EP0190591A1 - Motor assembly capable of functioning at a high speed - Google Patents

Abstract

In order to increase the maximum functioning speed of a stepping motor (1), with which a control circuit (2) is associated, it is equipped with a coil (5), the resistance and the self-inductance of which have values which are lower than the optimum values which they should have in order that the output of this motor has a maximum value when it is supplied by continuous drive pulses of given voltage and of an optimum duration which corresponds to this voltage and to the optimum values of the resistance and of the inductance of the coil. Furthermore, in order to avoid the motor not having this maximum value or roughly this value when the motor functions at low speed, the latter is then controlled by drive impulses of the same voltage and of the same duration as the continuous impulses but chopped. <IMAGE>

Description

The present invention relates to an engine assembly capable of operating at high speed and usable, in particular but not exclusively, in an electronic watch with analog display.

More specifically, the invention relates to a motor assembly comprising a stepping motor and its control circuit, which makes it possible to obtain a rotation speed significantly increased compared to that which the stepping motors currently used in watchmaking can achieve.

Note that here the expression "step by step" means that the motor is designed so that its rotor can move in spurts by stopping in one or more well-defined rest positions. This does not preclude the possibility of rotating it continuously, that is to say without the rotor stopping when it passes through a rest position and even without being braked at this place.

In the case of a watch which has a seconds hand driven by a stepping motor and in which the correction is done mechanically, it suffices to have a motor which takes a number of steps per second equal to that which must make the second hand to go from one division to the next. This number of steps is most often equal to one and is perfect for two or six. This therefore means that the motor must only take a maximum of six steps per second, which does not present any difficulty.

However, a problem already arises when, even in a watch without a seconds hand, you want to make corrections or time zone changes by directly controlling the engine. In this case, it is necessary to be able to run the motor at a frequency significantly higher than 6 Hz, that is to say six steps per second. However, currently, with Lavet type stepper motors for example, conventionally controlled, it is practically impossible to go beyond 64 Hz and even to guarantee proper operation of the motor at more than 50 Hz. we take the case of a minute and hour hand watch in which it the engine must take two steps for the minute hand to advance by one division and if, in addition, we want to make a time zone change of six hours, which corresponds to the longest correction that we has to do if we manage to be able to rotate the motor in both directions, using for example the solution which is the subject of US patent 4 112 671 to control it, we see that, even with a frequency of 64 Hz, it will take fifteen seconds to make this single correction, which is already very long.

As for the electronic time setting in a watch with a second hand, this can only be envisaged if at least two motors are used to drive the hands, for example one for the second hand and a for the minutes and hours or one for the second hand and the minute hand and another for the hour hand. The use of a single motor would lead to exorbitant correction times.

Patent application EP 0 103 542 indicates an interesting solution which makes it possible to operate a Lavet type motor in forward and reverse directions at a speed much greater than 50 or 64 steps per second. This solution consists in applying to the motor first a driving impulse for launching in forward or reverse direction to effectively cause the engine to start in the desired direction of rotation, then a train of simple maintenance pulses of alternating polarity and of shorter duration than that of the starting pulse, and finally a single stop pulse of polarity opposite to that of the last pulse-of maintenance and longer than this one. In the case of forward travel, the start pulse is, like the others, a single pulse whereas, for reverse travel, this start pulse is formed of three elementary pulses of alternating polarity, the polarity of the first pulse. depending on the rest position initially occupied by the rotor. The role of the first elementary pulse is to start the engine in forward so that the rotor acquires sufficient energy to be able to take the first step in reverse under the effect of the following two pulses.

The instants of application of the drive and stop pulses relative to the start of the start pulse can either be determined in advance, taking into account the characteristics of the motor, or be fixed by detecting a quantity representative of the movement of the rotor, such as for example the voltage induced by the latter in the motor coil.

This way of controlling a conventional Lavet motor makes it possible to reach both a forward and a reverse frequency very close to the theoretical synchronism frequency which is of the order of 200 Hz and therefore leads to correction times reasonable for a watch showing only the minutes and hours but still far too long for a watch with a second hand. In addition, it requires a relatively complicated motor control circuit.

Another solution for increasing the maximum operating speed of an engine is the subject of the Swiss patent application filed by the applicant on May 4, 1984 under No. 2 180/84. This is mainly to replace the single coil or each coil of an existing stepper motor two or more windings and connect these windings in series to operate the motor at low speed and in parallel to rotate it high speed. Naturally, it is most often advantageous that the total number of turns of the windings and the sum of their resistances are equal or almost respectively to the number of turns and to the resistance of the coil that they replace, this so that despite the modification made to its winding the motor continues to have practically the same characteristics and the same performance when it rotates slowly. On the other hand, it is clear that, when the windings are connected in parallel, the resistance and self inductance of the assembly are lower than those of a single coil and that, consequently, the transfer of energy from the supply voltage source to the motor rotor is increased and facilitated. It follows that the duration or, more generally, the energy of the driving pulses to be applied to the winding can be reduced to a large extent and the maximum operating speed of the motor significantly increased.

In the case of a Lavet type motor, this solution alone makes it possible, like that of patent application EP 0 103 542, to reach a frequency close to 200 Hz but not requiring not such a complicated control circuit, even if it is intended to run the engine in both directions.

It also has at least two advantages. The first is that it can be applied to different types of stepping motors, especially among those currently used in the manufacture of watches or other timepieces. The second is that, combined with the well-known technique of servo-control, it effectively achieves the speed it takes to be able to equip a seconds hand watch with an electronic correction system and which must be at least 1000 steps per second.

Unfortunately, this solution also has drawbacks. On the one hand, it complicates the manufacture of the motor from the point of view of winding. On the other hand, it greatly increases both the number of output terminals and the number of power transistors that must be provided in the control circuit to supply the motor. For example, if we replace the coil of a classic Lavet motor with two windings, the number of output terminals goes from two to at least four and the number of transistors from four to at least seven. However, we know that power transistors occupy a large area on an integrated circuit board and the more terminals there are, the less reliable and more expensive.

