WO2000005593A1

WO2000005593A1 - Variable dielectric position transducer and method

Info

Publication number: WO2000005593A1
Application number: PCT/US1999/016744
Authority: WO
Inventors: Philip T. Dempster
Original assignee: Life Measurement Instruments Inc
Current assignee: Life Measurement Instruments Inc
Priority date: 1998-07-24
Filing date: 1999-07-23
Publication date: 2000-02-03
Anticipated expiration: 2001-01-24
Also published as: AU5126599A

Abstract

A variable dielectric transducer includes first and second electrically conductive plates (A, B) separated from a third electrically conductive plate (D) wherein the first plate (A) has a first capacitance (CA) relative to the third plate (D) and the second plate (B) has a second capacitive (CB) relative to the third plate (D); a fourth movable nonconductive plate (P) having a dielectric constant greater than unity interposed between the first, second and third plates (A, B, D) and covering a portion of the first and second plates (A, B); means for moving the fourth plate (P) between the first, second and third plates (A, B, C) wherein movement of the fourth plate (P) produces complementary changes in the first capacitance (CA) and the second capacitance (CB); and means for generating a transducer signal representative of the changed first and second capacitances by the use of nulling technique.

Description

VARIABLE DIELECTRIC POSITION TRANSDUCER AND METHOD

BACKGROUND OF THE INVENTION

There are a variety of applications for an accurate, contactless, and small position transducer. Most of these fall into the categories of control and monitoring. In the former, the position transducer is normally used as the feedback component of a servo system. An example would be a mirror positioning system that utilizes a linear motor such as a linear solenoid or voice coil system. Another example would be control of a butterfly type air valve, in which application a rotary transducer is used to measure angular position. An example of a monitoring system is liquid level detection using a float in conjunction with a linear or rotary transducer. Another example is a two-axis joystick. In this case, the transducer is configured to have two outputs to describe coordinate position. It would be desirable to provide a position transducer and method that combines high accuracy, low moving mass, virtually unlimited life, and very low cost. It also would be desirable that it be adaptable to sensing along two independent axes. Such a transducer would make possible many new applications not hitherto practical, due to high cost or performance limitations of existing transducers.

SUMMARY OF THE INVENTION

It is an object of the present invention to provide an improved position transducer, and more particularly a variable dielectric transducer and method capable of detecting a variety of movements. It is another object of the invention to provide an improved linear transducer. It is another object of the invention to provide an improved rotary transducer. It is yet another object of the invention to provide a transducer of two outputs representing coordinate position in a plane or other two- dimensional surface.

In two preferred embodiments, the position transducer includes first and second electrically conductive plates separated from a third electrically conductive plate. The first plate has a first capacitance relative to the third plate and the second plate has a second capacitance relative to the third plate. The position transducer includes a fourth moveable nonconductive plate, having a dielectric constant greater than unity, that is interposed between the first, second and third plates and that covers a portion of the first and second plates.

In one embodiment, the position transducer includes means for translating or rotating the fourth nonconductive plate between the first, second and third conductive plates, so that movement of the fourth plate produces approximately complementary changes in the first capacitance and the second capacitance, and means for generating a transducer signal representative of the changed first and second capacitances by means of a nulling technique.

In another embodiment, the position transducer includes an array of four fixed conductive plates separated from a fifth fixed conductive plate. Each of the four conductive plates has a capacitance relative to the fifth plate, which are denoted as Cl, C2, C3, and C4, respectively. The position transducer includes a sixth moveable nonconductive plate interposed between the group of four plates and the fifth plate.

This embodiment includes a means for moving the sixth nonconductive plate between the group of four conductive plates and the fifth conductive plate along each of two axes, so that with movement along one axis, the sum (Cl + C2) varies in approximately complementary fashion to the sum (C3 + C4), and with movement along the other axis, the sum (Cl + C3) varies in approximately complementary fashion to the sum (C2 + C4). Other objects, features and advantages of the present invention will become apparent from the following detailed description when taken in conjunction with the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention:

FIG. 1 shows a schematic representation of an arrangement of fixed conductive plates and moveable nonconductive plate according to the present invention.

FIG. 2 A shows a circuit including the variable capacitances formed between pairs of conductive plates shown in FIG. 1.

FIG. 2B shows some representative waveforms at various points of the circuit shown in FIG. 2A with movement of the nonconductive plate.

FIG. 3 shows a schematic representation of the arrangement of the fixed conductive plates and moveable nonconductive plate in a transducer suitable for two axis sensing.

FIG. 4 shows a circuit including the variable capacitances formed between pairs of conductive plates shown in FIG. 3.

FIG. 5 shows some representative waveforms at various points in the circuit of FIG. 4.

FIGS. 6A, 6B, 6C, and 6D show some possible arrangements of the fixed conductive plates and moveable nonconductive plate shown in FIG. 1 that are suitable for linear position sensing.

FIGS. 7A, 7B, 1C, and 7D show some possible arrangements of the fixed conductive plates and moveable nonconductive plate shown in FIG. 1 that are suitable for rotary position sensing.

