Note: Descriptions are shown in the official language in which they were submitted.
~ 163299
This invention relates to devices for obtaining precise
angular and rectilinear movement of an optical system element.
In optical systems, it is often necessary to control the
position or velocity of system elements such as light sources,
detectors, mirrors, lenses and optical fibers. Such positioning must
be extremely precise. For example, an 8 micron diameter laser beam,
coupled into a 10 micron optical fiber core with a misalignment of more
than one or two microns yields unacceptable coupling losses. In
addition to aligning the axes of the fiber and laser beam, it is
usually desirahle to be able to move the fiber axially towards and away
from movement of the laser source. Thus there is need for controlled
rectilinear movement of the fiber in three mutually perpendicular
directions.
A similar positioning problem arises when using mirror
deflectors, for example for laser scanning of optical discs. Here it
is necessary to angularly tilt the mirror and also to move the mirror
bodily to vary or maintain the optical path length of a light beam
incident on the mirror.
Known devices for positioning such optical system
2n elements include pairs of orthogonally-disposed screw adjusters, moving
coils, electromagnetic actuators, or piezoelectic devices, each pair
capable of moving the fiber in two mutually perpendicular directions.
Such devices do not readily permit either controlled movement in the
third orthogonal direction or anguldr movement of the elements.
Moreover, such arranqements are complex, delicate, and frequently both
difficult to set up and tedious to use.
1 ~632~3g
The present invention avoids some of the complexity and
delicacy of known arrangements for moving optical system elements by
using materials known as ferromagnetic fluids. As clearly set out in
United States patent ~o. 3,488,531, ferromagnetic fluids are fluids
which become polarized in the presence of a magnetic field with
consequent pressure redistribution and, possibly, fluid motion.
Typically, ferromagnetic fluids are colloidal or colloid-like fluids
comprisinq a carrier liquid in which extremely fine particles are
suspended.
As indicated in the above-mentioned patent, ferromagnetic
fluids can be produced by grindinq a suitable ferromagnetic body and
colloidially dispersing ferromagnetic particles so that the combination
has particular fluid properties described for example by Rosenwig et
al, "Ferrohydrodynamic Fluids for Direct Conversion of Heat Energy",
Joint A.I.CH.E/lnstitute of Chemical Engineers Meeting, London, June
1965.
According to one aspect of the invention there is
provided a positioning device for accurately positioning an optical
system element, the positioninq device comprising a mass of
ferromagnetic fluid; restoration means tending to maintain the mass of
ferromagnetic fluid in a stable condition with a predetermined fluid
pressure distribution; a non-magnetic body in contact with the
ferromagnetic fluid mass; magnetic energizing means for energizing the
mass of ferromaqnetic fluid to redistribute pressure within the fluid
whereby to move the non-magnetic body rectilinearly along a first axis
and angularly about a second axis.
~ 163299
Preferahly the res-toration ~leans can include any one or
cornbination of permanent magnets, electromagnets, gravity, and a fixed
support or confining hody. Particularly for positioning a mirror, the
fixed support body can be a horizontal substrate. Thus the ferro-
magnetic fluid can he supported on the horizon-tal suhstrate with
enerqizing electromagnets below the substrate adapted to produce
localized magnetic field perturbations. If the perturbations are
distributed non-linearily and cause the ferromagnetic fluid -to
concentrate at at least three non-linearily distributed locations, then
a three-point support is provided for the mirror or other optical
system element. By individually varying the height of the
ferromagnetic tluid concentrations by differentially energizing the
electromaqnets, the mirror or other optical system element can be made
to tilt or can be bodily lifted relative to the suhstrate.
Particularly for supporting an optical fiber, the
confined body can be a cylinder within which the mass of ferromagnetic
fluid is retained to surround the fiber. The fiber can he mounted
within a specially shaped body which is itself immersed within the
ferromagnetic fluid. By arran~ing electromagnets around the hody, the
hody and the fiber held -thereby, can he made to move in an XY plane
perpendicular to the fiber axis. By further arranging (1roups of
individually energizahle electromagnets spaced along the axis of the
fiber, the body can be made to move such as to tilt the fiber axis.
The holding body can be so shaped that differential energization of
electromagnets spaced along the fiber axis can cause movement of the
fiber also along the Z axis.
~ 1~3~99
Preferably the positioning device includes a servo-loop
including an optical transducer for detecting motion of the optical
device, a feedback circuit controlled by the optical transducer, a
summing device for summing a signal from the feedback circuit and a
command signal, and a drive curcuit for independently driving the
electromagnets. The optical transducer can, for example, be a position
transducer or a velocity transducer. The optical input to the
transducer can be derived either from a primary optical beam controlled
by the positioning device or from a dedicated secondary optical beam
made dependent on the position of the optical system element controlled
by the positioning device. The optical transducer can include a
quadrant photodetector. The ferromagnetic fluid can be contained
within a flexible membrane.