The object of the present invention is to provide a solution for increasing the maximum operating speed of a stepping motor which does not have these drawbacks.

où T est le couple utile maximal que peut fournir le moteur, n le nombre de pas par tour effectués par son rotor, V la tension à laquelle il est alimenté et 1 le courant moyen qui lui est appliqué pendant une seconde; et un circuit de commande pour faire fonctionner ce moteur à petite et à grande vitesse en lui appliquant des impulsions motrices respectivement de fréquence basse et de fréquence élevée, les valeurs réelles de la résistance et de l'inductance propre de la bobine du moteur sont inférieures aux valeurs optimales et les impulsions motrices de fréquence basse sont telles que ce moteur ait un rendement de valeur sensiblement égale à la valeur maximale lorsqu'il fonctionne à petite vitesse.This object is achieved thanks to the fact that, in a motor assembly according to the invention which comprises a stepping motor provided with a rotor, a stator and a coil magnetically coupled to the stator and for which there are values optimal resistance and self inductance of the coil and an optimal duration of the driving pulses to be applied to this coil to control it which allow it to have a yield n of maximum value when it is supplied at a determined voltage, this return being defined by the relation

where T is the maximum useful torque that the motor can supply, n the number of steps per revolution made by its rotor, V the voltage at which it is supplied and 1 the average current which is applied to it for one second; and a control circuit for operating this motor at low and high speed by applying driving pulses of low frequency and high frequency respectively, the actual values of the resistance and of the self inductance of the motor coil are lower at optimum values and the driving pulses of low frequency are such that this motor has a value efficiency substantially equal to the maximum value when it operates at low speed.

Preferably, the low frequency driving pulses are pulses which have the optimum duration and the determined voltage in question and which are formed by trains of elementary pulses whose cyclic ratio allows the efficiency n of the motor to have substantially the value maximum when it operates at low speed.

Naturally, the simplest and most economical way of implementing the invention is to do as for the solution given in patent application CH 2 180/84, that is to say from a motor. step by step already developed and to replace the single coil or each coil of this motor with another coil which has a lower number of turns and which is produced from a wire of the same kind but with a larger diameter.

Moreover, the two solutions, that of the present invention and that of the Swiss patent application, in fact fall under the same principle since it is a question in both cases of reducing the resistance and the own inductance of the winding. 'an engine to increase its maximum operating speed by avoiding that this decrease causes a decrease in the performance of this engine when it has to turn slowly, which, in the case of an electronic watch, happens most often.

However, the invention does not make the winding operation more complicated and more expensive. On the contrary, it shortens it and makes it easier since there is a need for fewer turns and a larger wire can be used.

On the other hand, as long as one does not plan to rotate the motor at a speed greater than a certain limit beyond which it is no longer certain that the rotor will not take more steps than it should, the control circuit does not need to include more transistors power or output terminals as usual. As will appear later, it is only when it becomes necessary to provide braking means that the number of transistors increases but it takes even less than to be able to selectively connect two windings in series or in parallel.

• - la figure 1 représente schématiquement une forme possible de réalisation de l'ensemble moteur selon l'invention;
• - la figure 2 est un ensemble de trois diagrammes qui montrent la forme des impulsions motrices qui sont appliquées habituellement dans une montre à un moteur pas à pas monophasé bipolaire de type Lavet et de celles qui sont appliquées au moteur de l'ensemble moteur de la figure 1 pour le faire tourner à petite et à grande vitesse;
• - la figure 3 est un autre diagramme qui montre comment varie l'énergie électrique minimale à fournir au moteur pour qu'il fonctionne, à vide et en charge, en fonction de la durée des impulsions motrices qui lui sont appliquées;
• - la figure 4 est un schéma qui représente en détail un exemple de réalisation du circuit de contrôle et de hachage qui fait partie de l'ensemble moteur de la figure 1;
• - la figure 5 représente schématiquement une autre forme possible d'exécution de l'ensemble moteur selon l'invention qui utilise la technique de l'asservissement;
• - la figure 6 est un diagramme montrant comment le moteur de l'ensemble moteur de la figure 5 est alimenté lorsqu'il tourne à grande vitesse;
• - la figure 7 est un schéma de détail qui montre sous quelle forme peut éventuellement se présenter le circuit de contrôle et de hachage qui fait partie de l'ensemble moteur de la figure 5;
• - la figure 8 est un diagramme qui montre, avec deux exemples, comment varient la vitesse maximale de fonctionnement à vide et la consommation du moteur dans le cas d'un ensemble moteur comme celui de la figure 5, lorsqu'une résistance est branchée en série avec la bobine de ce moteur, en fonction de la valeur de cette résistance; et
• - la figure 9 montre en détail sous quelle forme se présente le circuit d'alimentation du moteur dans un ensemble moteur semblable à celui de la figure 5, lorsqu'il est effectivement prévu de brancher une résistance en série avec la bobine pour contrôler la vitesse de ce moteur.

The invention will be better understood on reading the following description of several embodiments, description made with reference to the appended drawing in which:

• - Figure 1 schematically shows a possible embodiment of the engine assembly according to the invention;
• FIG. 2 is a set of three diagrams which show the shape of the driving pulses which are usually applied in a watch to a bipolar single-phase stepping motor of the Lavet type and of those which are applied to the motor of the motor assembly of the Figure 1 to rotate it at low and high speed;
• - Figure 3 is another diagram which shows how the minimum electrical energy to be supplied to the motor varies so that it operates, unladen and loaded, as a function of the duration of the driving pulses applied to it;
• - Figure 4 is a diagram which shows in detail an embodiment of the control and chopping circuit which is part of the motor assembly of Figure 1;
• - Figure 5 shows schematically another possible embodiment of the engine assembly according to the invention which uses the technique of servo-control;
• - Figure 6 is a diagram showing how the engine of the engine assembly of Figure 5 is powered when it rotates at high speed;
• - Figure 7 is a detailed diagram which shows in what form may possibly appear the control and chopping circuit which is part of the motor assembly of Figure 5;
• - Figure 8 is a diagram which shows, with two examples, how the maximum speed of no-load operation varies and the consumption of the motor in the case of a motor assembly like that of FIG. 5, when a resistor is connected in series with the coil of this motor, as a function of the value of this resistance; and
• - Figure 9 shows in detail in what form the motor supply circuit is in a motor assembly similar to that of Figure 5, when it is actually planned to connect a resistor in series with the coil to control the speed of this engine.

It is already apparent from the above that the invention can be applied to all kinds of analog watches, from the simplest to the most sophisticated.

Furthermore, as we will be able to see later, it is often possible to start not only from an existing engine but also from an already known electronic watch circuit and to make some modifications to this circuit to achieve to an engine assembly according to the invention.

It is therefore best to illustrate the latter with simple examples which make it possible to highlight these modifications which, in principle, are always the same and which always remain within the reach of those skilled in the art.

