Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
~ l77s3n
BACKGROUND OF THE INVENTION
This invention relates to a circuit for electron-
ically controlling the rate of activation of an electro-
magnetic solenoid and, more particularly, the control of
the release of an elevator brake.
While this invention is primarily intended for,
and has been developed for use with an elevator brake magnet,
its principles are not limited to elevators. As will be
understood from the discussion below, this invention has
general applicability wherever it is necessary to obtain
smooth motion from a direct current electro-magnet.
~ n elevator brake is applied by springs which
force the brake shoes against the brake drum or disc to
prevent rotation of the hoist n,otor. The brake shoes, hhich
are attached to a plunger influenced by an electro-magnet,
~ 43g5
~ li7753n
are moved away from the drum or disc by the energization
of the brake electro-magnet, when it is required that the
hoist motor rotate. Although some slow speed elevator
systems depend entirely Oll the brake for stopping the ele-
vator at each normal stop, most systems use the brake onlyas a holding brake and occasionally as an emergency stop.
The present state of the art is such tha. the hoist motor
can be brought to a complete stand still, regardless of
the load on the car, prior to the de-energization of the
magnet that causes the application of the brake. Thus
it is not necessary to be concerned with the smoothness
of the application of the brake, because it has no effect
on the smoothness of the final stop as experienced by a
passenger in the elevator car. The same cannot be said
of the release of the brake when t~e car is about to start.
Although smooth starting performance is usually obtained
with ease when the weight of the car plus its load equals
the weight of the counterweight, considerable difficulty
can be experienced at other loads. The situation is somewhat
equivalent to starting an automobile on a hill, i.e. release
of the brake may allow the car to move even before the
motor is started.
The most obvious method to overcome the difficulty
of achieving smooth starting performance is to measure
the load `in the car while the doors are closing in preparation
for a trip. Then, the motor torque can be caused to assume
an appropriate value such that the hoist motor will not
move when the brake lifts. Once the brake has been lifted,
i 1` 24385
~ 377~3(~
the start can proceed normally with performance identical
to a start at balanced load. This method has disadvanta~es,
however. Accurate measurement of the load is difficult.
If the motor torque is brought up to the appropriate level
while the doors are closing, there is a potential hazard
because faulty operation of this system might apply sufficient
torque to rotate the motor in spite of the brake. There
must of course be some feedback signal related to the measure-
ment of torque, which might consist of a measurement of
the armature current of a D.C. motor, and failure of this
feed-back signal could permit maximum torque to be applied.
Further, if the motor torque is brought up to the appropriate
level only after the doors are fully closed, the start
may have to be delayed by a noticeable amount, thereby
reducing the performance of the elevator.
Another method for improving the smoothness of
starting is to use a very special design for the brake
electro-magnet. Experience has shown that extreme smoothness
in the motion of the brake shoes under the influence of
the magnet is required to get a smooth start. Any sudden
change in the braking force, when there is motor torque
or a w~ight ~nbalance trying to rotate the motor, results
in a rough start that is noticeable to a passenger in the
car. Brake magnets have the characteristic that as the
plunger on which the shoe is located moves toward its fully
energized position, it inherently reduces the air gap.
As the air gap decreases, less and less current is required
to produce a given ~orce. Thus a basic instability exists,
l 1 243~5
1 ~ 7753()
i.e. regardless of how slowly the c~rrent rises, a point
is reached where the reduced air gap causes increased force
which further reduces the air gap, and the brake moves
rapidly to the ful~y released position.'
By very careful design of the brake magnet, gen
erally by having nsteps" of increasing diameter on the
plunge-, it is possible to overcome this instability.
Such a brake has a smooth curve, not necessarily straight,
relating brake current to brake plunger travel. The inherent
inductance of the brake coil forces the current to rise
relatively slowly, and thus the brake lifts smoothly.
SUMMAR~ OF THE INVENTION
The purpose of this invention is to accomplish
electronically what was previously obtained by the special
mechanical design of the brake plunger. This purpose is
achieved by controlling the current activating the brake
magnet, or any other electro-magnetic solenoid, so that
the solenoid is applied at a constant rate.
