Control device for an electric motor

Abstract

A drive device 1000 for an electric motor, comprising: an inverter 3 that drives a motor 4; a voltage divider circuit 2 serving as a neutral point potential detection unit that detects a neutral point potential of a stator winding of the motor 4; and a controller 1 that estimates a rotor position of the motor 4 based on the detected neutral point potential and controls the inverter 3 based on the estimation result. A ground potential of the controller 1 is set to the negative potential or the negative potential of a DC voltage supplied to the inverter 3. The voltage divider circuit 2 detects the neutral point potential with respect to the negative potential or the positive potential.The controller 1 estimates the rotor position based on a difference between a first star point potential detected during the ON / OFF operation of the inverter 3 and a first fixed reference potential (2 / 3)Emax, and based on a difference between a second star point potential detected during the ON / OFF operation and a second fixed reference potential (1 / 3)Emax.

Description

TECHNICAL FIELDThe present invention relates to an electric motor driving apparatus that drives and controls a synchronous motor used for controlling the rotational speed of, for example, a fan, a pump, a compressor, a spindle motor, and the like, or controlling the torque in a position determining apparatus for a conveyor or a machine tool, an electric assist system, or the like, and further relates to an integrated type electric motor system, a pump system, a compressor system, and a position determining system, all of which include such an electric motor driving apparatus. A control device as described in the preambles of claims 1 and 2 is known from DE 10 2007 030 748 A1.PRIOR ARTPermanent magnet motors (i.e., synchronous motors) are compact and have high efficiency; such motors are widely used in various fields such as industrial, consumer electronics, automobiles, and the like. However, information on the position of the rotor of the motor is required for driving a permanent magnet motor, so that a position sensor has been necessary up to now.Recently, the position sensor has been often omitted, and it has become common practice to use sensorless control for the rotation speed control or torque control of a permanent magnet motor. By implementing sensorless control, it is possible to save the cost associated with the position sensor (i.e., the cost for the sensor itself, the cost incurred by the wiring for the sensor, etc.), and make the entire system more compact. In addition, it is advantageous that the system can be used in a low quality environment, etc., by saving the sensor. In the present practice, for the sensor-less control of a permanent magnet motor, either a method of driving the permanent magnet motor is performed in which an induced voltage (i.e., a speed-attributed voltage) generated by rotation of the rotor of the permanent magnet motor and this position information about the rotor is used, or a method of position estimation in which an estimate of the rotor position is calculated from a numerical model of the respective motor is used, and the like.However, there are also severe problems with these sensorless control methods. They occur at low speeds in the position detection methods. Most of the methods of sensorless control currently being put into practice are methods based on the induced voltage generated by the permanent magnet motor. Therefore, the sensitivity undesirably decreases when the motor is stopped or in a low-speed region where the induced voltage is low, and there is a risk that the position information will be omitted in the noise. Various strategies have been developed to solve this problem.In the invention described in JP 2010-74898 A, the position information is obtained by measuring the "neutral point potential", i.e., the potential at the connection point of the stator windings of the three phases. By detecting the neutral point potential in synchronism with the pulse voltages supplied to the motor from the inverter, it is possible to detect the voltage induced due to the imbalance of the inductances, and it is possible to obtain the potential change depending on the rotor position. The above-mentioned invention is thus distinguished by position information obtained during normal sine modulation of the voltages supplied to the motor by PWM (pulse width modulation). Here, the rotor position means the position of the permanent magnet attached to the rotor.FIG. 27 is a figure showing an example of a conventional synchronous motor electric drive system in which sensorless motor control is performed by detecting the neutral point potential of the permanent magnet motor. A controller 1K generates PWM signals for controlling a permanent magnet motor 4 based on the detected value of the neutral point potential. The PWM signals are supplied to an inverter 3 which controls the permanent magnet motor on the basis of the PWM signals.A virtual neutral point circuit 100 is connected in parallel with the permanent magnet motor 4. In order to detect the neutral point potential of the permanent magnet motor 4, a virtual neutral point potential Vnn is taken from the virtual neutral point circuit 100. A voltage dividing circuit 2 is provided to detect the neutral point potential Vn of the permanent magnet motor 4 using the virtual neutral point potential Vnn as a reference. The voltage-divided potential Vin generated by the voltage dividing circuit 2 is supplied to an A / D converter of the controller 1K via an isolation amplifier 101.FIG. 28 section (a) is a figure showing the output waveforms for different phases obtained from the ground line (Ni) of the inverter 3. During normal PWM operation, the output potentials of the three phases sequentially change in this manner. And at this time, the neutral point potential Vn and the virtual neutral point potential Vnn of the permanent magnet motor 4 change as shown in Section (b) of FIG. 28. Accordingly, since the impedances Z 3 of the virtual neutral point circuit 100 are the same, Vnn may take one of the four values VDC, (2 / 3) VDC, (1 / 3) VDC, and 0 depending on the switching state. Here, VDC means the DC voltage value of the inverter DC power supply 31.On the other hand, Vn changes substantially similarly to Vnn because the impedances of the three windings are the same for the three phases. However, the magnetic flux of the magnets of the permanent magnet motor 4 has an influence, and the inductance values of the three phases are somewhat changed thereby. Consequently, the inductance values of the three phases are non-uniform and depend on which phase (position angle) the rotor is located in, so that the value of Vn changes. The difference between Vn and Vnn as such represents information about the position of the rotor, whereby position determination without a sensor is possible. It is thus necessary to input the differences between the signals Vn and Vn to the controller 1K. For implementation, the neutral point potential is measured using the virtual neutral point Vnn as a reference.DE 10 2007 030 748 A1 describes a driving apparatus for an electric motor, comprising: an inverter that causes a plurality of switching elements to perform an ON / OFF operation and that converts a DC voltage from a DC power supply into an AC voltage for driving a three-phase synchronous motor; a neutral point potential detection unit that detects a neutral point potential of a stator winding of the three-phase synchronous motor; and a control unit that estimates a rotor position of the three-phase synchronous motor based on the detected neutral point potential and that controls the inverter based on the estimation result; wherein a ground potential of the control unit is set to the negative potential or the positive potential of the DC voltage supplied to the inverter; the neutral point potential detection unit detects the neutral point potential with respect to the negative potential or the positive potential; and the control unit estimates the rotor position based on a difference between a first neutral point potential detected by the neutral point potential detecting unit during the ON / OFF operation and a first fixed reference potential, and based on a difference between a second neutral point potential detected by the neutral point potential detecting unit during the ON / OFF operation and a second fixed reference potential.However, since the reference potential largely changes up and down depending on the circuit state of the inverter, it is accordingly essential that the potential in JP 2010-74898 A described above is input to the controller such as a microcomputer or the like via an isolation amplifier. Usually, the voltage level at the input of an A / D converter of a controller such as a microcomputer or the like is about several volts, and it must further be ensured that at least the ground of the control circuit is constant. For a motor use using a comparatively high voltage such as 100 V and above, it is indispensable that an isolation amplifier be used.An isolation amplifier is a component that is comparatively expensive, so that the cost is increased. Since the time for detecting the neutral point potential is an interval of an extremely short pulse shape, high responsiveness is required for the buffer amplifier itself. However, it is not possible to obtain a high-performance isolation amplifier at a low price, so that it is difficult to design the above-described method without any modification for application to an electric consumable such as a washing machine or a refrigerator or the like or also for application in a typical industrial use. From the above, the object of the present invention is to specify a control device for an electric motor which is inexpensive to produce and nevertheless avoids the problems of the prior art.SOLUTION OF THE OBJECTAccording to the present invention, this object is achieved by a control device for an electric motor having the features of claim 1. Dependent claims are directed to features of preferred embodiments of the invention.ADVANTAGEOUS EFFECTS OF THE INVENTIONAccording to the present invention, it is possible to detect the neutral point potential of a three-phase synchronous motor with high accuracy without using an isolation amplifier, and sensorless control can be realized at sinusoidal alternating currents from the very low speed region in the region of zero.BRIEF DESCRIPTION OF THE DRAWINGSFIG. 1 is a block diagram showing the structure of a driving apparatus for an electric motor according to a first embodiment of the present invention. FIG. 2 is a block diagram showing details of a position estimator / motor controller 11 shown in FIG. 1 ; FIG. 3 is a figure for explaining a voltage vector; FIG. 4 is a figure showing the relationship between the voltage vector and a PWM waveform actually output from an inverter 3; FIG. 5 is a figure showing the change of a neutral point potential; FIG. 6 is a waveform diagram illustrating the operation of various units; FIG. 7 is a block diagram showing the structure of a driving apparatus for an electric motor according to a second embodiment of the present invention; FIG. 8 is a figure for explaining a signal Vin2 (neutral point potential) obtained by switching an analog switch 6; FIG. 9 is a figure showing an example of an analog amplifier 5a; FIG. 10 is a block diagram showing the structure of a driving apparatus for an electric motor according to a third embodiment of the present invention; FIG. 11 is a block diagram showing the structure of a driving apparatus for an electric motor according to a fourth embodiment of the present invention; FIG. 12 is a block diagram showing the structure of a driving apparatus for an electric motor according to a fifth embodiment of the present invention; FIG. 13 is a block diagram showing the structure of a driving apparatus for an electric motor according to a sixth embodiment of the present invention; FIG. 14 is a figure showing the block diagram of internal processing by a digital controller 10F; FIG. 15 is a block diagram showing the structure of a driving apparatus for an electric motor according to a seventh embodiment of the present invention; FIG. 16 is a block diagram showing the structure of a driving apparatus for an electric motor according to an eighth embodiment of the present invention; FIG. 17 is a flowchart explaining the procedure for calculating an offset amount; FIG. 18 is a figure showing a waveform during offset adjustment; FIG. 19 is a block diagram showing the structure of a driving apparatus for an electric motor according to a ninth embodiment of the present invention; FIG. 20 is a flowchart showing an offset adjustment method performed in the ninth embodiment; FIG. 21 is a figure showing voltage instructions Vu*, Vv*, and Vw* during offset adjustment; FIG. 22 is a figure showing an integrated type electric motor system 41 according to a tenth embodiment of the present invention; FIG. 23 is a figure showing a pump system according to an eleventh embodiment; FIG. 24 is a figure showing a compressor system according to a twelfth embodiment; FIG. 25 is a figure showing a position determination system according to a thirteenth embodiment; FIG. 26 is a figure in which the positive potential of a DC voltage (the positive potential of a DC power supply 31) is used as the ground Ni; FIG. 27 is a figure showing an example of a conventional synchronous motor drive system; and FIG. 28 is a figure showing the driving waveform of this conventional synchronous motor driving system.DESCRIPTION OF EMBODIMENTSHereinafter, embodiments of the present invention will be explained with reference to the figures. Note that a driving apparatus for an electric motor according to the present invention is applied to a rotation speed control of a fan, a pump (a hydraulic pump or water pump), a compressor, a washing machine, a spindle motor, a disk drive, or the like, a positioning device of a conveyor or a machine tool, or an application in which a torque is controlled, for example, an electric assist system, or the like.- First Embodiment -

