GB2164164A - Step and repeat apparatus - Google Patents

Abstract

Step and repeat apparatus used for performance operations upon a generally planar component such as a microprocessor wafer comprises a support (1) for a workpiece, the support (1) being attached by means of an arm (2) to a mechanism adapted to move the support in two dimensions, the arm being pivotally secured to said mechanism at a location (3) remote from the support, and lifting means (19) arranged to move the support about the pivot (3) in a plane pshutters having a plurality of signal electrodes (eg, 46a & c in Fig. 4) crossed by a plurality of common @ <IMAGE>

Description

SPECIFICATION Step and repeat apparatus This invention relates to step and repeat apparatus of the kind used for performance of operations upon an approximately planar component. Such a component may be one of a large number each having features located in a not necessarily regular array on the planar surface. The apparatus may be used for performing operations repeatedly on successive components.

Examples of step and repeat apparatus include microprocessor wafer probes, printed circuit manufacturing apparatus laser trimmers and other apparatus for manufacture of hybrid circuits.

Wafer probe apparatus is required to move and accurately position a wafer in relation to a probe head. Once in position the wafer is raised into contact with the probe. Conventional apparatus comprises a first bed accurately movable in a first dimension, a second bed mounted upon the first and accurately movable in a second pependicular dimension and a support located upon the second bed. The support is raised by a lifting mechanism when the wafer is brought into contact with the probe.

The lifting mechanism is relatively sophisticated since it must raise the whole wafer a precise amount for performance of an operation upon the wafer. The position of the wafer is observed by use of a microscope. Difficulties in design arise in arranging for the microscope, beds and lifting mechanism to fit in the distance between a seated user's knees and eyes.

According to the present invention step and repeat apparatus comprises a support for a workpiece, the support being attached by means of an arm to a mechanism adapted to move the support in two dimensions, the arm being pivotally secured to said mechanism at a location remote from the support, and lifting means arranged to move the support about the pivot in a plane perpendicular to said two dimensions.

The lifting means may be remote from the said mechanism. Furthermore the mechanism is preferably not located beneath the microscope, reducing the knee-to-eye distance of the apparatus.

Avoidance of the need for stacking the said mechanism confers great advantages. The lifting mechanism may have a simple, lightweight design. The mechanism does not need to include relatively massive beds which are rigid to support the lifting mechanism. The mechanism may be concealed beneath a dust cover, whereas beds movable with the workpiece are difficult to cover. In addition concealment of the mechanism reduces noise. Furthermore, access to the workpiece is faciltated, for example for wafer handling apparatus, laser marking devices, probes, line width measurement devices etc. Manufacture of the apparatus is much easier with consequent savings in costs.

The invention is further described by means of example and not in any limitative sense, with reference to the accompanying drawing which is a diagrammatic view of a wafer probe apparatus.

The apparatus illustrated in the drawing (which is not to scale) comprises a support for a workpiece 1 located at the end of an arm 2 which is pivotally mounted by means of a hinge 3 remote from the support 1 but located in the horizontal plane of the latter when engaging the probe to ensure vertical movement of the wafer as it contacts the probe. The workpiece may be secured to the support by application of a vacuum from a suction device.

A counterbalance 4 is arranged so that the arm 2 and support 1 pivot about their centre of gravity. This allows the arm to have a relatively simple and inexpensive sheet metal construction.

The counterbalance of alternative embodiments of the invention may be replaced by a spring in order to reduce the inertia of the apparatus. A motor 5 drives a concealed chain loop (not shown) connected to the chuck or platter 6 of the support 1. The chuck 6 is rotatable about a vertical (z) axis; the motor 5 providing means for adjusting the angular disposition (o) of a workpiece secured to the chuck.

The hinge 3 is attached to mechanism which enables the arm, and hence the support, to be moved in two perpendicular dimensions (x and y).

