JP7823327B2 - Microlens array, diffuser plate for projection device, and projection device - Google Patents

Description

The present invention relates to a microlens array and a projection device.

A technology is known in which multiple concave lenses are formed on the surface of a transparent substrate and light is diffused using the refraction phenomenon (see, for example, Patent Documents 1 and 2). Figure 1 is a schematic diagram of a conventional lens array, with (A) a top view and (B) a vertical cross-sectional view. In a known diffuser plate having an array of concave lenses LNS, the depths of the concave lenses LNS vary as d1, d2, ..., and the central coordinate C of the concave lenses LNS varies in the in-plane direction. In (A) of Figure 1, dotted circles indicate equally spaced positions in the X-Y plane as a reference. Compared to the equally spaced positions, the central coordinate C of the concave lenses LNS is offset from the center of the equally spaced positions in both the X and Y directions, and the pitch varies as P1, P2, ... in the X-Y plane. Varying the depth d and pitch P of the concave lenses LNS prevents diffracted light from being generated in only a specific direction.

Patent No. 6424418 Patent No. 6680455

The inventors discovered that if the pitch of the lenses arranged in any one-dimensional direction in a microlens array varies, it becomes difficult to impart good cutoff characteristics to the diffused light profile. They also discovered that, depending on the degree of depth variation, diffracted bright spots may occur.

The objective of the present invention is to provide a microlens array that has good cutoff characteristics and suppresses the occurrence of diffraction bright spots.

In one aspect of the present invention, the microlens array comprises:
A glass substrate that is transparent to the wavelength used,
a plurality of aspherical lenses formed on a first surface of the glass substrate ;
and
a center of the aspherical lens is located at two or more different positions in a direction perpendicular to the first surface, and a value σ/μ obtained by dividing 1σ by an average value μ when the center positions of the aspherical lens in the direction perpendicular to the first surface follow a normal distribution is greater than 0.027 and less than 0.200,
When the average of all pitches in an area excluding a total of 25% from both ends of the aspherical lenses arranged in any one-dimensional direction within the plane of the first surface is set to 1, the absolute value σ _p of the maximum possible difference in relative pitch ratio is less than 0.075 .

A microlens array with good cutoff characteristics and reduced diffraction spot generation is realized.

FIG. 1 is a schematic diagram of a conventional lens array. 1A and 1B are a cross-sectional schematic view and a top view image of a microlens array according to an embodiment. 1 is a sample image of a microlens array according to an embodiment. XX' cross-sectional view of FIG. 3. 4 is a diffusion profile of the microlens array of FIG. 3. FIG. 10 is a diagram showing a simulation result of an aspherical lens array. FIG. 10 is a simulation diagram of a diffusion profile when the depth variation of an aspherical lens is changed. 10A and 10B are simulation diagrams of three-dimensional intensity distribution and two-dimensional intensity distribution when the depth variation of the aspherical lens is changed. This is a top view image of an actually fabricated sample. This is the actual measurement value of the lens depth along the line II' in FIG. 9A. This is the analysis result of the actual measurement values of FIG. 9B. FIG. 10 is a diagram showing the upper limit of the depth variation of an aspherical lens. 9B is a diffusion profile in the Y direction of the sample of FIG. 9A. FIG. 10 is a simulation diagram of a diffusion profile when the in-plane pitch variation of an aspherical lens is changed. FIG. 10 illustrates the change in cutoff band as a function of pitch variation. FIG. 10 is a diagram illustrating how to determine a cutoff band. 10A and 10B are simulation diagrams of two-dimensional intensity distributions when the pitch variation of the aspherical lens is changed only in the Y direction. FIG. 10 is a diagram showing changes in the relative pitch ratio of samples produced by changing the maximum pitch variation in the Y direction. FIG. 10 is a diagram showing the results of measurements of diffusion characteristics when the maximum pitch variation in the Y direction is changed. FIG. 16B is a simulation diagram of the diffusion characteristics corresponding to FIG. 16A. FIG. 16B is an enlarged view of a region of FIG. 16A. FIG. 10 is a diagram illustrating how to determine a slope as a cutoff characteristic. FIG. 18 is a diagram showing an analysis result based on FIG. 17 . 1 is a schematic diagram of a projection device to which a microlens array according to an embodiment is applied.

