Note: Descriptions are shown in the official language in which they were submitted.
32~5~
-- 1 --
FI~.LD OF TH~: INVENTION
This invention relates to electric motors operating on
the reluctance principle, this term being used in a broad
sense to refer to motors in which a changing electromag-
netic field is generated by a stator, and poles of a nor-
mally unwound ferromagnetic rotor move in that field
towards a minimum reluctance position whose angular loca-
tion is progressively altered by the changing electro-
magnetic field so as to produce continuous rotation of
the rotor. In principle, the functions of the rotor and
and stator can be interchanged, but in practice it is
usually more satisfactory for the electromagnetic field
to be produced by the stator since this eliminates the
necessity for slip rings or commutators, and ~his arrange-
ment will be assumed in the following specification andclaims. The polarization of the rotor may be induced in
soft magnetic matexial by the stator electromagnetic field,
as is usually the case in reluctance motors as commonly so
called, or the rotor poles may be permanently polarized by
permanent magnets comprised by the rotor, as in most stepper
motors and many forms of brushless direct current motor.
BACKGROUND OF THE ~NVENTION
Most electric motors have traditionally been provided with
both stator and rotor windings, even though in many induc-
tion motors the latter may be simplified to a "squirrelcage", and rely upon either conduction through commutators
or slip rings, or upon induction, to energize the rotor.
Induction motors normally require an alternating supply
for their operation, and are not in general well adapted
to variable speed operation since their optimum operating
speed is intimately xelated to the velocity of the rotat-
ing field generated by the alternating supply. Direct
: current motors on the other hand require some form of com-
; mutative switching of the supply to the rotor to pro ~de
~2~ 455
- 2 -
continuous rotation, and such commutators are expensive
to build and maintain, as well as a source of undesirable
broadband electrical interference. Control of such motors
where accuxate speeds or displacement con~rol i5 required
remains complex and difficult.
As a result, attention has been giventfor a wide range of
potential applications ranging from motors for consumer
electronic equipment to large appliance, traction and
industrial motors,to motors of the reluctance type in
which the current through stator windings is switched,
; usually in modern designs by solid state devices, so as
to produce a changing electromagnetic field which will
; result in pr~gressive angular movement of poles of a sta-
tor as it seeks a minimum reluctance position within the
field. This movement may be in the form of discrete steps,
i~dividually controlled, as in a stepper motox, ~r the
movement of the rotor may be sensed by some suitable means
to switch the current through the stator windings so as to
provide a free running mode in which successive steps or
impulses run together to provide continuous ro~ation.
Reyardless of the mode employed, the inductance of the
windings provides difficulties as they are progressively
switched, since it limits the rate of increase of the cur-
rent upon energization and the rate at which magnetic
Pnergy can be dispersed w~en no longer required, particu-
lar~y if excessive potentia~6 are not be induced in the
wlndlng~ .
One widely used approach to the second of the above prob-
lems has been to utilize so-called "free-wheeling" diodes
connected across the various windings. When external cur-
rent to a winding is interrupted, the diode provides an
alternative path for the current induced in the winding by
the collapsing magnetic field, and the current thus recir-
culates until the field is fully collapsed, giving a slow
all in current. The rate of collapse can be increased by
x~
incorporating a resistive element in the circuit, but this
reduces efficiency. Such a resistive element can also be
used to assist rapid build up of the field, by acting as
a current limiting device which permits application of
higher energization potentials than would otherwise be
possible. In many actual or potential applications of
such motors, efficient operation and high torque over a
wide range of speeds is required, and to attain these
objectives it is necessary to achieve rapid current rise
and fall times in the windings without unnecessary dissi-
pation of energy as heat so that the fields of the stator
and rotor can be maintained in optimum relationship. If
rise and fall times are too slow, there will either be
overlap with different windings producing opposing fields
at some stages in the cycle, or the speed and/or torque
obtainable will be limited.
One approach to the prob~em of obtaining rapid fall times
has been to regenerate current from the stator windings
to the supply. Thus in United States Patent No. 4,229,685
issued to Meier, the freewheeling diodes are supplemented
by diodes which divert current through a regulator circuit
and back to the supply thus rQcovering the energy stored
by the field generated by a winding following de-
energization of the latter whi~st assisting in rapid col-
lapse of the field. In order to promote rapid build upof the magnetic field, however, Meier utilizes a ch~pping
current regulator to limit current through the motor
winding, which also serves to select a particular winding,
together with a secondary switch which takes the free-
wheeling diode out of circuit except when that winding isenergized. Such a system requires that the supply poten-
tial to the motor be high enough to provide the desired
rate of current build up in the windings, and also
; requires the use of chopping regulators capable of sus-
taining the supply potential. The Meier patent refers to
a stepping motor which can be operated in free running
~ 2~245S
-- 4 -- -
mode. A somewhat similar arrangement is described in
U. S. Patent No. 4,459,519 i~sued to Erdman. This relates
to a motor with a permanent magnet rotor apparently pri-
marily intended for refrigeration systems, and whilst a
different system is used for regulating the current in
the windings, the rate of current build up is still limited
by the supply potential~ Yet further similar arrangements
as applied to various configurations of motor having mag-
netic rotors of both homopolar and heteropolar construc-
tions are described in U. S. Patent No. 3,826,966 issued
to Nagasaka et al. Yet a further arrangement operating
upon this principle is shown in U. S. Patent No. 3,748,554
issued to McDonald.