FIG. 8 shows another arrangement of the fixed conductive plates and moveable nonconductive plate according to the present invention.

FIG. 9 shows a circuit including the variable and fixed capacitances formed between pairs of conductive plates shown in FIG. 8.

FIGS. 10A, 10B, and IOC show some possible arrangements of the fixed conductive plates and moveable nonconductive plate of FIG. 8 that are suitable for linear position sensing.

FIG. 11 shows some of the possible configurations of fixed conductive plates and moveable nonconductive plate of FIG. 3 that are suitable for sensing position in a plane.

FIG. 12A shows an exploded view of some of the components of the present invention as configured for linear sensing.

FIG. 12B shows a side view of the implementation of FIG. 12A.

FIGS. 13A and 13B show an implementation for rotary sensing with the present invention.

FIGS. 14A and 14B illustrates the exceptional linearity and repeatability of a linear transducer example according to the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

Reference will now be made in detail to the preferred embodiments of the invention, examples of which are illustrated in the accompanying drawings. While the invention will be described in conjunction with the preferred embodiments, it will be understood that they are not intended to limit the invention to those embodiments. On the contrary, the invention is intended to cover alternatives, modifications and equivalents, which may be included within the spirit and scope of the invention as defined by the appended claims.

The present invention uses the principle that the capacitance between two conductive plates which are separated by an air gap is increased if the air in between is replaced with a nonconductive material having a dielectric constant greater than unity. By definition, if the space between the two plates is entirely filled, the proportion by which capacitance increases is the dielectric constant of the nonconductive material.

A position transducer according to the present invention includes at least two separated conductive plates having a sheet of nonconductive material that may partially fill the space between those two plates. If the sheet of nonconductive material is moved so that more of the space between two separated conductive plates is filled, the capacitance between the two conductive plates will increase. The position transducer provides a transducer signal or signals representative of the changed capacitance between one or several pairs of plates.

Although a transducer that simply consisted of pairs of conductive plates with a moveable sheet of nonconductive material in between would be practical, it would also suffer from various problems, chief among them, dependency on temperature, critical dependency on dimensions, and susceptibility to stray capacitance . It would be desirable to configure a transducer so that both baseline and span were inherently independent of temperature. In addition, it would be desirable to configure the transducer so that its calibration were independent of the dielectric constant of the moveable sheet of nonconductive material between the plates. It would also be desirable to configure the transducer so that close tolerances were not required for good performance. It would also be desirable that the transducer be insensitive to stray capacity.

Several preferred embodiments will be described. A first embodiment is a symmetric configuration with an inherently stable midscale output, and cancellation of some non-linearities, but whose span is dependent on spacing and dielectric constant. The second main embodiment is a refinement of the first that addresses dependency on both spacing and dielectric constant of the moveable plate, for more critical applications. The third main embodiment senses position along two axes. This embodiment may also include the refinements of the second embodiment, and thus reduce dependency on spacing and dielectric constant of the moveable plate.

The three embodiments may be further subdivided into multiple embodiments, as will be seen. For example, the first and second embodiments may be configured for either linear or rotary movement. Particular examples involving the use of printed circuit boards will be shown.

First Embodiment (Single Axis)

With reference to FIG. 1, electrically conductive plates 101 and 102 are separated from a third electrically conductive plate 103. Moveable nonconductive plate 104, which has a dielectric constant greater than unity, is interposed so that it fills part of the space between plates 101 and 103, and part of the space between plates 102 and 103. The capacitance of plate 101 relative to plate 103 is denoted by

C101, and the capacitance of plate 102 relative to plate 103 by C102. As plate 104 is moved so that the space between plates 102 and 103 is more completely filled, and the space between plates 101 and 103 is less completely filled, capacitance C102 increases and C101 decreases. By appropriate choice of plate geometry, the variation of capacitances C101 and C102 may be a linear function of the movement of plate 104, and the sum (C101 + C102) may remain approximately constant. By different choice of geometry, other (non-linear) functions relating capacitance and movement may be obtained as desired by the application designer.

FIG. 2A shows a circuit that includes the various plates shown in FIG. 1. Clock circuit 112 (not detailed) generates squarewave E_clk which operates double pole electronic switches 108, 109, and 110. The frequency of the squarewave could be 5 kHz, but other frequencies or asymmetric waveforms would also work. Reference voltage V_r, which may be a power supply voltage or generated in some other manner, is connected to one pole of switch 108 while the other pole of switch 108 is connected to the output voltage E₀ of integrator 111. Similarly, switch 109 has one pole connected to an electrically neutral point and the other to E₀. It can be seen that the action of switch 108 produces an excitation squarewave E^, at point B in FIG. 2A, whose amplitude is given by E,, = V_r - E₀, and the action of switch 109 produces an excitation squarewave E_a at point A of FIG. 2A, whose amplitude is given by E_a = E₀. If 0 < E₀ < V^ the two squarewaves will be 180 degrees out of phase.