In an embodiment, in a microlens array in which multiple aspherical lenses are formed on a first surface of a substrate transparent to the wavelength being used, the center positions of the aspherical lenses are varied in a direction perpendicular to the first surface, thereby suppressing pitch variation of the aspherical lenses within the first surface. Specifically, as described below, pitch variation relative to the average of all pitches in an area excluding both ends of 25% of the aspherical lenses arranged in any one-dimensional direction within the first surface is suppressed to less than 7.5%, more preferably 5.0% or less. Suppressing pitch variation within the aspherical lenses improves the cutoff characteristics of the diffusion profile. Furthermore, diffraction bright spots are suppressed by varying the center variation of the aspherical lenses, i.e., lens depth variation, by a predetermined percentage or more relative to the average value.

Figure 2 shows an example of a microlens array 10 according to an embodiment. Figure 2 (A) is a schematic cross-sectional view of the microlens array 10, and (B) is a top view.

As shown in Figure 2 (A), the microlens array 10 has multiple aspherical lenses 13 formed on a first surface 101 of a substrate 11 that is transparent to the wavelength being used. In this example, the aspherical lenses 13 are concave lenses with parabolic surfaces. The centers 14 of the aspherical lenses 13 are located at two or more different positions in a direction perpendicular to the first surface 101, which in this example is the depth direction of the substrate 11. In other words, the depth d of the aspherical lenses 13 varies.

On the other hand, the pitch P between the centers 14 of the aspherical lenses 13 is substantially constant. "Substantially constant" means that the pitch variation is suppressed to be smaller than a predetermined variation. In this specification, the pitch variation refers to the absolute value of the maximum possible difference in the randomly distributed relative pitch ratios over the entire pitch in a region of the lenses in any one-dimensional direction, where the average pitch in the region excluding both end 25% of the entire pitch in that one-dimensional direction is set to 1, and this is represented as _σp . In the microlens array 10, when focusing on the lens arrangement in any one-dimensional direction, the pitch variation _σp of the aspherical lenses 13 in the region excluding both end 25% is suppressed to less than 0.075, i.e., the variation from the average pitch is suppressed to less than 7.5%.

In Figure 2(B), the white dots in the center of each aspherical lens 13 indicate the lens center. The microlens array 10 includes a square lattice arrangement of aspherical lenses 13, with a nearly constant pitch in both the X direction (horizontal direction on the paper) and the Y direction (vertical direction on the paper). The absolute value of the pitch correlates with the diffusion angle in that direction. In the example of Figure 2(B), the aspect ratio of the pitch in the X and Y directions is set to greater than 1, but when forming a square projected image, the pitch of the aspherical lenses 13 may be designed to be the same in both the X and Y directions. In either case, the pitch variation of the aspherical lenses 13 in one-dimensional directions such as the X and Y directions is kept to less than 7.5%, and more preferably to 5.0% or less, resulting in a nearly regular lens arrangement in the in-plane direction.

By regularly arranging the aspherical lenses 13 in the XY plane, the cutoff characteristics of the diffusion profile of the microlens array 10 can be improved. In the direction perpendicular to the XY plane (i.e., the first surface 101), by varying the depth positions of the centers 14 of the aspherical lenses 13 of the microlens array 10 by a certain degree or more, the occurrence of diffracted bright spots can be suppressed. Below, we will explain in detail (1) the variation of the aspherical lenses 13 in the direction perpendicular to the first surface 101, and (2) the regularity of the in-plane arrangement of the first surface 101.

<Variation in a direction perpendicular to the first surface>
Fig. 3 is a top view image of a sample of the microlens array 10 according to the embodiment, and Fig. 4 is a schematic diagram of the X-X' cross section of Fig. 3. As shown in Fig. 3, a plurality of aspherical lenses 13 are regularly arranged in the XY plane. In the Z direction orthogonal to the XY plane, the centers 14 of the aspherical lenses 13 are located at two or more different positions, and therefore the conic coefficients, radii of curvature, etc. of the aspherical lenses 13 are different.

As shown in the cross-sectional profile of Figure 4, the pitch P between the centers 14 of the aspherical lenses 13 is almost constant, but the depth varies as d1, d2, d3, ... The difference Δd between the depths d1 and d2 is approximately 8 μm in the sample of Figure 3.

