JPS6260491A - Method of detecting position of rotor for brushless dc motor - Google Patents

Abstract

PURPOSE:To obstruct a passage of a commutation spike to a control circuit, and to drive a motor efficiently at high load torque and high speed by sample- holding an induced-voltage signal, synchronizing with a commutation spike signal. CONSTITUTION:Voltage at the neutral point of armature windings Lu, Lv, Lw for a motor 2 is divided by resistors R1, R2, and inputted to a gate circuit 4. The gate circuit 4 is controlled so as not to be conducted during the generation of a commutation spike by a commutation spike signal from a commutation spike detecting circuit 5. An induced voltage signal passing through the gate circuit 4 is charged to a capacitor C1. Voltage charged to the capacitor C1 is transmitted over a controller 3 for a semiconductor commutator device 1 through an operational amplifier OP1.

Description

Detailed Description of the Invention [Field of Industrial Application] The present invention is directed to a brushless direct current motor (hereinafter referred to as a motor) having a magnetic rotor, which is capable of controlling the induced voltage generated in the armature winding by the rotation of the magnetic rotor. The present invention relates to a method for detecting the position of a magnet rotor from an induced voltage in a control device that is driven using the electromotive force.

[Conventional technology]

Conventionally, in the position detection means of the magnet rotor of a control device that drives this type of motor, the magnetic detection means such as a Hall element is installed inside the motor, but there are problems with structural and manufacturing complexity and heat resistance. In order to solve these problems, I
Many methods have been proposed that utilize the induced voltage generated in a de-energized armature winding as the magnet rotor of a motor rotates.

One of them is to obtain the induced voltage from the neutral point of the star-connected three-phase armature windings of the motor, and to obtain the signal of the amplitude change of the induced voltage by delaying the phase by 90 degrees in electrical angle. A method of obtaining a position detection signal by comparing at the center of the amplitude has been proposed.

An example of this is shown and explained in FIG. 10. FIGS. 11 and 12 are timing charts showing the waveforms of each part in FIG. 10.

1 is a semiconductor commutator device, which includes energization control elements 1a to 1;
f is connected to an 8-phase bridge.

Furthermore, flywheel diodes H to D6 are connected in antiparallel to each of the energization control elements.

2 is a motor, and 3 is a control circuit.

Now, each energization control element of the semiconductor commutator device 1 is controlled to sequentially energize 120° of electrical angle and cut off electrical angle of 240° by the drive signal shown by Va - VT in FIG. 11, thereby driving motor 2. Then, due to the rotation of the magnet rotor of the motor 2, the armature winding Lu of the motor 2 is de-energized.
, Lv, and Lw, an induced voltage is generated in sequence.

Therefore, at the neutral points of the armature windings Lu, Lv, and Lw, an induced voltage appears around the voltage higher than the voltage supplied to the semiconductor commutator 1.

The induced voltage is divided by resistors 1'L, , R2 to obtain an induced voltage signal Vo.

The phase of the induced voltage signal V is delayed by 90 degrees through capacitor integration in terms of amplitude to obtain a soft signal V:, and the shift signal V is compared at the center of the amplitude and logicalized to form a logical position detection signal V. , is obtained.

As shown in FIG. 11, the rising and falling timings of the logical position detection signal V coincide with the timing of energization of each armature winding of the motor 2, so that the logical position detection signal V By appropriately processing V, it is possible to obtain the drive signal Va - Vf necessary to continuously rotate the motor 2, and as a result, the motor 2 is operated in a feedback manner.

[Problem that the invention seeks to solve]

However, the case where the motor can be operated by the signal processing described below is when the load applied to the motor 2 is light, and when the load is heavy and the current flowing through the motor 2 increases, the 6 The commutation spike that occurs each time the energization control element is turned off becomes conspicuous, making it impossible to obtain a normal logical position detection signal V9.

For example, in the energization command state of the drive signals Va and vr, the power supply from the semiconductor commutator device 1 to the motor 2 is as follows: energization control element 1a → armature winding Lu → neutral point N → armature winding Lw - energization This is done through the control element if.