A further problem which frequently arises in the design of
brushless DC motors is that of turning off the switchingsemiconductors utilized to provide control of the current
~ supplied to the field windings. The most readily avail-
; able and economical semiconductors for the purpose are thy-
ristors which have a controlled turn on ability but usually
can only be turned off by reducing the current through the
device to near zero. Furthermore, when turn off is
achieved, stored energy in the inductive circuits being
controlled can give rise to high potential spike~ which
can destroy the semiconductors if not properly controlled.
For this reason commutation circuits have been developed
for use in such applications which are essentially of
ring counter configuration in that the turn on of the
device controlling one winding is utilized to discharge
one plate of a capacitor connected to the supply to the
previously turned on device so as momentarily to divert
the current to that device to the other plate of the
capacitor and thus interrupt the current flow through the
device for long enough that it switches off. Once it is
switched off, recharging of the capacitor occurs, thus
taking up some of the energy from the collapsing field of
the associated winding.
1~8Z~5
-- 5 --
Although the capacitors used in ~uch circuits can contri-
bute to the transfer of surplus ener~y from one winding to
the next, this is not their primary purpose, and the
arrangement is only useful in cases where the supply to a
following winding can be turned on before ~hat to a pre-
vious winding is terminated. Examples of such arrangements
may be found in Vnited States Patents Nos. 3,611,081 issued
to Watson, and 4,445,077 issued to Kirschner.
In United States Patent No. 3,444,447 issued to Newell, an
arrangemen~ is described for improving the rise and fall
times of currents in the windings of a step motor. Firstly,
the supply is utilized to charge capacitors associated with
control circuits for each winding, ~he circuit beinq
arranged and the capacitor being switched so that its
charge potential is added to the supply potential when the
associated winding is energized, thus initially boosting
the supply potential and improving the current rise time.
Additionally, as described with reference ~o Figures 7 to
9, an arrangement using diodes and/or autotransformers is
utilized ~o transfer energy from the collapsing field of
a winding which has just been turned off to boost the po-
tential applied to a winding that has just been turned on,
thus improving both rise and fall times and improving
efficiency. The first of the techni~ues disclosed by
Newell provides a degree of boost which is substantially
constant regardless of operating conditions, whilst the
second technique is applicable only where the turning on
of one winding is simultaneous with the turning off of
another.
30 In United Sta~es Patent No. 3,486,096, issued to Van
; Cleave, windings of a stepper motor are transformer coupled
in pairs, and the switching means for each winding is
operative to block current flow in a forward direction
only. One or more diodes are placed in series with a D.C.
supply so that current can flow from the supply in a forward
1~3X4~5
-- 6 --
direction only, and unswitched ends of the windings, or
pairs of them, are connected to a capacitor or capaci~ors
whose other plates are grounded. When forward current
through a winding i8 interrupted, a current in the reverse
dixection is induced in the winding coupled thereto, and
charges the associated capacitors to a high potential,
whilst the field produced by the original winding rapidly
collapses. When a switching device again permits forward
current through a winding connected to the capacitor, the
high potential charge on the capacitors assists rapid cur-
rent build up in that winding. The primary purpose o the
arrangement is to speed up operation and protect the
switching device; efficiency is evidently not a concern
since resistors are placed in series with the supply to
limit current. Moreover, the de~ice is applicable only
to motors having a suitable winding arrangement so that
transformer action may be utilized to reverse the direc-
tion of curren~ flow in the windings during energy recovery.
