Apparatus and method for controlling brushless DC motor

[0001] This application claims the benefit of Korean Patent Application No. 2003-977, filed Jan. 8, 2003, in the Korean Intellectual Property Office, the disclosure of which is incorporated herein by reference.

[0002] 1. Field of the Invention

[0003] The present invention relates, in general, to an apparatus and method for controlling motors and, more particularly, to an apparatus and method for controlling a brushless direct current motor, which minimizes torque ripple.

[0004] 2. Description of the Related Art

[0005] As is well known to those skilled in the art, a brushless Direct Current (DC) motor employs a rectifying circuit including switching devices instead of mechanical elements, such as a brush and a commutator. The brushless DC motor is characterized in that the replacement of the brush due to abrasion is not necessary, and little electromagnetic interference and drive noise exist. A control apparatus driving the brushless DC motor is described in detail with reference to FIG. 1.

[0006] FIG. 1 is a block diagram of a conventional apparatus to control a brushless DC motor. As shown in FIG. 1, a power converting device including a converter 104, a capacitor 108 and an inverter 106 converts Alternating Current (AC) power supplied from an AC power supply unit 102 to pulse-shaped 3-phase AC power, and supplies the 3-phase AC power to a brushless DC motor (BLDC motor) 110. Of U, V and W phase currents of the 3-phase AC power supplied to the brushless DC motor 110 from the inverter 106, U and V phase currents are detected by first and second current sensors 112a and 112b. The U and V phase current information detected by the first and second current sensors 112a and 112b, respectively, is provided to a control unit 114 and inverter control signals provided to the inverter 106 are based on the detected U and V phase current information. A position and a speed of a rotator of the brushless DC motor 110 are detected by a position/speed detecting unit 116. The detected position/speed information of the rotator is further provided to the control unit 114 and inverter control signals provided to the inverter 106 are based on the detected position/speed information. That is, the control unit 114 controls the speed of rotation of the brushless DC motor 110 with reference to the phase current information inputted from the current sensors 112a and 112b and the rotator position/speed information inputted from the position/speed detecting unit 116. The control unit 114 outputs inverter control signals to control phase commutation times of the 3-phase AC power outputted from the inverter 106 and the intensities of phase currents thereof, thus allowing the rotation speed of the brushless DC motor 110 to comply with a speed command.

[0007] Torque of the brushless DC motor may be expressed by the multiplication of an induced voltage by a current. In phase commutation periods of the 3-phase AC power, phase currents are temporarily decreased, thus generating torque ripple. Since the torque ripple is a cause of noise generation and vibration, a plan to minimize the torque ripple is required.

[0008] To minimize the torque ripple of the brushless DC motor, phase currents temporarily reduced during the phase commutation periods must be provided with compensation. The compensation must be performed only during the phase commutation periods. If the phase commutation periods are not accurately detected, undercompensation or overcompensation of the phase currents occurs. Accordingly, to minimize the torque ripple, accurately detecting the phase commutation periods is important. In the related art, current sensors, such as hall effect sensors, are used to detect the phase commutation periods of the brushless DC motor. However, using the current sensors to detect respective phase currents, as shown in FIG. 1, increases a manufacturing cost of a brushless DC motor control apparatus, thereby imposing a heavy burden to both a manufacturer and a user.

[0009] Accordingly, it is an aspect of the present invention to provide an apparatus and method for controlling a brushless direct current (DC) motor, which accurately detects phase commutation times and periods of 3-phase AC power and compensates for a reduction of currents occurring during the detected phase commutation periods so as to minimize torque ripple due to a temporary reduction of phase currents occurring during the phase commutation periods of 3-phase currents supplied to the brushless DC motor.

[0010] Additional aspects and/or advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.

[0011] The above and/or other aspects are achieved by providing an apparatus for controlling a brushless direct current DC motor, including a power converting unit to convert commercial alternating current (AC) power to polyphase AC power and supply the polyphase AC power to the brushless DC motor. A rotator position/speed detecting unit detects status information of a rotator of the brushless DC motor. A terminal voltage detecting unit detects variations of terminal voltages of the polyphase AC power supplied to the brushless DC motor. A control unit detects phase commutation periods of the polyphase AC power using the status information of the rotator provided from the rotator position/speed detecting unit and the terminal voltage variation information of the polyphase AC power provided from the terminal voltage detecting unit. Further, the control unit controls the power converting unit to supply to the brushless DC motor a compensation voltage used to constantly maintain a mean voltage of a non-commutation phase of the polyphase AC power during the detected phase commutation periods.

