The present invention relates to a method of starting an
electric brushless motor.
Reference is made to US 5572097, which discloses a method
of starting three phase dc brushless motors for tuning rotating data media such
as computer hard disc drives, CD ROM drives, floppy disc drives, and the like. During
startup, an align output is communicated to a circuit for detecting the position
of the rotor. Furthermore, an additional step involving an initial line energization
sequence is required before a double commutation is performed.
A brushless motor is provided as an electric rotating machine
where the energization of three-phase stator windings for driving a rotating member
(referred to as a rotor hereinafter) is switched from one to another whenever the
rotor rotates through 120 degrees of the electric angle. Such a conventional brushless
motor has commonly a position detector element such as a Hall device for detecting
the rotating position of the rotor. Recently, another type of brushless motor which
includes no position detector element has been developed in response to the demand
for down-sizing of the brushless motor.
For example, a brushless motor is disclosed in Japanese
Patent Publication (Heisei)5-24760 where, in view of any two different phases of
the three-phase stator windings being energized in a sequence, the voltage induced
at the remaining not-energized phase is measured and used for calculating the rotating
position of the rotor. As the brushless motor produces non of the induced voltage
at the startup stage which is used for calculating the rotating position of the
rotor, its rotor has slightly be driven by forced commutation. The forced commutation
means that any two desired phases of the stator, e.g. U and V, are energized regardless
of the position of the rotor (which is hence referred to as one-phase energization
hereinafter). The position of the rotor is detected from the induced voltage and
then a common procedure of the energization will follow in relation to the detected
The positional relationship between the rotor and the stator
when they stop their movement as the motor has been deenergized is determined by
attracting and repulsing forces of the magnets . For example, when the motor is
an outer rotor type brushless motor having three-phase stator windings, its positional
relationship between the rotor and the stator is expressed by six different pausing
modes, p1 to p6, shown in Fig. 13. Fig. 13 illustrates an arrangement of a primary
part of the brushless motor in addition to the six pausing modes of the position
relationship between the rotor and the stator of which the movement stops as the
motor has been deenergized.
As shown in Fig. 12, the counter clockwise direction is
the forward direction Rs of the rotor while the clockwise direction is the reverse
direction Rr. The stator 100 and the rotor 200 of the brushless motor are disposed
inward and outward respectively. The stator 100 has magnetic poles 300 of U, V,
and W phase. The magnetic poles 300 incorporate windings. The rotor 200 has a row
of permanent magnets m1, m2, m3, ... of which the polarity alternates between N
and S along the circumference.
A movement of the rotor from the initial pausing mode p1
to p6 when is driven by forced commutation between U phase and W phase without initial
magnetization will be explained. When an electric current is supplied through U
phase to W phase, the U phase is magnetized to positive (N) pole and the W phase
is magnetized to negative (S) pole.
At the initial pausingmode p1, the magnet m2 at S is attracted
by the U phase at N but repulsed by the W phase at S. This causes the rotor 200
to rotate at a maximum torque in the forward direction Rs. At the initial pausing
mode p2, the U phase at N attracts the magnet m2 at S but repulses the magnet m3
at N hence allowing the rotor 200 to rotate at the maximum torque in the forward
direction Rs. At the initial pausing mode p3, the attraction between the U phase
at N and the magnet m2 at S is balanced with the attraction between the W phase
at S and the magnet m1 at N. This permits no movement of the rotor 200.
At the initial pausing mode p4, the magnet m2 at N is attracted
by the W phase at S while the magnet m1 at S is repulsed by the same. This causes
the rotor 200 to rotate in the reverse direction Rr. At the initial pausing mode
p5, the U phase at N attracts the magnet m3 at S but repulses the magnet m2 at N
hence allowing the rotor 200 to rotate further in the reverse direction Rr. At the
initial pausing mode p6, the repulsion between the U phase at N and the magnet m2
at N is balanced with the repulsion between the W phase at S and the magnet m1 at
S. This permits no movement of the rotor 200.
As described, the startup torque may be generated non or
too small at the initial pausing modes p3 and p6 thus disallowing the brushless
motor to start up. In particular, when the brushless motor is linked to a heavy
load and thus required to generate a large torque, this disadvantage will be significant.
For example, the motor for starting an internal combustion engine, even if its output
is great, may fail to generate a desired level of the startup torque because the
friction in the engine is too high. At the initial pausing modes p4 and p5, the
rotor rotates in the reverse direction and fails to generate a desiredmagnitude
of the induced voltage needed for detecting the position of the rotor, hence inhibiting
any normal energizing action. More particularly, by force commutation, the motor
when remains free in the movement can be rotated in the forward direction two times
out of six trials or at 1/3 of the probability.
