TECHNICAL FIELD

The present invention relates in general to miniaturized mechanical drive systems and in particular such miniaturized mechanical drive systems driven by electromechanical actuators.

BACKGROUND

Traditional drive systems for miniaturized mechanical devices typically consist of electromagnetic (EM) motors, gear-heads and screw/ nut converters. The gear-head is used mainly to decrease the rotational speed. At the same time, the torque will be increased. The screw/ nut converters are needed to transform a rotational to a linear movement. This mechanism will also operate as a gear-down unit and allows the device to hold the parts to be moved in the application in a fixed position without over-loading the motor or gear-head. With the proper thread pitch the load on the nut will be held by friction and therefore does not result in any torque in the gearbox. These solutions are however not suitable in all applications.

There are several applications where a linear movement is needed and where the dynamic force during driving as well as the static force during holding should be essentially the same. One example would be to move towards a sensible surface and apply a given force, stop at the surface and guarantee that the static force will not exceed the given maximum force even when the motor is turned off. As can be understood from the discussion further above, with typical thread pitches used in theses applications there is little or no force feed-back to the motor and a screw/ nut solution would therefore not be useful without adding some force controlling device. Hence the force could become very large, when the electromagnetic motor is turned off. A force controlling device would on the other hand increase the size, weight, complexity and cost, which in many cases makes such solution unsuitable in modern applications.

Furthermore, the use of gear-heads and screw/ nut converters also introduces a play in the operation of the drive system. Positioning accuracy is influenced detrimentally. In applications with high positioning accuracy requirements, additional arrangements for ensuring the actual position are needed, e.g. utilizing measurements and feed-back systems, which increase the complexity.

SUMMARY

An object of the present invention is to provide a drive system suitable for miniaturization, having a reliable force control. This object is achieved by drive systems and methods for driving thereof according to the enclosed independent patent claims. Preferred embodiments are defined by the dependent claims. In general words, in a first aspect, a drive system comprises a housing, an ultrasonic motor, a wheel, an object to be moved and a linear bearing arrangement. The ultrasonic motor is attached to the housing. The wheel is journalled with respect to the housing, allowing a rotation of the wheel around an axis. The wheel has a first wheel portion and a second wheel portion being rigidly attached to each other and arranged in axial relationship with respect to each other. The first wheel portion has a larger maximum radius than the second wheel portion. The ultrasonic motor is arranged for allowing application of a first driving force at a surface of the first wheel portion by means of a friction coupling for causing the wheel to rotate. The first driving force is directed in a tangential direction, with respect to the wheel. The object to be moved is arranged in mechanical contact with a peripheral surface of the second wheel portion for allowing a transfer of a second driving force from the second wheel portion to the object to be moved. The linear bearing arrangement is arranged between the object to be moved and the housing to allow a relative linear displacement there between. In a second aspect, a method for operating a drive system comprises provision of voltage signals to an ultrasonic motor attached to a housing of the drive system for moving a drive pad. The drive system further comprises a wheel, an object to be moved and a linear bearing arrangement. A first driving force is applied between the drive pad of the ultrasonic motor at a surface of a first wheel portion of the wheel by means of a friction coupling for causing the wheel to rotate around an axis. The first driving force is directed in a tangential direction, with respect to the wheel. A second driving force is transferred from a peripheral surface of a second wheel portion to the object to be moved, causing a relative linear displacement between the object to be moved and the housing.

One advantage of the present invention is that it allows for designing small and simple drive systems with well controlled maximum dynamic and static forces. Other advantages are further discussed in connection with the detailed description of the different embodiments further below.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention, together with further objects and advantages thereof, may best be understood by making reference to the following description taken together with the accompanying drawings, in which:

FIG. 1 is an illustration of an embodiment of a drive system;

FIG. 2 is a close-up of the drive system shown in Fig. 1;

FIG. 3 is a schematic drawing of a vibrator useful in a drive system;

FIG. 4 is a side view of an embodiment of a drive system;

FIG. 5 is a flow diagram of steps of an embodiment of a method for operating a drive system;

FIG. 6 is a schematic illustration of a linear bearing arrangement;

FIG. 7 is a cross-sectional view of the linear bearing arrangement of

Fig. 6; FIG. 8 is an illustration of an embodiment of a drive system utilizing a gear drive;

FIG. 9 is an illustration of the possibility to integrate position sensors;

FIG. 10 is a schematic illustration of a high density placement of linearly moving objects; and

FIG. 1 1 is a schematic illustration of a high density packing of linear moving objects in a plane.

