Electric motor driver and drive control system thereof

TITLE OF THE INVENTION

ELECTRIC MOTOR DRIVER AND DRIVE CONTROL SYSTEM THEREOF

BACKGROUND OF THE INVENTION

FIELD OE THE INVENTION

The present invention relates to an electric motor driver and a drive control system for controlling a drive of an electric motor such as a brushless DC motor at any rotation number .

DESCRIPTION OF THE PRIOR ART

Recently, m constructing a motor driver for driving an electric motor such as a compressor for use in an airconditioning system, it becomes increasingly necessary to reduce a power consumption in view of global environmental protection. As a power-saving technology in a motor driver, an inverter is widely used for driving a high efficient electric motor such as a brushless DC motor at any rotation number .

Furthermore, as a motor driving technology, a 20 sinusoidal-wave driving technology using a sinusoidal waveform current is remarked because of having a high efficiency and a low noise, in place of a rectangular-wave driving technology using a rectangular-waveform current.

In the case where an electric motor such as a compressor is used in an air-conditioning system, s ince it is difficult to mount a sensor for detecting a rotation position of a rotor of the electric motor, there has been employed a posit ion-sensorless type of a sinusoidal-wave driving technology which drives the electric motor by estimating a rotation position of tlie rotor by a method without using a rotation position sensor. in addition, as a method of estimating a rotation position of a rotor, there is used a known method of estimating an induced voltage of an electric motor. (See, for example, Patent

Document 1: Japanese Patent Laid-Open Publication No. 2000350489)

Fig. 9 illustrates a system construction which implements a position-sensorless type of a sinusoidal wave driving technology. In Fig. 9, reference numeral I designates a DC power supply, 2 designates an inverter, 3 designates a brushless motor, 4 designates a stator, 5 designates a rotor, 6 designates a controller, and reference numerals 7v and 7w designate current sensors. In this construction, the brushless motor 3 comprises the stator 4 and rotor 5, where the stator 4 has three phase windings 4u, 4v and 4w which are Y-connected at a neutral point as their centering point, and the rotor 5 has a magnet mounted thereon. A U-phase terminal 8u, a V-phase terminal 8v and a W-phase terminal 8w are connected to a non-connection end of the U-phase winding 4u, a non*5 connection end of the V-phase windi ng 4v and a nonconnection end of the W-phase winding 4w, respectively.

<td></td><td> The inverter 2</td><td> has three</td><td> series</td><td> circuits</td><td> for</td><td> a IJ phase,</td>

<td></td><td> a V phase and a W</td><td colspan="2"> phase, which are</td><td> connected</td><td> m</td><td> parallel.</td>

<td> 5</td><td> Each of the series</td><td> circuits</td><td colspan="2"> includes a pair</td><td> of</td><td> switching</td>

elements connected in series on the upper and lower streams of a current. A DC voltage output from the DC power supply is applied to the three series circuits. The series circuit for the U phase has a switching element 12u on the upstream side and a switching element 13u on the downstream side. The series circuit for the V phase has a switching element 12v on the upstream side and a switching element

13v on the downstream side. The series circuit for the W phase has a switching element 12w on the upstream side and a switching element 13w on the downstream side.

Furthermore, freewheel diodes 14u, 14v, 14w, 15u, 15v and 15w are connected in parallel to the switching eiements

12u, 12 v, 12w, 13u, 13v and 13w, respectively. The terminals 8u, 8v and 8w of the brushless motor 3 are respectively connected to an interconnection point between the switching elements 12u and 13u, an interconnection point between the switching elements 12v and 13v, and an interconnection point between the switching elements 12w and 13w of the inverter 2.

The currents flowing through the phase windings 4v and

4w in the brushless motor 3 are detected by the current sensors 7v and 7w and the detected current values are supplied to the controller 6. The controller 6 estimates an induced voltage based on the supplied current values and outputs a control signal for controlling the drive of the inverter 2. Thus, the control in driving the brushless motor 3 is performed in the above circuit constitution.

As described above, in the conventional positionsensorless type of a sinusoidal wave driving technology implemented by estimating the induced voltage, at least two current: detectors such as current sensors must be used to detect phase currents of an electric motor to be driven, which increases a cost in constituting a motor driver.

