DE102016115056B3 - Method and device for preventing faulty commutations for a brushless motor - Google Patents

Abstract

The invention relates to a method and a device for preventing erroneous commutations within a method for fully automatic determination of the commutation of a brushless DC motor (M) during operation. An essential step is the integrating detection of the EMF for determining the magnetic flux at a motor connection (U) within one or more freewheeling intervals. This is followed by determining a first point in time (t1) after the start of one of these freewheeling intervals on the basis of the detected value of the magnetic flux by means of a first partial method which determines a first point in time (t1) in the relevant freewheeling interval which corresponds to an ideal commutation point in the freewheeling interval concerned would correspond to an ideal engine in an undisturbed operation. In addition, the determination of a second point in time (t2) in the freewheeling interval in question after the start of the relevant freewheeling interval is effected by a second partial method, which is different from the first partial method and which is designed such that the second point in time (t2) in the freewheeling interval concerned freewheeling interval in an undisturbed operation for an ideal engine at the same time before the first time (t1) must lie in the relevant freewheeling interval. If this sequence of times (t1, t2) is not adhered to, the commutation takes place at the second time (t2) as an emergency operation, otherwise at the first time (t1).

Description

introduction

The invention relates to a method for error-free control of the stator coils of a brushless motor with an excitation based on a permanent magnet located in the rotor. Typically, when determining the commutation time, there are incorrect determinations which can lead to considerable disturbances of the engine operation. The invention is concerned with the detection of a specific error situation in the determination of the commutation time and its elimination by a run-flat mode.

• • In einem ersten Zustand ist der obere Schalter geschlossen, wodurch der betreffende Motoranschluss mit der positiven Versorgungsspannung verbunden wird.
• • In einem zweiten Zustand ist der untere Schalter geschlossen wodurch der betreffende Motoranschluss mit der unteren Versorgungsspannung verbunden wird.
• • In einem dritten Zustand sind beide Schalter geöffnet, wodurch der zugeordnete Motoranschluss (U, V, W) nicht bestromt ist. Die betreffende Halbbrücke des Ansteuerblocks (St) ist passiv.
• • Das gleichzeitige Schließen beider Schalter ist als vierter, nicht zulässiger Zustand verriegelt, um Querströme durch den oberen und unteren Schalter in Form eines Kurzschlusses auszuschließen.

Such a brushless motor typically includes a rotor that itself contains a permanent magnet having at least two magnetic poles. In this case, the even number of magnetic poles should be distributed substantially symmetrically about the axis of rotation of the motor, which is only approximately the case in reality. The rotor is rotatably mounted in a magnetic field generated by stator coils outside the rotor. The magnetic field generated by the stator coil usually has only an approximate rotational symmetry about the rotor axis. This magnetic field is generated by the superposition of the fields of the several different stator coils, which are typically grouped symmetrically about the axis of rotation of said rotor, which in reality also only approximates. When suitably controlled, these stator coils produce a magnetic rotating field with a magnetic axis related to the rotation axis, which again only approximately corresponds to the mechanical axis of rotation of the rotor. This rotating magnetic rotating field then follows the rotor due to its permanent magnetization with its magnetic poles. When controlling the stator coils of such electric motors with rotors having a magnetic field excited by one or more permanent magnets, the BLDC motors already mentioned, a block commutation can be used. In order to generate a progressive magnetic field, preferably at least three stator coils are necessary. The control of the at least three stator coils is effected by at least three associated half-bridges, each containing an upper and a lower power switch, preferably each a power transistor. This control of the power transistors takes place as synchronously as possible relative to the angular position of the rotor relative to the position of the stator coils in order to maximize the efficiency. The rotor position can be detected according to the prior art by means of sensors or based on the electromotive force induced in the stator coils of the motor. Be on this 1 already referred to here, which schematically shows such a system according to the prior art. The stator coils of such a motor (M) preferably have three motor terminals, here designated U, V and W, on. Let U be the first motor connection, V the second motor connection and W the third motor connection. A drive block (St) sets the supply voltages at predetermined commutation intervals (Φ ₁ to Φ ₆ , 2 ) to the respective motor terminals (U, V, W) or leaves the motor terminals (U, V, W) in some of the predetermined commutation intervals (Φ ₁ to Φ ₆ , 2 ) even energized, whereby at the respective motor terminals (U, V, W) in the respective commutation intervals (Φ ₁ to Φ ₆ , 2 ) the electromotive force (EMF) causes a measurable voltage curve. As already described, each half bridge has an upper switch for connecting the respective motor terminal (U, V, W) to a positive supply voltage and a lower switch for connecting the respective motor terminal (U, V, W) to a negative supply voltage. This can not be in 1 drawn control, which is typically located within the drive block (St), connect the respective motor terminal (U, V, W) with the upper or lower supply voltage. The upper and lower switches can now be switched independently of each other. Only three of the four possible switching states of the upper and lower switches of a single half-bridge are permitted in each case.

• • In a first state, the upper switch is closed, which connects the relevant motor connection to the positive supply voltage.
• • In a second state, the bottom switch is closed, which connects the relevant motor connection to the lower supply voltage.
• • In a third state both switches are open, whereby the assigned motor connection (U, V, W) is not energized. The relevant half-bridge of the drive block (St) is passive.
• • Simultaneous closing of both switches is locked as a fourth, non-permissible state in order to exclude cross currents through the upper and lower switch in the form of a short circuit.

Due to the drive method, the block commutation, one of the three half bridges for driving the relevant motor connections is always in the passive state, which corresponds to the described third state, in which the relevant motor connection is not actively energized. During this third state, the course of the EMF as a phase voltage against a reference potential, for example ground, is visible at the relevant motor connection (U, V, W). This third state is due to block commutation always at one of the three motor connections (U, V, W) during each of the six cyclically repeated commutation intervals (Φ ₁ to Φ ₆ , 2 ) in front. The first state is also always at one of the three motor connections (U, V, W) during block commutation during each of the six cyclically repeated commutation intervals (Φ ₁ to Φ ₆ , 2 ) in front. In the case of block commutation, the second state is always at one of the three motor connections (U, V, W) during each of the six cyclically repeated commutation intervals (Φ ₁ to Φ ₆ , 2 ) in front. The phase voltage at the motor terminal (U, V, W), which varies in a commutation interval (Φ ₁ to Φ ₆ , 2 ) is just in the third high-impedance state of the sides of the driving half-bridge, as is known from the prior art, can be used for the position determination of the rotor. 2 shows the corresponding voltages at the three motor terminals (U, V, W) in the six cyclically repeated commutation intervals (Φ ₁ to Φ ₆ , 2 ).

The following table gives the states (states 1-3) of the half-bridges in the different commutation intervals (Φ ₁ to Φ ₆ , 2 ) in this example of the prior art again. U V W Φ ₁ 1 3 2 ₂ 1 2 3 Φ ₃ 3 2 1 ₄ 2 3 1 Φ ₅ 2 1 3 Φ ₆ 3 1 2

• • in dem ersten Kommutierungsintervall (Φ₁) am zweiten Motoranschluss (V) gemessen werden und
• • in dem zweiten Kommutierungsintervall (Φ₂) am dritten Motoranschluss (W) gemessen werden und
• • in dem dritten Kommutierungsintervall (Φ₃) am ersten Motoranschluss (U) gemessen werden und
• • in dem vierten Kommutierungsintervall (Φ₄) am zweiten Motoranschluss (V) gemessen werden und
• • in dem fünften Kommutierungsintervall (Φ₅) am dritten Motoranschluss (W) gemessen werden und
• • in dem sechsten Kommutierungsintervall (Φ₆) am ersten Motoranschluss (U) gemessen werden.

Thus, in the prior art, in an exemplary three-phase motor, the emf may be due to six different measurement constellations

• • be measured in the first commutation interval (Φ ₁ ) at the second motor connection (V) and
• • be measured in the second commutation interval (Φ ₂ ) at the third motor connection (W) and
• • be measured in the third commutation interval (Φ ₃ ) at the first motor connection (U) and
• • be measured in the fourth commutation interval (Φ ₄ ) at the second motor connection (V) and
• • be measured in the fifth commutation interval (Φ ₅ ) at the third motor connection (W) and
• • be measured in the sixth commutation interval (Φ ₆ ) at the first motor connection (U).

Frequently, during this measurement, the so-called zero crossing of the EMF is used, in which the latter changes its sign relative to a reference potential. The internal time base for carrying out the commutation is regulated in such a way that this zero crossing takes place exactly in the middle of the commutation interval at that motor connection (U, V, W) which is currently in the third state.

It should be briefly mentioned here that the voltage at the motor connection (U, V, W) is also referred to as phase voltage.

Alternatively, in the third state, the course of the EMF itself, that is to say the course of the voltage at the motor connection (U, V, W), can be used for the determination of the commutation time. Since the speed of the rotor affects only the amplitude of the EMF, but this is otherwise a function of the course of the magnetic flux over the angular position, the integral of the EMF over the time from the zero crossing to the following commutation is an engine constant by specifying an upper Boundary for the integral can be reversed so a commutation with a fixed angular distance to the zero crossing directly, d. H. without the detour over a time base.

This measurement of the EMF is carried out by an EMF evaluation device (EMKA), which in 1 is drawn. This measures in the six commutation intervals (Φ ₁ to Φ ₆ , 2 ) corresponding to the respective commutation interval (Φ ₁ to Φ ₆ , 2 ) associated Meßkonstellation the EMF and generates for each of the motor terminals (U, V, W) each have a separate commutation signal (A ₁ , A ₂ , A ₃ ). The commutation signals (A ₁ , A ₂ , A ₃ ) are in 2 roughly drawn. Likewise, the rotor position is plotted as a parameter of the X-axis.

With each edge change on a commutation signal (A ₁ , A ₂ , A ₃ ), a control logic within the drive block (St) changes state. Alternatively, it is possible accordingly 2 derive the commutation interval (Φ ₁ to Φ ₆ ) by a static logic from the commutation signals (A ₁ , A ₂ , A ₃ ). For example, the first phase can be represented as the AND connection of the negated first commutation signal (A ₁ ) with the negated second commutation signal (A ₂ ) and the third commutation signal (A ₃ ). Analogously, the other phases can be determined. However, it has been shown that preferably between the commutation intervals (Φ ₁ to Φ ₆ ) briefly and asynchronously after each commutation interval (Φ _i with 0 <i <7) a commutation intermediate interval (Φ _i with 0 <i <7) associated with this commutation interval ( Φ _i 'with 0 <i <7) is inserted, in which the switches of the half-bridges, which change their circuit state, are turned off to safely exclude cross-currents. In this respect, it makes sense if the total number of actual states of a finite state machine in the control block (St) is twelve instead of six.