For this reason, the motor assembly which is represented diagrammatically in FIG. 1 and the elements of which could belong to a watch without a seconds hand, consists only of a bipolar single-phase stepping motor 1, of Lavet type, and d '' a control circuit 2 provided to rotate this motor in one direction, that indicated by the arrow F, and at two different speeds, a low speed when the watch is to operate normally and a high speed to allow setting at the fast hour.

The motor 1 conventionally comprises a stator 3 which is shown in the figure as being in one piece but which could just as easily be produced in two parts, a rotor 4 comprising a cylindrical permanent magnet with diametrical magnetization, surrounded by the stator, and a coil 5 which surrounds a core 6 magnetically coupled to the stator 3.

With the exception of the coil 5, the elements of the motor 1 are supposed to have exactly the same geometric and physical characteristics as those of an engine of the same kind, currently used for the manufacture of analog watches and therefore do not need to be described in detail here. It suffices to specify that these characteristics are those for which it is generally used to supply the motor with driving pulses such as those which are represented by the diagram a in FIG. 2, that is to say pulses continuous, of alternating polarity and of a duration of 7.8 ms, this when the watch is intended to be equipped with a source of electrical energy supplying a voltage of 1.55 V, for example a battery money, and when it has no means to adapt the duration of the driving pulses to the momentary load of the motor.

This duration of 7.8 ms corresponds to an optimal or quasi-optimal efficiency of the motor when it truly operates in step, in its normal direction of rotation, taking into account the maximum torque that it must provide.

dans laquelle T est le couple utile maximal que doit être capable de fournir le moteur, n le nombre de pas par tour effectué par son rotor, V la tension à laquelle il est alimenté et 1 le courant moyen qui traverse son bobinage pendant une seconde, à vide, c'est-à-dire lorsque son rotor n'est soumis à aucun couple résistant dû à des éléments extérieurs.The yield in question here is that which is defined by the expression

in which T is the maximum useful torque that the motor must be able to supply, n the number of steps per revolution made by its rotor, V the voltage at which it is supplied and 1 the average current flowing through its winding for one second, unladen, that is to say when its rotor is not subjected to any resistive torque due to external elements.

It should be noted that this expression of the yield which is commonly used in watchmaking applies as well to motors with several coils as to motors which have only one. On the other hand, it is not linked to the shape of the driving pulses which serve to control the motor.

In the case of a bipolar single-phase motor, when it is known at what voltage this motor will be supplied and what maximum useful torque it will have to provide, it is easy to determine the pulse duration which corresponds to a maximum efficiency by varying this duration and by measuring the minimum energy E to be supplied to the motor which is equal to the product of the lower operating voltage by the consumption which is equal to the product of the average current I _m by the duration _T of the pulses.

As shown by curve C ₁ of the diagram in FIG. 3, this minimum electrical energy E _m begins to decrease rapidly when the pulse duration _T increases, then passes through a minimum and then increases relatively slowly. The duration _TD which corresponds to the minimum is precisely that for which the efficiency of the motor is maximum.

Naturally, this duration _TO, which will later be described as "normal", varies in fairly large proportions depending on the way in which the motor is produced, that is to say according to the shape and dimensions of the different elements, the materials which constitute them, etc. For example, for many engines of the Lavet type which are currently used for the manufacture of watches the duration τ ₀ is 7.8 ms but there are also some for which it is only 3.9 ms approximately.

Furthermore, it can be seen, still in FIG. 3, that when the motor is subjected to a certain load one obtains for the minimum operating energy another curve, C ₂ , substantially of the same shape as the curve CI and whose the minimum takes place for the same pulse duration τ ₀ which continues to correspond to a maximum efficiency of the motor.

Of course, with regard to stepper motors with several coils, determining the optimal duration of the driving pulses is generally more difficult since often it is not enough to apply a pulse to one of the coils to make it pass a step to the rotor. On the contrary, the different coils must be supplied simultaneously or consecutively and with driving pulses which are not necessarily simple pulses. At least some of these motor pulses can indeed be formed from two or more elementary pulses, of different polarity. On the other hand, some motors can operate by being controlled in several distinct ways which do not always lead to the same performance.

Despite this, it is still possible, if only by successive approaches, to find for the driving impulses which must be applied to each coil of a given motor a duration which makes it possible to obtain the best efficiency and this for each of the ways of controlling this motor if there are several.

To return to the motor assembly shown in FIG. 1, the originality of the motor 1 is that its coil 5 is produced so as to have a resistance and an inductance proper that are clearly lower than those which would allow the motor to effectively have a maximum efficiency if powered by continuous driving pulses of 1.55 V and 7.8 ms.

This reduction in resistance and self inductance compared to the values they should normally have can be, for example, by a factor of 4. This can be easily obtained by making the coil 5 with half the turns it should include it and using a wire whose diameter is / 2 times greater than it should be. In this case the motor coil is equivalent to the two identical windings of patent application CH 2 180/84 when these are connected in parallel.

The control circuit 2 comprises an oscillator 7, composed of a quartz resonator and its maintenance circuit, which delivers a standard frequency signal, of 32,768 Hz for example, to a frequency divider circuit 8 which provides at its output a signal whose frequency is the one at which the engine should run when the watch is operating normally. This frequency can for example be 1/30 Hz so that the minute hand advances at the rate of two jumps per minute.

The low frequency signal delivered by the divider 8 is transmitted, with other periodic signals of different frequencies taken from the outputs of intermediate stages of this same divider, to a waveform converter circuit 9 which is responsible for producing continuously , on the one hand, continuous pulses of 7.8 ms whose frequency is the same as that of the output signal of the divider and, on the other hand, pulses also continuous but shorter than the previous ones and dcnt the frequency is otherwise equal at least very close to the maximum frequency at which the motor can rotate. In the particular case previously considered where the resistance and inductance of the motor coil are four times lower than they should be, this frequency of short pulses can be for example 128 Hz and their duration by 3.9 ms.

The pulses of the two types generated by the converter circuit 9 are transmitted separately to a control and chopping circuit 10 which also receives the output signal (s) respectively from one or more intermediate stages of the frequency divider circuit 8 and a signal binary S coming from a contact actuatable by a manual control member such as a push button or a time-setting rod, which is provided with the watch and which is not shown in the figure. This circuit 10 is also connected to a circuit 11 for supplying the motor, conventionally constituted by a bridge of four power transistors in the diagonal of which the coil 5 is connected.

It will be assumed that when the control signal S is at logic level "0" the watch is operating normally and that, when it is at logic level "1", the watch is in rapid correction mode.

