Patent Application
20060086889
The invention concerns an opto-electronic arrangement to capture relative movements or relative positions of two objects. This arrangement includes at least one position-sensitive detector, and each position-sensitive detector is illuminated by a light emission device, to form a measuring cell. The invention also concerns a force and/or moment sensor, which makes use of such an arrangement. Finally, the invention concerns a PC keyboard which has the force and/or moment sensor according to the invention.
For the computer user, it is becoming increasingly important to control three-dimensional movements via a peripheral device. A three-dimensional deflection is captured by the peripheral device and described as a translation (X, Y, Z) and/or a rotation (A, B, C) in space. The most important component is the sensor, which can measure the deflection in up to six (6) degrees of freedom.
DE-36 11 337 A1 discloses an opto-electronic arrangement which is housed in a plastic sphere, and can simultaneously capture six components, namely translations along three axes and angular rotations around three axes. For this purpose, six light-emitting devices are arranged in one plane at essentially the same angular distances from each other. A permanently arranged slotted diaphragm is connected in front of each light-emitting device. The relative movements or relative positions are detected by light-sensitive detectors, which are arranged so that they can move relative to the arrangement of light-emitting devices and slotted diaphragms, and the detector axis of which runs perpendicularly to the direction of the slots. The arrangement requires a relatively low design cost, since the light-emitting devices, diaphragms and other electronic devices as required for control and analysis can be arranged using conventional soldering technology on a single board, which can be permanently connected to a first object. The position-sensitive detectors are connected to the second object. However, it is disadvantageous that the arrangement takes up a relatively large area. The reason is the relatively large spatial extent of the diaphragms and detectors which are arranged in a ring around the light emission devices. This sets a limit to the miniaturisation of the arrangement.
Without any claim to completeness, other documents which show the technical background for the invention are: DE-27 27 704 C3, DE-36 11 336 C2, DE-32 40 241 A1, U.S. Pat. No. 3,921,445 and U.S. Pat. No. 3,628,394.
Problem on Which the Invention is Based
Opto-electronic arrangements to capture relative movements or relative positions, as well as force and/or moment sensors which use such arrangements, gained significance in the past mainly in industrial applications. Examples are control of robots and measurement of forces on motor vehicle test and measuring stands. But commercially very interesting application possibilities for the arrangements and sensors exist mainly in the office sector and entertainment electronics. Here they have the function of an input device with which up to six components can be input, in contrast to a joystick, a mouse or a trackball, which in general allow only two components to be input. Simple, convenient input of six components, as is allowed by a force and/or moment sensor with an opto-electronic arrangement, is desirable, for instance, to control 3D design software and sophisticated computer games. However, input devices until now have been decidedly unwieldy because of the area/volume which they require, and this was essentially the reason they were not more widely used. Miniaturisation would allow them to be used in, for instance, games consoles, PC keyboards or notebook computers, and thus make a broad market penetration possible.
The typical 3D input devices are used for view manipulation of three-dimensional objects in 6 degrees of freedom simultaneously (6 DOF=3 translations and 3 rotations). The cap or sphere of the 3D input device is carried on springs, and allows arbitrary deflection in space (6 DOF). This group of input devices is aimed at customers with true 3D applications (6 DOF), such as Catia or other CAD applications.
As well as the true 6 DOF applications, there is also a large group of applications in which rotating an object is not desired. Examples of such applications are the Office products (Word, Excel, Powerpoint etc.) and image processing programs (Adobe Photoshop, Acrobat Reader etc.). The manipulated object is usually a two-dimensional master (“paper with writing or pictures”), and rotating the master is not desired. The customer still wishes to change the view, but only by shifting (pan—2 DOF) and enlarging/reducing (zoom—1 DOF) the object.
The aim of development for this customer group is to build an input device which is especially suitable for pan/zoom applications. In this way, the high cost of a full 3D sensor (6 DOF), in which the three rotational movements are simply ignored, could be saved.