3 shows an exemplary sub-device of the prior art for determining the first commutation signal (A ₁ ) for the commutation at the first motor terminal (U). This is part of the EMK evaluation device (EMKA). The focus in the following is focused on this first branch in the description. The person skilled in the art, however, will find it easy to write the writing on the two corresponding branches for the second motor connection (V) with the associated second commutation signal (A ₂ ) and for the third motor connection (W) with the associated third commutation signal (A ₃ ). which are arranged in parallel to transfer. In a first stage, a virtual star point signal (SpS) is generated by means of a star connection of three voltage dividers (SpT ₁ , SpT ₂ , SpT ₃ ) from the three voltages of the three motor connections (U, V, W). This represents the mean value of the voltages at the three motor terminals (U, V, W) and is therefore subtracted from the voltage at the first motor terminal (U) by means of a second summer (SU _2U ). The corrected voltage signal (U _korr ) of the first motor terminal (U) thus obtained is the difference between the phase voltage U and the mean value of the three phase voltages and is integrated in a first integrator (Int _1U ). Possibly. The integrator can also be designed as a filter. The first threshold signal is obtained (S _1U ). A comparator, more precisely a first comparator (CMP _1U ), compares this first threshold _signal (S _1U ) with a first default value (V _refU ) for the commutation and generates therefrom the first commutation signal (A ₁ ) which, as described in the said control _block (A _1U ). St) for the correct angle commutation of the first half-bridge, which is the first motor terminal (U) energized used.

For example, from the DE 10 2008 025 442 A1 , of the US 2007 0 282 461 A1 and the US 2014 0 062 364 A1 Devices are known which use the electromotive force (EMF) in the form of the phase voltages during the freewheeling intervals during commutation and use these for the regulation of the commutation itself. These technical teachings do not use the integral of the EMF and thus not a value curve of the magnetic flux. Such methods have the disadvantage of speed dependency.

From the DE 100 54 594 A1 For example, a system for optimal commutation is known. In the 3 DE 100 54 594 A1 three control branches are drawn, which essentially the 3 correspond to this disclosure. The first default value (V _refU ), the second default value (V _refV ) and the third default value (V _refW ), which may possibly coincide, are not shown, but for the skilled person by the term "comparator" in DE 100 54 594 A1 implicitly recognizable.

From the DE 100 54 594 A1 is known, the neutral point voltage (reference V _AN of DE 100 54 594 A1) by means of a resistor star circuit (see 5 DE 100 54 594 A1), then to integrate and with a filtered signal that the current level (reference IA of DE 100 54 594 A1 in 5 DE 100 54 594 A1) in one of the motor phases (reference PHASE A of DE 100 54 594 A1) to combine by subtraction and from the thus obtained intermediate signal by means of two comparison units a plurality of control signals (reference numerals S1, S2, S3 of DE 100 54 594 A1, see 5 DE 100 54 594 A1) and in a position signal calculator (see 5 DE 100 54 594 A1) to a signal (reference numeral S4 of DE 100 54 594 A1) to combine. The regulation of DE 100 54 594 A1 is thus based on the detected zero crossing of the phase current of a phase (here the phase A of DE 100 54 594 A1 in whose 5 ) in combination with the star voltage of the three phases (reference VA of DE 100 54 594 A1). In the technique of DE 100 54 594 A1, a shunt resistor is necessary for this purpose. The technical teaching disclosed in DE 100 54 594 A1 is concerned with the determination of a single rotor position within each of the temporal commutation intervals of the activation of a BLDC motor comprising 60 ° angles of rotation. In real operation, such an engine is susceptible to interference. Therefore, when using the technique disclosed in DE 100 54 594 A1, disturbances can occur due to incorrectly detected rotor positions too early, ie rotor angles that are too small and thus too early commutations.

Also from the DE 100 64 486 A1 For example, the determination of values at various times is known for the control of a BLDC motor. Again, there is no detection of a faulty, too early commutation and thus no emergency operation.

From the DE 10 2009 019 414 A1 a drive circuit is known which can detect the complete failure of individual phase circuits. An isolated disturbance of individual rotor position measurements, for example due to electromagnetic radiation, with the consequence of commutation times determined too early or with the consequence of additional commutation times within a commutation interval is not detected or compensated.

From the US 2008 0 252 240 A1 a device is known which determines an optimized commutation time by evaluating the EMF. 5 US 2008 0 252 240 A1 shows in the upper part the time profile of a phase voltage (reference symbol Vu of US 2008 0 252 240 A1). This phase voltage (reference symbol Vu of US 2008 0 252 240 A1) always shows the EMF (reference symbol Vu 'of US 2008 0 252 240 A1) if the associated driver has a high impedance. In the lower part of the 3 US 2008 0 252 240 A1, the energization of the other two phases is shown in time, the energization of the phase under consideration always takes place when the signal (reference Spwm of US 2008 0 252 240 A1) is at the lower value. If the considered phase has a high impedance, the other two phases are energized and the EMF can be measured at the considered phase.

The EMF is therefore always measurable at the considered phase voltage (reference symbol Vu of US 2008 0 252 240 A1) only when the other two phases are energized.

In a device or a method according to US 2008 0 252 240 A1, as described above, the EMF is then searched for the zero crossing in order to determine therefrom the optimum commutation time. This determination of the commutation time occurs, if it is not in time in a high-impedance period of the driver, by linear interpolation from a first time by a suitable time delay. The revelation of US 2008 0 252 240 A1 describes the determination of this time delay.

When the zero crossing is in the non-energized period, to which the phase voltage (reference symbol Vu of US 2008 0 252 240 A1 ) not on the EMF line (reference Vu 'of US 2008 0 252 240 A1 in 3 US 2008 0 252 240 A1), the zero crossing can only be estimated. This situation shows 5 US 2008 0 252 240 A1. This is done in the technique of US 2008 0 252 240 A1 as follows:
According to US 2008 0 252 240 A1, a measurement of the phase voltage of interest is always taken at observable times (reference symbol Vu of US 2008 0 252 240 A1). This results in a step-shaped signal (reference numeral Sdiff of US 2008 0 252 240 A1 in 5 US 2008 0 252 240 A1). At the same time, the temporal slope of the phase voltage (reference symbol Vu of US 2008 0 252 240 A1) is detected at these measurement times on the basis of two past measurements and a sawtooth curve (reference Sramp of US 2008 0 252 240 A1) is generated with this slope. As a result, the zero crossing can be linearly interpolated into the period of time by simple inverse addition of these signals (reference symbols Sdiff and Sramp of US 2008 0 252 240 A1), to which the EMF at the phase connection of interest can not be measured, since it is energized.

• 1. Das Verfahren der US 2008 0 252 240 A1 ist störanfällig, weil Störungen auf der EMK, also der interessierenden Phasenspannung (Bezugszeichen Vu der US 2008 0 252 240 A1) zu Ungenauigkeiten in der Berechnung des Nulldurchgangs führen. Damit ist das Verfahren der US 2008 0 252 240 A1 insbesondere bei niedrigen Drehzahlen mit geringer EMK besonders anfällig gegen Störungen, wie beispielsweise EMV.
• 2. Der eigentliche Kommutierungszeitpunkt muss durch Addition einer linear interpolierten Zeitspanne ermittelt werden. Die Linearität beruht auf der Bildung der ersten zeitlichen Ableitung in Form des Sägezahnsignals (Bezugszeichen Sramp der US 2008 0 252 240 A1). Diese lineare Interpolation setzt einen nicht beschleunigten Motor voraus. Somit funktioniert das Verfahren der US 2008 0 252 240 A1 nur bei geringen Beschleunigungen. Eine solche Regelung, wie die der US 2008 0 252 240 A1, ist daher nicht geeignet, hohe Lastdynamiken abzufangen.
• 3. Im Falle von Störungen werden auf Grund der Störungen zeitlich zu früh in einem betreffenden Kommutierungsintervall ermittelte Kommutierungszeitpunkte nicht als solche erkannt. Auch sind mehr als zwei Kommutierungen in einem Intervall aufgrund von Störungen möglich, was den endlichen Automaten der Kommutierungssteuerung empfindlich stören kann.

It is therefore a method that directly evaluates the zero crossing of the emf and determines the zero crossing by temporal linear interpolation. This has the following disadvantages.

• 1. The procedure of US 2008 0 252 240 A1 is susceptible to interference because disturbances on the EMF, ie the phase voltage of interest (reference symbol Vu of US 2008 0 252 240 A1) lead to inaccuracies in the calculation of the zero crossing. Thus, the method of US 2008 0252240 A1 is particularly susceptible to disturbances, such as EMC, especially at low speeds with low EMF.
• 2. The actual commutation time must be determined by adding a linearly interpolated time span. The linearity is based on the formation of the first time derivative in the form of the sawtooth signal (reference Sramp of US 2008 0 252 240 A1). This linear interpolation requires a non-accelerated motor. Thus, the method of US 2008 0 252 240 A1 only works at low accelerations. Such a regulation, such as US 2008 0 252 240 A1, is therefore not suitable for absorbing high load dynamics.
• 3. In the case of disturbances, commutation times determined too early in a respective commutation interval are not recognized as such due to the disturbances. Also, more than two commutations in an interval due to disturbances are possible, which can disturb the finite state machine commutation sensitive.

4 shows according to the prior art, the voltage at the first motor terminal (U) with an optimal commutation for a low, a medium and a high angular velocity. That the first motor terminal (U) associated first threshold signal (S _1U ), according to the preceding description and the 3 is generated, increases substantially square until it is equal to the first default value (V _refU ) for the commutation. The unshown, first commutation signal (A ₁ ) and the half-bridge associated with the first motor connection (U) are then switched. It is completely irrelevant whether the angular velocity is high or low.

5 shows the voltage at the first motor terminal (U) according to the prior art at too early commutation, too late commutation and an optimal commutation. The first threshold signal (S _1U ) associated with the first motor terminal (U) increases again substantially square until it is equal to the first default value (V _refU ) for the commutation. The unshown first commutation signal (A ₁ ) and the half-bridge associated with the first motor connection (U) then commutate. This happens in the left part of the figure 5 too early, in the middle part of the figure too late and in the right part of the figure at an optimal time. It should be clear that a position of the zero crossing in the middle of the respective Kommutierungsintervalls is optimal.