Aspherical lenses 13, which have a substantially uniform in-plane pitch on the first surface 101 but which vary in lens depth or height in a direction perpendicular to the first surface 101, are formed by wet etching a pre-treated glass substrate. Pre-treatment can involve irradiating a certain position on the glass substrate with pulsed laser light to modify a partial region within the glass substrate, creating a density distribution in the thickness direction at least at the position irradiated with pulsed laser light, or by forming a predetermined wedge-shaped depression in the surface of the glass substrate using a chemical or physical technique (see, for example, WO 2019/189225). Wet etching is performed so that no flat surfaces remain between adjacent aspherical lenses 13, as shown in Figure 4. If flat surfaces remain, light will travel straight without being diffracted, generating zero-order light and resulting in bright spots.

When the aspherical lens 13 is a convex lens, multiple concave surfaces with a uniform pitch and varying depths can be formed on the substrate using the method described above, and then the convex lens array can be formed using this substrate as a mold.

The aspherical lenses 13 of the microlens array 10 in Figure 3 have sufficient variation in the depth or height direction, but it has not been clarified how the degree of variation in the depth or height of the aspherical lenses 13 affects the diffraction bright spots. Therefore, we will examine how variation in the lens center position in a direction perpendicular to the first surface 101 affects the diffraction bright spots.

Figure 5 shows the diffusion profile in the Y direction of the microlens array 10 in Figure 3. The horizontal axis represents the diffusion angle (°), and the vertical axis represents the relative intensity. The diffusion profile of the microlens array 10 has a rectangular pulse shape. If typical frosted glass is used as a diffuser, the laser light will have a Gaussian diffusion profile in the far field. In contrast, the microlens array 10 in Figure 3 is designed to diffuse light into a rectangular field of view. In this example, the field of view (FOV) in the Y direction is 45°.

In the embodiment, the "FOV" of the microlens array 10 refers to the angular range in which the relative intensity of the diffusion profile is 0.5 or greater. The relative intensity is the intensity when the average intensity in the diffusion angle range of -10° to +10° is normalized to 1. The FOV of the diffusion profile of the microlens array 10 in the X direction measured based on this definition is 60°, and the FOV in the Y direction is 45° as described above. The steepness of the rise of the diffusion profile indicates the cutoff characteristics, but cutoff characteristics will be discussed later; here, we will consider the suppression of diffracted bright spots.

For a microlens array 10 that diffuses light into a rectangular field of view, it is desirable for the intensity distribution at the top of the diffusion profile to be flat. This is because if the intensity of the diffracted light is locally strong, diffracted bright spots will occur, causing uneven brightness in the projected image. There should be an appropriate range of depth variation where the top of the diffusion profile approaches flatness. Therefore, we conducted a simulation to investigate how the degree of depth variation of the aspherical lens 13 affects diffracted bright spots.

Figure 6 shows the simulation results for an aspherical lens array. The substrate 11 is transparent to visible light and is made of "D263 T eco" glass (manufactured by SCHOTT) with a refractive index of 1.5230 for the D-line. The diffusion pattern obtained in the optical measurement is observed on a screen 103 mm away, with a minimum measurement resolution of 0.5 μm, a wavelength of the measurement light of 940 nm, and a beam diameter of 3 mm. Next, multiple aspherical lenses 13 are placed on the substrate 11. The conic coefficient of the aspherical lenses 13 is set to -1, the pitch in the X direction to 100 μm, the pitch in the Y direction to 83 μm, and the pitch variation in the X and Y directions to 0. The depth variation of the aspherical lenses 13 is set to 2.00 μm. The depth variation is 1σ when the depth of the aspherical lenses 13 of the microlens array 10 follows a normal distribution, and is the variation (standard deviation) from the median or mean value.

The simulated FOV of the microlens array 10 in the X direction is 52.9°, and the FOV in the Y direction is 43.2°. When the depth variation of the aspherical lens 13 is set to 2.0 μm, the relative intensity is slightly higher at both ends of the top of the diffusion profile in the X and Y directions, but there is no large ripple (fluctuation) between the peaks at both ends of the top, and it is almost flat. When we say that the top of the diffusion profile is "flat," we mean that diffraction bright spots are suppressed in the intensity distribution, or that almost all of the ripples at the top of the diffusion profile fall within a relative intensity range of 0.8 to 1.2.

Next, under the above conditions, the diffusion profile is obtained by varying only the depth variation of the aspherical lens 13. Figure 7 shows the simulation results when the depth variation (1σ) of the aspherical lens 13 is varied in seven ways: 0.00 μm, 0.40 μm, 0.80 μm, 1.20 μm, 1.60 μm, 2.00 μm, and 2.40 μm.