Next, when the energization pattern of the semiconductor commutator device I changes, the energization control element 1a is cut off, and the energization control element 1a is switched off.
b becomes energized.

Therefore, the energization pattern at this time is the energization control element 1b-
The path is armature winding Lv - neutral point N - electric machine child winding Lw -> energization control element 1.

However, the armature winding Lu, which was in the energized state immediately before the energization pattern changed, generates a back electromotive force, generally called a commutation spike, in order to release the inductance energy stored by the energization after the energization is cut off. , armature winding Lu → neutral point N → armature winding LW → energization control element 1f →
The flywheel diode D4- forms a closed circuit called the winding Lu to allow a return current to flow.

In this example, the energization control element on the upper arm side of the semiconductor commutator device 1 is cut off, but in the case where the lower arm side is cut off, for example, from the energization state of the energization control elements 1b and 1f, the energization control element 1b and 1f are cut off. In the case of the energization pattern 1d, the armature winding Lw is de-energized, so an electromotive force is generated in the armature winding L%V, and the armature winding Lw→flywheel diode D3- is energized. A return current is passed through a closed circuit of control element tb→armature winding Lv→neutral point→armature winding Lw to release inductance energy.

Similarly, in other energization patterns, a commutation spike occurs at the moment when the conduction of the energization control element is interrupted until the inductance energy of the armature winding whose energization is interrupted is released.

As mentioned above, the armature winding in which current is cut off (this should generate an induced voltage due to the rotation of the magnet rotor), but the energy stored by energizing the armature winding is A commutation spike occurs due to the release of , and the back electromotive force of the commutation spike appears at the neutral point.

This commutation spike disappears once the stored energy is released from the armature winding, but the longer the stored energy, ie the higher the current flowing through the armature winding, the longer it will last.

Now, assuming that each armature winding has an inductance of Lx, the energy W stored in the armature winding of each phase is W, =-HLxi'...Song/Song...
Equation 1 (1 is the current flowing through the armature winding) is obtained, and increases in proportion to the square of the current flowing through the armature winding.

Furthermore, in this type of motor, the current flowing through the armature is approximately proportional to the load torque applied to the motor.

Therefore, as the load torque of the motor 2 increases, the width of the commutation spike increases as shown in the induced voltage signal v in FIG. 12, and as a result, the area where the induced voltage is generated becomes narrower.

When capacitor integration is performed on the induced voltage signal VO in such a state to shift the phase by 90°, the shift signal ■= is affected by commutation spikes near the center of the amplitude, as shown in Figure 12, VW. As a result, a waveform with a slope in the opposite direction to the originally required slope of the shift signal appears. Therefore, even if you attempt to obtain a logical position detection signal by comparing such waveforms at the center of their amplitudes, the waveform will be distorted as shown in the logical position detection signal vol in FIG. 13, and the normal motor 2
control becomes impossible.

Also, if the shift signal V≦ in FIG. 12 is used, motor 2
In order to drive the shift signal yS, it is necessary to apply a large hysteresis when comparing the shift signal yS.

For example, the hysteresis levels VHH and VH shown in FIG.
By setting L, the hysteresis level Vrm is a logic HIG.
Hysteresis level VFIL to H level is logic LO
It is possible to send out the drive signals Va to Vf for driving the motor 2 in the regular order if each change point level is set to the W level, but the time Δ shown in FIG. 12,
The logic position detection signal V is generated for the first time at a time later than the time when the shift signal ■ passes through the center of its amplitude. 2 occurs, the motor 2 is inevitably driven in a delayed phase with respect to the correct position of the magnet rotor, resulting in poor efficiency of the motor 2.

Moreover, when trying to rotate the motor 2 at high speed, the frequency of the induced voltage signal VW increases and the amplitude period becomes short.

However, the duration of the commutation spike is the same as the induced voltage signal V. As mentioned above, it has nothing to do with the frequency of motor 2, and is almost determined by the load torque of motor 2, so if the load torque is the same, the higher the speed, the narrower the range of induced voltage will be. .

A motor that obtains a foot signal cannot efficiently drive the motor at high load torque and high speed.