According to the present invention, there is provided a
method of starting an electric brushless motor of an outer rotor type which has
a multi-pole magnetic outer rotor that is joined directly to an engine for acting
as an engine starter and a set of first, second, and third stator windings arranged
at phase intervals of 120 electric degree to be energized in sequence of forced
commutation according to a voltage signal induced on the stator windings when the
rotor is rotated, comprising the steps of: supplying current to the first stator
winding via the second stator winding for a first action of initial alignment of
the rotor at the startup, supplying current to the third stator winding via the
second stator winding for a second action of initial alignment of the rotor; and
supplying current to the third stator winding via the first stator winding for starting
the forced commutation, wherein each of the steps are performed without detecting
a rotation position and/or a rotation speed of the rotor, and in each action of
the initial alignment, each of the windings are excited to displace the rotor to
a position different from the most recent position and stop thereat by the initiating
With at least some embodiments of the present invention,
the problem of providing a method of starting an electric brushless rotating machine
which can generate a great level of the startup torque with no use of rotor position
detecting elements is addressed. Another problem addressed by such embodiments is
to provide a method of starting an electric brushless rotating machine which can
securely rotate in the right direction at high probability when the forced commutation
The method to be described in detail hereinbelow allows
an electric brushless rotating machine to start with a maximum of the startup torque.
With the method to be described the duration of supply
of energy for the initial alignment is predetermined to such a length as to stabilize
the position of the rotor after movement caused by the initial alignment. Furthermore,
the duration of supply of energy for the initial alignment is based on the maximum
of time among the time which are needed to stabilize the positions of the rotor
after movement from the six positions where the rotor may be stopped spontaneously.
The electric brushless rotating machine can be minimized
in the loss during the supply of energy and its startup duration can be shortened.
An advantage that is obtained, having regard to the fact
that the rotor is joined directly to the engine, the machine then acting as an engine
starter motor, is that the startup action can remain tuned with stability while
not affected by a great power of startup friction.
BRIEF DESCRIPTION OF THE DRAWINGS
For a better understanding of the invention and to show
how the same may be carried into effect, reference will now be made, by way of example,
to the accompanying drawings, in which:
- Fig. 1 is a side view of one embodiment of an electric brushless rotating machine
in the form of an engine generator system according to the present invention;
- Fig. 2 is a cross sectional view taken along the line V-V of Fig. 1;
- Fig. 3 is a schematic view of the engine generator system;
- Fig. 4 is a flowchart for controlling the startup of the engine generator system;
- Fig. 5 is a flowchart for controlling the initial magnetization;
- Fig. 6 is an explanatory view showing a fist and a second action of the initial
- Fig. 7 is an explanatory view showing the fist and second actions of the initial
- Fig. 8 is a flowchart for controlling the forced commutation in the first embodiment;
- Fig. 9 is a flowchart for controlling the forced commutation in the second embodiment;
- Fig. 10 is a flowchart for controlling a normal energization;
- Fig. 11 is a diagram showing a stable duration of the rotor single-phase energized
at the initial pausing modes p1 to p6; and
- Fig. 12 illustrates a relationship between the stator and the rotor which are
One embodiment of the present invention will be described
in mode detail referring to the relevant drawings. Fig. 1 is a side view of an electric
brushless rotating machine of the embodiment in the form of an engine generator
system. Fig. 2 is a cross sectional view taken along the line V-V of Fig. 1.
The engine generator system 1 includes an engine 2 and
a generator 3. The generator 3 is a magnet type multi-pole power generator. The
engine 2 has a crank shaft 4 thereof supported by a bearing 6 installed in a side
wall 5a of a crank case 5 to extend at one end outwardly of the engine 2. A star-shaped
annular iron core 7 is fixedly mounted by bolts 8 to a boss region about the crank
shaft 4 of the side wall 5a of the crank case 5. The iron core 7 comprises an annular
center yoke portion 7a with twenty seven projections 7b extending radially from
the center yoke portion.
The projections 7b have three-phase alternate windings
provided thereon thus constituting a stator 8. The iron core 7 is multi-poled for
generating a large output of power and its center yoke portion 7a and projection
7b are decreased in the radial length thus contributing to the lower weight of the
The crank shaft 4 has a hub 9 of a forged member fitted
onto the distal end thereof. The hub 9 is linked to a flywheel 10 which also acts
as a rotor yoke. The flywheel 10 is a pressed member of a cup-like shape comprising
a disk portion 10a and a cylinder portion 10b. The disk portion 10a is fixedly joined
to the hub 9 so that the cylinder portion 10b encloses the outsides of the projections
7b of the iron core 7.