According to the simulation results in Figure 7, when the depth variation (1σ) is 0.00 μm, 0.40 μm, and 0.80 μm, the relative intensity fluctuates significantly at both ends of the top. However, when the depth variation is 1.20 μm or greater, the relative intensity remains between approximately 0.8 and 1.2. From this, the depth variation should be greater than 0.80 μm, preferably 1.20 μm or greater, more preferably 1.60 μm, even more preferably 2.00 μm or greater, and even more preferably 2.40 μm or greater. This keeps the relative intensity within the FOV at 1.2 or less, making the intensity distribution closer to uniform.

Figure 8 shows the three-dimensional intensity distribution (top row) and the two-dimensional intensity distribution in the X-Y plane (bottom row) when the depth variation (1σ) of the aspherical lens 13 is 0.00 μm, 0.80 μm, 1.60 μm, and 2.40 μm. When the depth variation is 0.00 μm or 0.80 μm, many peaks appear along the edge of the diffusion region, and stripe-shaped intensity differences appear within the diffusion region. The peaks along the edge and the stripes within the diffusion region are due to the occurrence of diffraction bright spots. In contrast, when the depth variation is 1.60 μm, the intensity distribution within the diffusion region, particularly in the center, is uniform. When the depth variation is 2.00 μm, the intensity distribution is uniform overall, including the edge region.

The simulation results in Figures 7 and 8 show that the effect of diffraction bright spots can be suppressed by increasing the depth variation (1σ) of the aspherical lenses 13 to more than 0.80 μm. The depth variation of the aspherical lenses 13 is preferably 1.20 μm or more, more preferably 1.60 μm or more, even more preferably 2.00 μm or more, and even more preferably 2.40 μm or more. On the other hand, if the depth variation is too large, the diffusion profile will not be flat-topped but will be closer to a Gaussian distribution shape. When realizing a rectangular diffusion area using a microlens array 10 in which the aspherical lenses 13 are arranged in a square lattice, it is preferable that the diffusion profile shape be flat-topped rather than Gaussian. As will be described below with reference to Figure 9D, the depth variation (1σ) is preferably less than 6.00 μm, and more preferably 4.00 μm or less.

Figure 9A is a top view image of an actually fabricated microlens array 10A, Figure 9B is the actual measured values of depth variation along line II' in Figure 9A, and Figure 9C is a histogram of the actual measured values of depth variation. The depth of 63 aspherical lenses 13 was measured along line II' in Figure 9A, and as shown in Figure 9B, the in-plane pitch was almost constant, and only the depth of the aspherical lenses 13 varied.

The median depth of the aspherical lenses 13 in microlens array 10A is 29.275 μm, the average is 29.486 μm, the maximum is 37.634 μm, the minimum is 25.139 μm, and the depth variation (standard deviation) is 2.643 μm. The depth variation that can suppress the effects of diffracted bright spots depends on its relative relationship to the average value, not on the absolute value. Therefore, the average value is μ, and the value obtained by dividing the standard deviation σ by the average value (σ/μ) is introduced as a new index. The value of σ/μ for the above microlens array is 0.090 (2.643/29.486).

Figure 9D shows the upper limit of lens depth variation. When the depth variation is 6.00 μm or more, the flatness of the top of the diffusion profile becomes insufficient. The new index value in this case is 0.203. In other words, if the new index value is less than 0.203, the flatness of the top of the diffusion profile is maintained. When the depth variation is 4.0 μm or less (the new index σ/μ value is 0.136 or less), the top of the diffusion profile becomes flatter.

Figure 10 shows the diffusion profile in the Y direction of the microlens array 10A in Figure 9A. When the value of the new index σ/μ is 0.090, the relative intensity of the top of the diffusion profile falls between 0.8 and 1.2, and has sufficient depth variation to suppress bright spots. This also applies when the microlens array 10A is a convex lens array; by providing appropriate height variation to the convex aspherical lenses, the variation in the relative intensity of the top of the diffusion profile can be kept between 0.8 and 1.2.

