DE102013014480A1 - Method for operating an electric motor - Google Patents

Abstract

The invention relates to a method for operating a brushless DC motor (2) on a DC operating voltage (Vin) by means of a fully connected bridge circuit (11) with semiconductor switches (Q). The method preferably has two operating phases, which are run through each time the motor (2) is switched on one after the other. A starting phase (21), in which the engine (2) is accelerated from standstill, and a running phase (22), which adjoins the starting phase (21). If the DC operating voltage (Vin) during the start phase (21) is less than a defined threshold voltage (VTL), a current control by means of a pulse width modulation, wherein the PWM is operated with a PWM frequency F1 (16). In the starting phase (21), when the DC operating voltage (Vin) is greater than or equal to a defined threshold voltage (VTL), the PWM frequency is reduced to a frequency F2 (17), the PWM frequency F2 being less than the PWM frequency F1 is. The transition from the start phase (21) into the running phase (22) takes place due to a predetermined transition condition.

Description

The invention describes a method for operating an electric motor on a bridge circuit with semiconductor switches.

The electric motor can be given in particular by a brushless DC motor. Alternatively, a DC motor with brushes or a synchronous motor, for example a stepper motor which can also be operated as a brushless DC motor, may be provided. Brushless DC motors (BLDC motors) are well known in the art. An advantage of BLDC motors is that they are commutated electronically. In this case, the alternating field for driving the individual motor phases is generated by an electronic inverter, whereby the motors are much less susceptible to wear than motors with brush commutators. The speed of the motor can be controlled by a pulse width modulation (PWM) of the operating voltages at the windings almost independent of the DC operating voltage. For this reason, it is possible to design BLDC motors to be usable over a wide operating voltage range, for example, from 6 VDC to 24 VDC. The motor windings can be designed so that the rated speed is already reached at a relatively low voltage, which is for example in the range between 6 VDC and 15 VDC. At higher voltages, the speed is controlled by a PWM.

The inverter for converting the operating DC voltage into the individual motor phase voltages can be realized for example as a bridge circuit with semiconductor switches, in particular as a full bridge circuit, which is also known as B6C bridge. In addition, there are numerous other inverter circuits that are well known to those skilled in the art. As a semiconductor switch in particular MOSFET switches can be used. The semiconductor switches are controlled by a control circuit, wherein the control circuit may in particular comprise a microcontroller or may be integrated in this. The bridge circuit and corresponding control circuits are widely known in the prior art, which is why details should not be discussed here.

For example, MOSFET switches have a characteristic switching time, which elapses until after the application of the gate voltage, the drain-source path turns on, that is, a current flows. The turn-on time of the MOSFET switch is for example 4.5 μs. Among other things, this switching time depends largely on the internal resistance RDSon of the drain-source path in the switched-on state. In general, this resistance is on the order of milliohms. By series dispersion, especially in low-cost MOSFETs, this resistance can be quite larger, which also extends the turn-on time.

The problem underlying the invention relates mainly to motors designed for a wide operating voltage range. Since the winding must be designed for the smallest specified voltage, the motor current and / or the speed must be controlled at higher voltages, for example by a pulse width modulation (PWM). In this case, the switching time of the MOSFET switch defines the shortest possible turn-on time of a PWM clock.

It may now happen that the voltage is so high that the motor current would be far above the limit or a required speed is exceeded. To limit speed and current to the desired limit and / or to regulate, a very small duty cycle of the PWM would be necessary and thus a very short turn-on time of the PWM clock. If the necessary switch-on time of the PWM clock is shorter than the shortest possible switching time of the semiconductor switch, the semiconductor switch can no longer switch and no current flows through the motor in this PWM cycle. In this case, the engine can not be started.

The object of the invention is therefore to provide a method for operating an electric motor, in which over a wide operating voltage range of the start and the controlled operation of the engine is ensured.

This object is achieved by the method according to the main claim.

The operation of the engine is divided into at least two operating phases, which are run through each time the engine is turned on. A starting phase in which the engine is accelerated from standstill, and a running phase, which follows the starting phase.

In the starting phase, when the operating DC voltage is smaller than a defined threshold voltage, current limiting and / or current regulation takes place by means of a pulse width modulation (PWM), wherein the PWM is preferably operated with a PWM frequency F1. The PWM frequency is usually greater than 16 kHz and is thus outside the audible frequency range, so that no disturbing noises. Particularly preferably, the PWM frequency is greater than 20 kHz. To limit and / or regulate speed and / or current, the duty cycle of the PWM can be changed, wherein a longer turn-on time of the PWM signal brings a higher current or a higher speed with it.