A group of related United States Patents, ~os. 3,560,817
20 and 3,560,818 issued to Amato, 3,560,820, 3,697,839 and
3,714,533 issued to Unnewehr, and 3,697,840 issued to
Koch, and all assigned to Ford Motor Company, relate to
various configurations of control circuits for reluctance
type motors, in each of which a tuned circuit comprising
capacitors and induc~ors (which may be or comprise the
motor winding) are used in conjunction with solid state
switching elements, utilizing resonance effects to increase
the effective potentials available to provi~e fast rise
and fall times, and to reverse the polarity of charge
; 30 received from the circuit when a primary supply is cut
off. Although there are differences between the arrange-
ments described in these various Ford patents their general
principle of operation relies on drawing current from the
primary supply in pulses of approximately half-sine wave
form. Since the period of the pulses is set at a
~8Z~
- 7 -
substantially constant magnitude by the reactive components
in the circuit, provision for differen* motor speeds is
provided by varying the numher of pulses delivered during
each energization phase of a winding, substantial continu-
ity of current flow in the winding between pulses beingobtained both by freewheeling effects and by charge rever-
sal and re-applica~ion of energy recovered during field
collapse. In some of the arrangements, the circuit is
operated so as to build up potential on a capacitor to a
level much greater than the supply, which potential is
applied so as to augment the magnitude of the current
pulses from the supply. In the Unnewehr Patent No.
3,714,533, it is disclosed that surplus energy from this
capacitor may be tapped off by suitably timed firing of an
SCR and returned to the supply if not required to drive
; the motor. ~arious methods for controlling the various
motors disclosed are discussed, in general invol~ing air-
ly complex control of the firing sequence of the several
controlled rectifiers associated wi~h each winding. In
each case, it appears that operation requires an inductor
in series with the supply additional to the motor winding,
and that the operating parameters of the circuit are
critically dependen~ upon the value of this inductor and
- also thoseofan energy storage capacitor. These same ele-
ments also limit the rate at which energy can be drawn
from the supply, since the resonant characteristics of the
load limit both the periods over which current can be drawn
from the supply and ~he rate of supply current rise and
fall.
30 United States Patent No. 4,025,831 disc~oses a motor having
in one embodiment plural stator windings and a permanent
magnet homopolar rotor in a physical arrangement somewhat
resembling the physical arrangement of the preferred em-
bodiment of the motor described hereinbelow. The control
system of the motor is however quite different, as is the
mode of operation, no special provision being made for
-
- 8 - ~ Z ~2 45~
improving current rise and ~all times in the stator
windings, or for recovering energy from collapsing stator
fields.
An object of the present invention is to provide a motor
of the general claæs discussed in which rapid rise and
fall of winding current can be obtained at timings approp-
riate to ensure effective developmen~ of motor torque
over a ~ide range of motor speeds, without the necessity
for expedients which are wasteful of energy (such as added
series resistance), without unduly restricting the rate
at which energy can be drawn from the supply to meet tor-
que demands, and without the necessity for highly sophis-
ticated control means for matching motor characteri.stics
to load requirements.
According to the invention an electric motor, of the type
having a stator with multiple sequentially energizable
phase windings and a rotor magnetized to seek a minimum
reluctance position within a progressively moving electro-
:~ magnetic field produced by said phase windings, first con-
trolled switching means in sexies relative to a D.C. power
,~
~ supply with at least a ~ of each phase winding~, and
1A~ means to control said first switching means to produce
said progressively moving electromagnetic field, the im-
provement wherein ~a) a charge storage capacitor is pro-
~ 25 vided or each such phase winding, with one terminal of
i said capacitor connected by a 1QW impedance path to said
supply, and the other terminal having ~irst and second
connections establishing al~ernative low impedance paths
to the w.inding, the first such connection being estab-
lished through first diode means oriented to permit low
impedance passage to said capacitor from the winding of
forward current generated by collapse of ~he field pro-
: duced by said winding after turn-off of the switching
means, and the second such connection being established
through second controlled switching means, (b) means are
provided to turn on said second switching means
9 ~3Z455
substantially simultaneously with said first switching
means to provide low impedance passage of curren~ from
said capacitor in the forward direction through said
winding to said fir~t controlled switching means, and
(c) second diode means are provided between the supply
and the çonnection to the winding from the second con-
trolled switching means, such as to presen a low imped-
ance path for fo~ward current from tne supply but a high
impedance to reverse current.
lo As compared to the Van Cleave arrangement the present in-
vention does not require any special arrangement or oper-
ating sequence o~ the motor windings to utilize the energy
recovered bv the capacitor, nor does the primary switclling
device need bidirectional current carrying capabilities,
and the ability to withstand, in blocking condition, the
additional potential applied by transformer action in the
winding~.
As compared to the Ford patents discussed above, the
values of reactive ~omponents in the present invention
; 20 limit neither the maximum current which can bo~ drawn fr~m
- the supply, nor the proportion of th~ active period of a
phase winding during which current may be drawn if neces-
sary. Essentially, the value of the capacitor in relation
to the inductance ~f the associated winding determines
the rates of current rise and ~all which can be achieved
in the ~inding, and the proportion of the active period
of a phase winding during which it is necessary for cur-
rent to be drawn from the supply. Under normal operation,
the current to energize a winding is supplied from the
capacitor, and current is only drawn from the supply in
the latter part of the period during which the primary
switching means is switched on, this current draw provid-
ing make-up for energy output to a load or dissipated by
motor losses. The charge on the capacitor can be tapped
by a suitable circuit so that the motor can operate as a
D.C. to D.C. up-converter, or energy recovery or
~ 10 - ~2~2d~55
regeneration braking under overrun conditions. Under
such overrun or braking conditions the energy stored in the
capacitor will be i~excess of that required to maintain
rotation of the motor and the excess may be recovered by
drawing current from the capacitor when its potential
exceeds a certain level.
