Sensorless rotor position determination

Abstract

The invention relates to a method for determining the position of a rotor (2) of an electric motor (1). The method comprises the following steps: - Measurement of a time interval of a cyclically occurring electrical signal of at least one motor phase of the electric motor (1) and - relating the cyclically occurring electrical signal to a load acting on the rotor (2) of the electric motor and thereby determining the position of the rotor (2) in relation to the mechanical load of the electric motor (1).

Description

[0001] The present invention relates to a method for determining the position of a rotor of an electric motor. The invention also relates to an electrical control unit for controlling an electric motor.

[0002] To control a brushless DC motor (BLDCM), it is necessary to know the rotor position. In principle, the rotor position can be determined using sensors. However, in certain applications, this is not practical. For example, in compressor applications, the use of sensors is not feasible due to the very harsh environment and also due to the cost.

[0003] Another method for determining rotor position is the so-called sensorless approach. In the compressor industry, the most widely used sensorless position determination method is the so-called "direct back-electromagnetic force method" (abbreviated to direct BEMF method), in which the zero crossings of the induced machine voltages are determined by reading the electrical motor feedback signals.

[0004] In principle, such a brushless DC motor control method (BLDCM control method) that uses zero-point crossings of the BEMF can be used to obtain information for better control of the speed and torque of the electric motor. However, using the BEMF to control the speed and torque of the electric motor in this way cannot tell the microprocessor where the maximum torque of an external device driven by the electric motor is located (in the case where the electric motor is used to drive a compressor, the microprocessor cannot therefore tell the location of the maximum torque of the compressor). Information about where the maximum torque of the mechanical load (e.g.a compressor) could, for example, be used to minimize vibrations and to park the rotor after the electric motor has stopped in order to achieve a correspondingly smoother start and acceleration.

[0005] When using electric motors 1, 5, 7 with more than two poles (e.g., electric motors 5, 7), the rotor 2 can be in more than one position relative to the electric / magnetic field 3 if the electric motor 1, 5, 7 is operated without positioning sensors. Therefore, it is not possible to determine / estimate an absolute position of the rotor 2 using the electric field 3 (or for the mechanical load associated with the electric motor 1, 5, 7; in the case where a compressor is driven by the electric motor 1, 5, 7, the compressor).

[0006] When the electric motor 1, 5, 7 is connected to a mechanical load (e.g. a compressor), the position of the rotor 2 can be used to improve the performance of the mechanical load (e.g. the compressor).

[0007] The position of the electric field 3 is always known because it is controlled by a microcontroller. For a two-pole electric motor 1 ( Fig. 1), the rotor 2 can only be in one position if it is aligned with the field 3. In other words, if the field 3 has moved 360° electrically, the rotor has also moved 360° mechanically. Fig. 1 shows the rotor 2, indicated in a position of a first point 4.

[0008] For a four-pole motor 5, field 3 must now move twice by 360° electrically = 720° electrically for rotor 2 to move 360° mechanically. There are now two possible positions, 4 and 6, for rotor 2. Fig. 2 shows that the rotor 2 can also be positioned at the second point 6 for a given electric field 3.

[0009] For a six-pole motor 7, the electric field 3 must move 3 * 360° electrically = 1080° electrically for the rotor 2 to move 360° mechanically. There are now three possible positions 4, 6, 8 for the rotor 2 for a given electric field 3. Fig. 3 shows that the rotor 2 can now be positioned at the first point 4 as well as at the second point 6 or at the third point 8.

[0010] Due to this ambiguity in the connection between the electric field and the mechanical position of the electric motor's rotor, it is often not possible to achieve sufficient control of the electric motor, which can be particularly problematic in connection with vibrations and degraded starting behavior. Such problems arise not only because of this ambiguity between the electric field and the mechanical position, but also because the rotor can be "lost," i.e., the rotor can be (or is) stopped (e.g., due to a particularly high mechanical external load) while the electric field continues to rotate. Of course, such a "loss" of the rotor is also disadvantageous.

[0011] JP H11-164 584 A discloses a brushless DC motor and a voltage control circuit. The voltage control circuit allows the DC motor to be started at a low load in order to bring the rotor into the best possible starting position.

[0012] DE 10 2006 043 683 A1 discloses a method for operating an electronically commutating electric motor. Test voltage pulses are applied to the various motor phases sequentially, and the resulting BEMF voltages are then measured. The rotor position is then determined from the resulting BEMF voltages using reference voltage values.

[0013] The generic WO 2005 / 119 898 A2 discloses a brushless DC motor comprising a rotor and a stator for use in an actuator for a flap or valve for controlling a gas or liquid volume flow. The zero crossings of the BEMF voltages of the motor phases are measured, and the rotor position is determined from this.

