GB2618358A

GB2618358A - A method of determining a position of a rotor of a brushless permanent magnet motor

Info

Publication number: GB2618358A
Application number: GB2206551.0A
Authority: GB
Inventors: Horvat Mate
Original assignee: Dyson Technology Ltd
Current assignee: Dyson Technology Ltd
Priority date: 2022-05-05
Filing date: 2022-05-05
Publication date: 2023-11-08
Also published as: WO2023214150A1

Abstract

A position of a rotor brushless permanent magnet motor (PMM) is determined by: exciting a phase winding of the motor over an excitation period E1, by applying a voltage to the phase winding using an inverter of the motor; and turning off the inverter for a time period. During the time period, multiple values, indicative of a back EMF induced in the winding, are measured, from which a phase of the back EMF induced in the winding is calculated. A zero-crossing point of back EMF induced in the winding is determined using the calculated phase of back EMF, wherein an aligned position of the rotor is determined when the back EMF is at the zero-crossing point. The phase winding may be freewheeled for first and second freewheel periods FW1, FW2. The first freewheel period starts at an end of the excitation period, while the zero-crossing of back EMF induced in the winding occurs during the second freewheel period. Zero-current clamping may be performed when the inverter is turned off for the time period. The zero-crossing point of the back EMF may be determined based on one or more of a calculated phase, amplitude and frequency of the back EMF. The phase of the back EMF may be calculated by integrating the measured values of back EMF. The phase winding may be commutated at a commutation time measured relative to the zero-crossing of back EMF induced in the winding.

Description

A METHOD OF DETERMINING A POSITION OF A ROTOR OF A BRUSHLESS PERMANENT MAGNET MOTOR

Field of the Invention

The present invention relates to a method of determining a position of a rotor of a brushless permanent magnet motor.

Background of the Invention

Knowledge of rotor position is important in order to commutate phase windings of a brushless permanent magnet motor at the correct times. A brushless permanent-magnet motor will often include a Hall-effect sensor, which outputs a signal indicative of the rotor position. Although the component cost of the sensor is relatively cheap, integrating the sensor within the motor often complicates the design and manufacture of the motor. Additionally, the signal output by the sensor is often susceptible to electromagnetic noise generated within the motor.

Sensorless schemes for determining indirectly a position of a rotor are known.

Summary of the Invention

According to a first aspect of the present invention there is provided a method of determining a position of a rotor of a brushless permanent magnet motor, the method comprising: exciting a phase winding of the motor for an excitation period by applying a voltage to the phase winding using an inverter of the motor; turning off the inverter for a time period; measuring, during the time period, a plurality of values indicative of back EMF induced in the phase winding; calculating a phase of the back EMF induced in the phase winding using the plurality of measured values; determining a zero-crossing point of back EMF induced in the phase winding using the calculated phase of back EMF induced in the phase winding; and determining an aligned position of the rotor of the brushless permanent magnet motor when the back EMF induced in the phase winding is at the zero-crossing point.

It is known that the back EMF of a brushless permanent magnet motor may have a substantially sinusoidal form, and it may be possible to obtain an amplitude and a frequency of back EMF induced in the phase winding from past measurement or simulation, or from real-time calculation. By calculating a phase of back EMF induced in the phase winding using the plurality of values indicative of back EMF induced in the phase winding, this information can then be used in conjunction with the known amplitude and frequency to provide a relatively accurate representation of a waveform of the back EMF induced in the phase winding. The representation of a waveform of the back EMF induced in the phase winding may then be used to determine a zero-crossing point of the back EMF induced in the phase winding, and hence to determine an aligned position of the rotor when the back EMF induced in the phase winding is at the zero-crossing point.

By measuring the plurality of values indicative of back EMF induced in the phase winding during a time period in which the inverter is turned off, increased accuracy may be achieved relative to measuring values of phase voltage and phase current during the excitation period. For example, at relatively low speeds of the motor, an excitation period may be relatively short in duration, and a rise in phase current during such an excitation period may be relatively quick. Such a short excitation period may not enable time for sufficient samples of phase current and phase voltage to be measured to ensure accuracy, whilst such a quick current rise time may result in noise. In contrast, at relatively low speeds of the motor a time period in which the inverter is turned off, for example to provide zero-current clamping, may be relatively long in comparison to the excitation period.

By measuring values indicative of back EMF induced in the phase winding when the inverter is turned off, more samples may be obtainable, which may lead to increased accuracy in determination of an aligned position of the rotor of the motor.

The method according to the first aspect of the present invention may also be advantageous as the method may enable zero-crossing points to be determined where they lie outside of periods of excitation of the phase winding. In particular, by utilising a representation of the back EMF induced in the phase winding in the manner described above, zero-crossing points can be determined irrespective of whether or not they lie within a period of excitation of the phase winding. This may enable efficient operation over a wider range of powers relative to, for example an arrangement where the rotor position is determined by hardware. In particular, a motor in which zero-crossing points can only be determined during periods of excitation may have a lower operating efficiency for a given power relative to a motor which is controlled in accordance with the method according to the first aspect of the present invention, as zero-crossing points which lie outside periods of excitation cannot be determined with any accuracy, leading to inefficient commutation.

Measuring a plurality of values indicative of back EMF induced in the phase winding during the time period, e.g. when the inverter is turned off, may comprise measuring a voltage across the phase winding of the motor. Where the inverter is turned off, there may be no applied voltage to the phase winding, and hence any voltage across the phase winding may be caused directly by the back EMF induced in the phase winding. Measuring a plurality of values indicative of back EMF induced in the phase winding during the time period may comprise directly measuring voltage values of back EMF induced in the phase winding during the time period.

The time period may be between successive excitations of the phase winding. 30 The method may comprise measuring a plurality of values indicative of back EMF induced in the phase winding across a plurality of time periods between a plurality of successive excitations, for example such that the method is utilised to determine a plurality of zero-crossings of back EMF induced in the phase winding.

The method may be carried out using software, for example rather than using hardware. Hence component number and/or overall cost of a control system to implement the method according to the first aspect of the present invention may be reduced, for example relative to a scheme where zero-crossings of back EMF are predicted or calculated using hardware.