[0017] FIG. 1 is a block diagram showing the structure of a driving apparatus for an electric motor according to a first embodiment of the present invention. The driving device for an electric motor is a device intended to drive a permanent magnet motor 4 (this being a three-phase synchronous motor). The electric motor driving apparatus 1000 of this embodiment (also for the other embodiments) includes a controller 1, a voltage dividing circuit 2, and an inverter main circuit 32, and an output pilot 33 provided to an inverter 3. It is noted that the driving device for an electric motor may also comprise a DC power supply 31. It can also be considered that the inverter 3 includes the DC power supply 31.

[0018] The DC power supply 31 is a power supply that supplies a DC voltage to the inverter main circuit 32. The inverter main circuit 32 is an inverter circuit including six switching elements Sup to Swn. MOSFETs or IGBTs or the like are used for the switching elements Sup to Swn. The output pilot 33 is a driver that directly drives the inverter main circuit 32.

[0019] Vin is the voltage-divided value of the neutral point potential of the permanent magnet motor 4 (hereinafter referred to as "motor"), and Vin is input to the controller 1. The controller 1 calculates an estimated value of the position of the rotor of the motor 4 by processing the signal Vin inside the controller 1, and generates PWM signals for driving the permanent magnet motor 4 at a desired rotational speed or torque. This estimation of the rotor position and generation of PWM signals is performed by a position estimator / motor controller 11. Besides the position estimator / motor controller 11, the controller 1 further includes an A / D converter 12, a subtracter 13, a signal changer 14, and reference voltage generators 15 aand 15 b.

[0020] The inverter 3 amplifies the PWM signals of the controller 1 by means of the output pilot 33, and controls the switching elements Sup to Swn of the inverter main circuit 32. the outputs of the inverter 3 are supplied to the three-phase stator windings (stator coils) of the motor 4; thereby, the motor 4 is controlled.The neutral point potential Vn of the stator winding of the motor 4 is lowered by the voltage dividing circuit 2 to the value Vin of the input level of the controller 1. The neutral point potential Vin is supplied to the controller 1. For this connection, the input range of the controller is assumed to be 0 to Emax. Note that in this embodiment, the ground Nm of the voltage dividing circuit 2, the ground Nc of the controller 1, and the ground Ni of the inverter 3 are all connected to each other.

[0021] The neutral point potential Vin supplied to the controller 1 is discretized by the A / D converter 12 provided inside the controller 1. The reference voltage generator 15a provided inside the controller 1 generates a reference voltage of (2 / 3)Emax, and the reference voltage generator 15b generates a reference voltage of (1 / 3)Emax. The reference voltages (2 / 3)Emax and (1 / 3)Emax are supplied to the subtracter 13 via the signal changer 14 and subtracted from the outputs of the A / D converter 12 by the subtracter 13. The result of this subtraction, i.e., Vin2, is supplied to the position estimator / motor controller 11. Then, depending on the circuit state of the PWM signals output from the position estimator / motor controller 11, the signal changer 14 performs the switching.(The Neutral Point Potential Employing Sensorless Control)

[0022] FIG. 2 is a block diagram showing details of the position estimator / motor controller 11 shown in FIG. 1, illustrating the principal part of the sensorless control block using the neutral point potential. It is to be noted that phase current sensors 30a and 30b for detecting the motor currents are provided, although this device is not shown in Fig. 1. In the example illustrated in FIG. 2, although the motor currents are detected by the phase current sensors 30 aand 30 b, it would be possible to configure so as to detect the current on the DC bus bar (i.e., the current on the DC power supply line).

[0023] The values detected by the phase current sensors 30 aand 30 bare input to the A / D converters 12 cand 12 dprovided in the position estimator / motor controller 11. The A / D converters 12 cand 12 ddisk the detected values of the phase currents Iu and Iw of the motor 4 and input them to the controller. The discretized values Iuc and Iwc of the phase currents Iu and Iw are converted by the coordinate converter 16 into Id and Iq values of the d-q coordinates representing rotation coordinate axes.

[0024] This d-q coordinate conversion per se represents a conventional type of conversion used in vector control of an AC motor, wherein the direction of magnetic flux of the permanent magnet is given by the d-axis, while the direction of torque current orthogonal to the d-axis is the q-axis. The phase angle θdc used in this coordinate conversion is supplied by a position estimator 17. The position estimator 17 is a device that calculates an estimation value of the rotor phase based on the change in the neutral point potential. Note that calculation of an estimated value for the rotor phase may be performed by applying a method as described in Patent Document #1 or the like.

[0025] The differences between the instruction values Id* and Iq* for the current values Id and Iq and the actual current values Id and Iq are calculated by the subtractors 18 aand 18 b. The current controllers 21 and 22 calculate voltage commands Vd* and Vq* so that the differences calculated by the subtractors 18 aand 18 b, or in other words, the deviation of the actual current values Id and Iq from their command values Id* and Iq* become zero, to perform the control. The voltage commands Vd* and Vq* output from the current controllers 21 and 22 are converted to three-phase values by an inverse d-q converter 23, and then converted to pulse-width modulated signals by a PWM generator 24.

[0026] Note that Id* and Iq*, i.e., the instruction values for Id and Iq, are output from an Id* generator 20 and an Iq* generator 19, respectively. While the Id* generator 20 normally outputs "zero" in a rodless machine type motor, when the motor is a salient pole machine, it is controlled to a negative value depending on the load. Although the Iq* generator 19 is a device that indirectly outputs a torque command and is described in FIG. 2 as being included in the position estimator / motor controller 11, it may be provided by a higher-level control unit such as a speed controller or a position controller or the like.

