coil 28C corresponds to the curve shown in fig. 28C. This last figure relates to a particular case where the natural oscillation frequency F0 of the mechanical oscillator is equal to the set frequency. As the braking generated by the braking pulses 50C is weak, the amplitude of oscillation of the resonator 6C is slightly greater than that occurring in FIGS. 28A and 28B where the braking pulses 56, respectively 57 generate more substantial braking. The pulses 50C do not generate a time phase shift in the oscillation of the mechanical resonator, since they have central symmetry relative to the instant of passage of the resonator 6C through an extreme angular position on the graph of the braking torque. It will be noted that the two parts TBet TAdes separate time intervals TP, occurring respectively on both sides of the instant of passage of the resonator 6C through an extreme angular position, are here equal since the natural frequency is equal to the set frequency. Thus the adjacent half-cycles A22 and A11 have the same duration.

As a reminder, the time intervals TPs are defined by the short-circuit pulses 58 which have, between their respective beginnings, a time distance DT determined by the reference time base. In the present example, the short-circuit pulses 58 are generated with a trigger frequency FD equal to the set frequency, so that the time distances DT are here equal to a set period T 0 c.

In the case of a natural frequency F0 too high, the first part TB of the remote time intervals TP is less than the second part TA and the braking pulses 56 generated during these remote time intervals, by the short- corresponding circuits, intervene substantially in first half-alternations A1i (almost entirely in the specific example represented), so that they decrease the frequency of the mechanical oscillator to synchronize it with the auxiliary oscillator of the time base of reference and thus impose on this mechanical oscillator the set frequency F0c. In the case of a too low natural frequency F0, the first part TB of the remote time intervals TP is greater than the second part TA and the braking pulses 57 generated during these remote time intervals, by the corresponding short-circuit pulses, intervene substantially in second half-alternations A22 (also almost entirely in the specific example shown), so that they increase the frequency of the mechanical oscillator to synchronize it with the auxiliary oscillator.

Claims (13)

1. Timepiece (2) comprising a mechanical movement (4) which includes: - a mechanism (12) indicating at least one time datum, - a mechanical resonator (6,6A, 6B, 6C) capable of oscillating along a general axis of oscillation around a neutral position corresponding to its state of minimum potential energy, - a maintenance device (18) of the mechanical resonator forming with this mechanical resonator a mechanical oscillator which is arranged to clock the progress of the indicator mechanism, - an auxiliary oscillator (35) forming a reference time base (22); the timepiece further comprising a synchronization device (20) arranged to slave the average frequency of the mechanical oscillator to a set frequency determined by the reference time base, the synchronization device comprising an electromagnetic braking device of the mechanical resonator, this electromagnetic braking device being formed of at least one coil (28,28C) and at least one permanent magnet (30,32; 60; 64,65; 84) which are arranged so that, in a useful operating range of the mechanical oscillator, an induced voltage is generated by the oscillation of the mechanical resonator between the two terminals of the coil in each alternation of this oscillation; the synchronization device being arranged to be able to temporarily decrease the impedance between the two terminals of the coil; characterized in that the synchronization device is arranged so as to reduce the impedance between the two terminals of the coil during distinct time intervals TP and in such a way that the beginnings of any two successive distinct time intervals, among said intervals of distinct times, have between them a time distance DT equal to a positive integer N multiplied by half of a set period T0c for the mechanical oscillator, i.e. a mathematical relation DT = N⋅T0c / 2, the synchronization device being arranged to determine by means of the reference time base the start of each of said separate time intervals so as to satisfy said mathematical relationship between the time distance DT and the setpoint period T0c. 2. Timepiece according to claim 1, characterized in that the synchronization device is arranged so as to periodically trigger said separate time intervals TP, which have the same value, and so that the trigger frequency FD is equal to twice a setpoint frequency F0c, equal by definition to the inverse of the setpoint period T0c, divided by a positive integer M, i.e. FD = 2⋅F0c / M, the value of the separate time intervals TP being less than the half setpoint period, i.e. TP <T0c / 2. 3. Timepiece according to claim 1 or 2, characterized in that the mechanical resonator is formed by a balance (8, 8A, 8B, 8C) oscillating around an axis of oscillation (34). 4. Timepiece according to claim 3, characterized in that the balance carries said at least one permanent magnet and a support (5) of the mechanical resonator carries said at least one coil. 5. Timepiece according to any one of the preceding claims, characterized in that the electromagnetic braking device is arranged so that an induced voltage is generated in said at least one coil substantially without interruption for any oscillation of the mechanical resonator in said useful operating range of the mechanical oscillator. 6. Timepiece according to claim 5, characterized in that the value of the separate time intervals TP is less than a quarter of the set period T0c, ie TP <T0c / 4. 7. Timepiece according to claim 4, characterized in that the electromagnetic braking device comprises a magnetic system carried by the balance and formed by a pair of bipolar magnets (64,65) with axial magnetization and opposite polarities, these two bipolar magnets being arranged symmetrically relative to a reference half-axis (62A) of the balance, this reference half-axis defining a zero angular position when the mechanical resonator is in its neutral position; and in that said coil has an angular offset relative to the zero angular position so that an induced voltage in this coil occurs substantially, when the mechanical oscillator oscillates in its useful operating range, in each alternation alternately before and after the passage of the mechanical resonator through its neutral position in this alternation, the extreme angular positions of the mechanical resonator in said useful operating range being, in absolute values, greater than said angular offset which is defined as the minimum angular distance between the zero angular position and the angular position of the center of the coil. 8. Timepiece according to claim 7, characterized in that, in the useful operating range of the mechanical oscillator, the separate time intervals TPs are substantially equal to or greater than time zones without induced voltage in said coil around the two extreme angular positions of the mechanical resonator. 9. Timepiece according to claim 7 or 8, characterized in that said angular offset is substantially equal to 180 °. 10. Timepiece according to any one of the preceding claims, characterized in that it comprises a supply circuit (44) formed by a storage capacity and by a rectifier circuit of a voltage induced in the coil by at least a permanent magnet when the mechanical resonator oscillates. 11. Timepiece according to claim 10, characterized in that the supply circuit is constantly connected, on the one hand, to a terminal of said coil and, on the other hand, to a reference potential of the device synchronization; and in that said at least one permanent magnet generating the induced voltage rectified by the rectifier circuit, the coil and the supply circuit are arranged so that, in the useful operating range of the mechanical oscillator, the electrical energy accumulated in the storage capacity is sufficient to power the synchronization device. 12. Timepiece according to any one of claims 7 to 9, characterized in that it comprises a supply circuit (44) formed by a storage capacity and by a rectifier circuit of a voltage induced in the coil by another pair of permanent magnets (66, 67) when the mechanical resonator oscillates, the other pair of permanent magnets having a middle axis (68) between its two permanent magnets and being momentarily coupled to the coil in each alternation of the oscillation of the mechanical resonator, said center axis being substantially offset from said angular offset relative to said reference half-axis (62A) so that this center axis is substantially aligned with the center of the coil when the mechanical resonator is in its rest position ; and in that the supply circuit is connected, on the one hand, to a terminal of said coil and, on the other hand, to a reference potential of the synchronization device at least periodically when the mechanical resonator passes through its neutral position. 13. Timepiece according to any one of the preceding claims, characterized in that the synchronization device is arranged so as to generate a short circuit between the two terminals of said coil during said separate time intervals.