While the relationship between brake current
and plunger travel has an inherent discontinuity, the rela-
tionship between magnetic flux and plunger travel does
not have any such discontinuity. While this relationship
is not linear, it does not matter because smooth application
of the brake can still be accomplished by controlling the
flux of the solenoid in a feedback circuit. Although it
is not convenient to directly measure the magnetic flux
in a magnet for this purpose, it is very easy to accurately
3~5
~7753~
measure the rate of change of magnetic flux and to integrate
this signal to arrive at a flux signal. This is done by
adding a second winding to the brake coil. This extra
winding can consist of a relatively few turns of very fine
wire that occupy very much less space than the main winding.
This extra winding will be referred to as the "sensingn
winding.
The voltage induced into this sensing winding~
and measureable at its terminalsg is precisely proportional
to the rate of change of magnetic flux in the main winding
forming the electro-magnet. Thus, if the brake coil is
energized in such a way that the integral of the voltage
in the sensing winding is constant, smooth operation must
occur. To accomplish this it is necessary that a variable
voltage be applied to the magnet coi~.
It is very efficient to employ a static switch-
ing device, such as a power transistor, to control the
brake energization. Then, smooth operation can be obtained
by controlling the switching on and off of the transistor,
which is connected in series with the brake coil, in accor-
dance with the voltage induced in the sensing winding.
The res~lt is the same as that achieved for a variable
voltage supply.
The switching transistor commonly used to control
~5 the brake current of elevator brake coils normally is turned
fully on until the brake current reaches the desired level,
and then switches on and off with the dwell ratio automatically
adjusted to hold the current at the desired level in spite
--5--
~ ~7753n
of changes in temperature, coil resistance and line voltage.
This same transistor could be used to achieve the smooth
brake release of this invention.
In accordance with the present invention, there is
provided a col~trol for the activation or deactivation of an
electro-magnetic device having a primary operating winding,
comprising a first switching means for controlling the flow
of current through the primary winding of the device, a sensing
winding, magnetically coupled to said primary winding, in
which a voltage related to the rate of change of flux in
said primary winding is induced, control means for controlling
the switching of said first switching means in response to a
predetermined change in flux; means coupled to receive said
induced voltage for producing a change of flux signal and for
supplying said signal to said control means; and biasing means
for altering said signal for producing a change of the average
flux level in said electromagnetic device at a controlled,
predetermined rateO
BRIEF DESCRIPTION OF T~E DRAWINGS
. .. .. _ . .... .. . . _
The foregoing and other features of the present
invention will be more readily apparent from the following
detailed description and drawings of illustrative embodi-
ments of the invention in which:
Figure 1 shows a circuit which illustrates the basic
principles for the control of the magnetic flux in an
electro-magnet; and
Figure 2 shows a circuit in which the release of
an elevator brake is controlled by this principle.
DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS
_ . _ .... _ . _
In Fig. 1 there is shown an electro-magnet with
a primary winding BCl and a sensing winding BC2. The primary
winding BCl is energized by current flowing through winding
BCl and power transistor 1~ from a suitable D.C. power supply
~ ~ 7753(~
level +Vl to a level OV~ A diode D6 is connected across
winding BCl to provide a path fQr the inductive current
to decay into when the transistor 18 turns off.
Transistor 18 can be switched on or off in accor-
dance with the output of operational amplifier 12 via an
amplifier comprised of operational amplifier 14, resistors
R13, R14, and R15, diodes D3, D4 and D5, and transistor 16.
- 6a -
~91~5
~77~3~
Power- for this amplifier, and for the remainder of the
circuit of Fig. 1 is obtained from a DC power s~pply which
provides a positive voltage +V2 and a negative voltage
-V3 at a suitable level, such as plus and minus 5 volts J
respectively, with respect to level OV, which is preferably
at ground potential.