By a zero-crossing point of back EMF induced in the phase winding is meant a point at which the value of the back EMF hits zero during a transition between a positive polarity back EMF value and a negative polarity back EMF value, or vice versa.

The method may comprisefreewheeling the phase winding for a first freewheel period and a second freewheel period, wherein the first freewheel period starts at an end of the excitation period, and the second freewheel period is such that the zero-crossing of back EMF induced in the phase winding occurs during the second freewheel period, and wherein the time period occurs between an end of the first freewheel period and a start of the second freewheel period and/or between an end of the second freewheel period and a start of a next subsequent excitation period.

By performing the second freewheel period about the zero-crossing of back EMF induced in the phase winding, defluxing of a stator core of the motor may be achieved. In particular, the back EMF induced in the phase winding is a derivative of the flux linkage in the stator core, and so when the back EMF induced in the phase winding is zero, the flux linkage is at a peak, which means a flux density of the stator core is at a maximum. By freewheeling around the back EMF zero-crossing, magnetic flux can be created in the phase winding that opposes the magnetic flux of the stator core, thereby reducing magnetic flux density in the stator core and reducing iron losses associated with the motor (with iron losses being proportional to the square of the flux density).

Measuring the plurality of values indicative of back EMF induced in the phase winding during the time period, e.g. when the inverter is turned off, may facilitate determination of the zero-crossing of back EMF induced in the phase winding where the second freewheel period is utilised. For example, in a scenario in which the second freewheel period is not utilised, and instead inverter turn-off occurs over a period in which the zero-crossing occurs, it may be possible to directly detect the zero-crossing by measuring the voltage across the phase winding in that period. However, freewheeling for the second freewheel period may inhibit such direct measurement, and so the method described above may find particular utility here.

The time period may comprise a first portion prior to the second freewheel period and a second portion post-the second freewheel period, for example with the first portion occurring between the end of the first freewheel period and the start of the second freewheel period, and the second portion occurring between the end of the second freewheel period and the start of the next subsequent excitation period.

The inverter may comprise a first pair of switches and a second pair of switches, the first and second pairs of switches movable between a plurality of switch configurations, and freewheeling for the second freewheel period may comprise placing the inverter in a first switch configuration where one of the first and second pairs of switches is closed and the other of the second and first pairs of switches is open.

By freewheeling where one of the first and second pairs of switches is closed and the other of the second and first pairs of switches is open, i.e. by freewheel using two closed switches, efficiency may be improved relative to an arrangement where freewheeling takes place by utilising a switch configuration where one of a pair of switches is open and the other of the pair of switches is closed.

For example, where the switches of the first and second pairs of switches comprise transistors, in an arrangement where freewheeling takes place by utilising a switch configuration where one of a pair of transistors is open and the other of the pair of transistors is closed, current may flow through a body diode of the open transistor to provide freewheeling. This may, however, result in greater losses than an arrangement in which current flows through two closed transistors to provide freewheeling, for example due to the voltage drop across the body diode, leading to reduced efficiency.

Furthermore, it has been found that freewheeling in the second freewheel period may lead to symmetry of current induced in the phase winding about the zero-crossing of back EMF. Such symmetry may enable the second freewheel period to be applied for a given time period, and may avoid the need to monitor current flowing through the phase winding to know when to end the second freewheel period. In particular, for a given start point it may be possible to predict where the second freewheel period will end, thereby avoiding the need to monitor current in the phase winding to know when to end the second freewheel period.

The first and second pairs of switches may comprise high-and low-side pairs of switches. Freewheeling may comprise freewheeling around either the high-side pair of switches or freewheeling around the low-side pair of switches.

The method may comprise performing zero-current clamping between an end of the first freewheel period and a start of the second freewheel period, for example performing zero-current clamping during the time period. By performing zero-current clamping when the current in the phase winding decays to zero at the end of the first freewheel period, current may be inhibited from flowing through the phase winding in an opposite direction to the back EMF induced in the phase winding at that time, which would otherwise produce a negative torque, thereby inhibiting acceleration of the motor.

Freewheeling in the first freewheel period may comprise placing the inverter in the first switch configuration, and performing zero-current clamping may comprise placing the inverter in a second switch configuration where the first and second pairs of switches are open, i.e. where the inverter is turned off. Freewheeling in the first freewheel period using the first switch configuration may provide greater efficiency than, for example, an arrangement where freewheeling takes place by utilising a switch configuration where one of a pair of switches is open and the other of the pair of switches is closed.

Freewheeling in the first freewheel period may comprise placing the inverter in the first switch configuration, and placing the inverter in a subsequent switch configuration where one of the first and second pairs of switches is open, and the other of the second and first pairs of switches has one switch open and one switch closed, performing zero-current clamping may comprise placing the inverter in the subsequent switch configuration. Freewheeling in the first freewheel period using the first switch configuration may provide greater efficiency than, for example, an arrangement where freewheeling takes place by utilising a switch configuration where one of a pair of switches is open and the other of the pair of switches is closed. However, transitioning to the subsequent switch configuration during the first freewheeling period may enable zero-current clamping to occur naturally at the end of the first freewheel period. In particular, in the subsequent switch configuration described above, current may be inhibited from flowing through the open switch once the current transitions past its zero-crossing from the previous freewheeling state. Such natural current clamping in the subsequent configuration may avoid the need to time the end of the first freewheel period to manually apply current clamping, eg by switching off both of the first and second pairs of switches at a determined end of the first freewheel period, and may ensure that there is no delay in current clamping that could otherwise lead to negative torque generation as previously described.

Freewheeling for the second freewheel period may comprise only placing the inverter in the first switch configuration. This may result in increased efficiency relative to an arrangement where freewheeling takes place by utilising a switch configuration where one of a pair of switches is open and the other of the pair of switches is closed.

Freewheeling for the second freewheel period may comprise placing the inverter in the first switch configuration, and placing the inverter in a subsequent switch configuration where one of the first and second pairs of switches is open, and the other of the second and first pairs of switches has one switch open and one switch closed. This may provide the increased efficiency of the first switch configuration relative to the subsequent configuration, as previously discussed, whilst allowing for current clamping to happen naturally, given the subsequent switch configuration, at the end of the second freewheel period, thereby avoiding the need to time the end of the second freewheel period, and/or avoiding the need to monitor the phase current during the second freewheel period, to know when to end the second freewheel period.

