Pattern exposure equipment

Abstract

SOLUTION: A pattern exposure apparatus comprises: a first light source device which emits a first beam; a second light source device which emits a second beam; a plurality of acoustic optical modulation elements which allow the first beam and the second beam to pass in series; a plurality of drawing units which one-dimensionally scans spot light being a diffraction beam of the first beam and the second beam to draw a pattern on a substrate; and a synthesizing optical system which synthesizes the first beam and the second beam such that the first beam and the second beam cross at a predetermined cross angle and pass in the acoustic optical modulation element of a first stage, where the first spot light by the first beam and the second spot light by the second beam are projected at mutually different positions on the substrate from each of the plurality of drawing units.SELECTED DRAWING: Figure 4

Description

[Technical Field]

[0001] The present invention relates to a pattern exposure apparatus that exposes a pattern of an electronic device or the like on a substrate with a drawing beam modulated in accordance with drawing data. [Background technology]

[0002] Conventionally, in the process of manufacturing fine electronic devices on a substrate, a photolithography process has been carried out, including an exposure process in which an exposure beam (such as a light beam or an electron beam) corresponding to the pattern of the electronic device (a pattern that defines the shape of a wiring layer, an electrode layer, a semiconductor layer, an insulating layer, etc.) is irradiated onto a photoresist layer on the substrate, and a development process in which the exposed substrate is developed to reveal a pattern consisting of remaining and removed parts of the photoresist layer. Known exposure devices used in this exposure process include a system that uses a photomask on which the pattern to be exposed is fixedly formed, and a maskless system that dynamically modulates the intensity of the exposure beam based on drawing data (CAD data) corresponding to the pattern to be exposed.

[0003] Japanese Patent Laid-Open Publication No. 2002-196270 discloses a pattern drawing device as a maskless exposure device, which modulates a laser light flux (beam) from a laser light source by an acousto-optic modulator, repeatedly deflects the modulated beam one-dimensionally by each reflecting surface of a rotating polygon mirror, and performs one-dimensional scanning while imaging the beam deflected by the polygon mirror into a spot light on a surface to be scanned via an imaging optical system including an fθ lens. Furthermore, Japanese Patent Laid-Open Publication No. 2002-196270 also discloses the provision of a light flux position detector for measuring fluctuations such as the inclination of the traveling direction of the beam emitted from the laser light source or the lateral shift of the emitted beam, and an optical member for correcting the deviation of the scanning position of the spot light due to the fluctuations.

[0004] In the pattern drawing device described in JP 2002-196270 A, a beam from one laser light source is supplied to one drawing unit including one polygon mirror and an fθ lens, but in a multi-drawing head type exposure device in which multiple drawing units are arranged so that a pattern drawn by a drawing line (scanning line) drawn by the main scanning of a spot light is seamlessly exposed in the main scanning direction, multiple laser light sources may be used. In this case, in addition to deviations in the exposure position of the pattern due to individual variations in the beams emitted from each of the multiple laser light sources, it is necessary to reduce the occurrence of seamless errors due to variations in the relative positions and inclinations of the beams from each of the multiple laser light sources. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Laid-Open No. 2002-196270 Summary of the Invention

[0006] A first aspect of the present invention is a pattern exposure apparatus comprising: a first drawing unit that draws a pattern on a substrate using a first beam from a first light source device; and a second drawing unit that draws a pattern on the substrate using a second beam from a second light source device, the apparatus comprising: a first optical splitter that is provided in the optical path of the first beam from the first light source device to the first drawing unit and splits a portion of the first beam as a first measurement beam; a second optical splitter that is provided in the optical path of the second beam from the second light source device to the second drawing unit and splits a portion of the second beam as a second measurement beam; a fluctuation detection optical unit that receives the first measurement beam and the second measurement beam and detects relative position fluctuation or relative tilt fluctuation between the first beam and the second beam; a first light guide system that forms an optical path for the first measurement beam from the first light splitter to the fluctuation detection optical unit; and a second light guide system that forms an optical path for the second measurement beam from the second optical splitter to the fluctuation detection optical unit.

