JPS5994136A - Method and apparatus for identifying one position freely selected by user from among many positions on surface of mechanical member - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION BACKGROUND This invention relates to contact rods, contact plates, etc. that determine the input power provided to an electrical device by means of contact points.

Traditionally, the point of contact on a surface has been determined by sensing the capacitance changes of individual capacitors arranged on the surface, ie, Hamfield pickup. Contact plate sensors have also used the reflection of sound waves at contact points on the surface and the interference of criss-crossing infrared beams. Such contact rods or contact plates are susceptible to malfunction due to accidental contact or near contact of the user or his/her clothing, and for this reason it is practical to ensure that the device does not feel any less light contact with the surface. A device that can define a force threshold is highly desirable. This can be said about contact rods and contact plates that respond to mechanical forces, and until now this problem has been solved by keyboards that have thin films that are elastically deformable on top of the contact rods and contact plates.

A keyboard that generates highly reliable and highly accurate input any number of times is difficult to manufacture and requires an extremely large number of interface connections, making it unnecessarily expensive.

[Summary of the invention]

The invention is used to identify a selected position from among multiple positions on the surface of a mechanical member such as a contact frame or a contact plate.

According to the invention, the position is selected by applying a force thereto, a force sensor connected to the member generates an electrical signal representing each component of the applied force, and From the relative amplitudes, the identity of the selected position, ie the position where the force was applied, is calculated.

[Detailed explanation]

In FIG. 1, the mechanical component in question is a contact frame 10 supported on fulcrums 11, 12, which may have contact points as shown on its upper surface. If the contact frame 10 is made of a transparent material, the lower display screen can be seen through it and the position of the contact point can be indicated on the display screen. First, in order to explain the present invention in detail, the contact frame 10 has no weight component below the plane of the paper, and the user selects a distance (3) from the fulcrum 11 and 12 of d and d2 respectively. If a downward force FA is applied to the contact point 13, FA is resolved into two downward force components (not shown) applied to the fulcrum 11.12. The contact frame 10 is kept in equilibrium by an upward reaction force F□, F2 applied to the fulcrum 11.12. That is, the force vector F, the scalar amplitude F2 of F2,
F2 has the following relationship with the scalar amplitude FA of the force vector FA.

F + F2 - FA (1) In order to prevent rotation of the contact frame 10 around the fulcrum 11, the clockwise moment FAd with respect to that point has the following relationship with the counterclockwise moment with respect to that point.

FAd knee F2 (d□10d2) (2) That is, the total clockwise moment and the total counterclockwise moment are equal. Substituting equation (1) into equation (2), we get (F
□10 F2) d work = F2 (d work + d2) (3)
, °, d-work/(d-work+d2)-F2/(F□10F2)
(4) Therefore, the distribution of the position of the contact point 13 to the distance between the fulcrums 11 and (4) 12 can be determined by calculations based on the measured values of the forces that the fulcrums 11 and 12 should apply to the contact frame io.

Equation (5) is obtained.

x = F / (F + F2) (5) In Figure 2, the mechanical components in question have fulcrums 21 at each of the four corners.
．． It is a rectangular contact plate 20 with 22, 23, 24.

If the touch plate 20 is transparent and rests on a display screen showing various touch points for user selection, it is better not to have a fulcrum in the center of the touch plate. First, in order to explain the present invention in detail, let us assume that the contact plate 20 does not have a downward component of weight and the user applies a downward force FA to the selected contact point 25. This force is applied to the fulcrum points 21, 22, 23. .24 (not shown), each of which rebalances the contact plate 20. The sum of the scalar amplitudes F□, F2, F3, and F4 of this upward force is equal to the scalar amplitude of the force applied by the user, thereby preventing translation and movement.

F construction + F2 + F3 + F4 = FA (6) (5) To explain the distance, use the fulcrum 21 as the origin of the device and x
Consider the y coordinate system. Furthermore, the y-axis is from fulcrum 21 to fulcrum 24
The component of the distance extending to and parallel to this shall be measured as a function of the distance between the fulcrums 21.24.

The y-axis extends from the fulcrum 21 to the fulcrum 22, and the distance component parallel to this is measured as a function of the distance between the fulcrums 21 and 22. If the force FA is applied to the point (x, y) on the contact plate 20 at this time, the equilibrium of the y-axis and the moment about the y-axis can be expressed by the following equation.

(F2・1)+(F3・1)−FA−y (7)
(F ・1)+(F ・1)=F−x (8
) 3 4 When the moment of force applied to the four corner fulcrums of the contact plate 20 that is not on the axis of the moment is scalar multiplied by the unit scalar 1 of the arm, the coordinates (X, y) where FA is applied are It can be calculated from the scalar amplitude as follows.

x=(F +F)/(F engineering+F2+F3+F4)
(9) 4 y-(F20F3)/(F-work+F2+F3-14'4)
(10) In Figure 3, the mechanical member in question is the fulcrum 21
．． 22.23° (6) It is a right square prism 30 supported by 24 and here assumed to have no weight. (The problem is that the shape of the casing of an electronic device is a rectangular parallelepiped or a rectangular prism.) The contact point in question is only on the top surface of the prism 30, and the force FA applied by the user is in the lateral direction. If there is no component force of the downward direction, the determination of the coordinates (x, y) of the application point of the user's force FA is completely based on the prism 3.
0 as a (thick) contact plate 20. The problem is that if the force applied by the user to the top surface of the prism 30 has a component force in the front-rear direction, this will generate a moment about the z-axis depending on the height of the prism 30 in two directions perpendicular to the xy plane. be. In equation (7), this moment is considered to be 0, assuming that the thickness of the contact plate 20 can be ignored. Similarly, even if the force applied by the user to the top surface of the prism 300 has a component force in the left-right direction, a moment about the ignored y-axis is generated in equation (8) because the thickness of the contact plate 20 can be ignored.

be understood.

