The present techniques generally relate to actuator assemblies and methods of manufacturing actuator assemblies. They are particularly, but not exclusively, concerned with manufacture of shape memory alloy (SMA) actuator assemblies.

SMA actuators are known for use in handheld electronic devices, such as cameras and mobile phones. Such actuators may be used for example in miniature cameras to effect focus, zoom or optical image stabilization (OIS). By way of example, W02007/113478 discloses an SMA actuator arrangement for a camera providing autofocus using a single SMA wire and WO2013/175197 discloses a compact SMA actuator arrangement for a camera providing OIS using four SMA wires. Further, WO2011/104518 discloses an SMA actuator arrangement comprising eight SMA wires capable of effecting both autofocus and OIS. In each of these disclosures, each SMA wire is fixed at its ends to a static part and a moving part, and the preferred method of fixing is crimping in which a crimp portion is crimped around the SMA wire to form a crimp holding the SMA wire.

WO-2016/189314 discloses a method of manufacture of an SMA actuator assembly by first making a sub-assembly in the form of a strut element comprising a sacrificial strut body and crimp portions crimped around the SMA wire under application of a controlled tension, so that the crimp portions hold the SMA wire therebetween. The crimp portions are attached to the static part and the moving part, respectively. Then, the sacrificial strut body is removed, leaving the crimp portions attached to the static part and the moving part, respectively. WO- 2016/189314 teaches that this provides advantages including the provision of tight control of the length of the SMA wire in the manufactured SMA actuator assembly, because the sacrificial strut body of the strut element holds the relative locations of the crimp portions and thereby maintains the length of the SMA wire therebetween.

An object of the present techniques is to provide new methods of assembling actuator assemblies.

A further object of the present techniques is to provide methods of assembling actuator assemblies which allow continuous manufacture of actuator assemblies. A further object of the present techniques is to provide automated methods of assembling actuator assemblies with actuator wires that cross.

Approaches of the present techniques aim to provide an actuator assembly and methods of controlling actuator assemblies which satisfies one or more of the above objects.

At their broadest, approaches of the present techniques provide methods of manufacturing SMA actuators and apparatuses for manufacturing SMA actuators which can manufacture actuators in which SMA actuator wires cross in an actuator.

A first approach of the present techniques provides a method of manufacturing SMA actuators, each actuator having two SMA actuator wires, each actuator wire being connected between a respective first connector at a first end of the actuator and a respective second connector at a second end of the actuator such that the actuator wires cross in the actuator, the method including the steps of: supplying a strip which has a plurality of actuator blanks formed therein in a continuous pattern, each actuator blank comprising the connectors of one of said actuators; and repeatedly, for each actuator blank in the strip, positioning the actuator wires in relation to the connectors and joining the actuator wires to the connectors. For adjacent pairs of actuator blanks in the strip (preferably all such pairs), the actuator wires may be supplied (or fed) via one actuator blank to the next actuator blank. In other words, the wires may be connected using a continuous wiring method. Accordingly, at at least one step of the method, there is actuator wire passing between an adjacent pair of actuator blanks. More specifically, there is actuator wire passing between connectors at adjacent ends of the pair of actuator blanks.

Supplying the actuator wires to the actuator blank may comprise drawing the actuator wires from a feed mechanism by movement of the strip relative to the feed mechanism and due to the actuator wires being connected to a previous actuator. In some examples, the

aforementioned actuator wire passing between the adjacent pair of actuator blanks may be cut once the actuator wire has been joined to the connectors of the next actuator blank as the wire feed will be maintained by the connectors at one end of the next actuator blank.

SMA actuators which have crossed actuator wires can make the greatest use of the available space in the actuator to improve the range of motion controllable by the actuator. However, the crossed actuator wires can cause difficulties in continuous manufacturing techniques because of their inter-relation. In particular, if the wires are laid consecutively, the second wire will need to be laid in a manner that avoids the already laid first wire, as well as potentially to navigate features of the strip and/or actuator blanks. The present techniques set out a number of ways in which these problems may be overcome.

In many arrangements of the present techniques, the connectors are crimps and the step of joining the actuator wires to the connectors includes closing a crimp so that an actuator wire is gripped between opposing parts of the crimp. It is generally important that the actuator wire is accurately positioned in the crimp before the crimp is closed so that the length of wire between crimps is accurately known and so the performance of the actuator can be accurately and reliably known and consistent between actuators.

However, to form effective crimps, the angle between the opposing parts of the crimp is generally acute as this allows the crimp to be closed in a single press operation. However laying a wire into a crimp with two opposing parts at an acute angle to each other can be difficult. Some of the arrangements of the present techniques seek to overcome this by providing ways for the actuator wires to be moved into (and retained in) the connectors prior to the closing of the connectors.

Positioning the actuator wires in relation to the connectors may comprise laying the actuator wires between the first and second connectors in a direction that is at least substantially along the length of the strip. In other words, the direction along which the actuator wires are laid is at an acute angle (an angle of less than 45 degrees) to a direction along the length of the strip (e.g. the direction of advancement of the strip between actuator blanks). In some examples, the method may include this feature in place of the aforementioned continuous wiring feature.

In some arrangements the step of positioning the actuator wires includes laying an actuator wire approximately along the direction between the first and second connectors (which will generally have an extent at least in the direction of advancement of the strip between actuators, as well as, optionally, an extent perpendicular to that direction to provide for the crossing of the wire from one side of the actuator to the other) and moving the actuator wire laterally so that it engages with the connectors. This can simplify the mechanism which supplies the actuator wire, and can allow the same (or similar) mechanism to lay both actuator wires. It may also allow both actuator wires to be laid at the same time (and/or by a single mechanism) and then moved into position for joining. Alternatively, it may allow a single actuator wire to be laid down the centre of the actuator before either end of the actuator wire is moved into position for joining.

The step of moving the actuator wire may include lifting the actuator wire over at least one protrusion from the strip. The strip will typically have a number of protrusions (including, but not limited to, the connectors themselves) which have already been formed, for example by pressing out elements from the strip. A wire which is being moved laterally may therefore need to navigate such protrusions in order to arrive at its final location. This may be easier than, and therefore preferable to, guiding the wire around such protrusions as it is being supplied.

Once an actuator wire has been moved laterally, it may be held in position. This may be done by the mechanism which has moved the wire, or may be done by other means For example, one or more protrusions from the strip may be formed which hold the actuator wire in position. The mechanism may position the wire past these protrusions and then release it, such that the wire passes around the protrusions which maintain its position relative to the connectors. These protrusions may be formed prior to the wire being provided, in which case it may be necessary to lift the wire over the protrusions (or otherwise guide it around the protrusions), or they may be formed after the wire has been moved laterally past the position of the protrusion.

The method may further include the step of, after j oining the actuator wires to the connectors of an actuator, forming one or more protrusions from the strip which link the first connectors or the second connectors. Such protrusions may comprise one or more sacrificial tabs which are used to hold the connectors of the actuator in a fixed position relative to each other until the actuator has been installed in an apparatus, at which point the tab(s) may be removed.

Preferably the actuators are manufactured such that each actuator wire in an actuator is slack between the connectors when the wire is at a temperature of 25°C. The present techniques provide for a number of different ways in which this slack may be provided. The presence of two crossing actuator wires in the actuator as it is being manufactured may cause particular difficulties for the introduction of slack as it can be necessary to move one of the actuator wires without affecting the other (which may, for example, already have been securely joined to its respective connectors).

