DE102018119723A1 - Method for driving an electric motor and electric motor - Google Patents

Abstract

A method for controlling an electric motor in start-up mode is proposed, wherein an induced voltage and a phase current are generated in phase windings of the electric motor with a method comprising: detecting a characteristic curve based on the induced voltage and / or the phase current, and controlling the electric motor in Dependency of the recorded characteristic.

Description

Field of the Invention

The invention relates to a method for controlling an electric motor and an electric motor, in particular for an actuator. Actuators are used in various applications, such as in valve actuators of a motor vehicle or in industrial or building automation systems.

One application of an actuator is, for example, a roller blind that can be moved parallel to an opening to be closed, a fan motor or a flap actuator in automotive applications, for example in an air conditioning system. Another application relates to a flap actuator in a motor vehicle, which can be designed to close an actuator in the form of a tank flap or a loading flap for charging a battery of an electric vehicle. Such flaps are designed, for example, to be pivotable about an axis of rotation.

Another application in automotive engineering is, for example, to close an opening for drawing in outside air by means of a ventilation flap constructed with slats, the slats being controllable in order to adjust the ventilation flap between a fully closed and a fully open state. For example, this can influence the engine compartment temperature as required. This is also important for applications in electric vehicles in which the temperature has a significant influence on the performance, charging and discharging of the batteries. In order to prevent ventilation flaps from adjusting due to external forces, e.g. airflow, the actuator can be equipped with a gear with a high reduction ratio or with a self-locking gear and / or the electric motor can be supplied with a holding current when it is at a standstill. The invention is not restricted to the applications mentioned.

It is an object of the invention to provide a method for controlling an electric motor and an electric motor, in particular for an actuator, which starts up as efficiently as possible from a standstill.

This object is achieved by a method according to claim 1 and by an electric motor according to claim 25. Embodiments of the invention are specified in the dependent claims.

In one aspect, the invention relates to a method for controlling an electric motor during start-up operation, wherein an induced voltage and a phase current are generated in phase windings of the electric motor, the method comprising: detecting a characteristic curve based on the induced voltage and / or the phase current; and controlling the electric motor as a function of the detected characteristic.

In particular, it can be provided that the detection of the characteristic curve comprises the detection or determination of zero crossings of the characteristic curve. In such configurations of the invention, the electric motor can be controlled as a function of the zero crossings determined. For example, it can be provided to detect zero crossings of a voltage induced in the phase windings. In particular, the phase current of the electric motor can be regulated as a function of the temporal position of the detected zero crossings. Additionally or alternatively, it can be provided that a speed or an acceleration of the electric motor is regulated as a function of the time position of the detected zero crossings. The speed or acceleration can be regulated in particular by changing a duty cycle of a pulse width modulation of the phase current.

In some configurations of the method, it is provided that the control of the electric motor comprises determining an optimized next commutation time as a function of the detected characteristic or the zero crossings of the detected characteristic. Consequently, the electric motor can be controlled in such a way that it is commutated at the optimized next commutation times.

In preferred embodiments of the method, it is provided that the optimized next commutation point in time is determined from a time interval A between two successive zero crossings and is set such that it occurs by a time period A '= A / n after the second of the two zero crossings, wherein 1.25 ≤ n ≤ 50 applies.

In some configurations of the invention, driving the electric motor further comprises calculating at least one theoretical commutation step depending on a desired acceleration or a setpoint value of the phase current of the electric motor. The electric motor can then be controlled depending on the theoretical commutation step. In particular, it can be advantageous if the theoretical commutation time is a so-called ideal commutation time which lies in the middle between two successive zero crossings of the induced voltage. Are not all operating parameters necessary to determine the ideal Commutation point in time are necessary, the ideal commutation point in time can be made on the basis of estimated values or maximum values. This allows the theoretical commutation time to be determined in such a way that the electric motor can be started under all operating conditions. For example, to determine the ideal commutation time, a maximum load to be driven by the electric motor can be assumed.

In some configurations, the optimal next commutation time can be determined by comparing the detected zero crossings and / or the previous commutation times with the theoretical, or ideal, commutation time. Depending on the operating state, for example depending on the load driven by the electric motor or other operating parameters, the optimal next commutation time can correspond to an ideal commutation time, or earlier (pre-commutation) or later (post-commutation) than the ideal commutation time. In particular, the optimized next commutation step can be determined both as a function of the detected characteristic and as a function of the theoretical commutation time.

In some configurations, the method comprises energizing phase windings of the electric motor in such a way that a first optimized next commutation time is a pre-commutation time that lies before the theoretical commutation time, in particular before the ideal commutation time. The pre-commutation time can, for example, be 2 ° to 30 ° electrically before the ideal commutation time. In some configurations, it is advantageous if the first optimized next commutation time has a pre-commutation in the range from 15 ° to 30 °, that is to say the first optimized next commutation time is 15 ° to 30 ° before the corresponding theoretical or ideal commutation time.

The method can be used in particular to control an electric motor designed as a three-phase brushless DC motor. In such embodiments of the method, a pre-commutation time can be at least in some commutation steps 3 ° to 15 ° electrically before the calculated end of the theoretical commutation step / ideal commutation step.

It is particularly preferred if the two commutation steps which follow the first commutation step having a pre-commutation have a smaller pre-commutation than this first commutation step which has a pre-commutation. Alternatively or additionally, it can be provided that these two commutation steps are successively approximated to theoretical commutation steps, for example with ideal commutation times, the precommutation being consequently reduced.

It is preferably provided that for at least two commutation steps following a first commutation step, the optimized next commutation point in time is determined on the basis of the zero crossings of the characteristic during the respective previous commutation steps.

In some embodiments of the method it is provided that at least one optimized next commutation time is determined such that a zero crossing of the characteristic lies in the middle or approximately in the middle between two successive commutation times. The method can be implemented, for example, in such a way that the optimized next commutation steps are ideal commutation steps, or that the first optimized next commutation steps have precommutation and subsequent optimized next commutation steps are carried out as ideal commutation steps.

In some preferred embodiments of the method, the electric motor is operated with block commutation, with phase windings being supplied with a constant voltage. It is particularly preferred if the block commutation is superimposed by pulse width modulation. A pulse width modulation can be used, for example, to regulate a torque of the electric motor, a speed or a phase current to a setpoint.