[0027] Next, an overview of a sensorless position algorithm using the neutral point potential of a permanent magnet motor will be given. In total, eight voltages (i.e., switching patterns) can be output from the inverter 3. For example, when the state of the arm for each phase of the inverter main circuit 32 is represented as "1" when the upper switching element is ON and the lower switching element is OFF, and is represented as "0" when the upper switching element is OFF and the lower switching element is ON, representations in which all phases are combined into one vector are obtained as shown in FIG. 3( a).

[0028] V(1,0,0) and so on in FIG. 3(a) indicate the switching states of the switching elements of the inverter 3: V(1,0,0) means that for the phases in the order of U, V and W, the upper element is AN, the lower element is AN and the lower element is AN. Including the two zero vectors V(0,0,0) and V(1,1,1), the inverter 3 can output a total of eight voltage patterns. This type of vector representation is obtained by converting the switching states into α-β coordinates. The relationship between these vector representations and the coordinate axes of the permanent magnet motor 4 is shown in FIG. 3(b). The d-q coordinate rotates counterclockwise, as opposed to the output voltage vectors located in a fixed coordinate system, because the d-q coordinates are rotating coordinates.

[0029] FIG. 4 is a figure showing the relationship between these vectors (also referred to as "voltage vectors") and the PWM waveforms actually output from the inverter 3. FIG. 4( a) shows the relationship between the three-phase voltage instructions (the outputs of the inverse d-q converter 23) and a triangular carrier wave. In FIG. 4(b), the output PWM pulses PVu, PVv and PVw are shown. Further, FIG. 4(c) shows the generated voltage vectors. Finally, FIG. 4(d) shows the changing potential generated at the neutral point of the motor 4 when these voltage vectors are used. It is noted that the waveforms in FIG. 4 are schematically depicted to some extent and the voltage instructions for the three phases Vu*, Vv* and Vw* represent DC voltage values. Although these are actually AC voltage values which vary sinusoidally, if the frequencies of their fundamental waves are low and the carrier frequency is sufficiently high, they can be treated as a DC voltage in the manner indicated.

[0030] The voltage vectors output from the inverter 3 include four kinds of vectors including two zero vectors as shown in FIG. 4(c). Further, as shown in FIG. 4( d), changes in the neutral point potential corresponding to the rotor position are outputted at the intervals between the non-zero vectors V( 1, 0, 0) and V( 1, 1, 0). It is assumed that this phenomenon is due to the influence of the magnetic flux on the permanent magnets of the rotor, and that the neutral point potential Vn0 varies due to these differences in the inductances since there are differences between the inductances of the different phase windings. Note that here, the amount of change in the neutral point potential generated when V(1,0,0) is applied is referred to as VnA, whereas the amount of change in the neutral point potential generated when V(1,1,0) is applied is referred to as VnB.

[0031] FIG. 5 shows the actually measured values of the changes VnA and VnB as a function of the rotor phase θd. In the case shown in FIG. 5, the detection circuit is designed to have an input range of 10 bits (1024) with a center value of 512. Note that the dependence on the electrical angle can be confirmed and the position estimation can be performed. The above-described position estimation device 17 calculates an estimated value for the rotor phase on the basis of the change values of the neutral point potential.

[0032] The estimated value θdc for the rotor phase can be obtained by the following calculation, for example. Here, the neutral point potentials VnA and VnB changing as in FIG. 5 are assumed to be two of three AC voltage phases Xu, Xv, and Xw as in the following equation (1), and VnC (=X) is considered to be the neutral point potential of the remaining phase. Here, Xw(VnC) can be derived from the relationship Xu+Xv+X=0.

[0033] Then, Xa and Xb are obtained by performing three-phase to two-phase conversion (α-β conversion) according to the following equation (2) on the three AC voltage values Xu, Xv, and Xw. Using the results, the estimated value θdc of the rotor position θd can be obtained according to the following equation (3). Note that "arctan" in Equation (3) means arctangent.

[0034] The method for generating the neutral point potential Vin will now be explained. As described above, the motor is PWM-controlled by the inverter 3, examples of waveforms being shown in Fig. 6(a). When the three-phase output terminals of the inverter 3 are viewed from the ground potential Ni of the inverter 3, the waveforms viewed take either zero or VDC (the DC value of the inverter 3). VDC is the input-side voltage value of the inverter 3. when the neutral point potential Vn of the motor 4 is viewed from the ground potential Ni, or in other words, when the potential Ni is taken as a reference, this potential changes between 0 and VDC.

[0035] By the voltage dividing circuit 2, a voltage dividing ratio (Z1 / (Z1+Z2)) is set so that a change in the neutral point potential Vn (0 to VDC) falls within the range of 0 to Emax. The range of the neutral point potential (0 to Emax) and the input range of the A / D converter 12 preferably substantially coincide. As a result, the neutral point potential Vin then becomes the waveform shown in Fig. 6(b). Although the neutral point potential Vin should basically take the values 0, (1 / 3)Emax, (2 / 3)Emax, and Emax at four points, actually, as described above, certain amounts of deviation depending on the position (i.e., phase) of the rotor are observed due to the influence of the magnetic flux inside the permanent magnet motor 4.

[0036] The neutral point potential Vin is discretized by the A / D converter 12, and the reference voltage (2 / 3) Emax or (1 / 3) Emax is subtracted therefrom by the subtracter 13. Which of these two reference voltages (2 / 3) Emax or (1 / 3) Emax is selected is determined by the PWM signals.

[0037] The signal SW1 of FIG. 6(c) is the switching signal input from the controller 1 to the signal changer 14. If the switching signal SW1 is 1, the signal changing device 14 switches to its position "1", if the switching signal SW1 is 0, the switch of the signal changing device changes to its position "0". If two or more of the three phases output from the inverter 3 are VDC, then the switch of the signal changer 14 is set to its position "1" and the reference voltage (2 / 3) Emax of the reference voltage generator 15a is selected. In all other cases, or in other words, when the number of phases equal to VDC is one or less, the switch is set to the position "0", and the reference voltage (1 / 3)Emax of the reference voltage generator 15 bis selected.

[0038] As a result, as shown in FIG. 6( d), the change amount Vin 2 of the neutral point potential Vin of the motor 4 can be obtained from the potential (2 / 3) Emax or the potential (1 / 3) Emax. The amount of change Vin2 of the neutral point potential Vin corresponds to the output of the isolation amplifier 101 shown in FIG. 27, thereby enabling estimation of the rotor position and motor control based on the neutral point potential.

[0039] It is to be noted that although in the structure shown in Fig. 1, the negative potential of the DC voltage supplied to the inverter 3 (i.e., the negative potential of the DC power supply 31) is used as the ground Ni of the inverter 3, alternatively, as shown in Fig. 26, it would also be possible to use the positive potential of the DC voltage (i.e., the positive potential of the DC power supply 31) as the ground Ni.

[0040] In this embodiment, as explained above, the voltage dividing circuit 2 serving as the neutral point potential detecting unit for detecting the neutral point potential is included, the ground potential of the controller 1 is set to the negative or positive potential of the DC voltage supplied to the inverter 3, and it is arranged to detect the neutral point potential using the ground potential as a reference. The fixed first reference potential and the fixed second reference potential, which are not affected by the ON / OFF operation of the inverter 3, are generated within the controller 1 as shown in FIG. 1, and it is provided to estimate the rotor position on the basis of the difference between the first neutral point potential and the fixed first reference potential and on the basis of the difference between the second neutral point potential and the fixed second reference potential. In this way, the standard potentials (i.e., the reference potentials) do not change greatly upward or greatly downward depending on the switching state of the inverter, as is the case with conventional devices in which the neutral point potential is detected using a virtual neutral point as a reference.

[0041] Thereby, it is possible to input the detected neutral point potential to the controller 1 such as a microcomputer or the like without interposing an isolation amplifier. Further, since it is possible to do without using an isolation amplifier having some instability in response, it is possible to detect the neutral point potential of a three-phase synchronous motor with high accuracy, and the sensorless drive can be performed with currents having a sinusoidal waveform appropriately at a very low speed region near zero speed.- Second Embodiment -

[0042] FIG. 7 is a block diagram showing the structure of a driving apparatus for an electric motor according to a second embodiment of the present invention. This electric motor driving apparatus shares many components with the electric motor driving apparatus shown in FIG. 1 according to the first embodiment. Specifically, the voltage dividing circuit 2 and the inverter 3 have the same configuration as shown in FIG. 1, while the configuration of the controller 1B is different from that of the controller 1. Note that the description of the internal structural components of the inverter 3 and the motor 4 is omitted.