Thus, starting from the prior art, this invention is based on the object of creating an arrangement for capturing the relative movements or relative positions of two objects, said arrangement allowing a more manoeuvrable design compared with known arrangements. For instance, the design of the arrangement could be more efficient and/or flexible, or require a smaller area. Also, the design of the arrangement could be more economical and/or be specially suitable for pan/zoom applications.
Additionally, the invention is based on the object of creating a force and/or moment sensor which also allows a more elegant design in comparison with known sensors. Finally, the invention is based on the object of creating an input device which is for use in the office and allows uncomplicated input of up to six force or torque components.
Solution According to the Invention
To achieve this object, the invention discloses an opto-electronic arrangement, which is defined by the features of claim 1, 10, 22, 29 or 35, to capture relative movements or relative positions of two objects. The invention also discloses a force and/or moment sensor, which is defined by the features of claim 42. Preferably, the force sensor is used as a pan/zoom sensor for image processing and other similar office applications. Finally, it also discloses a personal computer keyboard which is defined by the features of claim 53.
Structure and Further Development of the Solution According to the Invention
An opto-electronic arrangement for capturing relative movements or relative positions of two objects according to one form of the invention includes at least one position-sensitive detector, and is characterized in that the position-sensitive detector is illuminated by at least two light emission devices, to form two measuring cells with a common detector.
Preferably, each of the two measuring cells which are formed by a common detector has a slotted diaphragm which is arranged in the beam path of the corresponding light emission device, between the said light emission device and the position-sensitive detector. Each position-sensitive detector can be associated with two adjacent slotted diaphragms.
In a preferred version of the opto-electronic arrangement, a slot direction of at least one of the slotted diaphragms is aligned diagonally in relation to the light-sensitive part of the detector. In another preferred version of the opto-electronic arrangement, a light plane, which shines through at least one of the slotted diaphragms and falls on the detector, forms an angle with a plane of a light-sensitive part of the detector.
It is preferred that each detector is illuminated alternately (i.e. periodically) by a light emission device. A measurement value of the detector is read out simultaneously. In other words, the detector of each measuring cell is illuminated by only one light emission device at a particular time, and the measurement value of the detector is read out simultaneously.
Typically, the measuring cells with the common detector are arranged such that the beam paths which emanate from the light emission devices intersect and illuminate the same portion of the common detector in the plane of their intersection.
An opto-electronic arrangement according to a further form of the invention includes at least one position-sensitive detector, which is illuminated by a light emission device, to form a measuring cell, which also has a slotted diaphragm which is arranged in the beam path of the light emission device between the light emission device and the position-sensitive detector. This opto-electronic arrangement is characterized in that a light plane which shines through the slotted diaphragm and falls on the detector is oriented at an angle in relation to a light-sensitive part of the detector.
In a preferred version of the opto-electronic arrangement, the light plane forms an angle with a plane of the light-sensitive part of the detector. Preferably, a slot direction of the slotted diaphragm runs essentially perpendicularly to the light-sensitive part of the detector.
In an alternative version of the opto-electronic arrangement, a slot direction of the slotted diaphragm is aligned diagonally in relation to the light-sensitive part of the detector.
In a preferred version of this opto-electronic arrangement of the invention, the position-sensitive detector is associated with two adjacent slotted diaphragms, said position-sensitive detector acting as part of two different measuring cells. Preferably, each slotted diaphragm is illuminated by its own light emission device, so that each position-sensitive detector is illuminated by two light emission devices, to form one measuring cell with a common detector.
In a particularly preferred configuration, each of the two adjacent slotted diaphragms is illuminated by a respectively arranged light emission device. The two adjacent slotted diaphragms can together enclose an angle, and can also preferably have slots which are arranged perpendicularly to each other.
An opto-electronic arrangement to capture relative movements or relative positions of two objects according to yet another form of the invention includes at least one position-sensitive detector, each position-sensitive detector being illuminated by its own light emission device, to form a measuring cell. This opto-electronic arrangement is characterized in that the measuring cells are arranged in groups, so that the measuring cells of each group are essentially arranged parallel or perpendicularly to each other.