SUMMARY OF THE INVENTION

The present invention is made to solve the above problem and it is an essential object of the present invention to provide an electric motor driver and a drive control system which can detect a phase current with high precision in an inexpensive constitution and excellently drive an electric motor in a low-speed rotation region to a high-speed rotation region, having a high speed stability in the low-speed region.

In order to achieve the object mentioned above, a first aspect of the present invent ion provides an electric motor driver for driving a three-phase electric motor under controlling a drive based on each phase current flowing through each phase winding of the three-phase electric motor .

The electric motor driver includes: an inverter for converter a DC voltage to a three-phase AC voltage to drive the three-phase electric motor; a current detecting means for detecting a current appearing on an inverter busbar to determine each phase current flowing through each phase winding of the three-phase electric motor; an induced voltage estimating means for estimating an induced voltage of the electric motor from the voltage value output of the inverter and the phase current value determined by the current detecting means; a rotation position/speed detecting means for estimating a rotation position and a rotation speed of a rotor of said electric motor based on the estimated induced voltage value; position/speed PI calculating means for calculating rotation position/speed correction values through a proportional-plus-integral (PI) calculation based on the estimated rotation position/speed of the rotor with application of a target speed; a PWM signal generating means for generating a PWM signal for controlling saio inverter based on the correction values obtained by the speed Pl calculating means; and a duty correcting means for correcting a duty of the PWM signal generated by said PWM signal generating means.

In this construction, the duty correcting means corrects the duty without changing the PWM signal during detection of the inverter busbar current by the current detecting means.

Preferably, a speed proportional gain (KP) of the speed PI calculating means is variable in a multistep manner according to a speed in a low-speed region of the electric motor .

A second aspect of the present invention provides a drive control system for controlling a drive of a threephase electric motor. The drive control system includes:

an inverter for converter a DC voltage to a three-phase AC voltage to drive the three-phase electric motor; a current detecting means, which is provided on an inverter busbar, for detecting a current appearing on the inverter busbar to determine each phase current flowing through each phase winding of the three-phase electric motor; an induced voltage estimating means for estimating an induced voltage of the electric motor from the voltage value output of the inverter and the phase current value determined by the current detecting means; a rotation position/speed detecting means for estimating a rotation position and a rotation speed of a rotor of said electric motor based on the estimated induced voltage value; anguiai/speed PI calculating means for calculating angular/speed correction values through a proportional-plus-integral calculation based on the estimated rotation position/speed of the rotor with application of a target speed; a PWM signal generating means for generating a PWM signal for controlling the inverter based cn the correction values obtained by the angular/speed Pl calculating means; and a duty correcting means for correcting a duty of the PWM signal generated by the PWM signal generating means.

In this construction, the corrected duty information of the PWM signals outputted from the duty correcting portion is applied to the current detector so that the current detector determines which phase current of the three phases is appearing on the inverter busbar with reference to the corrected duty in format ion while detecting the inverter busbar current.

BRIEF DESCRIPTION OF THE DRAWINGS

These and other objects and features of the present invention will be readily understood from the following detailed description taken m conjunction with preferred embodiments thereof with reference to the accompanying drawings, in which like parts are designated by like reference numerals and in which:

Fig. 1 is a block diagram showing a system construction of an electric motor driver according to an embodiment 1 of the present invention;

Fig. 2 illustrates an example of time-related variations of phase current conditions of an electric motor;

Fig. 3 illustrates an example of variations of a PWM s igna1;

Fig. 4 illustrates current conditions flowing in the elec trie motor and an inverter when the PWM signal is varied as shown in Fig. 3;

Fig. 5 illustrates an example of variations of the PWM signal;

Fig. 6 illustrates current conditions flowing in the electric motor and an inverter when the PWM sicfnal is varied as shown in Fig .5;

Fig. Ί illustrates an example of variations of the PWM signal;

Fig. 8 illustrates set values of a speed proportional gain KP by a speed PI calculating means; and 20 Fig. 9 is a block diagram showing a conventional example .

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT

Hereinafter, a preferred embodiment of the present invention is described with reference to the accompanying drawings .

