 |
Description  |
|
|
FIELD OF THE INVENTION
This invention relates to position control devices and more particularly to
a bearing for use in a high resolution device for precisely maintaining
the position of an object or for precise small movements of an object.
BACKGROUND OF THE INVENTION
There are many applications where an object must be positioned with high
precision in the nanometer range. Such applications include scanning
tunneling microscopy, precision mechanically suspended linear slides and
XY stapes, such as those used in the manufacture of electronic circuit
chips, various precision optical applications, and diamond turning
machines. In such applications and others, the object must not only be
precisely positioned, but the position must be maintained regardless of
perturbations to the object resulting from air currents, vibration,
temperature variations and the like. In some applications, such a fine
position control mechanism may be used alone, while in other applications
it may be used as a fine motion control stage in a mechanism which also
includes a coarse motion control stage for positioning the device over
larger distances with a lower degree of resolution.
While various bearings have been utilized in the past in such applications,
including mechanical bearings, flexures, gas and liquid fluid bearings,
etc., magnetic bearings have been found preferable for such applications
for a number of reasons. Such bearings or suspensions have superior
controllable stiffness, have less cross-coupling between modes with
multiple degrees of freedom and are generally easier to control. They are
not subject to rough spots, friction or wear as with mechanical bearings,
and are also simpler to design mechanically, generally involving only a
single moving part. However, while magnetic bearings offer a number of
advantages, they also provide various control problems, particularly when
operating with nanometer or Angstrom resolutions.
One approach to developing a magnetic bearing for fine motion control is
discussed in an article entitled "Design Considerations for Ultra
Precision Magnetic Being Supported Slides" by A. H. Slocum and D. B.
Eisenhaure, NASA Magnetic Suspension Technology Conference, Hampton, VA
Feb. 2-4, 1988. However, the device described in this paper had a number
of limitations which adversely affect its performance.
First, while permanent magnets were used in this device to support the
weight of the object or platen to be moved, the gap for the permanent
magnet was substantially the same as the gaps for the electromagnets used
for positioning the platen or maintaining the position thereof. These gaps
were relatively small, in part because position sensing capacitive probes
were mounted within the coils of the electromagnets. This results in a
relatively large unstable frequency of suspension, thereby making the
control problem far more difficult. One of the objects of this invention
is to provide enhanced stability in a small motion system.
Another problem with the prior system is that it did not operate the
electromagnets in push/pull mode in all dimensions and for all degrees of
freedom. This inhibited the ability to correct for various errors and also
reduced stability of the system. The control problem was also complicated
by using magnets of different size. Other problems included mounting the
probes in the electromagnets which increased the size, and thus the
required currents, for the electromagnets and also restricted the gaps
available for these magnets. Position sensing was also restricted to less
than the full range of motion of the platen, resulting in instability when
the platen moved outside the sensor range. Further, the platen was a
hollow tube which exhibited significant resonance at one or more selected
frequencies. It is desirable that the platen not be designed to permit
such resonance.
The prior art system also was unable to compensate with the permanent
magnet for variations in platen weight which might occur, for example,
when a load was placed on the platen to have work performed thereon. Since
enhanced stability and control is achieved, and electrical usage is
minimized, where the platen weight is always being supported solely by the
permanent magnet, regardless of changes in the weight of the platen, a
need exists for a mechanism to permit the magnetic force of the permanent
magnet exerted o the platen to be adjusted to achieve this objective.
In addition, it is desirable in some applications that the system be damped
beyond the damping provided solely by the magnetic bearings. Since air
provides little damping, such additional damping could be provided by
suspending the platen in a viscous fluid, ferrofluid, or similar media for
improving damping and high frequency coupling between the platen and a
frame or housing in which the plater, is mounted. Where such fluid is
provided, it might also be utilized to carry the weight of the platen in
lieu of or in addition to the permanent magnet and to adjust for weight
changes of the platen. Where a ferrofluid is employed, it may also be
utilized, in conjunction with the electromagnets, to control platen or
object position in some or all degrees of freedom.