[Means to solve the problem]

The present invention has been made in view of the above, and includes a means for detecting a commutation spike and a means for holding an induced voltage signal, and is configured to hold the induced voltage signal in synchronization with the commutation spike signal. This structure prevents commutation spikes from passing through the control circuit.

〔Example〕

Hereinafter, FIG. 1 shows a configuration diagram of the present invention, and FIG. 2 shows waveforms of each part related to FIG. 1 for explanation. 4 is a gate circuit, 5
is commutation spike detection circuit, C1 is capacitor, OF,
is an operational amplifier.

Note that the same reference numerals as in FIG. 11 represent the same or corresponding parts.

Induced voltage signal obtained by dividing the voltage at the neutral point of the armature windings Lu, Lv, and Lw of motor 2 using resistor R1 and the tortoise.
, are input to the gate circuit 4.

The gate circuit 4 is controlled to be non-conductive during the occurrence of a commutation spike by a commutation spike signal v6 sent out by the commutation spike detection circuit 5.

The commutation spike detection circuit 5 detects commutation spikes without phase difference and sends out a commutation spike signal V.

On the other hand, the output of the gate circuit 4 is connected to the capacitor C, and the induced voltage signal vw that has passed through the gate circuit 4 is charged to the capacitor C1. Here, the time constant of charging the capacitor C1 is If it is sufficiently small with respect to the period, it can be charged to the voltage level of the induced voltage signal with almost no phase delay.

The voltage charged in the capacitor C1 is impedance-converted by the operational amplifier OP and sent to the control circuit 3 as a sample-and-hold signal (2).

Now, if the energization pattern of the semiconductor commutator 1 changes and a commutation spike occurs, the commutation spike detection circuit detects this without phase delay and generates a commutation spike signal V,
By sending out the signal, the gate circuit 4 blocks the passage of the signal.

At this time, a closed circuit for discharging the capacitor CI is not formed, and the voltage of the induced voltage signal V, which has been charged in the capacitor C1 until then, is held, and when the gate circuit 4 is next turned on, at that point are charged or discharged to the voltage level of the induced voltage signals v and l, and as a result,
The commutation spike component included in the induced voltage signal is not transmitted from the gate circuit 4 thereafter.

As explained above, a sample-and-hold signal V, which eliminates the influence of commutation spikes, is obtained with a waveform as shown in FIG.

If the waveform of the sample-and-hold signal V is, even if it is integrated by a capacitor, a shift signal (2), which is a nearly triangular-wave amplitude signal with a phase delay of 90 degrees with respect to the sample-and-hold signal (2), can be obtained. Unlike the conventional method, when the induced voltage signal (1) 8 is directly integrated by a capacitor, the waveform is not disturbed near the center of the amplitude.

Therefore, there is no need to provide hysteresis when comparing the sample and hold signal V at the center of the amplitude to obtain the logical position detection signal.

This means that the motor 2 can be operated with high efficiency without being affected by the magnitude of the load torque applied to the motor 2 or the high rotational speed.

In the explanation so far, the neutral point of the armature winding of the motor 2 is used as the source of the induced voltage signal, but the output terminal of the semiconductor commutator 1 is connected in parallel with the motor 2 in the same way as the armature winding of the motor 2. A similar effect can be obtained by using a virtual neutral point of the motor 2 obtained by connecting a resistor array in which resistors of equal value are connected in a star shape as the source of the induced voltage signal.

An example in this case is shown in FIG.

That is, a virtual neutral point equivalent to the neutral point of the motor 2 has a resistor Ru at the output terminal of the semiconductor commutator 1, respectively.

If one end of Rv and Rw are connected and the other ends of each are connected to each other to form a star connection, the common connection point of the resistors becomes the virtual neutral point of the armature winding of the motor 2.

It is well known that the virtual neutral point is exactly equivalent to the neutral point of the actual armature winding of the motor 2.

Now, the above-mentioned method for detecting the rotor position of a motor requires that commutation spikes caused by changes in the energization pattern of the semiconductor commutator can be detected without phase delay.

Therefore, a first embodiment of the commutation spike detection method will be described with reference to FIG. 4F and FIGS. 5 to 7.