(Embodiment 1)

<td></td><td> Fig. 1</td><td> shows</td><td> a system construction</td><td> implementing</td><td> an</td>

<td></td><td> embodiment</td><td> 1 of</td><td colspan="2"> the present invention which is</td><td> a</td>

<td> 5</td><td> combination</td><td> o f an</td><td> electric motor driver</td><td colspan="2"> and an electric</td>

<td></td><td> motor. In</td><td> Dig. 1</td><td> , reference numeia 1 1</td><td> designates a</td><td> DC</td>

power supply, 2 designates an inverter, 3 designates a brushless motor, 4 designates a stator, 5 designates a rotor, and reference numeral 6 designates a controller.

The brushless motor 3 comprises the stator 4 and the rotor

5, where the stator 4 has three phase windings 4u, 4v and

4w which are Y-connected at a neutral point as their centering point, and the rotor 5 includes a magnet mounted thereon and the magnet has magnetic poles N and S. A U15 phase terminal 8u is connected to a non-connection end of the U-phase winding 4u, that is, to the side opposite to the neutral point. Similarly, a V-phase terminal 8v and a W-phase terminal 8w are connected to the non-connect ion end of the V-phase winding 4v and the non-connection end of the

W-phase winding 4w, respectively.

The inverter 2 has three series circuits connected in parallel for the U phase, the V phase and the W phase, each of which includes a pair of switching elements connected in series on the upper and the lower streams of a current. A

DC voltage output from the DC power supply 1 is applied to these series circuits cf the inverter 2 through a voltage detector 16 which detects a voltage value applied to the inverter. The series circuit for the U phase includes a switching element 12u on the upstream side and a switching element 13u on the downstream side.

Similarly, the series circuit for the V phase includes a switching element 12v on the upstream side and a switching element 13v on the downstream side. In addition, the series circuit for the W phase includes a switching element 12 w on the upstream side and a switching element

13w on the downstream side. Furthermore, freewheel diodes

14u, 14v, 14w, 15u, 15v and 1 Sw are connected in parallel to the switching elements 12u, 12v, 12w, 13u, 13v and 13w, respectιvely.

IS Tlie terminals 3u, θ v and 8w of the brushless motoi 3 are connected to an interconnection point between the switching elements 12u and 13u, an interconnection point between the switching elements 12v and 13 v, and an interconnection point between the switching elements 12w and 13w of the inverter 2, respectively.

The DC voltage applied from the DC power supply 1 to the inverter 2 is converted to a three-phase AC voltage by the series circuits comprised of the switching elements in the inverter 2 to thereby drive the brushless motor 3. In order to implement a target speed applied from the outside, a correction value is calculated based on a difference between the target speed and a present speed, and in order for the inverter to output a calculated voltage to be applied to the brushless motor, a PWM signal is generated in the controller 6 for controlling the drive of the switching elements in the inverter 2.

The controller 6 includes a PWM signal generating portion 9 for generating a PWM signal (PWM) to drivecontrol the switching elements of the inverter 2 , and a duty correcting portion 19 for correcting the PWM signal which is converted to a drive signal (DRV) by a base driver for electrically driving the switching elements. Then, the drive signal is applied from the base driver 10 to the inverter 2 to drive the switching elements 12u, 12v, 12w,

13u, 13v and 1 3w.

Further, the controller includes a current detector 11, which is provided on an inverter busbar, for detecting an inverter busbar current to determine a phase current (iu, iv, iw rpresented by is) of the brushless motor 3, and an 20 induced voltage estimating portion 17 for estimating an induced voltage (eu, ev, ew represented by es ) of the brushless motor 3.

Furthermore, the controller 6 includes a rotor position/speed estimating portion 18 for estimating a rotation position in angle Θ m of a polar and a rotation speed com of the rotor of the brushless motor 3, an angle Pl calculating portion 20 for calculating a correction value (CRA) by performing a proper tional-plus-mtegral calculation based on information of the polar position in 5 angle of the rocor, and a speed PI calculating portion 21 for calculating a correction value (CRS) by performing a proportional-plus-Integra1 calculation based on the estimated speed of the rotor.

Next_z an operation of the controller 6 is described in brief below. An induced voltage es=(eu, ev, ew) of the brushless motor 3 is estimated by the induced voltage estimating port ion 17 based on the information of the phase current is=(iu, iv, iw) of the brushless motor 3 aetected by the current detector 11 on the inverter busbar, an output voltage Vs± calculated by the PWM signal generating portion 9, and an inverter applied voltage detected by the voltage detector 16.