SUMMARY OF THE INVENTION
Thus, it is a primary object of this invention to provide an improved
magnetic bearing suitable for use in a fine position control or movement
system which affords high stability by supporting the weight of the platen
or other object to be moved with a first means regardless of the weight of
the platen, which utilizes separate electromagnets connected in push/pull
fashion to control or compensate for motion of the platen in N degrees of
freedom and which uses detectors independent of the electromagnets to
detect movement for effecting desired corrections. The first means for a
preferred embodiment is a permanent magnet having a gap with the platen
which is greater than that for the electromagnets.
More particularly, this invention provides a bearing which has a platen, a
frame in which the platen is positioned and a mechanism for suspending the
platen in a predetermined position in the frame. The suspension mechanism
includes a permanent magnet having a first air gap or other suitable means
for carrying the weight of the platen, and electromagnet means having a
second air gap and preferably operated in push/pull fashion for
controlling the position of the platen in the frame with N degrees (for
example six degrees) of freedom. The first air gap is preferably larger
than the second air gap. The suspension mechanism also includes detectors
for the position of the platen in the frame in the N degrees of freedom
and controls responsive to the detection by the detectors of a movement by
the platen for selectively energizing the electromagnetic means to restore
the platen to the predetermined position.
Where a load may be applied to the platen or removed from the platen, or
the weight of the platen otherwise changed, the magnetic force exerted by
the permanent magnet on the platen or the other means for carrying the
weight of the platen, may be altered, preferably by altering the first air
gap, to permit the permanent magnet or other means to carry the weight of
the platen regardless of any changes in such weight. The adjusting is
preferably accomplished by changing the air gap in response to the
detection of a current applied to a weight supporting electromagnet to
zero out such current and is preferably operative only for sustained
currents to the weight supporting electromagnets.
A means may also be provided in at least selected ones of the air gaps to
improve damping and high frequency coupling between the frame and platen.
This damping element or means may, for example, be a viscous fluid or
ferrofluid in at least the air gaps. Such fluid may also be utilized to
support the weight of the platen, either in addition to or instead of the
permanent magnet. Where ferrofluid is utilized, it may also be utilized in
conjunction with the electromagnets to move the platen.
The permanent magnet may be a single permanent magnet which interacts with
an iron bar, one of which is attached to the top of the platen and the
other one of which is attached adjacent thereto to the frame. Either in
addition to or instead of the permanent magnet operating in an attraction
mode, two permanent magnets may be attached in repulsion mode, one to the
bottom of the platen and the other adjacent thereto to the frame.
Where the position of the platen is controlled with L linear degrees of
freedom and R rotational degrees of freedom, there would be at least one
electromagnet for each linear degree of freedom and at least one
additional electromagnet for each rotational degree of freedom. For a
preferred embodiment, with L and R both equal to three, the electromagnet
means includes at least one electromagnet for each of two opposite sides
of the platen and at least two electromagnets for the remaining two sides
of the platen, the electromagnets on each side being on opposite sides of
a top/bottom center line for the side. At least four electromagnets are
provided for the top of the platen and for the bottom of the platen, the
four electromagnets for each of the top and bottom being in opposite
quandrants thereof. Electromagnets on opposite sides are positioned
opposite each other and all of the electromagnets are preferably identical
so as to permit the electromagnets to operate in push/pull fashion. For
each electromagnet, there is a corresponding iron insert embedded in the
platen which is spaced from the electromagnet by the second gap. The iron
inserts are preferably slightly larger than the corresponding
electromagnets to permit constant magnetic coupling to be maintained over
the full permitted range of movement of the platen in the frame and the
electromagnets and iron inserts are preferably positioned near each end of
each of the platen sides and near each corner of the platen top and platen
bottom. Ferrofluid in the second gaps may be substituted for the iron
inserts. Motion in a given degree of freedom may be achieved by
selectively energizing one or more of the electromagnets.
The platen is preferably formed as a honeycomb, foamed or other
hollowed-out cellular structure of a material which is light, stiff, has a
relatively low thermal expansion and has good damping characteristics.
Such material would preferably be a composite material.