In addition to the branch described so far within the EMF evaluation (EMKA) for the first motor connection (U), there is typically a second branch for the second motor connection (V) with associated individual elements (SU_2V, INT_1V, CMP_1V) and signals (V_corr, P_1V , V_RefV) and a third branch for the third motor terminal (W) with associated individual elements (SU_2W, INT_1W, CMP_1W) and signals (W_corr, P_1W, V_refW).

In order to achieve a commutation for the same angular position of the rotor with different motors, the respective default values (V _refU , V _refV , V _refW ) of the respective motor connections (UVW) must be adapted in each case. Since the course of the magnetic flux in relation to the angular position of the rotor usually not known and can not be determined from the data sheet of the engine, the respective default values (Vref _U , Vref _V , Vref _W ) in the prior art must first be determined experimentally , The respective default values (V _refU , V _refV , V _refW ) must be varied during operation until the commutation is carried out at the desired time. Typically, however, the default values are equal (V _refU = V _refV = V _refW = V _ref ). It is clear to the person skilled in the art that the device described above in space division multiplexing can also be used in time division multiplexing, ie that only one branch in the EMF evaluation (EMKA) has to be realized if the values of a motor connection (U, V, W) in the commutation intervals (Φ ₁ to Φ ₆ , 2 ) in which the associated half-bridge is in the states one or two, can be buffered and the values can be loaded into the corresponding branch associated with the motor terminal (U, V, W) which is in the commutation intervals (Φ ₁ to Φ ₆ , 2 ) has an associated half-bridge, which in the relevant commutation interval (Φ ₁ to Φ ₆ , 2 ) is currently in the third state. Therefore, it makes sense if parts of the device are realized in their function by means of a microcontroller or signal processor or other computer. In this respect, the various elements described above can also be combined into one or a few elements.

Object of the invention

It is the object of the invention to enable fault detection in the automatic determination of the commutation time for the respective motor connections (U, V, W) and emergency running in the event of a fault in the determination of the commutation time. In this case, the defects recognized in the prior art, in particular those of the DE 100 54 594 A1 , be avoided.

This object is achieved by a method according to claim 1 and an apparatus according to claim 5.

Description of the invention

The fully automatic determination of the commutation time (S _2U ) according to the invention is carried out by means of integration of the emf in the freewheeling intervals. From the prior art, it is known to the person skilled in the art that this is a method based on the magnetic flux, which therefore is independent of the rotational speed and thus also independent of the acceleration.

The invention thus relates to a method for preventing erroneous commutation and / or Kommutierungszeitpunkte within a method for fully automatic sensorless determination of the commutation of said brushless DC motor. In contrast to the revelation of DE 100 54 594 A1 Here, the temporal sequence of several timing signals is evaluated, the be produced according to the invention in different ways for redundancy purposes. As a result, in contrast to DE 100 54 594 A1, an additional robustness is created. The method according to the invention therefore comprises the commutation and energization of the at least three phases (U, V, W) in commutation intervals (Φ ₁ to Φ ₆ ) by a drive block (St), the energization of the first phase (U) in a subset of the commutation intervals ( Φ ₁ , Φ ₂ , Φ ₄ , Φ ₅ ) by the drive block (St) and the non-energizing of the first phase (U) in the other Kommutierungsintervallen (Φ ₃ , Φ ₆ ) by the drive block (St). This is followed by integrally detecting the EMF to determine the magnetic flux at the first phase (U) within one or more freewheeling intervals that are one or more of the commutation intervals (Φ ₃ , Φ ₆ ) in which the first motor terminal (U) passes through Activation block (St) is not energized, as well as generating a virtual neutral point signal (SpS) from the electrical potential curves at least three phases (U, V, W) by a first sub-device (SpT ₁ , SpT ₂ , SpT ₃ , SU _1U ) and the Formation of a corrected voltage signal (U _korr ) as the difference between the electric potential profile of the first phase (U) and the virtual neutral point signal (SpS). Like in the DE 100 54 594 the integration of the corrected voltage signal (U _korr ) for detecting the magnetic flux during said freewheeling _intervals follows a fourth threshold signal (S _2U ) through a fourth integrator (Int _2U ). In contrast to DE 100 54 594 A1, however, additionally and according to the invention, the determination of a first point in time (t ₁ ) after the start of the freewheeling interval in question is based on the determined flow by means of a first partial method which determines a first point in time (t ₁ ) Ideal commutation in an ideal engine would correspond in an undisturbed operation and at the same time - and this is the difference according to the invention to DE 100 54 594 A1 determining a second time (t ₂ ), due to a technically deviant, second determination sub-process time after the beginning of concerned freewheeling interval. The second determination method for determining this second time (t ₂ ) is designed so that in an undisturbed operation in an ideal engine this second time (t ₂ ) should be before the first time (t ₁ ). That this is possible was inventively recognized and used according to the invention for fault detection. At the same time it is proposed in this document to use an emergency operation by using the second time (t ₂ ) instead of the first time (t ₁ ) for the commutation. A further step of the method according to the invention is therefore the suppression of a commutation in a real operating case for the first determined time (t ₁ ), if the second determined time (t ₂ ) is later than the first determined time (t ₁ ) and the substitute commutation to second second point (t ₂ ) in this case. If there is no error, according to the invention, the commutation takes place in a real operating case at the first determined time (t ₁ ), when the second determined time (t ₂ ) is earlier than the first determined time (t ₁ ). The first time (t ₁ ) is thus the regular commutation time and the second time (t ₂ ) is the Ersatzkommutierungszeitpunkt for emergency.

The aforementioned subdevices thus carry out the following steps, as already described above: To determine the first instant (t ₁ ), the fourth threshold signal S _2U is compared with a reference potential 0 and / or a reference variable 0 by means of a first comparator (CMP _1U ) and the generation of a first commutation event signal (A ₁ '), in particular as a trigger signal (TS) for the commutation, which marks the first time (t ₁ ). This is done depending on the result of this comparison.

In order to determine the second instant (t ₂ ), the fourth threshold signal (S _2U ) is compared with a reference _potential (V _Ref2 ) and / or with a reference value (V _Ref2 ) by means of a fourth comparator (CMP _2U ) for generating a first set signal (S _2U ). S _U ), which marks the second time (t ₂ ). This is done depending on the result of the comparison by the fourth comparator (CMP _2U ).

The error detection and correction described above is done by setting the first memory output (EN _U ) of a first memory (RSFF _U ) in response to the first set signal (S _U ) to a status with the property "enabled" and generating a commutation signal (A ₁ ) for commutation by a logical operation (AND _U ) when the first memory output (EN _U ) is in a status with the property "enabled" and when the value of the fourth threshold signal (S _2U ) is the value of the reference signal (V _Ref2 ) and / or crosses in terms of value as well as the resetting of the first memory output (EN _U ) of the first memory (RSFF _U ) with or after the signaling of a commutation by the commutation signal (A ₁ ) to a second status, the property "enabled" not having. In contrast to DE 100 54 594 A1 Thus, a rotor position is not determined, but according to the invention instead of in dependence on a time sequence of the first time (t ₁ ) and the second time (t ₂ ) determines the commutation. However, this time sequence is not evaluated in DE 100 54 594 A1, which is why a device according to DE 100 54 594 A1 is unable to detect and correct errors.

In a variant of the method, the generation of a first sign signal (Sig _U ) by means of a first sign unit (Sgn _U ) from the corrected voltage signal (U _korr ), which indicates the sign of the corrected voltage signal (U _corr ) replaces the subsequent integration of this first Sign signal (Sig _U ) for integrating detection of the EMF in the freewheeling intervals to a fourth threshold signal (S _2U ) by a fourth integrator (Int _2U ) as a whole on one side of the previously described integration of the corrected voltage signal (U _corr ) for detecting the magnetic Flow during said freewheeling intervals to a fourth threshold signal (S _2U ) by a fourth integrator (Int _2U ) on the other side.

This method can of course be implemented in a suitable general device for preventing erroneous commutation and / or Kommutierungszeitpunkte within a device for commutation of a brushless DC motor (M) to a fully automatic sensorless certain commutation. This device comprises first a drive block (St), which commutes the at least three phases (U, V, W) in Kommutierungsintervallen (Φ ₁ to Φ ₆ ) and energized or de-energized, where he assigned the first phase (U) in Commutation (Φ ₁ , Φ ₂ , Φ ₄ , Φ ₅ ) energized and in other associated commutation (Φ ₃ , Φ ₆ ), the freewheeling intervals, not energized, and second, a first sub-device (SpT ₁ , SpT ₂ , SpT ₃ , SU _1U ), which generates a virtual neutral point signal (SpS) from the electrical potential curves of the at least three phases (U, V, W). Furthermore, the device according to the invention comprises a second sub-device (SU _2U ), which generates a corrected voltage signal (U _korr ) as a difference between the electric potential profile of the first phase (U) and the virtual star point signal (SpS), and a third sub-device for determining the Value of the level of magnetic flux during said freewheeling intervals on the basis of the corrected voltage signal (U _korr ) in the form of a fourth threshold signal (S _2U ). A fourth sub-device is also part of the device. This fourth sub-device is used to determine a first point in time in the form of an edge of a first commutation event signal (A ₁ '), after the start of one of these free-wheeling intervals within the relevant free-wheeling interval on the basis of the determined value of the level of the magnetic flux by means of a sub-method performed by the fourth sub-device, which determines a first instant (t ₁ ) that would correspond to an ideal commutation instant in an ideal motor in undisturbed operation. A fifth sub-device is used to determine a second time (t ₂ ) in the form of an edge of a set signal (S _U ), which is within the relevant free-wheeling interval after the start of the respective free-wheeling interval by a second, performed by the fifth sub-device sub-process. The second partial method is designed so that the second time (t ₂ ) must lie in an undisturbed operation in an ideal engine at the same time before the first time (t ₁ ). A sixth sub-device therefore suppresses according to the invention a commutation based on the first commutation event signal (A ₁ ') in a real operating case when the second determined time (t ₂ ) is later than the first determined time (t ₁ ) and performs a substitute commutation to the second Second point (t ₂ ) in this case for the emergency by the invention. This sixth sub-device commutates the control of the motor at the first determined time (t ₁ ) in the form of an edge of the first Kommutierungsereignissignals (A ₁ '), when the second determined second time (t ₂ ) in the form of an edge of the set signal (SU), in time before the first determined time (t ₁ ). This ensures that not plausible disturbances unlike DE 100 54 594 A1 can be detected and intercepted sensibly.