Still looking at FIG. 2A, conductive plate 103 is connected to the input of high impedance buffer amplifier 107. Neglecting stray capacitance and other sources of leakage currents, charge conservation requires that the voltage on plate 103 be a squarewave such that AC currents through capacitance C101 of plate 101 to plate 103, and capacitance C102 of plate 102 to plate 103, are equal in magnitude and opposite in sign. For any given combination of C 101 and C102, there is a unique voltage y, such that if E₀ = y, the AC voltage on plate 103 will be zero. Conversely, if E₀ is not equal to that unique voltage y, the AC voltage on plate 103 will be non-zero. Any AC voltage on plate 103 is amplified in its power by buffer amplifier 107, whose output drives capacitor 140. The other side of capacitor 140 is alternately grounded and connected to the input of integrator 111 by switch 110, synchronous with the AC voltage on plate 103, because all three switches 108, 109, and 110 are driven by E_dk. The effect is that capacitor 140 will deliver a charge to integrator 1 11 during each cycle that is equal to the amplitude of the voltage at the output of buffer 107 times the value of capacitor 140, and that the sign of this charge will change the output E₀ of integrator 111 in a direction tending to drive E₀ towards the voltage y which would result in no AC voltage on plate 103. If plate 104 remains in a single position, the difference between ideal voltage y and E₀ will become arbitrarily small. In a practical example, step response to change in position of plate 104 may be about one millisecond. However the transducer may be designed to respond faster or slower by choice of component values and clock frequency.

It may be seen that if E₀ = y, then E₀ must satisfy:

(1) C102 * (V_r - E_o) = C101 * E_o

Whence

(2) E_o = ((C102 / (C101 + C102)) * V_r

If (C101 + C102) is constant, and if C 102 varies linearly with the position of plate 104, E₀ will be a linear function of position and reference voltage V_r. Some waveforms of the circuit are illustrated in FIG. 2B. As the position of dielectric plate 103 is moved, the excitation voltage E^, decreases, and the excitation voltage E_a increases. The error voltage E_π shown in FIG. 2B results in the producing of a transducer signal E₀ which is representative of the changed capacitances C101 and C102. Two critically important features of the circuit described above should be particularly noted: First, in the case that a linear transducer is being implemented, significant nonlinearities are cancelled due to the complementary action of C101 and C102. If, for example, the alignment between plates 101 through 103 is non-parallel due to production tolerance, the resulting non-linearity will be reduced. To illustrate: Keeping in mind equation (2) and referring to FIG. 1, if the distance between plates

102 and 103 is less than the distance between plates 101 and 103, and plate 104 is translating or rotating in the direction of plate 102, the incremental change in output due to increase in C 102 becomes larger, but at the same time denominator (C101 + C102) also increases, tending to reduce the error. In addition, all nonlinearities expressable as even harmonic distortion of output versus input movement, will tend to be canceled by the symmetric geometry of plates 101 and 102.

Second, operation depends on a null condition or signal at the input of buffer amplifier 107. The capacitances of plates 101 and 102 with respect to plate

103 may be well under 1 picofarad. The null condition in terms of the capacitances and signals means that the signals applied to conductive plate means are such that, given the capacitances associated with each plate of the plate means, the voltage capacitively induced on the reference plate is zero. In other words, when the reference plate is at 0 volts (null), the currents through the two capacitors are equal in magnitude and opposite in sign, cancelling each other out. The effect is that as capacity diminishes due to movement of the non-conductive plate, its associated voltage increases and visa versa.

If operation depended on measurement of non-null output at plate 103, stray capacitance would act as a shunt, significantly affecting magnitude of output. As plate 103 is at a null when correct output is achieved, stray capacitance to neutral structures has little effect, so accuracy is not affected.

Second Embodiment (single axis)

The first embodiment, although adequate for many applications, may have the following deficiency: span of output is dependent on spacing of the conductive plates, as well as on the thickness and dielectric constant of plate 104. This creates the potential for unwanted temperature dependencies and variable performance due to manufacturing tolerances. A method to reduce these dependencies will now be described.

Referring to FIG. 9, these dependencies can be largely cancelled by the addition of summing inverting amplifier 115 and capacitor 1 16. It should be noted that in FIG. 9, capacitors 105 and 106 refer to the capacitance of plate 101 to plate 103 and of plate 102 to plate 103, respectively, as shown in FIG. 8. Resistances 113, 114, 122, and capacitor 116 may be chosen so that, in the absence of moveable nonconductive plate 104, the current through capacitor 116 would exactly null the current through ClOl or C102. An example of appropriate values is: resistance 113 = resistance 1 14 = resistance 122, and capacitor 116 = the capacitance due to air gap of ClOl and C102. There are, of course, an infinite number of combinations of values that would fulfill the requirements. Because the effects of capacitances ClOl and C102 in their air filled state are effectively canceled, when the space between plates 101 and 103 and the space between plates 102 and 103 are partially filled by plate 104, the effective capacitance of ClOl and C102 is that part of the capacitance that is due to the presence of nonconductive plate 104.