In the second case, the simplest, the control and chopping circuit 10 is content to transmit as such the short pulses which come from the waveform converter circuit 9 to the supply circuit 11, so that the coil re - then receives continuous driving pulses of high frequency and alternating polarity like those which are represented by diagram b of FIG. 2.

On the other hand, in the first, circuit 10 does not only transmit the 7.8 ms pulses which are applied to it on circuit 11. It chops them at a determined rate, using the signal or signals that it receives from the circuit frequency divider, so that the driving pulses applied to the coil are no longer continuous but formed by trains of elementary pulses like those which are represented by the diagram c of figure 2, the elementary pulses of each train having a polarity opposite to that of the pulses that make up the previous train. Note that diagram c shows only five elementary pulses for each motor pulse when there are in fact a very large number of them.

The rate at which circuit 10 chops the 7.8 ms pulses from converter circuit 9 or, which comes to the same thing, the duty cycle of the trains of elementary pulses applied to the coil is that which allows the motor 1 to have at low speed if not exactly at least very substantially the same efficiency as if the resistance and the inductance proper of its coil had their usual values and if the latter was supplied by normal driving pulses, i.e. 7.8 ms continuous pulses.

Naturally, the more the resistance and the inductance of the coil move away from the values which they would normally have, the more the duty cycle of the trains of elementary pulses which compose the driving impulses, i.e. the ratio between the duration of these elementary impulses and their period, must be small.

In the example envisaged above where the resistance and the inductance are each reduced to a quarter of their usual value, this duty cycle is equal to 0.5.

FIG. 4 shows how the control and chopping circuit 10 can be produced in this particular case.

This figure also shows the coil 5 and the supply circuit 11 formed as already indicated by a bridge of four transistors. In each branch of this bridge is a p-type transistor 12, respectively 14, in series with an n-type transistor 13, respectively 15. The sources of the p-type transistors 12, 14 are connected to the positive terminal + V of the source of the watch supply voltage and those of the n type transistors 13, 15 at its negative terminal, the coil 5 being connected between the common point of the drains of the transistors 12 and 13 and that of the drains of the transistors 14 and 15 .

To return to circuit 10, this comprises two AND gates 16 and 17 with two inputs to which the pulses of 7.8 ms and short pulses of 3.9 ms are applied, for example, produced by the converter circuit waveform 9 (see Figure 1). Gate 17 also receives the control signal S and gate 16 the inverted signal S.

The outputs of these two AND gates are connected to two inputs of an OR gate 18 whose output is connected, firstly, to the input _e of a T type flip-flop which changes state whenever its input goes from logic level "0" to level "1" and, on the other hand, to the first inputs of two AND gates 20 and 21 which have them three, the second inputs of these gates 20 and 21 being connected respectively to the outputs Q and Q of the flip-flop 19.

As the 7.8 ms pulses must be chopped to 50%, the circuit needs to perform this chopping only a periodic signal supplied by the frequency divider circuit 8. This signal whose frequency can be for example of 1024 Hz is applied to an input of an OR gate 22 which receives the signal S on another input and the output of which is connected to the third inputs of AND gates 20 and 21, the output of the first, 20, of these gates being connected to the grids of the transistors 12 and 13 of the supply circuit 11 and that of the second, 21, to the grids of the transistors 14 and 15.

When the watch is operating normally, the signal S being at logic level "0", the AND gate 16 is open to the 7.8 ms pulses which it receives and the OR gate 22 transmits the periodic signal which is applied to it at the third inputs AND gates 20 and 21. However, the short pulses are blocked by AND gate 17.

Outside the periods of application of the 7.8 ms pulses to the circuit, the output of the OR gate 18 being at level "0", the AND gates 20 and 21 remain blocked as well as the two transistors of type n 13 and 15 of the circuit 11. On the other hand, the p-type transistors 12 and 14 are conductive and short-circuit the coil 5 of the motor. As regards the flip-flop 19, its output Q can then be either at level "0" or at level "1".

Assuming that we are in the first case, that is to say that the output Q is at "0", when a pulse then appears at the input of the AND gate 16, this output Q of the rocker passes, at the same time as the entry T of the latter, at level "1". As a result, the AND gate 20 becomes conducting for the periodic signal which it receives on its third input. On the other hand, the AND gate 21 remains blocked because the logic states of its first and second inputs have simply been inverted.

For 7.8 ms, each time the third input of the AND gate 20 and, consequently, its output pass to level "1", the transistor 12 is blocked while the transistor 13 becomes conductive and a current flows in the coil , in the direction of arrow F '. At on the contrary, between two pulses of the periodic signal, the transistor 13 turns off again while the transistor 12 becomes conductive again and the coil 5 is short-circuited.

After 7.8 ms, the input T of the flip-flop returns to "0" and the periodic signal is no longer transmitted by gate 20.

When a new pulse appears at the input of gate 16, the Q output of flip-flop 19 goes to level "0" while its Q output goes to level "1" and it is this time gate AND 21 which transmits the periodic signal from the OR gate 22. This signal makes it possible to control the transistors 14 and 15 of the supply circuit so that the coil 5 is alternately traversed by a current in the opposite direction to that of the arrow F ' and short-circuited.

After this new 7.8 ms pulse, the circuit finds itself in the same situation as at the beginning.

If at a given time the signal S passes to level "1", the AND gate 16 blocks the 7.8 ms pulses which it receives while the AND gate 17 lets through the short pulses of high frequency which are applied to it. These pulses are transmitted alternately by AND gates 20 and 21, respectively to transistors 12 and 13 and to transistors 14 and 15 which control the passage of current in the coil sometimes in the direction of arrow F ', sometimes in the opposite direction. Furthermore, since the output of the OR gate 22 remains at level "1" as long as the signal S is also there, the driving pulses applied to the coil are continuous.

As long as the signal S changes logic level outside the periods when the motor receives driving pulses, the latter always have the correct polarity to rotate the rotor because, given the way circuit 10 is produced, a driving pulse is always of opposite polarity to that of the previous one.

However, if the change occurs while gate 16 or gate 17 is transmitting a pulse from the converter circuit, early enough that the rotor has not had time to complete a step or acquire a sufficient energy for this, it goes back and the first motor pulse following this change no longer has the correct polarity for rotate it. It is necessary to wait for the second pulse so that the rotor takes another step.

