Patent Application
20080157707
The invention relates to an electric motor and to a method of controlling it.
It is often required of electric motors that they be low in cost and quiet.
It is therefore an object of the invention to make available a novel electric motor and a novel method of controlling it.
This object is achieved, according to the invention, by distinguishing between a current rise subperiod, a current-constant middle subperiod and a current-drop subperiod during each commutation period, and measuring induced voltage only during the current-constant middle subperiod. The fact that current flow occurs with a substantially constant current enables an adaptation of the commutation duration as a function of a sensed voltage signal, in order to adapt the commutation duration to the rotation speed of the rotor. This enables commutation without additional rotor position sensors, and results in an inexpensive motor.
A preferred refinement is to regulate current rise and current drop gradually so that the signal traces form ramps. Because the commutation duration is ascertained, it is possible to carry out the commutation initiation process and the commutation completion process in the form of ramps. This smooth switching-on and shutoff decreases motor noise, and makes possible a quieter motor.
A further preferred embodiment is to calculate current target values in a digital controller, which applies those values to a current regulator, which in turn controls semiconductor switches in series with the windings. With such a method, rotation speed regulation with an electric motor according to the present invention is possible.
According to a further aspect of the invention, the object is achieved by an electric motor with two winding strands which are energized in alternation, with induced voltage being monitored in the currently non-energized winding strand. An electric motor of this kind allows a method according to the present invention to be carried out, and results in a low-cost and quiet motor.
Further details and advantageous refinements of the invention are evident from the exemplifying embodiments, in no way to be understood as a limitation of the invention, that are described below and depicted in the drawings. In the drawings:
The respective upper ends 161, 181 of strands 16 and 18 are connected via lead 20 to link circuit voltage UZK, which can be picked off via a measurement node MP_UZK 24. Link circuit voltage UZK is generated by a power supply 22 from operating voltage +UB, e.g. from an alternating line voltage or from a battery.
The lower end 162 of first strand 16 is connected via a MOSFET 40 and a measuring resistor 42 to ground GND. The potential at the lower end 162 of first strand 16 is picked off via a measurement node MP144. The potential between MOSFET 40 and resistor 42 is picked off via a node 46, and delivered through a lead 50 to a current regulator I_RGL148. Current regulator I_RGL148 is connected via a lead 52 to a microprocessor μC 32 that delivers a target value signal I_SOLL1 to current regulator I_RGL148. Current regulator 48 is connected via a lead 54 to the gate terminal of MOSFET 40 in order to control the latter.
In the same fashion, the lower end 182 of second strand 18 is connected to ground GND via a MOSFET 60 and a measuring resistor 62. The potential at the lower end 182 of second strand 18 is picked off via a measurement node MP264. The potential between MOSFET 60 and resistor 62 is picked off via a node 66, and delivered through a lead 70 to a current regulator I_RGL268. Current regulator I_RGL268 is connected via a lead 72 to microprocessor μC 32 that delivers a target value signal I_SOLL2 to current regulator I_RGL268. Current regulator 68 is connected via a lead 74 to the gate terminal of MOSFET 60 in order to control the latter.
Target current value signals I_SOLL1 and I_SOLL2 are preferably specified as analog voltage signals or as PWM (Pulse Width Modulated) signals.
Microprocessor μC 32 is connected via a lead 80 to a rotation direction indicator circuit “DIR DIG” 82, via a lead 84 to a “U1>0?” circuit 86 for detecting the sign of voltage U1, and via a lead 88 to a “U2>0?” circuit 90 for detecting the sign of voltage U2. Rotation direction indicator circuit “DIR DIG” 82 is connected to measurement node MP144, the “U1>0?” circuit is connected to measurement nodes MP144 and MP_UZK 24, and the “U2>0?” circuit is connected to measurement nodes MP264 and MP_UZK 24.
Operating data such as, for example, a target rotation speed n_s are delivered to microprocessor μC 32 via a bidirectional data bus 92, and the program executing in microprocessor μC 32 controls the rotation speed (n_CTRL), commutation (COMMUT), and input/output (I/O).
Examples of component values:
Rotor 12 is driven by the fact that current flows alternatingly in strands 16 and 18. The current is controlled by MOSFETs 40 and 60, and current regulation takes place by way of current regulators 48 and 68. Electric motor 10 according to the present invention works in sensorless fashion, i.e. no rotor position sensor such as, for example, a Hall sensor, is provided. The rotation direction is determined via rotation direction indicator circuit 82 from the potential at measurement node MP1 or MP2, and commutation (i.e. the alternation between current flow in the first and the second strand) is effected by measuring and evaluating voltages U1 and U2.