Freewheeling for the second freewheel period may comprise placing the inverter in a first switch configuration where one of the first and second pairs of switches is open, and the other of the second and first pairs of switches has one switch open and one switch closed. This may allow for current clamping to happen naturally, given the first switch configuration, at the end of the second freewheel period, thereby avoiding the need to time the end of the second freewheel period, and/or avoiding the need to monitor the phase current during the second freewheel period, to know when to end the second freewheel period.

Freewheeling in the first freewheel period may comprise placing the inverter in a second switch configuration where one of the first and second pairs of switches is closed and the other of the second and first pairs of switches is open, and placing the inverter in the first switch configuration, and performing zero-current clamping comprises placing the inverter in the first switch configuration.

Freewheeling in the first freewheel period using the second switch configuration may provide greater efficiency than, for example, an arrangement where freewheeling takes place by utilising the first switch configuration where one of a pair of switches is open and the other of the pair of switches is closed. However, transitioning to the first switch configuration during the first freewheeling period may enable zero-current clamping to occur naturally at the end of the first freewheel period. In particular, in the first switch configuration described above, current may be inhibited from flowing through the open switch once the current transitions past its zero-crossing from the previous freewheeling state. Such natural current clamping in the first switch configuration may avoid the need to time the end of the first freewheel period to manually apply current clamping, eg by switching off both of the first and second pairs of switches at a determined end of the first freewheel period, and may ensure that there is no delay in current clamping that could otherwise lead to negative torque generation as previously described.

The method may comprise commutating the phase winding at a commutation time measured relative to the zero-crossing of back EMF induced in the phase winding. For example, the method may comprise advancing commutation of the phase winding relative to the determined zero-crossing point, and/or the method may comprise commutating the phase winding synchronously with the determined zero-crossing point, and/or the method may comprise retarding commutation of the phase winding relative to the determined zero-crossing point.

Determining an aligned position of the rotor of the brushless permanent magnet motor when the back EMF induced in the phase winding is at the zero-crossing point may comprise determining a future aligned position of the rotor when the back EMF is at a future zero-crossing point. For example, determining a zero-crossing point of the back EMF induced in the phase winding using the calculated phase of back EMF induced in the phase winding may comprise determining a future zero-crossing point of the back EMF induced in the phase winding.

The back EMF induced in the phase winding may comprise a sinusoidal waveform, for example having an amplitude and a frequency. Determining a zero-crossing point of the back EMF induced in the phase winding may comprise utilising any or any combination of a calculated phase of back EMF induced in the phase winding, an amplitude representative of the amplitude of back EMF induced in the phase winding, and a frequency representative of the frequency of back EMF induced in the phase winding.

An amplitude representative of the amplitude of back EMF induced in the phase winding may comprise a pre-determined amplitude, and/or a frequency representative of the frequency of back EMF induced in the phase winding may comprise a pre-determined frequency. For example, a pre-determined amplitude and/or a pre-determined frequency may be obtained by prior measurement or simulation, and may be stored in memory of a controller of the brushless permanent magnet motor. An amplitude representative of the amplitude of back EMF induced in the phase winding may comprise a calculated amplitude, for example calculated in real time, and/or a frequency representative of the frequency of back EMF induced in the phase winding may comprise a calculated frequency, for example calculated in real time. The amplitude representative of the amplitude of back EMF induced in the phase winding, and/or the frequency representative of the frequency of back EMF induced in the phase winding may be speed-dependent. For example, a higher speed of rotation of the rotor of the brushless permanent magnet motor may result in a larger amplitude and/or frequency.

Calculating a phase of back EMF induced in the phase winding may comprise integrating the measured values of back EMF induced in the phase winding to obtain a relationship representative of integrated back EMF.

Integrating the measured values of back EMF induced in the phase winding may comprise integrating the measured values between boundaries set by the phase angle at the beginning and end of the time period.

Calculating a phase of back EMF induced in the phase winding may comprise equating integrated back EMF to an integral of a sinusoidal waveform representative of back EMF induced in the phase winding. Calculating a phase of back EMF induced in the phase winding may comprise equating a relationship representative of integrated back EMF with an integral of a sinusoidal waveform representative of back EMF induced in the phase winding.

According to a second aspect of the present invention there is provided a brushless permanent magnet motor comprising a phase winding, an inverter, and a controller configured to perform a method according to the first aspect of the present invention.

According to a third aspect of the present invention there is provided a data carrier comprising machine-readable instructions for the operation of one or more controllers of a brushless permanent magnet motor to perform the method according to the first aspect of the present invention.

According to a fourth aspect of the present invention there is provided a vacuum cleaner comprising a brushless permanent magnet motor according to the second aspect of the present invention.

According to a fifth aspect of the present invention there is provided a haircare appliance comprising a brushless permanent magnet motor according to the second aspect of the present invention.

Optional features of aspects of the present invention may be equally applied to other aspects of the present invention, where appropriate.

Brief Description of the Drawings

Figure 1 is a first schematic view illustrating a motor system; Figure 2 is a second schematic view illustrating the motor system of Figure 1; Figure 3 is a table indicating switching states of the motor system of Figures 1 and 2; Figure 4 is a flow diagram illustrating a method according to the present invention; Figure 5 is first schematic illustration of a current waveform obtained via use of 20 the method of Figure 4; Figure 6 is a second schematic illustration of a current waveform obtained via use of the method of Figure 4, alongside switch configurations of an inverter of the motor system; Figure 7 is a third schematic illustration of a current waveform obtained via use of the method of Figure 4, alongside switch configurations of an inverter of the motor system; Figure 8 is a fourth schematic illustration of a current waveform obtained via use of the method of Figure 4, alongside switch configurations of an inverter of the motor system; Figure 9 is a schematic illustration of a vacuum cleaner comprising the motor system of Figures 1 and 2; and Figure 10 is a schematic illustration of a haircare appliance comprising the motor system of Figures 1 and 2.