Apparently, this possibility for the rotor to miss a step is especially troublesome for the transition from high to low speed because it means that, immediately after being reset, the watch can delay by, for example, half a -minute if the pulse frequency of 7.8 ms is 1/30 Hz.

In fact, this is not really a drawback because it is clear that the very simple motor assembly which has been described could hardly be used as it is in a watch. The frequency of short pulses is indeed too great for it to be possible to send only a small, determined number to the motor or to interrupt a correction at the exact moment when the minute hand reaches the desired position. . This motor assembly which only allows significant corrections to be made in an approximate manner should therefore be supplemented by means for producing correction pulses whose low frequency could be fixed or variable as a function of the speed of actuation of the same control member. command that one that allows for quick and coarse corrections or another and whose number could be perfectly controlled. In addition, it would be necessary to modify the control and chopping circuit 10 which would also receive these pulses so that it can transmit them at the desired time to the power supply circuit 11 of the engine.

As any correction started at high speed should be completed using these additional means, the fact that the rotor misses a step when it starts to rotate rapidly and / or when the motor receives the first low frequency correction pulse would be unimportant.

However, if we nevertheless wanted to avoid that motor pulses could be applied unnecessarily to the motor, we could easily add to the modified control and hash circuit or include in an added correction circuit means to refuse a level change of the control signal S during the duration of these pulses or to delay their effects, means which would become necessary if the operation at high speed of the motor were used not only for simple updates the time, but also for time zone changes or to indicate by the hands something other than the current time, for example a memorized alarm time.

It was previously stated that it was almost always possible to start from an already known control circuit to produce an engine assembly according to the invention. It is easy to prove it.

For example, there is in US Pat. No. 3,901,022 a circuit which makes it possible to apply to the single-phase and unidirectional motor of a watch either low frequency driving pulses of 7.8 ms when the latter is operating normally, or pulses of the same duration and of 32 Hz when setting the time. Overall, the circuit which has been described is distinguished from it only by the fact that the low frequency pulses are chopped and that the correction pulses are shorter and of significantly higher frequency. Since the formation of pulses of determined duration and frequency poses no problem and the technique of chopping the pulses for controlling a stepping motor is now well known, in particular in the field of watchmaking , it is very easy for a person skilled in the art to modify the circuit of the American patent to result in another circuit which may not be exactly that of FIG. 1 but which will make it possible to control the motor 1 in the same way.

Another example: there are already many watches with which it is possible to advance the minute hand and, as a result, the hour hand at high speed or slowly to set the time. So, rather than completing the circuit which has been described by means which make it possible to make slow but precise corrections, we could very well replace circuit 2 by that of one of these known watches by making the same modifications. than those which have been indicated previously with regard to US Pat. No. 3,901,022. We would end up with an equivalent circuit.

Similarly, one could start from already marketed watch circuits which use the so-called rocking technique, described in US Pat. No. 4,112,671, to rotate a Lavet type motor in both directions if one wanted to associate with motor 1 a control circuit which allows corrections to be made in both directions, advance and delay.

Of course, many other examples could be cited to show that what has been said is true, but it is not necessary.

However, it should nevertheless be specified that, in certain known watches and in particular those which are provided with an automatic adaptation system of the energy of the driving pulses to the load of the motor, the chopping of the pulses already exists. There would therefore be no need to introduce it to result in an engine assembly in accordance with the invention, it would suffice to change the duty cycle (s) of the elementary pulses which constitute the driving pulses.

• Nombre de spires de la bobine : N = 26600
• Résistance de la bobine : R = 7400 Ω
• Perméance du circuit : A = L/N2 = 40 nH (L : inductance propre de la bobine)
• Valeur maximale du coefficient de couplage : γ₀ = 0,25 µNm/At
• Moment d'inertie du rotor : J = 3,5 · 10-12 _kg_m ²
• Couple de frottement sec : C_s = 3,5 · 10^-8 Nm
• Coefficient de frottement visqueux : f = 10^-10 Nm/rd/_s

For a classic Lavet type engine with the following mechanical and electrical characteristics:

• Number of coil turns: N = 26600
• Coil resistance: R = 7400 Ω
• Circuit permeance: A = L / N2 = 40 nH (L: self inductance of the coil)
• Maximum value of the coupling coefficient: γ ₀ = 0.25 µNm / At
• Moment of inertia of the rotor: J = 3.510-12 _k g _m ²
• Dry friction torque: C _s = 3.510 ^-8 Nm
• Viscous friction coefficient: f = 10 ^-10 Nm / rd / _s

Maximum value of the positioning torque: T ₂ = 0.35 µNm we usually obtain, for a supply voltage of 1.55 V and continuous driving pulses of 7.8 ms, a maximum useful torque T _u of 0.22 µNm and an average current I _m of approximately 105 µA, therefore an optimized efficiency substantially equal to 42.4%. In this case, the maximum frequency at which the motor can be controlled is, as already indicated, of the order of 50 or 60 Hz.

In the particular case envisaged above where the coil of this motor is replaced by another which has a resistance and a self-inductance four times lower, one obtains exactly the same results by supplying the motor with pulses chopped to 50%. On the other hand, with continuous pulses of 3.9 ms, the useful torque goes to 0.30 µNm, the average current to 518 µA and the maximum motor operating frequency is doubled.

As in general and in particular in the case of a watch there is no reason to have a greater useful torque at high speed than at low speed, it is still possible to reduce the duration of the continuous pulses up to that this torque takes about the same value as for the chopped 7.8 ms pulses, which happens for a duration of about 3 ms. This leads to an additional increase of about 30% in the maximum frequency and increases its value to 130 or 150 Hz.

If we refer to the numerical example given in patent application CH 2 180/84, it can be seen that the values which have just been indicated are the same as those obtained when the coil is replaced 26,600 turns and 7,400 n of the motor by two identical windings of 13,300 turns and 3,700 n each and when these windings are connected in series or in parallel to rotate the motor respectively at low and high speed.

So here again, we approach the theoretical frequency of synchronism of the conventional motor from which we started, which is about 200 Hz, and nothing prevents further reduction of the resistance and inductance of the coil for get an even higher maximum frequency.

On the other hand, it is entirely possible to combine the solution of the invention with that which is described in patent application EP 0 103 542 and which, as already said, also makes it possible to achieve a maximum frequency very close to the synchronism frequency.

où N désigne ici encore le nombre de spires de la bobine du moteur, γ₀ la valeur maximale du coefficient de couplage et U_i la tension induite dans la bobine par le mouvement du rotor, il suffit de diviser par deux le nombre de spires pour la doubler.Since the latter is given by the formula:

where N again designates the number of turns of the motor coil, γ ₀ the maximum value of the coupling coefficient and U _i the voltage induced in the coil by the movement of the rotor, it is enough to divide by two the number of turns for double it.