(7) FA - (FA-X・IX) + (FA, -1y) + (F
A, 2) α1) Therefore, equations (6), (7), (
8) can be rewritten in the correct form as follows for a square prism with considerable height in two directions.

F□-20F2□z10F3-Z10F4, =FA-z
O's (F21・1) + (F3−7・1
)-(F'A-7・y)+(FA-7・1) (17
)(F3-2・1)+(F4-z・1)−(FA-2・
x) + (FAx・1)C18).

From these formulas, FA-9, FA-e, Fl, F2,
It is clear that determining F3 and F4 is necessary to calculate the xy coordinates.

FA-7, FA-E are fulcrum points 21.22, 23.24
F knee 2, F2-z-3-z-"4-z only becomes F
l-y and Fl-X・F2-y and F2-x-F3-y”
3-x・F4-y and "4-x" can also be measured by installing a force sensor that measures in the same way (8). Therefore, to obtain the xy coordinates of point 31 where the user applied FA, Using equations (16), (17), and (18), the following two equations based on the fact that the prism 30 is restrained from moving forward, backward, left, and right in parallel can be solved.

F□, 10F2-y1F3-y1F4-y-FA-A
α9) Fl-X 10F2-X+F3-X+
F4-X-FA-X (20) FA
It is also possible to construct a structure in which force sensor 7 and FA-e can be easily sensed by one force sensor each. For example, the fulcrum 21.
22.23% 24 was placed on a stand below the prism 30, the shear force between the stand and the bottom of the prism was measured, and FA-7 and FA were directly measured.
-x can also be determined. This will be described in detail later. By combining and solving equations (16) to (20), the following equation representing the coordinates (X, y) of the user's force application point 31 is obtained.

(9) In Fig. 4, the mechanical member in question is also a right square prism 30, but the contact point in question is not on its top surface as in Fig. 3, but on the right side of the figure. It's on the front. In this case as well, it is assumed that the sensors at the fulcrums 21, 22, 23 and 24 resolve their supporting forces into the x, y, and z directions. Here, the XZ coordinates of the application point 32 of the force FB are determined.

(24) and (25) by assuming constraints on the translation of the prism 30 in the X direction, rotation around the y axis, rotation around the y axis, and translation in the X direction, respectively.

% formula % ) (25) ) (10) F knee
(27) Fl-20F2
-Z1F3-z1F4-z-FB-2 (28) When these equations are solved for the coordinates (X, Z) of point 32, the following equation is obtained.

From the above results, it is possible to determine the coordinates of the contact point on any side of the prism 30, and it is easier to select the origin of the Cartesian coordinate system at one vertex of that side than to select it at the vertex of the opposite side. It is concluded that it is simple.

The contact point array can be placed on any number of sides of the prism 30. Theoretically, by applying force to one surface, the fulcrum 21.22
．． Although it is possible to make the force sensor in 23.24 indistinguishable from the force applied to another surface, in practice the component of force perpendicular to the contact surface is the component of force parallel to that surface (11). Since it will be larger, first F□-E+F2-x+F3
□10F, □ and F□-90F2-y10F3□+F4-y
By comparing this with F□-20F2, 10F3-Z10F4-Z, it is determined which of the three pairs of faces of the prism 30 it touched, and the sign of the maximum value determines the opposite side. You can tell which side of the surface it touched. This procedure can be conveniently simplified if there is not an array of contact points on all sides of the prism 30.

It is contemplated that there may be many other mechanical components involved in implementing the invention. A contact plate supported at only three points can be used, but if there is no horizontal solution (x, y)L of the contact point, unlike in the case of four-point support, it is unclear which shelf the contact plate has contacted. Such problems can usually be solved by using more than one mechanical element to aid in the analysis of the contact points. In other words, in the case of the bookshelf type structure above, it is sufficient to sense the force applied to each shelf separately. It goes without saying that the sensing method of the present invention can be combined with other sensing methods (12) to eliminate uncertainty.

In the above explanation, the contact frame 10, the contact plate 20, and the prismatic column 30
It does not take into account that it has its own weight acting as a weight or that there is some weight that is set or applied to it like a weight for other purposes. The above-described scheme can be considered a linear scheme in which a superimposed force application is possible, ie the force sensor responds linearly or substantially to the force applied to it. The calculation of the contact point described above can therefore be performed on the basis of the differentiation of the force exerted on the sensor during the contact period. Separate the relatively fast-changing pulse-like force that is the source of the component force that cancels out the force applied by the user from the relatively slow-change DC-like component force that cancels out the force applied by the user. , the former component force can be used to calculate the coordinates of the contact point, and the latter component force can be discarded.