[0012] The above and/or other aspects are achieved by a method of controlling a brushless DC motor supplied with power from a power converting unit to convert commercial AC power to polyphase AC power. In the control method, the brushless DC motor is driven through a supply of a non-commutation phase current thereto. A starting time of a phase commutation is monitored using position information of a rotator. A compensation voltage is supplied to the brushless DC motor to minimize torque ripple due to the phase commutation when the phase commutation starts while the phase commutation starting time is monitored. When the phase commutation is complete, the supply of the compensation voltage is stopped.

[0025] Reference will now be made in detail to the present preferred embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to like elements throughout.

[0026] FIG. 2 is a block diagram of an apparatus for controlling a brushless DC motor according to an embodiment of the present invention. As shown in FIG. 2, a power converting device includes a converter 204, a DC-link capacitor 208 and an inverter 206. The power converting device converts AC power supplied from an AC power supply unit 202 to DC power, converts the DC power to 3-phase AC power, and provides the 3-phase AC power to a brushless DC motor 210. The converter 204 converts the AC power to the DC power, and the inverter 206 converts the DC power to pulse-shaped 3-phase AC power.

[0027] A terminal voltage detecting unit 216 detects phase terminal voltages U, V and W of the 3-phase AC power supplied to the brushless DC motor 210 from the inverter 206, and provides terminal voltage information U_C, V_C and W_C of the phases to a control unit 214. A position/speed detecting unit 246 obtains position/speed information F/G of a rotator of the brushless DC motor 210, and provides the position/speed information F/G to the control unit 214. The position/speed detecting unit 246 detects a position of the rotator using a position sensor, or phase commutation information of the phase terminal voltages U, V and W supplied to the brushless DC motor 210. The control unit 214 generates inverter control signals P1 to P6 to control a rotation speed of the brushless DC motor 210 with reference to the terminal voltage information U_C, V_C and W_C of the 3-phase AC power, provided from the terminal voltage detecting unit 216, and rotator position/speed information F/G of the brushless DC motor 210, provided from the position/speed detecting unit 246. The inverter control signals P1 to P6 generated by the control unit 214 are used to control commutation times and current amounts of respective phases U, V and W of the 3-phase AC power outputted to the brushless DC motor 210 from the inverter 206. The inverter control signals P1 to P6 allow the rotation speed of the brushless DC motor 210 to comply with a speed command H inputted from an outside.

[0028] FIG. 3 is a circuit diagram of the 3-phase full-bridge inverter 206 in the brushless DC motor control apparatus of FIG. 2. Referring to FIG. 3, respective switching devices, which are transistors Q1 to Q6, are switched on/off in response to the 3-phase AC signals P1 to P6 each having a 2-phase excitation pattern, which will be described later in FIG. 4, thereby commuting phase currents.

[0029] FIGS. 4A to 4F are graphs showing the inverter control signals to drive the inverter 206 in the brushless DC motor control apparatus of FIG. 2. In this case, FIGS. 4A to 4F show the inverter control signals P1 to P6 which are implemented in a form of the 3-phase AC signals each having the 2-phase excitation pattern and used in a rear end unipolar pulse width modulation. In the rear end unipolar pulse width modulation, a pulse width is modulated by a turn-on/off operation of a single switching device in a latter 600 interval (out-going phase) of a 120° conduction period.

[0030] FIG. 5 is a block diagram of the control unit 214 in the brushless DC motor control apparatus of FIG. 2. As shown in FIG. 5, an adder 502 arranged at an input end of the control unit 214 calculates a difference between the speed command H provided from the outside and speed information G provided from the position/speed detecting unit 246, and provides the difference to a speed control unit 504. The speed control unit 504 generates a first current control signal A to allow the rotation speed of the brushless DC motor 210 to comply with the speed command H in consideration with the difference between the speed command H and the speed information G. The first current control signal A outputted from the speed control unit 504 is provided to a control signal converting unit 510. The control signal converting unit 510 converts the first current control signal A outputted from the speed control unit 504 to internally generate a second current control signal B (not shown in FIG. 5, refer to FIGS. 8A to 8D). The control signal converting unit 510 selects one of the first current control signal A and the internally generated second current control signal B (not shown), and outputs the selected signal as a third current control signal C. In this case, a selection is performed in response to a commutation period detection signal D provided from a commutation period detecting unit 516.