To carry out the sub-method according to the invention, the device according to the invention comprises further sub-devices in special forms. By way of example, the third subdevice can be realized by means of a fourth integrator (Int _2U ), which integrates and / or filters the corrected voltage signal (U _korr ) during said freewheeling intervals into a fourth threshold signal (S _2U ). The fourth sub-device can be compared by a first comparator (CMP _1U ), which compares the fourth threshold signal (S _2U ) with a reference potential (0) and / or a reference variable (0), and a first commutation event signal (A ₁ '), in particular as trigger signal. Signal (TS), for the commutation, depending on the result of this comparison generated realized. The fifth sub-device can be realized by a fourth comparator (CMP _2U ), which compares the fourth threshold signal (S _2U ) with a reference _potential (V _Ref2 ) and / or a reference value (V _Ref2 ) and generates a set signal (S _U ). The sixth subdevice preferably comprises a first memory (RSFF _U ) with a first memory output (EN _U ), which by means of the first set signal (S _U ) by the fourth comparator (CMP _2U ) depending on the result of this comparison to a status with the Property "enabled" sets or does not set, and wherein the first memory output (EN _U ) with or after the signaling of a commutation by the commutation signal (A ₁ ) to a second status that does not have the property "enabled"; is reset. A logic operation device within the sixth sub-device, in particular an AND gate (AND _U ), which performs a logic operation between the first memory output (EN _U ) of the first memory (RSFF _U ) and the first commutation event signal (A ₁ ') and the commutation signal (A ₁ ) as a result of this link is generated when the first memory output (EN _U ) is in a status with the property "enabled" and when the fourth threshold signal (S _2U ) reaches and / or crosses the reference potential (0) and / or reference (0) also part of the sixth sub-device.

In the context of the development of the invention, as already described, it was recognized that the processing of the sign can be sufficient. The device according to the invention therefore comprises in a specific embodiment a first sign unit (Sgn _U ) which generates a first sign signal (Sig _U ) from the corrected voltage signal (U _korr ) which indicates the sign of the corrected voltage signal (U _korr ). In this embodiment of the invention, however, the device according to the invention comprises a third sub-device for determining the value of the magnetic flux on the basis of the additional first sign signal (Sig _U ) instead of as previously described on the basis of the corrected voltage signal (U _korr );

Description of the figures

1 shows a schematic interconnection of a drive device of the prior art and has already been described in the introduction as belonging to the prior art.

2 shows the exemplary three commutation signals (A ₁ , A ₂ , A ₃ ) and the associated voltage waveforms at the motor terminals (U, V, W) schematically for several commutation intervals (Φ ₁ to Φ ₆ ) and was already in the introduction as the State of the art associated described.

3 shows an exemplary branch within the EMF evaluation (EMKA) for the first motor terminal (U) for generating the first commutation signal (A ₁ ) associated with the first motor terminal (U) and has already been described in the introduction as belonging to the prior art.

4 shows the quadratic rise of the first threshold signal (S _1U ) and the voltage at the first motor terminal (U), when the associated half bridge of the drive block (St) is in the high impedance third state, for different angular velocities and was already in the introduction than the state of Technique associated described.

5 shows the voltage at the first motor terminal (U), when the associated half-bridge of the drive block (St) is in the high-impedance third state, for various values of the first preset value (V _u ) according to the invention, and the voltage waveform of the first threshold signal (S _1U ) and already described in the introduction as belonging to the prior art.

6 shows a realization of the integration method in which an additional integrator (Int ₂ ) has been added. The drawing does not represent a solution to the problem. The device according to the invention is more extensive.

7 shows the waveform for the corrected voltage signal (U _korr ), the first threshold signal (S _1U ) and the fourth threshold signal (S _2U ) for typical operating _cases , the potential of the reference signal (V _ref2 ) and the first time (t ₁ ) to which Commutation is to take place and the second time (t ₂ ) at which the commutation for the remainder of the respective Kommutierungsintervalls (Φ ₁ to Φ ₆ ) is allowed.

8th shows a device according to the invention for the first motor connection (U).

9 shows a device according to the invention accordingly 8th for the first motor terminal (U) with the difference that only the sign (Sig _U ) of the corrected voltage signal (U _korr ) of the first motor terminal (U) is integrated.

10 shows 1 with the three branches (ZW ₁ , ZW ₂ , ZW ₃ ).

11 shows a device according to the invention for the second motor connection (V).

12 shows a device according to the invention accordingly 11 for the second motor terminal (V) with the difference that only the sign (Sig _V ) of the corrected voltage signal (V _korr ) of the second motor terminal (V) is integrated.

13 shows a device according to the invention for the third motor connection (W).

14 shows a device according to the invention accordingly 13 for the third motor terminal (W) with the difference that only the sign (Sig _W ) of the corrected voltage signal (W _korr ) of the third motor terminal (W) is integrated.

In the following the invention with reference to the figures from 5 including, which do not correspond to the prior art, explained in more detail.

6 shows an exemplary sub-device according to the invention for determining the first commutation signal (A ₁ ) for the commutation at the first motor terminal (U). It concerns a first branch (ZW ₁ ) of the EMK evaluation (EMKA) of the 1 and 10 , In this respect, the in 3 described sub-device by this new sub-device according to the invention in at least one branch of the EMF evaluation (EMKA), which typically contains such a branch for each motor terminal (U, V, W), replaced. In a first stage of the subdevice according to the invention, a virtual neutral point signal (SpS) is again generated by means of the already known and unchanged star connection of three voltage dividers (SpT ₁ , SpT ₂ , SpT ₃ ) from the three voltages of the three motor connections (U, V, W) , This virtual neutral point voltage (SpS) is again, as before, deducted from the voltage at the first motor terminal (U) by means of the known second summer (SU _2U ). The thus obtained corrected voltage signal (U _korr ) of the first motor terminal (U) is integrated in a first integrator (Int _1U ) back to a first threshold signal (S _1U ). Possibly. In a specific embodiment of the invention, the first integrator (Int _1U ) can also be designed here as a filter. In this case, the first integrator (Int _1U ) preferably integrates only positive values of the corrected voltage signal (U _korr ) of the first motor connection (U) in the commutation intervals (Φ ₁ to Φ ₆ ), in which the associated half-bridge of the first motor connection (U) is high-impedance , In the example described here, these are the third commutation interval (Φ ₃ ) and the sixth commutation interval (Φ ₆ ). The first integrator (Int _1U ) is typically reset immediately before or at the beginning of such a commutation interval (Φ ₃ , Φ ₆ ), for example, by the drive block (St) or other control. For example, in order to integrate only the positive values of the corrected voltage signal (U _korr ) of the first motor terminal (U), there are two possibilities: Firstly, with the aid of a first limiter (B _U ), only positive signal components of the corrected voltage signal (U _corr ) of the first motor terminal (U) to the first integrator (Int _1U ) as a limited, corrected voltage signal (U _korr ') of the first motor terminal (U) and otherwise the first integrator (Int _1U ) by the first limiter (B _U ) a zero value to be delivered. This is in 6 shown.

Secondly, it is possible with the aid of a fourth integrator (Int _2U ) to integrate the corrected voltage signal (U _korr ) of the first motor terminal (U) into a fourth threshold signal (S _2U ).

Otherwise, that's right 6 with the 3 match.

Experiments have shown that the use of the first limiter (B _U ) is particularly preferable.

7 shows the curves of the corrected voltage signal (U _korr ), the first threshold signal (S _1U ) at the output of the first integrator (Int _1U ), and the fourth threshold signal (S _2U ) at the output of the fourth integrator (Int _2U ) 6 for a too early commutation (a), too late commutation (b) and a correct commutation (c) of the connected motor (M).

According to the invention, it has been recognized that, with correct commutation, the fourth threshold signal (S _2U ) at the output of the fourth integrator (Int _2U ) reaches its zero crossing at a first instant (t ₁ ), which in the ideal case of ideal operation and ideal motor corresponds to the optimal commutation time , It has also been recognized according to the invention that it makes sense to use this zero crossing of the fourth threshold signal (S _2U ) at the first time (t ₁ ) at the output of the fourth integrator (Int _2U ) also for the suppressed commutation of the motor (M). In order to be able to detect a fault, a second time (t ₂ ) is determined by means of the voltage of a reference signal (V _Ref2 ) in this example, always before the first time (t ₁ ) within the relevant commutation interval (Φ ₁ to Φ ₆ ) and after its beginning in time. According to the invention, this second time (t ₂ ) is used to release the commutation at the first time (t ₁ ) determined. For this purpose, it is detected when the voltage profile of the threshold signal (S _2U ) crosses the voltage of the reference signal (V _Ref2 ). This time marks the said second time (t ₂ ). The crossing of the two voltage curves is recorded for the remainder of the respective commutation interval in a memory (RSFF _U ), which allows the commutation at a first time (t ₁ ) for the remainder of the respective commutation or until the first commutation within the relevant commutation. Only at the end the commutation interval or better with the commutation of this memory is reset and a commutation by further, erroneously determined as a result of faults first time points (t ₁ ) is reliably suppressed for the remainder of the relevant Kommutierungsintervalls. In the event of an error, when the first time (t ₁ ) takes place before the second time (t ₂ ), the commutation is triggered only by setting the memory at the second time (t ₂ ).

One possible realization is in 8th shown. It is true until the output of the fourth integrator (Int _2U ), the fourth threshold signal (S _2U ), with the lower branch of 6 match. A first comparator (CMP _1U ) compares this fourth threshold signal (S _2U ) with zero, ie typically the reference potential, and thus generates the first integrator (Int _2U ) at the first instant (t ₁ ) of the zero crossing of the fourth threshold signal (S _2U ) first commutation event signal (A1 '). In addition, a device according to the invention also has a further branch which _comprises a fourth comparator (CMP _2U ) which compares the fourth threshold signal (S _2U ) with a reference signal (V _Ref2 ), a first memory (RSFF _U ), whose first memory output ( EN _U ) is set by the fourth comparator (CMP _2U ) at the second time (t ₂ ) and reset by an outer first reset signal (R _U ), which typically correlates to the end of the respective free-wheeling interval or, preferably, the commutation time, as well as one first AND operation (AND _U ). This branch has the following function: Off 7 It can be seen that the fourth threshold signal (S _2U ) of the fourth integrator (Int _2U ) is close to zero at the beginning of the commutation interval. In this period immediately at the beginning of the commutation, it might be possible that superimposed disturbances lead to undesired commutation. In order to prevent this, according to the invention, the fourth comparator (CMP _2U ) is inserted. It switches if, at a second time (t ₂ ), the fourth threshold signal (S _2U ) of the fourth integrator (Int _2U ) has exceeded the threshold in the form of the reference signal (V _{Ref 2} ). The reference signal (V _Ref2 ) serves as a fault threshold. All signals below this threshold are considered disturbances. Thus, the reference signal (V _Ref2 ) can be set as a rough global threshold and does not have to be adapted to different applications or motors. If the threshold has been exceeded in the form of said reference signal (V _Ref2 ), the first memory (RSFF _U ) is set by the first set signal (S _U ) generated by the fourth comparator (CMP _2U ) at the second time (t ₂ ) , In conjunction with the first AND operation (AND _U ), the commutation is released at this second instant (t ₂ ), so that from then on, a signal change of the first comparator (CMP _1U ) at its output, the commutation event signal (A ₁ '), would lead to a commutation in the form of a first commutation signal (A ₁ ) at a first time (t ₁ ). Preferably, time-synchronized with this commutation in the form of a first commutation signal (A ₁ ), the first memory (RSFF _U ) via the first reset signal (R _U ) is reset. The first reset signal (R _U ) thus typically depends inter alia on the first commutation signal (A ₁ ).