Further features of ~he invention, and further explana-
tion of its construction will become apparent from the
following description of an exemplary embodiment with
reference to the accompanying drawings.
Figure 1 is a front elevation of the stator and rotor of
apparatus embodying the invention;
Figure 2 is a section taken along line 2-2 in Figure l;
Figures 3a and 3b ar~ views similar to Figure 2 showing
various positions of the rotor and stator and useful in
explai.ning the operation o the apparatus;
Figure 4 is a circuit diagram useful in explaining the
switching operation that takes place in the practice of
the present invention;
Figure 5 is a circuit diagram of trigger pulse generating
and switching circuitry that may be used in practicing
the invention; and
i
Figure 6 is a circuit diagram showing an arrangement
; alternative to ~hat shown in Figure 4.
Referring to Figure 1, and sometimes to Figure 2 apparatus
embodying the present invention includes a rotor 10 which,
in the embodiment shown, includes a rotatable shaft 11 on
which a permanent magnet 12 is mounted and fixed, e.g. by
keying or by any other suitable means, and on which two,
spaced apart toothed wheels 13 and 14 are mounted and
fixed, again by keying or by any other suitable means.
The ~oothed wheels are identical to each other and, in
the embodiment shown, each have six teeth spaced 60 apart.
This is not critical, however, and th~ number of teeth may
vary widely. The toothed wheels are mounted with respect
to each other on shaft 11 50 that the teeth of the toothed
wheels align with each other.
Each toothed wheel is made of a magnetizable material,
i.e., a ferromagnetic material, e g., steel, and thus each
tooth of each toothed wheel constitutes either a north ox
a south pole, all of the teeth on one toothed wheel being
of the same polarity and all of the te~th on the other
toothed wheel also being of the same polarity but of oppo-
site polarity to that of the teeth of the first-mentioned
toothed wheel. In Figure 2 the six teeth of toothed wheel
14 are shown as north poles Nl ~o N6 inclusive.
Many varia~ions in the rotor are possible. For example,
individual permanent magnets may be employed in place of
one permanent magnet and ferromagnetic toothed wheels, or
the rotor may be magnetized by fields produced by windings
on the stator. Moreover, heteropolar a~ well as homopolar
rotor pole configurations may be ~tilized with suitable
stator winding configurations. In the example described,
however, the rotor will have permanent magnet means mounted
on and fixea to a rotatable shaft, the permanent magnet
means having a plurality of spaced apart north poles and a
plurality of spaced apart south poles, and both homopolar
sets of poles will be movable in two circular paths, one
of which is shown at 15 in Figure 2.
In the embodiment of the invention shown, the stator 16
consists of two spaced apart stationary plates 17 and 18
that happen to be of square configuration and that are held
in fixed, parallel relationship with respect to each other
by suitable spacers ox fastening devices 19; four electro-
magnets 20a, 20b, 20c and 20d; and holders 24 for the
; electromagnets.
1~, i7d ~3;2 4L 5 ~;
Plates 17 and 18 may be made of aluminum, for example, as
may holders 24 and spacers or fastening devices 19. Other
suitable materials that are ~on-ferromagnetic also may be
employed.
Each electromagne~ 20a, 20b, 20c and 20d consists of a
ferromagnetic core 21a, 21b, 21c and 21d respectively and
a coil 22a, 22b, 22c and 22d respectively. Holders 24
are secured to plates 17 and 18 and cores 21a-21d fit into
openings provided in holders 24.
The number of electromagnets may be varied without depart-
ing from thls invention, but sufficient electromagnets
must be employed to make it possible to creast a progres-
sively charging magnetic field which, through interaction
with the permanent magnets of rotor 10, causes rotation
of rotor 10.
Electromagnets 20a - 20d are evenly spaced apart from each
other (at 90 in the illustrated embodiment) and, as best
shown in Figure 2, are mounted sufficiently close to the
circular paths travelled by the north and south poles of
rotor 10 that the pole~ are capable of inducing voltages
in the coils of the electromagnets and the electromagnets
are capable of magnetically attracting and/or repelling
the poles as the latter rotate past the electromagnets.
It will be understood, of course, that plates 17 and 18
carry bearings for shaft 11, and shaft 11 may be coupled
to any rotary equipment that is to be dri~en thereby.