[0031] A commutation time detecting unit 518 detects a time when each phase commutation starts, and provides the detected phase commutation starting time information t_C to the commutation period detecting unit 516. The commutation period detecting unit 516 detects accurate commutation periods using the phase commutation starting time information t_C, provided from the commutation time detecting unit 518, and the terminal voltage information U_C, V_C and W_C of respective phases, provided from the terminal voltage detecting unit 216, and generates the commutation period detection signal D using the detected commutation period information. That is, the commutation period detection signal D is activated only during the phase commutation periods to allow the control signal converting unit 510 to output the second current control signal B as the third current control signal C. During a 2-phase conduction period, the commutation period detecting signal D is inactivated, so the first current control signal A is outputted as the third current control signal C during the 2-phase conduction period. The third current control signal C, outputted to an inverter control unit 514 from the control signal converting unit 510, determines a time ratio of each of the inverter control signals P1 to P6 having the 2-phase excitation pattern, outputted from the inverter control unit 514.

[0032] FIG. 6 is a circuit diagram of the terminal voltage detecting unit 216 of FIG. 2. As shown in FIG. 6, the phase terminal voltages U, V and W supplied to the brushless DC motor 210 and a DC-link voltage V_DC across ends of the DC-link capacitor 208 are inputted to the terminal voltage detecting unit 216. The phase terminal voltages U, V and W, and the DC-link voltage V_DC are each divided in ratios of the resistors R1, R2, R3 and R4 constituting the terminal voltage detecting unit 216, and each pair of corresponding voltages determined as corresponding ratios of the resistors R1, R2, R3 and R4 are compared with each other by respective comparators 602, 604 and 606. The compared results of the pair of corresponding voltages are outputted as the terminal voltage information U_C, V_C and W_C. That is, each of the phase terminal voltages is compared with, for example, ½ of the DC-link voltage V_DC across the ends of the DC-link capacitor 208. If each of the phase terminal voltages is greater than V_DC/2, each terminal voltage information U_C, V_C, and W_C is outputted as a “1” state (a high logic level state), while if each of the phase terminal voltages is less than V_DC/2, each terminal voltage information is outputted as a “0” state (a low logic level state). A principle of obtaining the terminal voltage information U_C, V_C and W_C of the respective phases through the comparison of the phase terminal voltages U, V and W with V_DC/2, and detecting the phase commutation periods using the phase terminal voltage information U_C, V_C and W_C is described below.

[0033] FIGS. 7A to 7C are views showing directions of currents flowing through the inverter 206 during a phase commutation period. FIGS. 7A to 7C illustrate by stages current flows generated during a period in which a U-V phase (0 to 60° interval) is commuted into a U-W phase (60 to 120° interval). In FIG. 7A, a current flow of the inverter 206 in the U-V phase (0 to 60° interval) is depicted. During the 0 to 600 interval, transistors Q1 and Q6 are turned on to allow a U-V phase current to flow through the transistors Q1 and Q6.

[0034] In this state, the transistor Q6 is turned off and the transistor Q2 is turned on, so that the phase commutation from the U-V phase to the U-W phase begins, as shown in FIGS. 7B and 7C. FIGS. 7B and 7C illustrate current flows of an extinguished V phase current 702 and an ignited W phase current 704 during a phase commutation period. In FIG. 7B, transistors Q1 and Q2 are turned on, such that the newly ignited W phase current 704 appears on a negative (−) terminal of the DC-link capacitor 208. In FIG. 7C, the transistor Q1 is turned off and the transistor Q2 continues to be turned on. In this case, the extinguished V phase current 106 flows through a diode D3. Therefore, the V phase terminal voltage V has a same intensity (i.e., magnitude or instantaneous value) as that of the DC-link voltage V_DC until the extinguished V phase current decreases to “0”. After the phase commutation from the V phase to the W phase is complete, the terminal voltage V of the opened V phase terminal is expressed by the following Equations (1),

