Apparatus for Position Encoding

This invention generally relates to the field of optical measuring and testing and more specifically to apparatus for using interference fringe patterns for position encoding. There are two basic methods for producing fringe patterns: (1) an interferometric technique that utilizes interference phenomena, and (2) a Moire technique that utilizes shadow casting and/or pattern multiplication.

There are a wide variety of measuring and testing procedures that utilize interference fringe patterns and there are many ways to produce and control interference fringes. Generally, an interference fringe pattern is produced when at least two coherent beams of light are brought together and interact. When two coherent beams interact, they destructively interfere to produce dark spots or bands and constructively interfere to produce bright spots or bands.

Moire fringes are produced when two similar, geometrically regular patterns consisting of well defined clear and opaque areas are juxtaposed and transilluminated. Some examples of geometrically regular patterns used to generate Moire fringes include (1) Ronchi rulings, (2) sets of concentric circles, and (3) radial grids. The generation of Moire fringes can be considered as shadow casting; that is, the shadow of the first pattern falling onto the second pattern produces the Moire fringes. The mathematical function describing Moire fringes is obtained by multiplying the intensity transmissions or irradiances of the overlapped geometrically regular patterns.

Apparatus using Moire or interferometric techniques is used to position accurately two elements relative to each other. With interferometric techniques, an incoming beam of light is generally split into two parts. One part is reflected from a reference position; the other, from a movable element. The reflected beams are recombined to produce an output fringe pattern that "moves" as the movable element moves. In one example of a Moire technique, two high contrast Ronchi rulings of slightly different spatial frequencies are juxtaposed and transilluminated. One ruling is stationary while the other is movable in a predetermined plane. Photodetectors sense the variations in the light that passes through the gratings and produce signals that indicate the motion.

Certain disadvantages exist in apparatus that utilizes the interferometric techniques to form fringe patterns in position detection applications. For example, in such apparatus the two light beams generally travel through different light paths that contain distinct optical elements. If the elements in each path are not matched optically, aberrations distort the fringe pattern. Matched optical elements can eliminate the aberration problem; however, they significantly increase the overall expense of the apparatus. Moreover, this apparatus is subject to various outside influences, such as vibration and thermal change. These influences can cause fringe pattern motion or noise and lead to improper measurements.

Moire techniques also have many limitations. When small spacings and high accuracies are required, the geometrically regular patterns used to generate Moire fringes are quite difficult and expensive to produce. In applications where one ruling moves next to a fixed ruling, the spacing between the rulings must be held constant or errors result. Also, Moire fringes are localized, i.e., they exist in a very small region of space, and additional optical components are often required to image the Moire fringes into desired regions.

Recently, an amplitude grating and a spatially coherent, quasi-monochromatic light source have been used to generate interference fringes. An amplitude grating is a generally transparent to semi-transparent media whose opacity is altered in accordance with some spatially periodic pattern. An amplitude grating "breaks up" or diffracts an incoming beam of light into a series of diffracted cones or orders. The strength, or amount, of light in each order depends upon the exact shape of the periodic opacity of the amplitude grating. Although various diffracted orders could be approximately the same strength, scalar diffraction theory for a thin amplitude grating predicts that the dominant strength will lie in the zero order undiffracted light and that the strength of other diffracted orders will vary. Indeed, practical applications bear out this prediction.

In the apparatus in U.S. Pat. No. 3,738,753, issued June 12, 1973, light passes from a source through an amplitude grating to produce different order cones of diffracted light: for example, zero order and first order cones. To compensate for the different intensities, the diffracted light cones are reflected back through the grating. After the second passage through the grating, the zero order cone of the reflected first order cone and the first order cone of the reflected zero order cone have equal strengths and are combined to form a high contrast interference fringe field. This double pass system is quite stable because it closely approximates a common path interferometer. In a common path interferometer the interfering beams traverse the same optical path. Therefore, perturbations affect both beams simultaneously and do not distort the output fringe pattern which is sensitive only to differences between the two optical paths. However, problems in such a double pass system do occur because it is difficult to control grating substrate aberrations and mirror-grating separation.

Further improvements have been made with the advent of holographically produced amplitude gratings. Holographic amplitude gratings are produced by exposing a high resolution photographic emulsion to the precise interference pattern of a laser two-beam interferometer. During ordinary photographic processing, the photosensitive silver halide in the emulsion converts into opaque metallic silver to form the amplitude grating.

In U.S. Pat. No. 3,829,219, issued 1974 to Wyant, and U.S. Pat. No. 4,118,124 issued Oct. 3, 1978 to Matsuda, one such holographic grating, a double frequency holographic grating, produces a so-called "shearing" pattern. This grating is produced by sequentially exposing a single photographic emulsion to a first laser interference pattern of a first spatial frequency, f.sub.1, and then to a second laser interference pattern of a second spatial frequency, f.sub.2. Equal amplitude transmission modulations at both frequencies f.sub.1 and f.sub.2 are achieved by adjusting the exposure to the first and second laser patterns. Ordinarily, the two sequential exposures are identical, but if f.sub.1 and f.sub.2 are very different or if one laser pattern is in red light and the other is in green light, the sequential exposures must be compensated for the spectral and frequency responses of the photographic plate. These exposure adjustments to achieve equal amplitude transmission modulations in f.sub.1 and f.sub.2 are usually done by trial and error.