A block 7 from which the hinge 3 depends is arranged to slide in the (x) direction upon an accurately machined bar 8. A ball screw 9 driven by a motor 10 engages a ballnut contained in the block 7 to drive the block along the bar 8. End supports 11,12 of the bar 8 are arranged to slide in the (y) direction upon respective accurately machined bars 13,14. The bars 13,14 are aligned perpendicularly to the bar 8. A ball screw 15 actuated by a motor 17 and chain drive 16 drives the end support 11.

The (x, y) mechanism is located upon four accurately machined supports 18.

Movement of the chuck 6 about the pivot in the z direction is controlled by a lift mechanism 19 such as a threaded member, driven by a drive belt 20 and motor 21. The chuck 6 is not attached to the lifting mechanism but merely rests upon it.

The advantages of the invention are apparent from consideration of the steps involved in assembly of the apparatus.

In the first step four coplanar location sites 18 are machined on the base plate (not shown).

Prior art apparatus comprising x and y beds require entirely machined baseplates to support their weight.

The parallel shafts 13,14 are located in inexpensive recirculating ball bushings attached to the location sites 18. The shafts 13,14 may be simply aligned with a gauge.

The cross bar 8 is mounted on the end support slides 11,12 between the bars 13,14. In prior art devices a cross table is suspended between beds movable on an orthogonal axis.

The arm 2 is mounted upon the bar 8. Perpendicularity of the bar 8 may be easily checked against the accurately perpendicular grid of a wafer mounted on the support 1.

Maintenance of the x,y mechanism does not involve interference with the support 1, in contrast to prior art devices in which the support surmounts the x,y mechanism.

The lightweight construction of the x,y mechanism allows the use of lightweight recirculating ball screws for the drives. Prior art devices employ more expensive ball or lead screws.

The apparatus finds application in wafer probes, laser trimmers and other apparatus for manufacture of hybrid circuits and other electronic components.

In addition to the mechanical arrangement described above, the apparatus also incorporates a microprocessor control system, edge sensors to detect the edges of a wafer under test and markers for applying ink marks to defective chips. The apparatus is coupled in use to a host computer system adapted to test a chip connected to the contacts of the probe.

The control system serves to give a user manual and automatic control of the location of microscope stage and of the height and attitude of the chuck. The control system includes three sub-systems: a main sub-system and two motor control sub-systems. The main sub-system serves to actuate the user-system interface i.e. a keyboard and VDU display. Operation of the motor control sub-systems which actuate the motors is controlled by the main sub-system. The main sub-system also communicates with the host computer, controls the interface with the host computer and controls the edge sensing equipment and chip marker.

The motor control sub-systems are adapted to communicate with the host computer, and to generate signals for actuation of the respective motors. Acceleration and deceleration of motors and protective functions are also controlled by these sub-systems.

The keyboard of the user-system interface may have the following keys i) DlSP-this key allows the existence and termination of system; functions ii) HOME-this key returns the stage to the "HOME" or load positions; iii) PAUSE-this key permits the operator to "PAUSE" or suspend function operation; iv) VAC-this key permits the operator to activate or deactivate the vacuum hold down circuit.

v) LCL-this key permits the operator to interrupt the host computer in the event that the operation of the machine is impaired or malfunctioning.

vi) MARK this key has several functions dependent on a sub-function activated at the time in question.

vii) CONT-this key permits the operator to "CONTINUE" the function operation after previously being "PAUSED".

viii) DEL--this key permits the operator to "DELETE" a previously entered data digit.

ix) ENTER-this key permits the operator to "ENTER" a selected function or sub-function or to teminate data string entry.

x) "numeric"--these keys enable the operator to enter numeric data in order to select or modify index size, Z lift parameters etc.

The host system communicates with each motor control sub-system along a twenty six way bus. The bus (local bus) comprises the following: Data Out4 lines of binary coded decimal data.

Data ln-4 lines of binary coded decimal data.

Status ln-4 lines of co-processor status.

Status Out4 lines of main-processor status.

Drive Select Out8 lines permitting a maximum of 8 co-processors to share the bus at any one time.

Reference Supplies2 lines tying up to logic Ov to 5v supplies.

The data sent and received along the Local Bus consists of two main types of numeric values: i) Absolute position data (for both axis) ii) "Stop when" position data (for both axis).