Signal U_MP1 picked off via measurement node MP1 is delivered, through resistor 140 and (in order to filter interference voltage spikes) through the low-pass filter constituted by resistor 144 and capacitor 142, to the base of transistor 146. When signal U_MP1 is less than signal UZK, transistor 146 conducts. Conversely, when signal U_MP1 is greater than signal UZK, transistor 146 blocks. When transistor 146 blocks, the base of transistor 156 is pulled to ground, and the latter transistor likewise blocks. Lead 84 is thereby pulled to +5 V, and this means a High signal for μC 32. When transistor 146 conducts, on the other hand, resistors 148 and 154 then act as a voltage divider and raise the potential at the base of transistor 156. Transistor 156 becomes conductive as a result, and lead 84 is pulled to ground GND, which corresponds to a Low signal for μC 32.
The sign of the voltage
U1=U—MP1−UZK
is converted by circuit 86 into a digital signal U1_DIG. When U1>0 V, U1_DIG=High, and when U1<=0 V, U1_DIG=Low. This allows simple evaluation of voltage U1 by μC 32.
The “U2>0?” circuit 90 is preferably constructed in the same fashion.
Examples of component values:
Signal I_IST1 is delivered to the negative input of operational amplifier 174 through a resistor 180. The output of operational amplifier 174 is connected via a resistor 176 to the gate terminal of MOSFET 40. The negative input and the output of operational amplifier 174 are connected via a capacitor 178.
In this exemplifying embodiment, target value signal I_SOLL1 is specified by μC 32 as a PWM signal pwm. The PWM signal is smoothed by the low-pass filter constituted by resistors 162, 166, and 170 and capacitors 164 and 168, and delivered to the positive input of operational amplifier 174. Motor current I1 is measured via base resistor 42, and the potential at node 46 is delivered through resistor 180 to the negative input of operational amplifier 174. Operational amplifier 174 controls the gate terminal of MOSFET 40 via resistor 176, and thus performs a current regulation of current I1 in such a way that the potential at node 46 corresponds to target current value I_SOLL1.
The utilization of an analog current regulator allows the use of a simple μC 32, since the latter needs to carry out only the calculation of target current value I_SOLL1. Alternatively, a digital current regulator can also be used, with which actual current value I_IST1 is delivered to μC 32 in suitable form.
Current regulator “I_RGL2” 68 is preferably constructed in the same manner as current regulator “I_RGL1” 48.
Examples of component values:
In contrast to a “hard” switch-on and shutoff of current I1, switching on and shutting off current I1 in the form of a ramp generates less noise.
Four commutation periods (720° el.) are depicted, which extend between commutation instants t_K1 and t_K2, t_K2 and t_K3, t_K3 and t_K4, and t_K4 and t_K5. In general, the respective first commutation instant for a commutation period will be referred to hereinafter as t_KN, and the respective second commutation instant as t_Kn+1. The commutation duration of the respective commutation periods is referred to as T_K. Only one of stator strands 16 and 18 of winding 15 experiences current flow during a commutation period, so that the direction of the magnetic field generated by the current flow of winding 15 does not change during that commutation period. Currents I1 and I2 flow alternatingly through stator strands 16 and 18.
During each commutation period, one commutation completion operation 107, one operation 108 at substantially constant current flow, and one commutation initiation operation 109 take place. In this exemplifying embodiment, commutation completion operation 107 begins after commutation instant t_KN, and current I1 or I2 rises in the form of a ramp during commutation completion operation 107. The duration of the commutation completion operation is labeled T_KA. Commutation completion operation 107 is followed by a time phase 108 with constant current flow, for a duration T_KK. Following time phase 108 with substantially constant current flow is commutation initiation operation 109, during which (in this exemplifying embodiment) current I1 or I2 is decreased in the form of a ramp until it reaches a value of 0 V. The duration for the commutation initiation operation is labeled T_KE.