Mounted on shaft 11 is an apertured timing wheel 25 on
opposite sides of which are light sources 26 and photo-
detectors 27. These compon~nts constitute a source of
trigger pulses or timing pulses. Light sources 26 and
photodetectors 27 are mounted on a holder which can be
: rotated relative to the apertures in timing wheel 25 to
~282~
- 13 ~
vaxy the phasing of the trigger pulses.
It will be appreciated, of course, that many other types
of devices may be used for generating trigger pulses with-
out departing from the present invention. For example, a
S microswitch contacted by a projection on shaft 11 could
be used.
Referring now to Figure 5, the trigger pulse generating
and switching circuitry for coils 22a and 22c and for
coils 22b and 22d is shown along with electrical energy
utilization circuitry. Only the switching circuitry and
electrical energy utilization circuitry for coils 22a and
22c is shown in Figure 4. The switching circuitry and
electrical energy utilization circuitry for coils 22b and
22d is the same as that shown in Figure 4, as will Se
evident from Figure 5.
Since the trigger pulse generating circuitry, switching
circuitry and electrical utilization circuitry~ is the
same for the two sets of coils, it will be described in
detail only for coils 22a and 22c.
Shown within line 28 is a standard trigger pulse generating
circuit that provides trigger pulses on conductors 29 and
30, the former being connected to the gate electrode of a
gate turn off device GTOl and the latter being con~ected
via a transformer to the gate electrode of a silicon con-
trolled rectifier SCRl.
A D.C. power supply (represented by B and ground) isprovided with B+ being connected via a diode Dl, coils
22a and 22c and a protection diode D2 to the anode of gate
turn off device GT~l, the cathode thereof being grounded.
Provided that the pulse generating circuit is suitably
modified to provide appropriate switching waveforms and
potentials, the gate turn off devices may be replaced by
82a~5
bipolar or field efect transist~xs.
Also associated with coils 22a and 22c is an electrical
energy storage device which, in the embodiment shown in
Figures 4 and S, is simply a capacitor Cl.
One plate of the capacitor is connectea to a terminal of
the supply; in the example shown this is the B~ ~erminal.
Any connection is acceptable that will provide a low im-
pedance source or sink for capacitor charging and dis-
charging currents required to accommodate changes in
potential of the other plates. This other plate is pro-
vided with two alternative connections to opposite ends
of the phase winding comprising the coils 22a and 22c (in
the example shown in Figure 4). The first connection is
to that end of the winding electrically adjacent the
switching means provided by the gate turn off device GT01
and i~s associated pro~ection diode D2, and incorporates
the diode D3so that this connection can only accommodate
charging currents tending to increase the potential on the
associated plate of capacitor Cl. The second connection
is made to the other end of the winding via a thyristor
SCRl, which when triggered on will pass discharging cur-
rents from the associated plate of capacitor Cl. The
diode Dl prevents the thyristor from appearing when turned
on as a short circuit across capacitor Cl, and permits the
other end of the winding to rise to a potential above B .
Referring to Figure 5, the left and right halves of the
circuit shown are identical and essentially independent
except for sharing one bias circuit as a matter of con-
venience, and except that in the example shown they receive
input from separate sets of light sources 26 and photo-
detectors 27, spaced 90 apart in relation to the timing
wheel 25 so that they operate 90 out of phase with one
another, and the left half of the circuit incorporates
the coils 22a, 22c whereas the right half incorporates
~ ~Z~5
- 15 ~
the coils 22b, 22d.
Each triyger circuit 28 has an input amplifier Al connected
to a bridge formed by resistors R2, R3 and R4 (R3 and R4
being common to both amplifiers)and a phbtotransistvr
forming the photodetector 27. When apertures in the wheel
25 permit light from a light emitting diode fed through
resistor Rl and forming the light 26 to fall off the
photodetector, the collector to emitter resistance of the
latter falls and reverses the direction of imbalanc~ of
the bridge, thus causing the comparator Al to apply a
switching transition to the conventional push-pull output
circuit fo.rmed by complementary transistors TRl and TR2
and associated bias componentsR5, R6, R7, ~8, R9 and zener
diodes 21 and Z2. The output of this circuit is appli~d
via a current limiting protection resistor R10 to the
device GTOl via line 29, and via a differentiating capaci-
tor C2, a pulse transformer Tl, and a current limiting
protection resistor Rll to the gate of thyristor ~1. A
small inductance ~1 is located in series with the cathode
of the thyristor to limit. the rate of current increase
through the thyristor to within its specifications. This
inductance, and other components alrea~y mentioned, and
~k~ r~sistor R12, capacitor C3 and diode D4 associated
with the gate turn off device GTOl,whose purpose is solely
~e protection of associated components, do not signifi-
cantly alter the operating mode of the circuit, and they
will not be further discussed. The actual values of the
components utilized, and the se~ection of the semiconduc-
tors to be utilized, is heavily dependent upon the size
of the motor and the supply potential utilized, and the
necessity to operate within the specifications of the
available semiconductors under all anticipated operating
conditions.