According to the invention, current limiting and / or current regulation by means of pulse width modulation (PWM) takes place at least when the DC operating voltage (Vin) in the starting phase is greater than or equal to a defined threshold voltage (V _TL ), the PWM until reaching a predetermined Limit speed N _{G is operated} with a PWM frequency F1. The invention should not be limited to the fact that during the start phase, the DC operating voltage Vin> V _TL , but can also be used independently. In particular, the switching of the PWM frequency in each operating point can be done depending on Vin and / or speed of the engine.

The PWM frequency is preferably greater than 16 kHz and is thus outside the audible frequency range, so that no disturbing noises. Particularly preferably, the PWM frequency is greater than 20 kHz. For example, if the current speed is below a predetermined limit speed N _G , the PWM frequency is reduced to a frequency F2, where the PWM frequency F2 is less than the PWM frequency F1.

In a preferred embodiment of the method, however, the PWM frequency F1 is reduced to the frequency F2 only when a transition condition has been met, for example, when a limit speed N _{G has been reached} , when the operating DC voltage (Vin) is greater than or equal to a defined threshold voltage (V _TL ) is.

At a smaller PWM frequency, the duration of a PWM clock increases. The limitation and / or regulation of current and / or speed takes place only after the duty cycle of the PWM, regardless of the PWM frequency. At the same duty cycle, the turn-on time is longer with a smaller PWM frequency than with a higher PWM frequency. The invention makes this easy to use in that the PWM frequency is reduced until the turn-on time of the required duty cycle is greater than the switching time of the MOSFET switch.

The PWM frequency F2 can be reduced practically arbitrarily, and it makes sense in principle, if the PWM frequency F2 despite the frequency reduction is still above the audible frequency range.

As a result, the motor can be safely started at almost any operating DC voltages. Current and / or speed can be regulated and / or limited at any time.

The current consumption of the motor is greatest when starting from standstill and decreases rapidly with increasing number of revolutions. Therefore, the problem of the invention occurs mainly during the start-up phase when the duty cycle of the PWM due to the high current would have to be reduced so much that the turn-on time of a PWM clock is shorter than the switching time of the MOSFET switch is.

As soon as the current consumption has dropped so much that a current control by the PWM is possible again due to the larger duty cycle, it is possible to switch from the starting phase to the running phase of the motor. The transition from the starting phase to the running phase therefore takes place on the basis of a predetermined transition condition. For this purpose, it is preferably checked after each commutation, whether the transition condition is met. If yes, it will switch to the running phase.

In a development of the method, in the starting phase, when the operating DC voltage is greater than or equal to a defined threshold voltage and when a starting condition is met, the PWM frequency is increased again within the start phase from the reduced PWM frequency F2 to the PWM frequency F1 ,

The start condition may be, for example, the achievement of a certain number of commutation steps, or in particular the achievement of a predefined speed.

In an advantageous embodiment of the invention, the commutation of the motor is sensorless. The transition from the start phase to the run phase then takes place when the transition condition is met, for example when a predetermined number of successive zero crossings in the phase windings as a result of the so-called back electromotive force ("back electromotive force", BEMF) induced voltage can be reliably detected. For example, it may be provided that a certain number of consecutive zero crossings of this BEMF voltage have been reliably detected.

In general, the transition condition may also be given by a change of an operating mode of the electric motor, the operating modes differing, for example, in the type of regulation. In particular, it can be provided that the electric motor is operated by means of a speed control from reaching the transition condition, said speed control can be superimposed on a current control. It may also be advantageous, especially in fan applications, when in this phase of operation the electric motor rotates as fast as possible, so that only one speed control is provided in a fixed voltage and speed range. If a predefined voltage value has been exceeded or a previously defined speed has been exceeded, then a current control can be superimposed again.

If the motor position sensors are on, the commutation also takes place in the starting phase with the aid of this position information.

Especially with sensorless motors, however, no position information is available in the starting phase. Therefore, in a preferred embodiment of the invention in the start phase, the commutation according to a fixed commutation pattern. For this purpose, the commutation times of each commutation step can be stored, for example, in a table. The commutation times can be chosen so that a preferably linear acceleration of the rotor takes place. Of course, other acceleration characteristics can also be achieved. The specification of the commutation times according to a fixed commutation pattern ensures in most cases a starting of the motor in the desired direction of rotation. The energization of the motor windings can thus be done with an increasing frequency, without the position of the rotor must be known. Thus, a start is achieved in synchronous operation of the motor, the rotor thus accelerated by means of a forced commutation.