Voltages U1 at stator strand 16 and U2 at stator strand 18 can contain, in particular, the following components:
U1=U1—ind+L11*dI1/dt+I1*R1+L12*dI2/dt (1)
U2=U2—ind+L22*dI2/dt+I2*R2+IL21*dI1/dt (2)
where
U1=voltage at stator strand 16
U2=voltage at stator strand 18
U1_ind=voltage induced in stator strand 16 by the rotating permanently magnetic rotor 12
U2_ind=voltage induced in stator strand 18 by the rotating permanently magnetic rotor 12
L11=self−inductance of stator strand 16
L22=self−inductance of stator strand 18
I1=current through stator strand 16
I2=current through stator strand 18
R1=ohmic resistance of stator strand 16
R2=ohmic resistance of stator strand 18
L12=mutual inductance between stator strand 18 and stator strand 16
L21=mutual inductance between stator strand 16 and stator strand 18
When a constant current I1 flows through stator strand 16 and when current I2=0 (time phase 108), the time-dependent terms drop out of equations (1) and (2) and what remains is:
U1=U1—ind+I1*R1 (3)
U2=U2_ind (4)
In the same fashion, what applies when stator strand 18 has a constant current I2 flowing through it, and current I1=0, is:
U1=U1_ind (5)
U2=U2—ind+I2*R2 (6)
For a single-phase, two-strand motor with constant current flow through a first winding strand, the induced voltage U_ind can thus be sensed at the respective winding strand through which current is not flowing. During the commutation operation, on the other hand, such sensing would generally be impossible, or possible only very inaccurately, because of the changing current I1 or I2.
The commutation period between commutation instant t_K3 and commutation instant t_K4 will be considered below. During time phase 108 with constant current flow through stator strand 16, voltage U1 at stator strand 16 is made up, according to equation (3), of induced voltage U1_ind (depicted as line 104) and a magnitude I1*R1 that is constant because current I1 is constant. Voltage U1 therefore does not correspond directly to induced voltage U1_ind. But because current I2 is equal to 0 A, voltage U2 at stator strand 18 corresponds, during time phase 108 of constant current flow, to induced voltage U2_ind, the following being applicable because of the winding 15 selected according to
U1—ind=−U2—ind (7)
Voltages U1 and U2 rise slightly during time phase 108 of constant current flow because motor 10 is configured to generate an auxiliary reluctance torque. For a motor of this kind, induced voltages U1_ind and U2_ind are dependent (for a uniform rotation speed) on the instantaneous rotation angle phi_mech, since stator poles 131, 132, 133, and 134 are configured asymmetrically, as indicated very schematically in
In the exemplifying embodiment of
At the moment of the change in sign of voltage U2, the “U2>0” signal 111 jumps from Low to High. The instant of the change is referred to hereinafter as late commutation instant t_spat.
Commutation duration T_K can be corrected by subtracting late commutation duration T_spät from it. Late commutation duration T_spät is obtained from the time span between late commutation instant t_spät and the commutation instant t_KN+1 predetermined by commutation duration T_K. The correct commutation duration T_K is also obtained directly from the time span between commutation instant t_KN and instant t_spät.
An operation of this kind, in which voltage U1 or U2 assumes, during duration T_KK of constant current flow, a value from an area unsuitable for the selected operating mode of motor 10 for the particular commutation period, is referred to as a late commutation operation.
It is also apparent that voltage U1, and thus also voltage U2, exhibit large disturbances during ramps 107 and 109 because of the regulator and the changes in over time in the current, and said voltages are therefore unsuitable, or only poorly suitable for evaluating induced voltage U1_ind or U2_ind.
Because both I1=0 and I2=0 after commutation initiation operation 109, both U1 and U2 correspond to the induced voltage (cf. equations (3) and (4)). In this exemplifying embodiment, the induced voltage is measured via voltage U2. The latter exhibits a sign change at instant t_früh, and the “U2>0?” signal 111 changes from Low to High at instant t_früh.
Commutation duration T_K can be corrected by increasing it by an amount equal to early commutation duration T_früh. Early commutation duration T_früh is obtained from the time span between early commutation instant t_früh and the commutation instant t_KN+1 predetermined by commutation duration T_K. Instead of commutation instant t_KN+1 it is also generally possible to use the point in time at which commutation initiation operation 109 ends.
An operation of this kind, in which voltage U1 or U2 assumes, at the end of commutation initiation operation 109, a value from an area unsuitable for the device operating mode of motor 10 for the particular commutation period, is referred to as an early commutation operation.
It is also apparent that voltage U1 exhibits large interference spikes during commutation initiation operation 109, making it very difficult or impossible to evaluate the induced voltage during commutation initiation operation 109. The interference during commutation initiation operation 109 also occurs as a result of the work of the current regulator that converts current I1 to the value I1=0 V in a predetermined form.