In some contact point identification schemes, a force may also be applied in a direction perpendicular to the direction in which the force sensor senses the force so that the force is not sensed. This is possible, for example, when the contact frame or contact plate is placed on a vertical plane (13).

When operating on the derivative of the applied force, the fulcrum of the contact frame or contact plate can be determined by taking the absolute value of the differential force before calculating the contact point, as long as the contact is made in the area between the fulcrums. can touch the side surface, i.e. the opposite surface,
In addition, by preserving the sign of the differential force throughout the calculation, it is also possible to calculate contact points other than the area between the supporting points. For example, it is also possible to detect a "cantilever type" contact position at both ends of a contact frame where both ends protrude from a pair of fulcrums.
This type of calculation is also useful for calculating contact points on a square contact plate supported by only three of the four corners. This shows a method for detecting which part of the user applied force.

A component of this force, opposite the force Fl, F2, is transmitted by a plunger 41° 42 moving through a guide 43 (shown in cross section), supported by a base plate 46 (shown in cross section) and filled with hydraulic fluid. Compress bags 44.45.

In order for the contact frame 40 to be able to rotate against the plunger 41.42 during the application of a vertical force, the plunger 41.
It is necessary that the connection of contact frame 40 to 42 be flexible, such as by a plastic adhesive, a spring, a loose fit, a Teflon ball-and-socket joint, or the like. The pressure within the bags 44, 45 is transmitted via tubes 47, 48 to pressure sensors 51, 52, respectively;
The sensors 51, 52 convert changes in applied force into changes in electrical signals proportional to F□, F2, respectively.

The high-frequency electrical noise components included in the output signals of the pressure sensors 51 and 52 have a sufficiently high roll-off frequency and the force F□
, F2 are smoothed by similar low-pass characteristics of the broadband amplifiers 53, 54, respectively. The output signal of the pressure sensor 51.52 is also applied as the input signal of the narrowband amplifier 55.56. Similar low-pass characteristics of the amplifiers 55 and 56 suppress high-frequency electrical noise components, and the power F□, F
It responds to changes in the value of 2.

The output responses of wideband amplifier 53 and narrowband amplifier 55 are applied to the non-inverting and inverting input terminals of differential input amplifier 57, respectively. This causes the (15) output signal of amplifier 57 to be proportional to the change in force F□ by a scaling factor G, resulting in a common mode rejection performance of amplifier 57 that suppresses the response to the bias value of F□.

The output responses of the wideband amplifiers 53 , 54 are summed in a summing circuit 58 and the output responses of the narrowband amplifiers 55 , 56 are summed in a summing circuit 59 . The sum of these responses is applied to the non-inverting and inverting inputs of differential input amplifier 60, respectively, to produce a force F.
□, yielding an output signal proportional (with a scaling factor G) to the sum of the changes in F2.

A coring amplifier 61, which is a direct-coupled differential input amplifier that cuts off the lower end of a positive value signal and cuts off the upper end of a negative value signal, receives the output signal of the amplifier 60 at its non-inverting input, and this signal is applied to its inverting input. It reacts only when a predetermined noise margin signal is exceeded. The output of this coring amplifier 61 is applied to a peak detector 62.

This detector 62 senses when the signal reaches its maximum inflection point and generates a keying signal, which causes the sample-and-hold circuits 63 and 64 to input ΔF and (F
(16) Δ(F□+F2), that is, a signal proportional to the change in F□+F2 is sampled. Here, ΔF□ and △(F +
F2) is held as it is until sufficient force is again applied to the contact frame 40 so that the coring amplifier 61 receives sufficient input from the differential amplifier 60 and exceeds the noise margin.

The held Δ(F + F2) from the sample-and-hold circuit 64 is applied to the ladder-type resistor voltage divider 65,
Here, Olo, 1△(F□10F2), 0.2△(F
□10F2), 0.3△(F□10F2), 0.4△(F□
10F2), 0.5△ (F□10F2), 0.6△ (F□10F2), 0.7△ (F + F2), 0.8Δ (Fi10F
2), 0.9△ (F engineering + F2) and △ (F□10F2)
generates voltage values that define the area limits for each window voltage comparator in area comparator 66 of . 10 of area comparators 66
The one of the window comparators whose window contains ΔF□ provides a signal to the encoder 67 representing which of the touch points on the touch frame 40 have been touched. Encoder 67 typically converts the output signal of area comparator 66 into a 4-bit binary
Encode into 0 base power. As is known to those skilled in the art, between each valid area of the response of the area comparator 66, the indicated contact (17) is used to reduce false indications that occur when touching the contact frame 40 at a point too close to the edge of the touch point. A protective band can also be provided.

It goes without saying that the contact frames operated according to the invention can be arranged in the form of a keyboard, and such an arrangement is attractive in reducing the number of mechanical parts of the keyboard. Multiplex the sample-and-hold circuits in the circuits associated with the various contact frames to use the same ladder resistor divider and area comparator to create one encoder for each key signal and the board comparator output for the entire keyboard. It is also possible to receive encoded output.