Upon illumination with spatially coherent, quasimonochromatic light, this double frequency grating produces two first order light cones of equal strength, one light cone being associated with each of the f.sub.1 and f.sub.2 frequencies. These two first order light cones interact to form a very stable, high contrast fringe pattern. Such a double frequency holographic shearing interferometer also is a common path interferometer. It is simple to construct. However, in this interferometer it is necessary to separate the zero order cone from the interacting first order cones. This separation requirement limits the f/number of the input light cone and the amount of shear obtainable. Moreover, if the two first order cones have high diffraction angles an astigmatic distortion of the output fringe field exists. In addition, the efficiency, or ratio of output fringe field power to input power, is only about 2%.

For many years people have bleached photographically recorded amplitude gratings to obtain "phase gratings". One basic type of such bleaching, known as volume bleaching, chemically converts the opaque silver in the photographic emulsion into a transparent, high index silver salt. A second type of bleaching, known as tanning, chemically removes the developed silver within the emulsion and leaves a void. A tanned phase grating has a corrugated surface. Whereas an amplitude grating selectively absorbs light, a bleached phase grating selectively introduces phase delays across the input light beam. As a result, a phase grating is much more efficient than an amplitude grating; that is, the ratio of first order power to input power is greater.

However, bleached gratings are generally characterized by substantial problems. They are very noisy and also may deterioriate physically back into amplitude gratings upon extended exposure to light. Bleached gratings also have a lower spatial frequency response than amplitude gratings. Although volume bleached gratings are less noisy and have a higher spatial frequency response than their tanned counterparts, they generally are weaker and less efficient.

The efficiency of a volume bleached grating can be increased by increasing its thickness. However, any substantial increase in thickness drastically changes the basic diffraction properties of the grating. Any amplitude or phase grating can be considered optically thick when the optical thickness of the emulsion is more than five times the grating spacing. A grating can be considered optically thin if the optical thickness of the emulsion is less than half the grating spacing. Properties of thick gratings are accurately predicted by electromagnetic theory while properties of thin gratings are described by scalar diffraction theory. For example, a thick phase grating output consists of only the zero order and one first order diffracted cone. In addition, diffraction takes place only for a plane wave input at a certain specified angle with respect to the grating. On the other hand, a thin grating of the same spacing produces multiple orders (i.e. the 0, .+-.1, .+-.2, .+-.3, etc. orders) with either a spherical wave or plane wave input at an arbitrary angle with respect to the grating.

Distinctions between optically thin amplitude and optically thin phase gratings are accurately predicted by scalar diffraction theory. When a pure sinusoidal amplitude transmission perturbation exists in a thin amplitude grating, only the zero and .+-.1 diffracted orders exist. When a pure sinusoidal phase perturbation occurs in a thin phase grating, many orders (e.g., the 0, .+-.1, .+-.2, .+-.3, and other orders) are observed. The strengths of the phase grating orders are proportional to the normalized Bessel functions [J.sub.n (m/2)].sub.2, where n is the order number (e.g., n equals 0, .+-.1, .+-.2, . . . ) and m is the strength or size of the phase perturbation in radians. When the amplitude grating perturbation departs from a pure sinusoidal form, additional diffracted orders are generated. The strengths of these additional orders are directly related to the strengths of the Fourier components associated with the grating perturbation function.

With a phase grating, the diffracted orders associated with a non-sinusoidal phase perturbation are predicted by convolving the individual outputs from each Fourier component of the phase perturbation. Such a multiple convolution reveals complicated phase relationships between multiple orders associated with just one particular Fourier component. In addition, diffracted orders corresponding to sum and difference frequencies are generated when the phase perturbation consists of more than one fundamental spatial frequency. For example, one might consider bleaching the previously discussed double-frequency holographic grating to improve its poor efficiency. Although bleaching will increase the overall efficiency of such a grating, the bleached grating, in accordance with the convolutional operation, produces sum and difference frequency diffraction cones that are in addition to and that interact with the desired fundamental frequency diffraction cones. It is then possible for the sum and difference frequency diffraction cones to destroy the fringe field.

Therefore, it is the object of this invention to provide an improved holographic phase grating for producing a high contrast interference pattern.

Another object of this invention is to provide an improved holographic grating that is useful in a number of applications including position detection.

Yet another object of this invention is to provide apparatus for accurate detection of position information.

In accordance with this invention, a spatially coherent light source is positioned adjacent to one side of a holographic grating and directs light to the grating along a light axis. The grating produces a first diffraction and a second diffraction of different order that have substantially equal strength and that overlap thereby to produce a high contrast, low-noise interference pattern in the area of overlap of the two diffractions. By the term "different order", I mean diffractions whose order numbers have different absolute values. Photodetectors are disposed on the opposite side of the grating to receive this interference pattern. Relative motion between the photodetection means and the grating in a plane that is normal to the light axis is readily detected and accurately measured by the photodetectors.

There are various embodiments that incorporate either a source of white light or of quasi-monochromatic light in confirmation with holographic gratings having particular characteristics. Position encoders for different applications are depicted. In one embodiment, a quasi-monochromatic light source and grating produce a number of overlapping diffractions. Two have substantially equal strength and another has nearly equal strength. The resulting separated fringes are in space quadrature and are all relatively high in contrast and low in noise.