[0014] It is therefore an object of the invention to provide a method for determining the position of a rotor of an electric motor that is improved compared to known methods. A further object of the invention is to provide an electrical control unit for controlling an electric motor that is improved compared to known electrical control units.

[0015] The invention solves both of these problems.

[0016] It is proposed to carry out a method for determining the position of a rotor of an electric motor, wherein the rotor has more than two poles, with the features of claim 1.

[0017] The process includes, among other things, the following steps: - measuring a time period of a cyclically occurring electrical signal from at least one motor phase of the motor in question, and - relating the cyclically occurring electrical signal to a mechanical load affecting the rotor of the electric motor, - and thereby determining the position of the rotor in relation to the mechanical load of the electric motor.

[0018] The basic idea utilizes the fact that the rotor of an electric motor is not subjected to a uniform mechanical load while rotating. This effect can occur even when no mechanical load is connected to the electric motor, because small fluctuations normally arise in the stator of the electric motor during one revolution of the rotor. In particular, when an external load is connected to the motor, it has been noticed that, at least in a large proportion of technical applications, the mechanical force on the rotor of an electric motor is not constant. In a fairly large proportion of technical applications, the electric motor is even connected to a cyclically repeating mechanical load. Examples of this are compressors.In the case of a compressor, the mechanical position of the rotor during one revolution is influenced by the piston's backlash (where the piston's backlash varies during such a revolution, thus exerting a different load on the electric motor's rotor). This varying backlash is typically significantly influenced by the pressure of the gas pumped by the compressor, which is located in a pump cavity with a cyclically variable volume. Such a design is often used for compressors. The term "cyclically repeating mechanical force" does not necessarily mean that two cycles are very similar. On the contrary, variations over time can occur, particularly during a start-up phase of the electric motor. Compressors are also examples of this.Here, when the compressor is first started, the pressure level (and thus the counterforce development) is often different during the first and / or first revolutions compared to later revolutions. However, the varying mechanical load on the rotor (regardless of its actual origin) will normally cause more or less pronounced variations in the rotational behavior of the rotor, and thus also in certain electrical signals from the electric motor over different time periods. Therefore, if (at least) a time period of (at least) a certain characteristic of the cyclically occurring electrical signal from at least one motor phase of the electric motor is measured, a relationship can be established with the force acting on the rotor of the electric motor, so that the position of the rotor in relation to the mechanical load of the electric motor can be determined.The electrical signal can originate from a wide variety of electrical signals. In particular, voltages, currents, counter-electromotive force voltages, crossings of certain voltage potentials, and the like can be used. A particular advantage of the method proposed here is that the ambiguity between the position of the electric field and the mechanical position of the rotor can be overcome. This is particularly advantageous when a motor with more than two poles (e.g., four poles, six poles, eight poles, ten poles, etc.) is used.

[0019] Furthermore, the mechanical load on the rotor of the electric motor is a cyclical mechanical load. In this case, variations in the mechanical load on the rotor are typically quite large, so that a better position measurement can be achieved. An example of a cyclical mechanical load is a compressor. A cyclical mechanical load does not necessarily mean that two consecutive cycles are essentially the same. In particular, some variations may arise over time. In particular, a cycle during the start-up phase of the electric motor may differ significantly from a cycle after the end of the start-up phase, particularly in terms of quality and / or quantity.

[0020] The method mentioned includes the step of relating the mechanical load to the position of the electric motor's rotor. In this case, it is possible to draw conclusions not only about the rotor's position relative to the electric motor's mechanical load, but also about the rotor's "absolute" position. This can be advantageous.

[0021] Additionally, it is proposed to implement the method in such a way that said at least one cyclically occurring electrical signal is a voltage and / or a current. Initial tests have shown that such signals are relatively easy to obtain and contain a wealth of information.

[0022] The period of the cyclically occurring electrical signal of at least one motor phase of the electric motor is the demagnetization time of that motor phase. Initial tests have shown that the demagnetization time is particularly valuable when drawing conclusions about the rotor position, both in relation to the mechanical load and the absolute position.

[0023] The aforementioned time period, namely the demagnetization time, is related to the current flowing through at least one winding of the electric motor's stator. In this case, the position accuracy can usually be improved.

[0024] Furthermore, it is possible that the time span of the cyclically occurring electrical signal of at least one motor phase of the electric motor is the time between two consecutive zero point crossings. Again, initial tests have shown that this signal is usually relatively easy to achieve and contains a wealth of information. A "zero point crossing" does not just mean the crossing of "the" zero potential (earth potential), but also the crossing of a specific potential, particularly (but not necessarily) when the electric motor is not earthed (galvanically isolated). For example, if an electric motor is powered by a DC voltage source, it is possible that the "zero potential" is essentially half the potential between the two contacts of the DC voltage source.