It is attractive to use a sampled data method to calculate the coordinates of the contact points according to the present invention, and this calculation can be performed, for example, on Intel Corp.'s MC8-4.
It is suitable for the capabilities of standard microprocessors such as the 8 series. Using a sampled data method like this,
There is no need to match the amplifier gain and filter response to ensure proportionality of the response to F□ and F2'i or the sum of them and Fl (18). One example of this method is suitable for use with electronic chessboards where, when moving a piece, you press down on the four corners that the piece currently resides on, and then press down on the new four corners to which it is to be moved.

FIG. 6 is an exploded view showing the structure of the chess board and force sensor. An electrically grounded metal plate serving as a common upper plate of four capacitor structures each having a rectangular metallized area 71, 72, 73, 74 provided on the upper surface of the printed circuit board 75 as a lower plate. A chessboard pattern is drawn on the metal plate 70, which is separated from the printed circuit board 75 by helical compression springs 76 at its four corners. spring 76
is held in place by a bolt 77 that passes through a corner aperture in plate 70 and a matching through hole in printed circuit board 75, and is compressed by a lock nut 78 at the end of bolt 77.
The entire printed circuit board 75 is pushed up with a constant force. A vibration damping material (not shown) surrounds the spring to prevent the plate from bouncing. (Actually, it is also possible to surround the periphery of the plate 70 with a spongy material, such as that used for door filling, and replace the spring 76 altogether.) In this way, each (19) square metallization of the square array A region is maintained under each quadrant of a chessboard pattern on a square board 70. Printed circuit board inserts 80.81 (sizes slightly exaggerated in FIG. 6) are connected by metallization to each pair of rectangular metallization areas 71.72.73.74. In addition, a compression spring 76 is crimped to this insertion.
Metallized areas 71, 72 . on the printed circuit board 75 that would ground the board 70 if the spring were of a conductive material.
Electrical contact can be made to two ground metallization areas 82, 83 surrounding the rectangular array of 73, 74. Alternatively or in addition to such a ground plug, a short conductive wire 84, one end of which is soldered to plate 70, can be soldered to one section of the ground metallization 82.83.

Electrically grounded plate 70 and metallized areas 71, 72.
73.74 has a value according to the following well-known formula:

C-ε. A/d (31) where A is the area of the smaller capacitor plate, d is the average distance between approximately parallel plates, and ε. is the dielectric constant of space.

This capacitance can also be formed by reducing the dimensions of the rectangular metallization area (20) so that it lies only under the four corners of the chessboard; Since it increases to tens to hundreds of pF in a common size range, it is extended to the bottom of each quadrant of the chessboard. The rolling blade on the surface of the chess board is decomposed into F
The distances d, d2, d3, and d4 of each associated capacitance C, C2, C3, and C4 are reduced, respectively.

This change in d, △d, △d2, △d3, △d4, is at that moment F, F2 ' F3 ' ``Change in 4 - △F, -
It is considered that it is proportional to ΔF2, -ΔF3, and -ΔF4.

Subtracting each of these relationships, expressed in logarithmic form, yields the following equation that represents or approximates the actual behavior.

△C□/C work=-△d□/d□=ΔF work/F e
(3rd △C2/C2--△d2/d2-ΔF2/F2(
33) △C, /C3=-△d3/d3=ΔF3/F3(
3◇△C4/C4=-△d, /d,=ΔF'47F,
(35) (21) Since it is necessary to compensate for the non-linearity of the relationship between capacitance and force, it may not be possible to actually use a spring whose force is proportional to the amount of compression. It is within the skill of those skilled in the art.

FIG. 7 shows a simple and economical way to measure capacitance at the input of a microprocessor used to calculate the points of contact with the chessboard by a chess piece held by a player. C1, C2, C3, C4 are electrical ground plates provided by plate 70 of FIG. 6 and metallized areas 71, 72 .
73.74 is a capacitor with electrically floating plates given respectively. These capacitors C1, C2, C3,
The floating plate of C4 is the input pin I of the microprocessor μP.
It is directly connected to PI, IF5, IF5, and IF5, and is connected to the positive operating voltage +V via resistors R1, R2, R3, and R4, respectively. Resistors R1, R2°R3, R4 (7) The respective resistance values R, , R2, R3, and R4 have the same magnitude, for example, 10 MΩ. Since the voltage +V is optically isolated from ground potential, capacitors CI, C2, C3°C4 are connected in a fairly short time to n-channel field effect transistors (hereinafter FE).
(referred to as T) Qll, C21, C31, (22) can be charged to a voltage exceeding the threshold voltage (vT) of C41.

Microprocessor μP has each input stage as its human pin IP
I, IF5. IF5. It is of a type that conventionally has an open drain type bidirectional structure following IF5.

The gates 91.92.93.94 of FET Qll, C21, C31, and C41 are made conductive by Norculus at the specified clamping time, and the capacitors 01%C2, 03%C4
The charges are transferred to the bleeder resistors R1, R2, and R3, respectively.
, R4. The breather microprocessor μP during clamping is a source-grounded FET C12, C2.
Count the time required for one of the capacitors C1, C2, C3, C4 associated with the gate of a selected one of C2, C32, C42 to be recharged to a voltage above the dark value voltage of that FET.