Furthermore, a polar position in angle Θ in and a rotation speed ω m of the rotor of the brushless motor 3 are estimated by the rotor position/speed estimating portion 18. The PWM signal for controlling the drive of the brushless motor 3 is generated by the PWM signal generating portion 9 based on correction values (CRA, CRS) .

The first correction value CPA is obtained through a proportional-plus-integral calculation performed by the angle PI calculating portion 20 based on the information 6 m of the polar position in angle of the rotor. The second correction value CRS is obtained through the proper tionalplus-integral calculation performed by the speed Pl calculating portion 21 based on the estimated rotation speed turn with reference to a target speed ω*. Thus, the rotor speed is controlled so as to become the target speed by information such as deviation of the estimated speed ω m

<td></td><td> of the ro tor</td><td> 5 from the</td><td> target</td><td> speed ω</td><td> * applied frcm</td><td> the</td>

<td> 10</td><td> outside .</td><td></td><td></td><td></td><td></td><td></td>

<td></td><td> Then, an</td><td> operation</td><td> o f the</td><td> induced</td><td> voltage estimat</td><td> mg</td>

<td></td><td> portion 17</td><td colspan="2"> is described.</td><td colspan="2"> The current detector</td><td> 11</td>

determines the phase currents iu, iv and iw (represented by is) flowing through the phase windings by detecting the

<td> inverter busbar current,</td><td> and</td><td> the detected phase</td><td> current</td>

<td> values are supplied to</td><td> the</td><td colspan="2"> induced voltage estimating</td>

<td> portion 17. In addition,</td><td> t he</td><td> output voltage Vs+ ca</td><td> lculat ed</td>

<td colspan="3"> by the PWM signal generating portion 9 and a</td><td> voltage</td>

<td> applied to the inverter</td><td> which</td><td> was detected by ihe</td><td> voltage</td>

detector 16 are supplied to the induced voltage estimating portion 17. Thus, the phase voltages (vu, vv and vw) applied to the phase windings are calculated by the induced voltage estimating portion 17 based on the above supplied information .

In theory, the induced voltage values eu, ev and ew (represented by es) induced in the phase windings are calculated from the above values of the phase currents (iu,

<td> iv, iw)</td><td> and the</td><td> phase voltages</td><td> (vu,</td><td> vv, vw) by</td><td> the</td>

<td> following</td><td> equations</td><td> (1) , (2) and</td><td> (3) .</td><td colspan="2"> In the following</td>

<td> equations,</td><td> R is :</td><td> resistance and</td><td> L is</td><td> inductance,</td><td> and</td>

<td> d ( i u) / d t,</td><td> d (i v)/dt</td><td colspan="2"> and d (iw) /dt are time</td><td> differentia</td><td> t ions</td>

<td colspan="3"> of the phase currents iu, iv and iw.</td><td colspan="2"> , respectively.</td><td></td>

<td> eu </td><td> vu - R·iu</td><td> - L·d(iu)/dt</td><td> .. ί 1)</td><td></td><td></td>

<td> ev =</td><td> vv - R·iv</td><td> - L-diiv)/dt</td><td> .. (2)</td><td></td><td></td>

<td> ew ~</td><td> vw - R·iw</td><td> - L d (iw)/dt</td><td> .. (31</td><td></td><td></td>

When the equations (1), ¢2) and (3) are developed in detail, the following equations (4), (5) and (6) are obtained.