The simultaneous use of the two solutions therefore makes it possible to reach a maximum frequency of the order of 400 Hz and even more.

Unfortunately, this is still insufficient for a watch with a second hand because a correction of six hours would then take more than 50 seconds.

In order to obtain acceptable correction times, it is necessary to use both the invention and the well-known technique of servo-control.

FIG. 5 schematically shows one of the simplest forms in which the engine assembly according to the invention can then appear. This possible embodiment corresponds to that which has been described previously, that is to say that the motor assembly shown in FIG. 5 simply consists of a Lavet type motor 1 'identical to that of the motor assembly of Figure 1 and on which it is therefore not necessary to return and a control circuit 2 'provided to rotate this motor in only one direction, at low and high speed.

Like circuit 2 of the motor assembly of FIG. 1, the control circuit 2 'comprises, connected one after the other, an oscillator 7', a frequency divider circuit 8 ', a converter circuit waveform 9 ', a control and chopping circuit 10' which receives a binary control signal S 'and a power supply circuit 11' of the motor.

The oscillator 7 'and the power supply circuit 11' are identical to those of the circuit 2. On the other hand, the divider circuit 8 'can comprise fewer stages than the divider circuit 8 and provide an output signal of higher frequency and equal to 1 Hz for example. On the other hand, the converter circuit 9 'is simpler than the circuit 9 because it no longer has to supply the control and chopping circuit 10' with 7.8 ms pulses at the frequency of the output signal of the divider circuit 8 ', for the operation of the motor 1' at low speed.

In addition to the elements 7 'to 11', the control circuit 2 'comprises a circuit 30 for permanently detecting, when the rotor 4' is rotating, a parameter representative of the instantaneous position thereof and for applying a short pulse to the control and chopping circuit 10 'each time this rotor passes through a position angular well determined or by the opposite position. These positions can be those of static equilibrium but it is preferable that they are located before, that is to say that each of them is between a position of static equilibrium and that of the two positions of equilibrium with current which is angularly the closest or is confused with the latter. As for the detected parameter, it can be the intensity of the current in the coil 5 ', the voltage induced in the latter by the movement of the rotor, the variation of magnetic flux in the stator 3', etc.

The circuit 30 can take one of the many forms which have been disclosed in the context of systems for adapting the energy of the driving pulses applied to a motor under its load. For example, for the detection of the induced voltage, the circuit which is described in US Pat. No. 4,446,413 can be used and for the flux variation that which appears in US Pat. No. 4,430,007.

Note that this detection circuit 30 has been shown in the figure as being connected to the terminals of the coil 5 ′, but it could also be connected to one or more points of the supply circuit 11 ′. It all depends in fact on the parameter chosen and the way in which the circuit is made.

When the control signal S 'is at logic level "0", the motor rotates step by step at low speed and the control circuit 2' then operates in exactly the same way as circuit 2 in FIG. 1 except that the pulses chopped driving motors are applied to the motor at a slightly higher frequency and the control and chopping circuit no longer blocks short pulses of high frequency originating from the waveform converter circuit but pulses of the same frequency as the driving pulses , produced by the detection circuit 30.

Naturally, it would also be possible not to operate this detection circuit as long as the engine is running at low speed, but it would then be necessary to provide, between it and the coil 5 ′ or the supply circuit 11 ′, a controlled switching system by the signal S '.

As we will not see below, the fact that the signal S 'can pass from logic level "0" to level "1" while a pulse driving is applied to the engine would have much more unfortunate consequences than in the case of the engine assembly of Figure 1. It will therefore be assumed that the control circuit 2 'includes means not shown which avoid this.

Thus, when the signal S 'passes to level "1", the control and chopping circuit 10' controls the supply circuit 11 'so that the coil 5' is traversed by a current of opposite direction with respect to that which passed through it for the duration of the last driving impulse and the rotor then begins to rotate. Furthermore, from this moment, the circuit 10 'blocks the 7.8 ms pulses which it receives from the converter circuit 9'.

At the moment when the rotor reaches one of the chosen reference positions, the detection circuit 30 sends a pulse to the circuit 10 'which acts on the supply circuit to reverse the direction of the current in the coil and continue to do turn the rotor.

When the latter passes through the other reference position, the detection circuit applies a new pulse to the control and chopping circuit which again changes the direction of the current in the coil and so on as long as the signal S 'is at level "1".

In other words, the coil 5 ′ is then supplied with driving pulses of voltage V, of alternating polarity, which follow one another without interruption, the start of one of these pulses being confused with the end of the previous one.

This is illustrated by the diagram in FIG. 6 which represents, in algebraic value, the voltage applied to the coil as a function of the angle a of rotation of the rotor. This parameter was chosen preferably over time because, in this case, the duration of the motor pulses is not constant. Indeed, from the moment the rotor begins to rotate, its speed increases, which means that this duration decreases and it only becomes constant from the moment when the rotor speed reaches a maximum value which it then keeps, this of course provided that the signal S 'remains at level "1" long enough.

When the signal S 'returns to level "0", the control and chopping circuit controls the supply circuit so that the coil 5 'is short-circuited as it was just before the level of the signal S' goes to "1" and, from this moment, it starts again to transmit the low frequency pulses of 7.8 ms by them chopping and blocking pulses from the detection circuit.

Shorting the motor coil effectively brakes the rotor, but considering the speed at which it can rotate, there is little chance that this will be enough to prevent it from making one or more half - additional turns after the signal S 'has returned to "0". Furthermore, if a low frequency drive pulse is applied to the coil while the rotor continues to rotate, this can, depending on its polarity and the moment it appears, help to brake the rotor or, on the contrary, delay its immobilization. . In addition, it is not at all sure that the driving impulse which comes after the rotor has stopped has the right polarity to make the motor take a step. The engine assembly which has just been described overall can therefore only be suitable for very simple watches and provided that it is completed, like that of FIG. 1, by means allowing slow corrections to be made.

A solution will then be indicated to stop the rotor at the desired time.

FIG. 7 shows in detail a way of making the control and chopping circuit 10 ′ in the particular case where the driving pulses applied to the motor are chopped with a duty ratio equal to 0.5, as well as the supply circuit 11 'with its four power transistors 12', 13 ', 14' and 15 'to control the passage of current through the coil 5'.