F E T That the gate voltage of a selected one of C12, C22, C32, C42 has reached its threshold voltage is sensed by the corresponding one of its drain electrodes 96.97%98.99; It is selectively connected to the counting circuit of the microprocessor μP. Next, Figures 8 to 13
Consider the flow of information through the microprocessor μP to obtain the 4-bit X-coordinate of the binary 4-bit X-coordinate of the point touched (23) on output 100 for the flowchart shown.

FIG. 8 shows the main program flow through the microprocessor μP of FIG. After the start connection 101, an initialization process is performed in which initial values are set in the registers for the fast average, slow average, average, and plus Δ signals.

Next, a force measurement process is performed, in which 4 of the chessboard-shaped contact plates 70
The corner forces are measured. The subroutine for this is the ninth
This is shown in the figure and will be described later.

The next step 104 in FIG. 8 is the calculation of fast and slow averages. In this process, a continuous fast average and a continuous slow average of each sample of the four forces measured are calculated as described below with respect to FIG.

For example, the running fast average of one sample of these forces measured is the last two running averages of that sample, and the running slow average is the last two running averages of that sample. These fast and slow averages are analogous to relatively wideband and relatively narrowband low-pass source (24) waves in continuous signaling in sampled data signaling.

A summation process then follows, with the most recent 4 of the 4 forces measured.
The sums of the four consecutive fast averages are combined to produce a continuous fast sum signal, and the sums of the four most recent consecutive slow averages of the four forces measured are combined to produce a continuous slow sum signal. In the Δ force calculation step 106, a Δ force signal is generated according to the fast sum signal minus the slow sum signal minus the noise margin. In this process, the calculation of the amount by which the difference in all the forces applied to the contact plate 70 exceeds the threshold value is completed. The noise margin signal is a threshold level that determines the degree of coring performed in the calculation of the Δ force signal.

Next, a determination 107 is made as to whether the Δ force is positive or not. If negative, it indicates that there is no information regarding the force applied by the user to the contact plate 70, and the routine loop returns to the force measurement process. If it is affirmative, it indicates that there is a high probability that the user applied force to the contact plate 70, and a positive value is held as Δforce, which is compared with the previously held Δforce value in determination 109 to determine Δ. It is determined whether there is an increase or decrease in force. In the case of an increase determination, the Δ force updated in step 111 is held as the previous value for the next comparison in determination 109 (25), and in step 112 the XY position of the force applied by the user is calculated.

In the case of a decrease determination, the Δforce updated in step 110 is held as the previous value for the next comparison in decision 109, and the routine loop returns to the force measurement step 103.

The calculation of the xy position of the force applied by the user in step 112 is discussed below with respect to FIGS. 11, 12, and 13. Next, the XY position of the point where the user touched the contact plate 70 is displayed or otherwise utilized in step 113, and then the routine loop returns to the force measurement step 113.

Judgment 107 shows that the force applied over a short period of time is significantly greater than the force applied over a long period of time, and is only used when suppressing the pie-scattering caused by the weight of the chessboard-shaped contact plate 70 or the weight of the chess pieces on it. Affirmation is made. Judgment 10
9 can continue to recalculate the xyX coordinates of the force applied by the user until the user begins to remove the force, so that the last calculation in this series is performed with the largest value of the force measured, Errors due to electrical noise are extremely small (26). However, the first calculation of a new touch point is done almost immediately without the user having to wait for the optimal conditions for the calculation.

In some embodiments of the invention, it may be desirable to determine the Δforce, which is a measure of the force applied by the user, from the calculation step 106 and use it in conjunction with the contact point position calculated in the display or other utilization step 113. be.

This may be desirable, for example, if a contact plate 70 other than a chessboard selectively controls the amplitude of a signal supplied to, for example, a loudspeaker array. The process up to calculating the Δ force was explained as being performed in one direction using unsigned operations, but by selecting an appropriately large pie scale and performing these steps using signed operations, it is possible to calculate the force applied by the user. Information about the direction of can also be obtained from the sign of the Δ force. In this case, the process of taking the absolute value comes before the process 108.

FIG. 9 shows capacitors C1, C2, and C in step 103.
3. It is a flowchart showing the outline of the measurement of force by charging C4. A step 122 of selecting the first of the four corners after the inlet connection 121 is performed, followed by a step 123 of discharging the selected capacitor (27). Qll in Figure 7,
If C21, C31, and C41 are n-channel FETs, this process is performed by making their gates all positive to ground. Then the reference clock cycle F=
Step 124 is performed to start charging the capacitor selected at O. After a one clock cycle delay in incrementer step 125, a charge complete determination 126 is made and capacitor C1,
A selected voltage among C2, C3, and C4 is compared with a threshold level to see if this level has been reached. If not, repeat this part of the subroutine to measure the number of lock cycles required to charge the selected capacitor;
After charging the selected capacitor for one more clock cycle, decision 126 is made again. When a positive decision 126 is reached, the current value of F represents the time required to charge the selected capacitor, which depends on the capacitance of that capacitor or the spacing of its plates;
Therefore, the component force applied to one of the plates of the capacitor is accumulated in the process 127 of storing F. Next, a judgment 128 is made to see if all four corners have been completed, and (28) the exit connection 129 is reached only if it is affirmative, but if it is negative, a step 130 is performed to select the next corner, and the subroutine is returned to step 123 to connect the capacitor CI, Draw the value of F representing the force measured for the next selected one of C2, C3, and C4.