Detailed Description of the Invention

A motor system, generally designated 10, is shown in Figures 1 and 2. The motor system 10 is powered by a DC power supply 12, for example a battery, and comprises a brushless permanent magnet motor 14 and a control circuit 16. It will be recognised by a person skilled in the art that the methods of the present invention may be equally applicable to a motor system powered by an AC power supply, with appropriate modification of the circuitry, for example to include a rectifier.

The motor 14 comprises a four-pole permanent-magnet rotor 18 that rotates relative to a four-pole stator 20. Although shown here as a four-pole permanent magnet rotor, it will be appreciated that the present invention may be applicable to motors haying differing numbers of poles, for example eight poles. Conductive wires wound about the stator 20 are coupled together to form a single-phase winding 22. Whilst described here as a single-phase motor, it will be recognised by a person skilled in the art that the teachings of the present application may also be applicable to multiphase, for example three-phase, motors.

The control circuit 16 comprises a filter 24, an inverter 26, a gate driver module 28, a current sensor 30, a first voltage sensor 32, a second voltage sensor 33, and a controller 34.

The filter 24 comprises a link capacitor Cl that smooths the relatively high-frequency ripple that arises from switching of the inverter 26.

The inverter 26 comprises a full bridge of four power switches 01-04 that couple the phase winding 22 to the voltage rails. Each of the switches 01-04 includes a freewheel diode. As illustrated in Figure 2, the switches 01 and 03 comprise a pair of high-side switches, and the switches 02 and 04 comprise a pair of low-side switches.

The gate driver module 28 drives the opening and closing of the switches 01-04 in response to control signals received from the controller 34.

The current sensor 30 comprises a shunt resistor R1 located between the inverter and the zero-volt rail. The voltage across the current sensor 30 provides a measure of the current in the phase winding 22 when connected to the power supply 12. The voltage across the current sensor 30 is output to the controller 34 as signal, I_SENSE. It will be recognised that in this embodiment it is not possible to measure current in the phase winding 22 during freewheeling, but that alternative embodiments where this is possible, for example via the use of a plurality of shunt resistors, are also envisaged.

The first voltage sensor 32 comprises a voltage divider in the form of resistors R2 and R3, located between the DC voltage rail and the zero-volt rail. The voltage sensor outputs a signal, V_DC, to the controller 34 that represents a scaled-down measure of the supply voltage provided by the power supply 12.

The second voltage sensor 33 comprises a pair of voltage dividers constituted by resistors R4, R5, R6, and R7, that are connected either side of the phase winding 22. The second voltage sensor 33 provides a signal indicative of back EMF induced in the phase winding 22 to the controller, as bEMF.

The controller 34 comprises a microcontroller having a processor, a memory device, and a plurality of peripherals (e.g. ADC, comparators, timers etc.). In an alternative embodiment, the controller 34 may comprise a state machine. The memory device stores instructions for execution by the processor, as well as control parameters that are employed by the processor during operation. The controller 34 is responsible for controlling the operation of the motor 14 and generates four control signals 51-54 for controlling each of the four power switches 01-04. The control signals are output to the gate driver module 28, which in response drives the opening and closing of the switches 01-04.

During normal operation, the controller 34 estimates the position of the rotor 18 using a sensorless control scheme, ie without the use of a Hall sensor or the like. In particular, zero-crossings of back EFM induced in the phase winding 22 can be estimated to estimate aligned positions of the rotor 18. The details of such a control scheme will be discussed in more detail hereafter. With knowledge of the position of the rotor 18 in normal operation, the controller 34 generates the control signals 51-54.

Figure 3 summarises the allowed states of the switches 01-04 in response to the control signals 51-54 output by the controller 33, and such allowed states may be referred to as switch configurations here. Hereafter, the terms 'set and 'clear' will be used to indicate that a signal has been pulled logically high and low respectively. As can be seen from Figure 3, the controller 34 sets Si and S4, and clears S2 and S3 in order to excite the phase winding 22 from left to right.

Conversely, the controller 34 sets S2 and S3, and clears Si and S4 in order to excite the phase winding 22 from right to left. The controller 34 clears Si and S3, and sets S2 and S4 in order to freewheel the phase winding 22. Freewheeling enables current in the phase winding 22 to re-circulate around the low-side loop of the inverter 26. In the present embodiment, the power switches 01-04 are capable of conducting in both directions. Accordingly, the controller 34 closes both low-side switches Q2,04 during freewheeling such that current flows through the switches 02,04 rather than the less efficient diodes.

Conceivably, the inverter 26 may comprise power switches that conduct in a single direction only. In this instance, the controller 34 would clear Si, S2 and 53, and set S4 so as to freewheel the phase winding 22 from left to right. The controller 34 would then clear Si, S3 and S4, and set S2 in order to freewheel the phase winding 22 from right to left. Current in the low-side loop of the inverter 26 then flows down through the closed low-side switch (e.g. Q4) and up through the diode of the open low-side switch (e.g. 02).

Appropriate control of the switches 01-04 can be used to drive the rotor 18 at speeds up to or in excess of 100krpm during normal operation, for example in a steady-state mode. In particular, the phase winding 22 can be excited and freewheeled sequentially, with commutation of the phase winding 22 occurring between successive excitations of the phase winding 22.

When the rotor 18 is driven, the back EMF induced in the phase winding 22 is a derivative of the flux linkage in the stator 20, and so when the back EMF induced in the phase winding 22 is zero, the flux linkage is at a peak, which means a flux density of the stator 20 is at a maximum. This can lead to relatively high iron losses associated with the motor 14, with iron losses being proportional to the square of the flux density.

A method 100 to mitigate for such iron losses is illustrated in the flow diagram of 30 Figure 4. The method 100 comprises exciting 102 the phase winding 22 of the motor 14 for an excitation period El, wherein exciting the phase winding 22 comprises applying a voltage to the phase winding 22. The method 100 comprises freewheeling 104 the phase winding 22 for a first freewheel period FW1 and a second freewheel period FW2, and commutating 106 the phase winding 22 at a commutation time measured relative to a zero-crossing of back EMF induced in the phase winding 22. The first freewheel period FW1 starts at an end of the excitation period El, and the second freewheel period FW2 is such that the zero-crossing of back EMF induced in the phase winding 22 occurs during the second freewheel period FW2.