At instant t_früh, a change in the “U1>0?” signal 112 from High to Low is detected. After this detection, in this exemplifying embodiment, commutation completion operation 107 is performed, i.e. current flow through stator strand 18 begins. Instant t_früh is preferably selected as first commutation instant t_KN+1 for calculating the next commutation instant t_KN+2.
Voltage U1 is preferably measured via two measurement nodes MP1344 and MP2346 that are arranged at the opposite ends of stator strand 316, and at which voltages U_MP1 and U_MP2 are present. Voltage U1 is calculated as
U1=U—MP2−U—MP1 (8).
For the check as to whether a late commutation operation exists, the I1*R1 component is subtracted from voltage U1. Either the target value for the corresponding current regulator can be used as a value of current I1, or it is ascertained by a measuring apparatus for current.
Late commutation area 140 can be defined in different ways. Two late commutation voltage areas 140′ and 140′ are presented as further exemplifying embodiments. In contrast to late commutation area 140, late commutation area 140′ is not open toward the top, but ends at a maximum voltage. This allows, if applicable, a simpler evaluation circuit. Late commutation voltage area 140′, on the other hand, begins not at 0 V but at a negative (or positive) voltage. This can be utilized, for example, for earlier detection of a late commutation operation. Detection occurs here at instant t_140″, which is located earlier in time than instant t_140. A shift of this kind can furthermore, for example, take into account an offset of voltage U2 that can occur in a single-strand motor as a result of component I1*R1.
An early commutation voltage area can be defined in the same fashion for the early commutation operation.
If a rotor standstill is identified in step S276, then in the “START_ROT” step S278 the rotor is caused to move by current flow. The main loop begins in step S286, and a check is made as to whether the rotor is still moving. If this is not the case, execution branches back to step S276. If the rotor is rotating, however, then in the “PERIOD—1” step S288 current flow is performed for the first commutation period. The PERIOD—1 routine is set forth in more detail in
In the “PERIOD—2” step S290, current flow is performed for the second commutation period, i.e. in the opposite direction. In the “n_CTRL” step S292, the rotation speed regulation calculation operation takes place. This is presented in more detail in
In the “OTHER” step S294, further steps necessary for operation of the motor take place. For example, input/output is performed, and error signals are outputted in the event of an error.
After step S294, execution branches back to step S286 and the next current flow takes place.
In a commutation completion operation 107, current I1 is elevated in four steps (N_KA=4) from the value I1=0 A to the value corresponding to target value I_SOLL. This is followed by a time phase 108 during which a constant current flow occurs at the value I1=I_SOLL. Following this is commutation initiation operation 109, during which current I1 is decreased in four steps (N_KE=4), in ramped fashion, from the value I1=I_SOLL to the value I1=0.
In this exemplifying embodiment, duration T_KA of commutation completion operation 107 and duration T_KE of commutation initiation operation 109 are calculated from commutation duration T_K. Commutation completion duration T_KA and commutation initiation duration T_KE are selected so that they each occupy 10% of the total commutation duration T_K. Time phase 108 of constant current flow occupies the remaining 80% of the commutation duration. In general, the values T_KA and T_KE are selected as follows:
T
—
KA=f
—
KA*T
—
K (9)
T
—
KE=f
—
KE*T
—
K (10)
where
T_KA=commutation completion duration
f_KA=proportional factor for the commutation completion duration
T_K=total commutation duration
T_KE=commutation initiation duration
f_KE=proportional factor for the commutation initiation duration.
The proportional factors f_KA and f_KE are preferably adapted to the particular motor type and the particular intended application of electric motor 10, and can be specified to μC 32, for example, by control unit 94 via interface 92 (cf.
If a late commutation operation is not taking place in the time phase of constant current flow, execution branches respectively from step S310 to step S312. Step S312 checks whether the time phase of constant current flow should continue to be implemented. This is done by way of the PHASE_CONST variable, which is previously set to 1 and which, upon expiration of the time entered in timer TIMER1, is set to 0 by the “TIMER1_INTERRUPT” interrupt routine S250 of
If, on the other hand, a late commutation operation is taking place during the time phase of constant current flow, execution then branches from step S310 to the “RESET_TIMER1” step S314. In this step, timer TIMER1 is reset so that an interrupt is no longer triggered. In step S316, late commutation duration T_LATE is then calculated from the difference between the present time t_TIMER and the starting instant of commutation completion operation t_KA. A correction of commutation duration T_K additionally takes place, by subtracting therefrom the late commutation duration T_LATE. Execution thereupon branches to step S320, and the “RAMP1_DOWN” commutation initiation operation S320 is initiated.