Although the commutation according to a fixed commutation pattern makes sense in sensorless motors, it can of course also be used in motors with position sensors.

Depending on the application of the motor, it may well happen that the rotor is moved by external forces, such as a fan by acting on the fan blades wind. In this case, this external excitation also cause a direction of rotation opposite to the desired or intended direction of rotation of the rotor. In extreme cases, it may happen that even during commutation with a fixed commutation pattern, the rotor continues to run in the wrong direction. Or the motor current increases through the necessary direction reversal over the limit limit. In both cases, an error occurs, which is detected by the control circuit and leads to switching off the motor.

In a development of the method according to the invention, therefore, a holding phase is always run through in time before the starting phase, in which the rotor is moved to a defined starting position and held there for a predefined holding time. This is done for example by applying two motor windings with a constant current. Alternatively, a fixed voltage or a predetermined PWM pattern, for example, with a fixed turn-on time of 30%, can be applied to one or more motor windings.

Holding the rotor at the stop position ensures that the rotor actually stops and does not move before starting the starting phase. As a result, the engine can now be reliably started in the desired direction of rotation in any case.

This facilitates the starting of the motor in the correct direction of rotation, in particular in sensorless motors, where at the beginning no position information by the counter electromotive force are available.

Advantageously, the switching of the PWM frequency F1 to F2 and vice versa takes place before and / or during the holding time in order, for example, to achieve the most accurate current regulation possible during the holding phase.

This method is particularly advantageous in combination with a commutation in synchronous operation according to a fixed commutation pattern.

The method according to the invention is explained in more detail below on the basis of exemplary embodiments with reference to the accompanying drawings.

It shows:

1 a cross section through a fan with a motor which is operated by a method according to the invention,

2 a circuit diagram of a control circuit for the fan of 1 with a full bridge circuit,

3 a flowchart of a first embodiment of the method according to the invention,

4 a flowchart of a second embodiment of the method according to the invention according to the 3 with an additional holding phase,

5 a flowchart of a third embodiment of the method according to the invention according to the 4 , with an additional start condition in the starting phase.

The 1 shows a cross section through a fan 1 in which the invention is explained by way of example. Of course, the invention is not for use with a fan 1 limited. Rather, the invention can be used in any brushless electric motor and brushed DC motor.

The fan in the embodiment has a three-phase, brushless electric motor (BLDC motor) 2 on, which is commutated without a sensor. For rotor position determination, therefore, there are no sensors, in particular no Hall sensors, but the zero crossings of the voltage induced by the counter-electromotive force are detected. The sensorless operation of a BLDC motor is basically known, which is why its operation should not be further explained here.

The engine has a stator 3 with 12 stator poles 4 and a rotor 5 with 9 magnetic poles, passing through permanent magnets 7 are formed. The number of poles of the engine are exemplary only and are in no way critical to the invention. At the rotor 5 is an impeller 8th arranged with a wave 6 is connected and generates an air flow upon rotation. The motor 2 is in a housing 9 arranged, which directs the air flow, carries the stator and supports the shaft by means of ball bearings. Because the fan 1 Here, by way of example only, as a possible application is shown, will not be discussed further features of the fan.

The motor is designed for an operating DC voltage range between 6 VDC and 22 VDC. To operate the three-phase motor 2 at a DC operating voltage Vin is an inverter 11 necessary. Such an inverter 11 is exemplary in 2 shown. The three motor windings U, V, W of the BLDC motor 2 are interconnected in a star connection. Alternatively, a delta connection can be used. In the example is the inverter 11 designed as a full bridge circuit. For each motor phase U, V, W, the inverter indicates 11 two semiconductor switches, such as in the form of MOSFET switches Q, which are connected in series between the DC operating voltage Vin and ground GND. The motor phases U, V, W are each connected to connection points between the two switches.

The MOSFET switches Q are individually through the control circuit 12 controlled, the control circuit 12 for example, is given by a microcontroller. In the example, the gate inputs of the MOSFET switches Q through the signals BHn through the control circuit 12 connected.

The illustrated inverter 11 is well known in the art, for example, as a B6C circuit, which is why circuit details are not further explained here.