This invention is pointed out with particularity in the appended claims. The above and further objects and advantages of this invention may be better understood by referring to the following description taken in conjunction with the accompanying drawings.

FIG. 1 is a diagram that illustrates the apparatus for producing a holographic grating that forms a part of this invention;

FIG. 2 is a chart that depicts the various basic steps for processing the holographic grating;

FIG. 3 is a diagram of an interferometer for producing fringe patterns;

FIG. 4A is a diagram of an embodiment of a position encoder constructed in accordance with this invention; FIGS. 4B and 4C depict typical fringe patterns that are produced by the apparatus shown in FIG. 4A;

FIG. 5A is a diagram of an alternate embodiment of a position encoder constructed in accordance with this invention for sensing motion along orthogonal axes;

FIG. 5B depicts typical fringe patterns that are produced by the apparatus shown in FIG. 5A;

FIG. 6 depicts the fringe patterns produced by the apparatus of FIG. 4A as modified to provide quadrature signals;

FIG. 7A is a diagram of a position encoder for producing stable quadrature position signals; FIG. 7B depicts the fringe patterns that are produced by the position encoder in FIG. 7A;

FIG. 8A is a diagram of another embodiment of a position encoder constructed in accordance with this invention; FIG. 8B depicts the fringe patterns that are produced by the position encoder in FIG. 8A;

FIG. 9A is a diagram of yet another embodiment of a position encoder constructed in accordance with this invention; FIG. 9B is a detailed pictoral view of a light source used in FIG. 9A; and FIG. 9C depicts the fringe patterns that are produced by the position encoder in FIG. 9A.

FIG. 1 depicts, in diagrammatic form, the arrangement of apparatus necessary for exposing a photographic plate during the production of a holographic phase grating. The holographic phase grating produced in accordance with the arrangement shown in FIG. 1 and the procedures outlined in FIG. 2 are essential to the operation of the diverse embodiments of the invention that are shown in the other Figures. Specifically, this apparatus includes a laser source 10 which directs light along an axis 11. The other apparatus in FIG. 1 splits the light into parts that travel over two separate paths and are then brought back together to expose a photographic plate 12.

A conventional beamsplitter 13 separates the light into two parts. A first part travels along a first path that includes mirrors 14 and 15 for reflecting the light into an objective lens and pinhole 16, thereby to produce a spherical wave that emanates from a point source at the pinhole. The wave appears in a cone 17 and is directed toward the photographic plate along an axis 18. The second path established by the beamsplitter 13 includes a mirror 20 and an objective lens and pinhole 21 that produce a spherical wave cone 22 that emanates from a point source at that pinhole along an axis 23. The light waves from these two point sources combine; they destructively interfere to produce dark bands and constructively interfere to produce bright bands at the photographic plate 12.

The photographic plate 12 mounts on a rotary table which positions the photographic plate 12 and accurately establishes an angle .theta. between the axes 18 and 23. The spatial frequency, .xi., of the interference pattern at plate 12 is closely approximated by the equation ##EQU1## where .lambda. is the laser wavelength. Although the fringes produced at the plate 12 are slightly hyperbolic, they are excellent approximations to rectilinear bands and therefore are shown as such in various Figures. Increasingly better approximations to rectilinear bands are achieved by increasing the distance along the axes 18 and 23 between the plate 12 and the pinholes 16 and 21, respectively.

The apparatus diagrammed in FIG. 1 has been used to manufacture gratings having the desirable properties that characterize my invention. The equipment is simple and relatively inexpensive. For example, the laser 10 can comprise a TEM.sub.00 mode laser; the beamsplitter 13, a conventional variable density beamsplitter that enables the intensity of the two beams to be equalized. The mirrors 14, 15 and 20 are standard planar mirrors. The objective lens comprises a conventional 10.times. microscope objective, and the pinhole matches that objective lens. The distances 18 and 23 are approximately 2 meters. With this specific arrangement, I am able to obtain a 500 line-per-millimeter interference fringe pattern over a 3".times.3" area with maximum fringe displacement error of about 0.00254 millimeter.

Once the apparatus in FIG. 1 is arranged, the emulsion on the photographic film can be exposed to the interference pattern as shown as Step 1 in FIG. 2. During this exposure step, certain controls must be exercised to assure a holographic grating of good quality. For example, the exposure should be made in an environment that is not subjected to vibrations. Thermal disturbances should be minimized as any air flow between the beamsplitter 13 and the photographic plate 12 can distort the resulting fringes. In applications where very high densities and minimal distortions are required the distances along axes 18 and 23 must be increased to 5 or even 10 meters. Precise determinations of .theta. and .lambda. must be made. Although this basic apparatus can be used to produce highly accurate holographic phase gratings, the maximum accuracy ultimately then will be determined by the accuracy of angular measuring equipment, the stability of the single frequency laser, the optical table stability, and the atmospheric and thermal controls that are exercised.