In this particular embodiment, the circuit 10 ′ includes all the elements of the circuit of FIG. 4 which are designated by the same references with, in addition, the signal ""'and which are connected together in the same way to this near that the output of the OR gate 18 'is not connected to inputs of the AND gates 20' and 21 'directly but via an OR gate 24 which also receives the signal S'. Furthermore, this OR gate 18 ′ has not only two inputs but three, the additional input being connected to the output Q of a monostable circuit 23 whose input TR also receives the signal S '. Finally, for the AND gate 17 ′, the short pulses of the waveform converter circuit are replaced by the pulses coming from the detection circuit 30. Note that if it had been planned to operate the latter only when the engine must run at high speed, this door 17 'would not have happened.

For the operation of the circuit, when the signal S 'is at level "0", there is nothing changed with respect to the circuit of FIG. 4 except that the AND gate 17' now blocks the pulses produced by the circuit of detection.

When the signal S 'goes to level "1", the AND gate 16' which receives the 7.8 ms pulses from the converter circuit is blocked while the monostable circuit 23 produces a short pulse which is transmitted by the OR gate 18 'at the entrance to the weighbridge 19'. Assuming that, just before, the outputs Q and Q were respectively at levels "0" and "1", these then pass to levels "1" and "0". From this moment, all the inputs of the AND gate 20 'are at level "1" and its output is too. On the other hand, the AND gate 21 'connected to the output Q of the flip-flop is blocked. Consequently, as soon as the signal S 'goes to level "1", a current flows in the coil 5' in the direction of the arrow F '.

When the first pulse from the detection circuit is transmitted by the gates 17 'and 18' to the input T of the flip-flop 19 ', the levels of the outputs Q and Q of this latter are reversed as well as those of the outputs AND gates 20 'and 21' and the direction of the current in the coil changes.

At the second pulse supplied by the detection circuit, it is again the AND gate 20 ′ which has its output at level "1", which causes the current to change direction again and so on.

When the signal S 'returns to level "0", the monostable circuit 23 emits no pulse and everything then takes place as for circuit 10 in FIG. 4.

We will now see what could have happened with practically all the usable detection circuits, if the possibility of passing to level "1" of the signal S 'during the application of a 7.8 ms pulse to the circuit control and hash had not been removed.

Let us assume, still considering the circuit 10 'of FIG. 7, on the one hand, that the signal S' actually passes to level "1" while the gate 16 'is in the process of transmitting a pulse and, on the other hand , that this gate 16 'and the monostable circuit 23 both react instantly to the rising edge of the signal S' applied to their input and that the flip-flop 19 'does not perceive the inversion of the levels of the outputs of these elements. In this eventuality, the current which has started to pass through the coil 5 'in one direction continues to do so after the signal level S' has gone to "1" and if the rotor has already exceeded the reference position by the time this change occurs, it is placed in its equilibrium position with current and it remains there as long as the level of the signal S 'is not reduced to "0". Naturally, if the rotor has not yet passed through the reference position there is no problem.

The same thing happens if the AND gate 16 'reacts less quickly than the monostable circuit.

On the other hand, if the output of gate 16 'goes to "0" before that of the monostable circuit goes to "1", the levels of outputs of flip-flop 19' reverse as do the direction of the current in the coil . If, at the moment when this occurs, the rotor has not rotated more than 135 ° or acquired sufficient kinetic energy to do so, it goes back to stabilize in its other equilibrium position with current and nothing does not change as long as the signal level S 'remains at "1".

Obviously, the user of the watch would very quickly realize that the hands are not advancing and would repeat its operation. Since the chances of the same thing happening a second time are minimal, the engine would start to run properly. Nevertheless, it is better to remove this drawback, especially since it can be done in a very simple way.

For a motor whose different parameters A, x ₀ , J, C _{S '} f and T ₂ have the values which were indicated previously and whose coil has a resistance and an inductance eigen four times weaker than those which it should normally having the fact of rotating the rotor continuously using the servo-control technique makes it possible to obtain a maximum no-load speed, that is to say for a useful torque T zero, of approximately 620 steps by second (in fact it would be more correct to say U-turns rather than not). This is still too little for a second hand watch.

on obtient, avec une tension non plus de 1,55 V mais de 6 V, une vitesse maximale à vide de 2075 pas par seconde. Pour un couple utile d'environ 0,20 uNm, cette vitesse est sensiblement égale à 1800 pas par seconde, ce qui est suffisant étant donné qu'une correction de six heures ne prend plus alors que douze secondes dont le même temps que pour une montre sans aiguille de secondes munie d'un moteur classique dont la vitesse maximale est, comme on l'a déjà indiqué, d'environ 60 pas par seconde et dans laquelle l'aiguille des minutes se déplace normalement à raison de deux sauts par minute.On the other hand, with an engine whose characteristics are as follows:

we obtain, with a voltage not more than 1.55 V but 6 V, a maximum no-load speed of 2075 steps per second. For a useful torque of approximately 0.20 uNm, this speed is approximately equal to 1800 steps per second, which is sufficient given that a correction of six hours only takes twelve seconds, the same time as for one watch without a second hand fitted with a conventional motor, the maximum speed of which, as already indicated, is around 60 steps per second and in which the minute hand moves normally at the rate of two jumps per minute .

On the other hand, it can be seen that with the servo the average consumption per step of the motor is slightly lower than when the latter really operates step by step.

It is clear that a rotor rotating at such speeds cannot stop suddenly as soon as the engine stops being supplied. So to be able to stop this rotor at the desired time, it must first be braked.

An effective solution for achieving this braking consists in connecting a resistor of adequate value in series with the coil for a certain time before interrupting the supply of the motor.

Naturally, braking is all the more important the higher the resistance value and this value must be chosen according to the parameters of the coil, the speed at which it is planned to rotate the rotor and the time during which we plans to connect this resistor. It will generally be of the same order of magnitude as that of the coil and it may be, for example, substantially equal to the difference between the resistance which the coil should normally have in order to obtain optimum efficiency in step-by-step operation at low speed and the one she really has.

Curves I and II of the diagram in FIG. 8 represent the variation of the maximum no-load speed v of the rotor as a function of the value R ′ of the resistor which is connected in series with the coil for the two motors which have been chosen as examples.

In the first case (curve I), this speed which is 624 steps per second in the absence of resistance increases to 305 steps per second for a resistance of 1400 n, that is to say it is little almost halved. This is already enough for the rotor to stop without taking any additional steps after the power supply to the motor has been interrupted. For a resistance of 5000 n, which corresponds substantially to three times the resistance of the coil, the speed is no more than 118 steps per second.