To understand why cancellation of the effects of air capacitance decreases dependence on spacing of plates 101, 102 and 103, and dielectric constant of plate 104, imagine that ClOl and C102 had value zero in the absence of plate 104, and some positive value C with plate 104 interposed. Dependent on position of plate 104, 0< ClOl < C_, 0 < C102 < C_x, and ClOl + C102 = C_x. Inspection of equation (2) indicates that output E₀ varies between 0 and V_r, as C 102 varies between 0 and C_x. This remains true no matter what the dielectric constant of plate 104 or the spacing of the plates.

Capacitor 1 16 may be constructed to be part of the structure of ClOl and C102. Referring now to FIG. 8, plate 131 is aligned opposite plate 104, as are plates 101 and 102. Plates 101, 102, and 131 might be mounted on the same insulating structure, for example, epoxy fiberglass board. As such, variations in spacing due to temperature or manufacturing tolerance will affect all three capacitances equally, maintaining cancellation of air capacitance. A disadvantage is that the transducer becomes larger. Where size is critical, capacitor 1 16 may be a standard component, or it may be created on a circuit board by juxtaposition of conductive areas, usually on opposite sides of a circuit board.

The point in the circuit represented by plate 103 and the input of buffer amplifier 107 is a very high impedance point in the circuit. It is desirable to carefully shield this part of the circuit from external fields, yet at the same time keep stray capacity to ground to a low value. Referring to FIGS. 12A-B and 13A-B for physical examples and to FIG. 9 for the circuitry, guard 132 may be placed behind plate 103 with guard 132 driven by the output of buffer amplifier 107. In FIG. 9 this connection is indicated by an arrow pointing to the capital letter 'G'. Although the capacitance between plate 103 and guard 132 may be large in comparison with ClOl, C102, and capacitor 116, that capacitance has virtually no effect, because buffer amplifier 107 maintains this plate at the same potential as plate 103. Although guard 132 is not grounded, it is not significantly affected by external fields, because the output of buffer amplifier 107 is of low impedance.

Although the above circuits utilize squarewaves driving plates 101 and 102, and a particular form of synchronous detection (switch 110, capacitor 140, integrator 1 11), it would be appreciated by one skilled in the art that sine waves or other waveforms, as well as detectors of different design, could be substituted without altering the fundamental operation of the circuit. The essential operation of the circuit is that the detector output alters the voltages applied to plates 101 and 102 in accordance with equations (1) and (2) in an arrangement that drives the voltage on plate 103 towards a null condition.

As an example of a variation within the spirit of the invention, a separate amplifier and capacitor could be used to null the air capacitance ClOl and C102. As another example, it would be clear to replace the ground of switch 109 with some other voltage, for example with (-V_r). In this modification, the output voltage in the center position of plate 104 would be zero. Buffer amplifier 107 is shown as an operational amplifier. It could be some other type of device, for example a source follower FET. Other integrators besides the simple circuits shown in FIGS.

2A and 9 could be used. In fact, integrator 111 could be replaced with a high gain amplifier that had a 6db/octave roll off at high frequencies, with no significant degradation of performance.

Mechanically, it will be apparent to one skilled in the art that the first two embodiments may be configured to respond to a variety of movements. The most common types of movement are linear translation and rotation, but it is not intended that the invention be restricted to these two movements. An example of a movement other than linear or rotary is a trajectory constrained to the surface of a sphere. Because linear and rotary applications will be the most common, several embodiments of these two types of movement will now be described.

Turning to the linear case, FIGS. 6A, B, C, and D show four arrangements of nonconductive plate 104 for the first preferred embodiment. FIGS. 10A, B, and C show three arrangements of nonconductive plate 104 for the second preferred embodiment. Note that these figures do not show conductive plate 103, since to do so would obscure the relationship of plates 101, 102, and 104. These arrangements all have in common that as plate 101 is uncovered by plate 104, plate 102 is correspondingly covered and vice versa. One skilled in the art would be able to devise other variations within the spirit of the invention.

FIGS. 7 A, B, C, and D show four arrangements for rotary sensing for the first embodiment. As in the case of FIGS. 6A-D and 10A-C, and for the same reason, plate 103 is not shown.

In FIG. 7A, plates 101 and 102 are symmetric half disks and nonconductive plate 104 is a half disk that covers a portion of each of plates 101 and 102. With continuous rotation, the output of such a transducer is a symmetric sawtooth waveform that repeats every 360 degrees. There are two linear regions of 180 degrees each. In most applications, movement would be restricted to one of the two linear regions.

In FIG. 7B, plates 101 and 102 are nesting spirals and nonconductive plate 104 is a section of a disk. This configuration may have geometry allowing linear response to rotation greater than 180 degrees.

FIG. 7C is a variation in which plate 104 is a spiral and plates 101 and 102 are pie-shaped sections. Like FIG. 7B, this configuration may have geometry allowing linear response to rotation over more than 180 degrees.

FIG. 7D is similar to FIG. 7A except that plates 101 and 102 have been shaped so as to produce a sinusoidal output with continuous rotation. This is just one example in which geometry of the plates may be designed to create a specific functional relationship between motion and transducer output.