[0025] A further preferred embodiment can be achieved by establishing a relationship between the time period and the load acting on the rotor of an electric motor. Here, too, initial tests have shown that particularly accurate measurements can be made in this way. Therefore, the result derived from this position can be correspondingly accurate.

[0026] In particular, it is possible that the method is used to determine the rotor position and to park the rotor after stopping. In this case, a subsequent start can be improved and / or carried out more quickly. Both options can be advantageous.

[0027] In particular, it is possible that the method is used to improve the starting behavior of the electric motor. In this case, noise and / or the time required for starting the electric motor can be reduced.

[0028] It is further preferred that the method be used to reduce vibrations. In this case, noises that are unpleasant to a person and / or can shorten the service life of the electric motor can be reduced.

[0029] Additionally, it is possible to compare a large number of electrical signals over time during the process. This allows short-term fluctuations to be compensated for, thereby improving the accuracy of the process.

[0030] Furthermore, an electrical control unit for controlling an electric motor is provided, wherein the electrical control unit is constructed and arranged to carry out a method according to one of the preceding claims. In this way, the electrical control unit can have the same characteristics and advantages as described above, at least analogously. In addition, it is possible to modify the electrical control unit as described above, at least analogously.

[0031] In particular, an electric motor assembly is provided which has at least one electric control unit as described above. Such an electric motor assembly can have the same characteristics and advantages as the above-proposed electric control unit and / or the above-described method, at least analogously.

[0032] Furthermore, it is possible for the electric motor assembly to comprise at least one fluid pump unit, in particular at least one compressor. Such use of the electric motor assembly (or the motor used therein) will typically yield particularly advantageous results.

[0033] The present invention and its advantages will become clearer through the following description of possible embodiments of the invention, with reference to the accompanying figures. The figures show: Fig. 1 the rotational behavior of a design of a two-pole electric motor in a schematic drawing; Fig. 2 the rotational behavior of a design of a four-pole electric motor in a schematic drawing; Fig. 3 the rotational behavior of a design of a six-pole electric motor in a schematic drawing; Fig. 4 a typical behavior of the torque of a compressor versus the rotor position of a compressor driven by an electric motor; Fig. 5 an example of the temporal behavior of phase voltages over time for three different phases of a six-pole electric motor used to drive a compressor; Fig. 6 Test results from initial tests showing the electrical currents versus time for three different phases of a six-pole electric motor used to drive a compressor; Fig. 7 first tests of the demagnetization time via the mechanical rotation of the rotor of an electric motor used to drive a compressor; Fig. 8 first tests of a commutation time via the mechanical rotation of the rotor of an electric motor used to drive a compressor.

[0034] The method and / or the electrical control circuit proposed here utilize the fact that a compressor does not have a uniform load over a whole revolution and thus loads the electric motor 1, 5, 7 differently as a function of the mechanical position of the rotor 2 during cycles performed by the piston.

[0035] By measuring the demagnetization time (cf. Fig. 7), which is related to the current, which in turn is related to the load, it is possible to determine the position of the rotor 2.

[0036] Fig. Figure 4 shows test results showing that the maximum load 9 is just before TDC (top dead center, the piston is in its upper position) and that it is possible to determine this position. Graph 10 shows the torque (along the ordinate) versus the rotor position (along the abscissa).

[0037] Because there are 18 commutations for one revolution, the demagnetization time (cf. Fig. 5) for the respective demagnetization events 11, the counter electromagnetic force (BEMF) is measured whenever a commutation ends and the phase is de-energized (zero crossing event 13). The demagnetization times of the respective demagnetization events 11 differ. Fig. 6 for the respective motor phases and / or over time (i.e., for different rotor positions). For each phase voltage, there is a maximum demagnetization time 12 that occurs repeatedly.

[0038] The Fig. 6 and Fig. 7 also shows the results of a test, where the results show the torque from the compressor and how it affects the speed, current, demagnetization time and commutation time. Fig. The curve shown in Figure 6 shows the electrical currents 14a, 14b, 14c for the three different motor phases of a six-pole electric motor 7 over one revolution (cf. Fig. 4). In the curves in Fig. 7, on the other hand, are the demagnetization times for one of the three different motor phases of a six-pole electric motor 7 over one revolution (cf. Fig. 4). Finally, the curve in Fig. 8 the commutation times 16a, 16b, 16c for the three different motor phases of a six-pole electric motor 7 over one revolution (cf. Fig. 4). For increased accuracy, it is suggested that the demagnetization time 15 be stored for each commutation and evaluated over a series of revolutions. This "averaging over time" can, of course, also be used for other signals.