eu = vu

- R iu

- (la i La) d ( iu)/dt

- Las-cos (20m)d(iu)/dt

- Las iu d ( cc-s (20m) }/dt + 0.5-La-d(ivl/dt

- Las-cos(20m - 120°) d(iv)/dt

- Las-iv’dfcos (20m - 120°) } / dt + 0.5·Lad(iw)/dt

- Las-cos(20m - 120^Q) d(iw)/dt

- Las-iw-d (cos(20m + 120°) )/dt ... (4) e v = vv

R iv

- (la + La) · ό(iv)/dt

- Las cos (20m + 120°)-d(iv)/dt

- Lasivd(cos (20m + 120°) }/dt + 0.5La·d(iw)/dt

- Lascos (20m) d(iw)/dt

- Las i.w d {cos (20m) } /dt + 0.5-La-d(iu)/dt

<td> - Las·</td><td> cos(20m - 120°) -d(iu)/dt</td>

<td> - Las </td><td> iud{cos (20m - 120°)}/dt ... (5)</td>

<td> ew = vw</td><td></td>

- R iw

<td> - (la</td><td> -t- La) d ( iw) /dt</td>

<td> - Las </td><td> cos (20m - 120°) -d(iw) /dt</td>

<td> - Las </td><td> iw d{cos (20m - 120°)|/dt</td>

+ 0.5La·d I iuJ/dt

<td> - Las </td><td> cos (20m + 120°) -d(iu) / dt</td>

<td> - Las </td><td> iu · d {cos (20m + 120°)}/dt</td>

<td> + 0.5·</td><td> Lad(iv)/dt</td>

<td> - Las </td><td> cos (20m) ·d(iv)/dt</td>

<td> - Las </td><td> iv dίcos (20m) )/dt ... (6)</td>

Here, d/dt designates a time differentiation and an electrical angular speed converted from an estimated speed mm is used as d0/dt appearing in the differential calculation referring to a trigonometric function.

Furthermore, d(iu)/dt, d(iv)/dt, and d(iw)/dt are obtained by a first-order Euler approximation. In addition, the wphase current value iw, the u-phase current value iu and the v-phase current value iv have a relation represented in an equation (14) to be mentioned later. Here, R designates a resistance per a phase of the winding, la designates a leakage inductance per a phase of the winding, La designates an average value of an effective inductance per a phase of the winding and Las designates an amplitude of the effective inductance per a phase of the winding.

In the operation of the induced voltage estimating portion 17, the following equations (7), ¢8) and (9) which simpli fies the equations (4 ) , (5) and (6) are used.

According to the simplification, it is assumed that the phase current values iu, iv and iw are sinusoidal waves and the phase currents ι u, iv and iw are obtained from a current command ampli tude la and a current command phase βΤ and simplified.

eu = vu i R · ia sin (θτη + βΤ) + 1.5-(la + La ) ·cos(9m + βΤ)

- 1.5Las·cos(9m - βΤ) ... (7) ev - vv f R- ia sin (9m + βΤ - 129°) + 1.5·(la + La)-cos(9m + βΤ- 120°)

- 1.5 Las · cos (9m - βΤ- 120°) ... (3) ew = vw + R-ia-sin(0m + βΤ- 240°) + 1.5 - (la + La) - cos (0m + βΤ- 2 4 0°)

- 1.5Lascos (0m - βΤ- 240°) ... (9)

Next, an operation of the rotor position/speed estimating portion 18 is described beiow. The position in angle β ru and the rotation speed ω π of the rotor 5 are estimated using the estimated induced voltage values es=(eu, ev and ew) estimated by the induced voltage estimating portion 17. In specific, an angle 0m estimated by the rotor position/speed estimating portion 18 is corrected using an error of the induced voltage to converge on a true value, whereby the estimated speed mm is obtained.

First, induced voltage reference values (eum, evm and

<td> 15</td><td> ewm) of the phases are obtained by eum = em-sin(0m + βΤ) evm = emsin(0m + βΤ - 120°)</td><td colspan="2"> the following equations.</td>

<td></td><td> ewm - em- sin (0m r βΤ - 240°)</td><td></td><td> (10)</td>

<td></td><td> Here, an induced voltage</td><td> amplitude</td><td> value em is</td>

<td> 20</td><td> obtained by corresponding to the</td><td> amplitude</td><td> values of the</td>

<td></td><td> estimated induced voltages eu,</td><td> e v and</td><td> ew. Then, a</td>

deviation ε between the obtained induced voltage reference values tesrn) and the induced voltage estimated values (es) is calculated.

As shown in the following equation (11), the deviation e is obtained by subtracting the induced voltage reference value esm from the induced voltage estimated value es.

ε = es - esm (i 1) (here, s represents the phases u, v and w)

The estimated angle 9m becomes the true value when the deviation becomes zero. Therefore, the estimated angle 0m is subject to the proportional-plus-integral calculation performed by the angle Pl calculating portion 20 in such a manner that the deviation ε becomes zero and the resultant correction value (CRA) is transmitted from the angle PI calculating portion 20 to the PWM signal generating portion

9. In addition, the est imated speed mm is obtained by calculating a variation value of the estimated angle Om in the rotor position/speed estimating portion 18 and the estimated speed com is transmitted to the speed PI

<td> calculating</td><td> port ion</td><td> 21 .</td><td></td><td></td>

<td> In the</td><td> speed</td><td colspan="2"> PI calculating portion 21, in</td><td> order to</td>

<td> implement a</td><td> targe t</td><td> speed cd,</td><td colspan="2"> using a difference Δω between</td>

<td> the target</td><td> speed</td><td> ω^τ and</td><td> the estimated speed</td><td> com, the</td>

correction value (CRS) is obtained by Equation (12).