Adaptation of the arrangements of FIGS. 7A-D to the second configuration is not shown in the figures. Many ways to add a compensating plate 131 to those arrangements will occur to those skilled in the art. For example, plate 131 could be a ring-shaped flat conductor surrounding plates 101 and 102. Often, in order to conserve space in the rotary case, capacitance 116 would be realized by a component or by juxtaposition of conductive areas on a printed circuit board.

A preferred linear configuration of either the first or second embodiments is an assembly comprising two printed circuit boards. These boards, separated by simple spacers and three or four electrical connections that may be the spacers themselves, contain the entire transducer, including the circuitry. For example, the second embodiment will be illustrated. With reference to FIGS. 12A and 12B, circuit board 124 may be a one or two layer board in which plates 101, 102, and 131 are etched onto layer 120 by standard printed circuit board techniques.

Circuit board 125 is a multi-layer board of typically four circuit layers, indicated as structures 127, 117, 118, and 119 in FIG. 12B. Layer 127 has plate

103 etched onto it by standard techniques. Layer 117 includes guard 132. Extra space is dedicated to ground for shielding. Layers 118 and 119 are both used for circuit traces, and layer 119 also carries the electronic components.

The bottom and top circuit boards are fixed to each other by standoffs 126, so that clearance exists to insert nonconductive plate 104. In this example, electrical connections are integrated with standoffs 126.

A variation would use three circuit boards or two circuit boards and an ASIC (application specific integrated circuit). The bottom circuit board would be as described. The middle circuit board would contain plate 103 and guard 132. The top circuit board would contain the electronics circuit. Alternatively, the top board would be replaced by an ASIC and the middle board would have a few traces and components. These boards could be mounted in a sandwich arrangement as in the above configuration, using one of a variety of electrical interconnections.

A configuration of the first embodiment for the rotary case is shown in FIGS. 13A and 13B. The arrangement is very similar to the linear case. The rotary configuration shown is that of FIG. 7A.

As with the linear case, the bottom board 124 contains plates 101 and 102, whereas the top board 125 contains plate 103, the guard, and the circuitry. Moveable plate 104 is mounted on a shaft 128 that passes through bushing 121 mounted to lower board 124. The bottom and top circuit boards are fixed to each other by standoffs (not shown), so that clearance exists to insert nonconductive plate 104.

Third Embodiment (Two Axis)

With reference to FIG. 3, imagine four conductive plates 1, 2, 3, 4 in a rectangular array that is fixedly arranged in a plane, an additional fixed conductive plate 5 arranged in a parallel plane behind plates 1 through 4 with a space in between, and a moveable nonconductive plate 6 that is constrained to move in a plane parallel to and between the other two planes. Further imagine that nonconductive plate 6 partially but not completely covers plates 1 through 4, and that it has a dielectric constant greater than unity.

Because of their proximity, each conductive plate 1 through 4 has a capacitance relative to plate 5, which are denoted Cl through C4, respectively. The more of any of plates 1 through 4 is covered by moveable plate 6, the larger will be its associated capacitance, since the dielectric constant of plate 6 is greater than unity.

By inspection of FIG. 3 it may be noted that as plate 6 moves directly upwards, sum (Cl + C2) increases whereas sum (C3 + C4) decreases. It is also true that sums (Cl + C3) and (C2 + C4) remain approximately constant. Similarly, for movement directly to the right, sum (C2 + C4) increases, sum (Cl + C3) decreases, and sums (Cl + C2) and (C3 + C4) remain approximately constant.

Now imagine a voltage Y proportional to ((Cl + C2) - (C3 + C4)) / (Cl + C2 + C3 + C4), and a voltage X proportional to ((C2 + C4) - (Cl + C3)) / (Cl + C2 + C3 + C4). Voltage X would be an analog of horizontal position and voltage Y an analog of vertical position.

A method of generating such voltages X and Y will be described. The method is stable, simple, and uses no transformers or inductors. The method is based on the fact that it is possible to apply four non-zero AC voltages to each of plates 1 through 4, with phase and amplitude such that the voltage on plate 5 is null.

Looking now at the schematic in FIG. 4 and the waveforms in FIG. 5, gates

21, 24 and flip flops 22, 23 create quadrature squarewaves A, B; i.e., A and B are 90 degrees out-of-phase. Squarewave A controls S.P.D.T. (Single Pole Double Throw) analog switches 12, 13 so that at the common point of switch 12 there is a squarewave whose amplitude is equal to reference voltage V+ minus output voltage X, and at the common point of switch 13 there is a squarewave whose amplitude is equal to output voltage X minus reference voltage V-. Note that these two squarewaves are 180 degrees out of phase. The squarewave of switch 12 is inverted by amplifiers 8, 11 and applied to plates 2, 4. Similarly, the squarewave of switch 13 is inverted by amplifiers 9, 10 and applied to plates 1, 3.

