Digitizer for position sensing

• US 4,928,256 A
• Filed: 03/16/1988
• Issued: 05/22/1990
• Est. Priority Date: 03/16/1988
• Status: Expired due to Fees

• Current Assignee: Calcomp Technology, Inc. (GTCO Calcomp, Incoporated)
• Original Assignee: Ametek Incorporated
• Inventors: Poole, Dan R., Parnell, James A.
• Primary Examiner(s): Gruber, Felix D.
• Application Number:
  US07/168,984
• Time in Patent Office: 797 Days
• Field of Search: 178/18, 178/19, 178/20, 364/559, 364/560, 364/561, 364/562, 364/570
• US Class Current: 700/167
• CPC Class Codes: G06F 3/046 by electromagnetic means

Abstract

Apparatus and method for determining position, tilt and offset of a pointer with respect to a grid of parallel conductors with which it is coupled by obtaining measured responses between each of a group conductors and the pointer and fitting an expression for the response to said measured responses.

Claims

1. A method for determining the positional coordinates of a pointer movable relative to a planar pad having spaced conductors arranged along orthogonal axes thereof, said spaced conductors and pointer being energized to produce responses which form a response curve dependent upon the distance between said pointer and said energized spaced conductors, said method comprising the steps of:

(a) storing information functionally related to the characteristic shape of said response curve;

(b) measuring said responses produced when said pointer is at an unknown positional coordinate relative to said planar pad; and

(c) mathematically fitting said measured responses to a functional expression which represents said characteristics shape of said response curve to determine said positional coordinates of said pointer.

2. The method of claim 1 wherein said mathematical fitting is performed by a least-squares technique.

3. The method of claim 1 where in said functional expression is a polynomial.

4. The method of claim 3 wherein said polynomial is partially determined by measuring said response at a plurality of predetermined positional coordinates to obtain a plurality of calibration values and fitting said plurality of calibration values to said polynomial to derive therefrom values for the coefficients of s aid polynomial which represent said characteristics shape of said response curve.

5. A method for determining the positional coordinates of a pointer movable relative to a planar pad having spaced conductors arranged along orthogonal axes thereof, said spaced conductors and pointer being energized to produce responses which form a response curve dependent upon the distance between said pointer and said energized spaced conductors, said method, for each axis comprising the steps of:

(a) measuring said responses at a plurality of predetermined positional coordinates to obtain a plurality of calibration values;

(b) mathematically fitting said plurality of calibration values to a polynomial to obtain therefrom a plurality of polynomial coefficients for said polynomial which define a polynomial response function representing said characteristic shape of said response curve;

(c) measuring responses produced when said pointer is at an unknown positional coordinate relative to said planar pad; and

(d) mathematically fitting said measured responses to said polynomial response function to determine therefrom said unknown positional coordinate of said pointer.

6. The method of claim 5 wherein said polynomial response function is represented by a series of quadratic expressions, each member of said series representing a portion of said polynomial response function.

7. The method of claim 5 wherein said mathematical fitting is performed by a least-squares technique.

8. The method of claim 5 wherein said polynomial response function, for each axis, as a function of the position x_i of each of said spaced conductors is expressed as a function

9. The method of claim 8 wherein segments of f(x) and g(x) are approximated by a series of parabolas to represent said characteristic shape of said response curve.

10. The method of claim 9 wherein said series of parabolas are mathematically fitted to said plurality of calibration values by a least-squares technique to obtain polynomial coefficients thereof which determine said polynomial response function.

11. The method of claim 8 wherein said measured responses are fitted to said function by a least-squares technique which minimizes the squared differences between said measured response and said function with respect to C₁, C₂, C₃ and C₄ to thereby obtain values of C₁, C₂, C₃ and C₄, determining therefrom the positional coordinate of said pointer by combining the value of C₂ so obtained with a previously determined coordinate of said spaced conductor which is nearest said pointer.

12. The method of claim 9 wherein said measured responses are fitted to a series of parabolas which approximate different segments of said function by minimizing the squared differences between said measured responses and said series of parabolas with respect to C₁, C₂, C₃ and C₄ to obtain values of C₁, C₂, C₃ and C₄ , determining therefrom the positional coordinate of said pointer by combining the values of C₂ so obtained with a previously determined coordinate of said spaced conductor which is nearest said pointer.