In a preferred version of this opto-electronic arrangement, the measuring cells also each include a slotted diaphragm which is arranged in the beam path of the light emission device between the light emission device and the position-sensitive detector, a detector axis of the position-sensitive detector being aligned essentially perpendicularly to a slot direction of the slotted diaphragm. The detector axes of the position-sensitive detectors in each group of measuring cells are preferably arranged parallel to each other.
According to another form of the invention, an opto-electronic arrangement to capture relative movements or relative positions of two objects includes at least two position-sensitive detectors, each position-sensitive detector being illuminated by its own light emission device, to form a measuring cell. This opto-electronic arrangement is characterized in that all position-sensitive detectors and light emission devices are arranged in a common plane, and that the measuring cells are arranged parallel to Cartesian axes. The measuring cells can therefore be arranged essentially parallel to each other and/or essentially perpendicularly to each other.
In a preferred version of this opto-electronic arrangement, the measuring cells also each include a slotted diaphragm which is arranged in the beam path of the light emission device between the light emission device and the position-sensitive detector, a detector axis of the position-sensitive detector being aligned essentially perpendicularly to a slot direction of the slotted diaphragm.
In a preferred version of the opto-electronic arrangement of the invention, an element of each measuring cell, consisting of light emission device, slotted diaphragm and detector, is movable relative to the other two elements. The movable element is preferably arranged in the centre of rotation of the measuring cell, so that the measuring cell can generally only (i.e. exclusively) capture translatory movements. In principle, therefore, this measuring cell cannot capture rotational movements. Rotations can be captured only if the movable element is at a distance from the centre of rotation. If this distance from the centre of rotation is zero or minimal, the measuring cell is “blind” or “almost blind” to the rotational movement.
According to another form of the invention, an opto-electronic arrangement to capture relative movements or relative positions of two objects includes at least one position-sensitive detector, each position-sensitive detector being illuminated by a light emission device, to form a measuring cell, and the measuring cell also has a slotted diaphragm which is arranged in the beam path of the light emission device between the light emission device and the position-sensitive detector. One element of the measuring cell, consisting of light emission device, slotted diaphragm and detector, is movable relative to the other two elements, and the measuring cell can capture exclusively translatory movements. The movable element of the measuring cell may be arranged in the centre of rotation of the measuring cell. Preferably, the movable element of each measuring cell is arranged in the centre of rotation of the corresponding measuring cell.
In a preferred version of the opto-electronic arrangement of the invention, the arrangement includes at least three measuring cells, preferably from three to six measuring cells or even more than six measuring cells.
In a preferred version of the opto-electronic arrangement of the invention, at least one measuring cell consisting of light emission device, slotted diaphragm and detector is provided with a movable light emission device, this measuring cell having a greater working range or movement range.
In a possible extension of the invention, all light emission devices, preferably infra-red light-emitting diodes (ILEDs) and position-sensitive detectors, preferably position-sensitive infra-red detectors, are arranged in a common (first) plane.
According to another aspect of the invention, a force and/or moment sensor is provided, which is characterized by an opto-electronic arrangement according to the invention to capture relative movements or relative positions of two objects. The two objects preferably consist of a first plate and a second plate, which are elastically joined to each other and movable relative to each other.
The 3D input devices according to the invention can be equated to a force and/or moment sensor. The translatory movements (X, Y, Z) correspond to the forces (Fx, Fy, Fz), and the rotational movements (A, B, C) correspond to the moments (Mx, My, Mz). A pan/zoom sensor corresponds to a force sensor (Fx, Fy, Fz), since the pan/zoom sensor can capture only translatory movements (X, Y, Z).
Other preferred arrangements of the invention are disclosed in the independent claims and in the following description of embodiments.