In the same fashion, squarewave B controls S.P.D.T. analog switches 14, 15 so that at the common point of switch 14 there is a squarewave whose amplitude is equal to reference voltage V+ minus output voltage Y, and at the common point of switch 15 there is a squarewave whose amplitude is equal to output voltage Y minus reference voltage V-, and these two squarewaves are 180 degrees out of phase. The squarewave of switch 14 is inverted by amplifiers 8, 9 and applied to plates 1 and 2. The squarewave of switch 15 is inverted by amplifiers 10, 11 and applied to plates 3 and 4.

It should be noted that for each squarewave A and B, two adjacent plates are driven in one phase and the opposite two plates are driven 180 degrees out of phase.

It will be seen that each plate is driven by a composite waveform that is the sum of two squarewaves in quadrature phase relationship. Representations of these waveforms are to be found at the bottom of FIG. 5.

These waveforms are cyclical around four transitions. The Roman numerals at the top of FIG. 5 represent the four states marked by these transitions. It will be noted that the magnitude of the transition on plates 1 through 4 that occurs between states I and II is dependent on the outputs of switches 14, 15. The magnitudes of the transitions on plates 1 through 4 that occur between states IV and I is dependent on the outputs of switches 12, 13.

Still looking at FIG. 5 and the four waveforms on plates 1 through 4, it can be seen that at each transition, there are two positive and two negative shifts in voltage. The relative size of each shift changes if one of output voltages X or Y changes, depending on which transition is being examined. Because all of plates 1 through 4 are capacitively coupled to plate 5, the negative and positive shifts tend to cancel in terms of resultant voltage on plate 5. In fact, for a given position of dielectric plate 6, there is a unique voltage X which creates a null voltage change between states IV and I on plate 5. This is because a change in voltage X increases the amplitude of one of the squarewaves and decreases the amplitude of the other, and these two squarewaves are 180 degrees out of phase. Similarly, there is a unique voltage Y that creates a null voltage change between states I and II on plate 5.

Looking first at the transition between states IV and I: The change in voltage on plate 5 may be considered to represent an error in output voltage X, since there is a unique voltage for X at which the change will be zero, and we would like voltage X to be this value. The voltage on plate 5 is amplified by unity voltage gain amplifier 16. The output of amplifier 16 drives capacitor 29 whose 5 other connection is alternately grounded and connected to integrator 19 by analog switch 17. That switch is controlled by waveform C which is generated as the NAND result of the non-inverted outputs of flip flops 22, 23. When this waveform is negative, capacitor 29 is connected to the integrator, and when it is positive, it is connected to ground. The effect is to deliver a charge to the integrator that is o proportional to any change in voltage on plate 5 that occurs in the transition from state IV to state I (see FIG. 5). That charge drives the integrator in a fashion that changes voltage X in the direction that tends to reduce the error. For a given position of plate 6, voltage X will approach the value that corresponds to a null transition on plate 5 with an arbitrarily small error.

Looking next at the transition between states I and II: The change in voltage on plate 5 may be considered to represent an error in output voltage Y, since there is a unique voltage for Y at which the transition will be zero, and we would like voltage Y to be this value. The voltage on plate 5 is amplified by unity voltage gain amplifier 16. The output of amplifier 16 drives capacitor 30 whose other connection is alternately grounded and connected to integrator 20 by analog switch 18. That switch is controlled by waveform D which is generated as the NAND result of the inverted output of flip flop 22 and the non-inverted output of flip flop 23. When this waveform is negative, capacitor 30 is connected to integrator 20, and when it is positive, it is connected to ground. The effect is to deliver a charge to the integrator that is proportional to any step change in voltage on plate 5 that occurs in the transition from state I to state II (see FIG. 5). That charge drives the integrator in a fashion that changes voltage Y in the direction that tends to reduce the error. For a given position of plate 6, voltage Y will approach the value that corresponds to a null transition on plate 5 with an arbitrarily small error.

The voltage X and Y that would be required to achieve null voltage changes on plate 5 may be calculated. In particular,

(3) ((V+) - X)(C2 + C4) - (X - (V-))(C 1 + C3).

Whence

(4) X = ((V+)(C2 + C4) + (V-XC1 + C3)) / (Cl + C2 + C3 + C4).

If, for example, V+ = -(V-) = 1 , then (5) X = ((C2 + C4) - (Cl + C3)) / (Cl + C2 + C3 + C4).

Similarly, under the same assumptions,

(6) Y = ((Cl + C2) - (C3 + C4)) / (Cl + C2 + C3 + C4).

As dielectric plate 6 moves to the right, sum (C2 + C4) increases whereas sum (Cl + C3) decreases. Thus output voltage X increases in response to this movement. On the other hand, sums (Cl + C2) and (C3 + C4) remain constant, so voltage Y remains constant.

Similarly, as dielectric plate 6 moves upward, sum (Cl + C2) increases whereas sum (C3 + C4) decreases. Thus output voltage Y decreases in response to this movement. On the other hand, sums (C2 + C4) and (Cl + C3) remain constant, so voltage X remains constant.