The detectors should include at least one detector probe for each of the
degrees of freedom, the probes being mounted to the frame and spaced from
the platen. At least one of the probes would be positioned to detect
movement in each lateral direction of motion and at least two of the
probes would be positioned to detect each of roll, pitch and yaw. The
probes could be capacitive probes, optical interferometers, inductance
probes or other suitable detectors.
The foregoing and other objects, features and advantages of the invention
will be apparent from the following more particular description of
preferred embodiments of the invention as illustrated in the accompanying
drawings.
IN THE DRAWINGS
FIG. 1 is a partially broken away top side perspective view of a bearing
employing the teachings of this invention.
FIG. 1A is an enlarged cut away side view of a portion of the bearing shown
in FIG. 1.
FIG 2 is a partially broken away top view of the bearing shown in FIG. 1.
FIG. 3 is a partially broken away end view of a bearing of the t shown in
FIG. 1 with two modifications.
FIG. 4 is a sectional side view of a bearing of the type shown in FIG. 1
with an additional modification.
FIG. 5 is a sectional side view of a portion of a bearing of the type shown
in FIG. 1 with a still further modification.
FIG. 6 is a schematic block diagram of a circuit suitable for use in
controlling the device shown in FIG. 5.
FIG. 7 is a partially cut away side view of an embodiment of the invention
which provides unobstructed access to a work surface on the top of the
platen.
FIG. 8 is a partially cutaway top view for the embodiment of the invention
shown in FIG. 7.
FIG. 9 shows a device of the type shown in FIGS. 7 and 8 being utilized as
a fine position stage in a semiconductor positioning mechanism.
DETAILED DESCRIPTION
Referring to the figures, and in particular to FIGS. 1, 1A, and 2 which
illustrate a first embodiment of the invention, the bearing 10 of this
invention includes a platen 12 which is mounted in a frame 14. Platen 12
has a sample mounting or work area 16 on its top surface. A pair of
permanent magnets 18 are affixed to the top surface of the platen at
substantially the center near each end thereof. Magnets 18 may be secured
to the platen by gluing or other suitable means. The platen also has iron
inserts 20 embedded in it at a plurality of locations on each face of the
platen. For the preferred embodiment shown in the figures, there is an
iron insert 20 near each corner of the top and bottom faces or surfaces of
the platen and an iron insert 20 near each end of each side of the platen.
Iron inserts 20 are preferably laminar to avoid eddy currents and are
thick enough to provide good magnetic coupling.
Platen 12 for the preferred embodiment has a solid rectangular shape which
is preferable, but not essential, for permitting motion control in all
dimensions. In order to avoid resonances, instead of the platen being a
hollow tube structure as in some prior art devices, the platen is formed
with a honeycomb structure which may be seen in the cutaway portion of
FIG. 1A or with a foamed or other structure having hollowed out cells.
Since the soft iron inserts 20 do not provide significant structural
support, and it is desirable that the platen be as stiff as possible to
avoid errors caused by bending thereof, the material for the platen should
be as stiff as possible while still being light in weight and having good
damping characteristics. Since the platen may be exposed to substantial
temperature variations in operation, it is preferable that the material
for the platen also have relatively low thermal expansion characteristics.
Examples of materials suitable for this application include carbon fiber,
various ceramics such as Zerodur ceramic or Invar metallic alloy.
Frame 14 may be a solid structure formed of a nonmagnetic material and, for
certain embodiments to be discussed hereinafter, at least a portion of the
frame must be solid to retain fluid therein. Frame 14 can also be an open
bar or panel structure, particularly on its top side, to permit access to
the platen. There are also applications where the platen needs to be in a
high pressure, low pressure or other sealed environment, in which case
frame 14 would have walls which serve as the walls of such a sealed
chamber.
An electromagnet 22 is mounted in frame 14 adjacent each of the iron
inserts 20. For the preferred embodiment, all of the electromagnets 22 are
identical and corresponding electromagnets adjacent opposite faces of
platen 12 are directly opposite each other and operate in push/pull
fashion to control the position of the platen. Each of the electromagnets
22 for the embodiment shown in the figures consists of an E shaped iron
core which is preferably laminated, having electrical wiring wrapped
around its center section. The physical size and electrical power of the
electromagnets will vary with application. Each iron insert 20 is
sufficiently larger than the corresponding electromagnet 22 so that the
electromagnet will have iron insert under all portions thereof for the
full range of movement of the platen in the N directions or degrees of
movement for the platen.