[0007] A second aspect of the present invention is a pattern exposure apparatus comprising: a first light source device that emits a first beam; a second light source device that emits a second beam; a plurality of acousto-optical modulation elements that pass the first beam and the second beam in series; and a plurality of drawing units that use diffracted beams of the first beam and the second beam generated from the plurality of acousto-optical modulation elements as spot light and scan the spot light one-dimensionally to draw a pattern on a substrate. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a front view showing a schematic overall configuration of a pattern exposure apparatus according to a first embodiment. [Figure 2] 1. FIG. 3 is a perspective view showing a schematic internal configuration of the imaging unit MU1 as a representative of the imaging units MU1 to MU6 shown in FIG. [Figure 3] 1. FIG. 2 is a perspective view showing the arrangement of drawing lines SL1 to SL6 set on a sheet substrate P supported by a rotating drum DR shown in FIG. 1, and the arrangement of alignment systems ALGn (ALG1 to ALG5). [Figure 4] 2 is a top view of the optical configuration in the beam switching unit BDU shown in FIG. 1, viewed in a plane parallel to the XY plane. [Figure 5] FIG. 5 is a perspective view showing the arrangement of optical members in the vicinity of the optical path from the laser light source 10B shown in FIG. 4 to the first acousto-optical element (AOM) AM6 for switching. [Figure 6] 5 is a perspective view showing a specific positional relationship between the triangular mirror 33 and the detection unit 34 shown in FIG. 4. FIG. [Figure 7] 7 is a diagram schematically showing the state of beams MBa and MBb projected onto the imaging surface of the first imaging element 34C shown in FIG. 6. FIG. [Figure 8] 7 is a diagram schematically illustrating the state of spot light of beams MBa and MBb projected onto the imaging surface of second imaging element 34G shown in FIG. 6. FIG. [Figure 9] FIG. 6 is a perspective view showing an example of a specific optical configuration of correction optical system 11B shown in FIGS. 4 and 5. [Figure 10] 6 is a perspective view illustrating the parallel shift of the beam LBb in the optical path from the laser light source 10B to the first-stage acousto-optic modulation element AM6 shown in FIG. 5. FIG. [Figure 11] 10B is a diagram showing in an exaggerated manner the states of the drawing beams LB2, LB4, and LB6 directed toward the even-numbered drawing units MU2, MU4, and MU6, respectively, when the beam LBb from the laser light source 10B is shifted in parallel in the −Y direction. [Figure 12] 10B is a diagram showing in an exaggerated manner the states of the drawing beams LB2, LB4, and LB6 directed toward the even-numbered drawing units MU2, MU4, and MU6, respectively, when the beam LBb from the laser light source 10B is shifted in parallel in the +Z direction. [Figure 13] 10 is an exaggerated view showing the states of the drawing beams LB2, LB4, and LB6 directed toward the even-numbered drawing units MU2, MU4, and MU6, respectively, when the beam LBb incident on the first-stage acousto-optic modulation element AM6 is tilted. [Figure 14] FIG. 14 is an exaggerated view showing the states of the drawing beams LB2, LB4, and LB6 directed toward the even-numbered drawing units MU2, MU4, and MU6, respectively, when the beam LBb incident on the first-stage acousto-optic modulation element AM6 is tilted in a direction perpendicular to the direction described in FIG. 13. [Figure 15] Figures 15A to 15C are diagrams explaining the incidence state and diffraction efficiency of beam LBb from laser light source 10B incident on the first-stage acousto-optical modulation element AM6 of the beam switching unit BDU, where Figure 15A is a diagram of acousto-optical modulation element AM6 viewed in the XZ plane of the Cartesian coordinate system XYZ, Figure 15B is a diagram of acousto-optical modulation element AM6 viewed in the XY plane of the Cartesian coordinate system XYZ, and Figure 15C is a graph schematically showing the change in intensity of beam LB6 (first-order diffracted beam) with respect to the incidence angle θz in the diffraction direction of beam LBb incident on acousto-optical modulation element AM6 and the incidence angle θy in the non-diffraction direction. [Figure 16] FIG. 11 is a perspective view showing the state of two beams on an optical path from an acousto-optic modulation element AM6 at the first stage of a beam switching unit BDU to an epi-illumination mirror IM6 according to the second embodiment. [Figure 17]17 is an exaggerated view showing the state of two beams LB6a and LB6b passing through the optical path from the reflected light mirror IM6 shown in FIG. 16 through the optical path adjusting section BV6 to the lens LGa in the imaging unit MU6. [Figure 18] FIG. 10 is a diagram showing an example of an optical path for guiding beams from four laser light sources 10A1, 10A2, 10B1, and 10B2 to first-stage acousto-optic modulation elements AM6 and AM1, which are applied to the second embodiment. [Figure 19] 19 is a diagram schematically showing how two spot lights SPa and SPb projected onto a sheet substrate P are scanned in the second embodiment shown in FIGS. 16 to 18. FIG. [Figure 20] FIG. 10 is a perspective view showing a modified example of the imaging unit MU1 (and the same for MU2 to MU6) shown in FIG. [Figure 21] 18 is a perspective view showing a modification of a part of the optical path adjusting unit BV6 shown in FIG. [Figure 22] FIG. 10 is a diagram of a modified example that schematically illustrates the optical paths of four measurement beams when four laser light sources are used. [Figure 23] 23 is a perspective view showing the positional relationship of a fluctuation optical detection system (a triangular mirror 33' and a detection unit 34) in the case of the modified example of FIG. 22. FIG. DETAILED DESCRIPTION OF THE INVENTION