As may be observed by an inspection of equations (4) and (6), output voltage is proportional to capacitive change, (C2 + C4) - (Cl + C3) or (Cl + C2) - (C3 + C4), relative to total capacity (Cl + C2 + C3 + C4). Therefore, output is dependent on plate spacing and the thickness and dielectric constant of plate 6. If it were possible to remove those effects of capacitive coupling to plate 5 that are due to air gap only, dependency on the properties of plate 6 and the spacing of the plates would be largely nullified.

A method for doing this will be described that is essentially identical to that described for the second main embodiment. With reference to FIG. 4, inverting amplifier 27 applies the inverted sum of the voltages on plates 1 through 4 to capacitor 28. The other end of this capacitor is connected to a point electrically common to sensing plate 5. If, for example, the capacitance of capacitor 28 were just equal to the sum of Cl through C4 and the gain of amplifier 27 exactly equal to -0.25, cancellation of the effects of air coupling would be achieved. Capacitor 28 may be a part of the structure of the transducer so that its capacitance varies automatically with the spacing of plates 1 through 4 from plate 5, or it may be a conventional capacitor. Full, partial, or no cancellation may be appropriate, depending on the design requirements for a given application and the tolerances and materials used.

Now turning to FIGS. 11A-D, configurations of the fixed and moving plates 1-6 will now be discussed. In many applications, plates 1 through 4 will be rectangular or square as seen in FIG. 11A. Movable nonconductive plate 6 may be rectangular in form and mechanically accessed by slender extensions as illustrated. Another approach is illustrated in FIG. 1 IB. Here plate 6 consists of a strip with a rectangular hole. In this case, as the plate moves upward, sum (Cl + C2) decreases whereas sum (C3 + C4) increases. For horizontal movement, capacitive change is as described above. So output X is as described previously, but output Y is inverted relative to up and down movement. FIG. 11C illustrates an arrangement in which attachment to moveable plate 6 is by a central rod normal to the plate. Plates 1 through 4 have a cutout for protrusion of the rod. FIG. 1 ID is a non-orthogonal version of FIG. 4C illustrating that coordinate axes need not be constrained to a 90- degree relationship .

As will be appreciated by one skilled in the art, the configurations shown in FIGS. 11A-D are far from exhaustive. They are intended to be illustrative of the fact that the invention is not dependent on some particular arrangement or shape of plates 1-6.

Turning again to FIGS. 3 and 4, plate 6 is a very high impedance point in the circuit. It is desirable to shield this part of the circuit from external fields, yet at the same time keep stray capacity between this part of the circuit to ground at a minimum. In a preferred embodiments, plate 7 is placed behind plate 5 as seen in

FIG. 3, and this plate is driven by the output of buffer amplifier 16. Although the capacitance between plates 6 and 7 may be large in comparison with Cl through C4, it has virtually no effect because buffer amplifier 16 maintains this plate at the same potential as plate 5. Although plate 7 is not grounded, it is not significantly affected by external fields because the output of amplifier 16 has a low impedance.

In summary of all the described embodiments, the position transducer according to the present invention includes several air gap capacitors which have their values varied by movement of an insulator between them of dielectric constant greater than unity, and means to convert the capacitor values into one or two transducer outputs that are representative of the changed capacitance. The position transducer includes a circuit that feeds back signals based on output voltage or output voltages to the capacitors such that a null voltage at their electrical junction results. The position transducer may include a guard system to shield sensitive parts of the transducer. The position transducer may further include a circuit board system, two examples of which were described above and are illustrated in FIGS. 12A-B and 13A-B. The position transducer may further include conducting and nonconducting plates in the forms shown in FIGS. 6A-D, 7A-D, 10A-C, and 11A- D. The position transducer may further include an air capacitance nulling system. As described above, the present invention provides a position transducer which combines high accuracy, low moving mass, virtually unlimited life span and very low cost for many new applications that were not practical before now because of high cost, low performance, or limitation to one axis of available transducers.

As an illustration of the performance possible, data collected from a transducer according to the present invention is presented in FIGS. 14A and 14B. The example transducer is a linear, single axis transducer of the second main embodiment. Plate configuration is illustrated in FIG. 10A. Data was collected over a displacement range of 0.350 inches in 0.050-inch increments, and five complete test runs were conducted. FIG. 14A illustrates the excellent linearity of response, and FIG. 14B demonstrates excellent repeatability by inspection of mean voltage output and standard deviation over the repeated trials. The foregoing descriptions of specific embodiments of the present invention have been presented for purposes of illustration and description. They are not intended to be exhaustive or to limit the invention to the precise forms disclosed, and it should be understood that many modifications and variations are possible in light of the above teaching.

The embodiments were chosen and described in order to best explain the principles of the invention and its practical application, to thereby enable others skilled in the art to best utilize the invention and various embodiments with various modifications as are suited to the particular use contemplated. It is intended that the scope of the invention be defined by the Claims appended hereto and their equivalents.

Claims

WHAT IS CLAIMED IS:

1. A position transducer circuit comprising:

electrically conductive plate means separated from an electrically conductive reference plate wherein the plate means has a first capacitance relative to the reference plate and a second capacitance relative to the reference plate;

a nonconductive plate having a dielectric constant greater than unity, the nonconductive plate interposed between the plate means and the reference plate;

means for effecting movement between the nonconductive plate and the electrically conductive plate means to produce complementary changes in the first capacitance and the second capacitance, and

means for generating a transducer signal representative of the changed first and second capacitances, in which the electrically conductive plate means is electrically driven so as to produce a null signal at the reference plate representative of a zero value between the reference plate and the circuit neutral point.