Operational amplifier 12 is operated as a toggle
device so that its output has only two states; one state
is positive with respect to level OV and the other state
is negative. When the output of operational amplifier
12 is positive, transistor 18 is turned off because the
output of operational amplifier 14 is also positive and
current through resistor R13 flows into the base of transistor
16 to turn it on. When transistor 16 is turned on, transistor
18 is turned off because its base is held slightly negative
due to the forward drop in voltage across d;odes D4 and
D5 from current flowing through resistor R15.
When the output of operational amplifier 12 is
neqative, transistor 18 is turned on because the output
2Q of operational amplifier 14 is also negative and this turns
transistor 16 off. ~Jhen transistor 16 is off, current
flo~s f~om +V2 through resistor R14 and through diodes
D4 and D5 into the base of transistor 18, which causes
it to turn on.
An ~nderstanding of the remainder of Fig. 1 can
best be obtained by first assuming that potentiometer Rl
has its slider exactly in the miàale so that it is at the
same potential as line OV. ~lso, it is ass~med that the
! 1 2~3~5
~ ~ 7753()
c~rrent through winding BCl is at some intermediate level,
neither zero nor maximum. Under these conditions, the
circuit oscillates at a fre~uency determined by capacitor
Cl, resistor R3, voltage +Vl and the turns ratio between
the solenoid winding BCl and the sensing coil BC2. For
satisfactory operation,-this frequency must be high enough
to achieve substantially steady current in winding BCl, with
the alternating component relatively small. The inductance
of the electro-magnet coil, of course, has the effect of
smoothing out the current in spite of the very high alter-
nating component in the voltage across the coil. Generally,
for typical magnets, a very wide range of frequencies is
acceptable.
This oscillation can be explained by starting
with the condition where transistor ~8 is turned on. Then
current is increasing in winding BCl and a negative voltage
is induced in sensing winding BC2. This voltaye is applied
to an integrator comprised of operational amplifier 10,
resistors R3 and R8, and capacitor Cl. Diodes Dl and D2
are for protection only and do not affect the normal operation
of this integrator. The output of this integrator swings
in a positive direction at a rate proportional to its input
voltage, and when it reaches a predetermined positive voltage,
as determined by resistors R9 and Rll, the output of opera-
tional amplifier 12 toggles abruptly to its positive state.This causes transistor 18 to turn off, as described earlier.
When transistor 18 turns off, the current in
winding BCl decays through diode D6, and the voltage ind~ced
1 37753()
in winding sC2 changes to positive. This causes the output of
the integrator to swing in a negative direction at a rate
proportional to its input voltage, which is the voltage induced
in sensing winding BC2. When the output of the integrator
reaches a predetermined negative voltage, again determined by
resistors R9 and Rll, the output of operational amplifier 12
toggles abruptly to its negative state. This turns transistor 18
on again, and thus one complete cycle of the oscillation has
occurred. Further similar cycles occur in repetition.
Since the voltage induced in sensing winding BC2 is
proportional to the rate of change of flux in the electro-magnet,
the integration of this voltage results in a measurement of the
actual change in flux. Each time transistor 18 turns on, the
flux rises in the electro-magnet b~ a small, but specific amount.
Each time transistor 18 turns off, the flux drops by an identical
amount. Thus, on the average, the flux remains constant.
The preceding description was based on the assumption that
potentiometer Rl had its slider at the same potential as OV. If
it is assumed that the slider is moved toward voltage ~V2 so as
to put a positive voltage on resistor R4, the output of the
integrator will be biased so that it tends to swing faster in the
negative direction and slower in the positive direction. This
causes the increase in flux, when transistor 18 is turned on, to
be greater than the decrease in flux when transistor 18 is turned
; ~ 24385
~ 7 7~53 ()
off. This is perhaps easier to understand by stating that
the off time of transistor 18 will be shortened and the
on time lengthened by the bias applied to the integrator.