In order to produce a phase grating with special properties that enable the construction of the various disclosed embodiments, it is first necessary to produce an amplitude grating. Given the various properties of commercially available photographic emulsions and developers, a thin emulsion photographic plate and a chemically compatible developer are selected. A process of heavily overexposing and underdeveloping the emulsion reduces the optical thickness of the processed emulsion to a fraction of its original physical thickness. Thus, by utilizing the controls set forth in steps 1 and 2 of FIG. 2, one produces an amplitude grating characterized by having:

1. an optically thin emulsion conforming to scalar diffraction theory;

2. a specific form for the absorbtion function which converts to a correspondingly specific phase transmission function after bleaching; and

3. a specific amplitude or strength of the absorbtion function which converts to a specific peak to peak phase modulation after bleaching.

Specific plate types, exposures, development times and developers are discussed later.

Once the development of step 2 is complete, the photographic plate is washed in an acid short-stop solution in step 3. The solution contains an acid hardener. A two-minute treatment in a hardening bath produces acceptable results.

In step 4 the emulsion of the photographic plate is fixed and hardened. A standard fixing bath and acid hardener have been used successfully, the plate being immersed in the bath for about ten minutes.

Next (step 5) the emulsion is prewashed for thirty seconds and hypo-cleared in a hypo clearing bath for about two minutes. In step 6 the emulsion is washed (e.g., twenty minutes in filtered water) and then soaked in a methanol bath until all residual sensitizing dye is removed (step 7). Once the methanol bath has been completed, the plate is dried in a light blow air drying operation.

All the foregoing steps are conventional photographic processing steps that utilize commericially available chemicals. Upon completion of step 7 an amplitude grating has been produced. Steps 8 and 9 then convert this amplitude grating into a phase grating having the desired characteristics.

More specifically, after the photographic plate is dried thoroughly in step 7, it is bleached during step 8 in a bromine vapor until the plate is clear. Once the bleaching operation has been completed, the plate is rinsed in a methanol bath to remove residual Br.sub.2 and dried thoroughly by a light blow air drying operation in step 9.

It now will be beneficial to discuss certain characteristics of these holographic phase gratings that are particularly desireable. First, the exposure and development times and the emulsion have been chosen to produce "thin" gratings. As a specific example, I have made 393.7 line-per-millimeter gratings on Kodak 131-01 plates according to the foregoing processing procedure using an average exposure of 200 ergs/cm.sup.2 and a development time of 15 seconds in standard Kodak D-19 developer at 80.degree. F. Uniform development is achieved by using a large development tank and rapid manual agitation of the plate. After complete processing in accordance with the steps of FIG. 2, the resulting thin phase grating diffracts both input spherical waves as well as input plane waves; as previously stated, a thick grating diffracts only input plane waves incident at a particular angle with respect to the grating.

Measurements have shown that a thin phase grating manufactured according to the foregoing process has a pure sinusoidal phase transmission function whose peak-to-peak phase delay produces equal strength zero and .+-.1 diffraction orders. The 200 ergs/cm.sup.2 exposure produces an average amplitude transmission of approximately 0.45 for the developed, but unbleached, Kodak 131-01 plates. Experimental data has confirmed that a pure sinusoidal phase transmission function is maintained when the thin grating has an average amplitude transmission of 0.5 or less in its developed but unbleached state. The strength or peak-to-peak phase delay of the final phase grating is adjusted by controlling the initial exposure (Step 1, FIG. 2) within the limits set by an average amplitude transmission of 0.5 (measured after Step 7 in FIG. 2). A very weak phase grating produced with low exposure levels exhibits a strong zero order diffraction, a weak first order, and an even weaker second order. Stronger gratings produced with higher exposure levels exhibit increasingly more powerful first and second order diffraction and decreased zero order diffraction. Equal strength zero and .+-.1 diffraction orders or equal strength zero and .+-.2 diffraction orders are achieved by a trial and error adjustment of the initial exposure.

As another example, a 393.7 line-per-millimeter grating can be produced on "AFGA" 10E75 plates. Specifically, an average exposure of 2000 ergs/cm.sup.2 and a development time of 30 seconds in standard Kodak D-19 developer produces an optically thin amplitude grating with a second order amplitude transmission component that is about half the strength of the first order component. Control of exposure and development times can produce an average amplitude transmission of about 0.55 in the developed but unbleached state. Bleaching yields an optically thin phase grating characterized by a second order distortion consisting of about half the total phase perturbation function. The strength or peak to peak phase delay of the final phase grating can be further adjusted by controlling the initial exposure (Step 1, FIG. 2), within the limits set by an average amplitude transmission of 0.55 (measured after Step 7 in FIG. 2). This adjustment can produce a preferred grating in which the zero and .+-.2nd diffractions are equal. This grating produces fringes of equal contrast which are framed within overlap areas common to , .+-.1 diffracted cones and to .+-.1, .+-.2 diffracted cones.

The advantages of such a thin phase grating that produces two different diffraction orders of equal strength will now become apparent in the following discussion of an interferometer that utilizes such a phase grating.

Referring now to FIG. 3, an interferometer is depicted in schematic form that includes a helium neon laser 30 which directs light along an axis 31 to a negative lens 32. The negative lens 32 expands the beam slightly so that it completely fills a microscope objective 33. The microscope objective 33 focuses this light at a focal point FP displaced a distance Z.sub.1 from a holographic grating 34 constructed as described above. The laser 30, negative lens 32 and microscope objective 33 constitute a source of a quasi-monochromatic diverging spherical wave that emanates from the focal point FP. In one embodiment, the cone from the focal point FP is an f/2 cone.