After the commutation initiation operation is complete, in step S322 a variable t_KE is set to the present time t_TIMER and an EARLY_COMMUT variable is set to 0. Step S324 checks whether an early commutation operation exists. This is done, for example, by way of voltage U1, and a check is made as to whether said voltage is in the EARLY_AREA early commutation area. In the case of an early commutation operation, execution branches to step S326 and the EARLY_COMMUT variable is set to 1 in order to indicate an early commutation operation. Execution then branches back to S324. As soon as voltage U1 is outside the EARLY_AREA early commutation area, execution branches to step S328. In the case of an early commutation operation, execution branches to step S330, where early commutation duration T_EARLY is calculated from the difference between present time t_TIMER and the time t_KE stored in step S322. A correction of commutation duration T_K is additionally performed, by increasing it by a value equal to early commutation duration T_EARLY. Execution thereupon branches to the end S332.
In step S202, commutation completion duration T_KA is calculated from commutation duration T_K (cf. description of
In step S208 the loop variable i is incremented by 1, and step S210 checks whether all N_KA steps have not yet been carried out. If Yes, execution jumps back to step S204 and the next step of ramp 107 is generated. After all N_KA steps have been carried out, the “RAMP1_UP” routine S200 is ended.
The corresponding “RAMP2_UP” and “RAMP2_DOWN” routines for specifying target value I_SOLL2 for regulator I_RGL268 correspond to routines S304 of
Also delivered to block 408 are a Kommut1 signal from a block 410, and a Ramp signal from a block 412. The Kommut1 signal specifies when current flow is to occur through the first strand; the Ramp signal specifies the ramp shape, which is a function of commutation duration T_K; and the signal from block 406 specifies the amplitude of the ramp-shaped commutation signal occurring in block 408, in order to influence the rotation speed. Block 408 is configured as a multiplier.
The commutation signal generated by block 408 is delivered to a block 414. Block 416 makes available a signal that corresponds to voltage U42 at base resistor 42 of
Blocks 428, 430, 432, 434, 436, 438, and 440 correspond to blocks 408 to 420, and control signal ISte112 for second stator strand 18 is generated therein.
Rotation speed regulation is implemented by the fact that the control output signal of PI controller 404 is delivered to multiplier 408 or multiplier 428, thereby determining the magnitude of the ramp current. A predetermined elevation in target rotation speed n_s would then, for example, cause the signal delivered to multiplier 408 to become greater, which results in a higher target current value and thus a higher motor current I1 or I2. The result is that rotor 12 rotates faster, and an adaptation of commutation duration T_K takes place until the electric motor exhibits a rotation speed n_i corresponding to target rotation speed n_s.
The rotation speed of the motor is thus determined by an interaction between rotation speed controller 404 and current regulator 414.
Many modifications are of course possible within the scope of the invention.
In a simpler configuration, for example, the rotation speed controller in
An electric motor according to the present invention is preferably used to drive and/or decelerate a fan.
Because motor 10 is configured to generate an auxiliary reluctance torque, the rotation direction can be ascertained in area 108 of constant current flow from the slope of voltage U1102, of voltage U1_ind 104, of voltage U2103, and/or of voltage U2_ind (cf.
In the case of the present motor, voltage U1102 is rising, and the derivative of voltage U1 (which corresponds to the slope) is likewise positive. For a rotation in the opposite direction, conversely, the slope or derivative of voltage U1 would be negative.
The rotation direction measurement can be performed at least once after or during startup, or it can also occur at predetermined intervals.
The invention is not limited to the exemplifying embodiments that are depicted and described, but rather encompasses all embodiments that function identically, within the context of the invention.
| Number | Date | Country | Kind |
|---|---|---|---|
| 10 2005 011 263.3 | Mar 2005 | DE | national |
This application is a section 371 of PCT/EP06/01815, filed 28 Feb. 2006, published 8 Sep. 2006 as WO 2006-092 265-A, and further claims priority from German application DE 10 2005 011 263.3, the contents of which are hereby incorporated by reference.
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/EP06/01815 | 2/28/2006 | WO | 00 | 10/3/2007 |