Embodiments are shown in the following figures. Components with identical or similar functions are marked with the same reference symbols.
a-5c show possible changes of the parameters of the measuring cell without functional effect;
a, 6b show a measuring cell of an opto-electronic arrangement according to the invention, with rotation of the diaphragm about the vector LEDdir;
a-11c show the structure of measuring cells of an opto-electronic arrangement according to the invention, multiple measuring cells being combined with each other, i.e. the measuring cells have a common position-sensitive detector;
a, 12b show a variation of the opto-electronic arrangement according to
a, 13b show the structure of an opto-electronic arrangement according to the invention, which is suitable for measuring six degrees of freedom;
a-14c show the structure of another opto-electronic arrangement according to the invention, which is suitable for measuring six degrees of freedom;
a-16c show a pair of adjacent diaphragms for an opto-electronic arrangement according to the invention;
a shows an opto-electronic arrangement according to the invention, consisting of three pairs of measuring cells which are combined with each other, and have the diaphragms according to
b shows the opto-electronic arrangement according to
Optical Sensor
Sensors for capturing a three-dimensional deflection are built up of optical elements. The arrangement of a LED (light emittent diode), a diaphragm and a PSD (position sensitivity detector) as the measuring cell of a complete sensor has proved itself. In
A LED emits a light cone, which strikes a slotted diaphragm, and the remaining light plane beyond the diaphragm intersects a one-dimensional PSD. The intersection point of the light plane with the PSD can be described by a scalar factor λ, which indicates the signed distance of the intersection point on the PSD from the resting (initial) position. Subsequently, the factor λ is understood as the determined voltage of the PSD. An important property results from the arrangement of the three optical elements into a measuring cell. The measuring cell captures particular movements (X, Y, Z, A, B or C), and cannot simultaneously measure other movements. Thus each individual measuring cell can be seen as the sensor for particular movements. The sum of all captured movements gives the measurement space of the complete sensor.
Parameters of a Measuring Cell
For the precise description of the measuring cell, the positions of the LED, diaphragm and PSD are required. To give the position, in the case of the LED the source of the generated light is used. In the cases of the diaphragm and PSD, the centre of the optical element is used. This is not absolutely necessary, but makes the further calculation clearer and has the effect that the scalar factor in the resting position has the value λ=0. Additionally, the direction of the slot in the diaphragm and the direction of the position-sensitive area of the PSD are required.
In the definition of the parameters, some assumptions apply. The light cone of the LED throws its light on the diaphragm, and the resulting light plane intersects the PSD in the whole working range.
For the later considerations, it is useful to define the viewing direction of the LED. It is given by the LED position and diaphragm position, and by the LED position and PSD position. It is assumed that the three points (LED, IRISpos and PSDpos) are arranged so that they are in a straight line.
The vector of the viewing direction LEDdir is standardised to the length 1. The standardisation to the length 1 also applies to the direction of the slotted diaphragm and the direction of the light-sensitive area of the PSD.
The thickness of the slotted diaphragm and of the position-sensitive area is seen as ideally thin. When the intersection of the light plane with the PSD is idealised, the result is an intersection point, not an intersection plane. The magnitude λ indicates the distance of the intersection point from the resting position. The result is positive values for the magnitude λ if the intersection point moves from the resting position in the direction PSDdir, and negative values for the opposite deflection. Obviously, the magnitude λ can be defined arbitrarily differently, and the resting position does not necessarily have to be in the centre. A different definition affects the calculation/working range of the individual measuring cells, but not the basic function or the arrangement of multiple measuring cells.
In
Later, the distance of the intersection point from the resting position (magnitude λ) is indicated by a voltage U1 . . . 6 of the associated PSD. The greater the amount of the voltage, the greater is the distance of the intersection point from the resting position. The sign of the voltage indicates on which side (PSDdir) of the resting position the intersection point is.
Calculation of the Intersection Point
The measuring cell captures the movement of the three optical elements relative to each other. The value λ is determined. It is assumed that one optical element (LED, diaphragm or PSD) moves, and the other two elements are in fixed positions. The case of two optical elements moving can be transferred to the case of one optical element moving, provided that the movable elements move in the same way (rigidly coupled). The result is three different scenarios:
Captured Movement
The vector Translate indicates the displacement of the movable optical element. The matrix Rotate describes the rotation of the movable optical element about the co-ordinate origin (e.g. with the roll, pitch, yaw angles). In the resting position, the vector Translate is 0 and the matrix Rotate equals the identity matrix.