In Fig. 4, 6 is a three-phase full-wave rectifier circuit, OP2゜OP
, are comparators, and 53 is an AND circuit.

In addition, the same reference numerals as those in the figures described above represent the same or corresponding parts.

The commutation spike detection method shown in this figure utilizes the fact that the back electromotive force of the commutation spike is generated under conditions that allow the return current to flow through the semiconductor commutator.

In other words, we focused on the fact that when a commutation spike occurs, the level of the electromotive voltage that causes the commutation spike either exceeds the voltage of the power supply voltage supplied to the semiconductor commutator or becomes lower than the ground point of the power supply voltage. It is something to do.

First, this will be explained using FIG.

Figure 5 shows the semiconductor commutator 1 and motor 2 shown in Figure 4.
This figure extracts and illustrates the closed circuit part in a certain energization pattern, where (a) is a power supply closed circuit diagram, and (b) is the return flow when the energization control element on the lower arm side is cut off from (a). This diagram shows a closed circuit diagram, and FIG. 2(e) shows a reflux circuit diagram when the energization control element on the upper arm side shuts off.

Now, as shown in FIG. 2(a), the energization control elements 1a and 1f in the semiconductor commutator 1 are electrically connected, current flows through the armature windings Lu and Lw of the motor 2, and the armature windings Lu
Energy is stored in and Lw.

Next, if the energization pattern changes and the energization control element 1f on the lower arm side is cut off, at least the armature winding Lw
The stored energy passes through the cut-off flywheel diode D on the upper arm side of the same phase, which is most likely to circulate, and passes through the energization control element 1a, which is continuously in a conductive state, to the armature winding Lu. After that, a reflux circuit is formed in which the reflux returns to the armature winding Lw.

In this case, the back electromotive voltage of the commutation spike generated by the armature winding Lw must necessarily exceed the power supply voltage in order to form a freewheeling circuit.

Conversely, if the energization pattern changes and the energization control element 1a on the upper arm side is cut off, at least the armature winding L
The energy stored in u passes through the armature winding Lw, which is the easiest to circulate, and passes through the energization control element 1f.
A reflux circuit is formed in which the current returns to the armature winding Lu via the cut-off flywheel diode D4 on the lower arm side of the same phase.

In this case, in order to form a freewheeling circuit, the voltage at the freewheeling end point of the armature winding Lu needs to be a voltage lower than the grounding point of the power supply voltage, that is, a negative voltage with respect to the power supply voltage.

The above-mentioned freewheeling circuit is formed in the same pattern for each armature winding, so if three-phase full-wave rectification is performed using this as the voltage appearing in each armature winding as shown in Figure 4, three The rectified positive voltage ■J at the positive output terminal of the phase full-wave rectifier circuit 6 and the rectified negative voltage V at the negative output terminal are based on the power supply negative voltage τ, which is the grounding point of the power supply electrode voltage of the semiconductor commutator device 1. , as shown in Figure 7.

Therefore, by comparing the rectified positive voltage vJ and the power supply voltage ■, and also by comparing the rectified negative voltage V; The second commutation spike signal V,2
and the first commutation spike signal v61 when the energization control element of the lower arm is cut off are obtained, respectively.

If both of the first and second commutation spike signals Vol and VS2 are logically synthesized so that the gate circuit is in a cut-off state, a gate signal V shown in FIG. 7 is obtained.

In this example, since the gate signal V is configured to shut off the gate circuit 4 at a logic LOW level, the first commutation spike signal v1! and the second commutation spike signal V02, the gate signal v
I got 6.

Here, it is assumed that the forward voltage drop of the flywheel diode that allows the freewheeling current due to the commutation spike to pass through.

This is because if the forward voltage drop of each flywheel diode used in the semiconductor commutator device is lower than the forward voltage drop of each diode of the three-phase full-wave rectifier circuit 6, commutation spikes cannot be detected. It is for this purpose.

However, the forward voltage drop of the flywheel diode is generally not lower than the forward voltage drop of the diode of the three-phase full-wave rectifier 6.