In further refinements of the method, this includes the application of a holding current to the electric motor before the first commutation step is carried out. This allows the rotor of the electric motor to be aligned, so that particularly efficient and fast starting is possible. In such configurations, the first commutation step in particular can also be shorter than an ideal commutation step. At the beginning of the start-up phase, the method therefore preferably also includes calculating at least the first theoretical commutation step as a function of a desired acceleration of the electric motor; and the selection of a first commutation step, which is shorter than the calculated first theoretical commutation step, for controlling the electric motor in the start-up phase. For example, with a 60 ° commutation, in each case two of three motor phases are energized, the first commutation step has a length of less than 60 ° electrical, for example a length of 30 ° electrical. In general, the method can consequently be designed such that the length of the first commutation step, which is selected in the start-up phase, is 50% or approximately 50% of the length of the first theoretical commutation step.

In particularly advantageous refinements of the method, this relates to a start-up phase which is designed as a synchronous start-up phase. In such synchronous start-up phases, the electric motor is driven by the specification of theoretical commutation steps. The present method can now be used to optimize the synchronous start-up phase by dynamically modifying at least one theoretical commutation time by means of the at least one optimized commutation step. In some configurations, the theoretical commutation times of the synchronous startup are predetermined and stored in a table. The table can then be read out by a motor controller and the electric motor can be controlled / commutated accordingly. It can also be provided that the theoretical commutation times of the synchronous start-up are calculated during operation. This has the advantage that the theoretical commutation times are determined as a function of the last commutation point in time, and / or a current phase current, and / or a current load, and / or a setpoint value of the phase current, and / or a desired acceleration / setpoint acceleration of the electric motor can be. A particularly efficient start-up can thus be achieved by calculating the theoretical commutation steps during operation and adapting at least one theoretical commutation step by means of an optimized next commutation step.

For example, such a dynamically adapted synchronous startup can take place as follows. First of all, an optional holding current can hold and / or pre-align the stationary rotor. In a next step, a first commutation step can take place, for example depending on a desired acceleration. The acceleration can be adjusted, for example, using pulse width modulation. In the case of a pre-aligned rotor, this first commutation step can have half the length of an ideal commutation step. The electric motor can then be accelerated on the basis of commutation times of the theoretical commutation steps.

For example, at least one or more commutation times are selected in accordance with the theoretical commutation steps. As soon as a characteristic curve of the induced voltage or phase current can be reliably detected, the theoretical commutation times of the synchronous starting can be adjusted on the basis of the characteristic curve. For example, a commutation time of a theoretical commutation step can be adapted by an optimized next commutation time as soon as two zero crossings of the characteristic curve, for example the induced voltage, have been detected. The optimized next commutation point in time can be determined as a function of the characteristic curve, in particular its zero crossing. Furthermore, the adaptation can take place as a function of ideal commutation times. For example, the commutation times of the theoretical commutation steps can be determined as ideal commutation times, assuming a maximum load that can be driven by the electric motor. It can also be provided that an ideal commutation time corresponding to the actually applied load is determined taking into account the detected characteristic. The optimized next commutation time can then be determined as a function of this ideal commutation time. In addition, one or more previous commutation times can be taken into account to determine the optimized next commutation time.

Furthermore, in order to determine the optimized next commutation point in time, a weighting of a point in time calculated on the basis of the characteristic curve and / or the commutation point in time of the theoretical commutation step can be carried out. For example, it can also be specified that the optimized next commutation time may only lie by a maximum electrical angle before and / or after the commutation time of the theoretical commutation step. If this maximum angle is exceeded, the calculated value can be replaced by the maximum value or by another predetermined value. It is preferably provided that the optimized next commutation time is a maximum of 30 ° electrical, in particular a maximum of 15 ° electrical, before or after the commutation time of the theoretical commutation step to be adapted.

Furthermore, it can be provided in the method that the optimized next commutation time is at most 15 ° electrically before or after the theoretical commutation time to be adjusted.

In some developments of the method it is provided that from the synchronous start-up phase to sensorless controlled motor operation is switched after at least one of the following conditions is met: reaching a predetermined number of commutation steps; reaching a predetermined number of zero crossings; reaching a predetermined speed; and detecting a back EMF that is above a predetermined threshold.

The optimized next commutation times can also be determined depending on the previous commutation times. Consequently, a determination can also be made as a function of at least one preceding commutation point in time, the detected characteristic curve and a next theoretical commutation step.

According to the configurations described above, particularly efficient operation of the electric motor can thus be ensured. By adapting the control of the electric motor on the basis of the characteristic curve during start-up operation, this operating phase can be carried out particularly efficiently. In the case of relatively small loads, a completely sensorless controlled start can be provided in extreme cases. At least a synchronous start-up can be improved in such a way that either the theoretical commutation times are retained and, for example, regulation, for example to a setpoint value of the phase current, can be carried out as a function of the detected characteristic curve. Thus, for example, the current consumption of the electric motor can be reduced and thus the electromagnetic compatibility of the electric motor can be improved. Greater acceleration or driving of a larger load can also be made possible. In other configurations, a synchronous start-up can be designed dynamically, so that by adapting the synchronous commutation times by means of the optimized next commutation times, a very fast start-up of the electric motor and thus, for example, a very quick change to a sensorless controlled operation is possible. The switch to sensorless operation can be possible, for example, after only one to five commutations of the synchronous start-up. A combination of the possible configurations of the method can also be provided.

In a further aspect, the invention relates to an electric motor comprising phase windings and a control circuit, the control circuit being set up to control the electric motor using the method described above.

The electric motor, or the control circuit, can further comprise a comparator and / or an analog-digital converter for detecting zero crossings of a characteristic curve which is based on an induced voltage and / or a phase current of the electric motor.