The operation of the apparatus can best be understood by
referring to Figures 3a and 3b.
1~824L55
-- 16 --
For purposes of explanation it will be assumed that the
poles of toothed wheel 14 are north poles and thus that
the poles of toothed wheel 13 are south poles. It will
also be assumed that rotation of rotor 10 in a counter-
clockwise direction, as shown by arrow 33, has alreadystarted.
When taothed wheel 14 is in the position shown in Figure
3a, silicon controlled rectifier SCRl and gate turn off
switch device GTOl are of and no current from the D.C.
power supply flows through coils 22a and 22c. As a con-
sequence, electromagnets 20a and 20c are de-energized.
Nevertheless, cores 21a and 21c attract poles Nl and N4
respectively of toothed wheel 14 as these seek a minimum
reluctance position. This attraction coupled with the in-
ertia of rotor 10 carries rotor 10 to the posi~ion shownin Figure 3bo As poles Nl and N4 mo~e past de-energized
coils 22a and 22c respectively from the position shown in
Figure 3a to the position shown in Figure 3b, currents
are induced in the coils as a result of movement of poles
Nl and N4 past them. The induced current is in a sense
such as to charge capacitor Cl through its first connec-
tion via diode D3.
Once the position of Figure 3b has been assumed, an aper-
ture in wheel 25 passes diode 26 resulting in a trigger
pulse on conductor 29 which triggers gate turn off device
GTOl into conduction. At the same time a trigger pulse
on conductor 30 triggers silicon controlled rectifier SCR1
into conduction. Initially the charge on capacitor Cl
will be sufficiently large that diode Dl will be reverse
biased and capacitor C1 discharges via thyristor SCRl
through coils 22a and 22c and gate turn off device GTOl.
Eventually the charge on capacitor Cl may decrease suffi-
ciently that diodé Dl will no longer be reverse biased
and current from the D.C. power supply flows in coils 22a
and 22c. The current, first from capacitor Cl and then
- 17 -
from the supply causes ~1e poles of electromagnets 20a and
20c that are adjacent poles Nl and N4 to become north
poles. The resulting magnetic repulsion between electro-
magnet 20a and pole Nl and between electromagnet ~ 20c
and pole N4 causes continued counterclockwise rotation of
rotor lO. Silicon controlled rectifier SCRl turns off in
this period as the discharge current from the capacitor
falls to zero. After poles Nl and N4 have rotated a pre-
determined amount beyond electromagnets 20a and 20c res-
pectively, the wheel 25 shuts off light ~alling on photo-
transistor 27 from diode 26, applying a negative potential
to line 29 and turning off the device GTOl. This results
in the D.C. power supply being disconnected from coils 20a
and 20c, whereupon a similar cycle is repeated but this
time with poles N6 and N3 interacting with electromagnets
20a and 20c respectively. Intermediately, a similar cycle
occurs, involving the lower half of the circuit of Figure
5, the poles N2 and N5 and the electromagnets 2~b and 20d
respectively. In addition, the same cycle will have ~een
repeated previously but with poles N3 and N6 and electro-
magnets 20~ and 20d respectively.
Of course the same sequence of events is occurring with
respect to toothed wheel 13 and elec~romagnets 20a - 20d
except that the opposite poles of electromagnets 20a - 20d
adjacent toothed wheel 13 are alternately energized to
form south poles rather than north poles.
In general, the poles on the rotor wil~ always tend to seek
positions such as to minimize the reluctance of the may-
netic circuits set up between the rotor and the stator.
3~ This reluctance will be a minimum in the case of poles
adjacent de-energized electromagnets when the poles and
magnets are angularly aligned. However, when all of the
electromagnets are de-energized, the seeking forces will
tend to cancel each other out because of the relative
configurations of the rotor and stator. In the case
~8~5S
- 18 -
already des~ribed of electromagnets energized so that
their poles have the same polarity as adjacent poles of
the rotor and therefore repel one another, the reluctance
will be a minimum when the electromagnet pole is midway
between two rotor poles. In the case o~ electromagnets
energized so that their poles have the opposite polarity
to adjacent poles of the rotor and therefore attract one
another, the reluctance will again be a minimum when the
electromagnet pole is aligned with a rotor pole. The
lC connection of the coils 22a, 22b, 22c, 22d may be oriented
for operation in either of these two modes, i.e. ei~her
repulsion operation or attraction operation. By providing
duplicate sets o~ coils on each electromagnet, with a
first winding on a particular magnet in series or parallel
with a second winding on an adjacent magnet, both modes
may ~e used simultaneously to obtain greater torque from
a given motor configuration. Whilst the sense of the con-
nections to the coils and the orientation of the disc 25
may be altered to accommodate th~se various modes of oper-
ation, the mode of operation of ~he circuit of Figures 4and 5 remains substantially the same.