The result of this biasing of the integrator is that the
flux is increased at each cycle of the oscillation by an
amount proportional to the positive voltage on the slider
o~ potentiometer Rl. By similar seasoning, a negative
voltage on the slider of potentiometer Rl causes the flux
to be decreased at each cycle by an amount proportional
to this negative voltage.
There is, o~ course, a limit as to how quickly
the flux can be reduced. If too high a rate is demanded,
transistor 18 will turn off completely, and the rate of
decay will be determined by the L/R ratio of the electro-
magnet and winding BCl. There is, however, no limit asto how slowly the flux can be reduced. Similarly, there
is a limit as to how quickly the flux can be increased
because if too high a rate is demanded, transistor 18 t~rns
on continuously; but, no corresponding limit on how slowly
the flux can be increased, exists.
Figure 2 contains all of the components of Fig. l,
but additions have been made to make it suitable for control-
ling the energization of an elevator brake coil in order
to achieve a smooth release of the brake.
Brake contactors Bl and B2 are required to de-
energize the brake coil for emergency stops. Contacts
B1-1, Bl-2, B2-l and B2-2 are shown in series with the
brake winding BCl. These contacts would be all closed
--10--
2~3~5
~ 30
to run, and open for an emergency stop and perhaps also
at every normal stop, if desired.
Optical GOUpler 22, analogue switch 26 and resistor
R12 have been added to the circuit of Fig. 2 to provide
a means of turning transistor 18 off to de-energize the
brake coil. Optical coupler 20, analogue switch 24, and
resistors R5, R6 and R7 have been added to bypass the flux
control feature for the initial part of the energization in
order to get a faster brake release. Further, potentiometer
Rl has been connected to level OV through resistor R2 to
give a suitable range of adjustment for positive voltages
only, since this circuit only controls the flux when it
is increasing.
When the elevator brake is not being energized,
the two optical couplers have no current flowing through
their light-emitting diodes, and both photo-transistors
are turned off. Resistor R12 thus holds the control input
of analogue switch 26 at ground potential, and the internal
solid state connection is as depicted by the mechanically
e~uivalent blaae of a toggle switch. This puts positive
voltage from line +V2 into the input of operational amplifier
14 and ~his turns transistor 18 off. Similarly, resistor
R7 holds the control input of analogue switch 24 at oround
potential, ana this connects resistors R5 and R6 into the
integrator circuit to force the integrator output to a
nesative value sufficient to make the output of operational
amplifier 14 toggle to the negative condition.
~ I 7753n
When the brake is required to lift, the contactors sl and
B2 are energized, and current is applied through terminals T3
and T4 of optical coupler 22 by circuits not shown in Fig. 2.
The illumination of the light emitting diode in circuit 22
causes the photo-transistor to turn on. This applies a positive
voltage to the control input of switch 26 which causes its
internal solid state connection to switch to the alterrlate
input, which is connected to the output of operational
amplifier 12. Since the output of amplifier 12 is presently
negative, transistor 18 turns on, and brake current starts to
flow. The control of flux, however, is not yet in effect.
After the brake current has risen to a predetermined value,
as determined by circuits not shown in Fig. 2, current is
supplied to terminals Tl and T2 of optical coupler 20. This
causes analogue switch 24 to connect the input of the integrator
into the feedback circuit for sensing winding BC2 so that the
further increase of flux proceeds at a rate determined by the
setting of potentiorneter Rl.
For general industrial use, the delay of the start of flux
control may not be required. For elevator brake control,
however, this delay is desirable so that the brake current rises
initially as fast as possible to minimize delays in starting.
The flux control delay, however, cannot be too great because the
current sensing means must cause the switch to flux control to
occur before the previously described magnet instability occurs.
Without this flux control delay feature, there may be too much time
~ 4385
~ 117753(~
delay between the initial energization o~ the brake and
the start of brake plunger movement.
Although the example of Fig. 2 is particularly
useful for elevator brake release, the invention is in
no way limited to elevator brakes, and is equally appiicable
to the control of the rate of activation or release of
any direct current electro-magnet.
-13-