Emhodiments of the invention will now be described, by
way of example, with reference to the accompanying drawings in which:-
Figure 1 is a perspective view with part cut awayof one embodiment of positioning device in which ferromagnetic fluid is
supported on a horizontal substrate;
Figure 2 is a sectional elevation of a second emhodiment
2n of positioning device;
Figure 3 is a cross sectional view of a positioning
device for positioning an optical fiber;
Fiqure 4 is a longitudinal section of the Figure 3
positioning device; and
Figure 5 is a block schematic diagram of a control
circuit applicable -to positioning devices according to the invention.
~ ~6.329~
Referring in detail to Figure 1, there is shown a mirror
1n resting on a mass 30 of ferromagnetic fluid. The ferromagnetic
fluid is supported on a ceramic substrate 14 which is itself mounted
over an electroma~net assembly 16. The electromagnets can be energized
to redistribute the ferromagnetic fluid mass and so alter the position
of the mirror 10.
The electromagnetic assembly 16 is linked to a central
permanent magnet 18 having a vertical magnetic axis. The magnet 18
supports an upper cruciform soft iron yoke 20 and is itself supported
by a lower cruciform yoke 22. The yokes 2n and 22 have vertical limbs
24 and 26 respectively on which the substrate 14 is mounted. A pole
pair 27 is established in the space hetween each pairing of a limb 24
and a limb 2fi. Windings 28 surround the limbs 24 and 2fi and are
connected to a power supply and control circuit (not shown). The mass
of ferromagnetic fluid consists essentially of four fluid
concentrations 30 which rest upon the upper surface of substrate 14.
The fluid concentrations or masses 30 are located one above each of the
four pole pairs 27 and are retained in place by the residual magnetic
field produced at the pole pairs 27 by the permanent magnet 18. A
suitable ferromagnetic fluid is obtainable from Ferrofluidics Inc.
under the specification No. A06.
The mirror 10 rests upon the four fluid masses 30, the
strength of the residual magnetic field of magnet 18 being sufficient
to suspend the mirror at a predetermined height above the substrate 14.
Essentially, the magnetic particles within the fluid are attracted
towards any region of high magnetic field strength and cause a
consequent pressure distribution within the fluid. If the windings 28
~ 163299
are energized, the magnetic field existing at pole pairs 27 is
increased and the particles within the mass 30 overlying any particular
pole pair are attracted towards the region of maxirnum field strength.
In consequence, the fluid bunches more closely over the pole pairs 27
causinq an upward shift of the mirror away from the substrate 14.
Conversely, if the magnetic field is weakened, the fluid masses are
lo~Jered by a restoring force attributable to a combination of gravity
and surface tension. The mirror 10 is thereby lowered. If opposed
pairs of the windings 28 are differentially energized or are excited in
1n opposite directions, the mirror tilts. The individual ~indings 28 can
be excited so as to oppose the field due to the magnet 18.
Alternatively, the magnet 18 can be dispensed with and each winding 28
maintained partially energize(l to a threshold level sufficient to
support the mirror 10 at a desired height above the substrate 14.
In the absence of a permanent magnet such as the maqnet
18, the ferromaqnetic fluid mass 12 must be otherwise retained on the
substrate 14. Referring in detail to Figure 2, an electrornagnetic
assembly 32 here comprises a soft iron toroid 34 havinq four identical
circumferentially equispaced windings 3h. A glass substrate 38 rests
2n on the toroid 34 and ferromagnetic fluid 4n contained within a flexible
pouch 42 rests upon the substrate 38 and itself supports a mirror 44.
The pouch ensures that no loss of performance is incurred through
wetting of the glass substrate 38 by the fluid 40. Moreover, the pouch
42 permits use of a magnetic fluid having a volatile suspension liquid.
Equal energization of the four windings 36 directs the
magnetic fluid 40 into a generally torodial form. As shown in Figure
2, differential variation in the excitation level of the windings 36
l 163299
causes one side of the fluid toroid to increase in height compared with
a diametrically opposed part causing the mirror 44 to tilt. Suitable
control of relative winding excitation permits tilting of the mirror 44
in any direction. Additional windings could be provided at intervals
around the toroid, but it will be appreciated that full angular and
height control is possible using only four coils with suitable control
circuitry.
Figures 3 and 4 illustrate a third embodiment of the
invention which is particularly suitable for displacing an elongate
ln hody such as an optical fiber both in mutually othogonal directions and
angularly about two othogonal axes.
Two axially spaced sets 50, 51 of radially extending
electromagnets 52, are mounted hetween inner and out~er cylinders 54 and
56 respectively, made of non-magnetic material. An optical fiber 58
extends along the axis of the cylinder 54 and is held within a plastic
holder 60. The holder has a generally cylindrical configuration, being
thinner at its center 62 than at its ends. Flexible radial membranes
63 at the ends of cylinder 54 seal against the fiber 58 and confine a
magnetic fluid mass 64 to the interior of the cylinder 54. The holder
2n 60 is immersed within the fluid 64. As in the previous embodiments,
the electroma~nets 52, which are positioned some distance in from the
ends of cylinders 54 and 56, are connected to drive circuits (not
shown) enabling windings 57 of the electromagnets 52 to be selectively
energized.