Finally, an actuator with such an electric motor and with a reduction gear, which has a reduction ratio between an input and an output of the reduction gear in the range from 1: 100 to 1: 1000, is to be provided.

list of figures

• 1 zeigt eine schematische Darstellung eines Elektromotors und einer zugehörigen Steuerschaltung gemäß einem Beispiel;
• 2A bis 2D zeigen schematische Darstellungen der Rotorposition in einem dreiphasigen Elektromotor bei 60° Blockkommutierung, gemäß einem Beispiel zur Erläuterung des Hintergrunds der Erfindung;
• 3A zeigt schematisch einen möglichen Steuersignalverlauf zur synchronen Anlauf-Kommutierung eines Elektromotors, gemäß einem Vergleichsbeispiel;
• 3B zeigt schematisch einen möglichen Verlauf des Motorstroms bei der synchronen Anlauf-Kommutierung des Vergleichsbeispiels der 3A;
• 3C zeigt schematisch einen möglichen Verlauf der induzierten Spannung bei der synchronen Anlauf-Kommutierung des Vergleichsbeispiels der 3A;
• 4A zeigt schematisch einen möglichen Steuersignalverlauf zur synchronen Anlauf-Kommutierung eines Elektromotors, gemäß einem Beispiel;
• 4B zeigt schematisch einen möglichen Verlauf des Motorstroms bei der synchronen Anlauf-Kommutierung des Beispiels der 4A;
• 4C zeigt schematisch einen möglichen Verlauf der induzierten Spannung bei der synchronen Anlauf-Kommutierung des Beispiels der 4A;
• 5A und 5B zeigen schematische Darstellungen von Komparatorsignalen zur Erkennung von Nulldurchgängen relativ zu Kommutierungszeitpunkten für die Anlauf-Kommutierungen der 3A bis 3C bzw. 4A bis 4C;
• 6 zeigt schematisch eine abgewandelte synchrone Anlauf-Kommutierung abhängig von erfassten Nulldurchgängen, gemäß einem weiteren Beispiel.

The invention is explained in more detail below by means of examples with reference to the drawings.

• 1 shows a schematic representation of an electric motor and an associated control circuit according to an example;
• 2A to 2D show schematic representations of the rotor position in a three-phase electric motor at 60 ° block commutation, according to an example to explain the background of the invention;
• 3A shows schematically a possible control signal curve for synchronous start-commutation of an electric motor, according to a comparative example;
• 3B shows schematically a possible course of the motor current in the synchronous start-commutation of the comparative example of FIG 3A ;
• 3C shows schematically a possible course of the induced voltage in the synchronous start-commutation of the comparative example of the 3A ;
• 4A schematically shows a possible control signal curve for synchronous startup commutation of an electric motor, according to an example;
• 4B shows schematically a possible course of the motor current in the synchronous start-commutation of the example of 4A ;
• 4C shows schematically a possible course of the induced voltage in the synchronous start-commutation of the example of 4A ;
• 5A and 5B show schematic representations of comparator signals for detecting zero crossings relative to commutation times for the start-up commutations of the 3A to 3C respectively. 4A to 4C ;
• 6 shows schematically a modified synchronous startup commutation depending on detected zero crossings, according to another example.

Detailed description

1 shows a schematic representation of an electric motor 10 with a control circuit 12 to control the electric motor. In the present example, the electric motor 10 as a three-phase brushless DC motor. In other examples, motors with a different number of phases and stepper motors instead of brushless DC motors can be used. The electric motor basically has a stator and a rotor arranged in the stator. An output shaft of the electric motor can be coupled to a transmission, the output of which is connected to an actuator.

In one example, the electric motor is a sensorless, brushless DC motor (BLDC motor) that does not use external position sensors, for example Hall sensors, to determine the motor position. The control of the commutation of such electric motors is usually based on the knowledge of the position of the rotor relative to the stator. If no position sensors are available, the voltage induced by the electric motor through the electromotive force (BACK-EMF or BEMF voltage) can be evaluated during operation.

Since no BEMF voltage can be measured when the motor is at a standstill and there is therefore no information about the rotor position, synchronous start-up operation can be selected for starting the motor from standstill by specifying the first commutation steps without position feedback. The first commutation steps can, as explained in the following, be calculated on the basis of a desired acceleration and fixedly set in order to continuously increase the speed of the electric motor until a minimum speed is reached at which a BEMF voltage can be detected. If a BEMF voltage is sufficiently high to reliably identify the rotor position of the electric motor, a switch can be made to a sensorless control algorithm for controlling the electric motor.

During synchronous start-up operation, for example, a synchronous ramp can be selected in such a way that successive commutation steps are shortened in each case in order to achieve a desired speed increase. Depending on the motor, algorithm, load and possibly other requirements, the speed can be increased linearly, exponentially or simply as quickly as possible.

The control circuit 12 of the electric motor 10 is set up to carry out a method according to the invention. The control circuit 12 in this example comprises a B6 bridge circuit with six semiconductor switches B1 to B6 which can be designed, for example, as a MOSFET switch. The semiconductor switch B1 to B6 can be bridged by free-wheeling diodes (not shown). The semiconductor switch B1 to B6 are controlled by a control device 14 , for example a microcontroller. By closing (switching on) and opening (blocking) an upper bridge switch B1 . B3 . B5 and a lower bridge switch B2 . B4 . B6 can phase windings U . V . W from the control circuit 12 controlled electric motor 10 energized or separated from the power source, so that the electric motor can be commutated accordingly. The power source is in 1 schematically by a battery voltage VBAT and earth GND illustrated. In the example it is assumed that the electric motor is used in the automotive sector and is powered by a battery.

For example, a six-step block commutation can be implemented, in each of which two of the three phase windings U . V . W are energized and each commutation step, that is to say the time span between two commutation times, corresponds electrically to a rotation of the rotor of 60 °. The semiconductor switch B1 to B6 can also be controlled with a superimposed pulse width modulation (PWM). A frequency in the range from 1 to 100 kHz, in particular in the range from 10 to 25 kHz, is typically selected for the PWM. This can be done by the electric motor via the pulse duty factor of the PWM 3 effected torque and / or a rotational speed of the rotor, or the actuating speed of an actuator coupled thereto, are set. The commutation steps can then take place electrically at a distance of approximately 60 °. The ideal commutation time is then 30 ° electrically distant in the middle between two zero crossings of the BEMF. In other configurations, a 12-step commutation can also be carried out, in which two or three phases alternate U . V . W of the electric motor are supplied.

The B6 bridge circuit is via a buffer capacitor C from a voltage source with the input voltage VBAT provided. Furthermore, the B6 bridge circuit is via a resistor R with earth GND connected.