DETAILED DESCRIPTION

Throughout the drawings, the same reference numbers are used for similar or corresponding elements.

A drive system according to the present invention is based on an ultra- sonic motor (UM). The UM is, as such known, to operate well in many micro mechanical system applications. By operating the UM in the ultrasonic range, the audible noise is reduced considerably. The UM is combined with an integrated gear-head and rotating-to-linear converter based on friction contacts in the same unit. Such a combination gives a total compact solution of the above mentioned problems.

Fig. 1 illustrates a part of an embodiment of a drive system 1. The drive system 1 comprises a housing 20 of which only a part is illustrated. The housing constitutes the backbone, to which the other parts of the drive system are attached. An ultrasonic motor 30 is attached to the housing 20. In the present embodiment, the ultrasonic motor 30 is supported by a flexible printed circuit board 24, which in turn is rigidly fastened by supports 26 to the housing 20. Since the flexible printed circuit board 24 is directed parallel to an intended main motion direction X, the ultrasonic motor 30 is maintained relatively rigidly in this direction, while the ultrasonic motor 30 is more flexible in a direction perpendicular to the flexible printed circuit board 24. A wheel 40 is journalled with respect to the housing 20, thereby allowing a rotation of the wheel 40 around an axis A. The wheel 40 has at least two wheel portions rigidly attached to each other. A first wheel portion 41 is arranged in axial relationship with respect to a second wheel portion 46. In other words, the wheel portions are provided side by side on a common axis and are rotating together with the same rotational speed when the axis is turned. The first wheel portion has a larger maximum radius than the second wheel portion.

The ultrasonic motor 30 is arranged for allowing application of a first driving force at a surface of the first wheel portion 41 by means of a friction coupling. In the present embodiment, the surface of the first wheel portion at which the first driving force is applied is an axially directed surface. In the present embodiment, the ultrasonic motor 30 acts on two opposite axial surfaces 44 of the first wheel portion 41. The first driving force is directed in a tangential direction T with respect to the wheel 40. Such a force will cause the wheel 40 to rotate around the axis A.

In Fig. 2, the parts around the ultrasonic motor 30 are illustrated somewhat enlarged. The first wheel portion 41 and the flexible printed circuit board 24 are drawn with broken lines as if they were transparent, in order to better visualize the relations with the ultrasonic motor 30. In this embodiment, the ultrasonic motor 30 is of a so-called "twin-type", where two vibrators 31 are provided on opposite sides of the first wheel portion 41. In other words, the ultrasonic motor is a twin motor, arranged for application of the first driving force at two opposite axially directed surfaces of the first wheel portion. The vibrators 31 are friction coupled to the surface 41 of the first wheel portion 41 by a respective drive pad 34. The vibrators 31 are caused to vibrate in such a way that the top of the drive pad 34 performs an elliptical motion in a plane of the tangential direction T and perpendicular to the surface 41. In such a way, the first wheel portion 41 is influenced by the first driving force in the tangential direction T, which causes the wheel 40 to rotate. The vibrators 31 are in the present embodiment soldered directly on the flexible printed circuit board 24.

In a presently preferred embodiment, the ultrasonic motor is of a type described e.g. in the published International Patent application WO 2004/001867 Al .

In alternative embodiments, other ultrasonic motor geometries can be utilized. For example, a single vibrator can be used, acting either on one of the axial surfaces of the first wheel portion 41 or on the peripheral surface 43 of the first wheel portion 41.