By performing the second freewheel period FW2 about the zero-crossing of back EMF induced in the phase winding 22, defluxing of the stator 20 of the motor 14 may be achieved. In particular, by freewheeling around the back EMF zero-crossing, magnetic flux can be created in the phase winding 22 that opposes the magnetic flux of the stator 20, thereby reducing magnetic flux density in the stator 20 and reducing iron losses associated with the motor 14.

An exemplary current waveform 200 and back EMF waveform 202 in accordance with the method 100 are illustrated schematically in Figure 5.

Initially, a voltage is applied to the phase winding 22 by closing switches Q1 and 04, i.e. by exciting the phase winding 22 from left-to-right, such that current is driven through the phase winding 22 and increases during the first excitation period El. Here the switches Q1-Q4 are in a switch configuration where one of the pair of high-side switches, i.e. 01, is on and the other of the pair of high-side switches, i.e. Q3, is off, and one of the pair of low-side switches, i.e. 04, is on and the other of the pair of low-side switches, i.e. Q2, is off.

When the excitation period El has expired, or when the current flowing through the phase winding 22 reaches a threshold value, the phase winding 22 is freewheeled for the first freewheel period FW1. Freewheeling in the first freewheel period FW1 can take place in a number of ways, as will be described in more detail hereafter.

The first freewheel period FW1 ends when the current in the phase winding 22 decays to zero, and zero-current clamping is performed for a first clamping period Cl. This may inhibit current of an opposite polarity to the back EMF from flowing in the phase winding 22 at the same time, which would otherwise result in negative torque generation. It will be appreciated that zero-current clamping can be achieved via different switch configurations of the inverter 26, as will be described in more detail hereafter.

At the end of the first clamping period Cl, the phase winding 22 is freewheeled for the second freewheel period FW2. Freewheeling in the second freewheel period FVV2 can take place in a number of ways, as will be described in more detail hereafter. The second freewheel period FW2 is timed such that the zero-crossing of back EMF induced in the phase winding 22 occurs during the second freewheel period FW2, and as illustrated in Figure 5 the second freewheel period FW2 is generally symmetric around the zero-crossing of back EMF induced in the phase winding 22. Sensorless methods of estimating zero-crossings of back EMF induced in the phase winding 22 can estimate future zero-crossings, and hence the end of the first clamping period Cl or the start of the second freewheel period FW2 can be determined relative to an estimated future zero-crossing of back EMF induced in the phase winding 22.

At the end of the second freewheeling period FW2, zero-current clamping is performed for a second clamping period C2. This may inhibit current of an opposite polarity to the back EMF from flowing in the phase winding 22 at the same time, which would otherwise result in negative torque generation. It will again be appreciated that zero-current clamping can be achieved via different switch configurations of the inverter 26, as will be described in more detail hereafter.

At the end of the second clamping period C2, the phase winding 22 is commutated, and a voltage is applied to the phase winding 22 by closing switches Q3 and 02, i.e. by exciting the phase winding 22 from right-to-left, such that current is driven through the phase winding 22 and increases during a second excitation period E2. Here the switches 01-04 are in a switch configuration where one of the pair of high-side switches, i.e. 03, is on and the other of the pair of high-side switches, i.e. Ql, is off, and one of the pair of low-side switches, i.e. 02, is on and the other of the pair of low-side switches, i.e. 04, is off.

The sequence of excitation, freewheeling, and clamping described above can be repeated over a number of excitations of the phase winding 22 as desired to appropriately drive the motor 14.

As noted above, a variety of switch configurations of the inverter 26 can be utilised to achieve freewheeling and zero-current clamping.

A first set of switch configurations for use in the method 100 described above is illustrated schematically in Figure 6, where arrows illustrate flow of current around the inverter 26.

As above, the first excitation period El is achieved by applying a voltage to the phase winding 22 by closing switches 01 and 04, i.e. by exciting the phase winding 22 from left-to-right, such that current is driven through the phase winding 22 and increases during the first excitation period El. Here the switches Q1 -Q4 are in a switch configuration where one of the pair of high-side switches, i.e. Ql, is on and the other of the pair of high-side switches, i.e. 03, is off, and one of the pair of low-side switches, i.e. 04, is on and the other of the pair of low-side switches, i.e. 02, is off. This is indicated as switch configuration 1 in Figure 6.

The first freewheel period FW1 is achieved by opening switches Q1 and Q3, and closing switches Q2 and 04, i.e. by so-called body freewheeling or dual device freewheeling. Here the switches are in a switch configuration where each of the pair of high-side switches, i.e. 01 and 03, are open, and each of the pair of low-side switches, i.e. Q2 and 04, are closed. This is indicated as switch configuration 2 in Figure 6. Dual device freewheeling may provide greater efficiency than, for example, so-called single device freewheeling or diode freewheeling, as a greater voltage drop may be experienced across a diode of a respective power switch in comparison with a voltage drop across a body of a respective power switch.

The first clamping period Cl is achieved by opening all switches, i.e. all of 01Q4, of the inverter 26, such that the inverter 26 is turned off. Here the switches are in a switch configuration where each of the pair of high-side switches, i.e. 01 and 03, are open, and each of the pair of low-side switches, i.e. 02 and 04, are open. This is indicated as switch configuration 3 in Figure 6.

The second freewheel period FW2 is achieved by opening switches 01 and 03, and closing switches 02 and Q4, i.e. by so-called body freewheeling or dual device freewheeling. Here the switches are in a switch configuration where each of the pair of high-side switches, i.e. 01 and 03, are open, and each of the pair of low-side switches, i.e. 02 and 04, are closed. This is indicated as switch configuration 4 in Figure 6. As above, dual device freewheeling may provide greater efficiency than, for example, so-called single device freewheeling or diode freewheeling, as a greater voltage drop may be experienced across a diode of a respective power switch in comparison with a voltage drop across a body of a respective power switch.

As indicated by the arrows in Figure 6, current flows in the opposite direction 30 around the low-side of the inverter 26 in the second freewheel period FVV2 when compared to the first freewheel period FW1.