Another preferred realization of the device is in 9 shown. This realization does not integrate via the reflected EMF, but over their sign. In the other function, this embodiment is identical to that of 8th , The device of 9 is thus distinguished from the in 8th in that the fourth integrator (Int _2U ) does not integrate the corrected voltage signal (U _korr ) to the fourth threshold signal (S _2U ), but instead precedes the fourth integrator (Int _2U ) with a first sign unit (Sgn _U ), which is a first Sign signal (Sig _U ) is generated, which is used instead of the corrected voltage signal (U _korr ) for integration by the fourth integrator (Int _2U ).

The 8th and 9 represent possible implementations for a first branch (ZW ₁ ) within the EMF evaluation (EMKA), which relates to the first motor phase (U).

10 shows 1 with the three branches (ZW ₁ , ZW ₂ , ZW ₃ ). The control block (St) or another controller activates the integration in the integrators of the branch, its associated half-bridge at the associated motor terminal (U, V, W) in the particular commutation interval (Φ ₁ to Φ ₆ ) each just in the high-resistance phase located. As explained above, the states of the commutation signals (A ₁ , A ₂ , A ₃ ) thereby reflect the current commutation interval (Φ ₁ to Φ ₆ ) in the form of a binary logic vector.

The 11 and 12 refer to the second motor phase (V). They correspond in function and structure to the 8th and 9 , The 11 and 12 represent possible implementations for a second branch (ZW ₂ ) within the EMF evaluation (EMKA), which refers to the second motor phase (V). Since the function of the individual components is analogous to that of the components in the 8th and 9 is, this description is not repeated here.

The 13 and 14 refer to the third motor phase (W). They correspond in function and structure to the 8th and 9 or the 11 and 12 , The 13 and 14 represent possible realizations for a third branch (ZW ₃ ) within the EMF evaluation (EMKA), which relates to the third motor phase (W). Since the function of the individual components is analogous to that of the components in the 8th and 9 respectively. 11 and 12 is, this description is not repeated here.

Advantages of the invention

The invention avoids an experimental determination of the commutation times by integration of the emf for determining a value for the level of the magnetic flux. The method is thus able to adapt itself to different engines. This also allows automated adjustment of the parameter in the manufacture of products based thereon.

The procedure adjusts itself again in each commutation in such a way that the right case in 7 results. It uses the lower signal of the curves in 7 , Thus, there is neither a transient nor a parameter that slowly approaches a target value. There is also no adjustment process. The process thus runs correctly right from the start without transient and control processes.

For example, series deviations in the engine to be used can be compensated. Even a realization of different products, which differ only with respect to the motor used, is possible without additional adjustment effort.

The invention is otherwise very well usable for the control of BLDC motors by block commutation in sensorless operation based on the evaluation of the magnetic flux. When using the magnetic flux as the integral of the emf, in contrast to the commutation based on the zero crossings of the emf, the adjustment between the angular position and the internal time base is omitted. A time base is therefore no longer necessary. Rather, a commutation takes place here directly without further calculation steps on the basis of the course of the EMF. The method thus offers greater stability and a better response to dynamic changes in the angular velocity of the rotor. In contrast to DE 100 54 594 A1 evaluates the technique disclosed here several times (t ₁ , t ₂ ) within a Kommutierungsintervalls and checks them for plausibility. As a result, faulty states are detected and are in contrast to the technology of DE 100 54 594 A1 in the event of an error, not used to determine the commutation time. In addition, a run-flat is made possible in these cases of error.