The inverter 11 also has a measuring circuit 13 for measuring the operating _DC voltage (Vin), on the one hand with measuring resistors R _SHUNT and on the other hand with an analog-to-digital converter input (ADC) of the control circuit 12 connected is.

Furthermore, a virtual star point MTC is provided which consists of a resistance network, which is connected to the supply lines of the three motor phase. The virtual star point is used to determine the voltage level, which serves as a zero crossing for the BEMF voltage.

In the example, the control circuit comprises 12 the inverter 11 , the virtual star point and the measurement circuit 13 and is on a circuit board 10 in the motor housing 9 arranged. Alternatively, at least parts of the above circuits may be externally located outside the control circuit.

The BLDC motor 2 in the example is operated with a current control. In this case, a fixed current limit is predetermined, for example by the maximum current carrying capacity of the MOSFET switches or the motor windings. The current regulation is realized by a pulse width modulation (PWM), in which the MOSFET switches Q are additionally switched within a commutation step with a PWM frequency. To control the current, the motor current is measured and the duty cycle of the PWM adjusted so that at the current operating DC voltage Vin, the current is below the current limit. A duty cycle of 100% in this context means that the voltage is continuously on during a PWM cycle. At a duty cycle of 0%, the voltage would always be off. At 50% duty cycle, the turn-on time and turn-off time are each half a PWM cycle. In addition to the current regulation, there is preferably a current limitation which prevents, for example, when blocking the rotor, that the current increases beyond the maximum current limit.

Instead of the current control, it is also possible to regulate a speed control or another operating parameter, the control always being effected by the PWM.

For current regulation to be possible, the PWM frequency must be substantially greater than the commutation frequency, so that at least some PWM cycles occur within a commutation step. The PWM frequency is preferably above 16 kHz, so that no audible switching noise. In the example, the PWM frequency at 23 kHz and the commutation frequency at 630 Hz. Both, however, depends inter alia substantially on the number of poles of the motor and the rated speed, which is why the above operating parameters are in any way limiting.

The problem underlying the invention is that the switching time of the MOSFET switches is greater for large operating DC voltages Vin than the short turn-on time of the PWM necessary for the current regulation. In this case, the MOSFET can not turn on before the PWM is already off again, causing the motor to not start.

This problem will now be explained by the in the flow chart of 3 shown, solved inventive method. By applying the DC operating voltage Vin, the engine is started 14. At each engine start, the inventive method is run through again.

The operation of the engine 2 is divided according to the invention in two phases of operation. Immediately after switching on the DC operating voltage Vin, the starting phase begins 21 of the motor. Therein, first in an optional step, the DC operating voltage Vin is compared with a predetermined threshold voltage V _TL 15. In the example, this threshold voltage V _TL is 6.4 VDC.

If the DC operating voltage Vin is smaller than the threshold voltage V _TL , the motor becomes 2 operated with a current control by pulse width modulation (PWM) with a PWM frequency F1, as described above 16. The PWM frequency F1 is for example 23 kHz.

However, if the DC operating voltage Vin is greater than or equal to the threshold voltage V _TL , the PWM frequency is reduced to a PWM frequency F2 17. In the example, the reduced PWM frequency F2 is about 19 kHz. Depending on the application, the PWM frequency F2 can also be selected smaller, for example 10 kHz.

The regulation of the current takes place via the duty cycle of the PWM and is independent of the PWM frequency. At a high operating DC voltage Vin, the duty cycle must be very small, whereby the turn-on time of the PWM clock is short. If the turn-on time were shorter than the switching time of the MOSFET switches, the motor could not start.

At a lower PWM frequency F2, the switch-on time is longer with the same duty cycle than with a higher PWM frequency F1. As a result, the duty cycle can be selected smaller than the higher PWM frequency. The current can still be regulated even at higher operating DC voltages Vin. The motor can be started safely with this method even at high operating DC voltages Vin.

As already explained above, high currents occur in particular during acceleration from standstill. As the speed increases, the motor current generally drops rapidly.

As the current decreases, so does the duty cycle of the PWM. As soon as the motor current has dropped so far that a current control by the PWM due to the larger duty cycle would also be possible again with the normal PWM frequency F1, can in the running phase 22 change. For this purpose, for example, it is checked after each commutation whether a transition condition is fulfilled 18, which is the transition to the running phase 22 signaled.