The drive system 1 of course also comprises an object 10 to be moved. The object 10 to be moved can be of almost any size and shape and only a part of the object 10 to be moved is illustrated in Fig. 1. The object 10 to be moved is arranged in mechanical contact with a peripheral surface 48 of the second wheel portion 46. The mechanical contact takes place via a driving surface 12 of the object 10 to be moved. The mechanical contact thus allows a transfer of a second driving force from the second wheel portion 46 to the object 10 to be moved. In the present embodiment, the object 10 to be moved is arranged for a linear motion, i.e. a translation, but in more general cases, the motion of the object to be moved may also be e.g. rotational or according to any other predetermined path. The type of path is typically determined by the type of bearing. In this embodiment (however not illustrated in Fig. 1) a linear bearing arrangement is arranged between the object 10 to be moved and the housing 20 to allow a relative linear displacement there between. The bearing arrangement will be discussed in more detail further below.

Fig. 1 thus illustrates a drive mechanism of an object 10 to be moved, in this particular embodiment a linearly moving member, using a friction coupled wheel 40 between the object 10 to be moved and the ultrasonic motor 30. The object 10 to be moved is friction coupled with the wheel 40 at a position of the wheel 40 where the radius is reduced, i.e. at the second wheel portion 46. The second wheel portion 46 has in this embodiment a circular cylindrical peripheral surface and the object 10 to be moved is coupled by a friction coupling to the circular cylindrical peripheral surface. Preferably, the circular cylindrical peripheral surface is provided with a friction material 47, having a high and well determined friction coefficient. The wheel 40 with its journal bearing is pressed against this friction coupling with a spring arrangement 22, in this embodiment a mechanical spiral spring. In other words, the object 10 to be moved is pressed against the second wheel portion by means of a spring arrangement 22. The ultrasonic motor 30 has in this embodiment two vibrators 31 held by the flexible printed circuit board 24 and is pressed against the axial sides 44 of the wheel 40 at which the first driving force is applied with another spring arrangement, in this embodiment a motor spring 38.

The drive pads 34 of the vibrator are friction coupled with the axial sides 44 of the wheel 40.

The first driving force is thus applied at a distance from the axis that is larger than the maximum radius of the second wheel portion 46. This makes it possible to design the drive system to be very compact as well as offering a gear-down functionality.

A compact solution is thus to use an ultrasonic motor (UM) with an integrated gear-head and rotating-to-linear converter based on friction in the same unit, as shown in fig. 1. In Fig. 1, a linear UM is driving a wheel 40 close to the periphery by friction coupling, giving a controlled maximum torque. The wheel 40 has also a small radius pin as bearing to reduce size and cost. Due to the radius difference the friction torque from the bearing will be negligible in relation to the torque the UM 30 will generate. In principle also a rotational UM could be used since these motors always use a friction engagement with the part to be driven. It should be realized that UM's 30 typically has a limited speed operation range and to optimize the speed range a gear down unit is often needed. To transform the rotational movement of the wheel 40 in Fig. 1 into a linear movement of the object 10 to be moved, here also referred to as a rod, a second rotational-to-linear conversion is used. The rod is friction engaged with the wheel 40 at another axial position of the wheel where the radius is smaller. This will result in a gear-down of the rod speed in relation to the linear UM speed. The use of one and the same wheel 40 for both gear-down and force controlling coupling makes the device very compact indeed.

Often it is not possible to use direct drive of the rod from the UM's since the speed ranges are too limited. It should, however, also be remembered that the friction coefficient of a UM typically varies with operation and from that point of view the friction coupling in the UM is not the most ideal force controlling mechanism. The use of a completely separate friction coupling of the outgoing rod is therefore preferred in these cases. Note the spring 9 in Fig. 1 that is used to control the normal and hence the friction force of the outgoing rod. This spring 9 can be exchanged to control the maximum dynamic and static force.

It should also be realized that the mass of all of the moving components are important for the resulting forces on the rod. Apart from friction forces there will always be inertial forces, F=ma, where m is mass and a is acceleration, and by reducing the size and density of the moving parts these inertial forces will be minimized. To achieve low forces, low density materials such as reinforced plastic are preferably used in the moving parts.

The mass or weight of a whole driving system is nowadays more and more important. It could be related to minimization of the inertial forces, but could also be related to the portability of the unit. The total mass could make it possible to carry or handle the device longer times without negative effects from an ergonomic point of view. The reduction in mass could also make it possible to add another drive mechanism in e.g. a space application. As part of weight and mass reduction, it is convenient to decrease all parts as can be seen in Fig. 1. In the present embodiment, the ultrasonic motor 30 only consists of two vibrators 31 pressed against the drive wheel 40 with a motor spring 38 at each planar side. The vibrators will be further discussed here below.