FIG. 10 shows a routine that constitutes step 104 of the main program flowchart of FIG. Step 132 is performed starting at one corner after the inlet connection 131. In the guarantee step 133, the slow average of the force measurements for the selected corner is determined by subtracting that value divided by 2 (as obtained by a simple bit place shift) from the previous value of this slow average and calculating the force measurement. Calculated by adding the updated value of

8-bit decomposition was used to measure the F value, but the slow average 2
Perform 16-bit decomposition to accommodate division by 8. In the initial setting step 102 of the main program flow in FIG.
The initial value of the slow average for each force measurement value can also be given by the blade side constant value itself. After calculating the low-speed average for one corner, presence/absence determination 134 of other corners is performed, and here it is determined whether the low-speed average has been calculated for all four corners. If yes, the next corner (29) is taken, step 135 is performed, and the subroutine returns to step 133 to calculate the slow average for the next corner.

If negative, start high-speed average calculation. In this calculation, processes 136, 137, 138, and 139 are respectively processes 132 and 132.
．． Similar to 133.134.135, but with process 137
differs from process 133 in that the average is not a long time (28 samples) average but a short time (2 samples) average. Add the previous value of the fast average and the updated value of the force measurement and divide this sum by two (as can be done with a bit position shift). Judgment 138
When is positive, the slow average and fast average values have been calculated for all four corners and reach the exit connection 140.

FIG. 11 is a flowchart of the routine of the xy position calculation step 112 of the main program flow of FIG. Entry connection 141 to this routine begins with slow and fast averaging for force measurements in each of the four quadrants of contact plate 70. The low speed average for each of the lower left, lower right, upper left, and upper right quadrants of the contact plate 70 shown in FIG. 6 is represented by the acronym SB.
L, SBR, STL.

Denoted by STR, the corresponding fast average is F B L, F
B.R.

(30) Expressed as FTL and FTR. In step 142, a variable R2, which is the sum of the forces on the left half of the contact plate 70, is calculated, and in step 143, the sum of the forces on the right half of the contact plate 70 is calculated. The next location call connection 144 is step 142 of the flowchart of FIG. Variable R2 generated in 14.3
, R3, calls the subroutine of FIG. 12 which calculates the X coordinate of the point at which the user applies force to the contact plate 70. This subroutine extrapolates the sum of the left and right forces on the contact plate 70 from the center of the quadrant capacitor to the left and right edges of the contact plate 70, and actually calculates the maximum of that extrapolation by the sum of the left-most and right-most extrapolations. The rightmost quotient is calculated to obtain the variable R5 corresponding to X. In step 145 of the routine of FIG. 11, R5 is stored as X.

Next, calculate the X coordinate of the user's force application point. Process 146
In step 147, R2 is calculated which is equal to the sum of the subtractions of the forces on the lower half of the contact plate 70, and in step 147, R3 is calculated which is equal to the sum of the subtractions of the forces on the upper half of the contact plate 70. The subroutine of FIG. 12 continues via location call connection 148, and step 149 stores the resulting value of R5 as the X coordinate. Once X, y have been calculated, exit connection 150 leads to (31) step 113 of the main program flow of FIG.

FIG. 12 is a flowchart of a subroutine for calculating the location of an applied force. When positioning connection 151 is called, variable R5 is first set to O in the first step 152 of the subroutine. After that, capacitor 71.
Theoretical support forces at the four corners of the contact plate 70 are calculated equal to the actual support forces resolved into each center of 72%73.74.

Figure 13 shows the idea underlying this calculation.

R2 and R3 are actually measured forces, and R2' and R3' are theoretical forces. The cross section 153 of the contact plate 70 is along its y-axis or y-axis, and its inclination is due to the force applied by the user, which is not shown for simplicity. Cross section 1 of contact plate 70
The distance from the center of 53 to both ends is 2I! is capacitor 71
．． It is twice the distance l to the center of 72.73.74.

Because of the inclination angle of section 153, R2 and R3 are each greater than their average AV by R2-AV and smaller by AV-R3 over a distance l from the center of section 153. The size is AV+2 (R2-AV) and AV, respectively.
-2 (AV-R3) force R2/, R3' cutting (32) In addition to the distance of 21 from the center of the surface, R2' + R3' =
The same slope can be obtained by R2+R3.

Returning to FIG. 12, in step 154 AV = (R20R3/
2 is calculated, and in step 155, the variable A (the difference between R2' and AV in FIG. 13) is calculated as being equal to (2XR2)-AV.

In judgment 156, it is determined whether the calculated A is 0 or positive and valid. If A is negative and invalid, 0 is given to the value obtained in AK process 155 in step 157, and 0 is given in step 158. R
2 correction value (i.e., the value corresponding to R2' in Figure 13)
is calculated as being equal to (A + R2)/2. Process 159, determination 160, process 16 for determining correction value R3
1.162 is the process 15 used to determine the correction value of R2
5, determination 156, and steps 157 and 158.