[0043] In the first embodiment described above, it is arranged that the neutral point potential Vin obtained from the voltage dividing circuit 2 is input to the A / D converter 12 in the controller 1 unchanged, and the difference between its value after the discretizing and a reference voltage is obtained. In the case of such a structure, if the resolving power of the A / D converter 12 is not sufficiently high, the required detection accuracy cannot be secured for the following reasons.

[0044] The amount by which the neutral point potential changes depends on the position of the rotor and is only a few percent of the voltage VDCof the voltage supply for the inverter 3. Also, if the accuracy is improved as the resolution capability of the A / D converter 12 is increased, there is a disadvantage in that the cost is increased thereby. In addition, when a very fast A / D converter having high resolution is used, there is a problem that the electric power consumed by the A / D converter 12 itself undesirably increases. Therefore, in this embodiment, a structure is provided which enables improvement of the detection resolving power without changing the accuracy of the A / D converter 12.

[0045] The controller 1B of FIG. 7 includes an analog amplifier 5 a, reference voltage generators 7 aand 7 b, an analog switch 6, and a digital controller 10B. The digital controller 10B is a digital controller internally equipped with an A / D converter and includes the A / D converter 12 and the position estimator / motor controller 11 of the first embodiment.

[0046] The signal Vin (i.e., the neutral point potential) input from the voltage dividing circuit 2 to the controller 1B is input as an input signal In1 to the input "+" of the analog amplifier 5a. On the other hand, the output value of the reference generator 7a (the reference voltage (2 / 3)Emax) or the output value of the reference generator 7b (the reference voltage (1 / 3)Emax) is applied to the input "-" of the analog amplifier 5a via the analog switch 6. The analog amplifier 5a is a device for amplifying the difference between the signal In1 applied to its input "+" and the signal In2 applied to its input "-" with the result being outputted as the signal Vin0. The signal Vin0 output from the analog amplifier 5a is input to the digital controller 10B.

[0047] It is to be noted that as the analog amplifier 5a, for example, a differential amplifier as shown in FIG. 9 may be used. The differential amplifier shown in FIG. 9 comprises three operational amplifiers 50a, 50b and 50c and resistors 51a, 51b, 52a, 52b, 53a, 53b and 54.

[0048] Switching of the analog switch 6 is effected by a switching signal SW1 input from the controller 1B. The switching signal SW1 corresponds to the signal shown in Fig. 6(b). When the SW1=1 inputted to the analog switch 6, the reference voltage (2 / 3) Emax is applied to the analog amplifier 5a. On the other hand, when a SW1=0 is input to the analog switch 6, the analog amplifier 5a receives the reference voltage (1 / 3) Emax.

[0049] FIG. 8 shows a waveform similar to that shown in FIG. 6 for explaining the signal Vin2 (i.e., the neutral point potential) obtained by switching the analog switch 6. Vin, which is the input signal In1 input to the input "+" of the analog amplifier 5a, changes as shown in Fig. 8(a). This is the potential shown in Fig. 6(b). When the switching signal SW1 shown in FIG. 8(b) is input to the analog switch 6, the output of the analog switch 6, in other words, the signal In2 becomes the input "-" of the analog amplifier 5a as shown in FIG. 8(c). This results in an output Vin0 of the analog amplifier 5a as shown in Fig. 8(d).

[0050] The output Vin0 of the analog amplifier 5a must be compatible with the input range (0 to VADmax) of the A / D converter of the digital controller 10B. It is therefore necessary that the center of the input voltage applied from the analog amplifier 5a to the A / D converter be offset so as to match the center voltage point of the A / D converter input portion. This offset can be realized by setting the potential at the connection point of the resistor 51a (R4) to VADmax / 2. The gain of the analog amplifier 5a should also be adjusted so that the amount of change in the neutral point potential after amplification is kept within the input range of the A / D converter. Although the amount of change of the neutral point potential largely depends on the characteristics of the magnetic circuit of the motor 4, it is assumed that a gain of about 5 to 50 is suitable.

[0051] As described above, according to this second embodiment, similarly advantageous operational characteristics as in the case of the first embodiment can be obtained. In addition, in this second embodiment, it is provided to use the analog amplifier 5 awhich is a differential amplifier and to discretize the signals corresponding to the difference between the neutral point potential and the two reference potentials with the A / D converter of the digital controller 10B after amplification. Therefore, a change in the neutral point potential can be detected with high accuracy even with an A / D converter of the digital controller 10B whose accuracy is not particularly high, whereby it is possible to realize driving at low rotational speeds in a suitable manner even without a position sensor. In a conventional device, such a realization is difficult.- Third Embodiment -

[0052] FIG. 10 is a block diagram showing the structure of a driving apparatus for an electric motor according to a third embodiment of the present invention. This electric motor driving apparatus of the third embodiment shares many components with the electric motor driving apparatus of the second embodiment shown in FIG. 7. In FIG. 10, the voltage dividing circuit 2 and the inverter 3 correspond to those shown in FIGS. 1 and 7, whereas the controller 1B of FIG. 7 is replaced with the controller 1C.

[0053] In the second embodiment described above, the difference value of the neutral point potential Vin is obtained by switching the reference voltage of the analog amplifier 5 awith the analog switch 6; however, in this third embodiment, two analog amplifiers 5 aand 5 bare provided outside the digital controller 10C.

[0054] Instead of providing the analog switch 6, the number of analog amplifiers is increased, thereby making the circuit arrangement simpler in some cases. For example, in a digital controller including a microcomputer or the like, since it is not uncommon to provide an internal A / D converter with a plurality of input channels, this is because this alternative of signal generation for switching a signal is considered more troublesome. Moreover, since, in the context of analog amplifier circuits, there are individual housings in which a plurality of isolation amplifiers are accommodated, the implementation is accordingly simple and easy.

[0055] In the controller 1C shown in FIG. 10, by adding the analog amplifier 5b, the analog amplifier 5a functions as a differential amplifier using as a reference (2 / 3)Emax, whereas the analog amplifier 5b functions as a differential amplifier using as a reference (1 / 3)Emax. The output of the analog amplifier 5a is input to the input channel ch0 of the A / D converter of the digital controller 10C, while the output of the analog amplifier 5b is input to the input channel ch1 of the A / D converter. Depending on the circuit state of the PWM signals, it is determined which of the signals on channel ch0 and channel ch1 is selected and A / D converted by the digital controller 10C. This makes it possible to read a waveform completely equivalent to that shown in FIG. 4 for the second embodiment into the digital controller 10C.

[0056] As explained above, this third embodiment can provide similar favorable operation results to the first and second embodiments described above. In addition, in the third embodiment, it is possible to omit the analog switch 6 shown in FIG. 7 and detect the change of the neutral point potential with high accuracy in a similar manner to the case of the second embodiment, so that it is possible to realize the motor control at very low speeds suitably without a position sensor. In a conventional device, such a realization is difficult.

[0057] In a digital controller comprising a microcomputer or the like, it is not uncommon to provide an internal A / D converter with a plurality of input ports; moreover, there are analog amplifier circuits in which a plurality of individual amplifiers are accommodated in a single case. This makes it possible to easily realize a more favorable circuit having a structure in which two differential amplifiers (i.e., analog amplifiers 5 aand 5 b) are included, which corresponds to two reference potentials.- Fourth Embodiment -

[0058] FIG. 11 is a block diagram showing the structure of a driving apparatus for an electric motor according to a fourth embodiment of the present invention. This electric motor driving apparatus of the fourth embodiment shares many components with the electric motor driving apparatus of the third embodiment shown in FIG. 10. The voltage dividing circuit 2, the analog amplifiers 5 aand 5 bof the controller 1D, and the digital controller 10C correspond to the elements shown in FIG. 10. In this embodiment, as shown in FIG. 11, a DC voltage dividing circuit 34 is provided in an inverter 3, and a buffer amplifier 8 and a buffer voltage dividing circuit 9 are provided inside the controller 1D.