Based on the above considerations, diffraction bright spots can be suppressed by varying the center positions of the aspherical lenses 13 of the microlens arrays 10 and 10A to two or more different positions in a direction perpendicular to the surface on which the aspherical lenses 13 are formed. The top of the diffusion profile can be made flat by ensuring that the positional variation in the direction perpendicular to the lens formation surface, i.e., the value (σ/μ) obtained by dividing the depth variation (1σ) by the average value (μ), is greater than 0.027, preferably 0.041 or greater, more preferably 0.054 or greater, even more preferably 0.068 or greater, and even more preferably 0.081 or greater, but less than 0.200.

<Regularity of in-plane arrangement of aspherical lenses>
Next, we will consider the regularity of the lens arrangement on the first surface 101 of the substrate 11. A regular arrangement of the aspherical lenses 13 within the two-dimensional plane is advantageous for cutoff characteristics, as will be described later. The cutoff characteristics of the microlens array 10 refer to the abruptness of the change in whether light is diffused or blocked in a predetermined direction.

11 is a simulation diagram of the diffusion profile when the pitch variation of the aspherical lens 13 is changed. As described above, when the average pitch in each of the X and Y directions is set to 1, the pitch variation indicates the absolute value of the maximum possible difference in the randomly distributed relative pitch ratios over the entire pitch in a predetermined region (region excluding 25% at both ends) of the first surface 101, and this is denoted as σ _P (see FIG. 2 ). In the simulation, the FOV in the Y direction is set to 45°, and the pitch variation in the Y direction is changed to 0.0%, 5.0%, 7.5%, 10.0%, 15.0%, 20.0%, and 25.0% to calculate the diffusion characteristics.

A partial area on the positive side of the diffusion profile is shown enlarged. The greater the in-plane pitch variation, the more gradual the rise or fall of the diffusion profile becomes, and the worse the cutoff characteristics become. In particular, when the in-plane pitch variation exceeds 10.0%, the steepness of the base of the diffusion profile is lost.

Figure 12 is a plot of the cutoff band as a function of pitch variation from the simulation results of Figure 11. In this embodiment, the cutoff band is the angular width required for the intensity of the diffusion profile to change to a predetermined level. A more precise definition of the cutoff band will be explained with reference to Figure 13.

Figure 13 is a diagram explaining how to determine the cutoff band. In this embodiment, the "cutoff band" is the angular width required for the relative intensity to change between 0.2 and 0.8. As defined in the study of depth variation in the aspherical lens 13, the relative intensity of the diffusion profile is the intensity when the average intensity (light amount) in the diffusion angle range of -10° to +10° is normalized to 1. The angular widths at which the relative intensity changes from 0.2 to 0.8 or from 0.8 to 0.2 are found on both the negative and positive sides of the diffusion profile, and the larger of the two angular widths is taken as the cutoff band.

The smaller the cutoff band, the steeper the rise and fall of the diffusion profile and the better the cutoff characteristics. The larger the cutoff band, the gentler the rise and fall of the diffusion profile and the worse the cutoff characteristics.

Returning to Figure 12, the greater the pitch variation of the aspherical lens, the larger the cutoff band (°), so it is desirable to keep the pitch variation small and maintain a small cutoff band. In the simulation example of Figure 12, the pitch variation of the aspherical lens 13 is desirably less than 0.075 (or 7.5%), at which point the cutoff band begins to increase sharply, and more desirably 0.050 (or 5.0%) or less.

Figure 14 shows the results of a two-dimensional electromagnetic field simulation when the pitch variation of the aspherical lenses 13 of the microlens array 10 is changed only in the Y direction. When the pitch variation in the Y direction is 0.0% and 5.0%, there is little bleeding in the vertical direction of the paper, i.e., the Y direction. When the pitch variation in the Y direction is 10%, bleeding appears in the Y direction. The greater the pitch variation, the more pronounced the bleeding in the Y direction. Prominent bleeding means that the base of the diffusion profile is sagging. Because the axes of the X and Y directions can be defined arbitrarily, the same applies to pitch variation in the X direction.

Referring to the simulation results in Figures 12 and 14, it can be seen that it is desirable for the in-plane pitch variation of the microlens array 10 to be less than 7.5%, and more preferably 5.0% or less.

Up to this point, we have examined the cutoff band as a cutoff characteristic based on simulations. Next, we will examine the cutoff characteristics based on an actually fabricated sample.