In certain arrangements the step of positioning the actuator wires includes passing the actuator wire around one or more protrusions which cause the actuator wire between the first connector and the second connector to deviate from the direct path between those connectors. By causing the wire to deviate from the direct path, the length of wire between the connectors is greater than required, which introduces slack in the wire when the protrusions are removed.

The protrusions can form part of the strip itself, in which case they can be removed once the wire has been joined to its respective connectors. This can simplify the process of adding slack to the wires in each actuator and the protrusions can be formed identically for each actuator as part of the processing of the strip (for example at the same or similar points to the formation of the connectors on the strip).

In other arrangements, the protrusions can be provided on a die and the method further includes the steps of, before positioning the actuator wire, introducing the die into the strip so that the protrusions are positioned between the first and second connectors and, after joining the actuator wire, removing the die from the strip. This can provide for more consistency in the amount of slack to be added as the same die (or set of dies) can be used for each actuator. However, the die needs to remain in place until the wires are joined to the connectors, so it may have to move with the strip between different steps of the process.

Other approaches to adding/creating slack in the actuator wires in each actuator include the further step, after the step of positioning the actuator wires and before the step of joining the actuator wires to the connectors, of increasing the amount of actuator wire between each connector. In this manner, whilst approximately the right amount of wire for a taught (no slack) connection of the wire between the connectors can be provided initially, the wire in the actuator can then be drawn out to provide for the desired slack.

In some arrangements, the increasing of the amount of actuator wire includes applying a force on the actuator wire perpendicular to the direction of the laying of the wire. In order to accurately control the amount of slack being added and/or to prevent any drawing out of the actuator wire from causing the wire to move from the connectors (for example to be pulled out of the jaws of crimps), the method may further include preventing movement of the actuator wire in the direction of the applied force in a region proximate to one or more of the connectors.

The force may be applied mechanically (e.g. by contact between a part of the manufacturing apparatus), or may be applied by a non-contact force, such as a vacuum/suction.

Preferably the force is not applied to any other actuator wire already present in the actuator. This allows slack to be introduced into a second wire that is laid in an actuator without affecting a first wire that is already present in the actuator and may, for example, already have had the desired amount of slack provided for and/or may already been joined to its respective connectors, such that additional force on that wire may affect its position or performance or damage it.

A second approach of the present techniques provides an apparatus for manufacturing SMA actuators, each actuator having two SMA actuator wires, each actuator wire being connected between a respective first connector at a first end of the actuator and a respective second connector at a second end of the actuator such that the actuator wires cross in the actuator, the apparatus comprising: a strip feed mechanism arranged to guide an elongate strip through the apparatus, the strip having a plurality of actuator blanks formed therein in a continuous pattern, each actuator blank comprising the connectors of one of said actuators; a wire feed mechanism arranged to supply an SMA actuator wire and position it in relation to the connectors; and a joining mechanism arranged to join the actuator wires to the connectors.

For adjacent pairs of actuator blanks in the strip, the apparatus may be configured to supply the actuator wires via one actuator blank to the next actuator blank.

SMA actuators which have crossed actuator wires can make the greatest use of the available space in the actuator to improve the range of motion controllable by the actuator. However, the crossed actuator wires can cause difficulties in continuous manufacturing techniques because of their inter-relation. In particular, if the wires are laid consecutively, the second wire will need to be laid in a manner that avoids the already laid first wire, as well as potentially to navigate features of the strip and/or actuator blanks. The present techniques set out a number of apparatuses which may overcome these problems.

In many arrangements of the present techniques, the connectors are crimps and joining the actuator wires to the connectors includes closing a crimp so that an actuator wire is gripped between opposing parts of the crimp, for example by using a stamp or press tool. It is generally important that the actuator wire is accurately positioned in the crimp before the crimp is closed so that the length of wire between crimps is accurately known and so the performance of the actuator can be accurately and reliably known and consistent between actuators.

However, to form effective crimps, the angle between the opposing parts of the crimp is generally acute as this allows the crimp to be closed in a single stamping/press operation. However laying a wire into a crimp with two opposing parts at an acute angle to each other can be difficult. Some of the arrangements of the present techniques seek to overcome this by providing ways for the actuator wires to be moved into (and retained in) the connectors prior to the closing of the connectors.

When positioning the actuator wires in relation to the connectors, the wire feed mechanism may be configured to lay the actuator wires between the first and second connectors in a direction that is at least substantially along the length of the strip.

In some arrangements the wire feed mechanism is arranged to lay an actuator wire approximately along the direction between the first and second connectors (which will generally have an extent at least in the direction of advancement of the strip between actuators, as well as, optionally, an extent perpendicular to that direction to provide for the crossing of the wire from one side of the actuator to the other) and the apparatus further comprises an engagement portion which is arranged to engage with an actuator wire and move it so that it engages with the connectors. This arrangement can simplify the wire feed mechanism, and can allow the same (or similar) mechanism to lay both actuator wires. It may also allow both actuator wires to be laid at the same time (and/or by a single mechanism) and then moved into position for joining. Alternatively, it may allow a single actuator wire to be laid down the centre of the actuator before either end of the actuator wire is moved into position for joining. The engagement portion may be arranged to lift the actuator wire over at least one protrusion from the strip. The strip will typically have a number of protrusions (including, but not limited to, the connectors themselves) which have already been formed, for example by pressing out elements from the strip. A wire which is being moved laterally may therefore need to navigate such protrusions in order to arrive at its final location. This may be easier than, and therefore preferable to, configuring the apparatus so that the wire feed mechanism can guide the wire around such protrusions as it supplies the wire.

A number of suitable engagement portions exist. Preferably the engagement portion is able to engage with the actuator wire without knowing its precise location. The engagement portion may therefore be significantly larger than the diameter of the wire. Preferably the engagement portion, once it has engaged the wire, holds the wire in a precisely known location. This can assist in accurately positioning the wire and in precisely controlling the amount of wire between the connectors.

In one arrangement the engagement portion includes a gripper having a pair of opposing arms, the arms collectively defining a holding portion between them when the gripper is in a closed state, wherein the holding portion constrains the movement of a wire contained therein in directions perpendicular to the extent of the wire. The gripper can thus close around the wire and keep it in an accurately-known lateral position.

In one arrangement the engagement portion includes a head portion which is arranged to engage with the actuator wire and a crank mechanism to which the head portion is connected and the crank mechanism is arranged to cause the head portion to rotate as it moves laterally, such that the head portion follows an arcuate path between the two extremes of its motion. Thus the head portion can pick up the wire and lift it (for example over an obstacle or protrusion) whilst also moving it laterally. The head portion may be, for example, a roller or notched pin.

Once an actuator wire has been moved laterally, it may be held in position. This may be done by the engagement portion which has moved the wire, or may be done by other means. For example, one or more protrusions from the strip may be formed which hold the actuator wire in position. The engagement portion may position the wire past these protrusions and then release it, such that the wire passes around the protrusions which maintain its position relative to the connectors. These protrusions may be formed prior to the wire being provided, in which case it may be necessary to lift the wire over the protrusions (or otherwise guide it around the protrusions), or they may be formed after the wire has been moved laterally past the position of the protrusion.

The apparatus may thus further include one or more press tools arranged to form one or more protrusions from the strip after the actuator wire has been moved by the engagement portion, wherein the protrusions are arranged to hold the actuator wire in position after it has been laterally moved.