[0035] $\begin{array}{cc}\begin{array}{c}V=E_v+\frac{V_{DC}}{2}\left(\mathrm{Q1}=\mathrm{ON}\right)\\ V=0\left(\mathrm{Q1}=\mathrm{OFF}\right)\end{array}& \left(1\right)\end{array}$
where E_V is a back electromotive force (EMF) of the V phase. Since an intensity of the back EMF E_V is less than “0” immediately after the phase commutation is complete, the V phase terminal voltage is less than V_DC/2. That is, the terminal voltage V of the opened V phase terminal is greater than V_DC/2 during the phase commutation period, and is less than V_DC/2 after the phase commutation is complete. Therefore, the phase commutation period may be accurately detected by comparing each of the phase terminal voltages with V_DC/2 using a comparison circuit shown in FIG. 6.

[0036] FIGS. 8A to 8D are waveform diagrams showing input/output signals of the terminal voltage detecting unit 216 of FIG. 6 such that respective phase terminal voltages U, V and W are depicted together with the phase terminal voltage information U_C, V_C and W_C thereof. As shown in FIGS. 8A to 8D, each phase terminal voltage information U_C, V_C and W_C is in a “1” state during intervals in which each of the phase terminal voltages U, V and W is greater than V_DC/2. Further, each phase terminal voltage information U_C, V_C and W_C is in a “0” state during intervals in which each of the phase terminal voltages U, V and W is less than V_DC/2.

[0037] In FIGS. 8A to 8D, three phase commutation times #1, #2 and #3, which represent commutation starting times of the W, V and U phases, respectively, are depicted. As shown in FIGS. 7A to 7C, the terminal voltage of an extinguished phase has the intensity of “0” or V_DC until a current of the extinguished phase decreases to “0” after each phase commutation begins. Therefore, a terminal voltage of the phase extinguished immediately after phase commutation takes place is compared with V_DC/2, and a period in which a compared result is maintained at “1” or “0” is a phase commutation period.

[0038] In FIGS. 8A to 8D, after the commutation time #1, the W phase terminal voltage is maintained at V_DC, whereby an output of the comparator 606 of FIG. 6 is maintained at “1” for a time period t_W. The time period t_W is the phase commutation period of the W phase. Further, after the commutation time #2, the V phase terminal voltage is maintained at “0”, whereby an output of the comparator 604 is maintained at “0” for a time period t_V. The time period t_V is a phase commutation period of the V phase. Similar to this, after the commutation time #3, the U phase terminal voltage is maintained at V_DC, whereby an output of the comparator 602 is maintained at “1” for a time period t_U. The time period t_U is a phase commutation period of the U phase. Therefore, the commutation period detecting unit 516 of FIG. 5 recognizes each phase commutation starting time using the phase commutation starting time information t_C provided from the commutation time detecting unit 518, and allows the control signal converting unit 510 to output the third current control signal C capable of preventing voltage variation due to the phase commutation for time periods t_U, t_V and t_W, each representing a period from the phase commutation starting time to a completion of the phase commutation, thus minimizing torque ripple generated during the phase commutation periods.

[0039] FIG. 9 is a block diagram of the control signal converting unit 510 of the brushless DC motor control apparatus of FIG. 5. As shown in FIG. 9, the control signal converting unit 510 outputs one of the first current control signal A provided from the speed control unit 504 and the internally generated second current control signal B as the third current control signal C through a switch 910. The second current control signal B is generated through the following process. A first amplifier 902 with a gain of ¾ amplifies the first current control signal A, and provides an amplified result to an adder 908. A second amplifier 904 with a gain of ½ amplifies a signal A_MAX which is a maximum intensity of the first current control signal A and provides an amplified result to the adder 908. A signal generator 906 generates a signal K_eω/2V_DC, and provides the signal K_eω/2V_DC to the adder 908. In this case, K_e is a back electromotive force constant, ω is a rotation speed of a rotator of the brushless DC motor 210, and V_DC is a voltage across the DC-link capacitor 208. The signal added by the adder 908 is variably amplified by a Variable Gain Control (AGC) amplifier 914 to generate the second current control signal B. During a conduction period, a contact point of the switch 910 is connected to a terminal S1, so that the first current control signal A is outputted as the third current control signal C. Further, during the phase commutation period, the contact point of the switch 910 is connected to a terminal S2 by an activated commutation period detection signal D, so that the second current control signal B to compensate for a mean voltage variation generated during the phase commutation period is outputted as the third current control signal C.