LIST OF REFERENCE NUMBERS

1

first process step

2

second process step

3

third process step

4

fourth process step

5

fifth process step

7

seventh process step

8th

eighth process step

9

ninth process step

11

eleventh process step

• A ₁ 'first commutation event signal
• A ₁ first commutation signal for the drive block (St). The first commutation signal is generated by the EMF evaluation (EMKA).
• The first commutation signal determines the time of the next voltage commutation by the drive block (St). The voltage commutation relates to the half-bridge of the control block, the upper and lower switch to the first motor connection (U) are connected. The signal is assigned to the first motor connection (U). The first commutation signal is only generated by means of the first AND operation (AND _U ) from the first commutation event signal (A ₁ ') when the first memory output (EN _U ) of the first memory (RSFF _V ) is set.
• A ₂ 'second commutation event signal
• A ₂ second commutation signal for the drive block (St). The second commutation signal is generated by the EMF evaluation (EMKA). The second commutation signal determines the time of the next voltage commutation by the drive block (St). The voltage commutation relates to the half-bridge of the control block, whose upper and lower switches are connected to the second motor terminal (V). The signal is assigned to the second motor connection (V). The second commutation signal is only generated by means of the second AND operation (AND _V ) from the second commutation event signal (A ₂ ') when the second memory output (EN _V ) of the second memory (RSFF _V ) is set.
• A ₃ 'third commutation event signal
• A ₃ third commutation signal for the control block (St). The third commutation signal is generated by the EMF evaluation (EMKA). The third commutation signal determines the time of the next voltage commutation by the drive block (St). The voltage commutation relates to the half-bridge of the control block whose upper and lower switches are connected to the third motor connection (W). The signal is assigned to the third motor connection (V). The third commutation signal is only generated by means of the third AND operation (AND _W ) from the third commutation event signal (A ₃ ') when the third memory output (EN _W ) of the third memory (RSFF _W ) is set.
• AND _U first AND link. The first AND gate generates the first commutation signal (A ₁ ) from the first commutation event signal (A ₁ ') when the first memory output (EN _U ) of the first memory (RSFF _U ) is set and the first commutation event signal (A ₁ ') Commutation time signaled.
• AND _V second AND link. The second AND gate generates the second commutation signal (A ₂ ) from the second commutation event signal (A ₂ ') when the second memory output (EN _V ) of the second memory (RSFF _V ) is set and the second commutation event signal (A ₂ ') Commutation time signaled.
• AND _W third AND link. The third AND gate generates the third commutation signal (A ₃ ) from the third commutation event signal (A ₃ ') when the third memory output (EN _W ) of the third memory (RSFF _W ) is set and the third commutation event signal (A ₃ ') Commutation time signaled.
• B _U first limiter. The first limiter generates the limited corrected voltage signal (U ' _korr ) of the first motor terminal (U) from the corrected voltage signal (U _korr ) of the first motor terminal (U). He sets the limited corrected voltage signal (U ' _corr ) of the first motor terminal (U) to zero, when the corrected voltage signal (U _corr ) of the first motor terminal (U) is negative. This limit can also be inverted given a suitable choice of sign of all components of the system. It is essential, therefore, that the limiter only pass one polarity of the corrected voltage signal (U _korr ) of the first motor terminal (U) and the other polarity maps to zero. This device part is assigned to the first motor connection (U).
• B _V second limiter. The second limiter generates the limited corrected voltage signal (V ' _corr ) of the second motor terminal (V) from the corrected voltage signal (V _corr ) of the second motor terminal (V). He sets the limited corrected voltage signal (V ' _corr ) of the second motor terminal (V) to zero, when the corrected voltage signal (V _corr ) of the second motor terminal (V) is negative. This limit can also be inverted given a suitable choice of sign of all components of the system. It is essential, therefore, that the second limiter only pass one polarity of the corrected voltage signal (V _korr ) of the second motor terminal (V) and that the other polarity maps to zero. This device part is assigned to the second motor connection (V).
• B _W third limiter. The third limiter generates the limited corrected voltage signal (W ' _korr ) of the third motor terminal (W) from the corrected voltage signal (W _korr ) of the third motor terminal (W). He sets the limited corrected voltage signal (W ' _corr ) of the third motor terminal (W) to zero, when the corrected voltage signal (W _corr ) of the third motor terminal (W) is negative. This limit can also be inverted given a suitable choice of sign of all components of the system. It is essential, therefore, that the third limiter allow only one polarity of the corrected voltage signal (W _corr ) of the third motor terminal (W) to pass, and that the other polarity be mapped to zero. This device part is assigned to the third motor connection (W).
• CMP _1U The first comparator compares the first threshold signal (S _1U ) with a reference potential (0) for the commutation and generates therefrom a first commutation event signal (A ₁ ') that controls the commutation of the half-bridge of the drive block (St) with the first Motor connection (U) is connected. This device part is assigned to the first motor connection (U).
• CMP _1V The second comparator compares the second threshold signal (S _1V ) with a reference potential (0) for the commutation and generates therefrom a second commutation event signal (A ₂ ') which controls the commutation of the half-bridge of the drive block (St) with the second Motor connection (V) is connected. This device part is assigned to the second motor connection (V).
• CMP _1W The third comparator compares the third threshold signal (S _1W ) with a reference potential (0) for the commutation and generates therefrom a third commutation event signal (A ₃ ') which controls the commutation of the half-bridge of the drive block (St) with the third Motor connection (W) is connected. This device part is assigned to the third motor connection (W).
• CMP _2U The fourth comparator compares the fourth threshold signal (S _2W ) with the reference signal (V _Ref2 ) for the commutation and generates from this the first set signal (S _U ). This device part is assigned to the third motor connection (W).
• CMP _2V The fifth comparator compares the fifth threshold signal (S _2V ) with the reference signal (V _Ref2 ) for the commutation and generates from this the second set signal (S _V ). This device part is assigned to the second motor connection (V).
• CMP _2W The sixth comparator compares the sixth threshold signal (S _2W ) with the reference signal (V _Ref2 ) for the commutation and generates from this the third set signal (S _W ). This device part is assigned to the third motor connection (W).
• EN _U first memory output of the first memory (RSFF _U )
• EN _V second memory output of the second memory (RSFF _V )
• EN _W third memory output of the third memory (RSFF _W )
• F _U first constant factor (F _U ) for adjusting the transient response of the first branch (ZW ₁ ) by a multiplier included in the fourth integrator (Int _2U ). The factor is assigned to the first motor connection (U).
• F _V second constant factor (F _V ) for adjusting the transient response of the second branch (ZW ₂ ) by a multiplier included in the fifth integrator (Int _2V ). The factor is assigned to the second motor connection (V).
• F _W third constant factor (F _W ) for adjusting the transient response of the third branch (ZW ₃ ) by a multiplier included in the sixth integrator (Int _2W ). The factor is assigned to the third motor connection (W).
• EMK electromotive force. The EMF can be measured in the freewheeling intervals on the freewheeling motor connections as a voltage curve.
• EMKA EMF evaluation. The EMF evaluation generates the commutation signals (A ₁ , A ₂ , A ₃ ) for the control of the commutation time of the half bridges of the drive circuit (St). This generation of the commutation signals (A ₁ , A ₂ , A ₃ ) takes place as a function of the voltages at the motor terminals (U, V, W) and the commutation intervals (Φ ₁ to Φ ₆ ). The first commutation signal (A ₁ ) is generated as a function of the terminal voltage at the first motor terminal (U) in the third commutation interval (Φ ₃ ) and / or in the sixth commutation interval (Φ ₆ ). The second commutation signal (A ₂ ) is generated as a function of the terminal voltage at the second motor terminal (V) in the first commutation interval (Φ ₁ ) and / or in the fourth commutation interval (Φ ₄ ). The third commutation signal (A ₃ ) is generated as a function of the terminal voltage at the third motor terminal (W) in the second commutation interval (Φ ₂ ) and / or in the fifth commutation interval (Φ ₅ ).
• Int _1U first integrator. The first integrator forms in the prior art by integration of the corrected voltage signal (U _korr ) of the first motor terminal (U) or by integration of the limited corrected voltage signal (U ' _korr ) of the first motor terminal (U) an associated first threshold signal (S _1U ) , This device part is assigned to the first motor connection (U).
• Int _1V second integrator. The second integrator forms in the prior art by integration of the corrected voltage signal (V _korr ) of the second motor terminal (V) or by integration of the limited corrected voltage signal (V ' _korr ) of the second motor terminal (V) an associated second threshold signal (S _1V ) , This device part is assigned to the second motor connection (V).
• Int _1W third integrator. The third integrator forms in the prior art by integration of the corrected voltage signal (W _korr ) of the third motor terminal (W) or by integration of the limited corrected voltage signal (W ' _korr ) of the third motor terminal (W) an associated third threshold signal (S _1W ) , This device part is assigned to the third motor connection (W).
• Int _2U fourth integrator. Integrating the corrected voltage signal (U _korr ) of the first motor terminal (U), the fourth integrator according to the invention forms the fourth threshold signal (S _2U ). Of the The value of the integration can be adjusted before output with a first constant factor (Eu) to adjust the transient response by a multiplier included in the fourth integrator (Int _2U ). This device part is assigned to the first motor connection (U).
• Int _{2 V} fifth integrator. By integrating the corrected voltage signal (V _korr ) of the first motor terminal (V), the fifth integrator according to the invention forms the fifth threshold signal (S _2V ). The value of the integration can be adjusted before output with a first constant factor (F _V ) for adjusting the transient response by a multiplier included in the fifth integrator (Int _2V ). This device part is assigned to the second motor connection (V).
• Int _2W sixth integrator. The sixth integrator according to the invention forms the sixth threshold signal (S _2W ) by integration of the corrected voltage signal (W _corr ) of the third motor terminal (W). The value of the integration can be adjusted before output with a first constant factor (F _W ) for adjusting the transient response by a multiplier included in the sixth integrator (Int _2W ). This device part is assigned to the third motor connection (W).
• M exemplary BLDC motor
• Φ ₁ first commutation interval. In this commutation interval, the upper switch of the half-bridge within the drive circuit connected to the first motor terminal (U) on one side and the upper supply voltage on the other side is closed. At the same time, in this commutation interval, the lower switch of the half-bridge within the drive circuit connected to the third motor terminal (W) on one side and the lower supply voltage on the other side is closed. In addition, both switches of the half-bridge within the drive circuit, which are connected to the second motor terminal (V), are opened. Therefore, the electromotive force (EMF) can be measured as the phase voltage at the second motor terminal (V) in this commutation interval.
• Φ ₂ second commutation interval. In this commutation interval, the upper switch of the half-bridge within the drive circuit connected to the first motor terminal (U) on one side and the upper supply voltage on the other side is closed. At the same time, in this commutation interval, the lower switch of the half-bridge within the drive circuit connected to the second motor terminal (V) on one side and the lower supply voltage on the other side is closed. In addition, both switches of the half-bridge within the drive circuit connected to the third motor terminal (W) are opened. Therefore, the electromotive force (EMF) can be measured as the phase voltage at the third motor connection (W) in this commutation interval.
• Φ ₃ third commutation interval. In this commutation interval, the upper switch of the half-bridge within the drive circuit connected to the third motor terminal (W) on one side and the upper supply voltage on the other side is closed. At the same time, in this commutation interval, the lower switch of the half-bridge within the drive circuit, which is connected to the second motor terminal (W) on one side and the lower supply voltage on the other side, is closed. In addition, both switches of the half-bridge within the drive circuit, which are connected to the first motor terminal (U), are opened. Therefore, the electromotive force (EMF) can be measured as the phase voltage at the first motor connection (U) in this commutation interval.
• Φ ₄ fourth commutation interval. In this commutation interval, the upper switch of the half-bridge within the drive circuit connected to the third motor terminal (W) on one side and the upper supply voltage on the other side is closed. At the same time, in this commutation interval, the lower switch of the half-bridge within the drive circuit, which is connected to the first motor terminal (U) on one side and the lower supply voltage on the other side, is closed. In addition, both switches of the half-bridge within the drive circuit, which are connected to the second motor terminal (V), are opened. Therefore, the electromotive force (EMF) can be measured as the phase voltage at the second motor terminal (V) in this commutation interval.
• Φ ₅ fifth commutation interval. In this commutation interval, the upper switch of the half-bridge within the drive circuit connected to the second motor terminal (V) on one side and the upper supply voltage on the other side is closed. At the same time, in this commutation interval, the lower switch of the half-bridge within the drive circuit, which is connected to the first motor terminal (U) on one side and the lower supply voltage on the other side, is closed. In addition, both switches of the half-bridge within the drive circuit connected to the third motor terminal (W) are opened. Therefore, the electromotive force (EMF) can be measured as the phase voltage at the third motor connection (W) in this commutation interval.
• Φ ₆ sixth commutation interval. In this commutation interval, the upper switch of the half-bridge within the drive circuit connected to the second motor terminal (V) on one side and the upper supply voltage on the other side is closed. At the same time, in this commutation interval, the lower switch of the half-bridge within the drive circuit connected to the third motor terminal (W) on one side and the lower supply voltage on the other side is closed. In addition, both switches of the half-bridge within the drive circuit, which are connected to the first motor terminal (U), are opened. Therefore, the electromotive force (EMF) can be measured as the phase voltage at the first motor connection (U) in this commutation interval.