The host computer communication system provides a capability of exercising control and the ability to set up or modify wafer parameters remote from the system itself i.e. on-line.

The message interchange via either of the two communication systems available fall into two categories namely: i) "Passive" commands-these are commands that result in no change in position of the stage after command execution.

The ATE/tester interface readily permits the user to link the probing system and test system.

A system ancillary port permits the monitoring of the edge sensing system and activation of the three inker drive circuits.

A password protection system prevents accidental or unathorised access to the data contained within the apparatus.

The apparatus incorporates means for detecting the edge of a wafer and for minimising the time spent locating the chips on the wafer.

The traditional approach to this has been the use of a two crossed probe assembly known as an edge sensor, such as a break-on-contact switch. The switch opens as it contacts the wafer.

Opening of the contact pair indicates that the probing fixture is positioned above the wafer. If the contact pair remains closed this indicates that the probing fixture has stepped over the edge of the wafer. This action of stepping over the edge of the slice should then cause the stage to drive forward one index size in the Y axis.

Preferred apparatus in accordance with this invention is provided with the ability to determine the location of a complete device for test relative to the location of the preceding device.

In order to ascertain whether the system needs to increase or decrease to the left or right, the length of a row of complete devices to be probed, the system first determines which of four quadrants of the wafer the probing fixture is in.

The four quadrants may be defined as: A Upper, left of centre9 o'clock to 12 o'clock on a clock face B B Upper, right of centre12 o'clock to 3 o'clock on a clock face C Lower, left of centre6 o'clock to 9 o'clock on a clock face D Lower, right of centre3 o'clock to 6 o'clock on a clock face.

If the stage is in quadrant A relative to the probing fixture position and if the direction of travel is to the left then there is no further calculation required until quadrant B is reached. If the direction of travel is to the right, then when the edge is predicted for the next index right, the stage must move forward one index and shorten the row from the left.

If the stage is in quadrant B relative to the probing fixture position and if the direction of travel is to the right then there is no further calculation required until quadrant A is reached. If the direction of travel is to the left then when the edge is predicted for the next index left, the stage must move forward one index and shorten the row from the right.

If the stage is in quadrant C relative to the probing fixture position and if the direction of travel is to the left then no further calculation is required until the quadrant D is reached. If the direction of travel is to the right then when the edge is predicted for the next index right, the stage must move forward one index and increase the row from the left.

If the stage is in quadrant D to the probing fixture position and if the direction of travel is to the right then there is no further calculation required until quadrant C is reached. If the direction of travel is to the left then when the edge is predicted for the next index left, the stage must move forward on index and increase the row from the right.

The basic method for determining where to go next involves the questions: a) is the stage left or right of centre? b) which direction is it moving? c) from a) and b) above-does the length of the row increase or decrease? In most instances the system will already know the current direction of travel and only a) and c) would need to be calculated.

The following are examples of the two algorithms required.