[0040] A gain K of the AGC amplifier 914 that outputs the second current control signal B is controlled in response to a gain control signal E provided from the commutation period detecting unit 516. The gain K is used to prevent a current ripple due to overcompensation generated when the phase commutation is completed within one cycle of a PWM carrier signal. The gain K has a value between 0 to 1, and the value thereof is determined depending on a ratio of a width of the phase commutation period to a cycle of the PWM carrier signal inputted to the inverter control unit 514. For example, if the cycle of the PWM carrier signal is 250 μs and the width of the phase commutation period is 125 μs, the gain K is 0.5. At this time, the second control current signal B is defined by a value obtained by multiplying a time ratio of each of the inverter control signals P1 to P6 used to control the transistors Q1 to Q6 of the inverter 206 by the gain K. If the cycle of the PWM carrier signal is 250 μs and the width of the phase commutation period is 300 μs, that is, if the phase commutation period is longer than one cycle of the PWM carrier signal, the gain K is 1 during one cycle of the PWM carrier signal immediately after the phase commutation, and the gain K is 0.2 during the next cycle of the PWM carrier signal. In this way, if the phase commutation period is longer than the one cycle of the PWM carrier signal, the commutation period detection signal D controls the switch 910 to output the second current control signal B as the third current control signal C during two cycles of the PWM carrier signal. Generally, the time ratio of each of the inverter control signals P1 to P6 has a value between 0 to 1, so a limiter 912 limits the time ratio of the third current control signal C to a range from 0 to 1.

[0041] As described above, to control the time ratio of each of the inverter control signals P1 to P6 is important so as to compensate for the mean voltage variation during the phase commutation period and is described below in detail. First, if the U-V phase is conducting, a mean phase voltage V_M1 applied to each of conducting U and V phases is defined by the following Equation (2),

[0042] $\begin{array}{cc}{V}_{\mathrm{M1}}=\frac{V_{DC}·T_1}{2}\left(0<T_1<1\right)& \left(2\right)\end{array}$
where V_DC is a voltage across the ends of the DC-link capacitor 208 in parallel with the inverter 206, and T₁ is a time ratio of the inverter control signal P6 applied to the transistor Q6 of the inverter 206. A mean voltage V_M2 applied to a non-commutation U phase during a phase commutation period from the U-V phase conduction to a U-W phase conduction is expressed by the following Equation (3),

[0043] $\begin{array}{cc}{V}_{\mathrm{M2}}=\frac{V_{DC}}{3}\left(2T_2-1\right)-\frac{E_U+E_V+E_W}{3}\left(0<T_2<1\right)& \left(3\right)\end{array}$
where T₂ is a time ratio of the inverter control signal P1 applied to the transistor Q1 of the inverter 206 during the phase commutation period from the U-V phase to the U-W phase, and E_U, E_V and E_W are back electromotive forces of U, V and W phases, respectively.

[0044] As indicated in Equations (2) and (3), if the time ratios T₁ and T₂ of the inverter control signals P6 and P1 during the 2-phase conduction period and the phase commutation period, respectively, are the same, the mean voltage V_M1 applied to the non-commutation U phase is temporarily varied to V_M2 due to the phase commutation. The temporary variation of the mean voltage causes ripple in the non-commutation phase current. Torque of the brushless DC motor 210 is proportional to the non-commutation phase current during the phase commutation interval, so the ripple of the non-commutation phase current due to the phase commutation causes a torque ripple. Therefore, to reduce the torque ripple due to the phase commutation, the time ratio T₂ of the inverter control signal P1 provided to the inverter 206 during the phase commutation period is adjusted, as shown in Equation (4), so as to allow a constant current without a current ripple to flow through the non-commutation phase during the phase commutation period, that is, V_M1 to be V_M2.

[0045] $\begin{array}{cc}{T}_2=\frac{1}{2}+\frac{3}{4}{T}_1+\frac{E_U+E_V+E_W}{2V_{DC}}& \left(4\right)\end{array}$

[0046] Provided that an intensity of each of the back electromotive forces E_U, E_V and E_W of U, V and W phases is constant and is proportional to a rotation speed of the brushless DC motor 210 during the phase commutation period, Equation (4) may be simplified as the following Equation (5).