The transition condition for switching from the start phase to the normal mode or the running phase can be met, for example, if a predetermined speed is exceeded, a predetermined motor current is exceeded or a predetermined number of commutation steps has been passed. The transition condition, however, can also be given in particular by a reliable detection of the zero crossings of the BEMF voltage which is generated in the motor windings. For example, after successfully recognizing some, for example, between 4 and 20, consecutive expected zero crossings of the BEMF voltage may be switched to the running phase of the motor.

In the example, the motor is sensorless and the transient condition is satisfied when five consecutive zero crossings of the counter electromotive force induced BEMF voltage are detected. This is usually the case after about 20 commutation steps. On the other hand, if the transition condition is not met up to a tolerance time, there is probably an error, such as a stall, so that the engine can not start. In such a case, the control circuit stops the engine and outputs an error message.

Normal operation 19 in the running phase may include, for example, a current control with PWM as described above or a speed control. In particular, the PWM frequency during the running phase 22 F1 with F1> F2.

The 4 shows a development of the method of 3 which ensures safe starting in the desired direction of rotation even in such cases, in which the rotor of the fan is already moved by external forces or is in a position between two stator poles. The procedure is essentially the same as in 3 shown method. In addition, this method has a holding phase 20 on, which at every start before the start phase 21 is going through.

In the holding phase 20 becomes the rotor 5 moved to a defined starting position 23 and held there for a predefined holding time T _H. In the example, the hold time T _{H is} one second. However, it can be longer or shorter. For this purpose, two motor windings are subjected to a constant current. This causes the rotor to move 5 in a defined starting position. In the case of sensorless commutation, there is normally no position information at standstill due to the lack of rotor position sensors. Since the rotor is now at a known starting position, the commutation can begin at this starting position. As a result, starting the engine is much easier or only possible.

Holding the rotor in the starting position ensures that the rotor is stationary and not already rotating, especially possibly in the opposite direction of rotation. This always ensures that the engine starts in the right direction of rotation.

Furthermore, a forced commutation in the example follows a fixed commutation pattern in synchronous operation. Because the rotor position by the in the holding phase 20 If the starting position is known, the commutation pattern can start from exactly this starting position. This also ensures a safe starting of the rotor 5 favored in the right direction of rotation. Due to the commutation pattern, the subsequent commutation times are now known.

The commutation pattern may be, for example, in a table in the control circuit 12 be deposited. As already mentioned above, the transition condition is usually met after about 20 commutation steps. For safety, the table contains the commutation times of preferably 50 to 200 commutation steps. If, after completion of the complete commutation pattern table, the transition condition has not yet been met, there is probably an error and the motor is stopped and / or an error is signaled.

The commutation pattern provides a preferably linear acceleration of the rotor 5 why the commutation strokes are getting shorter quickly. Furthermore, it can be recognized that with increasing speed, the motor current drops rapidly. However, the commutation times of the commutation pattern may alternatively be selected according to another acceleration characteristic.

The transition from the starting phase 21 in the running phase 22 For example, after detecting, for example, five consecutive zero crossings, the BEMF voltage induced by the counterelectromotive force occurs. In the running phase, a speed control and / or a current control, which is why the motor current can also be kept almost constant here.

The 5 shows a flow diagram of a preferred embodiment of the method according to the invention, which is essentially the in 4 corresponds to the method shown. In the starting phase 21 However, if the DC operating voltage is above the threshold voltage, it is checked here, whether a start condition is met 24 , If yes, it will be during the launch phase 21 In the example, the start condition is met when the speed N of the rotor has exceeded a limit speed N _G. Since at higher speeds the current is usually lower, a shorter turn-on time of the PWM clock is no longer a problem and therefore possible. In the example, the limit speed N _G is 2800 U / min. In other applications, however, this limit speed may be higher or lower. After switching to the frequency F1 during the start phase, the transition condition to the transition to the running phase can again be interrogated, in particular the successful recognition of several successive, expected zero crossings of the BEMF voltage.

Of course, the additional starting condition can also without the holding phase 20 in the process of 3 to be available.

This process is summarized by way of example in the table below. Vin ≤ V _TL Vin> V _TL Vin> V _T N ≤ N _G N> N _G holding phase current control current control current control 150 mA 150 mA 150 mA PWM <100% PWM <100% PWM <100% start-up phase current control current control current control PWM <100% PWM <100% PWM <100% F1 = 23 kHz F2 = 19 kHz F1 = 23 kHz up phase current control current control current control PWM <100% PWM <100% PWM <100%

In some embodiments of the invention, it can also be provided during the running phase that the duty cycle of the PWM can also assume values of less than 100% for the case Vin> V _TL . As a result, in particular, a speed control during the running phase can be made.