co~ord~left: check left co-ordinates ; if error move to next for right call calc~next~left ; ove to next left comp use~x,0 ; null index? jz left~fail~a ; branch yes call set~up~x ; prepare x sub ax,factor~x ; make destination ic left~a~fail~a ; branch out of limit call which ; which sector ic ok~left~a ; branch positive mov dest~x,ax ; save destination offset call prepare~y pos ; set up y limits call centre~down ; chuck centre - dest down jc pos~pos~quad ; brach++quadrant mov dest~down,ax ; save down offset call centre~up ; chuck centre - dest up ic pos~neg~quad ; branch ± quadrant mov dest-up,ax ; save it call up~squared ; (dest~up) x (dest~up) call rad~up ,((radius) x (radius) - (dest~up) x (dest~up)) ic left~fail~1 ; branch out of radius call test~dest~x ; (test - (dest~x) x (dest~x)) Ic left~fail~l ; branch out of limit ok~left~a: call index~left ; index left ok left: mov al,Offh ; ok flag left test end: cop al, Offh ;det flags ret left~fall~a: imp left~fail~2~: pos~neg-quad: call do~cal~up ; start calc ic left~fail~1 ; branch error call test~dest~x ; (test-(dest~x) x (dest~x) ic left~f ail~1 ; branch out of limit call down~squared ; (dest~down) x (dest~down) call rad~down ; ((rad) x (rad) - (dest~down) x (dest~down)) jc left~fail~2 ; branch error call test~dest~x ; (test - (dest~x) x (dest~x)) jc left~fail~2 ; branch out of limit jmp short ok~left~a ; loop passed pos~pos~quad: call do~cal~down ; start talc jc left~fail~2 ; branch error call test~dest~x ; ( test - (dest~x) x (dest~x)) jc left~fail~2 ; branch out of limit jmp short ok~left~a ; loop passed left~fail~1: call calc~next~right ; correct cmp skip~it,Offh ; skip this? jnz left~fail~skip~1 ; branch no call down~fit ; ok to go down? jnz left~fail~1a ; branch no left~fail~skip~1: mov al,O ; flag change direction imp left~test~end ; exit left~fail~1a: call calc~next~right ; calc next right index cmp use~x,0 ; null index? jz end~of~slice~left ; branch error call centre~next~x ; chuck centre - next x jc end~of~slice~left ; branch yes call down~fit ; ok to go down? jnz left~fail~1a ; branch no and try again call index~right ; do index mov al,0 ; flag change direction imp left~test~end ; exit end~of~slice~left: cal index~right ; do it pop ax ; waste return imp go~load ; exit auto left~fail~2: call calc~next~right ; correct left~fail~2~x: clip skip~it,Offh ; skip it? jnz left~fail~skip~2 ; branch yes call calc~next~down ; calculate next down call index~down ; do it left~fall~2a: call calc~next~left ; calculate next left index cip use~x,0 ; null index? jz left~fail~3: ; branch yes call set~up~x ; prepare x sub ax,factor~x ; make destination jc left~fail~3 ; branch out of limit call centre~dest~x ; chuck centre - dest x call do~fail~lr ; calc a bit jc left fail~3 ; branch error call test~dest~x ; (test-(dest~x) x (dest~x)) jc left~fail~3 ; branch out of limit imp short left~fail~2a ; loop for more left~fail~3: call calc~next~right ; correct left~fail~3a: call index~left ; do it left~fail~skip~2: mov al,0 ; set change direction imp left~test~end ; loop back co~ord~right: ; check right co-ordinates ; if error move to next for left call calc~next~right ; calculate next right cmp use~x,0 ; null index? jz right~fail~a ; branch yes call set~up~x ; prepare x add ax, factor~x ; make destination call which~x ; which sector? Inc ok~right~a ; branch negative not ax ; two s complement inc ax mov dest~x,ax ; save it call prepare~y~pos ; set up y limits call centre~down ; chuck centre - dest down jc pos~pos~quad~right ; branch t quadrant mov dest~down,ax ; save down offset call centre~up ; chuck centre - dest .p ic pos~neg~quad~right ; branch ± quadrant mov dest~up,ax ; save it call up~squared ; (dest~up) x (dest~up) call rad~up ; ((rad) x (rad) - (dest~up) x (dest~up)) jc right~fail~i ; branch error call test~dest~x ; (test - (dest x) x (dest x)) Ic right fail~1 ; branch out of limit ok~rightz~a: call index~right ; index right ok~right: mov al,Offh ; ok flag right~test~end: cip al,Offh ; set flags ret right~fail~a: imp right~fail~2~x pos~neq~quad~right: call de~cal~up ; start calculation jc right~fail~1 ; branch error call test~cpsi~x ; (test - (dest~x) x (dest~x) Ic right~fail~l ; branch out of limit call down squared ; (dest down) x (dest~down) call rad~down ; ((rad) x (rad) - (dest~down) x (dest down)) jc right~fail~2 ; branch error coll test~dest~x ; (test - (dest~x) x (dest~x)) jc right~fail~2 ; branch out of limit jntp short ok~right~a ; loop passed pos~pos~quad~right: call do~cal~down start calc Ic right~fail~2 branch error call test~dest~x ; (test - (dest~x) x (dest~x)) jc right~fail~2 ; branch out of limit jmp short ok~right~a ; loop passed right~fail~1: call calc~next~left ; correct call donwn~fit ; ok to go down? jnz right~fail~1a ; branch no eov al,O ; flag change direction jmp right~test~end ; exit right~fail~1a: call calc~nex:~left ; calculate next left index cup use~x,0 ; null index? jx end~of~slice~right ; branch error call centre~next~x ; chuck centre - next x inc end~of~slice~right ; branch yes call down~fit ; ok to go down? jnz right~fail~1a ; branch no and try again call index~left ; do it mov al,0 ; flag change direction imp right~test~end ; exit end~of~slich~right: pop ax uaste return imp go~load ; exit auto right~fail~2: call calc~next~left ; correct right~fail~2 x: call calc~next~down ; calc next index down call indes~down ; do it right~fail~2a: call calc~next~right ; calculate next right cmp use~x,0 ; null index? jz right~fail~3a ; branch yes call get~up~x ; prepare x add ax, factor~x ; make destination mov jest~x, ax ; save it call dest~x~centre ; do it call do~fail~lr ; caic a bit jc right~fail~3 ; branch error call test~dest~x ; (test - (dest~x) x (dest x)) jc right~fail~3 ; branch out of limit jap short right fail? a ; loop for more right~fail~3: call calc~next~left ; correct right~fail~3a: call index right ; do it mov al,0 ; set change direction Imp right~test~end ; loop back do~call~up: not ax ; two's complement inc ax mov dest~up, ax ; save dest. up call up~squared ; (dest~up) x (dest~up) jmp r+d~up ; ((radius) x (radius) - (dest~up) x (dest~up)) do~cal~down: not ax one's complement inc ax ; two's complement mov dest~down, ax ; save it call down squared ; (dest~down) x (dest~down) jmp rad~down ; ((rad) x (rad) - (dest down) x (dest~down)) do~fail~lr: mov dest~x,ax ; save it call prepare~y~pos ; set up y co-ordinates call down centre ; dest down - chuck centre moy dest~down~ax save it call down~squared ; prepare down limit jmp rad~down ; create now limit public down~fit,up~fit down fit: call calc~next~down ; calculate next down index cmp use~y,0 ; null index? jz down~fail~a ; branch yes call set~up~v ; prepare y sub ax,factor3 ; make destination jc down~fail~a ; branch out of limit call which~v ; which sector? ic ok~down~@ ; branch positive nov dest~y,ax save it call prepare~x~pos ; set up x limits call centre~left ; chuck centre - dest left jc pos~pos~quad~down ; branch ++ quadrant mov dest~left,ax ; save left offset call centre~right ; chuck centre - dest right jc pos~neq~quad~down ; branch + quadrant call left~squared ; (dest~left) x (dest~left) call rad~left ; ((rad) x (rad) - (dest~left) x (dest~left)) jc down~fail ; branch error call test~dest~v ; test - (dest~v) x (dest~v)) Ic down fail ; branch out of limit ok~down~a call index~down ss do it ok~down nov al,Offh ; ok flag down~test~end cmp al,Offh ; set flags ret down~fail~a: jmp down~fail~v pos~pos~quad~down: call right~centre ; dest right - chuck centre iov dest~right,ax ; save offset call right~squared ; (dest~right) x (dest~right) call rad right ; ((rad) x (rad) - (dest~right) x (dest~right) jc down~fail ; branch error call test~dest~v ; (limit - (dest~v) x (dest~v)) jc down fail ; branch out of limit jFp short od~down~a ; else pass ok.