In step 163, a calculated value count is set. This is X
or the number of precision bits that are the binary part that should represent the X coordinate. In the wear call process 164, the wear connection 171 of the subroutine of FIG. 14 is called and returns R4 to 1 bit. In step 165, the stored R5 is shifted to a higher position by one bit, and in step (33) 166, R4 is added to update R5 to have one bit more precision. In step 167, the count is decreased by 1.

In count determination 168, it is determined whether the new count is 0 or not. If negative, the subroutine is returned to the software call process 164, and if affirmative, R5 is returned to the connection 169.
to be applied.

FIG. 14 is a flowchart of a subroutine that calculates each bit of the binary portion specifying the X or X coordinate of the point to which the user applied force. FIG. 15 shows the geometrical concept underlying the operations used in this calculation. For example, let the X coordinate be the subject of calculation. The combined differential force R on the left side of the contact plate 70
2 is placed at the right end (i.e., 1) of a beam of unit length, and the combined difference force R3 on the right side is placed at the left end (i.e., 0). If we draw the moment triangle, which is well known in elementary mechanics,
The hypotenuse is the X coordinate X□ corresponding to the point where the user applied force
intersect at point 170. Comparing the sizes of R2 and R3, the highest bit of the binary part indicating this coordinate is
It can be seen that when R is 3 or more, it is O, and when it is less than R3, it is 1.

(34) Distance from X origin 0. The highest bit of the IX part is R3
It can be determined by comparing and (R2-R3)/2. This shares vertex 170 and opposite side R3 and (R2
- Two triangles with R3)/2 are similar, and the ratio of the heights from R3 and (R2-R3)/2 to 170 is equal to the ratio of R3 and (R2-R3)/2. It is. The next highest bit of the X coordinate is R2 (R2-R3)
If it is larger than /2, set it as 1. Continuing this calculation, 17
Move the left and right vertical lines that are the base of the triangle with 0 as the vertex, compare the bases on this vertical line, determine the ratio of their heights, and define other decomposition bits in the binary part X□ .

This operation is a crude division operation.

Returning to the flowchart in FIG. 14, wear connection 171
In the next subtraction process, A = R2-R3 is determined. If the sign check 173 of the next result is negative, step 174 is started and A is updated to -A/2, further step 175 sets the remaining R3 equal to A, and step 176 sets R4 to 1. It is determined that the return connection 180 is reached. If it is positive, step 177 is started and A is updated to A/2 (35), then in step 178 the remaining R2 is set equal to A, and in step 179 R4 is determined to be 0 and returned. Connection 180 is reached.

FIG. 16 shows one embodiment of the present invention, in which a rectangular parallelepiped housing 200
The TV receiver with supporting metal wire 201.202.20
3.204 is suspended within the outer box (not shown). Support lines 201 and 202 at the upper front corner of the housing 200 are connected to support rods 205 that run left and right over the opening left on the front of the outer box so that the television display screen on the front of the housing 200 can be seen.
The support wires 203 and 204 in the upper rear surface are connected to strain gauges 206 and 207 supported by a support rod 208 provided at the rear of the housing 200 at the same height as the support rod 205. The support rod 209 is attached to the housing 2 a little further back inside the outer box.
It is installed at a height slightly higher than the bottom of 00, and the other 2
Two strain gauges 211 and 212 are attached, and a roller bearing caster attached to the lower corner of the back surface of the housing 200 is in contact with each front plate.

The strain gauges 206, 207, 211, and 212 are of the resistive bridge type, each with an electronic preamplifier and a microb (36
) Provide electrical input to the processor 215 to determine the location of viewer contact on the faceplate in front of the television display surface on the front of the housing 200. The viewer touches the faceplate when playing a video game displayed on the television receiver and interacts with certain television display computer graphics production operations. The microprocessor 215 is typically housed in a chassis (not shown) for gaming and other support electronics packed into the space below the support rods 205, 208 on the housing 200, and half of its outputs 216 are sent to the housing. It specifies the X coordinate of the contact point of the face plate on the front side of the body 200, and the rest consists of a plurality of bit outputs specifying the X coordinate.

Support lines 201, 202, 203, 204 are several inches long and are capable of bending moment forces of each lower back corner of housing 200 about an axis passing through its front corner caused by a viewer touching the faceplate of a television display surface. It is designed to react only to Strain gauges 211 and 212 also respond only to components of lateral shear forces transmitted perpendicular to their front plates. Therefore, (37) the number of force sensors required for positioning the contact point on the front surface of the rectangular parallelepiped is reduced to '4. This structure also features inexpensive, frictionless support that eliminates the problem of unrestricted movement of the casing.

As far as the determination of the abscissa of the point touched by the viewer on the television display surface faceplate is concerned, the microprocessor 2
15 proceeds in the same way as for positions along the contact frame. Each support line is a strain gauge 211.21 of the component force perpendicular to the display surface.
Since the transmission to 2 is not restricted, the horizontal position is the strain gauge 211.
．． 212 by dividing the response to the difference in force transmitted to one of them by the response to the difference in force transmitted to both.