[0059] In the above-described driving apparatuses for an electric motor of the first to third embodiments, reference voltages for the neutral point potential of the motor 4 are provided independently of each other, respectively, and the change amounts of the neutral point potential are detected with high accuracy. However, in these methods, there is a common problem that the voltage value VDCof the DC power supply 31 for the inverter 3 varies to a quite considerable extent due to a change in the load of the motor 4 or the like. Since the DC power supply 31 is obtained by matching a conventional power supply, the DC power value in an inverter for industrial use or an inverter for a so-called "white goods" type consumer electronic device such as refrigerators, washing machines, changes when the voltage of the conventional power supply fluctuates, also associated therewith.

[0060] As in the driving devices for an electric motor of the first to third embodiments, when reference voltages that do not relate to the supply voltage for the inverter are used, the influence of variations in the voltage of the voltage supply is not reflected in the reference voltages. Accordingly, since the amounts of change in the neutral point potential are very small as described above, a change in the neutral point potential due to a change in the supply voltage can be easily confused with a change due to the position of the rotor, which has an influence on the estimation of the rotor position.

[0061] In order to solve this problem, the DC voltage dividing circuit 34 in this fourth embodiment is configured to detect variations in the DC power supply of the inverter, and it is provided to input the DC power line information for the inverter 3 (hereinafter referred to as "VDC information) to the controller 1D. As shown in FIG. 11, the DC voltage dividing circuit 34 provided on the DC bus line is connected in parallel with the DC power supply 31. The voltage VDCof the DC power supply 31 is divided by the DC voltage dividing circuit 34, and a voltage (a potential at which the ground Ni is used as a reference) determined by the voltage division ratio (Z4 / (Z3+Z4)) is input to the buffer amplifier 8 of the controller 1D as VDC information.

[0062] The voltage serving as VDC information is amplified by the buffer amplifier 8 and divided into the reference voltage (2 / 3)Emax and the reference voltage (1 / 3)Emax by the buffer voltage dividing circuit 9. The reference voltages (2 / 3) Emax and (1 / 3) Emax are input to the analog amplifiers 5a and 5b similar to those shown in FIG. 10, and similar differential amplification is performed.

[0063] In the fourth embodiment, since the reference voltages (2 / 3) Emax and (1 / 3) Emax are thus generated based on the DC voltage on the DC bus of the inverter 3, the reference voltages also decrease accordingly when the power supply voltage drops, for example. In addition, since the voltage-divided value Vin of the neutral point potential also decreases in accordance with the decrease in the supply voltage, it is accordingly possible to detect the output of the differential amplifiers (i.e., the differential amplifiers 5a and 5b) as a "difference". This makes it possible to reduce the influence on the position estimation due to variations in the supply voltage.

[0064] In this fourth embodiment, as explained above, the DC voltage dividing circuit 34 is provided which generates a voltage divided potential by dividing the voltage of the DC power supply 31, and the two reference potentials (the reference potentials (2 / 3)Emax and (1 / 3)Emax) are generated on the basis of the above voltage divided potential by the buffer voltage dividing circuit 9. Thereby, it is possible to detect changes in the neutral point potential of an electric motor with high accuracy without using a single amplifier or an A / D converter or an analog switch having high resolution capability and without errors occurring due to a fluctuation in the power supply of the inverter, so that it is possible to suitably realize very low-speed control without providing a position sensor, which is difficult to realize in a conventional apparatus.- Fifth Embodiment -

[0065] FIG. 12 is a block diagram showing the structure of a driving apparatus for an electric motor according to a fifth embodiment of the present invention. The electric motor driving apparatus of the fifth embodiment shown in FIG. 12 has many components in common with the electric motor driving apparatus of the fourth embodiment shown in FIG. 11 : The voltage dividing circuit 2, the analog amplifiers 5 aand 5 b, and the digital controller 10C correspond to the elements shown in FIG. 11. In this embodiment, a DC voltage dividing circuit 34E is provided to the inverter 3 in place of the DC voltage dividing circuit 34, whereas the buffer amplifier 8 and the buffer voltage dividing circuit 9 used in the structure of Fig. 11 have been omitted.

[0066] For realizing a drive system for an electric motor, it is desirable to design the circuit configuration as simple as possible, and accordingly, by providing such a configuration as described above, a circuit configuration even simpler than the circuit structure of the fourth embodiment can be provided.

[0067] In the DC voltage dividing circuit 34E shown in FIG. 12, voltage dividing points are generated at two points by using three resistors. The voltage dividing ratios are set so that the voltage divided values are (2 / 3) Emax and (1 / 3) Emax. In this embodiment, the two voltage information thus generated by voltage division are to be used as reference voltages from the analog amplifiers 5a and 5b provided in the controller 1E.

[0068] Thus, exactly the same operation as in the fourth embodiment described above is possible. While a buffer amplifier is required in the fourth embodiment, no buffer amplifier is required in this embodiment. Note that, although the wiring is more complicated due to the connections of the voltage dividing resistors, it does not pose a great problem when the present embodiment is applied to a "generator-integrated electric motor system" in which the motor 4 and the electric motor driving device 1000 are integrally constructed. A more important aspect is that this embodiment is preferable from the viewpoint of implementation because the number of components can be reduced.

[0069] As explained above, according to this fifth embodiment, a DC voltage dividing circuit 34E is provided which divides the voltage of the DC power supply 31 and generates two voltage divided potentials as the two reference voltages (2 / 3)Emax and (1 / 3)Emax. Thereby, it is possible to detect a change in the neutral point potential of a synchronous motor with high accuracy by a simple detection circuit without using a single amplifier or an A / D converter or an analog switch having high resolution capability and without errors occurring due to a fluctuation in the power supply of the inverter, so that it is possible to suitably realize very low speed control without providing a position sensor, which is difficult to realize in a conventional apparatus.- Sixth Embodiment -

[0070] FIG. 13 is a block diagram showing the structure of a driving apparatus for an electric motor according to a sixth embodiment of the present invention. The electric motor driving apparatus according to the sixth embodiment shares many components with the electric motor driving apparatus according to the third embodiment shown in FIG. 10 : the voltage dividing circuit 34 shown in FIG. 11 is added to the structure of FIG. 10, and the VDC information from the voltage dividing circuit 34 is applied to the digital controller 10F of the controller 1F.

[0071] The digital controller 10F is provided with three A / D converter input channels ch 0 to ch 2. The outputs of the analog amplifiers 5a and 5b are input to the input channels ch0 and ch1 in a similar manner to the case of Fig. 10, and the VDC information from the voltage dividing circuit 34 is applied to the input channel ch2.

[0072] FIG. 14 is a block diagram showing internal processing of the digital controller 10F. The digital controller 10F includes a position estimator / motor controller 11F, A / D converters 12 ato 12 c, adders 13 aand 13 b, a DC reference value generator 121, a multiplier-divider 122, a "1" generator 123, a subtractor 124, an amplifier G 125, and gain coefficients 126 and 127. Note that in this block diagram, in practice, other than the A / D converters 12 ato 12 c, the constituent elements are expected to be realized in the form of software.

[0073] Before explaining the block diagram of FIG. 14, the influence of the above-described voltage variations of the DC power supply will be described in detail. The output values of the analog amplifiers 5a and 5b can be expressed as given in Equation (4) below. In equation (4), Vn0represents the output of the amplifier, G the gain of the amplifier, Vin the voltage-divided value of the neutral point potential of the motor 4, Vinb the reference voltage, and Eb the bias value during the A / D converter input.

[0074] In the case of the structures shown in FIGS. 11 and 12, the influence of the voltage fluctuations of the DC voltage supply is also determined at the reference potentials. Thus, when the reference value of the DC voltage is referred to as VDC 0 and the actual DC voltage is referred to as VDC, and the amount of change of the voltage is considered, the above-described voltage Vn 0 has the value Vn 1 given by the following formula (5). When the equation (4) and the equation (5) are compared, it is not a problem if the gain varies to some extent.

[0075] However, when the reference voltages are fixed and like the reference voltages output from the reference voltage generators 7 aand 7 bof FIG. 10 do not have a relationship with the DC voltage, the output values of the analog amplifiers 5 aand 5 bare given by the following equation (6) to Vn 2: and the error Ve with respect to Vn 1 of the equation (2) is given by the equation (7):Thus, the proper value of Vn1 can be obtained by correcting Vn2 because Vn1 is given by the following equation (8).