ΚΡΔω + ΚΙ'Δω · · (12) (KP : proportional gain, KI : integral gain)

The obtained correction value (CRS} is transmitted from the speed PI calculating portion 21 to the PWM signal generating portion 9. in the PWM signal generating portion

Ιϋ

9, a voltage V* to be output is calculated based on Che correction values (CR?., CRS) from the angle PI calculating portion 20 and the speed PI calculating portion 21. Then a voltage Vs* to be applied to each phase (here s representing phases u, v and w) is obtained fiom the voltage value V* as following Equation (13).

Vu’ = V*-sin(0m + βΤ)

Vv* - V*-sm(Gm + βΤ - 120°)

Vw* = V*-sin(0m + βΤ - 240°) ··· (13)

Furthermore, in order for t.he inverter 2 to output the obtained voltage Vs* (s : phases u, v and w) to be applied to each of the phases, the PWM signal is corrected by the outy correcting portion 19 and outputted as the drive signal (DRV) through the base driver 10 to be applied to each of the switching elements 12 u, ]2v, 12w, 13u, 13v and

13w. In specific, the duty correcting portion 19 corrects a duty of the PWM signal generated by the PWM signal generating portion, without changing the PWM signal during the detection of the inverter busbar current by the current detector 11. Then, each of the switching elements 12u, 12v,

12w, 13u, 13v and 13w are electrically driven by the drive signal (DRV) to generate a sinusoidal alternate current to be applied to each of the phases, thus driving the brushless motor.

According to the present embodiment, the estimated ε between the angle 6m is provided using the deviation estimated induced voltage value es and the induced voltage reference value esm in the rotor position/speed estimating portion 18, and the sinusoidal driving of the brushless motor 3 is thus implemented by flowing the sinusoidal phase currents .

Here, the phase current of the brushless motor 3 appearing in the current flowing on the inverter busbar is described with reference to Figs. 2 through 6. Fig. 2 illustrates a condition of the phase current flowing in each phase winding of the brushless motor 3 every section of electrical angles. In the section of the electrical angles 0 to 60°, it is shown that the phase currents flow from the non-connect ion ends to the neutral point in the (Jphase winding 4u and the W-phase winding 4w and from the neutral point to the non-connect ion end in the V-phase winding 4v.

In tne section of the electrical angles 68 to L20^u, it is shown that the currents flow from the non-connect, ion end to the neutral point in the U-phase winding 4u and from the neutral point to the non-connection ends in the V-phase winding 4v and the W-phase winding 4w. Thus, it is shown that conditions of the phase currents flowing thrcugh the phase windings are varied every electrical angles 60°.

For example, when the PWM signals generated by the PWM signal generating portion 9 at the electrical angle 30° in Fig. 2 vary as shown in Fig. 3 (U, V, W, X, Y and Z are signals for driving the switching elements 12u, 12v, 12w,

13u, 13v and 13w, respectively which are described by an active Hi), there appears no current flow (zero) at a timing (1) , there appears a current flowing through the biphase winding 4w at a timing (2) and there appears a current flowing through the V-phase winding 4v at a timing (3) on the inverter busbar as shown in Fig. 4.

Furthermore, when the PWM signals generated by the PWM signal generating portion 9 are varied at the electrical angle 30° in Fig. 2 as shown in Fig. 5, there appears no current flow (zero) at the timing (1), there appears a current flowing through the U-phase winding 4u at the timing (2) and there appears a current flowing through the

V-phase winding 4v at the timing (3) on the inverter busbar as shown in Fig. 6.

Thus, it can be seen that the phase currents appear on the inverter busbar according to the conditions of the switching elements 12u, 12v, 12w, 13u, 13v and 13w in the inverter 2.