Calculation of a Translatory Movement
The above equations are further decomposed. The rotational portion is transferred to the translatory portion. A rotational movement can be captured by the measuring cell only because the rotation also causes a displacement, because of a lever.
The rotation is “transferred” to the Translate vector, and then also includes the translatory movement which occurs because of the rotation of the movable portion. This translatory portion can occur only if the movable part is not in the centre of rotation. The actual rotation of the movable part is ignored. The simplification of the portion Rotate*Translate≈Translate is applied.
The relative translatory movement of the movable part of the measuring cell is newly specified, and is thus:
Translate→Rotate·<movableportion>−<movableportion>+Translate
Subject to the condition:
0=IRISdir·(LED×PSDpos−IRISpos×PSDpos+IRISpos×LED)
λ=0 applies to the condition of no deflection (Translation=Rotation=(0 0 0)T). The following simplifications result for the above equations (E=identity matrix):
Changes with No Functional Effect on the Measuring Cell
The above equations describe the structure of a measuring cell quite generally. Because of the geometrical arrangement, it can be seen that parameters in the measuring cell can be changed with no change to the functioning of the measuring cell. Particular changes to one or more parameters of the measuring cell are thus insignificant for the actual function. The result is an additional “margin” for the arrangement of the measuring cell, resulting in a changed geometrical structure, but no effect on the function of the measuring cell.
In
In
In
Rotating the Light Plane About the LEDdir Vector
There are other changes to the arrangement of the measuring cell which affect its function. In these cases, the usual perpendicular or quasi-perpendicular arrangement is abandoned. The result of rotating the diaphragm about the LEDdir vector is that the light plane strikes the PSD only in a perpendicular or quasi-perpendicular direction.
In
should be understood as points of the individual optical elements, and the parameters
are the direction vectors of the measuring cell, with the property |IRISdir|=|PSDdir|=1.
Rotating the Light Plane About the IRISdir Vector
A further change to the measuring cell is achieved by rotating the light plane about the IRISdir vector.
In
Rules for Design of an Optical 3D Sensor
Group Formation
From the individual measuring cells, a complete 3D sensor (pan/zoom 3 degrees of freedom, or with 6 degrees of freedom) is to be built. The basic rule applies, that with N measuring cells at best an N-dimensional sensor can be built. The sensor is always seen in a Cartesian co-ordinate system which corresponds to the right hand rule. The aim of the following group formation is to create rules using which groups of measuring cells (one or more measuring cells) can capture particular degrees of freedom in Cartesian space.
1-Group
With the 1-group, a single measuring cell is arranged so that approximately only one degree of freedom is captured. The measuring cell can actually capture no rotation, which can only be measured if it also causes a displacement (translation because of rotation, “carrousel movement”).
Conversely, if the moved optical element (LED, diaphragm or PSD) is in or near the centre of rotation of the sensor, the measuring cell can only measure a translation.
In the next step, the above 3D sensor (pan/zoom) is changed again. Instead of the LEDs in the centre of rotation, the PSDs are now placed there. Although it would be possible to place three PSDs in the centre of rotation, only a single PSD is used here, but the single PSD is used by all three measuring cells (multiple use). Obviously, this cannot happen simultaneously, because the PSD can detect only one intersection point of a light plane. Three intersection points simultaneously result in an arithmetic mean which cannot usefully be processed further. However, it is possible to interrogate the measuring cells in succession, to switch the LEDs on at staggered times (without overlap), and to determine the intersection points on the PSD in succession.
In the first step, a 1-group is formed. Using it, the movement along a principal axis is determined (here along the X axis).