The second clamping period C2 is achieved by opening all switches, i.e. all of 01-04, of the inverter 26, such that the inverter 26 is turned off. Here the switches are in a switch configuration where each of the pair of high-side switches, i.e. 01 and 03, are open, and each of the pair of low-side switches, i.e. 02 and 04, are open. This is indicated as switch configuration 5 in Figure 6.

A second set of switch configurations for use in the method 100 described above is illustrated schematically in Figure 7, where arrows illustrate flow of current around the inverter 26.

As above, the first excitation period El is achieved by applying a voltage to the phase winding 22 by closing switches Q1 and 04, i.e. by exciting the phase winding 22 from left-to-right, such that current is driven through the phase winding 22 and increases during the first excitation period El. Here the switches 01-04 are in a switch configuration where one of the pair of high-side switches, i.e. Ql, is on and the other of the pair of high-side switches, i.e. 03, is off, and one of the pair of low-side switches, i.e. 04, is on and the other of the pair of low-side switches, i.e. 02, is off. This is indicated as switch configuration 1 in Figure 7.

The first freewheel period FW1 is initially achieved by opening switches 01 and 03, and closing switches 02 and 04, i.e. by so-called body freewheeling or dual device freewheeling. Here the switches are in a switch configuration where each of the pair of high-side switches, i.e. Ql and 03, are open, and each of the pair of low-side switches, i.e. 02 and 04, are closed. This is indicated as switch configuration 2 in Figure 7.

After a pre-determined amount of the first freewheel period FW1 has elapsed, the low-side switch 02 is opened, such that so-called single device freewheeling or diode freewheeling is performed. Here the switches are in a switch configuration where each of the pair of high-side switches, i.e. Q1 and 03, are open, one of the pair of low-side switches, i.e. 04, is closed, and the other of the pair of low side switches, i.e. 02, is open This is indicated as switch configuration 3 in Figure 7.

Whilst dual device freewheeling may be more efficient than single device freewheeling, single device freewheeling can achieve natural zero-current clamping at a transition in polarity of current induced in the phase winding 22, which may avoid the need to turn-off the inverter 26. In particular, the body diode of the low-side switch Q2 may allow current to flow in a first direction to achieve single device freewheeling around the low-side of the inverter 26, but may inhibit current flowing in a second, opposite, direction around the low-side of the inverter 26, thereby achieving zero-current clamping.

The first clamping period Cl in the example of Figure 7 is thereby achieved by maintaining the inverter 26 in a switch configuration where each of the pair of high-side switches, i.e. Q1 and Q3, are open, one of the pair of low-side switches, i.e. 04, is closed, and the other of the pair of low side switches, i.e. 02, is open. This is indicated as switch configuration 4 in Figure 7.

The second freewheel period FW2 is initially achieved by opening switches 01 and 03, and closing switches 02 and 04, i.e. by so-called body freewheeling or dual device freewheeling. Here the switches are in a switch configuration where each of the pair of high-side switches, i.e. 01 and 03, are open, and each of the pair of low-side switches, i.e. 02 and 04, are closed. This is indicated as switch configuration 5 in Figure 7.

After a pre-determined amount of the second freewheel period FW2 has elapsed, the low-side switch 04 is opened, such that so-called single device freewheeling or diode freewheeling is performed. Here the switches are in a switch configuration where each of the pair of high-side switches, i.e. 01 and Q3, are open, one of the pair of low-side switches, i.e. 02, is closed, and the other of the pair of low side switches, i.e. 04, is open. This is indicated as switch configuration 6 in Figure 7. As can be seen from Figure 7, current flows in opposite directions around the low-side of the inverter 26 between switch configurations 3 and 6.

The second clamping period C2 in the example of Figure 7 is then achieved by maintaining the inverter 26 in a switch configuration where each of the pair of high-side switches, i.e. 01 and Q3, are open, one of the pair of low-side switches, i.e. Q2, is closed, and the other of the pair of low side switches, i.e. 04, is open. This is indicated as switch configuration 7 in Figure 7.

A third set of switch configurations for use in the method 100 described above is illustrated schematically in Figure 8, where arrows illustrate flow of current around the inverter 26.

As above, the first excitation period El is achieved by applying a voltage to the phase winding 22 by closing switches 01 and 04, i.e. by exciting the phase winding 22 from left-to-right, such that current is driven through the phase winding 22 and increases during the first excitation period El. Here the switches 01-04 are in a switch configuration where one of the pair of high-side switches, i.e. Ql, is on and the other of the pair of high-side switches, i.e. 03, is off, and one of the pair of low-side switches, i.e. 04, is on and the other of the pair of low-side switches, i.e. 02, is off. This is indicated as switch configuration 1 in Figure 8.

The first freewheel period FW1 is initially achieved by opening switches 01 and 03, and closing switches 02 and 04, i.e. by so-called body freewheeling or dual device freewheeling. Here the switches are in a switch configuration where each of the pair of high-side switches, i.e. 01 and 03, are open, and each of the pair of low-side switches, i.e. 02 and 04, are closed. This is indicated as switch configuration 2 in Figure 8.

After a pre-determined amount of the first freewheel period FW1 has elapsed, the low-side switch 02 is opened, such that so-called single device freewheeling or diode freewheeling is performed. Here the switches are in a switch configuration where each of the pair of high-side switches, i.e. 01 and 03, are open, one of the pair of low-side switches, i.e. 04, is closed, and the other of the pair of low side switches, i.e. 02, is open. This is indicated as switch configuration 3 in Figure 8.

The first clamping period Cl in the example of Figure 8 is then achieved by maintaining the inverter 26 in a switch configuration where each of the pair of high-side switches, i.e. 01 and Q3, are open, one of the pair of low-side switches, i.e. 04, is closed, and the other of the pair of low side switches, i.e. 02, is open. This is indicated as switch configuration 4 in Figure 8.

The second freewheel period FW2 is achieved by closing switch 02 and opening switch 04, i.e. by switching the direction of single device or diode freewheeling. Here the switches are in a switch configuration where each of the pair of high-side switches, i.e. 01 and 03, are open, one of the pair of low-side switches, i.e. Q2, is closed, and the other of the pair of low side switches, i.e. 04, is open. This is indicated as switch configuration 5 in Figure 8.