For the second motor (curve II), the speed decreases from 2075 to 665 steps per second when the resistance value goes from 0 to 5000 n. This decrease is significant but not entirely sufficient. On the other hand, a resistance of around 7000 Ω could be suitable.

This same diagram also shows how the average consumption per step C expressed in nano-amps varies in both cases. When R 'increases from 0 to 5000 Ω, this consumption drops roughly from 800 to 1765 nA for the first motor (see curve III) and from 535 to 1000 nA for the second (see curve IV). It is therefore roughly doubled. However, as in a watch the engine is only supposed to run at high speed only very rarely and since, in addition, the braking period would often represent only a very small part of the time during which the engine would operate thus, this does not constitute not a drawback.

FIG. 9 shows how the supply circuit of a motor such as that of FIG. 1 or of FIG. 5 must be produced when this solution of the solution is actually used. additional resistance to brake the rotor. We can also consider that the resistance is part of this circuit.

We see that in the two conventional branches each formed by a p-type transistor 12 ", respectively 14", and an n-type transistor 13 ", respectively 15", and between which is connected the coil 5 "of the motor , a third branch is added in parallel with the two others and also consists of a p-type transistor 31 and an n-type transistor 32 in series, the additional resistor 33 of value R 'being connected between the junction point of the drains of these two additional transistors 31 and 32 and that of the drains of the transistors 14 "and 15".

When this resistor must be connected in series with the coil, the transistors 14 "and 15" remain blocked and it is then the transistors 31 and 32 which cooperate with the transistors 12 "and 13" to control the passage of current in one direction or in the other, both in the coil and in the resistor. Outside these periods, the transistors 31 and 32 remain permanently blocked.

Due to the presence of six transistors instead of four in the supply circuit and also because the control circuit is practically obliged to know the number of U-turns that the rotor must perform at high speed in order to be able to start brake it after a certain time, the control and chopping circuit can no longer have a shape as simple as that of FIG. 7. As it can be produced in many ways which are not directly linked to the invention and which are all largely within the reach of those skilled in the art, this circuit will not be described here. One can imagine for example that it is designed to control the supply circuit of FIG. 9, so that the rotor makes 1800 revolutions, that is to say that the hands of the watch advance exactly one hour each time it receives a control pulse produced when a specific action is exerted such as pressing, pulling or rapid rotation, on a manual control member.

Furthermore, it is quite possible to use the additional resistance not only to slow down the motor at the end of a period of operation at high frequency, but also to regulate the rotor speed throughout this period. In this case, the motor control circuit must also include a device for detecting the momentary speed of the rotor and the cores to connect the resistor in series with the coil for a determined time when this speed exceeds a certain threshold or better for connecting the resistance when the speed becomes greater than a first value and disconnect it as soon as the speed falls below a second value less than the first. Of course, this regulation system must be rendered inoperative at the time of the final braking phase.

It is also possible to provide several additional resistors for braking and / or regulating the speed of the rotor as well as means for selecting one or the other of these, in particular when it is a question of operating the engine not only at low speed and high speed but also at one or more intermediate speeds.

Finally, it is obvious that the invention is not limited to the various embodiments which have been described or envisaged.

For example, rather than chopping the low frequency drive pulses, we could reduce the height, that is to say power the motor with a lower voltage, or the duration or even combine these different solutions. The main thing is that these pulses are such that they make it possible to compensate as best as possible for the drop in efficiency of the motor due to the fact that the resistance of its coil or of each of its coils has values significantly lower than those they should normally have.

Furthermore, as has already been emphasized on several occasions, the invention can be extended to many kinds of engines. Apart from the various Lavet type motors, there may be mentioned as examples of possible applications the motors normally with a single coil but truly bidirectional which derive from motors of the aforementioned kind and which are the subject of patent CH 616 302 and of the application EP 0 085 648 or the two-phase motor, several forms of which are described in the patents CH 625 646 and CH 634 696.

In addition, it can be used in sectors other than watchmaking.

Claims (7)

1. Motor assembly comprising a stepping motor provided with a rotor, a stator and a coil magnetically coupled to the stator and for which there are optimum values of the resistance and the self inductance of the coil and an optimal duration of the driving pulses to be applied to this coil to control it which allow it to have a yield n of maximum value when it is supplied with a determined voltage, said yield being defined by the relation:

where T is the maximum useful torque that the motor can supply, n the number of steps per revolution made by its rotor, V the voltage at which it is supplied and I _m the average current applied to it for one second; and a control circuit for operating said motor at low and high speed by applying driving pulses of low frequency and high frequency respectively, characterized in that the actual values of the resistance and of the self inductance of said coil (5; 5 '; 5 ") are less than said optimum values and said low frequency driving pulses are such that the motor (1; 1') has a value output substantially equal to said maximum value when it operates at low speed. 2. Motor assembly according to claim 1, characterized in that the low frequency driving pulses are pulses which have said optimal duration and said determined voltage and which are formed by trains of elementary pulses whose duty cycle allows the yield n of the motor (1; 1 ') to have substantially said maximum value when said motor operates at low speed. 3. Motor assembly according to claim 1 or 2, characterized in that the high frequency driving pulses are pulses which have said determined voltage and a fixed duration, less than said optimal duration. 4. Motor assembly according to claim 1 or 2, characterized in that the real values of the resistance and of the inductance of the coil (5; 5 '; 5 ") are substantially equal to a quarter of said optimal values. 5. Motor assembly according to claim 4, characterized in that the high frequency driving pulses are pulses which have said determined voltage and a fixed duration, substantially equal to half of said optimal duration. 6. Motor assembly according to claim 1 or 2, characterized in that the driving pulses of high frequency are pulses which have said determined voltage and which follow one another continuously, the beginning of a pulse coinciding with the end of the previous one and with the instant when a parameter representative of the instantaneous position of the rotor (4; 4 ') of the motor (1; l') which is continuously detected by the control circuit (2; 2 ') when said motor operates at high speed reaches a determined value. 7. Motor assembly according to one of the preceding claims, characterized in that the control circuit (2; 2 ') is designed to be able to connect a resistor (33) in series with the coil (5; 5'; 5 " ) when the motor (1; 1 ') is operating at high speed, during a braking period of the rotor (4; 4') which precedes the end of the sending of high frequency drive pulses to said coil.

Images