The type of vibrators exemplified here above is built on double-bimorph elements, schematically illustrated in Fig. 3. A vibrator 31, in this embodiment a double-bimorph ultrasonic vibrator, is provided with a flexible drive pad 34 at the center. Each half of the beam of the vibrator 31 is built as a bimorph 39 and the two bimorph 39 parts of the beam can be controlled individually. The vibrator 31 is built by electromechanically active material, e.g. piezoelectric ceramics, interleaved between electrodes. A multilayer technique is used so that the electrical voltage can be reduced. Each half of the vibrator beam thus operates as a multilayer bimorph 39. In Fig. 3, the electromechanically active material is illustrated with broken lines as if it would have been transparent. This is however only for increasing the readability of the electrodes in the figure.

The bimorphs 39 are built with alternating electrode layers. In the right part of the figure, every second electrode 33 is connected to a signal phase S I . Likewise, every second of the electrodes 35 of the left bimorph 39 is connected to another signal, signal phase S2. To this en, the drive system further comprises a drive control unit 50, configured for supplying the ultrasonic motor with appropriate signal phases, SI and S2. Every other second electrode 36 or 37 is connected to a constant voltage, e.g. ground or VCC. These voltages are preferably also provided by the drive control unit 50. The electrodes 36, 37 passing through the entire volume are in the present embodiment contacted at the back side as illustrated in the figure. By this connection, one signal phase S 1 controls the bending of one half of the vibrator 31 element and the other signal phase S2 controls the other half of the vibrator 31 element. If soft piezoelectric material is used it is convenient to connect the electrodes 37 passing through the entire volume of one half of the vibrator 31 elements, in the figure the lower part, to high voltage VCC and the electrodes 36 passing through the entire volume of the other half of the vibrator 31 elements, in the figure the upper part, to GND (ground) .

The vibrator 31 element of the present embodiment is preferably designed to have two different moving modes in the point of contact, i.e. at the drive pad 34, making it possible to generate an elliptical moving trajectory at the point of contact with the wheel. The vibration modes can as described above be controlled with electrical signals and by e.g. supplying two phase-shifted drive signals to the vibrators, the direction of motion can be controlled by the phase-shift of the two drive signals. The drive pad 34 in Fig. 3 is a tube and the spring constant of the tube will be one of the design parameters to generate a resonance in the direction normal to the driven component, in this case the wheel. The second flexural resonance of the beam in combination with the tube height as a lever will generate the vibrations in the diving, tangential, direction of the wheel. In other words, the ultrasonic motor of the present embodiment comprises a vibrator 31 with a drive pad 34. where the vibrator 31 presents two different moving modes at the contact point of the drive pad 34. The two different moving modes together generate an elliptical moving trajectory. It is of course possible to use other types of vibrators that can generate movements in tangential and normal/ axial directions of the wheel as well.

The great advantage with the described vibrators is that the vibrator 31 element can be oriented along the tangential direction of the wheel and the space needed for the vibrator 31 is little. In Fig. 4, an embodiment of the support structure of the ultrasonic motor 30 is illustrated. The central position of the drive pad 34 on the vibrator 31 element further minimizes the space demands. To minimize the size and mass the vibrators 31 are in the present embodiment soldered directly on a flexible printed circuit board (FPC) 24 to be mechanically supported by the flexible printed circuit board 24. To avoid damping and undesired forces there are no other mechanical components keeping the vibrators at the proper tangential position of the wheel. The FPC 24 is held in position relative the housing 20 by two supports 26. The FPC supports 24 for holding the motor in the tangential and radial direction with respect to the wheel 40 are placed at a fairly large distance from the vibrator 31 of the UM 30 to minimize torques and forces from the FPC 24 to the vibrator 31.

The vibrators 31 are kept at the proper axial position of the wheel 40 by the motor spring 38, i.e. the vibrators 31 are moving with the wheel 40 in the axial position, perpendicular to the plane of Fig. 4. The relatively long FPC 24 by which the motor is supported makes it possible to have low tolerance demands on the wheel 40, bearing, housing 20 etc. Despite the bending flexibility of the FPC 24, the provided motor support still makes the tangential stiffness sufficiently high for the positioning demands in most applications.