[0009] A pattern exposure apparatus according to an aspect of the present invention will be described in detail below with reference to the accompanying drawings, showing preferred embodiments. It should be noted that the aspects of the present invention are not limited to these embodiments and include various modifications or improvements. In other words, the components described below include those that would be easily conceivable to a person skilled in the art and those that are substantially identical, and the components described below can be combined as appropriate. Furthermore, various omissions, substitutions, or modifications of the components can be made without departing from the spirit of the present invention.

[0010] [First embodiment] Fig. 1 is a diagram showing a schematic overall configuration of a pattern exposure apparatus according to a first embodiment. As shown in Fig. 1, the pattern exposure apparatus of this embodiment exposes various patterns corresponding to electronic devices (display devices, wiring devices, sensor devices, etc.) onto a photosensitive layer coated on a flexible, long sheet substrate P (hereinafter also simply referred to as substrate P) using a maskless method by scanning with spot light. Such pattern exposure apparatuses are disclosed, for example, in WO 2015 / 152218, WO 2015 / 166910, WO 2016 / 152758, WO 2017 / 057415, etc.

[0011] 1, the pattern exposure apparatus EX of this embodiment is installed on the floor of an installation location (such as a factory) parallel to the XY plane of an XYZ Cartesian coordinate system in which the Z axis is the direction of gravity. The exposure apparatus EX includes a rotating drum DR for stably supporting a sheet substrate P and transporting it at a constant speed, six imaging units MU1 to MU6 for imaging a pattern on a photosensitive layer of the sheet substrate P, a beam switching unit BDU for switching and distributing beams LBa and LBb from two laser light sources 10A and 10B to each of the imaging units MU1 to MU6 in a time-division manner, optical path adjustment units BV1 to BV6 for making the beams LB1, LB2, ... distributed by the beam switching unit BDU incident on each of the imaging units MU1 to MU6 at an adjusted angle, and a plurality of alignment systems ALGn (where n = 1 to 5) for detecting alignment marks on the sheet substrate P via an objective lens system OBL.

[0012] The rotating drum DR has a cylindrical outer circumferential surface with a constant radius from a rotation centerline AXo parallel to the Y axis of the XY plane, and a shaft Sft protruding coaxially from both ends of the rotating drum DR in the Y direction. The sheet substrate P is closely supported in the longitudinal direction along approximately half of the outer circumferential surface of the rotating drum DR and is transported at a constant speed in the longitudinal direction by the uniform rotation of the rotating drum DR due to the rotational torque of a rotation drive motor (not shown). The base material of the sheet substrate P is a resin material such as PET (polyethylene terephthalate) film, PEN (polyethylene naphthalate) film, or polyimide film. However, other materials such as glass formed into an ultrathin sheet of 100 μm or less and made flexible, metal material such as stainless steel formed into a thin sheet by rolling, or paper material containing cellulose nanofibers may also be used.

[0013] The multiple imaging units MU1 to MU6 are arranged in a line in the Y direction in the space above the rotating drum DR, but each of the odd-numbered imaging units MU1, MU3, and MU5 and each of the even-numbered imaging units MU2, MU4, and MU6 are arranged symmetrically with respect to a central plane Cp that is parallel to the YZ plane and includes the rotation center line AXo when viewed in the XZ plane. Each of the odd-numbered imaging units MU1, MU3, and MU5 is attached to the column frame BF of the apparatus main body so that the extension of the center line of the beam LB1 (LB3, LB5) projected onto the sheet substrate P is directed toward the rotation center line AXo and is inclined at an angle -θu from the central plane Cp when viewed in the XZ plane. Similarly, each of the even-numbered drawing units MU2, MU4, and MU6 is attached to the column frame BF of the device body so that the extension of the center line of the beam LB2 (LB4, LB6) projected onto the sheet substrate P is directed toward the rotation center line AXo and is tilted at an angle +θu from the center plane Cp when viewed in the XZ plane.