The figure also shows an exemplary connection of a voltage comparator 16 shown, which has three inputs for connection to one phase winding each, U . V , or W , and has an input for connecting to a base point of the bridge circuit. As the input signal of the voltage comparator 16 can serve a characteristic curve which corresponds to the induced voltage and / or the phase current in the phase winding of the electric motor. An output of the voltage comparator 16 is with an input of the microcontroller 14 the control circuit 12 connected. Instead of a comparator, three separate voltage comparators can be provided, each with two inputs, one input with a phase winding U . V or W is coupled. In principle, the phase windings U . V . W of the three-phase BLDC motor in a star connection or a delta connection. The second input of the voltage comparators 16 can thus, as shown, be coupled to the base point or to a second phase winding or to a star point or a virtual star point of the phase windings.

Additionally or alternatively, the control circuit may include an analog-to-digital converter for detecting the induced voltage or a phase current. In other configurations, the phase current can also be detected by means of a current mirror circuit implemented in the bridge circuit.

Basically, the electric motor is set up to carry out a method according to the invention. The method can be carried out as part of a computer program, for example an operating program / firmware of the control circuit. In particular, such a program can be stored in a non-volatile memory of a microcontroller in the control circuit and called up by the microcontroller. To control the electric motor, the microcontroller can be coupled to semiconductor switches of one or more bridge circuits via control outputs. In the case of a three-phase brushless DC motor, for example, the B6 bridge circuit can be provided. If, on the other hand, the method is used to operate a two-phase stepper motor, two bridge circuits, each with four semiconductor switches, can be provided.

The electric motor can be part of an actuator, in particular a flap actuator. Actuators, which are designed, for example, to adjust a flap, a blind, or a linearly adjustable actuator, can be found in different applications. For example, such actuators can be used to adjust a flap or a blind on a motor vehicle or in an industrial application or in building adjustment systems. Such actuators typically include electric motors which can provide a torque in the range from 0.1 Nm to 20 Nm. A flap actuator of a motor vehicle can, for example, provide a nominal torque in the range from 0.4 Nm to 4 Nm. The actuator can also be a participant of a field bus, for example part of a LIN bus (Local Interconnect Network). It can therefore also be provided that the control circuit can receive commands or information from the bus system and take them into account in the execution of the method.

The actuator can in particular also comprise a reduction gear. Such a transmission can comprise several transmission stages and, for example, have a reduction ratio between the output of the electric motor and the output of the actuator in the range from 1:20 to 1: 2000, in particular in the range from 1: 100 to 1: 1000. The electric motor can, for example, at speeds ranging from 500 min ^-1 to 20,000 min ^-1 are operated. The actuator can for example be designed for a speed of the output of the actuator in the range from 1 min ^-1 to 100 min ^-1 , in particular in the range from 2 min ^-1 to 30 min ^-1 . To detect the induced voltages, for example, a speed of the electric motor in the range from 200 min ^-1 to 800 min ^-1 may be necessary. Otherwise the amplitude of the induced voltage is too low, so that it, or the corresponding phase current, cannot be detected reliably. Therefore, the method can be applied particularly well to actuators whose actuator is driven by the electric motor via a reduction gear. In such electric motors, a comparatively high speed of the electric motor only causes a relatively low speed at the output. A high rotational speed of the rotor magnet can thus be achieved quickly, as a result of which the rotational speeds necessary for detecting the induced voltage and / or the phase current are reached relatively quickly.

2A to 2D schematically show three coils 22 . 24 . 26 according to the three phases U . V and W a three-phase brushless DC motor to illustrate the turning behavior of a rotor, which is schematically represented by a permanent magnet 20 is shown, depending on the energization of the phases U . V and W , The example shows a simplified representation of a synchronous start-up operation based on a situation in which a holding current is applied to the motor and the rotor is in a position that is aligned with a corresponding static field. In a synchronous start-up operation, fixed commutation times or commutation steps can be specified depending on a desired acceleration of the motor. The Commutation steps are calculated so that they correspond to the desired acceleration. In a simplified example, the first commutation steps can correspond to 8 ms, 6 ms, 5 ms and 4 ms, for example.

In 2A are just the phases V and W energized with a constant holding current, and the rotor 20 is corresponding to the coils 24 . 26 of phases V and W aligned. The two phases V and W generate a static field on which the rotor 20 aligns with the phases V and W generate two equally strong poles. The rotor position is therefore known.

In 2 B there is a first commutation step, with the current on the phases U and V is switched. The resulting commutation position is in 2 B shown. Because the rotor 20 was pre-aligned due to the holding current, it only has to rotate 30 ° electrically until the new commutation position is reached, instead of 60 ° electrically, as with a conventional six-step commutation or 60 ° block commutation in stationary operation. The rotor 20 The new commutation position therefore takes up after 50% of the time provided for this.

In 2C there is a next commutation step, the current on the phases W and U is switched. The resulting commutation position is in 2C shown. Because the rotor 20 in this case, if it is not controlled from a static position, but has already started up, it now rotates 60 ° electrically from the previous to the next commutation position, as in a conventional six-step commutation or 60 ° block commutation in stationary operation. The rotor 20 thus takes up the new commutation position within the time provided for this. This is followed by further commutation steps.

1. a. von der Rotorposition bei Anliegen des Haltestroms in den Phasen U, V, der bis zum Zeitpunkt t1 gilt,
2. b. mit einem ersten Kommutierungsschritt zum Zeitpunkt t1, um den Rotor von der statischen Halteposition zur ersten Rotorposition aufgrund der ersten Kommutierung auf die Phasen V, W zu drehen,
3. c. einem zweiten Kommutierungsschritt zum Zeitpunkt t2 von der ersten Rotorposition zur zweiten Rotorposition aufgrund der zweiten Kommutierung auf die Phasen W, U, und
4. d. mit einem dritten Kommutierungsschritt zum Zeitpunkt t3 von der zweiten Rotorposition zur dritten Rotorposition aufgrund der dritten Kommutierung auf die Phasen U, V, wobei der dritte Kommutierungsschritt zum Zeitpunkt t4 endet.

2D illustrates again the movement of the rotor 20:

1. a. from the rotor position when the holding current is applied in the phases U . V who by the time t1 applies,
2. b. with a first commutation step at the time t1 to move the rotor from the static stop position to the first rotor position due to the first commutation on the phases V . W to shoot
3. c. a second commutation step at the time t2 from the first rotor position to the second rotor position due to the second commutation on the phases W . U , and
4. d. with a third commutation step at the time t3 from the second rotor position to the third rotor position due to the third commutation on the phases U . V , with the third commutation step at time t4 ends.