The apparatus may further include a tool arranged to form one or more protrusions from the strip which link the first connectors or the second connectors. Such protrusions may comprise one or more sacrificial tabs which are used to hold the connectors of the actuator in a fixed position relative to each other until the actuator has been installed in an apparatus, at which point the tab(s) may be removed.

Preferably the apparatus is arranged to manufacture the actuators such that each actuator wire in an actuator is slack between the connectors when the wire is at a temperature of 25°C. The present techniques provide for a number of different ways in which this slack may be provided. The presence of two crossing actuator wires in the actuator as it is being manufactured may cause particular difficulties for the introduction of slack as it can be necessary to move one of the actuator wires without affecting the other (which may, for example, already have been securely joined to its respective connectors).

In certain arrangements the wire feed mechanism is arranged to position the actuator wires around one or more protrusions such that an actuator wire between the first connector and the second connector deviates from the direct path between those connectors. By causing the wire to deviate from the direct path, the length of wire between the connectors is greater than required, which introduces slack in the wire when the protrusions are removed.

The protrusions can form part of the strip itself, in which case the apparatus may further include a tool arranged to form said protrusions on the strip prior to the strip reaching the wire feed mechanism and a tool arranged to remove said protrusions after the strip has passed the joining mechanism. This can simplify the process of adding slack to the wires in each actuator and the protrusions can be formed identically for each actuator as part of the processing of the strip (for example at the same or similar points to the formation of the connectors on the strip).

In other arrangements the apparatus may further include a die on which said protrusions are formed, wherein the apparatus is arranged to introduce the die into the strip so that the protrusions are positioned between the first and second connectors prior to the strip reaching the wire feed mechanism and remove the die from the strip after the strip has passed the joining mechanism. This can provide for more consistency in the amount of slack to be added as the same die (or set of dies) can be used for each actuator. However, the die needs to remain in place until the wires are joined to the connectors, so it may have to move with the strip between different steps of the process.

In other approaches for adding/creating slack in the actuator wires, the apparatus may further include a wire extending mechanism arranged between the wire feed mechanism and the joining mechanism, the wire extending mechanism being arranged to increase the amount of actuator wire between each connector in an actuator. In this manner, whilst approximately the right amount of wire for a taught (no slack) connection of the wire between the connectors can be provided initially, the wire in the actuator can then be drawn out to provide for the desired slack.

The wire extending mechanism may be arranged to apply a force on the actuator wire perpendicular to the direction of the laying of the wire.

The wire extending mechanism may include a arm which is arranged to engage with the actuator wire in a region proximate to one of the connectors to prevent movement of the actuator wire in the direction of the applied force in that region. This can allow the wire extending mechanism to accurately control the amount of slack being added and/or to prevent any drawing out of the actuator wire from causing the wire to move from the connectors (for example to be pulled out of the jaws of crimps), The wire extending mechanism may apply the force mechanically (e.g. by contact between a part of the manufacturing apparatus), or may be applied by a non-contact force, such as a vacuum/suction.

Preferably the wire extending mechanism is arranged to apply the force to only a single actuator wire, even where multiple actuator wires are present in the actuator. This allows slack to be introduced into a second wire that is laid in an actuator without affecting a first wire that is already present in the actuator and may, for example, already have had the desired amount of slack provided for and/or may already been joined to its respective connectors, such that additional force on that wire may affect its position or performance or damage it.

In certain arrangements the wire extending mechanism includes two contact elements which are arranged to engage with opposite sides of the actuator wire at separate positions along the wire, and is arranged to apply the force to the actuator wire by moving the contact elements in opposite directions. For example, the contact elements may be pins or rollers which extend from a rotatable head, such that rotation of the head causes the pins or rollers to exert a lateral force on the actuator wire. Such contact elements can be introduced into the actuator between the actuator wires (if two wires are present) and thus only act on one of the wires.

In certain arrangements the wire extending mechanism includes a plunger which has a distal surface, the mechanism being arranged to bring the plunger into contact with the actuator wire in a direction perpendicular to the plane of the strip. The distal surface of the plunger may have a groove which is arranged to accommodate another actuator wire which is present in the actuator such that the plunger does not apply a force on the another actuator wire.

The actuators of the present techniques may in general be used in any type of device that comprises a static part and a moveable part which is moveable with respect to the static part. By way of non-limitative example, the actuator assembly may be, or may be provided in, any one of the following devices: a smartphone, a camera, a foldable smartphone, a foldable smartphone camera, a foldable consumer electronics device, an image capture device, a foldable image capture device, an array camera, a 3D sensing device or system, a servomotor, a consumer electronic device (including domestic appliances), a mobile or portable computing device, a mobile or portable electronic device, a laptop, a tablet computing device, an e-reader (also known as an e-book reader or e-book device), a computing accessory or computing peripheral device (e.g. mouse, keyboard, headphones, earphones, earbuds, etc.), a security system, a gaming system, a gaming accessory (e.g. controller, headset, a wearable controller, etc.), a robot or robotics device, a medical device (e.g. an endoscope), an augmented reality system, an augmented reality device, a virtual reality system, a virtual reality device, a wearable device (e.g. a watch, a smartwatch, a fitness tracker, etc.), a drone (aerial, water, underwater, etc.), an aircraft, a spacecraft, a submersible vessel, a vehicle, and an autonomous vehicle. It will be understood that this is a non-exhaustive list of example devices.

Actuator assemblies as described herein may be used in devices/sy stems suitable for image capture, 3D sensing, depth mapping, aerial surveying, terrestrial surveying, surveying in or from space, hydrographic surveying, underwater surveying, scene detection, collision warning, security, facial recognition, augmented and/or virtual reality, advanced driver- assistance systems in vehicles, autonomous vehicles, gaming, gesture control/recognition, robotic devices, robotic device control, touchless technology, home automation, medical devices, and haptics.

Embodiments of the present techniques will now be described by way of example with reference to the accompanying drawings in which:

Fig. 1 is a schematic side view of an SMA actuation arrangement;

Fig. 2 is a perspective view of a configuration of the SMA actuation arrangement;

Fig. 3 illustrates an SMA actuator apparatus when assembled;

Fig. 4 shows a continuous strip of SMA actuator blanks for use in manufacturing methods and apparatuses of the present techniques;

Fig. 5 shows an arrangement for continuous manufacturing of SMA actuators according to an example of the present techniques;

Fig. 6 shows a further arrangement for continuous manufacturing of SMA actuators according to an example of the present techniques; Fig. 7 shows a further arrangement for continuous manufacturing of SMA actuators according to an example of the present techniques;

Fig. 8 shows a further arrangement for continuous manufacturing of SMA actuators according to an example of the present techniques;

Figs. 9 and 10 show further alternative arrangements for the approach shown in Fig. 8;

Fig. 11 shows a further arrangement for continuous manufacturing of SMA actuators according to an example of the present techniques;

Fig. 12 shows a further arrangement for continuous manufacturing of SMA actuators according to an example of the present techniques;

Figs. 13 and 14 show arrangements for adding slack to SMA actuator wires according to examples of the present techniques;

Fig. 15 shows a further arrangement for adding slack to SMA actuator wires according to an example of the present techniques;

Figs. 16 to 18 show further arrangements for adding slack to SMA actuator wires according to examples of the present techniques;

Fig. 19 shows a further arrangement for adding slack to SMA actuator wires according to an example of the present techniques;

Fig. 20 shows a further arrangement for continuous manufacturing of SMA actuators according to an example of the present techniques;

Fig. 21 shows an alternative arrangement for continuous manufacturing of SMA actuators according to an example of the present techniques;

Fig. 22 shows a potential configuration of the apparatus of Fig. 20; Fig. 23 shows an arrangement which may be used in an apparatus or method according to the present techniques; and

Fig. 24 shows a gripper which may be used in an apparatus or method according to the present techniques.