• RSFF _U first memory. The first memory is typically an RS flip-flop, which is set when the fourth threshold signal (S _2U ) is beyond a reference signal (V _Ref2 ). The first memory is set by the first set signal (S _U ) to a first logical value and reset by the first reset signal (R _U ) to a second logical value. The reference signal (V _Ref2 ) may be specific to the first Be memory or equal to the reference signals of other memory. The RS flip-flop of the first memory is typically erased when the commutation is performed by means of the first commutation signal (A ₁ ) or when the time-out of the respective free-wheeling interval is reached. Preferably, the deletion is synchronous with the commutation time within the respective freewheeling interval, as this prevents further erroneous commutations within the respective freewheeling interval.
• RSFF _V second memory. The second memory is typically an RS flip-flop, which is set when the fifth threshold signal (S _2V ) is beyond a reference signal (V _Ref2 ). The second memory is set to a first logical value by the second set signal (S _V ) and reset to a second logical value by the second reset signal (R). The reference signal (V _Ref2 ) may be specific to the second memory or equal to the reference _{signals of} other memories. The RS flip-flop of the second memory is typically cleared when the commutation is performed by means of the second commutation signal (A ₂ ) or when the time-out of the respective free-wheeling interval is reached. Preferably, the deletion is synchronous with the commutation time within the respective freewheeling interval, as this prevents further erroneous commutations within the respective freewheeling interval.
• RSFF _W third memory. The third memory is typically an RS flip-flop, which is set when the sixth threshold signal (S _2W ) is beyond a reference signal (V _Ref2 ). The third memory is set by the third set signal (S _W ) to a first logical value and reset by the third reset signal (R _W ) to a second logical value. The reference signal (V _Ref2 ) may be specific to the third memory or equal to the reference _{signals of} other memory. The RS flip-flop of the third memory is typically cleared when the commutation is performed by means of the third commutation signal (A ₃ ) or when the time-out of the respective free-wheeling interval is reached. Preferably, the deletion is synchronous with the commutation time within the respective freewheeling interval, as this prevents further erroneous commutations within the respective freewheeling interval.
• R _U first reset signal. The first reset signal resets the first memory (RSFF _U ) to an initial value, a second logical value. Preferably, the first reset signal correlates with the commutation time within the respective free-wheeling interval or the end of the free-wheeling interval. The correlation with the commutation time is preferred.
• R _V second reset signal. The second reset signal resets the second memory (RSFF _V ) to an initial value, a second logical value. Preferably, the second reset signal correlates with the commutation time within the respective free-wheeling interval or the end of the free-wheeling interval. The correlation with the commutation time is preferred.
• R _W third reset signal. The third reset signal resets the third memory (RSFF _W ) to an initial value, a second logical value. Preferably, the third reset signal correlates with the commutation time within the respective free-wheeling interval or the end of the free-wheeling interval. The correlation with the commutation time is preferred.
• S _1U first threshold signal. The first threshold signal is generated by integration of the corrected voltage signal (U _korr ) of the first motor terminal (U) in the first integrator (Int _1U ). The signal is assigned to the first motor connection (U).
• S _1V second threshold signal. The second threshold signal is generated by integration of the corrected voltage signal (V _korr ) of the second motor terminal (V) in the second integrator (Int _1V ). The signal is assigned to the second motor connection (V).
• S _1W third threshold signal. The third threshold signal is generated by integration of the corrected voltage signal (W _korr ) of the third motor terminal (W) in the third integrator (Int _1W ). The signal is assigned to the third motor connection (W).
• S _2U fourth threshold signal. The fourth threshold signal is generated by integration of the corrected voltage signal (U _korr ) of the first motor terminal (U) in the fourth integrator (Int _2U ). The signal is assigned to the first motor connection (U).
• S _2V fifth threshold signal. The fifth threshold signal is generated by integration of the corrected voltage signal (V _korr ) of the second motor terminal (V) in the fifth integrator (Int _2V ). The signal is assigned to the second motor connection (V).
• S _2W sixth threshold signal. The sixth threshold signal is generated by integration of the corrected voltage signal (W _corr ) of the third motor terminal (W) in the sixth integrator (Int _2W ). The signal is assigned to the third motor connection (W).
• Sgn _U first sign unit. The first sign unit according to the invention determines the sign of the corrected voltage signal (U _korr ) of the first motor terminal (U). It outputs the first sign signal (Sig _U ). This device part is assigned to the first motor connection (U).
• Sgn _V second sign unit. The second sign unit according to the invention determines the sign of the corrected voltage signal (V _korr ) of the second motor terminal (V). It outputs the second sign signal (Sig _V ). This device part is assigned to the second motor connection (V).
• Sgn _W third sign unit. The third sign unit according to the invention determines the sign of the corrected voltage signal (W _korr ) of the third motor terminal (W). It outputs the third sign signal (Sig _W ). This device part is assigned to the third motor connection (W).
• Sig _U first sign signal. The first sign signal represents the sign of the corrected voltage signal (U _korr ) of the first motor terminal (U). The signal is assigned to the first motor connection (U).
• Sig _V second sign signal. The second sign signal represents the sign of the corrected voltage signal (V _korr ) of the second motor terminal (V). The signal is assigned to the second motor connection (V).
• Sig _W third sign signal. The third sign signal represents the sign of the corrected voltage signal (W _korr ) of the third motor terminal (W). The signal is assigned to the third motor connection (W).
• SdT marking the relevant figure as prior art
• SpS virtual star point signal. The virtual star point signal is preferably the sum of the first, second and third reduced clamp signals (U _r , V _r , W _r ) and is formed in the first summer (SU ₁ ). The signal is assigned to all motor connections (U, V, W).
• SpT ₁ first voltage divider. The first voltage divider reduces the voltage at the first motor terminal (U) by a factor of 1/3 to the reduced first terminal signal (U _r ). This device part is assigned to all motor connections (U, V, W).
• SpT ₂ second voltage divider. The second voltage divider reduces the voltage at the second motor terminal (V) by a factor of 1/3 to the reduced second terminal signal (V _r ). This device part is assigned to all motor connections (U, V, W).
• Spt ₃ third voltage divider. The third voltage divider reduces the voltage at the third motor terminal (W) by a factor of 1/3 to the reduced third terminal signal (W _r ). This device part is assigned to all motor connections (U, V, W).
• SSt Systemsteuerung It is typically a finite state machine as a flow control and / or a microprocessor with memory. In a particular embodiment of the invention, the system controller typically comprises, in particular, one or more analog-to-digital converters and optionally further memories which may have initial values for the fourth integrator (Int _2U ), the fifth integrator (Int _2V ) and / or _Secure the sixth integrator (Int _2W ). This initial value (V ₀₎ may optionally be generated in the form of three separate initial values specific to the respective branch (ZW _1, ZW _2, ZW _3). After switching off the supply voltage, however, the system would lose the respective initial value according to the invention and would have to perform another parameterization at the next restart. It therefore makes sense if the respective initial value according to the invention is saved in a non-volatile, preferably digital memory preferably within the system controller (SSt) and as a respective assigned specific initial value when the system is restarted in the fourth integrator (Int _2U ) or in the fifth Integrator (Int _2V ) or in the sixth integrator (Int _2W ) is loaded. This device part is typically assigned to all motor connections (U, V, W).
• St drive block. The drive block generates the signals for the three motor connections (U, V, W) from the commutation signals A ₁ , A ₂ , A ₃ . This block commutation drive circuit typically has three unshown half bridges. A first half-bridge is connected with its output to the first motor connection (U). A second half-bridge is connected with its output to the second motor terminal (V). A third bridge is connected with its output to the third motor terminal (W). Each of the jumpers typically has an upper switch which can connect the output of the respective half-bridge to an upper supply voltage and a lower switch which can connect the output of the respective half-bridge to a lower supply voltage. A simultaneous connection of upper and lower supply voltage to the respective output of a half-bridge is prevented by a latch circuit within the drive circuit for block commutation. In addition, the block commutation drive circuit has logic that can occupy at least six states. These six states correspond to the six commutation intervals (Φ ₁ to Φ ₆ ). With a predetermined edge of a commutation signal (A ₁ , A ₂ , A ₃ ), which may be falling and / or rising, the drive circuit changes state. This can lead to an asynchronicity of the commutation signals (A ₁ , A ₂ , A ₃ ).
• S _U first set signal. The first set signal is generated by the fourth comparator (CMP _2U ) and sets the first memory (RSFF _U ) to a first logical value.
• SU ₁ first summer. The first summer forms a virtual star point signal (SpS) from the first, second and third reduced terminal signal (U _r , V _r , W _r ). This device part is typically assigned to all motor connections (U, V, W).
• SU _2U second summer for the first motor connection (U). The second summer for the first motor terminal (U) subtracts the virtual neutral point signal (SpS) from the voltage signal of the first Motor terminal (U) and thereby forms the corrected voltage signal (U _corr ) of the first motor terminal (U). This device part is assigned to the first motor connection (U).
• SU _2V second summer for the second motor connection (V). The second summer for the second motor terminal (V) subtracts the virtual neutral point signal (SpS) from the voltage signal of the second motor terminal (V) and thereby forms the corrected voltage signal (V _korr ) of the second motor terminal (V). This device part is assigned to the second motor connection (V).
• SU _2W second summer for the third motor connection (W). The second summer for the third motor terminal (V) subtracts the virtual neutral point signal (SpS) from the voltage signal of the third motor terminal (W) and thereby forms the corrected voltage signal (W _korr ) of the third motor terminal (W). This device part is assigned to the third motor connection (W).
• S _V second set signal. The second set signal is generated by the fifth comparator (CMP _2V ) and sets the second memory (RSFF _V ) to a first logical value.
• S _W third set signal. The third set signal is generated by the sixth comparator (CMP _2W ) and sets the third memory (RSFF _W ) to a first logical value.
• t ₁ first time. The first time is within the respective freewheeling interval and time after the second time (t ₂ ) within the respective freewheeling interval, unless the determination of the first time point is disturbed. With the second time (t ₂ ), the commutation for the respective freewheeling interval is made possible synchronously to the first time. If the first time (t ₁ ) before the second time (t ₂ ) within the respective freewheeling interval in time, the commutation takes place at the second time (t ₂ ), otherwise typically at the first time (t ₁ ) within the respective freewheeling interval.
• t ₂ second time. The second time is after the beginning of each freewheeling interval and time before the first time (t ₁ ) within the respective freewheeling interval, unless the determination of the first time (t ₁ ) is not disturbed. With the second time, the commutation for the respective freewheeling interval is made possible. If the first time (t ₁ ) before the second time within the respective freewheeling interval, the commutation takes place at the second time (t ₂ ), otherwise typically at the first time (t ₁ ) within the respective freewheeling interval.
• U first motor connection of the exemplary BLDC motor
• U _r reduced first terminal signal. The voltage level is preferably lower by a factor of 1/3 than the voltage at the first motor terminal (U). The signal is assigned to the first motor connection (U).
• U _korr corrected voltage signal (U _korr ) of the first motor terminal (U). The corrected voltage signal (U _korr ) of the first motor terminal (U) is generated in the associated second summer (SU _2U ) by subtraction of the virtual star point signal (SpS) from the voltage signal of the first motor terminal (U). The signal is assigned to the first motor connection (U).
• U ' _korr limited corrected voltage signal (U' _korr ) of the first motor terminal (U). The signal is assigned to the first motor connection (U).
• V second motor terminal of the exemplary BLDC motor
• V _r reduced second terminal signal. The voltage level is preferably lower by a factor of 1/3 than the voltage at the second motor terminal (V). The signal is assigned to the second motor connection (V).
• V _Ref2 reference signal
• V _U reference voltage for the first motor phase (U)
• V _V reference voltage for the second motor phase (V)
• V _W reference voltage for the third motor phase (W)
• V _korr corrected voltage signal (V _korr ) of the second motor terminal (V). The corrected voltage signal (V _korr ) of the second motor terminal (V) is generated in the associated second summer (SU _2V ) by subtracting the virtual neutral signal (SpS) from the voltage signal of the second motor terminal (V). The signal is assigned to the second motor connection (V).
• V ' _korr limited corrected voltage signal (V' _korr ) of the second motor terminal (V). The signal is assigned to the second motor connection (V).
• W third motor connection of the exemplary BLDC motor
• W _r reduced third clamping signal. The voltage level is preferably lower by a factor of 1/3 than the voltage at the third motor terminal (W). The signal is assigned to the third motor connection (W).
• W _cor corrected voltage signal (W _corr ) of the third motor terminal (W). The corrected voltage signal (W _corr ) of the third motor terminal (W) is generated in the associated second summer (SU _2W ) by subtraction of the virtual neutral signal (SpS) from the voltage signal of the third motor terminal (W). The signal is assigned to the third motor connection (W).
• W ' _corr limited corrected voltage signal (W' _corr ) of the third motor terminal (W). The signal is assigned to the third motor connection (W).
• ZW ₁ first branch within the EMF evaluation (EMKA) for generating the first commutation signal (A ₁ ) from the EMF at the first motor terminal (U) during the third commutation interval (Φ ₃ ) and during the sixth commutation interval (Φ ₆ ). This device part is assigned to the first motor connection (U).
• ZW ₂ second branch within the EMF evaluation (EMKA) for generating the second commutation signal (A ₂ ) from the EMF at the second motor terminal (V) during the first commutation interval (Φ ₁ ) and during the fourth commutation interval (Φ ₄ ). This device part is assigned to the second motor connection (V).
• ZW ₃ third branch within the EMF evaluation (EMKA) for generating the third commutation signal (A ₃ ) from the EMF at the third motor terminal (W) during the second commutation interval (Φ ₂ ) and during the fifth commutation interval (Φ ₅ ). This device part is assigned to the third motor connection (W).