When the spherical wave from the point source at the focal point FP strikes the grating 34, it produces a number of cones of diffraction. According to scalar diffraction theory, the strength of the diffracted cones is governed by the Bessel function [J.sub.n (m/2)].sup.2 where n is the diffraction order number and m is the grating transmission function peak-to-peak phase delay in radians. The previously specified exposure and development times for a Kodak 131-01 plate yield a value of m=2.870 at .lambda.=6328 .ANG.. The zero and first order diffraction cones are of equal intensity because [J.sub.0 (1.435)].sup.2 =[J.sub.1 (1.435)].sup.2. Moreover, the diffraction angles are such that the zero order cone overlaps both first order cones while the first order cones merely abut each other. At some point at a distance Z.sub.2 from the grating 34, an output such as is shown in FIG. 3 is produced. The zero order cone appears as planar circle 35, first order cones appear as planar circles 36A and 36B; and second order cones appear as planar circles 42A and 42B. Areas 37A, 37B, 43A and 43B are areas of overlap and the fringes are produced in those areas. Moreover, the fringes in the areas 37A and 378 are out of phase with each other. Thus, if the centrally located fringe in area 37A is a dark band, the corresponding fringe in area 37B is a light, or bright, band. By "light" and "dark" bands, I do not means bands having the same intensity across the band, as the bands are shown in the drawings. The fringe intensity actually varies smoothly and is proportional to the square of a sine function, although the eye may perceive distinct alternating bands under some illumination conditions.

The 180.degree. phase shift between the fringes in areas 37A and 37B is a direct result of having a pure sinusoidal phase transmission function associated with grating 34. When the phase transmission function of 34 departs from a pure sinusoid, the fringes in the areas 37A and 37B will have some other phase relationship not equal to 180.degree.. The 180.degree. phase shift is not essential to the production of high contrast fringe patterns; but it is important in a position detecting application where quadrature electrical signals are derived from the central fringes. Control of the grating transmission function form is achieved by selecting the proper combination of emulsion, developer, exposure and development time as previously discussed.

The interferometer shown in FIG. 3 has several properties. If the distance Z.sub.1 is varied, the number of fringes within the overlap areas 37A and 37B changes. Specifically, decreasing the distance Z.sub.1 decreases the number of fringes that appear in the overlap areas. As Z.sub.1 is varied, fringes "flow" into or out of the areas 37A and 37B. Although this "fringe flow" may cause the central fringes to widen or narrow, it does not move the central fringes; they remain located at the centers of their respective areas. The importance of this central fringe behavior with Z.sub.1 variations will be discussed later. If the grating 34 is moved in a plane that is normal to the axis 31 and perpendicular to the direction of the fringes, all the fringes in the areas 37A and 37B appear to slide through those areas, but the number of fringes in those areas remains unchanged. If the distance Z.sub.2 varies, the number of fringes also remain the same, but in this case their sizes change, the fringe widths becoming smaller as Z.sub.2 decreases. The interferometer parameters are related by the equation:

Where T is the fringe period in overlap regions 37A, 37B, 43A and 43B, .xi. is the spatial frequency of the grating 34 defined by equation (1) and Z.sub.1 and Z.sub.2 are the positive distances shown in FIG. 3.

The holographic grating interferometer in FIG. 3 is very stable and free of fringe distortion from outside influences because it is essentially a common path interferometer. Atmospheric changes, air currents and thermal instabilities do not distort the fringes. Moreover, the intensity of the light in each of the diffraction cones is much greater than that usually obtained from amplitude gratings because the phase grating essentially is transparent and relies entirely on time delays within the grating 34 to produce the diffraction cones. As a result the overall brightness of the fringe pattern is increased. Moreover, as the intensities in each of the zero-order and first-order cones are equal, the destructive and constructive interferences tend to be complete so the dark bands are essentially black while the bright bands are essentially twice as bright as the average light. Thus the grating enables the production of a simple common path interferometer that produces bright, high-contrast fringes.

If the interferometer in FIG. 3 is constructed with a holographic grating constructed as previously discussed with the AGFA 10E75 plate, a different transmission function peak-to-peak phase delay results because a nonsinusoidal phase perturbation is introduced into the grating. If the strengths of the zero and .+-.2 order diffractions are approximately equal, the various interferences produce four areas of overlap that are all high-contrast areas from the standpoint of conventional detectors. The fringes in areas 37A and 37B will be in phase while the fringes in areas 43A and 43B will be 180.degree. out of phase. More importantly, the fringes in areas 37B and 43B have a 90.degree. phase relationship. As a result, the interferometer uses a single source of quasi-monochromatic light in combination with the AGFA 10E75 plate to project four fringe patterns. The central fringes are in space quadrature. Thus, the positioning information in the central fringes is independent of variations in the Z.sub.1 distance.

The foregoing properties lay a basis for understanding applications of a bleached phase grating in a variety of position encoders. In such encoders the Z.sub.1 and Z.sub.2 distances remain essentially unchanged, but the grating is allowed to move thereby to shift the fringes.

In accordance with this invention, the interferometer shown in FIG. 3 is readily adapted to use in a position control system. As previously indicated, the fringes in the areas of overlap 37A and 37B in FIG. 3 move through the area of overlap in the direction of motion of the grating. Moreover, if the distance Z.sub.1 between the focal point FP and the grating remains constant, the number of fringes in the area of overlap remain constant. On the other hand, if the distance Z.sub.2 varies, then the number of fringes within the area of overlap remains the same but the area of the overlap varies, as would be expected in a projection type system.