[0076] In the above equations Vinb, since the two reference voltages (2 / 3) means Emax and (1 / 3) means Emax, each of them should be corrected according to equation (7).When equations (7) and (8) are represented as a block diagram, FIG. 14 is obtained.

[0077] In other words, the value of the DC voltage VDCis A / D converted and quantized by the A / D converter 12 cand input to the multiplier-divider 122 as the VDCof the equation (5). The reference value VDC0is also input from the DC reference value generator 121 to the multiplier-divider 122, and VDC / VDC0is calculated by the multiplier-divider 122. Then, VDC / VDC0is subtracted from the value 1 input from the "1" generator 123 by the subtracter 124, and the result is given to the amplifier G125. Accordingly, G·{1-(VDC / VDC0)} is output from the amplifier G125, and the errors Ve given by the equation (7) above are output from the amplifier coefficients 126 and 127. The errors Ve are added by the adders 13a and 13b to the outputs of the A / D converters 12a and 12b (i.e., to Vin2 of Equation (6)). Thereby, neutral point potentials Vin 1 and Vin 2 at which the influence of the voltage variations has been corrected are input to the position estimator / motor controller 11F.

[0078] In this sixth embodiment, as explained above, a DC voltage dividing circuit 34 is provided which generates a voltage divided potential by dividing the voltage of the DC power supply 31, and it is provided to correct a difference signal based on this voltage divided potential which has been A / D converted, so that the influence of the change in the power supply voltage is reduced, and estimate the position of the rotor based on this corrected difference signal. It is thereby possible to correct changes in the neutral point potential of the electric motor with high accuracy with a simple detection circuit, so that no error occurs even in the case of a fluctuation in the supply voltage to the inverter 3. This makes it possible to suitably realize motor control at very low speed without providing a position sensor, which is difficult to realize in a conventional apparatus.- Seventh Embodiment -

[0079] FIG. 15 is a block diagram showing the structure of a driving apparatus for an electric motor according to a seventh embodiment of the present invention. The electric motor driving apparatus according to the seventh embodiment shares many components with the electric motor driving apparatus according to the fifth embodiment shown in FIG. 12, and isolation amplifiers 101a and 101b are added to the structure of FIG. 12. In other words, the outputs of the analog amplifiers 5a and 5b are input to the digital controller 10C via the isolation amplifiers 101a and 101b.

[0080] In the case of the structure shown in FIG. 27 in which the neutral point potential of the permanent magnet motor is detected by using a virtual neutral point potential as a reference, an isolation amplifier is essential because the virtual neutral point potential greatly changes over the range of the power supply voltage. Accordingly, since the reference potential of the virtual neutral point potential (i.e., the virtual neutral point potential) greatly fluctuates, it is necessary to implement countermeasures against noise by giving sufficient attention to the implementation of the circuit so that the isolation amplifier can operate in a stable manner.

[0081] In this embodiment, the reference potential is extremely stable because a fixed value which does not depend on the circuit state of the inverter 3 is used as the reference potential instead of the virtual neutral point potential. For this reason, as shown in FIG. 15, since the isolation amplifiers 101 aand 101 bare used, the countermeasures against noise to be realized are extremely simple, and the result is a detection circuit resistant to noise. Thereby, even if it is not possible to avoid an increase in cost because these isolation amplifiers are provided, it is possible to realize isolation from the digital controller 1G, so that the advantage is great from the viewpoint of safety.

[0082] As explained above, according to the seventh embodiment, it is possible to detect the neutral point potential without using an A / D converter having high resolution capability or comprising an analog switch; moreover, an influence due to the change of the power supply to the inverter or due to noise is hardly noticed, so that it is possible to suitably realize motor control at very low speed without providing a position sensor, which is difficult to realize in a conventional apparatus.- Eighth Embodiment -

[0083] FIG. 16 is a block diagram showing the structure of a driving apparatus for an electric motor according to an eighth embodiment of the present invention. The electric motor driving apparatus according to the eighth embodiment shares many components with the electric motor driving apparatus according to the fifth embodiment shown in FIG. 12. It is obtained by adding phase current sensors 30a and 30b that detect the phase currents of the motor 4. In this embodiment, the distinguishing feature is the structure of the digital controller 10H, and these blocks are mainly realized in the form of software, except for the A / D converters and the parts generating the PWM signals.

[0084] The problem which occurs in practice with respect to the circuit for detecting the neutral point potential of the motor 4 is the fault detection characteristic of analog circuits. In particular, it is necessary to correct the offset value, and it is impossible to calculate an accurate estimate of the mover position based on the neutral point potential when this correction is not performed. As will be explained below, this embodiment differs in a structure related to offset adjustment.

[0085] First, the structure of the controller 1H will be explained. As explained in connection with the fifth embodiment, the A / D converters 12 aand 12 bare devices that discretize the amount of change in the neutral point potential. In addition, the A / D converters 12 cand 12 dare integrated into the controller and discretize the detected values of Iu and Iw of the motor 4.

[0086] The change amount of the neutral point potential changes with respect to the rotor phase θd as explained above, as illustrated in FIG. 5. Although the detection circuit is designed so that, for the waveform of Fig. 5, the center value of the variation component having an input range of 10 bits (i.e., 1024) is 512, it is noted that an offset amount actually exists. If the offset amount is not corrected, there is a problem that an estimation error occurs when an estimated value for the mover position is calculated. In this embodiment, therefore, offset correction is automatically performed.

[0087] Before the actual operation of the magnetic motor 4, an offset compensation step is first performed. Now, the changeover switches 14a, 14b and 14c shown in Fig. 16 are changed over to their "1" position. During normal operation, the changeover switches 14a, 14b and 14c are at the "0" position. Only the changeover switch 14d remains at its "0" position.

[0088] When the changeover switches 14 ato 14 dare brought into this state, a phase command that keeps the rotor fixed is output from the θds generator 26. By switching the changeover switches 14 band 14 cin their "1" position, the command Is generated by the current command generator 28 for position determination and a zero signal generated by the zero generator 29 are input to the adders 18 aand 18 b. In other words, the inputs Iq* and Id* to the adders 18 aand 18 bbecomes Iq*=0 and Id*=Is. Thereby, only the current Id flows through the motor, and the rotor is pulled to the phase where the current Id flows, and stops here. Here, the so-called "DC position determination" is performed.

[0089] After performing the position determination, the changeover switch 14 dis switched to its "1" position, and an instruction (a voltage instruction) is output from the offset adjustment instruction generator 27 so that the voltage vectors V(1,0,0) and V(1,1,0) are alternately applied. In other words, voltage commands Vu*, Vv*, and Vw* such as those shown in FIG. 4( a) are input from the offset adjustment command generator 27 to the PWM generator 24. Thereby, it is possible to obtain the change amount of the neutral point voltage when the rotor is positioned at θd. The neutral point potential is detected while the voltage command outputted from the offset adjustment command generator 27 is changed and the position of the rotor is always slightly changed, thereby obtaining a waveform as shown in FIG. 5. Finally, it is possible to obtain the offset amount from the average value of the waveform shown in FIG. 5.

[0090] FIG. 17 is a flowchart illustrating the calculation of the offset amount. Fig. 18 shows the waveform of offset adjustment. In FIG. 17, in step S 01, θds=0is output from the θds generator 26, the current Id=Is flows, and the rotor is moved, whereby the rotor position is set to the position θd=0. Then, in step S02, by switching the changeover switch 14d to its "1" position, two voltage vectors (for example, the voltage vectors V(1,0,0) and V(1,1,0) shown in Fig. 4) are applied, and the amount of change of the neutral point potential at that time, i.e., at θd=0, is acquired.

[0091] Next, in step S 03, a displacement width θds for varying the value of θds is set. The smaller the value of this displacement width θds, the better the accuracy in the calculation of the offset amount; on the other hand, the longer the time period required for recording the offset amount, the smaller the displacement amount is selected. Therefore, the adjustment width is preferably about 10° to 30° of the electrical angle.

[0092] Now, in step S 04, the displacement width δθ is added to the previous value θds(k-1) of θds, and the position of the mover is set to θds(k). Then, in step S 05, the changeover switch 14 dis switched to its "1" position, and the two voltage vectors are applied, and the amount of change in the neutral point potential at θds(k) is acquired. The recorded change amount of the neutral point potential is stored in a memory provided by the digital controller.