Eighteen neodymium magnets 11 having higher magnetism are
circumferentially mounted on the inner side of the cylinder portion 10b of the flywheel
10 thus constituting a magnetic rotor 12 of an outer rotor type. The rotor 12 has
the magnets 11 aligned tightly on the inner side of the cylinder portion 10b to
have a sufficient mass and can hence function successfully as the flywheel.
A cooling fan 13 is mounted to the disk portion 10a of
the flywheel 10. The cooling fan 13 has a set of blades 13b provided upright and
arranged circumferentially on one side of an annular base 13a thereof. The annular
base 13a is fixedly mounted to the outer side of the disk portion 10a of the flywheel
10. The cooling fan 13 is enclosed in a fan cover 14 which provides a cooling air
passage 14a extending from the outer side of the flywheel 10 to the engine 2.
Fig. 3 is a schematic view of the engine generator system
1. The generator 3 is driven by the (internal combustion) engine 2 to generate a
three-phase alternating current. The alternating current output of the generator
3 is full-wave rectified to a direct current by a converter 15 which comprises a
rectifying circuit having a group of semiconductor rectifier devices connected in
a bridge form. The direct current output of the converter 15 is then smoothed by
a capacitor smoothing circuit 16 and transferred to an inverter 17 where it is converted
into an alternating current at a desired frequency by the FET bridge circuit of
the inverter 17. The alternating current output of the inverter 17 is received by
a demodulation filter 18 where a lower frequency component (e.g. commercial frequencies)
is passed through. The alternating current passed through the demodulation filter
18 is transferred via a relay 19 and a fuse 20 to an output terminal 21. The relay
19 remains open at the startup of the engine 2 and is then closed when the engine
2 runs to a specific level.
The generator 3 in the engine generator system 1 is also
used as a starter for starting the engine 2. For the purpose, the generator 3 includes
a starter driver 22. A rectifying circuit 23 and a smoothing circuit 24 are provided
for supplying the starter driver 22 with a current for starting the engine 2. The
rectifying circuit 23 comprises a harmonic filter 231 and a converter 232. The harmonic
filter 231 has a fuse 20A and is connected by the fuse 20A to the output terminal
21. The output of the generator 3 is connected to, for example, a single-phase power
source 25 at 200 VAC and receives the alternating current from the source 25 for
the startup action. The alternating current is transmitted to the harmonic filter
231 where its harmonic is removed off, converted to a direct current by the converter
232, and received as a power supply via the smoothing circuit 24 by the starter
The starter driver 22 supplies the three-phase windings
of the generator 3 in a predetermined sequence with the current for starting the
engine 2. For sequentially supplying the windings with the current, a switching
device (FET) 221, a CPU 222, and a sensorless driver 223 employing no sensor (magnetic
pole detector) for detecting the location of the rotor 12. As the rotor rotates,
the sensorless driver 223 measures the location of the rotor from voltage signals
induced on the first, second, and third stator windings arranged at equal intervals
of a 120-degree phase difference and determines the energization of the stator windings.
Fig. 4 is a flowchart for controlling the startup of the
engine generator system 1. When the generator 3 starts operating after its free
pausing state, it may fail to have a desired startup torque during the forced commutation
due to the negative positional relationship between the rotor and the stator. Also,
the forward rotation may be interrupted. For compensation, Steps S1 and S2 conduct
the first and the second action of the initial magnetization for shifting the rotor
12 to its desired location relative to the stator so that the desired startup torque
is gained by the forced commutation and the forward rotation is encouraged. The
initial magnetization then allows the rotor 12 to move to the desired location for
gaining its maximum torque. The first and the second action of the initial magnetization
are different in the energizing phase but equal in the procedure (as will be described
later in more detail) . Even when the rotor and the stator remain at their free
pausing state or at any positional relationship (ranging from p1 to p6 in Fig. 13),
the two initial magnetizing actions can shift the rotor 12 to a desired position
for producing the maximum torque. If the duration of the initial magnetization is
too short, the rotor may rotate without steadiness and jog at its stop position.
The energizing period for the initial magnetization will hence extend until the
rotor is located with stability, i.e. substantially one second.