pos~neq~quad~down: not ax ; two s complement inc ax row dest~right,ax ; save dest. right call right~squared ; (dest~right) x (dest~right) call rad~right ; ((rad) x (rad) - (dest~right) x (dest~right)) jc down~fail ; branch error call test~dest~v ; (test - (dest~v) x (dest~v) jc down~fail ; branch out of limit call left~squared ; (dest~left) x (dest~left) call rad~left ; ((rad) x (rad) - (dest~left) x (dest~left)) jc down~fail ; branch error call test~dest~y ; (test - (dest~y) x (dest~y)) jc down~fail ; branch out of limit jmp short ok~down~a ; loop passed These algorithms determine whether a single point on the wafer is located under the probing fixture.As the "die" or device to be tested is always two dimensional, the software needs to perform a "die" sized "fit" into the theoretical wafer area.

The edges of wafers are usually slightly non-circular for two reasons.

i) A "flat" is used for orientation and tooling purposes.

ii) Chamfering of the edge of the wafer.

Apparatus in accordance with this invention may overcome any difficulties caused by noncircular wafers by making the assumption that any "flat" on the wafer will adhere to sizes set down by industrial standards and that the chamfer can preclude the need to test part complete devices on the periphery of the wafer.

This problem is overcome by calculating whether it is possible to fit one and a half devices into the space remaining on the wafer, in an analogous manner to the previously listed algorithms. Use of this method elminates the need for multiple edge sensors and avoids the missing of complete chips which may occur in prior art apparatus if the wafer has an awkward aspect ratio.

The tips of a probe card are generally arranged to lie in a common plane. Three main reasons have been found for failure of probe cards: i) Lack of planarity between wafer surface and probing fixture caused solely by mechanical misalignment of the workpiece holder relative to the chuck surface.

ii) The probing system attempting to "probe" incomplete devices resulting in unequal stress across the probe. In most cases this has been caused by excess travel in use of the normal edge sensing technique.

iii) When using a simple "dead" lift operation of the chuck. That is to say a chuck lifting mechanism which is solenoid operated. This gives rise to excessive instantaneous force being applied to both wafer and probing fixture. Also the resultant "switch bounce" of the probe tips gives unpredictable contact characteristics which dramatically effect the results obtained in parametric testing.

The present apparatus may overcome these problems as follows.

The provision of the edge sensing capability described above eliminates the risk of probing incomplete devices. By confirming the use of the edge sensing circuit purely as a height detector (to establish the amount of wafer touchdown as opposed to wafer presence), the chuck will only lift the amount required and will not "bounce" the probing fixture.

Mechanical accuracy of the chuck lifting mechanism is achieved by the unique "platter-lift" pusher assembly. It now follows that the incremental lift of the chuck needs to be of such a small magnitude as to correct for any small taper in the wafer caused by the sawing or slicing process.

The chuck lift on the system describd above may have a single step up or down resolution of approximately 0.00025 of an imperial inch (6.3X10 4cm). By using such a fine resolution in chuck lift, reliable and accurate probing force can be maintained.

The apparatus may be provided with the ability to store the last height location at which the wafer was detected. By operating a search window of +0.0005" and -0.0005" (+1.27X10-3cm) to allow for taper of the wafer, the possibility of damage to the probing fixture caused by a malfunctioning edge sensor can be precluded.

Failed devices are marked with a dot of ink which may be up to 0.030" (7.6X 10 2cm) in diameter. A further problem is caused by dragging the probing fixture tips through the ink dot, which may be 0.008"-0.009" (2.0X10 2-2.3X10 2cm) in height above the wafer surface causing the probe tipes to be coated with marking ink. This may cause partial or total electrical isolation between the testing system and the device under test.

In order to overcome this problem the system is provided with the ability to remember during its next index to an adjacent device whether or not the device has been marked (inked). If the current device has been inked the chuck is caused to lower by a pre-programmed amount in order to clear the offending "mountain" of ink. If the device has passed the testing stage, the chuck will only lower the required amount to ensure that the probe tips are clear of the wafer surface.

Another situation that can cause premature termination of the useful life of a probing fixture is fouling of the edge sensor.

To overcome this problem, physical points in the chuck lift cycle are established at which the edge sensor should have been expected to have opened and closed. Failure for the edge sensor to perform in the expected manner causes the chuck system to cease movement and the operator is advised of the problem. This prevents any further damage to the probing fixture and, or cour, to the wafer under test.

The step and repeat apparatus system software has an internal non-user accessed local bus.