Each electromagnet 22 is spaced from the corresponding iron insert 20 by an
air gap 26. As was previously discussed, the unstable natural frequency
increases undesirably as current is increased for a fixed air gap, or as
the air gap is decreased at fixed current. Conversely, the unstable
natural frequency decreases desirably as current is decreased at fixed air
gap or as the air gap is increased at fixed current. Thus, increasing the
size of the air gaps to approximately 0.030 inches would make the
suspension easier to control, and would also afford greater range of
motion where the system is being utilized as a small motion controller. In
other applications, the air gap could be larger or smaller as appropriate.
However, if permanent magnets 18 are utilized, as will be discussed later,
to carry the gravity load of platen 12, and the system is arranged so that
all electromagnets 22 are in push/pull pairs, then the system can run at
smaller air gaps with reduced bias currents. At small operating point
currents, the suspension open loop time constants can be made sufficiently
large even with smaller air gaps. This allows for the efficiency of force
production associated with small air gaps while retaining the advantage of
stable operation.
For each permanent magnet 18, there is an iron bar 28 mounted to frame 14,
which bar is large enough so that it will always be over the entire magnet
18 through the full range of motion of platen 12. Each of the iron bars 28
is spaced from the corresponding magnet 18 by an air gap 29. For preferred
embodiments of the invention, the air gap 29 is larger than the air gaps
26. This provides for stable operation afforded by larger permanent magnet
air gaps while still providing the operating efficiency resulting from
electromagnets operating at small air gaps.
Finally, a plurality of capacitive probes 3A-30F are mounted in frame 14.
Each probe 30A-30F is spaced from platen 12 by an air gap which is
dictated primarily by the range or sensitivity of the probe. There is a
capacitive probe 30A-30F for each direction or degree of motion of concern
for platen 12 with at least one of the capacitive probe 30A-30F being
utilized to detect each linear degree of movement and at least two of the
probes 30A-30F being utilized to detect each rotational degree of motion.
Thus, the probes 30A, 30B and 30C all detect linear motion in the up/down
direction, the X direction as shown in FIG. 1. Probes 30D and 30E detect
motion in the forward or Y direction and probe 30F detects linear motion
in the lateral or Z direction. Probes 30A, 30B and 30C also detect pitch
motion of the platen (i.e. rotation about the Z axis), while probes 30B
and 30C detect roll motion (rotation about the Y axis). Probes 30D and 30E
can be utilized to detect yaw motion (i.e. rotation about the X axis).
While a particular arrangement of the probes 30A-30F is shown in the
figures, this arrangement is primarily for purposes of illustration and
other arrangements are possible. For example, a probe may be provided
adjacent each electromagnet 22 for the three faces having probes; and
where two or more probes are provided for a single face, one or more of
such probes may be on the opposite face. All that is required is that
sufficient probes be provided to be able to detect each degree of motion o
concern. Further, while capacitive probes have been shown in the figures
for purposes of illustration, other probes known in the art, including
optical interferometers and induction or eddy current probes, might also
be utilized.
FIG. 3 is an end view of a bearing of this invention which illustrates two
variations on the embodiment of FIGS. 1 and 2. First, for the earlier
embodiments, the weight of the platen is supported solely by the
attraction of magnet 18 to iron bar 28. In FIG. 3, this support for the
weight of the platen is supplemented by a magnet or magnets 40 secured to
the bottom of the platen and a second magnet 42 of polarity opposite to
that of magnet 40 mounted to frame 14 opposite each magnet 40. Where two
magnets 18 are employed, as shown in FIGS. 1 and 2, it is preferable that
two magnets 40 also be employed which are positioned on the opposite side
of platen 12 from the magnets 18. The repulsive force between magnets 40
and 42 further serves to support the gravitational weight of platen 12.
While both an attractive and repulsive arrangement are shown in FIG. 3 for
supporting the weight of the platen, two magnets connected to operate in a
repulsive mode may be used alone to support the weight of the platen.