The second clamping period C2 in the example of Figure 7 is then achieved by maintaining the inverter 26 in a switch configuration where each of the pair of high-side switches, i.e. 01 and Q3, are open, one of the pair of low-side switches, i.e. Q2, is closed, and the other of the pair of low side switches, i.e. 04, is open.

This is indicated as switch configuration 6 in Figure 8.

Whilst described above in relation to low-side freewheeling, it will be appreciated that techniques used herein can also be implemented in conjunction with high-side freewheeling, where appropriate. Furthermore, although described above in relation to a continuous excitation period, it will be appreciated that techniques used herein may be utilised in conjunction with excitation schemes that utilise a split excitation period, for example with first and second excitation pulses separated by an intermediate freewheel period.

In each of the switch configurations of Figures 6, 7 and 8, the second freewheel period is applied such that a zero-crossing of back EMF induced in the phase winding 22 occurs during the second freewheel period FW2.

By performing the second freewheel period FW2 about the zero-crossing of back EMF induced in the phase winding 22, defluxing of the stator 20 of the motor 14 may be achieved. In particular, by freewheeling around the back EMF zero-crossing, magnetic flux can be created in the phase winding 22 that opposes the magnetic flux of the stator 20, thereby reducing magnetic flux density in the stator 20 and reducing iron losses associated with the motor 14. This can result in more efficient operation of the motor 14, and more efficient operation of a product in which the motor 14 is utilised.

However, it has been found that use of the second freewheel period FW2 can have implications for control schemes used to generate the control signals Si-S4. In particular, sensorless schemes have previously been proposed where voltage and current measurements made during excitation of the phase winding 22 are used to estimate zero-crossings of back EMF induced in the phase winding 22, and hence to estimate aligned positions of the rotor 18. Such a sensorless scheme is described in published GB patent application GB2582612. However, at relatively low speeds of the motor, the excitation periods can be short, for example as seen in any of Figures 5 to 8, which can lead to insufficient time to obtain necessary voltage and current samples to enable accurate estimation of back EMF zero-crossings induced in the phase winding 22.

One way to mitigate for this would be to instead utilise the bEMF signal from the second voltage sensor 33 to intermittently directly determine a zero-crossing of the back EMF induced in the phase winding 22, and hence to directly determine an aligned position of the rotor 18, to account for any errors that may have occurred during estimation of the zero-crossings of the back EMF induced in the phase winding 22.

However, such a "resynchronisation" of the position signal may be incompatible with the method 100 discussed above. In particular, to directly measure a zero-crossing of back EMF induced in the phase winding 22, the signal bEMF from the voltage sensor 33 would need to be periodically monitored by turning switches Q1-04 off, i.e., by turning off the inverter 26, and when the voltage measured by the voltage sensor 33 transitions from negative to positive or positive to negative determining a back EMF zero-crossing to occur. In accordance with the method 100, freewheeling of the phase winding 22 occurs such that a zero-crossing of back EMF induced in the phase winding 22 is within the second freewheel period 22. It is not possible to directly measure the back EMF via the second voltage sensor 33 when freewheeling occurs.

Thus a second method 300 in accordance with the present invention is illustrated in the flow diagram of Figure 9.

The method 300 comprises exciting 302 the phase winding 22 of the motor 14 for an excitation period by applying a voltage to the phase winding 22 using the inverter 26 of the motor 14, and turning off 304 the inverter 26 for a time period. The method 300 comprises measuring 306, during the time period, a plurality of values indicative of back EMF induced in the phase winding 22, and calculating 308 a phase of the back EMF induced in the phase winding 22 using the plurality of measured values. The method 300 comprises determining 310 a zero-crossing point of back EMF induced in the phase winding 22 using the calculated phase of back EMF induced in the phase winding 22, and determining 312 an aligned position of the rotor 18 of the brushless permanent magnet motor 14 when the back EMF induced in the phase winding 22 is at the zero-crossing point.

By measuring the plurality of values indicative of back EMF induced in the phase winding 22 during a time period in which the inverter 26 is turned off, increased accuracy may be achieved relative to measuring values of phase voltage and phase current during the excitation period. For example, at relatively low speeds of the motor 14, an excitation period may be relatively short in duration, and a rise in phase current during such an excitation period may be relatively quick. Such a short excitation period may not enable time for sufficient samples of phase current and phase voltage to be measured to ensure accuracy, whilst such a quick current rise time may result in noise. In contrast, at relatively low speeds of the motor 14 a time period in which the inverter 26 is turned off, for example to provide zero-current clamping, may be relatively long in comparison to the excitation period. By measuring values indicative of back EMF induced in the phase winding when the inverter is turned off, more samples may be obtainable, which may lead to increased accuracy in determination of an aligned position of the rotor of the motor 14.

In the context of the method 300, the time period can comprise one or more of the first clamping period Cl and the second clamping period C2, where the inverter 26 is placed in a switch configuration in which all switches Q1-Q4 are open, and hence the inverter 26 is turned off. This corresponds to either of switch configuration 3 or switch configuration 5 in Figure 6.

An illustrative voltage waveform 400 measured by the second voltage sensor 33 in the context of the method 300 is shown schematically in Figure 10. Here the measuring 306 of the plurality of values of back EMF induced in the phase winding 22 takes place prior to an excitation period E, during a clamping time period C in which the inverter 26 is turned off and zero-current clamping is performed. Such a clamping time period C is indicated by the shaded region under the voltage waveform of Figure 10.

Once obtained, the plurality of values of back EMF induced in the phase winding 22 are integrated to obtain an integral of back EMF, bEMF_int, as shown below: Ca bEM Fint = E (t)dt -a where -a and a are boundary values at the beginning and end of the time period in which the inverter 26 is turned off respectively. This equation can be used to obtain an estimated back EMF integral over the measurement interval, but this estimate needs to be normalised.

The back EMF induced in the phase winding 22 can also be fairly accurately approximated by a sinusoidal waveform having the following equation: E(t) = Asin(wt -cp) + noise (t) where E(t) is the back EMF, A is the amplitude of the back EMF, w is the angular frequency of the back EMF in radians per second, and cp is the phase of the back EMF in radians. Noise(t) represents any noise present in the back EMF signal.