13. A method for making an initial coarse determination of the position of a pointer relative to a planar pad having spaced conductors arranged along orthogonal axes thereof, said spaced conductors and pointer being energized to produce responses which form a response curve dependent upon the distance between said pointer and said energized spaced conductors, said method, for each axis, comprising the steps of:

(a) selecting a group of said spaced conductors located at one edge of said grid;

(b) measuring the responses for each conductor of said group;

(c) mathematically fitting the measured responses to a functional expression of the form

space="preserve" listing-type="equation">Y.sub.i =C.sub.1 f(x.sub.i +C.sub.2)+C.sub.4 g(x.sub.i +C.sub.2 )+C.sub.3

which represents the response Y_i for a conductor at coordinate x_i, where f(x) is an amplitude term having a gain coefficient C₁, g(x) is a tilt term having a tilt coefficient C₄, C₂ is the distance along said planar pad from said pointer to the nearest of said spaced conductors, and C₃ represents an offset term, said fitting being performed by a least-squares technique to determine the values of C₁, C₂, C₃ and C₄ which fit said measured responses for said group;

(d) storing the value of C₁ and the location of said group of conductors;

(e) successively selecting next groups of conductors each spatially offset from the previously selected group of conductors by one conductor spacing and determining and storing for each of said next groups the value of C₁ and the location of said group in accordance with the above steps (b), (c) and (d), for each of said successively selected groups of conductors comparing the value of C₁ to the previously stored value of C₁, retaining the greater of the two values of C₁ and the corresponding location of said group of conductors associated with said greater value until all the conductors along said axis have been so selected to determine therefrom which of said group of conductors corresponds to a maximum value of C₁ ; and

(f) determining the coordinate of said conductor nearest said pointer from the location of said group of conductors corresponding to said maximum value of C₁, which coordinate defines the coarse location of said pointer along each of said axes.

14. A method for determining the fine position of a pointer movable relative to a planar pad having sets of spaced conductors arranged along orthogonal axes thereof, said fine position being the distance along said planar pad of said pointer from a conductor previously determined to be nearest said pointer, said spaced conductor and pointer being energized to produce responses which form a response curve dependent upon the distance between said pointer and said energized spaced conductors, said method, for each axis, comprising the steps of:

(a) selecting a group of said spaced conductors including said conductor nearest said pointer;

(b) measuring the responses for each conductor of said group;

(c) mathematically fitting the measured responses to a function of the form

space="preserve" listing-type="equation">Y.sub.i =C.sub.1 f(x.sub.i +C.sub.2)+C.sub.4 g(x.sub.i +C.sub.2)+C.sub.3

which represents the response Y_i for a conductor at coordinate x_i, where f(x) is an amplitude term having a gain coefficient C₁, g(x) is a tilt term having a tilt coefficient C₄, C₂ is said fine position, and C₃ represents an offset term, said fitting being performed by a least-squares technique to determine values of C₁, C₂, C₃ and C₄ which fit said measured responses for said group, said value of C₂ determined therefrom corresponding to the fine position of said pointer along each of said axes.

15. The method in accordance with claim 14, including the further step of combining said fine position with a previously determined coarse position of said pointer corresponding to the coordinate of said conductor nearest said pointer to obtain the positional coordinate of said pointer.

16. The method in accordance with claim 15, wherein the coarse determination of the position of said pointer at successive locations of said pointer is calculated from a knowledge of its previous trajectory and velocity by using a predictor algorithm.

17. The method of claim 16 wherein said predictor algorithm is based on Kalman filter techniques.

18. An apparatus comprising:

(a) a planner pad having spaced conductors arranged along orthogonal axes thereof to form a grid;

(b) a pointer capable of being moved along said planar pad;

(c) energizing means to produce responses dependent upon the distance between said pointer and said energized spaced conductors which form a response curve; and

(d) a control system including a microprocessor having means for controlling said energizing means and means for determining the positional coordinates of said pointer along each axis as it is moved relative to said planar pad, by the steps of;

(i) storing information functionally related to the characteristic shape of said response curve;

(ii) measuring the responses produced when said pointer is at unknown positional coordinate; and

(iii) mathematically fitting said measured response to a functional expression which represents said characteristic shape of said response curve to determine therefrom said unknown positional coordinate of said pointer.

19. The apparatus in accordance with claim 18, wherein said microprocessor performs said mathematical fitting by a least-squares technique.

20. The apparatus in accordance with claim 18, wherein said microprocessor acquires and stores coefficients which express said characteristic shape of said response curve as a polynomial.

21. The apparatus in accordance with claim 20, wherein said microprocessor stores information functionally related to the characteristic shape of said response curve by measuring said responses at a plurality of predetermined positions of said pointer to obtain a plurality of calibration values which are then mathematically fit to said polynomial to derive therefrom values for polynomial coefficients which represent said characteristic shape of said response curve.