This is illustrated in Fig. 6(a), which shows the part of the flywheel diode located in the freewheeling circuit of the commutation spike caused by the interruption of the energization control element of the lower arm, and the part involved in the detection of the commutation spike at that time. (b) shows and explains the characteristics of a general diode.

D,I are flywheel diodes, D22 is a diode in the three-phase full-wave rectifying circuit, and OF2 is a comparator.When the freewheeling current IFI flows through the flywheel diode D2I, a forward voltage drop v,1 is generated.

Therefore, the voltage on the cathode side of the flywheel diode is Va, the voltage on the anode side is v:, and the comparator OF.

If the voltage at the non-inverting input terminal of is Vc, then V:=
Va-V,, ・・・・・・・・・・・・・・・・・・
・Formula 2 On the other hand, the diode D, 2 in the three-phase full-wave rectifier circuit
The current passing through is I,2, and the forward voltage drop is v,
If l, then VC= Va −V, 2 ・・・・・・・・・・・・
......Equation 3 Here, the current passing through the diode D22 (A2) is not intended to cause a current to flow through the diode D22, so the anode side of the diode D22 has a high innovidance. It may be possible to do so.

Therefore, I,, > I,, ・・・・・・・・・・・・
......Formula 4 General l and diode are approximately 0.6
It has a forward voltage drop of about V, and its magnitude increases as the current increases.

Therefore, I□)■, 2 ・・・・・・・・・・・・・・・
......For formula 5, as shown in Figure 6(b), -v, I〉■22 ......
・・・・・・・・・Equation 6 should be obtained, so as a result, V−〉 Vc ・・・・・・
. . . Equation 7 is obtained, and the comparator OF sends out a level HIGH.

As described above, commutation spikes can be detected by the freewheeling current passing through the flywheel diode.

Next, another commutation spike detection method will be described with reference to a second embodiment shown in FIG. 8, and a timing chart showing waveforms at various parts in FIG. 8 in FIG. 9.

This commutation spike detection method is a method of detecting the return current flowing due to the commutation spike, and uses a flywheel diode connected in antiparallel to each energization control element of a semiconductor commutator device as explained in FIG. ,
It is separated from the semiconductor commutator device and made independent as a flywheel diode circuit, so that the return current due to the commutation spike flows toward the semiconductor commutator device via the flywheel diode circuit. In FIG. 8, 1' is a semiconductor commutator device consisting only of energization control elements, 7 is a flywheel diode circuit, OF5 and OF are comparators, and RH.

Basically, it is a resistor that converts the circulating current into voltage.

The flywheel diode circuit 7 has a common resistor RH that converts the return current flowing toward the semiconductor commutator device 1' into a voltage, and the semiconductor commutator device 1' flows out between the two power supply lines of the semiconductor commutator device 1'. They are connected through a common resistor RL that converts the freewheeling current into voltage.

The return current flowing through the flywheel diode and the energization control element of the semiconductor commutator device 1' due to the commutation spike takes place in the same manner as described in FIG.

Therefore, in FIG. 8, even if any of the lower arm energization control elements is cut off due to a change in the energization pattern of the semiconductor commutator device 1', the return current flows through the seven life cycle diode circuit 7 to the resistor RH. A first return current IH shown in FIG. 9 flows, causing a voltage drop.

Conversely, as the energization pattern of the semiconductor commutator device 1' changes, even if any of the energization control elements on the upper arm side are cut off, the return current will flow through the semiconductor commutator device 1' via the resistance ratio. The second return current IL shown in FIG. 9 is 7
The current flows across the life cycle diode circuit 71, causing a voltage drop across the resistance.

The direction of the voltage drop across the resistor rib is from the flywheel diode circuit 7 side to the power supply line of the semiconductor commutator device 1', and the direction of the voltage drop across the resistor Rt is toward the ground of the power supply of the semiconductor commutator device. 7 life wheel diode circuit 7 from the line, therefore, if the presence or absence of voltage drop across these resistor ribs and ends is detected, the same process as in the first embodiment shown in FIG. , a gate signal V necessary for sampling and holding an induced voltage signal including commutation spikes can be obtained.