The local bus is physically configured as previously described. The local bus supports one controller interface and two motor control co-processors.

The local bus interface has the facility to read and write information to and from the selected co-processor at any mutually convenient time. This time is ascertaied by the main processor polling the ready/busy status line.

The main processor has the capability of initialising the current position of one or all the available co-processors, or the bus, with a value of 0-16,500 inclusive.

The main processor has the capability also of being able to interrogate (read) the position of the stage at any point of travel.

The main processor can command the coprocessor to go to a location on their respective axes simply by sending the desired location (0-16,500) as a destruction command.

The main processor has the facility of sending single four bit commands to the co-processor with the interpretations as listed below.

0000 : Traverse main fast clockwise 0001 : Traverse main medium clockwise 0010 : Traverse main slow clockwise 0011 : Single step main clockwise 0 1 0 0 : Traverse aux. until aborted clockwise 0101 : Single step aux. clockwise 0 11 0 : Next data is absolute position 0 111 Next data is absolute destination 1000 : Traverse main fast anti-clockwise 1001 : Traverse main medium anti-clockwise 1010 . Traverse main slow anti-clockwise 1011 : Single step main anti-clockwise 1100 : Traverse aux. till aborted anti-clockwise 1101 . Single step aux. anti-clockwise 11 10 : Drive to limit 'n' main axis 1111 : Drive to limit 'n' aux. axis

Claims (1)

1. Step and repeat apparatus comprising a support for a workpiece, the support being attached by means of an arm to a mechanism adapted to move the support in two dimensions, the arm being pivotally secured to said mechanism at a location remote from the support, and lifting means arranged to move the support about the pivot in a plane perpendicular to said two dimensions.

2. Apparatus as claimed in claim 1, wherein the lifting mechanism is remote from the said mechanism and adapted to move the support.

3. Apparatus as claimed in claim 1 or 2, including a probing fixture adapted to be engaged by a workpiece wherein the pivotal axis of the arm and the location of engagement of the workpiece and probing fixture are disposed in a common horizontal plane.

4. Apparatus as claimed in any preceding claim, including a probing fixture adapted to engage a succession of devices on a wafer, further comprising means for automatically moving the support to align and engage successive devices with the probing fixture and means adapted to determine the locations of sucessive complete devices on the wafer.

5. Apparatus as claimed in claim 4, further comprising safety means adapted to prevent engagement of the probing fixture with any incomplete devices at an edge of the wafer.

6. Apparatus as claimed in claim 5, including an edge sensor adapted to detect location of an edge of the wafer, safety means arranged when an edge is not detected to determine whether the edge sensor should have detected an edge of the wafer and further arranged to prevent movement of the support if an edge should have been detected

7. Apparatus as claimed in claim 5 or 6, including a memory adapted to store the vertical location of a device engaged by the probing fixture and means adapted to calculate the possible vertical locations of a successive device.

8. Apparatus as claimed in any claims 5 to 7, further comprising a marker adapted to apply markings to rejected devices, and means adapted to calculate and store the location of any such markings and further adapted to prevent contact of the probe ring with the markings.

9. Apparatus as claimed in any of claims 4 to 8, arranged to locate complete devices arranged in rows on a wafer relative to preceding devices on the wafer by performing the steps of: dividing the wafer into four quadrants, determining the location of the probe fixture in relation to the four quadrants, determining the location of an end of a row of devices on the wafer, calculating the location of the start of the succeeding row of devices and calculating the length of the succeeding row of devices.

10. Step and repeat apparatus substantially as hereinbefore described with reference to the accompanying drawings.

CLAIMS Amendments to the claims have been filed, and have the following effect: Claim 1 above has been deleted or textually amended. New or textually amended claims have been filed as follows

1. Electronic compound handling step and repeat apparatus comprising a support for a workpiece including a said component, the support being attached by means of an arm to a mechanism adapted to move the support in two dimensions, the arm being pivotally secured to said mechanism at a location remote from the support, and lifting means arranged to move the support about the pivot in a plane perpendicular to said two dimensions.