[0039] When constant current regulators are used in an embodiment not according to the invention, the demagnetization time is the same for all commutations, making it impossible to distinguish between the different commutations. Instead, the maximum load can be determined by analyzing the commutation time over one mechanical revolution, because the speed decreases as the load increases.

[0040] In particular, the invention can be described as follows: Method for determining the position of a rotor 2 of an electric motor 1, 5, 7, in particular an electric motor 1, 5, 7 for driving a compressor, wherein the rotor 2 drives a piston of the compressor and the method comprises the following steps: - measuring the demagnetization time 15 of the rotor 2, where the demagnetization time 15 is the time that elapses until the phase voltage is on the supply rail opposite the supply rail of the previous commutation phase, - relating the demagnetization time 15 to the current flowing through the windings of the stator of the electric motor 1, 5, 7, - relating the current to the load acting on the rotor 2 and / or in particular the piston of the compressor during cycles of the piston, and thereby determining the position of the rotor 2 relative to a load acting on the rotor 2, in particular shortly before the top dead center (TDC) of the piston.

[0041] Method according to the previous section, the method comprising the step of determining the position of a rotor 2 of a compressor, the rotor 2 having more than two poles.

[0042] Method according to one of the preceding two sections, the method comprising the step of determining the position of the rotor 2 for the purpose of reducing vibrations of the compressor.

[0043] Method according to one of the previous three sections, the method comprising the step of determining the position of the rotor 2 for the purpose of parking the rotor 2 after the stop.

[0044] Method according to one of the previous four sections, the method comprising the step of determining the position of the rotor 2 for the purpose of smoother starting and acceleration.

[0045] Method according to one of the previous five sections, the method comprising the step of storing the demagnetization time 15 for each commutation and evaluating the demagnetization time 15 over a series of revolutions of the rotor 2 of the compressor.

[0046] Further information can be found in applications DE 10 2012 006 491 A1 and DE 10 2012 006 495 A1, which were filed on the same day as the present application.

Claims

[1] Method for determining the position of a rotor (2) of an electric motor (1, 5, 7), the rotor (2) having more than two poles, the method comprising the following steps: - measuring a time period of a cyclically occurring electrical signal (13, 14, 15, 16) of at least one motor phase of the electric motor (1, 5, 7) and - relating the cyclically occurring electrical signal to a load (9) acting on the rotor (2) of the electric motor and thereby determining the position of the rotor (2) in relation to the mechanical load (9) of the electric motor (1, 5, 7), wherein the mechanical load (9) on the rotor (2) of the electric motor (1, 5, 7) is a cyclically occurring mechanical load (9), wherein the time period of the cyclically occurring electrical signal (13, 14, 15, 16) of at least one motor phase of the electric motor (1, 5, 7) is the demagnetization time (15) of at least one motor phase, wherein the time period, namely the demagnetization time (15), is related to the current (14) flowing through the at least one winding of the stator of the electric motor (1, 5, 7), characterized by that the mechanical load (9) connected to the motor (1, 5, 7) is a compressor driven by the motor (1, 5, 7) and the rotor (2) drives a piston of the compressor, wherein the compressor does not have a uniform load (9) over a complete revolution of the rotor (2) and thus loads the electric motor (1, 5, 7) differently as a function of the mechanical position of the rotor (2) during cycles performed by the piston, wherein the method comprises the step of relating the mechanical load (9) to the position of the rotor (2) of the electric motor (1, 5, 7), and wherein the absolute position of the rotor (2) is determined by measuring the demagnetization time (15) related to a current (14) which in turn is related to the mechanical load (9). [2] Method according to claim 1, characterized by that a correlation is established between the time period and the load (9) exerted on the rotor (2) of the electric motor (1, 5, 7). [3] Method according to one of the preceding claims, characterized by that the step for determining the position of the rotor (2) is applied to park the rotor (2) after the stop. [4] Method according to one of the preceding claims, characterized by that the method is used to improve the starting behavior of the electric motor (13, 14, 15, 16). [5] Method according to one of the preceding claims, characterized by that the process is used to reduce vibrations. [6] Method according to one of the preceding claims, characterized by that an average over time is determined from a plurality of cyclically occurring electrical signals (13, 14, 15, 16). [7] Electrical control unit for controlling an electric motor, characterized by that it is constructed and arranged to carry out the method according to any one of the preceding claims. [8] Electric motor arrangement with at least one electric control unit according to claim 7. [9] Electric motor arrangement according to claim 8 with at least one fluid pumping unit, namely at least one compressor unit.

Images