[0047] $\begin{array}{cc}{T}_2=\frac{1}{2}+\frac{3}{4}{T}_1+\frac{K_e·\omega }{2V_{DC}}& \left(5\right)\end{array}$

[0048] Consequently, when a 2-phase excitation type brushless DC motor 210 is controlled, the variation of the mean voltage applied to the non-commutation phase during the phase commutation period must be prevented so as to reduce the current ripple due to the phase commutation. For these operations, the time ratio of each of the inverter control signals P1 to P6 must be controlled in accordance with the above Equation (5) simultaneously with the phase commutation. If the time ratio of each of the inverter control signals P1 to P6 is T₁ during a 2-phase conduction period, the mean voltage variation of the non-commutation phase may be prevented only when the time ratio thereof is changed to a time ratio T₂ simultaneously with a start of the phase commutation and is again changed to a time ratio T₁ at a time when the phase commutation is completed (that is, when the phase current of an extinguished phase becomes “0”).

[0049] If the time ratio thereof is varied to the time ratio T₁ from the time ratio T₂, in advance, before the phase commutation is complete, a reduction of the mean voltage due to the phase commutation may not be of a sufficient compensation (undercompensation), thus causing the current ripple. If the time ratio thereof is continuously maintained at the time ratio T₂ even after the phase commutation is complete, the current ripple is caused due to an overcompensation. Therefore, to minimize the torque ripple due to the phase commutation, the phase commutation period must be accurately detected and the time ratio of each of the inverter control signals P1 to P6 must be maintained at the time ratio T₂ only during the phase commutation period.

[0050] FIGS. 10A to 10D are waveform diagrams showing the inverter control signals to control transistors Q₁, Q₂ and Q₆ and the PWM carrier signal of the brushless DC motor control apparatus of the embodiment of the present invention. In the case where the control unit 214 of FIG. 5 is used to reduce the torque ripple due to phase commutation, the PWM carrier signal must be synchronized with a turn-on time of a newly ignited switching device, as shown in FIGS. 10A to 10D, otherwise, the mean voltage applied to the non-commutation phase is varied at the turn-on time of the newly ignited switching device, thus causing the current ripple. As shown in FIGS. 10A to 10D, when the frequency of the PWM carrier signal is changed so as to synchronize the PWM carrier signal with the turn-on time of the newly ignited switching device, the time ratio of the PWM carrier signal must be maintained at a constant value by suitably adjusting a turn-on time of a PWM pulse. If only the frequency of the PWM carrier signal is changed while the turn-on time of the PWM pulse is constant, an excessive or insufficient input signal is applied to cause the current ripple. In order to synchronize the PWM carrier signal with an ignition time of a new phase, the carrier frequency of the PWM pulse applied to the extinguished phase may be changed immediately before the phase commutation based on the commutation time, as shown in FIGS. 10A to 10D. Further, a method of changing the carrier frequency of the PWM carrier signal by an angle of 60° may also be used.

[0051] FIG. 11 is a flowchart of a method of controlling the brushless DC motor according to the embodiment of the present invention. As shown in FIG. 11, a non-commutation phase current is supplied to drive the brushless DC motor 210 at operation 1102. A starting time of a phase commutation is monitored using position information of a rotator at operation 1104. If the phase commutation starts in operation 1106, a compensation voltage is supplied to the brushless DC motor 210 to minimize a torque ripple in operation 1108. If the phase commutation is complete in operation 1110, the supply of the compensation voltage is stopped.

[0052] As is apparent from the above description, an apparatus and method for controlling a brushless DC motor is provided, which can accurately detect phase commutation times and phase commutation periods of 3-phase AC power and compensate for a reduction of phase currents occurring during the detected phase commutation periods so as to minimize a torque ripple due to the temporary reduction of phase currents occurring during the phase commutation periods of 3-phase currents supplied to the brushless DC motor.

[0053] Although an embodiment of the present invention has been shown and described, it would be appreciated by those skilled in the art that changes may be made in the embodiment without departing from the principles and spirit of the invention, the scope of which is defined in the claims and their equivalents.