A specific embodiment of a position encoder that can be used in a wide variety of measurement and control functions is disclosed in FIG. 4A. In FIG. 4A, light emanates from a point source 100 of quasi-monochromatic, spatially coherent light. A holographically recorded, single-frequency phase grating 101 is mounted in a carrier 102 that moves in in the X direction of an XY plane that is orthogonal to the light, or Z, axis. Light from the source 100 is diffracted into equal strength zero and first order cones by the grating 101 that is supported in a carrier 102. The zero order distribution is represented as a planar circle 103 while the two first order distributions are depicted by planar circles 104 and 105. The fringes in the areas of overlap 106 and 107 are projected onto photodetectors 110 and 111 that generate input signals for a position detection circuit 112, such circuits being well known in the art.

Referring to FIG. 4B, the photodetectors 110 and 111 are horizontally oriented in the central fringe that is produced in each of the areas of overlap 106 and 107; i.e., on axes 120 and 121 respectively. As previously explained, controlling the form of the grating phase transmission function causes the fringes in one area of overlap to be 180.degree. out of phase with the fringes in the other area of overlap. As shown in FIG. 4B the photodetector 110 is aligned with a dark band at the central fringe position while the photodetector 111 is aligned with a bright band at the central fringe position. For purposes of the photodetection, this particular embodiment is particularly simplified if the light source 100 generates red or near infrared light as photodetection cells, such as photodiodes, are particularly sensitive in this region of the spectrum.

If the carrier 102 moves slightly to the right in FIG. 4A along the X axis, the fringes shift with it. After an incremental motion, the bands that impinge the photodetectors 110 and 111 in FIG. 4B shift to the positions shown in FIG. 4C. Now a bright band impinges the photodetector 110, and a dark band impinges the photodector 111. If the grating 101 has a phase pattern of 400 lines per millimeter, this binary change represents a translation along the X axis of approximately 0.000050 inch. Yet even with this accuracy, this apparatus is relatively easy to construct because the bands that impinge the photodetectors are relatively wide. For example, bands having a width of about 0.1 inch are obtained when Z.sub.2 equals about 2 inches and Z.sub.1 equals 0.001 inche. See equation (1). Bands of this width facilitate the placement of the photodetectors because their positions can be established with fairly loose tolerances.

This apparatus is essentially insensitive to any changes in the distance between the grating 101 and the photodetectors 110 and 111 along axes 120 and 121. As previously indicated, the size of the fringe fields 106 and 107 changes if Z.sub.2 changes, but the number of fringes within the fields does not change. Thus, in FIG. 4B the photodetectors 110 and 111 remain centered on their respective central fringes, notwithstanding any variations in the distance Z.sub.2 along axes 120 and 121.

The point source 100 shown in FIG. 4A comprises elements such as a laser 30, negative lens 32 and microscope objective 33 of FIG. 3. The laser 30 can comprise a low power TEM.sub.00 mode helium neon cylindrical or other like laser. The negative lens 32 and microscope objective lens 33 are movable along the axis 31 in FIG. 3 and focus the light at a focal point FP. The negative lens 32 expands the beam from the laser 30 slightly so as to completely fill the microscope objective lens 33 aperture with a uniform light distribution. A -4 mm focal length biconcave lens is a satisfactory negative lens. The microscope objective lens 33 is a conventional objective lens, a 10.times. N.A. 0.25 objective lens being satisfactory. With a source of this construction, the laser radiation is allowed to overfill the objective, thus producing a well bounded radiation field with radial symmetry as shown in FIG. 4A.

On the other hand, a simple laser diode could also be used by itself or in combination with the microscope objective 33. The shape of the laser diode radiating region is approximately rectangular instead of circular. Therefore, when the laser diode is used by itself, the zero and .+-.1 order distributions are not the radially symmetrical and sharply bounded circles 103, 104 and 105 in FIG. 4A. However, FIG. 4A does constitute an approximate representation of the actual irradiance distributions in the XY plane when the spatially coherent, quasi-monochromatic source comprises only a laser diode. When a laser diode and microscope objective are used in combination, the laser diode radiation is allowed to overfill the microscope objective. Thus, a well-bounded radiation field with greater radial symmetry is produced. With any such "source", the discussions concerning the various position encoders is totally valid and unaffected by the use of the idealized representations 103, 104 and 105, FIG. 4A.

While the apparatus in FIG. 4A is useful in making measurements in one direction, FIG. 5A discloses a carrier with a grating 123 that moves in both the X and Y directions. The grating 123 differs from the grating 101 in FIG. 4A and the difference is most readily understood by referring to FIGS. 1 and 2. In forming the grating 123, the photographic plate 12 is exposed as previously described with reference to step 1 in FIG. 2. However, the plate is then turned 90.degree. and exposed again before it is developed. This double exposure produces superimposed horizontal and vertical single frequency interference patterns.