Movement Vector
In
2-Group
In a 2-group, two measuring cells are combined with each other, so that each measuring cell can capture up to two movements along the axes (X, Y or Z). It must be possible to distinguish the two movements through the combination of the two measuring cells. This can be read off on the basis of the appropriate movement vectors. The movement vectors must not be equal BEW1≠BEW2, or expressed otherwise the volume of the tetrahedron (cross-product) which is stretched out by the movement vectors should be as great as possible (sufficient condition).
|BEW1×BEW2|=MAX>0
For the 2-group, the first measuring cell is combined with another measuring cell. The second measuring cell is attached laterally, so that the light plane strikes the PSD at 45°. It is thus able to capture the up and down movements along the Y axis as well as the movements along the X axis. The two measuring cells together form a 2-group, since each measuring cell can capture up to 2 degrees of freedom and the individual degrees of freedom can be uniquely deduced from the combination of the two captured movements. This relationship can be seen again later in the calibration matrix of the complete sensor (pan/zoom). The requirements for a 2-group do not make it necessary that a measuring cell captures only one movement direction (e.g. here along the X axis). A 2-group would also be given if the measuring cell 1 was arranged as the mirror image of the measuring cell 2. Such a combination is shown in
The third measuring cell must now capture at least the movement along the Z axis. A 1-group could achieve this. However, it can no longer be used here, since the PSD is already positioned along the X axis. A movement in the Z axis can be captured on the PSD only by a light plane which is rotated in the X/Z plane. This results in an arrangement of the third measuring cell in which the LED is displaced (e.g. along the Z axis) and the light plane falls on the PSD as desired through a rotated diaphragm.
Table 11b shows the calibration matrix, which because of the group formation can be very easily interpreted. To determine the movement along the X axis, only the first measuring cell is responsible. To determine this movement, only U1 is required. The voltage U2 (second measuring cell) captures the movement along the X axis in a similar way to the first measuring cell. The difference between the U2 and U1 voltages eliminates the X movement, and only the Y movement is left, and is captured only by the second measuring cell. The third measuring cell actually represents a 3-group, because it can measure translatory movements along all axes. By using the 2-group which is formed with the first two measuring cells, the already known movements along the X and Y axes can be eliminated. The factor for U1 eliminates the movement along the X axis for the first and second measuring cells. Additionally, with the factor for U2, the movement along the Y axis is computationally removed from the third measuring cell. Through the calibration matrix in the third row, only the movement along the Z axis is left, and is measured only by the third measuring cell.
Two further variations are shown in
In
However, a symmetrical arrangement is not absolutely necessary for group formation. Instead, its purpose is to obtain a simpler calibration matrix, and to construct the working range of the complete sensor symmetrically. The third measuring cell forms a further 2-group with the first 2-group (measuring cells 1 and 2), since the measuring cell cannot capture the movement along the Y axis.
The above examples show that numerous arrangements result in a pan/zoom sensor. For the basic functions, whether the diagonally incident light plane is at 45° or a different angle is not decisive. The angle of incidence affects the gained resolution and the working range of the movement to be captured. By placing the light plane diagonally (in two degrees of freedom, rotation about the LEDdir and IRISdir vectors), the measuring cell can also be used for “unfavourable” movements.
In the case of perpendicular or quasi-perpendicular incident light, these additional possibilities cannot be used.
Design of 3D Sensors with 6 Degrees of Freedom
In a similar way to the case of the pan/zoom sensor, a 3D sensor with 6 degrees of freedom is now constructed. The 1-groups are set first. In this example, the diaphragms will be the movable optical element. The diaphragms are positioned on the principal axes to form the 1-groups. In
The diaphragm of the first measuring cell is positioned on the X axis. This measuring cell can therefore capture exclusively movements along the X axis. It is suggested as a partner for a 2-group because the movement along the X axis can be completely calculated from a 2-group. The second measuring cell is positioned similarly. It can measure only the movements along the Z axis. So that the third measuring cell also forms a 1-group, its diaphragm is placed in the co-ordinate origin. It can therefore capture only the movements along the Y axis. With these three measuring cells, only the translatory movements are measured. Once each measuring cell is responsible for exactly one principal axis, it is only necessary to arrange the remaining three measuring cells in such a way that they can capture the rotational degrees of freedom.