2. A position transducer circuit comprising:

first and second electrically conductive plates separated from a third electrically conductive plate wherein the first plate has a first capacitance relative to the third plate and the second plate has a second capacitance relative to the third plate;

a fourth moveable nonconductive plate having a dielectric constant greater than one interposed between the first, second and third plates and covering a portion of the first and second plates;

means for moving the fourth plate between the first, second and third plates, wherein movement of the fourth plate produces complementary changes in the first capacitance and the second capacitance,

means for generating a transducer signal representative of the changed first and second capacitances, in which first and second conductive plates are electrically driven so as to produce a null signal at the third plate representative of a zero value between the reference plate and the circuit neutral point.

3. A position transducer as in Claim 2 wherein the complementary changes are substantially linear.

4. A position transducer as in Claim 2 wherein the first, second and third electrically conductive plates are shaped so that the complementary changes represent a particular non-linear function.

5. A position transducer as in Claim 2 wherein the fourth moveable nonconductive plate is shaped so that the complementary changes represent a particular nonlinear function.

0 6. A position transducer as in Claim 2 including one or more printed circuit boards, wherein at least one of the conductive plates is created in the printed circuit board process, and the means for generating the transducer signal is carried on the same printed circuit board.

5 7. A position transducer as in Claim 2 including a nulling means for substantially nulling the current that would pass through the first and second capacitances due to the effect of the air gap only.

8. A position transducer as in Claim 7 where the nulling means includes a nulling o plate arranged opposite to and separated from the third conductive plate.

9. Apparatus for generating a transducer signal representative of changed first and second capacitances, comprising:

an amplifier to detect and amplify the power of any imbalance signal at a reference plate point common to both capacitances;

a phase sensitive detector to convert the amplified imbalance signal into a DC signal dependent on phase and amplitude;

an integrator or high gain amplifier;

a means to convert the integrator output relative to a first reference voltage to a first AC output signal applied to the second capacitance; and

a means to convert the integrator or high gain amplifier output relative to a second reference voltage to a second AC output signal applied to the first capacitance.

10. The apparatus of Claim 9 including an amplifier that sums and inverts the second and first output signals present at the first and second capacitances, respectively, and whose output is capacitively coupled to the reference plate or a point common to the reference plate.

11. In a position transducer including electrically conductive plate means separated from an electrically conductive reference plate wherein the plate means has a first capacitance relative to the reference plate and a second capacitance relative to the reference plate, a nonconductive plate having a dielectric constant greater than unity, the nonconductive plate interposed between the plate means and the reference plate, the method comprising the steps of:

effecting movement between the nonconductive plate and the plate means to produce complementary changes in the first capacitance and the second capacitance wherein the sum of the first and second capacitances remains substantially constant, and

generating a transducer signal representative of the changed first and second capacitances, in which the electrically conductive plate means is electrically driven so as to produce a null signal at the electrically conductive reference plate.

12. A position transducer comprising:

electrically conductive plate means separated from an electrically conductive reference plate, wherein the plate means has a first, second, third, and fourth capacitance relative to the reference plate;

means for effecting movement between the nonconductive plate and the plate means to produce opposite changes between the sum of the first and third capacitance and the sum of the second and fourth capacitance as the plate is moved along one axis, and between the sum of the first and second capacitance and the third and fourth capacitance as the plate is moved along another axis; and

means for generating two transducer signals, one signal representative of the changes in the sum of the first and third capacitance and the second and fourth capacitance, and the other signal representative of the changes in the sum of the first and second capacitance and the third and fourth capacitance.

13. A transducer as in claim 12 in which the electrically conductive plate means includes a fifth capacitance relative to the reference plate.

14. A transducer as in claim 12 including a guard or shield plate parallel to the reference plate.

15. A transducer as in claim 12 including electronic means for generating the two transducer signals:

comprising means to generate two pairs of AC signals, the first pair representative of the difference between a first reference voltage and the transducer signal, the second pair representative of the difference between the transducer signal and a second reference voltage;

means for summing the two pairs of the AC signals and applying each sum to one of the four plates of the first conductive plate means;

means to convey the signal on the reference plate or a representation thereof to each of two synchronous detector means;

the first of said detector means responsive to the first pair of AC signals and the second said detector means responsive to the second pair of AC signals;

the output of the first detector means applied to a first integrator means and the output of the second detector means applied to a second integrator means; and

the output of first and second integrator means generating said transducer signals.

16. A transducer as in claim 15 including means to capacitively couple a signal to a point electrically common to said reference plate, said signal proportional to the inverted sum of the signals on said first conductive plate means.

A transducer as in claim 16 in which said means to capacitively couple a signal to a point electrically common to said reference plate comprises a plate mechanically spaced from said reference plate means.