[0093] In step S 06, a decision is made as to whether θds<2π or not. With respect to θds, in step S 01, θds is set to 0 (θds=0) and the displacement width δθ is added each time step S 04 is executed.In the case of θds<2π, an affirmative decision is made in step S 06, and steps S 04 and S 05 are repeated. On the other hand, if θds≥2π, a negative decision is made in step S 06, and control proceeds to step S 07. Since steps S 04 and S 05 are repeated again and again, the amounts of change VnA and VnB for a plurality of rotor positions in the range of 0<θds<2π are stored in the memory.

[0094] In S 07, from the plurality of change amounts VnA that have been taken in, the average value of the plurality of change amounts VnA is calculated using it as the offset amount. Then, in step S08, the average value belonging to the change amounts VnB is similarly calculated and used as the offset amount. The determination of the offset amounts is completed by the method steps shown in FIG. 17.

[0095] FIG. 18 shows the states SWa to SWc of the changeover switches 14 ato 14 c(FIG. 18( a) ), the state SWd of the changeover switch 14 d(FIG. 18( b) ), and the change of the phase θd of the rotor (FIG. 18( c) ) during the determination of the offset amounts. During the offset adjustment interval, the changeover switches 14a to 14c are always at their position "1"; the d-axis current Id continuously flows. The rotor of the motor rotates stepwise in accordance with the value θds generated by the θds generator 26. At the time when the rotor position is fixed, the changeover switch 14 dis switched to its position "1" and VnA and VnB, which are the change amounts of the neutral point potential, are taken in. After the offset adjustment is completed, the system proceeds to the normal control, whereby the changeover switches 14a to 14d are changed over to their normal "0" position.

[0096] According to the eighth embodiment, as explained above, an offset corrector 25 is provided which also corrects the offset component of the neutral potential before starting the motor 4 together with the detection of the neutral potential. Accordingly, since the respective neutral point potentials are detected while the rotor is successively brought into a plurality of rotational positions and the offset component is corrected based on these plurality of neutral point potentials which have been detected, it is possible to correct the offset component and realize motor control without a position sensor with high accuracy at low speed.- Ninth Embodiment -

[0097] A ninth embodiment of the present invention will now be explained. The electric motor driving apparatus according to the ninth embodiment shown in FIG. 19 shares many components with the electric motor driving apparatus according to the eighth embodiment shown in FIG. 16. The digital controller 10J shown in FIG. 19 is obtained by omitting the changeover switches 14 ato 14 c, the θs generator 26, the current command generator 28 for position determination, and the zero generator 29 in the digital controller 10H shown in FIG. 16. In place of the offset adjustment instruction generator 27 of FIG. 16, an offset adjustment instruction generator 27J that operates differently is also provided.

[0098] In the ninth embodiment described above, a neutral-point offset adjustment is performed by bringing the mover to a plurality of positions before the actual operation of the motor 4. This must be a system in which the rotor of the motor is freely movable. As a result, the method explained in connection with the ninth embodiment cannot be applied to the case of a system in which the motor cannot move freely, such as a conveyor belt or a robot arm.

[0099] In this embodiment, on the other hand, it is possible to perform the detection and adjustment of the offset of the neutral point potential detection circuit without moving the rotor of the motor 4. In the case of the structure shown in FIG. 19, in the offset adjustment, the changeover switch 14 dis switched to its "1" position, so that the three-phase voltage commands are supplied from the offset adjustment command generator 27J to the PWM generator 24. The application of the voltage instructions and the recording of the change in the neutral point potential takes place according to the flow chart shown in FIG. 20.

[0100] In step S 11, the offset adjustment instruction generator 27J outputs voltage instructions Vu*, Vv*, and Vw* so that the voltage vectors V(1, 0, 0) and V(1, 1, 0) are applied as in FIG. 4( c). The change amounts VnA and VnB of the neutral point potential generated at this time are detected and stored in the memory. It is noted with respect to the voltage instructions Vu*, Vv*, and Vw* that settings in FIG. 4(a) Vu*=-Vw*, and Vv*=0 are selected because torque is undesirably generated to some extent by the motor 4 when no voltages whose average value is zero are applied.

[0101] Then, in step S 12, voltage commands Vu*, Vv*, and Vw* are output from the offset adjustment command generator 27J, so that voltage vectors V(0, 1, 0) and V(0, 1, 1) are applied as shown in FIG. 21( a). Then, the change amounts (VnC and VnD) of the neutral point potential generated at this time are detected and stored in the memory. Subsequently, in step S 13, voltage commands Vu*, Vv*, and Vw* are output from the offset adjustment command generator 27J, so that voltage vectors V(0, 0, 0, 1) and V(1, 0, 1) are applied as illustrated in FIG. 21( b). Then, the change amounts (VnE and VnF) of the neutral point potential generated at this time are detected and stored in the memory. Finally, in step S 14, the average value of the change amounts VnA to VnF of the detected change amounts of the neutral point potential is calculated, and the result is stored as the offset amount.

[0102] This makes it possible to obtain an alternative value for the offset amount. Even if the detected values VnA to VnF represent information at a coarse resolution of 60°, it is possible to obtain an offset value by such a method. In addition, since the rotor is not moved in the above-described method, the method can be performed accordingly on an electric motor used in a system limited by the above-described condition.

[0103] As explained above, according to the ninth embodiment, there is provided an offset corrector 25 that corrects, before the start of the motor 4, the offset component of the neutral point potential generated when the neutral point potential is detected. It is provided to output a plurality of voltage instructions from the offset adjustment instruction generator 27J, which is the voltage instruction output unit, in order, and correct the offset component on the basis of the neutral point potentials detected during the output of these voltage instructions. Thereby, it is possible to obtain the offset component even for a system in which it is not possible to move the rotor, whereby motor control without a position sensor can be obtained with higher accuracy at low speed.- Tenth Embodiment -

[0104] FIG. 22 is for explaining a tenth embodiment, and shows the general configuration of an integrated type electric motor system 41 in which the electric motor driving device 1000 according to any one of the above-described first to eighth embodiments and the motor 4 are configured in an integrated configuration. FIG. 22( a) is an external perspective view of the integrated type electric motor system 41, whereas FIG. 22( b) illustrates the structure of the integrated type electric motor system 41. The electric motor integrated type system 41 is a device in which the motor 4 and the above-described electric motor driving device 1000 are integrated in a case 410. The housing 410 can also serve as a motor housing for the motor 4; it is alternatively also possible to provide the motor housing and the housing 410 separately from one another.

[0105] As shown in FIG. 22( b), the digital controller 10C is realized as a single integrated circuit, and the inverter 3 is controlled by the PWM pulse waveforms output therefrom. The inverter 3 and the digital controller 10C are realized on a printed circuit board, the wiring for supplying the U, V and W phase currents, and the wiring for detecting the neutral point potential are formed between the printed circuit board and the motor 4. By this integration, the wiring can be accommodated in the case 410. Due to this structure, the only terminals extending outward from the case 410 are the current supply 411 to the inverter and the transmission line 412 used for the speed instruction and the return to the operating state and the like.

[0106] Although it is also necessary in the present invention to take out the neutral point potential of the motor 4, the wiring for the neutral point potential becomes simple when the motor and the drive circuit are integrated in this manner. Accordingly, since sensorless position determination can be performed, it is also possible to provide an integrated system that is extremely compact as a whole and to make this system even more compact.eleventh embodiment -

[0107] FIG. 23 is a figure for explaining an eleventh embodiment, showing a pump system equipped with a driving device for an electric motor 1000 according to the above-described first to eighth embodiments and a motor 4. The pump system shown in FIG. 23 is a hydraulic system including a hydraulic pump 74 and used in a hydraulic transmission or a hydraulic brake system of an automobile. Note that the driving device 100 and the motor 4 may be formed separately from each other, although the pump system in FIG. 23 is realized using the integrated type electric motor system 41 shown in the tenth embodiment.

[0108] In the hydraulic system illustrated in FIG. 23, the hydraulic pump 74 is mounted on the engine 4. The hydraulic pressure in the hydraulic circuit 60 is controlled by the hydraulic pump 74. The hydraulic circuit 60 includes a tank 61 in which hydraulic oil is stored, a relief valve 62 that maintains the hydraulic pressure at or below a predetermined value, a solenoid valve 63 that switches the hydraulic circuit, and a cylinder 64 that serves as a hydraulic actuator.