Fig. 5 is a flow diagram of steps of an embodiment of a method for operating a drive system. The method begins in step 200. The drive system is a drive system comprising a housing, an ultrasonic motor, attached to the housing, a wheel, an object to be moved and a linear bearing arrangement. In step 210, voltage signals are provided to the ultrasonic motor for moving a drive pad. A first driving force is in step 212 applied between the drive pad of the ultrasonic motor at a surface of a first wheel portion of the wheel by means of a friction coupling. This is performed for causing the wheel to rotate around an axis. The first driving force is directed in a tangential direction with respect to the wheel. In step 214, a second driving force is transferred from a peripheral surface of a second wheel portion to the object to be moved. This causes a relative linear displacement between the object to be moved and the housing. The procedure ends in step 299.

The noise or sound of the drive system is often an important issue. To reduce the noise the use of a motor in the inaudible ultrasonic frequency range is preferably used here. The motor will hence make no audible noise as long as the motor is operating at fixed frequency and constant high speed. The gear down mechanism that has been described with help of Figs. 1-3 makes it possible to keep the audible noise low also at a fairly slow speed. It should be remembered that UM's typically has a speed at no load of several dm/s due to natural vibration speed of the piezoelectric materials used. When speeds in the range of mm/s are to be used, some means to reduce speed have to be used. Theoretically it is possible to reduce the speed of a vibrator by reduction of the signal amplitude or by signal pulse width modulation (PWM), which reduces the input power. However, the motor operation will typically be unstable in practice since various uncontrolled phenomena such as wear and wear debris will become more important at small vibration amplitudes.

A preferred embodiment, for use in applications demanding low speed, is one of or a combination of phase-shift control and "pulse train" control. Thus, in the method for operating a drive system described above, there will be an additional step of controlling the voltage signals that are provided to the ultrasonic by means of phase- shifted signals or by pulse train control. The phase-shift method is related to the previous mentioned method for changing the driving direction of a vibrator. When there are two more or less orthogonal movement modes of the vibrator part in contact with the driven component and when these modes can be controlled by different signals, the phase shift between these signals can be used to control the direction of the elliptical trajectory, or direction of a linearly moving contact point. When changing the direction, the phase shift is typically changed from e.g. +90° to -90° or vice versa. If instead the phase shift is changed to some intermediate value, the vibrating amplitude in the direction against the wheel surface, i.e. normal to the wheel, can be kept constant. At the same time, the vibration amplitude in the tangential driving direction of the wheel can be varied smoothly from a positive maximum amplitude to a negative maximum amplitude. In such a way, the frequency of the operation can be kept within the ultra- sonic range and the amplitude normal to the wheel surface can be kept constant, and still the driving speed can be changed. The pulse train control method is based on full amplitude signal generation, but sending only a number of signals, the "train", with regular intervals. For instance if square wave signals with the amplitude 5 V at 100 kHz are used, then it is possible to reduce the speed by sending a repetition of 4 square waves with full width with the duration 10 μβ each. This means four pulses with 5 μβ at 5 V and 5 ]is at 0 V. Then, there is a break for 20 \is before the sequence is repeated. This would be considered a pulse train of 4 out of 6 cycles. The pulse train technique is somewhat similar to the amplitude control and PWM methods but it has the great advantage that there will always be a movement also at low power input since there will always be a sufficient number of pulses to overcome e.g. wear debris.

A further advantage with both these approaches is that one and the same frequency is used reducing any noise generating alternations in the driving. Thus, in a preferred embodiments, voltage signals are controlled by phase- shifted signals or by pulse train control to cause a reduced output speed of the ultrasonic motor, when requested, i.e. an electronically controlled "gear- down". In order to achieve such a control, the drive control unit of the drive system is configured for controlling motor voltages to the vibrator by means of phase-shifted signals or by pulse train control. In particular, in a preferred embodiment, the drive control unit is further configured for causing a reduced output speed of the ultrasonic motor by the phase- shifted signals or by the pulse train control, when requested.