22. An apparatus comprising:

(a) a planner pad having spaced conductors arranged along orthogonal axes thereof;

(b) a pointer capable of being moved relative to said planar pad;

(c) energizing means to produce responses which form a response curve dependent upon the distance between said pointer and said spaced conductors; and

(d) a control system including a microprocessor having means for controlling said energizing means, and means for determining positional coordinates of said pointer along each axis as it is moved relative to said planar pad by the steps of;

(i) measuring said responses at a plurality of predetermined positional coordinates to obtain a plurality of calibration values which are stored therein;

(ii) mathematically fitting said plurality of calibration values to a polynomial to obtain therefrom a plurality of polynomial coefficients which are stored therein to define the characteristic shape of said response curve;

(iii) measuring said responses corresponding to an unknown positional coordinate of said pointer along said axis; and

(iv) mathematically fitting said measured responses to said polynomial to obtain therefrom said positional coordinate of said pointer along said axis.

23. The apparatus in accordance with claim 22 wherein said polynomial is represented by a series of quadratic expressions, said microprocessor storing information to mathematically fit each member of said series to a portion of the characteristic shape of said response curve.

24. The apparatus in accordance with claim 22 wherein said microprocessor stores information to perform said mathematical fitting by a least-squares technique.

25. The apparatus in accordance with claim 22 wherein said microprocessor stores information which expresses said characteristic shape of said response curve for each axis as a function of the position x_i of each of said spaced conductors by a function

26. The apparatus in accordance with claim 25 wherein said microprocessor stores information to represent segments of f(x) and g(x) by parabolas.

27. The apparatus in accordance with claim 26 wherein said microprocessor stores information to mathematically fit said series of parabolas to said plurality of calibration values by a least-squares technique to obtain said stored polynomial coefficients which represent the characteristic shape of said response curve.

28. The apparatus in accordance with claim 25 wherein said microprocessor stores information to fit said measured responses to said function by a least-squares technique which minimizes the squared differences between said measured responses and said function with respect to C₁, C₂, C₃ and C₄ to obtain values of C₁, C₂, C₃ and C₄, determining therefrom the positional coordinate of said pointer by combining the value of C₂ so obtained with a previously determined coordinate corresponding to said spaced conductor which is nearest said pointer.

29. The apparatus in accordance with claim 26 wherein said microprocessor stores information to fit said measured responses to a series of parabolas which approximate different segments of said function by minimizing the squared differences between said measured responses and said series of parabolas with respect to C₁, C₂, C₃ and C₄ to obtain values of C₁, C₂, C₃ and C₄, determining therefrom the positional coordinate of said pointer by combining the value of C₂ so obtained with a previously determined coordinate of said spaced conductor which is nearest said pointer.

30. An apparatus comprising

(a) a planner pad having spaced conductors arranged along orthogonal axes thereof;

(b) a pointer capable of being moved relative to said planar pad;

(c) energizing means to produce responses which form a response curve dependent upon the distance between said pointer and said spaced conductors; and

(d) a control system including a microprocessor having means for controlling said energizing means and means for determining an initial coarse position along each axis when said pointer if first placed onto said planar pad by the steps of;

(i) selecting a group of said spaced conductors located at one edge of said grid;

(ii) measuring the responses for each conductor of said group;

(iii) fitting the measured responses to a function of the form

space="preserve" listing-type="equation">Y.sub.i =C.sub.1 f(x.sub.i +C.sub.2)+C.sub.4 g(x.sub.i +C.sub.2)+C.sub.3

representing the response Y_i for a conductor at coordinate x_i, where f(x) is an amplitude term having a gain coefficient C₁, g(x) is a tilt term having a tilt coefficient C₄, C₂ the distance along said planar pad from said pointer to the nearest of said spaced conductors, and C₃ represents an offset term, said fitting being performed by a least-squares technique to determine the values of C₁, C₂, C₃ and C₄ which fit said measured responses for said group;

(iv) storing the value of C₁ and the location of said group of conductors;

(v) successively selecting next groups of conductors each spatially offset from the previously selected group of conductors by one conductor spacing and determining and storing for each of said next groups the value of C₁ and the location of the group in accordance with the above steps (ii) (iii), and (iv) for each of said successively selected groups of conductors comparing the value of C₁ to the previously stored value of C₁, retaining the greater of these two values of C₁ and the corresponding location of said group of conductors associated with said greater value until all the conductors along said axis have been so selected to determine therefrom which of said group of conductors corresponds to a maximum value of C₁ ;

(vi) determining the coordinate of said conductor nearest said pointer from the location of said group of conductors corresponding to said maximum value of C₁, which coordinate defines the coarse location of said pointer.