[0109] The hydraulic pump 74 generates the hydraulic pressure by the operation of the integrated type electric motor system 41 and controls the cylinder 64, which is a hydraulic drive. Since the hydraulic circuit is switched by the solenoid valve 63, the load on the hydraulic pump 74 changes, thereby disturbing the load on the integrated type electric motor system 41. Also, a load several times or more of the steady-state pressure is sometimes applied to the hydraulic circuit, and in some cases, the engine stops, which is undesirable. For this reason, it is intended that an excessively high hydraulic pressure that would cause a high load on the engine be relieved by the relief valve 62.- Twelfth Embodiment -

[0110] Fig. 24 shows a twelfth embodiment of the present invention. FIG. 24 is a figure illustrating an outdoor unit which is a compressor system used in an air conditioner of a room air conditioner or an air conditioner unit. The outdoor unit 80 of the air conditioner includes the above-described driving device for an electric motor 1000 and components such as a motor 4, a compressor 81, a fan, etc. Of these, the motor 4 constitutes the power unit for the compressor 81, which is mounted inside the compressor.thirteenth embodiment -

[0111] FIG. 25 illustrates a thirteenth embodiment of the present invention. FIG. 25 is a figure showing the entire block structure of a position determining system including the motor 4 controlling a position determining means 90. In FIG. 25, the position detector 90 is connected to the motor as a load. A speed controller 91 is connected to the controller 1 and has a function of a higher order generator. The actual rotational speed ωr is subtracted by the subtracter 93 bfrom a rotational speed command ωr* output from a position controller 92. The position controller 92 calculates Iq* so that the value of the difference (i.e., the deviation) becomes zero.

[0112] The position determination device 90 is a device using, for example, a ball screw or the like, and is adjusted by the position controller 92 so as to control the position to a predetermined position 0*. No position sensor is attached to the position determining device 90, and the position value θdc estimated by the controller 1 is used as it is. This makes it possible to execute position control in which it is not necessary to equip the position determination device with a position sensor.

[0113] The above-described embodiments may be realized individually or in combination. The reason for this is that the advantageous effects of each embodiment can be obtained either by itself or in synergistic combination with other embodiments. For example, it would be possible to apply the structure of FIG. 11 or 13 using VDC information or a structure as in FIG. 12 in which the supply voltage is divided to generate two reference voltages to a structure as shown in FIG. 7 in which the analog switch 6 is used. The structure shown in FIG. 16 or FIG. 19 relating to offset correction can also be applied to a driving device for an electric motor 1000 illustrated in a different figure from FIGS. 16 and 19. It is also noted that a structure as in FIG. 15 in which the isolation amplifiers are employed may also be used in an electric motor driving apparatus 1000 illustrated in a different figure from FIG. 15. Thus, in this case, since a high-power isolation amplifier is not required similarly to the case of FIG. 15, it is possible to avoid the cost increase as well as the uncertainty in performance due to the use of an isolation amplifier. In addition, provided that the essential features of the present invention are realized, the present invention is not limited to the above-described embodiments.

Claims

A driving device (100) for an electric motor (4), comprising: an inverter (3) that causes a plurality of switching elements (Sup, Svp, Swp; Sun, Svn, Swn) to perform an ON / OFF operation, and that converts a DC voltage from a DC power supply (31) into an AC voltage for driving a three-phase synchronous motor (4); a neutral point potential detection unit (2) that detects a neutral point potential of a stator winding of the three-phase synchronous motor (4); and a control unit (1) that estimates a rotor position of the three-phase synchronous motor (4) based on the detected neutral point potential and that controls the inverter (3) based on the estimation result; wherein: a ground potential (Nc) of the control unit (1) is set to the negative potential or the positive potential of the DC voltage supplied to the inverter (3); the neutral point potential detecting unit (2) detects the neutral point potential with respect to the negative potential or the positive potential; and the control unit (1) estimates the rotor position on the basis of a difference between a first neutral point potential detected by the neutral point potential detection unit (2) during the ON / OFF operation and a first fixed reference potential ((2 / 3) Emax), and on the basis of a difference between a second neutral point potential detected by the neutral point potential detection unit (2) during the ON / OFF operation and a second fixed reference potential ((1 / 3) Emax), wherein the control unit (1) includes: a selection unit that selects one of the first reference potential ((2 / 3) Emax) and the second reference potential ((1 / 3) Emax) in association with the ON / OFF operation of the plurality of switching elements (Sup, Svp, Swp; Sun, Svn, Swn); a differential amplifier (5a, 5b) to which the first and second neutral potentials detected by the neutral point potential detection unit (2) are input as a first input signal and a reference potential (Emax) selected by the selection unit is input as a second input signal, and which amplifies and outputs the difference between the first input signal and the second input signal; and an A / D converter (12a, 12b) which analog-to-digital converts the output signal of the differential amplifier (5a. 5b); wherein the mover position is estimated based on an output of the A / D converter (12a, 12b), and wherein the driving device (100) further comprises: a DC voltage dividing circuit (2) that generates first and second voltage divided potentials as first and second reference potentials (Emax) by dividing the voltage of the DC power supply (31); wherein the control unit (1) corrects the analog-to-digital converted difference signal based on the voltage divided potential so as to reduce the influence of variations of the power supply (31) on the difference signal, and estimates the mover position based on the corrected difference signal.The driving device for an electric motor (4) according to claim 1, wherein the control unit (1) comprises: a first differential amplifier (5a) that generates a differential signal for the difference between the first neutral point potential and the first reference potential ((2 / 3) Emax), and that amplifies and outputs this differential signal; a second differential amplifier (5b) that generates a differential signal for the difference between the second neutral point potential and the second reference potential (1 / 3) Emax), and that amplifies and outputs this differential signal; a first A / D converter (12a) that analog-digital converts an output signal of the first differential amplifier (5a); and a second A / D converter (12b) that analog-digital converts an output signal of the second differential amplifier (5b); and wherein the mover position is estimated based on the outputs of the first and second A / D converters (12a, 12b).The driving device (100) for an electric motor (4) according to any one of claims 1 or 2, wherein: an isolation amplifier (101a, 101b) is provided between the differential amplifier (5a, 5b) and the A / D converter (12a, 12b); and a signal output from the differential amplifier (5a, 5b) is input to the A / D converter (12a, 12b) via the isolation amplifier (101a, 101b).The driving device (100) for an electric motor (4) according to any one of claims 1 to 3, wherein: the control unit (1) includes an offset correction unit (25-29, 14a-14d) that corrects an offset component included in the neutral point potential before starting the three-phase synchronous motor (4).The driving device (100) for an electric motor (4) according to claim 4, wherein: the offset correction unit (25-29, 14a-14d) brings the mover into a plurality of mover positions in order and detects the corresponding neutral point potentials, and corrects the offset component on the basis of a plurality of detected neutral point potentials.The driving device (100) for an electric motor (4) according to claim 4, wherein: the offset correction unit (25-29, 14a-14d) includes a voltage command output unit (27) that sequentially outputs a plurality of voltage commands and corrects the offset component in outputting each voltage command based on the neutral point potential detected by the neutral point potential detection unit (2).An integrated type electric motor system comprising, housed in a common case (410), a driving device (100) for an electric motor (4) according to any one of claims 1 to 6, and a rotor and a stator of a three-phase synchronous motor (4) driven and controlled by the driving device (100) for an electric motor (4).A pump system comprising: a driving device (100) for an electric motor (4) according to any one of claims 1 to 6; a three-phase synchronous motor (4) driven and controlled by the driving device (100) for an electric motor (4); and a pump (74) for liquids driven by the three-phase synchronous motor (4).A compressor system comprising: a driving device (100) for an electric motor (4) according to any one of claims 1 to 6; a three-phase synchronous motor (4) driven and controlled by the driving device (100) for an electric motor (4); and a compressor driven by the three-phase synchronous motor (4).A position detection system comprising: a driving device (100) for an electric motor (4) according to any one of claims 1 to 6; a three-phase synchronous motor (4) driven (4) and controlled by the driving device (100) for an electric motor; and position detecting means slidably or rotationally driven by forward rotation or reverse rotation of the three-phase synchronous motor (4).

Images