Fig. 1 illustrates an SMA actuation arrangement 1 for a camera. The SMA actuation arrangement 1 comprises a support structure 30 having an image sensor 20 mounted thereon. A camera lens element 10 is suspended on the support structure 30 and is arranged to focus an image on the image sensor 20. The camera lens element comprises one or more lenses 11, a single lens being illustrated in Fig. 1 for clarity. The camera is a miniature camera in which the or each lens 11 has a diameter of no more than 20mm.

The SMA actuation arrangement 1 is for a miniature camera in which the camera lens element 10 is the movable element. However, it will be appreciated that this is only one example of the use of such an arrangement and that many other uses exist where it is desired to control the movement of a movable element relative to a stationary support structure.

Plural SMA actuator wires 50 are connected in tension between the support structure 30 and the camera lens element 10. The camera lens element 10 may be suspended on the support structure 30 exclusively by the SMA actuator wires 50. Alternatively, the camera lens element 10 may be suspended on the support structure 30 by a suspension system (not shown) that may have any suitable form for allowing movement of the camera lens element 10 with respect to the support structure 30 with the desired degrees of freedom, for example formed by flexures to allow movement in three dimensions, or formed by ball bearings or sliding bearings to allow movement in two dimensions while constraining movement in a third dimension, or constraining movement within a particular range of movement.

The SMA actuator wires 50 are in an arrangement in which the SMA actuator wires 50 are capable of driving movement of the camera lens element 10 with respect to the support structure 30 with plural degrees of freedom on selective contraction of the SMA actuator wires 50. The SMA actuator wires 50 may be configured to drive such movement as shown in Fig. 2 or Fig. 3 which show first and second configurations of the SMA actuation arrangement 1, or in general may have other configurations.

Fig. 2 illustrates schematically a configuration for an SMA actuator arrangement 1 in which eight SMA actuator wires 50 are provided. The SMA actuation arrangement 1 may have a construction as described in further detail in any of WO-2011/104518, WO-2012/066285 or WO-2014/076463, to which reference is made, and the contents of which are hereby incorporated in their entirety. However, an overview of the arrangement of SMA actuator wires 50 is as follows. For ease of reference, the z axis is taken to be the optical axis of the movable element 10 (in this arrangement a camera lens element) and the x and y axes are perpendicular thereto. In the case where the movable element 10 is a camera lens element, in desired orientation of the camera lens element 10, the optical axis of the camera lens element 10 is perpendicular to the image sensor 20 and the x and y axes are lateral to the image sensor 20.

Two SMA actuator wires 50 are provided on each of four sides of the movable element 10 in a 2-fold rotationally symmetric arrangement. Each SMA actuator wire 50 extends perpendicular to a line radial of the optical axis of the camera lens element 10, that is substantially perpendicular to the x axis or to the y axis. However, the SMA actuator wires 50 are inclined with respect the optical axis of the camera lens element 10, so that they each provide a component of force along the z axis and a component of force primarily along the x axis or primarily along the y axis.

Each SMA actuator wire 50 is connected at one end to the support structure 30 and at the other end to the camera lens element 10, selected so that in combination with the inclination of the SMA actuator wires 50, different SMA actuator wires 50 provide components of force in different directions along the z axis and different directions along the x axis or along the y axis. In particular, the pair of SMA actuator wires 50 on any given side of the camera lens element 10 are connected to provide components of force in opposite directions along the z axis, but in the same direction along the x axis or along the y axis. The two pairs of SMA actuator wires 50 on opposite sides of the camera lens element 10 are connected to provide components of force in opposite directions along the x axis or along the y axis.

Thus, the SMA actuator wires 50 are capable, on selective contraction, of driving movement of the movable element 10 with respect to the support structure 30 in translational movement with three degrees of freedom (i.e. along the x, y and z axes) and also rotational movement with three degrees of freedom (i.e. around the x, y and z axes). Due to the symmetrical arrangement, movement with each of the degrees of freedom is driven by contraction of different combinations of SMA actuator wires 50. As the movements add linearly, movement to any translational and/or rotational position within the six degrees of freedom is driven by a linear combination of contractions of the SMA actuator wires 50. Thus, the translational and rotational position of the movable element 10 is controlled by controlling the drive signals applied to each SMA actuator wire 50.

In the case where the movable element 10 is a camera lens element, translational movement along the optical axis of the camera lens element 10 (i.e. along the z axis) may be used to change the focus of an image formed by the camera lens element 10 and translational movement laterally of the optical axis of the camera lens element 10 (i.e. along the x and y axes) may be used to provide OIS.

Fig. 3 illustrates an SMA actuator apparatus 1 such as that described above with reference to Fig. 2 in a less schematic manner and showing more of the structural details of the apparatus when assembled. Like features have been given the same reference numerals in Fig. 3 as Fig. 2

It can be seen from the actuator apparatus 1 of Fig. 3 that the SMA actuator wires 50 on each side of the actuator cross in a pair to connect to four separate crimps 51 on each side. The pair of actuator wires 50 and crimps on one side of the actuator apparatus 1 that form a separable component of the actuator apparatus will be referred to herein as“an actuator” 55. In the arrangement shown in Fig. 3 two of the crimps 51 at one end of the actuator 55 are merged to create a common terminal 52, but this is not essential.

It is desirable to manufacture the actuators 55 separately from the movable element 10 and support structure 30 and subsequently assemble the actuator assembly 1. In particular this can allow the actuators 55 to be manufactured as a continuous strip, before being separated and assembled, which allows for increased manufacturing speed and productivity. Techniques for continuous strip manufacturing of SMA actuators are known. However, none of these address the difficulties associated with actuators in which a pair of SMA actuator wires has to cross, such as the actuators 55 shown in Fig. 3.

Further, it has been found that it is desirable that the SMA actuator wires in an actuator are manufactured such that there is slack in the wires between the crimps. This means that it is not necessary to apply a controlled tension to the SMA actuator wires while closing the crimp portions around the length of SMA wire, with the result that the manufacturing process is simplified, which can allow the speed of manufacturing to be increased and the unit cost of manufacture to be reduced. When reference is made to slack in an SMA actuator wire, it may mean that the wire is slack at a temperature of 25°C which is a typical ambient temperature. When a drive signal is applied to the SMA actuator wire in use in order to cause contraction, the temperature of the wire rises significantly above 25°C.

Generally-speaking, the terms“slack” or“slack wire” may mean a wire which has zero tension. Alternatively the terms“slack” or“slack wire” may mean a wire which has zero tension when the wire is unpowered. This may mean that a wire which has“slack” and which is held between two fixed positions (i.e. is mechanically coupled at two points along its length to one or more other elements, such as crimps or crimp portions) does not lie in a straight line between those positions when the wire is unpowered and at ambient temperature.

Typically when SMA actuators are manufactured, the SMA wire(s) will be pulled from a spool (in which the wire is under tension), and in some cases the wire may retain some tension due to hysteresis. Thus the terms“slack” or“slack wire” may mean that the wire is slack after any residual tension has been removed. The tension may be removed by, for example, stretching the wire at ambient temperature. One the tension is removed, if the length of the wire between two fixed positions is greater than the straight-line distance between those positions, then the wire may be considered to be“slack”.