During the holding phase a., Up to the point in time t1 , the induced voltage in the motor phases is zero. The rotor is aligned according to the stator field of the holding phase and stands still. The rotor is at the point in time due to the holding phase t1 at the beginning of the first commutation step already electrically aligned at 30 °. Due to the orientation of the rotor by the holding current, the first step therefore only has to go through 30 ° electrically, while the second and all further steps have to go through 60 ° electrically. However, the first commutation step in the case of commutation according to the prior art at the time t1 a torque on the rotor, which moves the rotor electrically by 60 °. Therefore, if the first commutation step is chosen to be as long as the subsequent commutation steps or longer, this can lead to a strong rotor oscillation, which can modulate on the phase currents induced in the motor. This is with respect to the 3A to 3C explained further.

The inventors have found that this rotor oscillation can occur particularly strongly during the first commutation steps in synchronous starting. Due to the rotor oscillation, which can depend on the inertia of the system, the load and the electrical power fed in, the synchronicity between the rotor position and the commutation is lost. It is no longer mandatory (deterministic) to which rotor position subsequent commutations take place. The rotor swing can even cause the motor to turn in the wrong direction until the next commutation is carried out, thus preventing a desired start-up because the rotor no longer follows the applied rotating field.

EMC (electromagnetic compatibility) can also suffer from this behavior. The vibrations of the rotor and the overall non-deterministic commutation during startup can increase the phase currents in the motor. This results in increased power consumption and poorer EMC or higher costs for the design of the EMC-relevant components.

The 3A . 3B and 3C illustrate such a scenario of synchronous start-up commutation with predetermined commutation steps depending on a desired acceleration. 3A shows exemplary commutation times; 3B schematically illustrates a possible course of the motor current; and 3C illustrated schematically the voltages applied and induced in the motor phases.

A first time of commutation K1 occurs at the time t1 on, a second commutation time K2 at the time t2 , a third commutation time K3 at the time t3 , a fourth time of commutation K4 at the time t4 and a fifth time of commutation K5 at the time t5 , The first commutation step lies between t1 and t2 , the second commutation step lies between t2 and t3 , etc. In a simplified example, the first four commutation steps K1-K2 . K2-K3 . K3-K4 . K4-K5 for example 8 ms, 6 ms, 5 ms and 4 ms. At one commutation time K1 . K2 . K3 . K4 . K5 the current supply to the phases is switched in each case, as with reference to the 2A to 2C explained.

In the example shown, it is again assumed that the first commutation at the time of commutation K1 takes place after a holding current has been applied to the motor and the rotor is in a static rotor position, which results from the alignment of the rotor with the magnetic field generated by the holding current. The rotor is therefore not where it would be at a commutation time during stationary operation, but rather a position that corresponds to the middle between two commutation times. Due to the relatively long first commutation step Ki-K2 , which was calculated on the basis of a desired acceleration, overshoot or even generation of an opposing torque of the rotor before the next commutation and consequently rotor oscillation can occur.

3B first shows the holding current In, which is before the first commutation time K1 and then the motor current during the commutation steps K1-K2 . K2-K3 , etc. One recognizes an overall very high motor current and a strong overshoot of the motor current, with strong current peaks in each case at the end of a commutation step, which results from the rotor oscillation. As a result, the motor's power consumption increases unnecessarily and the EMC deteriorates.

3C schematically illustrates those in the phases 20 . 24 . 26 . U . V . W , the applied and induced voltages of the motor starting from a zero value which is present during the application of the holding current I _H before the first commutation time K1 Is accepted. The holding current is in the phases V and W on. The tension in the phase V is zero during this time. A so-called “full block” control is also assumed, in which the pulse duty factor of a PWM is 100%. U + denotes the positive supply voltage, U denotes earth, and U _diode denotes a capped induced voltage due to a diode in the phase string. The induced voltages can in particular have a very low amplitude, so that reliable detection is difficult to achieve.

First you see him 3C that the voltage curve in all three phases U . V and W does not have a pronounced acceleration ramp and therefore zero crossings of the induced voltages are difficult to detect. During the first commutation step, between t1 and t2 , no zero crossing of an induced voltage can be determined. If zero crossings can be determined or "guessed", they are not in the middle between two commutation times, but occur significantly too early, with in 3C Zero crossings with ZC (zero crossing) are marked. The induced voltage shows that the rotor does not follow the rotating field. Since the first commutation takes too long, the first detected zero crossing occurs ZC1 not in the middle, but at the beginning of the second commutation step K2-K3 on. The same applies to the next zero crossing ZC2 , Furthermore, the amplitudes of the voltage profiles can be significantly smaller than shown, so that, for example, reliable detection of zero crossings is hardly possible for this reason either.

Basically, for an ideal commutation it applies that a zero crossing of the induced voltage should occur in the middle between two successive commutation times. Accordingly, the commutation times of the theoretical commutation times (theoretical commutation times) should be chosen so that they represent ideal commutation times. Of these, the commutation control of the prior art, which in the 3A to 3C is illustrated by way of example, far away.

It is possible to switch from synchronous start-up operation to sensorless asynchronous operation if, for example, a predetermined number, for example three zero crossings of the induced voltage have been reliably detected or if a predetermined number of commutation steps have been carried out or if a predetermined speed has been detected. The ramp of the induced voltage increases with increasing speed, so that zero crossings can be better recognized. A possible time for switching over to sensorless operation is in 3B exemplary at S shown. Due to the rotor swinging, as stated, the power consumption is increased and the EMC is deteriorated. The time is also delayed S of switching to sensorless operation.

The 4A . 4B and 4C illustrate a synchronous start-up commutation with predetermined commutation steps depending on a desired acceleration, the first interval being modified. 4A shows exemplary commutation times; 4B schematically illustrates a possible course of the motor current; and 4C schematically illustrates the voltages induced in the motor phases.