Referring again to FIG. 5A, when the grating 123 is illuminated with any point source of quasi-monochromatic, spatially coherent light, the grating produces five diffraction cones of interest that are approximated in FIG. 5A and are shown more clearly by a planar projection in FIG. 5B. The vertical phase pattern on the grating produces cones represented by circles 103 through 105 as previously described, these cones producing areas of overlap 106A and 107A. However, the horizontally disposed phase pattern produces a pair of first order cones in the vertical direction designated by reference numerals 114 and 115. Four areas of overlap are important. Arcuate wedge areas 106A and 107A correspond to the areas 106 and 107 shown in FIG. 4B that are independent of the influence exerted by the first order cones 114 and 115. Arcuate wedge areas 116A and 117A are formed by overlapping the zero order cone 103 and the first order cones 114 and 115, and they are independent of any influence by the first order cones 104 and 105. The photodetectors are aligned on the central fringe for each area of overlap. Photodetectors 110 and 111 aligned with the areas 106A and 107A respond to motion along the X axis as previously discussed. Photodetectors 120 and 121 are aligned with the areas 116A and 117A. They sense vertical motion along the Y axis. These four photodetectors are then coupled to the position detection circuit 124 that responds to these signals either for indicating XY motion or for providing an input to an XY positioning servo mechanism.

FIG. 6 illustrates, diagrammatically, apparatus that generates quadrature signals. Basically this apparatus employs the apparatus shown in FIG. 4A with the addition of photodetectors 125 and 126. Photodetectors 110 and 111 are shifted upwardly but remain positioned on the central fringes. The additional photodetectors 125 and 126 are positioned one-quarter fringe period to the right of each central fringe (i.e. 90.degree. out of phase or in a quadrature position). Thus, as will be readily apparent, the signals from these four photodetectors provide quadrature signals that inherently provide both position and direction information.

In each of the foregoing applications, it is assumed that the Z.sub.1 distance between the point source of light and the grating remain constant. As apparent, however, such a constant dimension might be difficult to achieve in some practical applications. FIGS. 7A and 7B disclose apparatus which is essentially insensitive to a reasonable range of variations in the Z.sub.1 dimension. This again is shown in connection with an apparatus for detecting translation along the X direction only. The resulting signals are conveyed to a position detection system 127 that includes an up-down counter to count the passage of fringes and to provide an accurate indication of motion.

More specifically, a laser light source 100 transmits light through the grating 101 to produce zero and first order cones 103, 104 and 105 with areas of overlap 106 and 107. As shown in FIG. 7B, photodetectors 110 and 111 then provide a first set of signals. Another light source 130 is disposed below the light source 100. It is positioned to produce a zero-order cone 133 and first-order cones 134 and 135 that lie below the diffraction cones produced by light from the source 100. The horizontal, or X, position of source 130 is adjusted to produce overlap areas 136 and 137 whose central fringes are 90.degree. out of phase with the central fringes of overlap areas 106 and 107. Photodetectors 140 and 141 are aligned with these central fringes in overlap areas 136 and 137. With this arrangement, quadrature signals are generated from four central fringes whose position is unaffected by variations in Z.sub.1. Increasing Z.sub.1 causes the width of the central fringes to decrease, but so long as the photodetector aperture can resolve the central finges, an accurate quadrature signal is produced.

FIG. 8A depicts a position encoder that generates quadrature signals from a single source in four separate areas of overlap. This decoder utilizes the interferometer in FIG. 3 which utilizes a thin holographic grating with a specific nonsinusoidal phase perturbation funcion containing a large second order distortion. As previously described, the central fringes in the four areas of overlap are in space quadrature, so the measurements are relatively insensitive to variations in Z.sub.1. This position encoder modifies the apparatus in FIG. 4A by adding two photodetectors and associated electronics for processing the quadrature position signals.

More specifically, a laser light source 100 transmits light through grating 151 to produce zero, first order and second order cones 35, 36A, 36B, 42A, and 42B with four areas of overlap, 37A, 37B, 43A, and 43B. Grating 151 with its specific, nonsinusoidal phase perturbation, generates the fringe patterns in the overlap areas as shown in FIG. 8B. If the zero and .+-.2 diffraction orders are equal, the areas in FIG. 8B will have substantially equal contrast. As previously discussed, the fringes in areas 43A and 43B are 180.degree. out of phase and the fringes in areas 37A and 37B are in phase, while the fringes in areas 43B and 37B are 90.degree. out of phase. Photodetectors 110 and 111 plus detectors 148 and 149 are located at the four central fringes. As the central fringes are in space quadrature, the signals from the corresponding detectors will also be in quadrature. Moreover, only source 100 is used as opposed to two sources in the apparatus in FIG. 7A.

In each of the foregoing embodiments, the light source produces quasi-monochromatic, spatially coherent light. FIG. 9A depicts a position encoder which includes a light source that produces white, spatially coherent light and which produces quadrature position signals.

Specifically, two spatially coherent, white light sources 150 and 151 produce two point sources at a focal plane FP. Light from these point sources is directed toward an array 152 of detectors through a holographic grating 153. Both point sources are located at the same distance, Z.sub.1, from the grating 153 in the focal plane. As a carrier 154 moves in an X-Y plane along an X axis, the grating 153 produces overlapped diffractions of different order. High contrast achromatic fringes are produced in the area of overlap.