Also, as far as determining the ordinate of the point at which the viewer touches the television display faceplate is concerned, a certain distance y from the axis of the front upper edge of the housing 200 on the faceplate of the display surface.
The value of the bending moment around that axis due to the force F4 applied below is FAy. This bending moment is equal to this bending moment F6゛Zo in the support line 203.204.
Equilibrium. However, Fo is the sum of the applied forces 2° is the axis passing through the joint between the support line 201.202 and the upper edge of the casing (38) and the support @203.204
and the axis passing through the joint with the upper edge. Since the value FA can be determined by the sum of the output responses of the strain meters 206 and 207, this value is determined by the sum of the output responses of the strain meters 206 and 207. It is found by the sum of the output responses of 202.
Divided by, y is its known dimension Y. It is obtained by the ratio of

The number of force sensors required can also be further reduced by placing one strain gauge midway between strain gauges 206 and 207 to measure Fo. That is, the support wires 203 and 204 are eliminated, and one support wire is replaced with the strain gauge and the upper edge of the casing 200 is connected to the middle of the connection point of the replaced support wires 203 and 204. The problem that arises with this configuration is to ensure that the replaced strain gauges do not respond to bending moments about the vertical axis. Strain meters 206 and 207 in the configuration of Figure 16
The response to such a bending moment is opposite, and is canceled out by the sum when calculating Fo.

Although the problem of contact rods and contact plates has been analyzed above using a beam mechanism with no constraints on the fulcrum, this invention can also be extended to contact rods, contacts (39), and contact plates that can be analyzed using a beam mechanism with constraints on the fulcrum. be able to. One example of this is where the contact plate is the top surface of a plastic television housing, and the top surface is constrained horizontally at the front, back, and sides. This restraint of the upper surface is due to an upward force through the front and rear walls of the casing and a bending moment that keeps the upper surface of the casing horizontal. The force applied by the user can be decomposed into forces and moments in opposite directions, and ordinary mechanical engineers can use the measured values of the strain gauges attached to the lower surface of the top wall of the housing as the basis for calculating the point of application of the forces. The well-known theory of area moment method for restrained beam processing can be applied between the front and back and both sides.

Force sensors used to resolve the force applied by the user and determine the point of contact can also be used in some schemes to determine the direction of application of that force and add further input to the user response system. You can also do that.

In the claims, Paiska can also be 0, and for binary numbers, "parallel supply" includes binary numbers that are supplied during the same calculation cycle even if the binary numbers are not supplied at the same time at a particular point in time. (40) Should be interpreted freely.

[Brief explanation of drawings]

1 and 2 are diagrams useful for understanding the method of calculating the point at which a user's force is applied to a contact rod or contact plate according to the present invention by proceeding from decomposition of the force at the fulcrum of such a mechanical member, 3 and 4 are diagrams useful for understanding the calculation according to the present invention of contact points on the top or side surfaces of a rectangular parallelepiped supported at the four corners of the lower surface, and FIG. 5 is an output display of the contact points along the contact bar. a schematic block diagram showing an apparatus embodying the invention for carrying out the
6 is an exploded view of a chessboard device having means for providing a variable capacitance responsive to a force applied by a user in accordance with the present invention; FIG. 7 is an exploded view of the chessboard device of FIG. 8 to 12 and 14 are flow chart diagrams illustrating the above calculations; FIGS. 13 and 15 Figure 1 is a diagram useful for understanding the process in the above flowchart.
Figure 6 is a perspective view showing how the television receiver is installed inside the outer box in order to activate the force sensor that calculates the point at which the user touches the face plate (41) of the television receiver according to the present invention. It is. 40.70... Mechanical member, 41.42... Fulcrum,
44.51.45.52.53...force sensor, 62~
67.103...Determination means. % Applicant RCSNY Corporation Agent Satoshi Shimizu and 2 others (42) Procedural amendment (Form 2) 1. Indication of the case Patent application No. 153992/1982 2, Title of the invention Utilization from multiple positions on the surface of a mechanical member Method of identifying any one position selected by the person and device address 306 Rockefeller-Frasa, New York, New York 10020, United States of America, Drawing to be corrected: Plane 7, Contents of amendment: Drawings, Fig. 8, Fig. 9, Fig. 10 Figure, Figure 11, Figure 12
The figures and Figure 14 will be replaced with those in the attached sheet (the contents of the drawings showing each flowchart will remain unchanged). Attached documents drawings 8.9.10.11.12.14 1 mail order
Up

Claims (2)

[Claims] (1) A method for identifying any one position selected by a user from among multiple positions on the surface of a mechanical member, the force sensing being connected to the member using a force applied to the selected position. A method comprising the steps of: generating an electrical signal representing each component of the force applied to the device; and calculating the identity of the location of the applied force from the relative amplitude of the electrical signal. . (2) A device for identifying any one position selected by a user from among multiple positions on the surface of a mechanical member, the device being connected to the mechanical member at each of its support points and applying voltage to the selected position. a force sensor that senses each component of the total force and generates an electrical signal representing each component; and a force sensor that senses each component of the total force and generates an electrical signal representing each component, and the user to whom the force is applied substantially in real time based on the amplitude of each of the electrical signals. (1) A device characterized in that it includes means for determining the identity of a selected position.