A first time of commutation K1 occurs at the time t1 on, a second commutation time K2 ' at the time t2 ' , a third commutation time K3 ' at the time t3 ' , and a fourth commutation time K4 ' at the time t4 'etc. The first commutation step lies between t1 and t2 ' , the second commutation step lies between t2 ' and t3 ' , etc. In this example it is assumed that the first four commutation steps K1-K2 ' . K2'-K3 ' . K3 'K4' , mathematically would be 8 ms, 6 ms, 5 ms and 4 ms if a desired acceleration is to be achieved. Here is the first commutation step K1-K2 ' reduced by 50% to 4 ms to take account of the fact that the rotor is pre-aligned due to a previously applied holding current, as with reference to FIG 2A to 2D was explained. How about 2D illustrated, in the case assumed there, the first commutation step for an ideal control should be electrical after only 30 ° at the time t2 be ended. If this connection is not taken into account, the rotor can overshoot over its ideal orientation in relation to the applied rotating field and is asynchronous at the time of the subsequent commutations K2 . K3 . K4 , ... The second commutation time K2 ' In the simplest case, this takes place with a pre-commutation of 50% of the calculated duration of the corresponding commutation step. At the respective commutation times K2 ' . K3 ' . K4 ' . K5 ' the current supply to the phases is thus switched earlier than in the comparative example.

It has been shown that a reduction in the first commutation step by half, even without further adjustments or pre-commutation, which will be explained in the following, can significantly improve the starting behavior of the electric motor.

In the example shown, it is again assumed that the first commutation takes place at the commutation point in time K1 = t1 after a holding current has been applied to the motor and the rotor is in a static rotor position, which is due to the orientation of the rotor to that caused by the Holding current generated magnetic field results. The rotor is therefore not where it would be at a commutation time during stationary operation, but at a position that corresponds to the middle between two commutation times. The first commutation step K1-K2 , which was calculated on the basis of a desired acceleration in order to prevent the rotor from overshooting before the next commutation, is therefore shortened. In the example, the commutation step is shortened by 50% compared to the arithmetic value in order to bring about a rotation of the rotor of approximately 30 ° electrically instead of 60 ° electrically. This can also be referred to as pre-commutation. As a result, the rotor is at the end of the first commutation step K1-K2 ' there or roughly where it should be located, as in 2 B shown.

4B first shows the holding current I _H which is before the first commutation time K1 and then the motor current during the commutation steps K1-K2 ' . K2'-K3 ' , etc. One recognizes an overall lower motor current and a significantly lower overshoot of the motor current than in the scenario of 3A , with no or significantly lower current peaks at the end of each commutation step. This can be achieved because the rotor swing is largely or completely avoided. This reduces motor power consumption and improves EMC.

4C shows schematically the in the phases 20 . 24 . 26 . U . V . W , the motor induced voltages based on a zero value during the application of the holding current I _H before the first commutation time K1 Is accepted. The holding current is also in the phases in this example V and W on. The tension in the phase V is zero during this time. Furthermore, a so-called “full block” control is again assumed.

First you see him 4C , especially compared to 3C , a more pronounced acceleration ramp in all three phases U . V and W , which allows it after the second commutation K2 ' To reliably detect zero crossings of the induced voltages. Furthermore, the zero crossings are largely in the middle between two commutation times. The induced voltage thus shows that the rotor follows the rotating field. The commutation is therefore at or close to the ideal commutation.

In practice, the acceleration and alignment of the rotor is dependent on external influences, such as a load on the motor. In a modification, the first shortened commutation step can therefore also be parameterizable, it being able to be further shortened without a load, for example, and less shortened with a load. The load can be determined, for example, by detecting the phase current. Also a division of the Shortening to several successive commutation steps is possible. This allows the motor start-up to be optimized from a standstill for various applications.

5A and 5B show schematic representations of comparator signals for detecting zero crossings relative to commutation times for the start-up commutations of the 3A to 3C respectively. 4A to 4C , The zero crossings ZC (zero crossing) of the induced voltage can be determined using the in 1 shown voltage comparator 16 determine. As the input signal of the voltage comparator 16 can serve a characteristic curve which corresponds to the induced voltage and / or the phase current in the phase winding of the electric motor. The output signals of the voltage comparator 16 can, for example, via IO inputs of the microcontroller 14 that is set up to carry out the method can be evaluated. The microcontroller 14 can generate corresponding interrupts from the comparator signals, which are used to control and / or regulate the electric motor 10 can be used. The 5A and 5B show comparator signals in the form of a pulse, its amplitude U _comp can assume a high or a low voltage level. Becomes a zero crossing through a voltage comparator 17 A point in time can be detected, for example, via the rising edge of the comparator signal t _ZC a zero crossing ZC1 . ZC2 , ... are recorded.

6 shows schematically a modified synchronous startup commutation depending on detected zero crossings ZC , according to a further example, which can be combined with the examples described above. In the 6 Modification shown is based on the synchronous start-up commutation, which in the 3A to 3C and 5A is shown. Accordingly, in 6 the commutation times with K1 to K5 designated. The modification could also correspond to the startup commutation of the 4A to 4C and 5B be applied. In particular, in the 6 an adaptation of the commutation times of the theoretical commutation steps by means of calculated optimized next commutation steps is explained. The 6 also describes variants in which the optimized next commutation step is determined on the basis of an ideal commutation step, which is calculated as a function of the detected characteristic. The in the 6 mentioned commutation times K1 to K5 correspond to theoretical commutation steps.

Furthermore shows 6 exemplary zero crossings ZC1 . ZC2 . ZC3 that can result from this commutation. There is a commutation point in stationary operation K1 . K2 , etc. in the middle between two zero crossings of the induced voltage, which arises due to the back electromotive force (BEMF). While the induced voltage is generated in all phases of the motor, it is sufficient to record it in the currently undetermined phase. Ideally, for example, the commutation time could be K2 lie in the middle between a zero crossing of the U phase and a zero crossing of the W phase.

Based on the calculated commutation times K1 to K5 are in 6 the zero crossings that define the theoretical commutation times EZC1 to EZC4 (EZC = expected zero crossing). Actually detected zero crossings are in 6 exemplary with ZC1 to ZC3 designated. In particular, the actual zero crossings ZC1 to ZC3 , as explained above, earlier than the expected zero crossings EZC1 to EZC3 occur, ie not in the middle between two commutation times, but at an earlier time. Depending on the application and, for example, an applied load, the actual zero crossings can also occur later than the expected zero crossings. The commutation times can be adapted depending on the detected zero crossings in order to determine optimized commutation times which are at least approximated to the theoretical or ideal commutation.