List of quoted writings

• DE 100 54 594 A1
• DE 100 64 486 A1
• DE 10 2008 025 442 A1
• DE 10 2009 019 414 A1
• US 2007 0 282 461 A1
• US 2008 0 252 240 A1
• US 2014 0 062 364 A1

Claims (5)

Device for preventing erroneous commutation and / or commutation within a device for commutation to a fully automatic sensorless commutation of a brushless DC motor (M) with at least three motor terminals (U, V, W), wherein the here as the first motor terminal (U) designated motor terminal any, exemplary motor connection of these at least three motor connections (U, V, W) is, with a. a drive block (St) commutes the at least three motor terminals (U, V, W) in Kommutierungsintervallen (Φ ₁ to Φ ₆ ) and energized or not energized, wherein the first motor terminal (U) in associated Kommutierungsintervallen (Φ ₁ , Φ ₂ , Φ ₄ , Φ ₅ ) are energized and in other associated commutation intervals (Φ ₃ , Φ ₆ ), the freewheeling intervals, not energized, and b. a first sub-device (SpT ₁ , SpT ₂ , SpT ₃ , SU _1U ), i. which generates a virtual neutral point signal (SpS) from the electrical potential curves of the at least three motor connections (U, V, W) and ii. a first voltage divider (SpT ₁ ) and iii. a second voltage divider (SpT ₂ ) and iv. a third voltage divider (SpT ₃ ) and v. a first summer (SU _1U ), and c. a second summer (SU _2U ) which generates a corrected voltage signal (U _korr ) as the difference between the electric potential profile of the first motor terminal (U) and the virtual star point signal (SpS), and d. a third sub-device for determining the value of the magnetic flux during the respective free-wheeling interval on the basis of the corrected voltage signal (U _korr ) in the form of a fourth threshold signal (S _2U ) and e. a fourth sub-device for determining a first time (t ₁ ) in the respective freewheeling interval in the form of an edge of a first Kommutierungsereignissignals (A ₁ '), after the start of one of these freewheeling intervals (Φ ₃ , Φ ₆ ) within the relevant freewheeling interval based on the determined value the magnetic flux by means of a first sub-method performed by the fourth sub-device, which determines a first time (t ₁ ) in the respective free-wheeling interval, which would correspond to an ideal commutation time in the respective free-wheeling interval in an ideal engine in an undisturbed operation and f. a fifth sub-device for determining a second time (t ₂ ) in the respective freewheeling interval in the form of an edge of a set signal (S _U ), which is within the respective freewheeling interval after the start of the freewheeling interval, by a second, performed by the fifth sub-device second A sub-method that differs from the first sub-method performed by the fourth sub-device and that is configured such that the second time (t ₂ ) in the respective free-wheeling interval in an undisturbed operation in an ideal engine simultaneously before the first time (t ₂ ) and must be after the start of each freewheeling interval and G. a sixth sub-device, which suppresses a commutation based on the first commutation event signal (A ₁ ') in a real operating case, when the second detected time (t ₂ ) is temporally in the respective free-wheeling interval after the first determined time (t ₁ ), and a substitute commutation to the second second point (t ₂ ) in the respective freewheeling interval in this case, wherein the sixth subdevice performs the commutation in a real operating case to the first determined time (t ₁ ) in the respective freewheeling interval when the second determined time (t ₂ ) in the form of the edge of the set signal (S _U ), temporally before the first determined time (t ₁ ) in the respective freewheeling interval in the form of the edge of the first commutation event signal (A ₁ '), and h. wherein the third subdevice comprises a fourth integrator (Int _2U ) which integrates and / or filters the corrected voltage signal (U _korr ) during the respective free-wheeling interval into a fourth threshold signal (S _2U ) as a signal for the value of the magnetic flux, and i. wherein the fourth sub-device comprises a first comparator (CMP _1U ) which compares the fourth threshold signal (S _2U ) with a reference potential (0) and / or a reference variable (0) and a first commutation event signal (A ₁ ') for the commutation, in Dependence on the result of this comparison, and j. wherein the fifth sub-device comprises a fourth comparator (CMP _2U ) which compares the fourth threshold signal (S _2U ) with a reference _potential (V _Ref2 ) and / or a reference value (V _Ref2 ) and generates a set signal (S _U ), and k. wherein the sixth subdevice comprises a first memory (RSFF _U ) having a first memory output (EN _U ), which by means of the first set signal (S _U ) by the fourth comparator (CMP _2U ), depending on the result of this comparison to a status with the Property "released" sets or does not set, and l. wherein the first memory output (EN _U ) is reset with or after the signaling of a commutation by the commutation signal (A ₁ ) or at the end of the respective free-wheeling interval to a second status that does not have the property "enabled", and m. wherein a logic operation device (AND _U ) is provided to perform a logic operation between the first memory output (EN _U ) of the first memory (RSFF _U ) and the first commutation event signal (A ₁ ') and the commutation signal (A ₁ ) as a result Create link if the first memory output (EN _U ) is in a status with the property "enabled" and if the fourth threshold signal (S _2U ) reaches the reference potential (0) and / or reference (0) and / or crosses in terms of value , Apparatus according to claim 1 with a. a first sign unit (SGN _U), which generates a first sign signal (Sig _U) from the corrected voltage signal (U _corr) indicating the sign of the corrected voltage signal (U _corr), and b. a third sub-device for determining the value of the magnetic flux in the form of the fourth threshold signal (S _2U ) during said freewheeling intervals based on the additional first sign signal (Sgn _U ), instead of as in claim 1 based on the corrected voltage signal (U _korr ). Method for carrying out with a device according to claim 1 and for preventing erroneous commutation and / or Kommutierungszeitpunkte within a method for determining the commutation of a brushless DC motor (M) with at least three motor terminals (U, V, W), which evaluates the magnetic flux, wherein the motor connection designated here by the first motor connection (U) is an arbitrary, exemplary motor connection of these at least three motor connections (U, V, W), comprising the steps a. Commutating and energizing the at least three motor terminals (U, V, W) in commutation intervals (Φ ₁ to Φ ₆ ) by a drive block (St); b. Energizing the first motor terminal (U) in a subset of the commutation intervals (Φ ₁ , Φ ₂ , Φ ₄ , Φ ₅ ) by the drive block (St); c. Not energizing the first motor terminal (U) in the other commutation intervals (Φ ₃ , Φ ₆ ), the freewheeling intervals, by the drive block (St); d. integrating the EMF to determine the magnetic flux at the first motor terminal (U) within one or more freewheeling intervals that are one or more of the commutation intervals (Φ ₃ , Φ ₆ ) in which the first motor terminal (U) is driven by the drive block (St ) is not energized; e. Generating a neutral point signal (SpS) from the electrical potential curves of at least three motor connections (U, V, W) by a first partial device (SpT ₁ , SpT ₂ , SpT ₃ , SU _1U ); f. Forming a corrected voltage signal (U _korr ) as the difference between the electrical potential profile of the first motor terminal (U) and the virtual neutral point signal (SpS); G. Integrating the corrected voltage signal (U _korr ) for detecting the value of the magnetic flux during a respective one-way interval into a fourth threshold signal (S _2U ) through a fourth integrator (Int _2u ); H. Determining a first time (t ₁ ) after the start of one of these freewheeling intervals (Φ ₃ , Φ ₆ ) as the freewheeling interval concerned on the basis of the detected value of the magnetic flux by means of a first partial method, which determines a first time point (t ₁ ) in the relevant freewheeling interval which would correspond to an ideal commutation time in the respective free-wheeling interval for an ideal engine in an undisturbed operation, i. wherein the first sub-method comprises comparing the fourth threshold signal (S _2U ) with a reference potential (0) and / or reference (0) by means of a first comparator (CMP _1U ) and generating a first commutation event signal (A ₁ ') representing the first Time (t ₁ ) marked in the relevant freewheeling interval, depending on the result of this comparison comprises; j. Determining a second time (t ₂ ) in the relevant freewheeling interval of the time after the beginning of the respective freewheeling interval is due to a second partial method, which is different from the first partial method and which is designed such that the second time point (t ₂ ) in the respective Freewheeling interval in an undisturbed operation for an ideal engine at the same time before the first time (t ₁ ) must lie in the relevant freewheeling interval, k. wherein the second partial method, the comparison of the fourth threshold signal (S _2U ) with a reference _potential (V _Ref2 ) and / or with a reference value (V _Ref2 ) by means of a fourth comparator (CMP _2U ) and the generation of a first set signal (S _U ) in dependence from the result of the comparison by the fourth comparator (CMP _2U ) and. l. wherein the first set signal (S _U ) marks the second time (t ₂ ) in the respective one-way interval; m. Suppressing a commutation in the relevant freewheeling interval in a real operating case for the first determined time (t ₁ ) in the relevant freewheeling interval, if the second determined time (t ₂ ) temporally after the first determined time (t ₁ ) in the relevant freewheeling interval and substitute Commutation to the second second point (t ₂ ) in the relevant freewheeling interval in this case by commutation in the respective freewheeling interval in a real operating case for the first determined time (t ₁ ) in the relevant freewheeling interval, if the second determined time (t ₂ ) in the relevant Freewheeling interval in time before the first determined time (t ₁ ) is in the relevant freewheeling interval. Method according to Claim 3, comprising the following steps for suppressing and triggering the commutation in the relevant freewheeling interval: a. Setting the first memory output (EN _U ) of a first memory (RSFF _U ) in response to the first set signal (S _U ) to a status with the property "enabled"; b. Generating a commutation signal (A ₁ ) for the commutation by a logical operation (AND _U ) when the first memory output (EN _U ) is in a status with the property "enabled" and when the value of the fourth threshold signal (S _2U ) the value reaches the reference signal (V _Ref2 ) and / or crosses in terms of value; c. Resetting the first memory output (EN _U ) of the first memory (RSFF _U ) with or after the signaling of a commutation by the commutation signal (A ₁ ) to a second status that does not have the property "enabled" or at the latest at the end of the relevant freewheeling interval. Method according to one or more of claims 3 to 4 comprising the steps a. of generating a first sign signal (Sig _U) by means of a first sign unit (SGN _U) from the corrected voltage signal (U _corr) indicating the sign of the corrected voltage signal (U _corr) and b. the integration of the first sign signal (Sig _U ) for integrally detecting the EMF in the freewheeling intervals into a fourth threshold signal (S _2U ) by a fourth integrator (Int _2U ); instead of step c. Integration of the corrected voltage signal (U _korr ) for detecting the magnetic flux during said freewheeling intervals to a fourth threshold signal (S _2U ) by a fourth integrator (Int _2U ).

Images