〔Effect of the invention〕

As described above, according to the present invention, in the control method for driving this type of motor using the induced voltage generated at the neutral point of the armature winding of the motor by rotation of the motor, commutation spikes can be eliminated without phase difference. Since the commutation spike is detected and sampled and held to hold the induced voltage signal while the commutation spike occurs, the commutation spike component included in the induced voltage signal is removed.

In the present invention, since the sample-and-hold signal obtained in this way is used as a signal for driving the motor, the capacitor integral (90° Shift when performing delayed phase] - No disturbance of the waveform near the center of the signal amplitude can occur.

Therefore, when obtaining a logical position detection signal by comparing the shift signal at the center of its amplitude, there is no need to provide hysteresis in consideration of the influence of commutation spikes. It is possible to obtain a logical position detection signal with a phase state necessary for driving.

Therefore, the motor can be efficiently operated and driven regardless of whether the motor is operated at high speed, low speed, high torque, low torque, or any combination thereof.

[Brief explanation of drawings]

FIG. 1 is a configuration diagram showing an embodiment of the present invention. FIG. 2 is a timing chart showing signals of various parts in FIG. 1 and waveforms of processed signals according to the invention for explaining FIG. FIG. 3 is a configuration diagram showing another embodiment of the present invention. FIG. 4 is a configuration diagram of a commutation spike detection circuit according to an embodiment of the present invention. 5(a), 5(b), and 5(e) are conduction pattern diagrams for explaining the operation in FIG. 4, and FIG. 7 is a timing chart showing waveforms at various parts in FIG. 4. FIGS. 6(a) and 6(b) are explanatory diagrams illustrating the establishment of the configuration diagram of the commutation spike detection circuit shown in FIG. 4. FIG. 8 is a configuration diagram of another commutation spike detection circuit according to an embodiment of the present invention. FIG. 9 is a timing chart showing the waveforms of signals at each part in FIG. FIG. 10 is a configuration diagram showing a conventional example. 11 and 12 are timing charts showing waveforms of various parts in FIG. 10 and waveforms of processed signals for explaining FIG. 10. 1...7 Semiconductor commutator device with flywheel diode 1'...Semiconductor commutator device without flywheel diode 2...Brushless DC motor 3... ...Control circuit 4...Gate circuit 5...Commutation spike detection circuit 6...
... Three-phase full-wave rectifier circuit 7 ... Flywheel diode circuit OF
+......Operation amplifier circuit OF, ~OP,...
・・・・・・Comparator t, ・・・・・・・Commutation spike duration 53 ・・・・・・・AND circuit 几・・・・・・・・・・Operation amplifier OP, , Op, Bias resistance number 1! ! 1 2m Figure 3 Figure 4 g5 (cL) (b) 115 Figure 7 Flash Figure 6 (, d) (b) Figure 9 Figure 10 Flash 11 I2! Figure 12

Claims (3)

[Claims] (1) A brushless DC motor having a semiconductor commutator device consisting of a magnet rotor, a three-phase armature winding connected in a star shape, and six energization control elements connected to a three-phase bridge, In a control method of driving by an induced voltage signal of a neutral point of a line or a virtual neutral point equivalent to the neutral point,
means for detecting a commutation spike signal generated in the armature winding when the energization control element is cut off, and means for sampling and holding an induced voltage signal, the induced voltage being synchronous with the commutation spike signal. A method for detecting the rotor position of a brushless DC motor, characterized in that it is configured to sample and hold signals. (2) The means for detecting the commutation spike signal includes a rectified positive voltage obtained by three-phase full-wave rectification of the terminal voltage of the armature winding and a positive power supply voltage of the power supply supplied to the semiconductor commutator device. The first commutation spike signal is obtained by comparing the rectified negative voltage obtained by three-phase full-wave rectification of the terminal voltage of the armature winding with the feeding negative voltage of the feeding power source. 2. The rotor position detection method of a brushless DC motor according to claim 1, wherein the rotor position detection method is configured to be a sum of the second commutation spike signal and the second commutation spike signal. (3) The brushless motor according to claim 1, wherein the means for detecting the commutation spike signal is configured to detect a return current using the armature winding as a power source. Rotor position detection method.