In order to maximize the mean torque available from thQ
motor and minimize losses, it is desirable that the
electromagnets be energized only when the rotor poles
adjacent the electromagnet poles are moving towards a
minimum reluctance position. If thP magnets are energized
during a period when the adjacent rotor poles are moving
away from a minimum reluctance position, a countertorque
will be produced during this period. This condition is
typical for example of a stepper motor operated in a dis-
crete stepping mode, in which the interaction of the
rotor and stator produces in each step first an accelerat-
ing torque as the rotor moves towards a new minimum reluc-
tance position, and then a holding torque as it reaches
and moves through this position. It is however undesir-
able in a motor intended for continuous running.
~282455
-- 19 --
Correct energi~ation of the electromagnets is not merely
a matter of correctly timing the switching of current to
the electromagnet windings since the latter posses sub-
stantial inductance, which moreover varies with the
reluctance o~ the magnetic circuit with which they are
associated. As a result, the rate of build up of current
in the winding, and thus the rate of energization, is
determined by the value of this inductance and the poten-
tial applied to the winding. As the current builds up,
so does the energy sto~ed in the magnetic circuit. In
order to de-energize the magnet, it is not sufficient
merely to interrupt the current through the winding,
since the stored energy must also be removed in some man-
ner. For efficient operation this energy should be
recovered and used productively. Furthermore, to obtain
a high specific power output from the mo~or, in the form
of good torque at high speeds of rotation, it is necessary
that both energization and de-energization of the magnets
be as rapid as possible.
In the arrangement described, and referring to Figure 4,
let it be assumed as a convenient starting point that the
switching device GTOl is switched on and current is pass-
ing (using the positive to negative current flow conven-
tion) from the supply B+ through diode Dl, coils 22a and
22c, diode D2 and device GTOl to the supply ground. The
current in the coils 22a and 22c results in a correspond-
ing magnetic flux in the maynetic circuits associated with
the coils . Device GTOl is now turned of f, interrupting
the cir~uit through the supply. The tendency of the flux
in the magnetic circuits to collapse induces potentials
in the coils such as to tend to oppose this collapse, and
these potentials result in the forward current in the
coils continuing through the alternative path provided by
the diode D3, the capacitor Cl, and the diode Dl. This
35 forward current continues until the potential across the
capacitor Cl equals that induced across the coils, by which
2~5
_ 20 --
time most of the available energy from the collapsing
magnetic field has been transferred to the capacitor Cl.
Assuming that thyristor SCRl remains in a blocking condi-
tion, the capacitor will then remain charged since it can-
not discharge thxough the diodes Dl and 33. The timerequired for this transfer of energy to the capacitor is
determined by the resonant frequency of the tuned circuit
formed by the windings 22a, 22c and the capacitor Cl,
being rather less than the period o~ one half cycle.
Oscillation of the circuit is suppressed by the diodes Dl
and D3, and therefore by suitable choice of capacitance
and inductance value, very rapid de-energization of the
magnets can be obtained, whilst the capacitor Cl can be
charged to a potential much greater than the supply
potential.
When energization of the coils 22a and 22c is again re-
quired, device GTOl is again switched on, and thyristor
SCRl is simultaneously switched on. If the potential at
the lower plate of Cl exceeds B , as will normally be the
case, diode Dl will be reversed biased, and current will
endeavour to flow from Cl and SCRl through coils 22a and
22c and device GTO2 to ground. ~he rate o~ build up of
current through the coils will depend on the potential
available at the lower plate of Cl, provided that its
upper plate has a low impedance path to ground, in this
case through a terminal of the supply. Since this avail-
able potential will usually be much greater than the
supply potential, the rate of current build up on the coils
will in turn be much greater than wou].d be the case were
only the regular supply potential available to produce
this build up.
The interaction o the rotor and stator to seek minimum re-
luctance position is an attempt to minimize the energy stored
: in the ~agnetic circuits, the energy released upon result-
ing relative movement ~eing available as mechanical energy
~ 32455
21 -
(di~regarding iron, copper and frictional losses). When
the motor is running, energy will be required from the
supply to supplement that available from the capacitor Cl
only to make up losses and to replace mechanical energy
delivered by the motor to a load. Each time the device
GTOl is switched on, current to the windings will be
initially supplied from the capacitor Cl. Whe~ the poten-
tial on the lower plate of Cl drops bel~w the supply
potential, make up current will then pass from the supply
through diode Dl until the device GTOl is switched off.
Under overrun conditions, the back EMF generated in the
coils may be such that potential on lower plate of Cl
never drops low enough to allow diode Dl to become forward
biased.