When the currents for the two phases are provided at the adjacent timings as described above, it is obvious that the currents of the respective three phases are obtained by a relation of following Equation (14):

iu+iv+iw=0 ---(14)

However, when the PWM signals generated by the PWM signal generating portion 9 are varied at the electrical angle 30° in Fig. 2 as shown in Fig. 7, there appears no current flow (zero) at the timing (1) and there appears only a current flowing through the V-phase winding 4v at the timing (3) on the inverter busbar. When the PWM signals of such variations are repeated, the currents m the respective three phases are not provided and the induced voltage estimating portion 17 cannot estimate the induced voltage and the brushless motor 3 cannot be driven.

In order to prevent the above problem, in the duty correcting portion 19, when it becomes necessary to detect a phase current flowing each of the phase windings in the brushless motor 3, the PWM signals generated by rhe PWM signal generating portion 9 are checked. In the event that the signal variation is as shown in Fig 7, the generated

PWM signals are corrected to those as shown m Fig. 3, for example .

Furthermore, when the inverter busbar current is shown as an example m Fig. 3, oscillation in the current is generated at the time of switching. In addition, a noise or the like is superposed in the current detector 11. In order to eliminate such influences, the PWM signal generating portion 9 has a low-pass filter or the like (not shown).

Although a speed proportional gain (KP) of the speed

PI calculating portion 21 may be variable in a linear manner accordi^f to a speed in a low-speed region of the electric motor, the stability and responsibility is deteriorated a little when a pulse width is small (at the time of a low speed) because of the above correction and the low-pass filter. Therefore, in a preferred embodiment, the speed PI calculating portion 21 may be adapted to set the values of the speed proportional gain KP m a multisuep manner according to a rotation speed ω as shown in Fig. 8. Thus, an appropriate speed proportional gain KP according to a load can be set to theieby prevent a hunting or an overcurrent from being generated.

When the speed proportional gain KP is increased more than the line (1) in Fig. 8, the control is too sensitive and the overcurrent flows in each of the switching elements in the inverter 2. When the speed proportiona1 gain KP is decreased to be less than the line (2), responsibility is lowered because of shortage of the gain and a hunting may be generated. However, those problems can be prevented by the detailed setting in the multi-step manner.

Furthermore, m a preferred embodiment, the speed PI calculating portion 21 may include a memory 21a such as 25 EEPROM or the like in which the speed proportional gain KP values can be stored in a readable and writable manner to and from the outside of the controller 6. Thus, the values according co a difference of loads caused by a capacity difference or the like of a compressor or the like using a three-phase electric motor are stored in the memory 21a to correspond to each load, whereby an electronic control unit can be shared without changing a ROM of a microcomputer.

In addition, duty information of the fWY signa 1 s output from the duty correcting portion 19 is applied to the current detector 11. While the current detector 11 detects the inverter busbar current, the current detector determines which phase current of the three phases in the brushless motor 3 is appearing on the inverter busbar.

Thus, the current detector 11 determines the respective currents flowing through the three phases based on the inverter busbar current with reference to the duty information. The resultant current values of the three phases are used in the estimate calculation of the induced voltage in the following induced voltage estimating portion .

As described above, by providing the current detector which detects the inverter busbar current and the duty correcting portion 19, the phase current flowing each of tne phase windings in the brush!ess motor 3 can be detected In addition, the values of the speed proportional gain KP are set in a multi-step manner according to a rotation speed by means of the speed PI calculating portion 21, the stability at a low speed region can be improved.

As described above, according to the electric motor driver of the present invention, there can be obtained an effect that a sinusoidal wave drive can be stably

<td> implemented</td><td> at a low</td><td> through</td><td> high speed range</td><td> i r&gt;</td><td> an</td>

<td> inexpensive</td><td colspan="2"> system construction</td><td> without providing</td><td> two</td><td> or</td>

<td> more current</td><td> detectors</td><td colspan="2"> on a wiring between the inver</td><td> ter</td><td> and</td>

<td> the electric</td><td> motor .</td><td></td><td></td><td></td><td></td>

<td colspan="2"> Furthermore, since</td><td> the speed</td><td> proportional gain</td><td> for</td><td> the</td>

speed PI control is stored in a readable and writable memory, various loads (differences of rhe three-phase electric motor or a product to be used) can be corresponded by merely varying the data of the memory, whereby a microcomputer and an electronic controller can be shared.