In the example of the 6 is the time interval between the first zero crossings ZC1 and ZC2 With A designated. If one assumes stationary operation, then a commutation point lies in the middle between two zero crossings of the induced voltage or a zero crossing lies in the middle between two commutation points in time. So the next commutation time would have to be in stationary operation K3 , after the second detected zero crossing ZC2 , at a time interval of A / n after the detected second commutation time ZC2 occur, where n = 2 was selected. This is in 6 as MK3 (MK = modified commutation time). If one also takes into account that the motor is in a start-up or acceleration phase and not in stationary operation, the modified commutation time could MK3 done even earlier.

Thus, for example, the time of commutation K3 depending on the times when the zero crossings ZC1 and ZC2 were recorded, adjusted. The modified commutation time MK3 based on A is calculated, then MK3 = ZC2 + A / n, with n = 2, and is therefore brought forward. If pre-commutation is also taken into account for the start-up or acceleration phase, the time of commutation can be changed MK3 brought forward become electrical, for example by up to 30 °. This results in a modified commutation time MK3PC (PC = pre-commutation). The pre-commutation can alternatively be described by a suitable choice of n, with n> 2.

A corresponding adjustment of the next commutation time can be based on the time interval B between the second and third zero crossing ZC2 and ZC3 respectively. The time of commutation is used for this K4 depending on the times when the zero crossings ZC2 and ZC3 were recorded, adjusted. The modified commutation time MK4 based on B is then calculated at MK4 = ZC5 + B / n, with n = 2, and is thus relative to the originally calculated point in time K4 advanced. If pre-commutation is also taken into account for the start-up or acceleration phase, the time of commutation can be changed MK4 be advanced, for example by up to 30 ° electrically or up to 15 ° electrically. This results in a modified commutation time MK4PC ,

The above example is based on the assumption of a 60 ° block commutation, in which two of the three phases are always energized in the course of an electrical revolution, so that the distance between two successive commutations is 60 ° electrical because six different combinations of energized phases are possible and be run through. With sensorless motor operation, the pre-commutation at the start of the start-up phase can be, for example, 15 ° to 30 ° electrically, 30 ° for example corresponding to the distance between a zero crossing and the next commutation time. After the first sensorless commutation step, the pre-commutation can be electrically reduced to 0-15 °. In various configurations, a constant pre-commutation angle or a pre-commutation angle that gradually approaches zero can be selected. The pre-commutation of a subsequent commutation step can in each case be smaller than the pre-commutation of the previous commutation step. The pre-commutation can be, for example, up to 0.5Δtzc, where Δtzc is the time interval between two successive zero crossings, in the above examples Δtzc corresponds to A or B, for example.

More generally, the time of the pre-commutation can be determined in proportion to a time interval ΔtZC between two previously detected, successive zero crossings. It can be provided that the time of commutation is a fraction of the time interval ΔtZC, e.g. A or B, before the ideal commutation time. For example, the commutation time of the first commutation step can be 0.03 ΔtZC to 0.25 ΔtZC, in particular approximately 0.125 ΔtZC, before the ideal commutation time. As a result, the electromagnetic fields of the stator and the rotor of the electric motor can be aligned relative to one another by means of the pre-commutation, and the electric motor can be operated efficiently and reliably.

In further refinements, it can be provided that second commutation steps following the first commutation step are determined by the control circuit in such a way that they have a shorter pre-commutation time than the first commutation step. For example, the time of the pre-commutation can again be proportional to a time interval ΔtZC2, e.g. B, between two previously detected, successive zero crossings. For example, the commutation times of the second commutation steps can be 0.015 .DELTA.tZC2 to 0.125 .DELTA.tZC2, in particular approximately 0.0625 .DELTA.tZC2, before the ideal commutation time.

If it is detected that zero crossings occur later than expected, commutation times can also be adjusted by moving the times behind the calculated values.

In terms of 6 The example described were the commutation times that occurred on the second commutation time K2 follow, only due to the time interval, e.g. A or B , determined by two previous zero crossings. In a modification, the modified commutation times can be calculated MK3 and MK4 the originally calculated commutation steps are also taken into account and / or a weighting is carried out. For example, an average could be formed from an originally calculated commutation step and a commutation step calculated on the basis of the distance between the zero crossings. In a further example, the commutation steps calculated on the basis of the distance between the zero crossings can be weighted with a constant factor or with a factor increasing or decreasing depending on the respective commutation step. This is particularly expedient because the acceleration and alignment of the rotor can depend on external influences, such as a load on the motor. This can be taken into account by parameterizing or weighting the commutation steps.

In a further embodiment, which can be combined with the variants described above, the second commutation time can already occur K2 based on a first detected zero crossing ZC1 be adjusted. For example, the time interval between the first commutation point in time kl and the first zero crossing ZC1 can be determined, it being assumed that this distance α / n, with n = 2, corresponds to approximately half a commutation step. An adjusted second commutation time can be derived from this K2α be calculated by a period of time α after the first commutation time K1 lies.

If pre-commutation is also taken into account for the start-up or acceleration phase, the time of commutation can be changed K2α be advanced even further, for example by up to 30 ° electrically. This results in a modified commutation time, which is not shown in the figures.

Finally in 6 shown as an example, starting from the modified commutation time MK3PC a next modified commutation time MK4PC can be determined. For this the time interval β between the modified commutation time MK3PC and the next zero crossing ZC3PC (in this case, a next zero crossing when providing a pre-commutation), and the next modified commutation time is set to MK4PC = ZC3PC + β / n, with n = 2.

The various approaches explained above for adapting the commutation times, with or without pre-commutation, can be combined and, in a corresponding manner, to the already optimized synchronous start-up procedure, which relates to the 3 to 3C and 5B has been explained. The pre-commutation can be set by setting an electrical pre-commutation angle before a next theoretical commutation time, for example from 15 ° to 30 ° electrically or 3 ° to 15 ° electrically. In a modification, the pre-commutation can be set by applying a factor to the respectively calculated commutation time interval by choosing n not equal to 2, for example 1.25 n n 50 50. The next commutation time in each case can thus be determined in relation to the previous zero crossing that it is not shifted exactly by half the previous time interval between two zero crossings, but by a slightly shorter or longer time span, depending on the desired acceleration, load or other operating conditions.