The measuring cell 4 captures the rotation about the Z axis (C value) as well as the movement along the Y axis. Similarly, the measuring cell 5 captures the movement along the Y axis and the rotation about the X axis (A value). The remaining rotation about the Y axis is measured by the measuring cell 6, which can also capture the movement along the X axis. The result is the following calibration matrix.
The calibration matrix shows the chosen arrangement very clearly. For instance, the movement along the X axis can be determined only by the first measuring cell (voltage U1), although the measuring cell 6 can also capture the movement along the X axis. Overall, the calibration matrix is very thinly populated. Table 13c shows the calibration matrix with very small values removed.
The errors of the calibration matrix for translation and rotation occur because of the linearisation which is applied there. However, because of the chosen arrangement, the exact model can also be applied very easily.
2-Group
For the next arrangement, 2-groups are formed immediately. The measuring cells in a 2-group are arranged so that two degrees of freedom of a 2-group are captured. In this way, the movable optical element no longer has to be arranged at the origin or along the principal axis.
Because of the lateral displacement of the measuring cell 2 to the measuring cell 1, the second measuring cell can also capture rotations about the X axis (movement A). However, because of the short distance to the axis, this is not particularly pronounced.
Another 2-group now captures two further degrees of freedom. It is positioned similarly to the first 2-group, but fitted rotated by 90°. The second 2-group is shown in
A 2-group which can capture the missing movements (Z and A) could be arranged along the Y axis. This could happen with the same arrangement as in the cases of the first two 2-groups. Since this would complicate the structure, the two remaining degrees of freedom are captured separately. Each measuring cell supplements the previously positioned 2-groups to form a 3-group.
The measuring cell 5 captures the A and Y movements. It thus supplements the first 2-group (measuring cells 1 and 2—Y/C) to form a 3-group. The equivalent happens with measuring cell 6, which captures movements Z and B. The second 2-group (measuring cells 3 and 4—X/B) becomes a 3-group, and can measure the movements X, B and Z.
3-Group
In
Starting from the above arrangement, two measuring cells are now combined. The two LEDs throw the light onto the same PSD. In other words, the PSDs of the two measuring cells are in the same place and have the same orientation. Thus one of the two PSDs is saved. The PSD is usually the most expensive optical element of the measuring cell.
For the calculations, two individual PSDs are still assumed. The arrangement is changed so that an adjacent LED shines on the PSD of the neighbour. So that the two light planes cause one intersection point on the PSD, the two PSDs are rotated. The two PSDs thus have the same orientation, which is rotated at 45° to both light planes. The light planes of the two measuring cells are at right angles to each other. The diaphragm is the movable optical element. It is arranged so that the LED of the partner measuring cell cannot throw its light plane onto the PSD through the wrong slotted diaphragm. The partner slotted diaphragm (“wrong slotted diaphragm”) is arranged so that the diaphragm is arranged in the direction of the partner LED and thus no light incidence is possible. The diaphragm uses the degree of freedom (see “Changes with no functional effect on the measuring cell”) on the one hand to be the correct slotted diaphragm for its own LED, and on the other hand to stand along the direction of the partner LED and thus shade the light. The diaphragm can be extended at the end, to ensure that no external light from a LED falls on the PSD.
The measuring cells 1, 3, 5 and the measuring cells 2, 4, 6 each form a 3-group. The movements X, Z and C are captured using the measuring cells 1, 3, 5. The measuring cells 2, 4, 6 are responsible for the movements Y, A and B.
An identically functioning 3D sensor can be obtained if all PSDs are rotated about the appropriate LEDdir vector with the same angle. The slotted diaphragms must be rotated correspondingly, so that the light planes again fall on the PSDs rotated by 45° (or a similar angle) and form measurable intersection points.