At Step S3, the forced commutation is carried out. The
forced commutation involves single-phase energization from the positional relationship
between the rotor and the stator when the maximum torque is gained after the second
action of the initial magnetization. The induced voltage from the non-energized
phase is measured through the forced commutation and then used for detecting the
position of the rotor 12. As the position of the rotor 12 has been determined from
the induced voltage detected, the procedure goes to Step S4 where a normal procedure
of the energization is carried out.
Fig. 5 is a flowchart for the initial magnetization (of
both the first and the second action). At Step S10, the FET 221 is driven for energization
of a predetermined phase. The first action of the initial magnetization energizes
from the V phase to the U phase while the second action of the initial magnetization
energizes from the V phase to the W phase. Step S11 follows where the initial value
of energization duty is increased by a predetermined rate (for example, 1 %). It
is then examined at Step S12 whether or not the rotor 12 stops at any initial location
(ranging from p1' to p6') in relation to the stator after a counter electromotive
force is produced. When the rotor 12 remains not moved, the counter electromotive
force is zero. Accordingly, the pausing of the rotor 12 at the initial location
can be determined when the counter electromotive force is zero. At this step, once
the counter electromotive force has been released, its value is examined whether
zero or not. If no counter electromotive force has been released, it is judged "no"
at the step. When "yes" at Step S12 is given, it is judged that the initial magnetization
has been completed and the procedure goes to the next step. More specifically, when
the first action of the initial magnetization is completed, the procedure goes to
the second action of the initial magnetization. When the second action of the initial
magnetization is completed, the procedure goes to the force commutation.
When it is judged "no" at Step S12, the procedure advances
to Step S13 where it is examined whether the energization duty of the FET 221 exceeds
an upper limit (e.g. 50 %) or not. If not, the energization is carried out at the
current duty (at Step S14) and the procedure returns to Step S11. When the rotor
12 fails to pause at the initial location with the duty reaching the upper limit
or the counter electromotive force has not yet been released, it is judged "yes"
at Step S13. This indicates a lockup state or an overloaded state and the duty is
turned back to zero at Step S15 before the procedure is terminated with fail (at
The first and the second action of the initial magnetization
will be explained in more detail referring to Figs. 6 and 7. The initial pausing
modes from p1 to p6 illustrated at the left end in Figs. 6 and 7 indicate the initial
location of the rotor 200 relative to the stator 100 when the generator stops spontaneously
as are identical to those p1 to p6 shown in Fig. 13. When the energization from
the V phase to the U phase is carried out for conducting the first action of the
initial magnetization, the polarity of the V phase is turned to N and the polarity
of the U phase is turned to S. This causes the permanent magnets m2 and m3 of the
rotor 200 at the initial relationship p1 to be attracted by the N pole of the V
phase and the S pole of the U phase respectively. As a result, the magnetic interaction
between the stator 100 and the rotor 200 is balanced thus holding the rotor 200
at the location p1'. When the positional relationship between the stator 100 and
the rotor 200 is at any of the locations p2 to p6, the rotor 200 is held by the
same effect at the locations p2' to p6'. As apparent, the location p4' among p1'
to p6' is different from the others p1' to p3', p5', and p6'.
When the energization from the V phase to the W phase is
carried out for conducting the second action of the initial magnetization, the polarity
of the V phase is turned to N and the polarity of the W phase is turned to S. This
allows the S pole and the N pole of the rotor 200 to be repulsed and attracted respectively
by the S pole of the W phase. As a result, the rotor 200 pauses with the permanent
magnet m2 at S and the permanent magnet m1 at N held by the N pole of the V phase
and the S pole of the W phase respectively. It is hence apparent that the pausing
mode p1" is established when the second action of the initial magnetization is carried
out at any pausing mode of p1' top6' determined by the first action of the initialmagnetization.
More particularly, all the different pausing modes p1 to p6 can be converged to
the single pausing mode p1" through the first and the second action of the initial
magnetization. The positional relationship between the stator and the rotor involves
the generation of a maximum of the startup torque in the revolution in the forward
direction when the U and W phases are shifted to the N and S poles respectively
by the succeeding forced commutation from the U phase to the W phase.
Accordingly, when the generator at the initial pausing
mode p1'' is driven by the forced commutation, it starts up with its rotor and stator
generating the maximum torque and can thus rotate without difficulty in the forward
The duration of the first and the second action of the
initial magnetization will now be explained referring to Fig. 11. Fig. 11 illustrates
the duration required before the rotation of the rotor becomes stable when the stator
and the rotor at any of the initial pausing modes p1 to p6 have been single-phase
magnetized. If the duration of the initial magnetization is too short, the rotor
may rotate unstable and create a rocking motion at its pausing location. As apparent
from Fig. 11, the duration from the startup of the initial magnetization and to
the rotor becoming stable is amaximumor substantially 0.7 second at the initial
pausing mode p5. It is hence desired that the duration for the initial magnetization
before the rotation of the rotor becomes stable is substantially one second in consideration
of a generous margin.