The integral of the noise component of the back EMF equation approximates to zero, and hence can effectively be ignored.

If we let Fs be the sampling frequency over a measurement interval from -h to h, we let s be the time in samples, and t be the time in seconds, such that s=Fst, then the value bemf_int, calculated from the measured values of back EMF, can also be written as the estimated integral of the sinusoidal back EMF waveform on the interval [-h, h] in samples: bE114 Fmt = E(s)ds = Asin (co -ye) ds If we substitute s=(Fs/ w)x, then we get cc) coco Fs Fs f hy.% bEM Fifa = Asin(x -(p)-dx = A - sin(x -yo)clx O.) to F Fs It can be seen that a normalisation constant for the integral given above is A.(Fs/ w), where Fs is the sampling frequency. It can also be observed that the integration limit hi) is half of the measurement interval expressed as the angle of Fs back EMF in radians.

The amplitude, A, depends linearly on the motor speed and is commonly expressed via the motor-specific constant M1 00K, which is the amplitude in volts at the speed of 100,000 RPM. This constant depends on the motor construction, varies slightly with temperature, and can be determined by characterisation during a resynchronisation phase of the motor 14. The amplitude is thus given by: iRPm 103. 60.f A = M.10- = M 2.105 where FRPM is the motor speed in RPM.

The normalisation constant for the integral of the back EMF therefore becomes: Fs Fs 103. 60. F, A w = A 2Trf = M 27.2.105 The expression (60Fs)/(2.105) equals the number of samples per electrical period at 100,000 RPM for a four-pole motor, i.e. at the speed for which the M constant is specified. This can be thought of as the frequency normalisation factor, whereas M.103 can be thought of as the amplitude normalisation factor.

Thus it can be seen from that for known values amplitude and frequency of back EMF, we can calculate the phase of the back EMF induced in the phase winding 22 using the following relationship: (0 p_ A f 'co sin(x -(p)dx w r., -h sinço = for unit amplitude, and a period of 2-rr.

From the integral of the back EMF mentioned above, we know that Fs h, fa E (t)dt -a A -fr " sin(x -(p)dx 0-) -hr, 2sin (hp) sirup = . 2sin (h T.) Then by utilising the values for the measured back EMF, and converting the argument of the denominator into radians, we can determine a value for sirkp.

The phase, cp, is then obtained by applying the arcsin function.

Once the phase has been calculated 308, known amplitude and/or frequency values stored in memory for the given rotor speed, or indeed calculated amplitude and/or frequency values for the given rotor speed, can be used in combination with the phase to determine zero-crossing points of back EMF induced in the phase winding 22, for example using a representation of the back EMF waveform.

2sin (h tf=.)) The zero-crossing points of back EMF induced in the phase winding correspond to aligned positions of the rotor 18. Information regarding zero-crossing points of back EMF are then used by the controller 34 to commutate the phase winding 22 of the motor 14 in a desired manner, be that advanced commutation relative to the zero-crossing point, synchronous commutation with the zero-crossing point, or retarded commutation relative to the zero-crossing point.

A vacuum cleaner 500 comprising the brushless permanent magnet motor 14 is illustrated schematically in Figure 11. A haircare appliance 600 comprising the brushless permanent magnet motor 14 is illustrated schematically in Figure 12.

Claims

Claims 1. A method of determining a position of a rotor of a brushless permanent magnet motor, the method comprising: exciting a phase winding of the motor for an excitation period by applying a voltage to the phase winding using an inverter of the motor; turning off the inverter for a time period; measuring, during the time period, a plurality of values indicative of back EMF induced in the phase winding; calculating a phase of the back EMF induced in the phase winding using the plurality of measured values; determining a zero-crossing point of back EMF induced in the phase winding using the calculated phase of back EMF induced in the phase winding; and determining an aligned position of the rotor of the brushless permanent magnet motor when the back EMF induced in the phase winding is at the zero-crossing point.
2. A method as claimed in Claim 1, wherein the method comprises: freewheeling the phase winding for a first freewheel period and a second freewheel period; wherein the first freewheel period starts at an end of the excitation period, and the second freewheel period is such that the zero-crossing of back EMF induced in the phase winding occurs during the second freewheel period; and wherein the time period occurs between an end of the first freewheel period and a start of the second freewheel period and/or between an end of the second freewheel period and a start of a next subsequent excitation period.
3. A method as claimed in Claim 1 or Claim 2, wherein turning off the inverter for the time period comprises performing zero-current clamping for the time period.
4. A method as claimed in any preceding claim, wherein the method comprises commutating the phase winding at a commutation time measured relative to the zero-crossing of back EMF induced in the phase winding.
5. A method as claimed in any preceding claim, wherein determining a zero-crossing point of the back EMF induced in the phase winding comprises utilising any or any combination of a calculated phase of back EMF induced in the phase winding, an amplitude representative of the amplitude of back EMF induced in the phase winding, and a frequency representative of the frequency of back EMF induced in the phase winding.
6. A method as claimed in any preceding claim, wherein calculating a phase of back EMF induced in the phase winding comprises integrating the measured values of back EMF induced in the phase winding to obtain a relationship representative of integrated back EMF
7. A method as claimed in Claim 6, wherein calculating a phase of back EMF induced in the phase winding comprises equating integrated back EMF to an integral of a sinusoidal waveform representative of back EMF induced in the phase winding.
8. A method as claimed in Claim 7, wherein calculating a phase of back EMF induced in the phase winding comprises equating a relationship representative of integrated back EMF with an integral of a sinusoidal waveform representative of back EMF induced in the phase winding.
9. A brushless permanent magnet motor comprising a phase winding, an inverter, and a controller configured to perform a method as claimed in any preceding claim.
10. A data carrier comprising machine-readable instructions for the operation of one or more controllers of a brushless permanent magnet motor to perform the method as claimed in any of Claims 1 to 8.
11. A vacuum cleaner comprising a brushless permanent magnet motor as claimed in Claim 9.
12. A haircare appliance comprising a brushless permanent magnet motor as claimed in Claim 9.