Five types of samples were created by setting the reference pitch in the Y direction to 84 μm and varying only the pitch variation. Sample a, with a pitch variation in the Y direction of 0.0%, was designed so that the centers of the aspherical lenses 13 were all equally spaced in the Y direction. Sample b, with a pitch variation in the Y direction of 5.0%, was designed so that the pitch was randomly distributed in the Y direction between a minimum pitch of 79.8 μm and a maximum pitch of 88.2 μm, with a maximum variation in the Y direction of 0.05.

Similarly, sample c with a pitch variation in the Y direction of 7.5%, sample e with a pitch variation in the Y direction of 15.0%, and sample g with a pitch variation in the Y direction of 25.0% were fabricated. Samples a, b, c, e, and g correspond to the pitch variations a, b, c, e, and g in the simulation of Figure 11.

Figure 15 plots the relative pitch ratios of approximately 40 rows (approximately 40 aspherical lenses lined up in the Y direction) of the five types of samples fabricated. The intended pitch variation in the Y direction was achieved in all samples. Diffusion characteristics were measured using these samples.

Figure 16A shows the measured diffusion characteristics of five samples. Figure 16B shows the pitch variations a, b, c, e, and g corresponding to Figure 16A extracted from the simulation results of Figure 11. When comparing the measured results of Figure 16A with the simulation results of Figure 16B, the overall diffusion profile trends are the same. In Figure 16A, the cutoff band cannot be calculated as accurately as in Figure 16B due to measurement resolution, and there is a concern that significant differences in cutoff characteristics due to pitch variations may be hidden by the above-mentioned uncertainty. Therefore, a new index of cutoff characteristics, a slope [deg/a.u.] that allows for fair comparisons, is introduced using the closest approximation of a specified relative intensity and diffusion angle.

Figure 17 is an enlarged view of the diffusion characteristics in Figure 16A, covering the diffusion angle range from -28° to -17°. The slope is determined for each of the measured results for the five samples.

Figure 18 is a diagram explaining how to determine the slope. The average intensity (light amount) within the diffusion angle range of -10° to +10° is normalized to 1. On the negative and positive sides of the diffusion profile, the points with the closest relative intensity values of 0.2 and 0.8 are connected, and the slope of the line is found. The smaller of the absolute values of the two found slopes is taken as the slope.

FIG. 19 shows the analysis results of the diffusion characteristics of samples a, b, c, e, and g in FIG. 17. From the measured values of the diffusion characteristics of each sample,
The closest relative intensity of 0.2 (1),
The diffusion angle at the closest value (1),
The closest relative intensity of 0.8 (2),
The diffusion angle at the closest value (2),
The slope is the difference between the closest relative intensities of 0.2 and 0.8 divided by the difference in the diffusion angle at each closest value.

There is no significant difference between when the in-plane pitch variation is 0.0% and when it is 5.0%. When the pitch variation exceeds 7.5%, the cutoff band increases, the slope decreases, and the cutoff characteristics clearly deteriorate. Therefore, in order to make the slope of the diffusion profile greater than 0.139 (for example, 0.14 or greater), it is desirable to keep the in-plane pitch variation below 7.5%.

As such, based on actual measurement data, it is confirmed that in-plane pitch variation of less than 7.5% is preferable, and 5.0% or less is even more preferable. This result is consistent with the simulation results in Figure 12.

11 to 19, the microlens array 10 is examined in the Y direction, but the same results are obtained in the X direction, and cutoff characteristics similar to those in the Y direction are obtained. In the microlens array 10, appropriate cutoff characteristics can be obtained by setting the pitch variation (σ _P ) relative to the average of the entire pitch in the area excluding 25% at both ends to less than 7.5% in both the X and Y directions. When light is diffused over a rectangular area, a projected image with little blur is obtained.

As mentioned above, the microlens array 10 has sufficient variation in the depth or height of the aspherical lenses 13. In addition to the variation in the depth or height direction, the pitch variation relative to the average pitch of a specified area of the aspherical lenses 13 within the plane (area excluding 25% at both ends) is kept to less than 7.5%, thereby suppressing diffraction bright spots and achieving a microlens array 10 with good cutoff characteristics.

<Application examples of microlens arrays>
20 is a schematic diagram of a projection device 20 to which the microlens array 10 of the embodiment is applied. The projection device 20 includes a light source 21, a lens 22, and a microlens array 10. The light source 21 is, for example, a light-emitting diode (LED). Light emitted from the light source 21 is collimated by the lens 22 to parallel light and then enters the microlens array 10. The microlens array 10 is arranged so that a first surface 101 on which an array of concave aspherical lenses 13 is formed faces the light incident side. In this example, the microlens array 10 is used as a diffuser.