LIST OF REFERENCE NUMBERS

Vin

Reverse Standoff

V _TL

threshold voltage

V _HL

upper threshold voltage

Q

MOSFET switch

AND MANY MORE

motor phases

ADC

Analog-to-digital converter input

BHN

Gate signals of the MOSFET switches

1

Fan

2

BLDC motor

3

stator

4

stator

5

rotor

6

wave

7

permanent magnet

8th

impeller

9

casing

10

circuit board

11

inverter

12

control circuit

13

measuring circuit

14-19

steps

20

holding phase

21

start-up phase

22

up phase

23-24

steps

Claims (15)

Method for operating an electric motor ( 2 ) at a speed (N) at a DC operating voltage (Vin) by means of a bridge circuit ( 11 ) with semiconductor switches (Q), characterized in that the method comprises a pulse width modulation (PWM), wherein the PWM is operated with a PWM frequency F2 ( 16 ), and wherein the PWM frequency is increased to a frequency F1 with F1> F2 ( 17 ) when the rotational speed (N) of the electric motor is greater than a predetermined limit rotational speed (N _G ). Method for operating an electric motor ( 2 ) at a speed (N) at a DC operating voltage (Vin) by means of a bridge circuit ( 11 ) with semiconductor switches (Q), characterized in that the method comprises a pulse width modulation (PWM), wherein the PWM is operated with a PWM frequency F1 ( 16 ), if the operating DC voltage (Vin) is less than a defined threshold voltage (V _TL ), and wherein the PWM frequency with a frequency F2 with F2 <F1 is operated ( 17 ), if the DC operating voltage (Vin) is greater than or equal to a defined threshold voltage (V _TL ). A method according to claim 2, characterized in that, if the DC operating voltage (Vin) is greater than or equal to a defined threshold voltage (V _TL ), the PWM is operated at the frequency F1, if the rotational speed (N) of the electric motor is greater than a predetermined limit Speed N _G is. Method according to one of the preceding claims, characterized in that after each switching on of the engine ( 2 ) are passed through at least two operating phases in succession, a start phase ( 21 ), in which the engine ( 2 ) is accelerated from standstill, and a running phase ( 22 ), which are in the starting phase ( 21 ), with the PWM frequency in the start phase ( 21 ) is operated according to one of the preceding claims and is operated in the running phase with the frequency F1. Method according to claim 4, characterized in that the transition from the starting phase ( 21 ) into the running phase ( 22 ) due to a given transition condition. A method according to claim 5, characterized in that the transition condition is met exactly when a predetermined speed (N _g ) is exceeded and / or when a predetermined motor current is exceeded and / or when a predetermined number of Zwangsskommutierungs steps has been passed through / and or when a predetermined number of consecutive zero crossings of the induced BEMF voltage have been detected. Method according to one of claims 4 to 6, characterized in that in the starting phase current regulation takes place by means of a pulse width modulation. Method according to one of claims 4 to 7, characterized in that in the starting phase ( 21 ) the commutation of the motor ( 2 ) is performed by a forced commutation after a fixed sequence of Kommutierungszeitpunkten. Method according to one of claims 4 to 8, characterized in that the fixed predetermined sequence of Kommutierungszeitpunkten is selected so that the engine ( 2 ) performs a linear acceleration. Method according to one of claims 4 to 9, characterized in that the commutation of the motor in the running phase is sensorless and that the transition condition for the transition from the Starting phase ( 21 ) into the running phase ( 22 ) is given by registering a predetermined number of consecutive zero crossings of the voltage induced by the counterelectromotive force. Method according to one of claims 4 to 10, characterized in that before the start phase ( 21 ) always a holding phase ( 20 ), in which the rotor ( 2 ) is moved to a defined start position and held there for a predefined hold time (T _H ) ( 23 ). Method according to claim 11, characterized in that the rotor ( 2 ) assumes the starting position by applying a constant current to at least two motor phases. Method according to claim 11, characterized in that the rotor ( 2 ) assumes the starting position by applying a constant voltage to two or more motor phases. Method according to claim 11, characterized in that the rotor ( 2 ) assumes the starting position by applying a predetermined PWM pattern to two or more motor phases. Method according to one of claims 1 to 14, characterized in that the electric motor ( 2 ) is a brushless DC motor.

Images