Second, in FIG. 3, the spaces 26 between the electromagnetics and platen 12
are shown filled with a fluid 44. This is done because air has relatively
poor damping characteristics, and damping of platen movement caused by
vibration or other perturbations on the platen can thus be enhanced by
providing an oil or other damping fluid in spaces 26. While not
specifically shown in FIG. 3, optimum damping can be obtained by immersing
the entire platen or substantially the entire platen in the damping fluid.
However, enhanced damping also results in increased response time for the
system in applications where the system is being used as a small motion
controller, and it may, therefore, be desirable to reduce the amount of
damping fluid utilized in such applications.
It is also possible to use ferrofluid as the fluid 44. When this is done,
iron inserts 20 may be dispensed with, movement of the platen being
controlled by applying current or additional current to selected
electromagnets 22 to attract additional ferrofluid into the gap between
such electromagnet and the platen. The controlled flow of ferrofluid into
the gap exerts a pushing force on the platen to effect the desired motion
thereof. With ferrofluid as the fluid 44, it may be necessary to maintain
a small current in electromagnets 22 to maintain the ferrofluid in the
gaps. Ferrofluid in the gaps may also be used to supplement rather than to
replace motion control resulting from the interaction of the electromagnet
and the iron inserts. If other fluid is utilized and the entire platen is
not immersed in the fluid, some other mechanism known in the art may be
required to retain the fluid in the gaps.
FIG. 4 illustrates still another variation on the invention wherein frame
14 is a solid chamber at least for the lower portion thereof which is
filled to a predetermined level with a viscous fluid 50 such as oil. The
average density of the fluid is greater than the average density of the
material forming platen 12 so that the platen floats at a predetermined
and controllable level in frame 14. The weight of the platen is thus
completely supported by fluid 50 and attractive or repulsive permanent
magnets such as the magnets 18, 40 and 42 are not required to support the
weight of the platen. Except for this difference, the embodiment of FIG. 4
functions in the same manner indicated for the previous embodiments.
FIG. 5 illustrates another variation on the embodiment shown in FIG. 1
which permits the weight of platen 12 to be supported by magnet 18 and
iron bar 28 regardless of changes which may occur in the weight of the
platen. Thus, where the platen is, for example, being utilized for
controlled positioning of a semiconductor device, placing the
semiconductor material to be operated on in sample mounting area 16 will
result in a predetermined increase in the weight of platen 12. Normally,
this additional weight would be carried by increasing the current flow in
the electromagnets 22 above the platen to restore the platen to its
initial position and maintain it therein. If the weight of the platen were
reduced, current would flow in the electromagnets below the platen to
restore equilibrium. However, having current flowing in the electromagnets
over an extended period of time results in increased current usage and
thus increases the operating costs of the system. It also results in heat
being generated which must be dissipated in order to avoid thermal induced
errors which can become significant when operating in the nanometer range.
Finally, current continuously flowing in some of the electromagnets
reduces the flexibility of the system to respond to movement
perturbations. It is, therefore, desirable that the weight of the platen
be at all times supported solely by the permanent magnet or magnets
regardless of changes in platen weight.
Therefore, for the embodiment of FIG. 5, metal plate 28 is mounted to frame
14 by lead screws 60, a rack and pinion or other suitable means which
permit the spacing between metal plate 28 and frame 14 (g ) to be altered,
thus altering the gap g.sub.O between permanent magnet 18 and metal plate
28. Varying the gap g.sub.O varies the magnetic force between magnet 18
and plate 28. Therefore, if the weight of platen 12 is varied, the lead
screws 60 or other suitable means may be operated by a suitable drive
mechanism 61 to vary gap g.sub.O in an appropriate direction so that the
magnetic force between magnet 18 and plate 28 is again sufficient to
support the new weight of platen 12. Where two magnets 18 and two plates
28 are employed, the gaps g.sub.O magnets would be concurrently adjusted.
While typically the adjustments for both magnets would be the same, they
may be different in situations such as where the load is placed off center
on the platen requiring a torque correction.
While FIG. 5 shows the adjustment being made for a magnet and plate
operating in an attractive mode, where magnets 40 and 42 are | | |