A further alternative definition of the terms“slack” or“slack wire” is that the wire is not under tension when the SMA wire is substantially martensitic.

It will be understood that any of the above definitions may be used in the present techniques to relate to a wire which is“slack”. Further details of the meaning of“slack” in this context and the potential advantages of having slack in SMA actuators are set out in PCT/GB2019/050072 and GB 1815673.7 which are co-owned applications of the present applicant. The contents of each of these documents is hereby incorporated by reference in their entirety and specifically in relation to the meaning of the term“slack”, the advantages of providing such slack, and the techniques by which such slack may be provided in an SMA actuator.

In all manufacturing techniques for SMA actuators, it is desirable to maintain very tight tolerances in the distance between crimps at either end of the actuator and in the length of the SMA actuator wires, in order to achieve good, and predictable, performance from the actuator assembly. Where, as in Fig. 3, the actuators 55 include a pair of crossed actuator wires 50 ensuring consistency between the pair of wires in an individual actuator is also important.

The arrangements below set out different ways in which a continuous manufacturing process for SMA actuators with crossed actuator wires may operate, as well as techniques for manufacturing SMA actuators in which the (potentially crossed) SMA actuator wires in the final actuator have slack in them. It will be appreciated that one or more of the arrangements set out below may be combined in any particular manufacturing process.

It will be noted from Fig. 3 that the actuators 55 shown in that figure come in“left-hand” and “right-hand” configurations (see, for example, the two actuators visible in Fig. 3 which are mirror images of each other). Whilst the present techniques will be described in relation to actuator pairs of this kind, this mirror image arrangement is not necessary and the techniques can be applied to actuators which are identical or to actuators where all the connections are configured differently. Actuator pairs of this kind can be easier to manufacture on an continuous basis as the wires can be continuously laced from bottom to top of one actuator pair, followed by from top to bottom of the subsequent actuator pair. In such an arrangement, each of the different actuator configurations will always be produced with the wirings moving in the same manner.

Fig. 4 shows an arrangement of mirrored pairs of blanks for actuators 55 forming a continuous strip 54 for manufacture, in which a single actuator wire 50 has been laced back and forth between the top 51a and bottom 5 lb crimps. A second wire (not shown) would be laced on the opposing diagonals to join the other two crimps 51c, 5 Id. Only one of the actuators 55 on the strip 54 has been labelled, but it will be seen that each actuator is identical, save for the mirror image arrangement.

It can be seen from Fig. 4 that the pair of crimps 51a, 51 d in the actuator 55 which do not form a common terminal 52 are joined by a sacrificial tab 56. This tab maintains the crimps 51a, 5 Id in the correct position relative to each other during the manufacturing process until the actuator 55 has been connected to the movable element 10 and the support structure 30, at which point it can be removed (thus arriving at the arrangement shown in Fig. 3, in which the tabs would have been on the comer of the assembly 1 closest to the viewer). From Fig. 3 it will be appreciated that the tabs preferably project out of the plane of the strip 54 in order that the two tabs which are positioned adjacent to each other in the assembly 1 do not interfere. The tab 56 is typically bent out of the plane of the strip when other forming processes are performed on the strip, such as the creation of the crimps 51.

These two wires 50 could be laid into the crimps 51 at the same time. Alternatively a first of the wires 50 may be laced into the crimps and the second wire added later in the process. Depending on the implementation of the process and the tooling, the crimps could be closed to grip the wire(s) at the same time or at different times for each crimp or each wire.

Specifically, a first wire 50 could be laced into a pair of crimps 51 in a particular actuator 55 and both crimps on that wire in that actuator closed. Alternatively, the crimps could be closed sequentially, or in an arrangement in which one crimp of a first actuator is closed at the same time as a crimp of a subsequent actuator. The second wire could be laced using similar approaches, either at the same time or sequentially (for example, the second wire could be laced into an actuator after one or both of the crimps of that actuator have been closed on the first wire).

The design of the actuators 53 requires the two wires 50 to cross each other without touching. This can be achieved by arranging the crimps 51 so that the crimps which receive one of the wires are at a different height to the crimps that receive the other wire. The optimal orientation of the wires 50 in the actuators 53 may also not necessarily be in a flat plane. Therefore the crimps 51 on the strip 54 may be arranged so that they are at different heights and/or in different orientations. One or more of the crimps may be mounted on an anvil which holds the crimp at an angle offset from the wire plane. Fig. 4 also shows a plurality of indexing locations 57 along the strip 54 which are used to accurately locate the strip as it moves through the manufacturing process.

Fig. 5 shows an arrangement for continuous manufacturing of actuators 55 having crossed actuator wires 50. From Fig. 4 it will be appreciated that the raised profde of the tab 56 and its necessary proximity to the crimps 51a, 5 Id can create difficulties for a continuous process of laying the actuator wire 50, particularly on the scales typically associated with SMA actuation assemblies, where an actuator 55 may be a few millimetres in width. Thus the tab 56 can interfere with either the actuator wire 50 itself, or the mechanism for laying the wire (not shown).

The arrangement shown in Fig. 5 addresses this problem by changing the order in which the tab 56 is formed compared to the actuator wires 50 being laid. In the arrangement shown in Fig. 5, the shape of the tab 56 is formed in the strip 54 along with the other components, but the tab 56 is not bent upwards until after the wires 50 have been laid. This may be done in the same action as the closing of the crimps 51, or in a separate step.

It is preferable that the tab 56 is folded up before the crimps 51 are closed to avoid adding further stresses to the separated crimps. Alternatively the crimp frets could be clamped to prevent relative movement under the tab 56 has been folded up. Fig. 5 shows how the actions of folding up the tab 56 and the closing of the crimps 51 can be performed with a 2-stage tool. A bi-directional stamping tool 61a, 61b acts to firstly push the tab 56 up and then, secondly to close the crimps 51.

Fig. 6A shows a further arrangement for continuous manufacturing of actuators 55 having crossed actuator wires 50. In order to route two actuator wires 50 into the continuous strip 54 of etched actuators 55, the two wires are fed from a feed (not shown) and are drawn by the movement of the strip 54 due to the wires 50 being clamped in closed crimps of downstream actuators (not shown). As the strip 54 indexes to the next actuator 55 along, a pair of rollers or pins 62 are actuated and engage with a respective one of the wires 50 between an adjacent pair of actuators 55a, 55b and pull it sideways so that it is positioned into the crimps 51 at adjacent ends of the pair of actuators. The crimps 51 adjacent the rollers 62 are then closed, either at the same time or sequentially whilst the rollers 62 are holding the wires 50 in place. Alternatively, the crimps of the first actuator 55a may all still be open and a second pair of rollers (not shown) may still be engaged with the wires 50 upstream of the first actuator 55a, such that the positioning of the wires 50 across the actuator 55a can be closely controlled. Then all the crimps 51 of the first actuator 55a can be closed at the same time.

Once the crimps 51 have been closed, such that the pull of the rollers 62 is no longer required, the rollers can return to their starting position. Once all crimps 51 in a particular actuator have been closed, the wires 50 can be cut as the wire feed will be maintained by the closed crimps at one end of the latest actuator 55b on strip 54.