In the arrangement shown, the upper plate of capacitor Cl
draws current from the supply during its discharge, and in
effect appears, so far as the coils are concerned, in
series with the load. 1~ the upper plate (as shown in
Figure 4) is connected to ground then it will draw no
current from the supply during discharge, but will draw
current during charging. Either of these connections may
be utilized, or any other connection which provides a low
impedance path between the upper plate and the supply.
For various reasons, it may be desirable to withdraw
: 25 energy from the capacitor Cl. Firstly, it may be necessary
to ~void the build up of excessive potentials across the
capacitor which might cause breakdown of the associated
: semiconductor devices or the capacitor itself. Seondly,
such withdrawal enables energy recovery from the motor
during overrun conditions or if the motor is being driven
to form a genera~or. Thirdly, it enables a measure of
speed control to be exercised by increasing the rise time
of the current in the windinys. Finally, it enables the
motor to be utilized as a step-up DC to DC converter,
since the output potential which can he o~tained can be
4~5
22 -
considerably higher than the supply potential. A possible
means for achieving such energy withdrawal is il~ustrated
in Figure 4 in broken lines, in the form of a suitably
controlled thyristor SCR2, having its anode connected to
the lower plate of Cl, and its cathode connected to a
load RL.
The timing of the trigger signals applied to the primary
switching devices such as GTOl and the secondary switching
devices such as SCRl is important to the attainment of
maximum power output. Some adjustment of the theore*ical
optimum positions is desirable in order to ensure a par-
ticular sense of rotation, and in order to facilitate
starting.
In practice, with apparatus of the type shown in Figure 2
operating in the mode first described, successful results
have been achieved by energizing the coil of each electro-
magnet to repel its associated pole of toothed wheel 14
about 7.5 past top dead centre and maintaining its coil
energized until the pole in questions has reached about
22.5 past top dead centre. These figures are exemplary
only and should not be construed as limiting. For example,
by delaying turn off of a winding, the rotor will be sub-
jected to a countertor~ue as it moves past its minimum reluc-
tance position, reducing the net transfe~ of kinetic e~ergy
by the rotor and thus reducing motor speed for a gi~en load.
It should be noted that as rotor 10 speeds up (for example
as result of increasing the voltage of the D.C. power
supply), it is necessary, in order to obtain energization
and de-energization at these angles, to move both photo-
cell 27 and light sources 26 relative to timing wheel25 so that photocell 27 is activated earlier in the
cycle. This is because of the finite time required for
the triggering current to build up. Movement of these
components can be effected manually or automatically. In
~245`~
-- 23 --
the latter case the holder for the photocell and light
source can be driven by a motor whose output shaft posi-
tion is responsive to changes in the speed of shaft 11.
While preferred embodiments of this invention have been
described herein, the invention is not limited thereto,
and those skilled in the art will appreciate that changes
may be made therein withvut departing from the spirit
and scope of the invention as defined in the appended
claims.
Thus, although in the embodiment described and in the
following claims, the connection to the capacitor Cl are
described as being made to the ends of the phase winding,
it is intended that functionally equivalent arrangements
be comprised wi~hin the scope of the invention. Thus the
connections to the capacitor could be made from an addi-
tional portion of the winding}transformer coupled to the
remainder.
:,
An example o~ such an arrangement is illustrated in Figure
6, showing the essentials of an alternative circuit which
may replace the circui~ of Figure 4. A number of the com-
ponents whose function is the same have been allotted the
same reference numerals. The windings 22a and 22c have
been represented simply by the reference 22, which may
represent either a pair of diametrically opposed stator
windings or a single winding accordiny to the configura-
tion of the motor. The gate turn off device has been
shown replaced by a switching transistor TR10 which will
operate similarly but requires a continuous signal at its
base electrode during the time that it is turned on. The
capacitor Cl has been shown with its alternative connec-
tion to ground rather than B~.
The abave changes essentially merely illustrate possible
^~
~'~a2~55
- 24 -
variants already discussed above. The most significant
difference between Figures 4 and 6 is the use of a secon~
dary winding or windings 23 to derive the chargîng cur-
rent for capacitor Cl. This winding (or windings) 23 is
pre~erably bifilar wound with the winding(.s~ 22 to pro-
vide clo~e coupling (although a closely coupled step-up
con~iguration could be used) and is isolated from the
primary windings. One end of the winding 23 is connected
to diode Dl which provides a unidirectional path from B~
through winding 23 to capacitor Cl. Although ~he other
plate of capacitor Cl may be connected to either B~ or to
ground, as in the case of the embodiment of Figure 4, the
other end of the winding 23 should be connected to B+,
either directly or through diode Dl, rather than to
ground, since otherwise a potential equal to the supply
potential must be developed across winding 23 be~ore the
capacitor can begin to charge.