In a device with restricted size and mass, the demands on the guiding system often become one of the critical issues. In Fig. 1 the linearly moving rod has typically to be guided with a minimum of friction losses without an addition of mechanical details. One embodiment for solving this in an elegant way can be seen in detail in Figs. 6 and 7. A linear bearing arrangement 60 is provided, arranged between the object 10 to be moved and the housing 20 to allow a relative linear displacement there between. The housing 20 is removed from Fig. 6 to facilitate the view of the object 10 to be moved. In Fig. 7, a cross- sectional view is illustrated, showing the interaction with the housing. In the present embodiment, the linear bearing arrangement 60 is based on a primary guiding of the longitudinal movement is accomplished with a first rod 61 , in this embodiment a first steel pin, and sliding bearings 63. In a preferred embodiment, where the object 10 to be moved is manufactured as a plastic body 14, the sliding bearings 63 are simply integrated in the plastic body 14, i.e. integrated plastic journal bearings. The small diameter of the pin or rod 61 in combination with a sufficiently large distance between the sliding bearings 63 avoids self-locking friction effects. In alternative embodiment, the sliding bearings can be provided in the housing instead.

The rotation of the object 10 to be moved around its own axis and radial movements are avoided with forks that are cross-aligned around another steel pin. For this purpose, a second rod 62, in the present embodiment a second steel pin, is provided in engagement of cross-aligned forks. A first fork 65 integrated in the housing 20 and a second fork 64 in the object 10 to be moved, with adequate tolerances. One fork is thus integrated in the housing 20, the other with the object 10 to be moved and the rod 62 is free to move within the fork openings. The fork solution with sufficient tolerances does not impose any additional friction during sliding and the cross- alignment restricts the movements efficiently without any need for tight tolerances. In other words, the linear bearing arrangement comprises two rods 61, 62 directed parallel to the main motion direction. A first rod 61 of said two rods is held in journal bearings 63 in the object 10 to be moved or the housing 20. A second rod 62 of the two rods is held in a first fork member 64 of the objectlO to be moved or the housing 20 and a second fork member 65 in the other one of the object 10 to be moved or the housing 20. The first fork member 64 and the second fork member 65 are cross-aligned.

In Figs. 6 and 7, note also the wheel 40 shape. Here the first wheel portion 41 with its large radius and the second wheel portion 46 with its smaller radius are easily seen. The driving surface 12 of the object 10 to be moved interacts with the friction material 47 of the second wheel portion 46, while the UM (not shown) interacts with the first wheel portion 41 at a position at a large distance from the axis.

There are several applications with strict size and weight demands where high precision positioning is to be combined with a locked drive rod or object to be moved, when the power is turned off. At first it might appear to be possible to solve this with the traditional electromagnetic motor, gear head and screw/ nut drive train, but these solutions suffer from inherent backlash when drive direction is changed. Even the best "zero-play", back-lash free gear head solutions will suffer from a difficult hysteresis making high precision movements very unreliable. Friction-based UM's do not have this problem since there is a direct drive on the rotor, rod or object to be moved by the friction based drive mechanism.

Main technical challenges in many applications are to maintain precision and stiffness even when the speed has to be reduced and the force has to be increased. The solution presented in Fig. 8 addresses these challenges. To reduce the speed and increase the force, a particular gear solution has been developed. The UM 30 is in the particular embodiment driving on to the axially directed surfaces 44 of a wheel 40, in this embodiment made of reinforced plastics. In this embodiment, the second wheel portion 46 is a gear wheel 49. The object 10 to be moved, in this embodiment a linearly moving rod, has a driving block 16 provided with corresponding gear rod 18 on one of the rod longitudinal sides. The gear rod 18 is provided in mechanical interaction with the gear wheel 49. In analogy with the EM motor and a gear head, there will indeed be a play or hysteresis associated with this gear-down mechanism. The UM 30 does, however, not need more than one gear-down step and this minimizes the uncontrolled movements to a minimum. To avoid play in the mechanism the gear wheel 49 is in this embodiment pressed against the gear rod 18 with a spring at each side, as can be seen in Fig. 8. The springs, being part of the spring arrangement 22, can be adjusted to optimize the precision in relation to the necessary forces in the application.