Interference between the rollers 62 can be avoided by having the rollers move in different horizontal planes, as shown in the cross-sectional view in Fig. 6B. Different configurations of the rollers 62 can be used to provide for the continuous wiring of the actuators 55 on the strip 54. For example, a single pair of rollers 62 at each station may be used and the rollers may be able to actuate in both directions perpendicular to the direction of advancement of the strip 54. Thus a particular one of the rollers 62 can engage with one wire 50 between a first pair of actuators 55, and with the other wire between the next pair of actuators.

However, as this would involve the rollers being vertically movable (in order to traverse to the opposite side of the wires), it may be preferable to have two pairs of rollers 62 at each station, with each pair configured to pull one of the wires 50 in opposite directions. This would also accommodate the usual relative arrangement of actuator wires 50 in actuators 55 wherein the same wires are always the“top” and“bottom” wires when viewed from the side, for example as in Fig. 6B.

Alternatively to providing multiple pairs of rollers 62 or having the rollers swap sides, the apparatus for manufacturing the actuators may proceed by only laying a single wire at a time, which can avoid the need to account for possible interference between the rollers 62.

The rollers 62 can also be actuated to move perpendicular to the plane of the strip 54 and thereby pull the wires 50 as accurately as possible into the crimps 51. Whilst rollers 62 as illustrated in Fig. 6 will generally provide for lower friction and therefore have less risk of damaging the wires 50, for small-scale devices pins or dowels will be equally effective and the contact sufficiently small for friction to not be an issue and so can be used instead.

In the arrangement shown in Fig. 7, the actuator wire(s) 50 are fed from a feed mechanism 63 which may be, for example, a capillary tube or a reel and pulley. The tip of the feed mechanism 63 adjacent to the actuator strip 54 during wire laying is manoeuvrable in both axes in the plane perpendicular to the direction of advancement or indexation of the strip.

This allows the feed mechanism to lift the wire over obstructing features (such as the sacrificial tab 56 discussed above) as the strip 54 advances. This freedom of motion of the feed mechanism 63 can also allow the feed of the wire 50 to align closer to the desired routing prior to crimping.

Fig. 8 shows a further arrangement for the continuous manufacture of actuators 55 using a strip 54 of connected actuators which is an alternative to the arrangement shown in Fig. 6 and described above.

In the arrangement in Fig. 8, additional tabs 64 are etched into the strip 54 of actuators 55 when the other features (such as the crimps 51 and their connectors) are created. These tabs 64 can be folded up before the strip reaches the wiring stages. At the wiring stages, the wires 50 can be fed around those tabs 64 which act to hold them in the correct orientation vis-a-vis the crimps 51 until the crimps are closed (for example using a press tool 65). Preferably, the tabs 64 are arranged so that the wire 50 is pulled into the jaw of the crimps 51 by the direction between the crimp 51 and the adjacent tab 64.

As a result of using the tabs 64, the feed mechanism 63 does not control the alignment of the wires 50 when the crimps 51 are closed, which can avoid pulling on the feed mechanism 63 (or on the feed of the wire 50 itself). As the tabs 64 hold both ends of a wire passing through crimps 51 of a single actuator 55, the crimps at each end can be closed simultaneously (for example as shown in Fig. 8).

After the wires 50 have been secured in the crimps 51, the tabs 64 can be removed and discarded and the wires cut. It is noted that there are other ways of forming the tabs 64, such as direct forming (without etching), or just by etching, and the manner of manufacture of the tabs 64 described above is not intended to be limiting.

The tabs 64 may be provided with rounded or slanted edges to avoid sharp points of contact with the wire 50.

The use of the tabs 64 avoids the need for a mechanism (such as the rollers/pins 62 described above) which pulls the wires 50 sideways. Such an arrangement may be more practical on smaller devices as it avoids the need to ensure accurate engagement with very fine wires.

The tabs can also avoid any friction between the rollers/pins and the wires and can allow the apparatus to separate (in distance terms) the actions of laying the wires 50 and the closing of the crimps 51 to allow access for the necessary mechanisms and avoid interference between parts of the assembly apparatus which are performing different tasks.

Figs. 9 and 10 show further alternative arrangements for the approach described and illustrated in relation to Fig. 8 above. The arrangements in Figs. 9 and 10 show alternative approaches for arranging for the wire 50 to pass around the tabs 64, which may allow for a simpler feed mechanism 63 to be used.

In the arrangement shown in Fig. 9, the wire 50 may be laid down the centre of the strip 54 and a hook 66 is used to scoop up the wire and loop it over the tab 64, before being lowered further below the plane of the strip 54 and then withdrawn, leaving the wire 50 around the tab 64. Preferably the tab 64 has dimensions such that the action of the hook 66 in lifting the wire does not cause it to come out of the adjacent open crimp 51. Alternatively, a separate feature may be used to hold the wire 50 in the crimp whilst the hook is moving the wire.

In the arrangement shown in Fig. 10, the need to lift the wire 50 is avoided by changing the order of the steps. The tabs 64 are etched into the strip 54, but are not formed up until the wiring stage. The wire 50 is laid into the area between the crimps 51 and a hook (or similar mechanism, not shown), such as that described above, is used to engage with the wire 50 and move it sideways so that it lies outside of the position of the unformed tabs 64a. The unformed tabs 64a are then folded up, for example using a press tool 67, to form a tab 64, and the hook can disengage from the wire, allowing it to move back towards the centre of the strip 54 and engage with the outside of the tabs 64.

Fig. 11 shows an alternative arrangement for the routing of an actuator wire 50 around a sacrificial tab 64 (or any other protruding feature of the strip 54). Such actions will typically require a complex motion in two axes.

In the arrangement in Fig. 11, the feed mechanism 63 is provided with a bobbin 70. The bobbin 70 is specially shaped to achieve the function of feeding the wire 50 so that it is looped over/around a protrusion such as a tab 64.

As the strip 54 is indexed through the apparatus, the wire is drawn down the conical section 71 and out through the lower end. The apparatus is arranged to move the strip until it reaches a stop position with the bobbin 70 approximately aligned with the tab 64 (or other obstacle). The bobbin 70 rotates, causing the hook 72 on the underside to scoop up the wire 50 and loop it around the tab 64. As the bobbin approaches 180 degrees of rotation from the starting point (the point where the wire was first engaged), the profile causes the wire to unhook and any spare slack in the wire can be drawn back into the feed mechanism, for example by the application of slight tension.

Depending on the configuration of the bobbin 70 and any other protruding features on the strip 54, it may be possible to lift the bobbin away from the strip during the indexing process as the strip moves forward through the apparatus.

Fig. 12 shows an alternative arrangement for routing the wire 50 around or over a tab 64 (or any other protruding feature of the strip 54). In this arrangement, as the strip 54 is indexed through the apparatus, the wire(s) 50 are crimped into the crimps 51 downstream of the portion shown. This causes the advancement of the strip to draw wire 50 from a feed mechanism 63 (which may be any known kind of feed mechanism, including rollers, dowels or a capillary) which has the ability to take up slack in the wire 50 after it has been drawn.

A rotating bobbin 75 is positioned over the wire and between a pair of actuators 55 adjacent the crimps 51 and the tabs 64. Its rotation axis is tilted from the direction perpendicular to the plane of the strip 54 and it has a complex surface profile to provide its functionality as described further below.