Further Variations for Arranging Measuring Cells
Co-Ordinate Transformation
The individual measuring cells are arranged in a specified Cartesian co-ordinate system. However, the definition of a co-ordinate system is arbitrary. The relationship between two co-ordinate systems is described by a linear co-ordinate transformation. The mapping ensures that the magnitude ratios are unchanged and the relationship of the elements to each other remains the same. Thus for a 3D sensor with 6 degrees of freedom, the co-ordinate system which is used can be arbitrarily defined in space. A 3D sensor can therefore be considered as equivalent if the co-ordinate system which is used can be transferred to a co-ordinate system described here using a linear co-ordinate transformation.
Different Movable Optical Elements
To operate a measuring cell, as well as the fixed optical elements a movable element is also required. In all previous arrangements, it is always assumed that this is of the same type (LED, diaphragm or PSD). Obviously, measuring cells with different movable elements can also be combined with each other. For instance, measuring cells can be arranged with movable diaphragms and movable PSDs. The above rules for arranging 3D sensors remain valid in this case.
Jointly Used Slotted Diaphragm
The movement which can be captured by a measuring cell is described by the movement vector, which is calculated from the product IRISdir×LEDdir. From this it can be seen that with one slotted diaphragm two different movement vectors can be formed, if the directions of the two LEDs are different.
Carrying Signals Via the Springs
It is possible to connect the movable optical element and the two fixed optical elements via wire springs. This connection can also be used for electrical cabling of movable and fixed parts of the sensor. Thus as well as a power supply various control signals can be carried. If the LEDs are the movable optical elements, they can be operated via the springs, for instance in a matrix arrangement.
Movable LEDs to Extend the Working Range
From the equations of “Calculation of a translatory movement”, another interesting property becomes clear, and experience confirms it. In the case of a measuring cell with a movable LED, the working range of the movable optical element can be influenced by the arrangement of the fixed optical elements.
In Equation 1 (LED movable), the distance vector PSD-diaphragm is related to the distance vector LED-diaphragm. If the diaphragm is positioned nearer to the PSD than to the LED, this enlarges the movement range of the LED. In the reverse case, the movement range of the LED is restricted, but the smaller movement range is then more finely resolved.
In Equation 2 (diaphragm movable), the distance vector LED-PSD is related to LED-diaphragm. Since the diaphragm must always be in front of the PSD, the distance LED-PSD is always greater than the distance LED-diaphragm. Therefore, in the case of a movable diaphragm, the result can only be a restriction of the movement range.
In Equation 3 (PSD movable), the distance vector LED-diaphragm is in both the numerator and the denominator. The movement range of the PSD is thus always equal, and corresponds to the maximum extent of the light-sensitive part of the PSD.
3D Sensor with More Than 6 Measuring Cells
To construct a 3D sensor with 6 degrees of freedom, at least 6 measuring cells are necessary. Obviously, more measuring cells than would actually be required can be used. This redundancy of the 3D sensor can be used to increase the precision of the sensor or to keep the sensor in operation even if one or more measuring cells fail. This applies equivalently to a pan/zoom sensor.
Appendix A
Example Calculation
Appendix B
In this embodiment of the invention, which is illustrated in
The sensor structure is therefore suitable only for pan/zoom applications, and not for applications with 6 degrees of freedom (6 DOF). The design aim for a pan/zoom sensor is therefore to relocate the movable element into the centre of rotation.
In this description, if it is said that a measuring cell can capture “generally only” or “exclusively” translatory movements, this means that the measuring cell or sensor can measure exclusively translatory movements, at least in a first approximation. Rotational movements can also have a small influence on the measurement. This part is small and therefore negligible, but nevertheless present. The result of the displacement and rotation of the sensor is that in the sensor the individual measuring cells slightly leave their ideal positions (e.g. the movable element is no longer exactly in the centre of rotation), so that small errors occur.
This situation is handled using the following method:
Method of determining relative movements or relative positions of two objects in an arrangement according to the invention, which can capture translatory and rotational movements or generally translatory movements only, with the steps:
| Number | Date | Country | Kind |
|---|---|---|---|
| 10 2004 051 565.4 | Oct 2004 | DE | national |