The guiding of the moving components in the embodiment given in Fig. 8 could be solved in analogy with the motion system in Figs. 6 and 7. Alternatively, in some cases a simplified ball/ roller bearing type of solution can also be used. A linear bearing typically consist of a track rail, slider and balls/ rollers with spacers. The track rail and sliders have longitudinally cut recession, e.g. v-shaped grooves, for the balls or rollers and when properly assembled these recessions will be the guiding surfaces. Since it seldom is space for a complete linear bearing, the motor components can instead have integrated recessions to act as the track rail and slider. These recessions and balls/ rollers will work as a guiding system and typically the motor housing will act as the track rail and the moving components as the slider. Thus, in such a solution, the linear bearing arrangement comprises grooves in the object to be moved and in the housing, and balls or rollers comprised in the grooves.

Another space saving solution, schematically illustrated in Fig. 8, is to guide a part of the object to be moved with an external or internal tube or rod. In Fig. 8, two tubes 66 are attached to the object 10 to be moved. Within each of these tubes, a rod 67 is provided. These rods 67 are attached (not shown) to the housing, directly or indirectly. In alternative solutions, rods or tubes could be attached to the object to be moved, while concentric tubes are provided attached to the housing. These solutions will resemble a telescope mechanism since the movable rod or tube will move axially away from the guiding stator or housing tube or rod in the longitudinal direction. Thus, the linear bearing arrangement 60 comprises in such a case of concentric tubes or a rod in a concentric tube.

Often, the application demands control of the absolute position and this is normally not possible with conventional UM's. The sensors that could fit into miniaturized linear drive systems are typically incremental and often two phase-shifted signals are received making it possible to tell in which direction the rod is moving. The numbers of steps determine the traveled distance and the "zero" position is obtained by an initial calibration where the object is moved to a fixed support at one extreme position. Using quadrature, i.e. dividing each step into the four quadrants that can easily be identified considering the signs of the two signals that can occur (+/+, +/-, - /- and -/+), the resolution will be ¼ of the step size. Using Hall sensors, the step size will be equal to the length of the magnetized regions. The resolution can be improved if the output signal is smoothly varying, e.g. sinusoidal signals, and by converting the signal amplitude to distance, so called interpolation, a higher resolution than the quadrature can be achieved.

Still, there are some applications where the object to be moved is not allowed to move to the "zero" position and in these cases an absolute encoder is needed. Austramicrosystems has a linear Hall sensor with ten bit digital output resolution and the chip scale package is only 1.46 x .1 mm. This sensor chip can measure the absolute position of a magnet with only two poles and for movements in the mm range the resolution will be in the pm range. In Fig. 9, it is illustrated how the sensor 70 and magnet 72 can be assembled into the housing (not shown) and into the driven components, respectively, without any noticeable size increase.

There are some applications with need for several linearly moving axes with a very narrow pitch. Typical EM's have their output axis in the center of a housing and the pitch will hence be equal to or larger than the diameter of motor, gear-head, screw and/ or nut. With ultrasonic motors and by instead position the axis or other object to be moved immediately inside a housing outer wall, advantages can be achieved. The object to be moved is furthermore configured to be moved essentially parallel to the housing outer wall. This makes it possible to place the axes of two different motors very close to each other. Even better is to use the housing corners. Using the motor design given in Fig. 10 the local pitch can be highly improved. Here, the object 10 to be moved is positioned immediately inside a housing outer wall corner 28. A high density placement of objects, in this embodiment linearly moving rods, can be achieved. The placement of the outgoing axis in a corner of the motor housing makes it possible to place two motors very close. An angular corner makes it possible to get high density placement also along a line as can be understood from Fig. 10.

Furthermore, if the motors are placed with their corners 28 towards each other, as illustrated in a top view in Fig. 1 1, it is also possible to get a local high density of linear axis within the plane. The corner 28 could be made sharper and it is therefore possible to further increase the density of linear axes. A 60 degree corner will e.g. allow 6 closely packed motors etc.

The embodiments described above are to be understood as a few illustrative examples of the present invention. It will be understood by those skilled in the art that various modifications, combinations and changes may be made to the embodiments without departing from the scope of the present invention. In particular, different part solutions in the different embodiments can be combined in other configurations, where technically possible. The scope of the present invention is, however, defined by the appended claims.