The bobbin has a hook 76 which engages with the wire 50 as the bobbin rotates. As the bobbin 76 rotates further, the angled axis causes the hook 76 to lift the wire over the tab 64. Immediately behind the hook 76, the bobbin has a skirt 77 which extends over the tab forming a shield and preventing the wire 50 from dropping. As the bobbin 75 rotates further, the wire 50 is draped around the skirt 77 and slides down it until it drops off the bottom of the bobbin and onto the strip 54 outside the tab 64. The profile of the hook 76 then causes the wire extending to the feed mechanism 63 to unhook and the feed mechanism rewinds the spare slack in the wire 50. The bobbin 75 stops with the hook approximately opposite the tab 64 and the strip can then index forward to the next stage. This sequence of steps is shown in Fig. 12B.

Again, depending on the configuration of the bobbin 75 and any other protruding features on the strip 54, it may be possible to lift the bobbin away from the strip during the indexing process.

A number of arrangement of apparatus for continuous manufacturing of actuators 55 which are arranged to introduce slack into the wires 50 between crimps 51 in an actuator will now be described. The exact amount of slack to be introduced is dependent on the design, but for a given design it needs to be very precisely controlled to avoid imbalances within an actuator 55 and inconsistencies between actuators. Naturally, any slack must be added before the crimps 51 are closed, so the wire-laying portions of the apparatus will ideally have a mechanism or operation which ensures that a precisely defined amount of wire is positioned between the crimps 51 prior to the crimps being closed.

Figs. 13 and 14 show two similar arrangements of the apparatus which allow for the introduction and control of the amount of slack in the wires 50 of each individual actuator 55.

In the arrangement shown in Fig. 13, a plurality of additional sacrificial tabs 80 are formed in the strip 54 during the etching process and are folded up so as to protrude from one side of the strip in a similar manner to the sacrificial tabs described above in relation to other arrangements. These additional tabs 80 form features that the wires 50 can be routed around to increase the wire length between the crimps 51 at either end of an actuator 55 by a predetermined and well-known distance. Once the crimps 51 have been closed (for example with a press tool 65), the tabs 80 can be removed and the wire between the crimps at either end of the actuator will be longer than the straight-line distance between the crimps creating the desired slack.

In order to achieve the optimal control of the length of wire and therefore of the slack in the finished actuator, the apparatus may be arranged to pull the wires taught around the tabs 80 and the crimps 51 before closing the latter. Due to the complexity of the routing of the wires 50 through the tabs 80 and the crimps 51, a highly manoeuvrable feed system 63 may be used to lay the wires.

In the arrangement shown in Fig. 14, a similar principle is used in that the wires 50 between the crimps 51 at either end of each actuator 55 are made to follow a convoluted or lengthened route to introduce the desired slack. However, the formation of additional features in the strip 54 (i.e. the tabs 80) adds complication and potentially adds additional forming steps.

Accordingly, in the arrangement shown in Fig. 14, a base plate 81 is provided which has a plurality of pins 82 extending upwards from its upper surface. The base plate 81, which is shown in isolation in Fig. 14B, can be introduced below the strip 54 at the appropriate stage in the manufacture, so that the wires 50 can be routed through the crimps 51 and around the pins 82 (which take the place of the tabs 80 in the arrangement described above in relation to Fig. 13). As well as avoiding additional manufacturing steps, the base plate 81 can ensure consistency in the routing of the wires and the amount of slack introduced. It is also easier to manufacture the pins 82 with rounded edges (compared to tabs 80 which are etched and folded from the strip 54) which can reduce the likelihood of damage to the wires 50.

The base plate 81 is likely to need to move with the strip 54 as an actuator 55 moves from the wire-laying station in the apparatus to the crimping station, before it can be withdrawn (or the strip lifted off) and returned to be re-used.

Fig. 15 shows a further arrangement of the apparatus which provides for the introduction of slack in the actuator wires 50 in each actuator 53. This arrangement is particularly suitable where the actuators 53 include a pair of actuator wires 50 that cross over and therefore the mechanism either has to be arranged to add slack to both actuator wires 50, or to take account of the potential obstruction of a further actuator wire that has already been installed in the actuator.

In the arrangement shown in Fig. 15, two retractable pins 83 are mounted on a mount 84 which sits below the path of the strip 54 as it is indexed through the apparatus. Fig. 15A shows a perspective view of the mount and the actuator which has been simplified to remove the components that are not involved in this operation. Fig. 15B shows a side perspective view of the mount 84. Once the actuator wire 50 (or wires) has been laid, and before the crimps 51 are closed, the mount 84 is raised so that the pins 83 are co-planar with the wire(s) 50. The mount 84 is then rotated clockwise or anti-clockwise (depending on which wire is being installed), thereby routing the wire 50 away from the straight-line route between the crimps 51 at either end of the actuator 53 and increasing the amount of wire between the crimps 51 by a pre-determined (and controlled) amount.

Once the crimps 51 have been closed, the mount 84 and pins 83 are rotated back to their original orientation and can be withdrawn downwards, leaving slack in the wire 50.

It will be appreciated that, with appropriate selection of the location of the pins 83 and the width of the mount 84, the same pins can be used to introduce the slack into each of the wires 50 passing through the actuator 53 by rotating in opposite directions for each wire.

In a development of this arrangement, as shown in Fig. 15C, further static pins 85 are provided adjacent to each of the crimps 51 which ensure that the wire 50 exits the crimp at a fixed angle and the introduction of the slack in the wire using the mount 84 and pins 83 does not cause the wire 50 to be pulled against the edges of the crimps 51.

Figs. 16 to 18 show further arrangements of apparatuses which provide for the introduction of slack in the actuator wires 50 of each actuator 53. As with the arrangement described above, these arrangements are particularly suitable where the actuators 53 include a pair of actuator wires 50 that cross over.

The additional mechanism of the arrangement shown in Fig. 16 includes a clamp 86 and a plunger 87 which are mounted on opposite sides of the track along which the strip of actuators 53 advances. Fig. 16A shows a plan view of the mechanism, whilst Fig. 16B shows a side view.

Once the wire 50 has been laid through the crimps 51, the clamp 86 is lowered so that the arms 88 at the bottom of the clamp are just above the wire 50. The arms 88 do not grip the wire 50, but prevent it from being lifted further than the position of the arms.

Once the clamp 86 is in place, the plunger 87 is raised and engages with the lower side of the wire 50, lifting it. As the wire is prevented from being lifted at the ends of the actuator by the arms 88, a pre-determined additional amount of wire 50 can be drawn into the actuator between the crimps 51. The crimps can then be closed and the plunger 87 and clamp 86 withdrawn, leaving a slack wire between the crimps.

In the arrangement shown in Fig. 16B, the plunger 87 is moved by a mechanical lever 89 and an adjustable end-stop 90 determines the maximum height of the plunger. However, it will be appreciated that other known mechanisms for actuating a plunger of this nature and controlling the extent of its movement could also be used.

For example, Fig. 17 shows an alternative arrangement of the mechanism shown in Fig. 16 in which the plunger 87 is driven by a pneumatic drive 91. In this arrangement an adjustable end-stop 92 is placed above the wire 50 to limit the upwards motion of the plunger and wire.

In both of the arrangements shown in Figs. 16 and 17, the plunger 87 needs to be adapted to prevent interaction with an earlier-placed wire in the actuator, which may already be fixed between closed crimps 51 with the desired amount of slack. Fig. 18 shows a potential configuration of the upper surface of the plunger 87 which allows for this.

The upper surface of the plunger 87 has a slot 93 which is approximately aligned with the likely position of the other wire 5 in the actuator. As a result, as the plunger 87 is raised, the first wire 50a passes through the slot 93 and is not raised by the plunger as it lifts the second wire 50b.