The microlens array 10 diffuses the incident parallel light in the X and Y directions over a predetermined FOV and projects it onto the screen 25. If a laser light source is used instead of an LED as the light source, the collimating lens 22 may be omitted. To obtain a color projection image, a microlens array 10 may be arranged for each of the red, green, and blue light sources, and the light emitted from each microlens array 10 may be combined using a prism or the like and projected onto the screen 25.

In the microlens array 10, the center positions of the multiple aspherical lenses 13 (see Figure 2) formed on the first surface 101 vary in the direction perpendicular to the first surface 101 (Z direction), but are arranged in a generally regular pattern within the plane of the first surface 101 (X-Y plane). In the area excluding the two 25% of the aspherical lenses 13 on the first surface 101, pitch variation is suppressed to less than 7.5%, and more preferably 5.0% or less, resulting in a projected image with a steep diffusion profile and minimal blurring. Furthermore, the variation in the Z direction of the aspherical lenses 13 suppresses diffracted bright spots, resulting in a uniform intensity distribution.

The present invention has been described above based on specific configuration examples, but the present invention is not limited to the above configuration examples. Each aspherical lens 13 of the microlens array 10 may be formed as a convex lens with a parabolic surface. In this case, the apex positions of the convex lenses are varied to suppress in-plane pitch variation to less than 7.5%. A microlens array including an arrangement of convex aspherical lenses 13 may be made of resin or the like using the microlens array 10 produced in the embodiment as a mold.

The microlens array 10 of this embodiment can be applied not only to projection devices, but also to lighting devices, imaging systems, etc. Wavelength selectivity can be achieved by tuning the pitch itself while suppressing pitch variation of the aspherical lenses within the plane. In this case, light of a specific wavelength can be diffused, making it suitable for application to color projection devices.

10, 10A Microlens array 11 Substrate 13 Aspherical lens 14 Center 20 Projection device 21 Light source 22 Lens 25 Screen 101 First surface

Claims (9)

A glass substrate that is transparent to the wavelength used,
a microlens array having a plurality of aspherical lenses formed on the first surface of the glass substrate,
the center of the aspherical lens is at two or more different positions in a direction perpendicular to the first surface,
the plurality of aspherical lenses include a plurality of aspherical lenses having different conic coefficients and/or different radii of curvature;
a value σ/μ obtained by dividing the value of the depth variation σ by an average value μ when the center positions of the aspherical lens in a direction orthogonal to the first surface follow a normal distribution is 0.068 or more and less than 0.200,
an absolute value σp of the maximum possible difference in relative pitch ratio is less than 0.075 when the average of all pitches in a region excluding a total of 25% from both ends of the aspherical lenses arranged in an arbitrary one-dimensional direction within the first surface is set to 1;
The diffusion pattern obtained in the optical measurement is observed on a screen 103 mm away, the minimum measurement resolution is 0.5 μm, the measurement light is parallel light with a wavelength of 940 nm, and the beam diameter is 3 mm. When the average light amount in the diffusion angle range of −10 degrees to +10 degrees of the diffused light of the microlens array is normalized to 1 and the relative intensity is plotted as a function of the diffusion angle, the maximum relative intensity in the angle range where the value of the relative intensity is 0.5 or more is 1.0 or more and 1.2 or less.
Microlens array. the aspherical lenses are arranged in a square lattice;
the absolute value σp is less than 0.075 in each of a first direction and a second direction that are orthogonal to each other within the plane of the first surface;
The microlens array of claim 1 . The absolute value σp is 0.05 or less.
The microlens array according to claim 1 or 2. The depth variation σ is equal to or greater than 2.40 μm and less than 6.00 μm.
The microlens array according to claim 1 . the slope between the relative intensity values of 0.2 and 0.8 is 0.14 or greater;
The microlens array according to claim 1 . the aspherical lenses provided on the first surface are continuous with each other without any flat surface between adjacent aspherical lenses;
The microlens array according to claim 1 . A diffusion plate for a projection device, comprising the microlens array according to claim 1 . A light source and
the microlens array according to claim 1 provided on the output side of the light source;
The microlens array is a projection device that diffuses and projects the light emitted from the light source. The aspherical lens is a concave lens, and the microlens array is disposed with the first surface facing the light source.
9. The projection device according to claim 8 .