31. An apparatus comprising:

(a) a planar pad having spaced conductors arranged along orthogonal axes thereof;

(b) a pointer capable of being moved along said planar pad;

(c) energizing means to produce responses which form a response curve dependent upon the distance between said pointer and said spaced conductors; and

(d) a control system including a microprocessor having means for controlling said energizing means and means for determining the distance of said pointer from a conductor previously determined to be nearest said pointer by performing, for each axis, the steps of;

(i) selecting a group of said spaced conductors including said conductor nearest said pointer;

(ii) measuring the responses for each conductor of said group;

(iii) mathematically fitting the measured responses to a function of the form

space="preserve" listing-type="equation">Y.sub.i =C.sub.1 f(x.sub.i +C.sub.2)+C.sub.4 g(x.sub.i +C.sub.2)+C.sub.3

which represents the response Y_i for a conductor at coordinate x_i, where f(x) is an amplitude term for said pointer having a gain coefficient C₁, g(x) is a tilt term having a tilt coefficient C₄, C₂ is the distance from said pointer to said nearest of said spaced conductors and C₃ represents an offset term, said fitting being performed by a least-squares technique to determine values of C₁, C₂, C₃ and C₄ which fit said measured responses for said group, said value of C₂ determined thereby corresponding to the distance of said pointer from said nearest conductor.

32. The apparatus in accordance with claim 31 wherein said microprocessor is further instructed to combine said value of C₂ with a previous coarse determination of the position of said pointer to obtain the positional coordinate of said pointer.

33. The apparatus in accordance with claim 32 wherein said microprocessor stores information to calculate a coarse position of said pointer as it is moved to successive locations from a knowledge of its previous trajectory and velocity by utilizing a predictor algorithm.

34. The apparatus of claim 33 wherein said predictor algorithm is based on Kalman filter techniques.

35. A method for determining the tilt angle of a pointer positioned on and movable over a planar pad having spaced conductors arranged along orthogonal axes thereof, said spaced conductors and pointer being energized to produce responses which form a response curve dependent upon the tilt angle and distance between said pointer and said energized spaced conductors, said method comprising the steps of:

(a) storing information functionally related to the characteristic shape of said response curve;

(b) measuring responses produced when said pointer is at an unknown position and tilt angle relative to said planar pad; and

(c) mathematically fitting said measured responses to a functional expression which represents said characteristic shape of said response curve to determine therefrom said tilt angle and position of said pointer.

36. The method in accordance with claim 35 wherein said functional expression is a polynomial expression determined by measuring said responses at a known tilt angle for a plurality of predetermined positional coordinates to obtain a plurality of calibration values, said plurality of calibration values being mathematically fitted to said polynomial expression to derive therefrom values for the coefficients of said polynomial expression which represent said characteristic shape of said response curve.

37. The method in accordance with claim 35 wherein said functional expression is

38. The method in accordance with claim 37, wherein said measured responses are fitted to said functional expression by minimizing the squared differences between said measured responses and said functional expression with respect to C₁, C₂, C₃ and C₄, determining thereby a value of C₂ which corresponds to said position of said pointer and a value of C₄ which corresponds to the tilt angle at said position of said pointer.

39. A method for determining the positional coordinate and tilt angle of a pointer movable relative to a planar pad having spaced conductors arranged along orthogonal axes thereof, said spaced conductors and pointer being energized to produce responses which form a response curve dependent upon the tilt angle and distance between said pointer and said energized spaced conductors, said method comprising the steps of:

(a) storing information functionally related to the characteristic shape of said response curve;

(b) measuring responses produced when said pointer is at an unknown positional coordinate and tilt angle relative to said planar pad;

(c) mathematically fitting said measured responses to a functional expression which includes parameters representing said positional coordinate and tilt angle to determine therefrom values for said parameters.

40. An apparatus comprising:

(a) a planar pad having spaced conductors arranged along orthogonal axes thereof;

(b) a pointer movable relative to said planar pad;

(c) energizing means to produce responses which form a response curve dependent upon the tilt angle and distance between said pointer and said spaced conductors; and

(d) a control system including a microprocessor having means for controlling said energizing means, and means for determining said positional coordinate and tilt angle of said pointer, for each axis as it is moved along said planar pad, by the steps of;

(i) storing information functionally related to the characteristic shape of said response curve;

(ii) measuring responses produced when said pointer is at an unknown positional coordinate and tilt angle relative to said planar pad;

(iii) mathematically fitting said measured responses to a functional expression which includes parameters representing said positional coordinate and tilt angle to determine therefrom values for said parameters.