Axial movement of the fiber 58 is produced by energizing
the windings of one set of electromagnets to a level greater than -the
other set. Thus if the magnetic field produced by the electromagnets
11632g9
5() is greater than that produced by the electromagnets 51, a
correspondin(l pressure differential is established within the magnetic
fluid 64. ~here the magnetic field is higher, the magnetic particles
entrained within the fluid 64 move towards the region of increased
field strength so producinq an increase in pressure on a tapered
surface 6h at the left-hand side of the holder 60 and a decrease in the
pressure at a tapered surface 68 on the right-hand side of the holder 60.
There is a resultant axial force tending to drive the holder 60 and the
fiber held by it towards the left.
Movement of the fiber in a plane perpendicular to its
axis is achieved by increasing the magnetic field strength on one side
of the holder compared to the magnetic field existing on the reverse
side. Again, magnetic fluid particles are drawn towards the region of
increased field strength causing the holder 60 to he displaced in -the
opposite direction. If both axially spaced sets 50, 51 of
electromagnets are equally energized, the holder moves laterally. If
the sets 50 and 52 are differentially energized, the holder and fiber
58 move to orientations in which the fiber axis is angularly inclined
to the longitudinal axis of the cylinder 54.
Other holder configurations and electromagne-t
arrangements can be employed to similar effect. For example, the
holder can be a straight cylinder with increased magnetic fluid
pressure adapted to act against the end of the cylinder in piston-and-
cylinder fashion. The holder/electromagnet assembly of the Figure 3
arrangement provides five degrees of freedom. By altering the
cylindrical form of the holder 60, a sixth degree of freedom can be
provided. The positioning device is then universal in nature in the
J 163299
sense that it can, within limitations set by device geometry, be moved
in any direction and to any orientation.
A movable body can have additional flexural support such
as sprinqs tending to restore the body to a stable position in the
absence of an applied magnetic field.
The arrangements described ahove for obtaining combined
rectilinear and angular movement of an op-tical system elernent have
numerous advantages. Since there is very little friction, extremely
small movements are possible. The devices can be made very small
makinq them particularly adaptable for optical systems used in
fiheroptics, The magnetic fluid provides inherent damping to eliminate
backlash. The fluid has low inertia allowing high speed response, this
being particularly important in dynamic control applications where
displacement varies continuously as in scanning. The positioning
devices, since they use very few mechanical parts, are qenerally more
reliable than known systems for moving optical system elements.
As previously recited, typical applications of these
positioniny devices are in positioning optical Fibers relative to
optical switches, lasers, photo-detectors, and other fibers.
Alternatively the positioning devices can be used to accurately
position the fiberoptic devices relative to a fiber. As well as
fiberoptic applications, the positioning devices may find application
in the control of mirror beam deflectors for laser beam scanning, in
for example optical disc memories. An important application is the use
of movable mirrors directly in a laser resonant cavity for the purposes
of static and controlled tuning of such structures. Generally the
invention is of advantaqe in optical work where a small positioning
~ 1632~
device is required with a span of about 10n rnicrons at a resolution of
less than l/1nOO and a bandwid-th of lKHz.
As previously mentioned, the invention is not limited to
controlling astatic displacement, but is equally applicable to dynamic
displacement control such as in beam scanning. The electromagnetic
nature of the apparatus permits relatively simple feedback control as
illustrated in the attached schematic view of a feedhack circuit in
Figure 5.
Referring in detail to Figure 5, an optical system
element 70 is moved between desired positions by a command si~nal 72
acting throuqh a driver circuit 74, a compensation circuit 7fi, and
winding 78 of an electromagnet. The electromagnet may be energized to
move the optical system element 70 in a manner described in any of the
embodiments of Figures 1 to 4. Presuminq that the optical system
element 70 is a mirror, then as shown, a secondary optical beam is
generated at a source 80 and is reflected back to a quadrant
photodetector 82. The output of the photodetector 82 is taken to a
processing unit 84 where a correction signal 86 is generated and is
summed with the command signal 72 at network 87. A modified command
2n signal is then fed to the driver circuit 74. A feedback circuit will
normally be associated with each of the windinqs so that movement in
any one degree of freedom can he achieved independently of the o-ther
degrees of freedom. The compensation circuits 76, which are associated
with each of the windings 78 are necessary in order to compensate for
differences in performance characteristics of the individual
electromagnets. Although as shown, a secondary radiation beam is used
for position control, it will be understood that part of a primary beam
~ 1632g9
being controlled by -the optical system and being used primarily for
another purpose, may be used to generate a correction signal.
The arrangements used have, for ease of design of the
servo-loop, all been limited to arrangements in which the
electromagnets are energized simultaneously to move the optical element
from one fixed position to another. Clearly, if the element is
intended to be kept in motion then the electromagne-ts can be energized
in such order as will give the element focus and reorientation desired.
Indeed at the expense of a more complex loop, the number of
electromagnets and the complexity of holder shape can be reduced. Thus
by appropriately ordering the firing of three electromagnet pairs
having orthogonally disposed axes, a spherical body can be successfully
moved in any direction and to any orientation.