The plunger 87 shown in Fig. 18 is circular in plan view. This allows the orientation of the slot to be rotated so that the plunger 87 can be used on consecutive left- and right-handed actuators 53 as they alternate along the strip 54 whilst providing identical amounts of slack in each of the wires in each of the actuators.

Fig. 19 shows an alternative arrangement of an apparatus which provides for the introduction of slack in the actuator wires 50 of each actuator 53. As with the arrangements described above, these arrangements are particularly suitable where the actuators 53 include a pair of actuator wires 50 that cross over.

In the arrangement shown in Fig. 19, a clamp 86 is used which is similar to the clamp described in relation to Figs. 16 & 17 above. However, in place of the plunger 87, a movable vacuum nozzle 94 is used to lengthen the wire 50 between the crimps 51. Once the clamp 86 is in place, the vacuum nozzle 94 is lowered towards the wire and an airflow induced through the nozzle. This sucks the wire 50 against the lower side of the nozzle 94 which is then lifted to the height desired to introduce the additional length of wire 50 between the crimps 51. The arms 88 of the clamp 86 also act to prevent the wire 50 from being lifted out of the open crimps.

The vacuum arrangement shown in Fig. 19 can allow the upper wire 50 in the actuator to be picked up by the nozzle 94 whilst a lower force is applied to the lower wire due to it being further from the end of the nozzle. As the upper and lower wires 50 will generally be laid in the actuators without changing their vertical relationship along the strip 54 of actuators 53, if the lower wire is laid first (as will usually be the case when the wires are laid from above the strip), then the upper wire can always be attracted in this manner and it is not necessary to take account of the changing orientation of the wires between left- and right-handed actuators.

Fig. 20 shows a further arrangement for the apparatus for continuous manufacturing of actuators 53. This arrangement is designed to improve the positioning of the actuator wires 50 in the crimps 51 prior to the closing of the crimps.

When laying wire 50 into crimps 51, it is desirable that the crimps are pre-formed so that the opening into which the wire is inserted is an acute angle. This allows the press tool that is used to close the crimp to fold them neatly into the desired final position without crumpling the crimps. In some arrangements, this is achieved by a two stage closure process in which a crimp which is formed with a 90 degree (or greater) opening is partly closed by a first pressing step, before being closed by a second pressing step, but that adds complexity to the process and can cause distortion to the crimp location.

In the arrangement shown in Fig. 20, the crimps 51 are formed with acute openings towards the interior of the actuator 53. However, this arrangement means that it is not possible to directly lay the wire 50 into the crimps 51 as the upper surface of the crimp will obstruct this. Accordingly, in the arrangement shown in Fig. 20, the wire 50 is initially laid into the gap between the crimps 51 at either end of the actuator 53. Then, after laying, the wire is moved sideways into the crimps 51 by rotating the feed mechanism 63 and a pair of rollers or pins 95 so that they are positioned outside the line between the crimps 51 at either end of the actuator, thereby pulling the wire under the angled upper surface of the crimps and into the fold of the crimps.

In the arrangement shown in Fig. 20, the feed mechanism 63 is a capillary tube which is sufficiently narrow to pass easily through the gap between the crimps 51, thus allowing the strip 54 to be indexed along. The rollers or pins 95 engage with each wire 50 after it has been laid in the actuator until the crimps on that wire are closed, following which the wire is released. The rollers or pins 95 may then be retracted so that they do not interfere with the indexing of the strip of actuators.

Preferably the rollers/pins 95 and the feed mechanism 63 rotate to move the wire 50 into the crimps 51. The rotation may be about the centre of the adjacent crimp, or could be offset depending on the orientation and dimensions of the design. In other arrangements, the rollers/pins 95 and the feed mechanism 63 can translate perpendicular to the direction of advancement of the strip 54. This will change the wire length between them, but this will be acceptable if the friction in the feed mechanism 63 is low. In yet further arrangements, grippers or hooks could be used to pull the wire 50 sideways into the crimps 51.

Fig. 21 shows a variation in the arrangement of Fig. 20 in which a roller/pin 95 is provided which operates in the centre of the gap between successive actuators 53. Two such rollers/pins 95 are provided so that each roller only pushes in one direction and two actuators 53 are therefore wired at the same time, before the strip 54 indexes by two steps. Fig. 22 shows one possible configuration of the apparatus in the arrangement of Fig. 20. In this arrangement, the feed mechanism 63 and the rollers/pins 95 are linked by an arm 96 which pivots about a pivot point 97 so that they move in concert to push/pull the wire 50 into the crimps 51. The axis of rotation of the arm 96 may be skewed from the plane of the etch, in which case the height of the feed mechanism 63 and the height of the rollers/pins 95 can be adjusted to match the position of the crimps 51.

Fig. 23 shows an arrangement which may be used in an apparatus for continuous manufacture of actuators (including but not limited to any of the apparatuses described herein) for moving an actuator wire 50. In particular the arrangement shown in Fig. 23 may be useful where access to the actuator wire 50 to lift or move it is obstructed. In particular it may allow for pick up of a wire where the exact position of the wire is unknown and therefore a roller which is significantly wider than the wire is needed. Such a roller may be difficult to position and manoeuver around the protruding features on the strip 54.

In the arrangement shown in Fig. 23, the roller 98 is mounted on a slider crank mechanism 99 which is shown in side-view. As the crank mechanism operates, the roller 98 lifts up (from below the wire), picks up the wire 50 and then rotates and translates sideways and

downwards to the desired location (for example to pull the wire into a crimp 51, or to lift it over a projection, such as a tab). This allows the motion of the roller to be kept within a small predefined volume (for example within the confines of the strip 54). It will be apparent that, by appropriate selection of the lengths and positioning of the elements of the crank mechanism, the motion of the roller can be adjusted to suit the particular application required.

Fig. 24 shows a gripper 100 that may be used as part of an apparatus for continuous manufacture of actuators 53. The gripper 100 is designed to deal with the situation that the positioning of the actuator wires 50 which may need to be moved during the manufacturing process (for example as set out in one or more of the above arrangements, but also for other reasons) can be variable due to the manufacturing tolerances of the etching of the crimps 51 and other features on the strip 54 of actuators.

The gripper has two elongate arms 101 which extend from a pivot point 102. The tips of the arms each have a prong 103 which is designed to scoop up the wire from a variable position. As the arms 101 close by rotating about the pivot point 102, a very small aperture 104 is formed which is just large enough for an actuator wire 50 to pass through.

Due to the small dimensions of the aperture 104, its position relative to the pivot point 102 (or any other part of the gripper 100) can be precisely known and vertical and horizontal movement of the wire 50 is effectively prevented. However, as the aperture is slightly larger than the cross-section of the wire 50, the wire can slide through the aperture, for example to allow the addition of slack as discussed above. The gripper 100 may be used to lift an actuator wire vertically and/or shift an actuator wire sideways in order to position it over or round an obstacle (such as the tabs described in some of the above arrangements). The gripper 100 may be additionally or alternatively used to pull a wire into a crimp without adding tension to the wire. Those skilled in the art will appreciate that while the foregoing has described what is considered to be the best mode and where appropriate other modes of performing present techniques, the present techniques should not be limited to the specific configurations and methods disclosed in this description of the preferred embodiment. Those skilled in the art will recognise that present techniques have a broad range of applications, and that the embodiments may take a wide range of modifications without departing from any inventive concept as defined in the appended claims.