In a further modification, a rule can also be set up according to which a detected zero crossing is only taken into account if it has a minimum distance from a previous commutation time or that a minimum minimum distance is always assumed in order to avoid a too high switching frequency of the commutation. The method assumes that a characteristic curve can be determined on the basis of the induced voltage and / or the phase current of the electric motor, which allows the determination of zero crossings. As explained above, this is possible due to the modification of the commutation control with a shortened first commutation step. If no zero crossings can be determined during the first commutation step or if the position of the zero crossings does not appear to be reliable, the calculated commutation times or the above-mentioned minimum distance can be used for the first commutation step.

In this way, successive commutation times can be determined until a condition for switching to sensorless synchronous operation is fulfilled. This condition can be, for example, reaching a predetermined number of commutation steps, detecting a predetermined number of zero crossings, reaching a target speed, reaching a target position of an actuator, etc. During the start-up phase, for example, the electric motor can be operated in a regulated manner. The rotational speed can be regulated, for example, by modulating a duty cycle of a PWM. become. A current limitation can also be provided.

LIST OF REFERENCE NUMBERS

10

electric motor

12

control circuit

14

control device

16

voltage

20

rotor

22, 24, 26

Do the washing up

C

capacitor

R

resistance

FP

nadir

Claims (27)

Method for controlling an electric motor (10) in the start-up mode, wherein an induced voltage and a phase current are generated in phase windings of the electric motor (10), the method comprising: Detecting a characteristic curve based on the induced voltage and / or the phase current, and Actuation of the electric motor (10) depending on the characteristic curve detected. Procedure according to Claim 1 , further comprising: determining zero crossings ZC1, ZC2 of the characteristic curve and actuating the electric motor (10) depending on the determined zero crossings. Procedure according to Claim 2 The phase current of the electric motor (10) is regulated as a function of the position in time of the detected zero crossings ZC1, ZC2. Procedure according to Claim 2 or 3 , A speed or an acceleration of the electric motor (10) being regulated as a function of the temporal position of the detected zero crossings ZC1, ZC2. Procedure according to Claim 4 , the speed or acceleration being regulated by changing a duty cycle of a pulse width modulation of the phase current. Procedure according to one of the Claims 1 to 5 , wherein the actuation of the electric motor (10) comprises: determining an optimized next commutation time as a function of the detected characteristic or the zero crossings of the detected characteristic; and commutating the electric motor (10) at the optimized next commutation time. Procedure according to Claim 6 , the optimized next commutation point in time being determined from a time interval A between two successive zero crossings ZC1, ZC2 and by a time period A '= A / n after the second ZC2 of the two zero crossings, 1.25 ≤ n 50 50. Method according to one of the preceding claims, wherein the actuation of the electric motor (10) further comprises: Calculating at least one theoretical commutation step depending on a desired acceleration or a desired value of the phase current of the electric motor (10); and Control of the electric motor (10) depending on the theoretical commutation step. Procedure according to one of the Claims 6 or 7 as well as after Claim 8 , the method further comprising: determining the optimized next commutation step as a function of the detected characteristic and the theoretical time of commutation. Procedure according to one of the Claims 8 or 9 , further comprising: energizing phase windings of the electric motor (10) during a first commutation step up to a pre-commutation point in time which lies before the calculated end of the theoretical commutation step. Procedure according to Claim 10 , wherein the electric motor (10) is a three-phase brushless DC motor and the pre-commutation is 3 ° to 15 ° electrically before the calculated end of the theoretical commutation step. Procedure according to one of the Claims 6 to 11 , wherein for at least two commutation steps following a first commutation step, the optimized next commutation time is determined on the basis of the zero crossings of the characteristic curve during the respectively preceding commutation steps. Procedure according to Claim 12 , wherein the two commutation steps following the first commutation step have a smaller pre-commutation than the first commutation step and / or wherein the commutation steps are successively approximated to theoretical commutation steps. Method according to one of the preceding claims, wherein the optimized next commutation point in time is determined such that a zero crossing of the characteristic curve lies in the middle or approximately in the middle between two successive commutation points in time. Method according to one of the preceding claims, wherein the electric motor (10) is operated with a block commutation, wherein phase windings are supplied with a constant voltage. Procedure according to Claim 15 , wherein the block commutation is superimposed on pulse width modulation. Method according to one of the preceding claims, further comprising: Applying a holding current to the electric motor (10) before executing the first commutation step. Procedure according to Claim 17 , further comprising: at the start of the start-up phase, calculating at least the first theoretical commutation step as a function of a desired acceleration of the electric motor (10); and selecting a first commutation step, which is shorter than the calculated first theoretical commutation step, for controlling the electric motor (10) in the start-up phase. Procedure according to Claim 18 , the length of the first commutation step, which in the Start-up phase is selected, is 50% or about 50% of the length of the first theoretical commutation step. Procedure according to one of the Claims 8 to 19 , wherein the start-up phase is a synchronous start-up phase with theoretical commutation steps and wherein the commutation times of the theoretical commutation steps are dynamically modified by means of the at least one optimized commutation step. Procedure according to Claim 20 , whereby the commutation times of the theoretical commutation steps are calculated in operation or are stored in a table. Procedure according to one of the Claims 20 or 21 , the optimized next commutation time being a maximum of 15 ° electrically before or after the commutation time of the theoretical commutation step to be adapted. Procedure according to one of the Claims 20 to 22 wherein switching from the synchronous start-up phase to a sensorless controlled motor operation after at least one of the following conditions is fulfilled: reaching a predetermined number of commutation steps; reaching a predetermined number of zero crossings; reaching a predetermined speed; the detection of a back EMF that is above a predetermined threshold. Procedure according to one of the Claims 6 to 23 , the optimized next commutation time being determined as a function of the preceding commutation times. Electric motor (10) comprising phase windings (22, 24, 26) and a control circuit (12, 14), the control circuit (12, 14) being set up to carry out the method according to one of the preceding claims. Electric motor (10) after Claim 25 , further comprising a comparator (16) and / or an analog-digital converter for recognizing zero crossings of a characteristic curve which is based on an induced voltage and / or a phase current of the electric motor (10). Actuator with an electric motor (10) Claim 25 or 26 and with a reduction gear, which has a reduction ratio between an input and an output of the reduction gear in the range from 1: 100 to 1: 1000.

Images