In order to achieve optimum results in the white light position encoder, energy in the zero order diffractions 160 and 170 must be minimized. According to scalar diffraction theory, for any wavelength an optically thin phase grating with a purely sinusoidal peak-to-peak phase delay of 4.8 radians extinguishes all energy in the zero order diffraction. However, for a given grating, achieving fixed characteristics, complete extinction of the zero order diffraction is not possible for a range of wavelengths since the phase delay of 4.8 radians is wavelength dependent. Therefore, the grating 153 is manufactured with a peak-to-peak phase delay of 4.8 radians at the wavelength associated with optimum detector efficiency.

To summarize, high contrast achromatic fringes are achieved when the grating 153 has the following characteristics:

1. The emulsion is optically thin and conforms to scalar diffraction theory;

2. The phase perturbation function is purely sinusoidal; and

3. The peak-to-peak phase delay is 4.8 radians at the wavelength of maximum detector sensitivity.

A grating with these characteristics can be exposed and developed according to the procedures of FIG. 2.

The apparatus of FIG. 9A constitutes a basic white light interferometer. In one embodiment, a 196.85 l/mm holographic grating 153 generates diffraction cones from source 150 and diffraction cones from source 151. Each of the sources 150 and 151 is identical. FIG. 9B depicts the source 150 that comprises a filament lamp 140, collecting lens 141 and pinhole 142. Lamp 140 emits white light; lens 141 collects flux from lamp 140 and maximizes the irradiance at pinhole 142. Pinhole 142 is chosen to have a width compatible with the existence of spatially coherent light at aperture 144 of microscope objective 145; a typical pinhole diameter being 5 to 15 microns. Objective 145 is operated at a numerical aperture of about 1.3 times the .xi..lambda. product of grating 153, .xi. being the grating spatial frequency and .lambda..apprxeq.5500 .ANG. for white light operation. Such a choice of numerical aperture maximizes the overlap areas 161 and 162. Filter 143 is a simple colored glass optical filter used to limit the spectral content of the white light radiation. Microscope objective 145 forms an image FP of pinhole 142. Limiting the discussion to only one source, when zero order distribution 160 is absent, the area 161 is formed by the overlap of +1 and -1 diffracted orders. The physical dimensions of 161 are maximized by adjusting the numerical aperture of objective 145 by means of aperture 144. Interference fringes formed in area 161 are achromatic since light of all wavelengths forms fringes of the same spatial frequency. The angle between +1 and -1 order diffractions within area 161 is larger for longer (red) wavelengths and smaller for short (blue) wavelengths. This effect is perfectly counterbalanced by the requirement of larger interaction angles for longer wavelengths and smaller interaction angles for shorter wavelengths to produce fringes of the same spatial frequency. Also important is the symmetrical diffraction of +1 and -1 order energy about optical axis 14, thereby causing the fringe patterns associated with each wavelength to be in register or in phase, as well as the same spatial frequency.

In this specific embodiment, the spatial frequency of the grating is 196.85 lines/mm. If the grating 153 translates one-half the grating spacing, the fringe pattern shifts one cycle. This corresponds to a one-period shift of the fringe pattern each time the carrier 154 moves 0.0001 inch along the X axis.

Referring now to FIGS. 9A and 9C, the light sources 150 and 151 produce two sets of diffractions. The light source 150 projects the +1 order 155 and -1 order 156 and the +2 and -2 orders 157 and 158 respectively on either side of the zero order position 160. Given that all the input energy in the zero order is diffracted out, the area of overlap 161 of the +1 and -1 orders 155 and 156 defines an area of high contrast, achromatic fringes. Similarly the light source 151 produces an area of overlap 162 of the +1 and -1 orders 163 and 164. The +2 and -2 orders 165 and 166 are also shown as is the nunal position of the zero order 167.

The photodetection array comprises detectors 170 through 173. The detectors 170 and 171 are aligned on out-of-phase fringes in the area of overlap 161, the detectors 172 and 173 are vertically aligned with the detectors 170 and 171 and are located in the area of overlap 162. Moreover, the light sources 150 and 151 are relatively displaced by a multiple of 0.0000125 along the X axis to place the central fringes in the areas of overlap 160 and 161 in space quadrature. Thus, the detectors produce space quadrature signals.

This detector array is relatively insensitive to variations in Z.sub.1. If Z.sub.1 can be maintained without significant variation, the detector array can be positioned in only one of the overlap areas, thereby eliminating the requirement for two light sources. On the other hand, if Z.sub.1 cannot be maintained at a constant or even approximately constant value, four sources could be located in quadrature positions along the X axis. Then one detector could be located at the central fringe position for each area of overlap.

In summary, there has been disclosed a basic inteferometer construction that utilizes a holographically recorded, single-frequency, bleached, phase grating for producing stable, high-contrast fringe patterns with high efficiency. Moreover, there has been disclosed a specific application of this interferometer for sensing or controlling the position of a mechanical element. In this application, the grating is moved in a plane that is orthagonal to a light axis so the fringe patterns move past detectors. The detectors convert the passage of the fringe patterns into signals that indicate discrete increments of motion by the grating.

It will be apparent from the foregoing discussion, however, that the specific embodiments of this invention that have been disclosed are merely representative. The basic principles can be employed in other embodiments with the attainment of some or all of the advantages of this invention. Therefore, it is an object of the appended claims to cover all such variations and modifications as come within the true spirit and scope of this invention.