Fig. 8 is a flowchart showing a procedure of the forced
commutation. At Step S20, the energization to a predetermined phase, e.g. from the
U phase to the W phase, is conducted. Step S21 follows where the duty of PWM is
gradually increased, for example, at steps of 1 %. It is examined at Step S22 whether
or not the current required for generating a torque of starting the engine or getting
over the upper dead point for the compression exceeds an upper limit determined
from the allowance for the energization (over-current).
When it is judged "yes" at Step S22 or the current exceeds
the upper limit, the procedure jumps to Step S24 where the duty is reduced, for
example, by 1 % to protect a relevant component or switching device in the driver.
At Step S25, the forced commutation is executed/continued at the 1% reduced duty.
When it is judged "no" at Step S22, the procedure moves
to Step S23 for examining again whether or not the internal combustion engine has
completed its full turning motion of predetermined times, for example, 10 times.
When so, it is judged at Step S23 that the number of revolutions by the forced commutation
is turned stable and the procedure for the force commutation is terminated before
returning back to the normal energization procedure shown in Fig. 10. It is also
possible to judge the rotating action from the number of revolutions per unit time
instead of the foregoing predetermined number.
As the duty of PWM is gradually increased to the predetermined
upper limit, the energization to any of the windings can be efficient without a
redundancy of the energizing current. Also, when the duty reaches its upper limit,
it can be decreased to inhibit over-current in the energization to the windings
thus permitting the continuous operation.
Fig. 9 is a flowchart showing a modification of the forced
commutation. It may be necessary in respect of the capability of the switching element
or the driver in the engine generator system 1 to carry out some times the forced
commutation with over-currents exceeding the upper limit and overcome or climb over
the upper dead point for the compression using a climb over torque. This is implemented
by the modification of the forced commutation. Like steps are denoted by like numerals
as those shown in Fig. 8 and will be explained in no more detail.
When it is judged at Step S22 that over-current is drawn,
the procedure goes to Step S31 for increasing the count by one. It is then examined
at Step S32 whether or not the count is greater than e.g. 10. When not, the procedure
moves to Step S25 for carrying out the forced commutation at the duty of over-current.
The forced commutation with over-current is continued until the count reaches 10.
When the over-current remains (as judged "not" at Step S32), the duty is decreased
by 1 % at Step S33 to eliminate the over-current.
Fig. 10 is a flowchart for the normal energization. At
Step S41, the duty is increased by 1 %. It is then examined at Step S42 whether
or not the duty reaches its upper limit or the over-current is drawn. When not,
the procedure goes to Step S43 where it is examined whether or not the number of
revolution is higher than a predetermined level (for example, 800 rpm). When it
is judged "yes" at Step S43 or the engine has started, the action of the stator
is completed. In other words, the procedure goes to Step S44 for turning the duty
to 0 %. When it is judged "yes" at Step S42, the procedure advances to Step S45
for decreasing the duty by 1 % to eliminate the over-current. Step S46 follows where
the action is executed at the decreased duty.
While the forced commutation is carried out but not limited
to after the first and the second action of the initial magnetization in the above
embodiment, it may be conducted after the first action of the initial magnetization.
At the initial pausing mode p4 shown in Fig. 7, the forced commutation after the
first action of the initial magnetization causes a reverse of the rotation. However,
the rotation can be in the forward direction with any of the initial pausing modes
p1 to p3, p5, and p6. More specifically, if the forced commutation is conducted
without the initial magnetization, the rotation in the forward direction can be
implemented two times out of six trials as described. When the forced commutation
follows the first action of the initial magnetization, the forward direction can
be effected five times out of six trials thus exhibiting a higher rate of the probability.
As apparent from the foregoing description, the electric
brushless rotating machine according to claims 1 to 3 can start with a maximum of
the startup torque even if it is equipped with no position detector. Also, the machine
allows the system to be simplified in the arrangement without declining the stability
in the startup action.
According to an embodiment of the present invention, the
startup action can be shortened in the duration while its energizing power is minimized
in the loss.
According to another embodiment of the present invention
the startup action is commenced at a maximum of the startup torque and can thus
remain tuned with stability while not affected by a great power of startup friction.