[0014] 1 to the column frame BF is disclosed, for example, in WO 2016 / 152758, in which each of the odd-numbered imaging units MU1, MU3, and MU5 is provided rotatable around a rotation axis LE1 (LE3, LE5) within a small angular range (for example, ±several degrees or less), and each of the even-numbered imaging units MU2, MU4, and MU6 is provided rotatable around a rotation axis LE2 (LE4, LE6) within a small angular range (for example, ±several degrees or less). An extension line of each of the rotation axes LE1 (LE3, LE5) and LE2 (LE4, LE6) is arranged so as to be perpendicular to the rotation center line AXo and to pass through the midpoint in the Y direction of a imaging line formed on the sheet substrate P by the spot light beams LB1 to LB6 scanned from each of the imaging units MU1 to MU6.

[0015] The internal configuration of each of the imaging units MU1 to MU6 includes multiple mirrors, multiple lenses, a rotating polygon mirror PM, and a telecentric fθ lens system FT, as disclosed in, for example, International Publication No. 2016 / 152758 or International Publication No. 2019 / 082850. The center lines of the beams LB1 to LB6 emitted from the optical path adjusting units BV1 to BV6 and incident on the corresponding imaging units MU1 to MU6 are set to be coaxial with the rotation axes LE1 to LE6, respectively. Note that within the imaging units MU1 to MU6, a Cartesian coordinate system XtYtZt is set, which is defined by the Zt axis set parallel to each of the rotation axes LE1, LE2, etc., and the Xt and Yt axes perpendicular to the Zt axis. Therefore, the Yt axis of the Cartesian coordinate system XtYtZt is parallel to the Y axis of the Cartesian coordinate system XYZ, and the Cartesian coordinate system XtYtZt is tilted by an angle θu (angle -θu or angle +θu) around the Y axis with respect to the XY plane of the Cartesian coordinate system XYZ.

[0016] In the beam switching unit BDU shown in FIG. 1, the beam LBa from the laser light source 10A is distributed sequentially to one of the odd-numbered imaging units MU1, MU3, and MU5 in a time-division manner, while the beam LBb from the laser light source 10B is repeatedly distributed sequentially to one of the even-numbered imaging units MU2, MU4, and MU6 in a time-division manner. As disclosed in International Publication No. 2016 / 152758, beam switching within the beam switching unit BDU is performed by an acousto-optical modulator (AOM). In this embodiment, relative fluctuations (lateral shift error and tilt error) between the beam LBa from the laser light source 10A and the beam LBb from the laser light source 10B are measured within the beam switching unit BDU, as will be described in detail later. The laser light sources 10A and 10B, as well as the acousto-optical modulator (AOM) and various optical components (mirrors, lenses, etc.) that make up the beam switching unit BDU, are mounted on an optical surface plate OBP.

[0017] FIG. 2 is a perspective view showing a schematic internal configuration of the imaging unit MU1, which is representative of the imaging units MU1 to MU6 shown in FIG. 1. The configuration of the imaging unit MU1 in FIG. 2 is almost the same as the configuration disclosed in International Publication No. 2016 / 152758, and will be briefly described below. Beam LB1 (a parallel beam with a diameter of 1 mm or less) from the optical path adjustment unit BV1 is incident on mirror M10 coaxially with rotation axis LE1, which extends parallel to the Zt axis. After being reflected at 90 degrees, it passes through a beam expander formed by lenses LGa and LGb, and is then reflected at 90 degrees by mirror M11 and enters polarizing beam splitter PBS. Since beam LB1 is linearly polarized in the Zt axis direction (S-polarized light), it is efficiently reflected by polarizing beam splitter PBS, reflected at 90 degrees by mirror M12, traveling in the −Zt direction, and reflected at 90 degrees by mirror M13, traveling in the +Xt direction. The beam LB1 reflected by the mirror M13 passes through a quarter-wave (λ / 4) plate QP and a first cylindrical lens CYa, and is then reflected by the mirror M14 to reach one reflecting surface Rp1 of the rotating polygon mirror PM.

[0018] The beam LB1 reflected by the reflecting surface Rp1 of the rotating polygon mirror PM is deflected within the XtYt plane by the rotation of the rotating polygon mirror PM and enters a telecentric fθ lens system FT having an optical axis AXf1 parallel to the Xt axis. A mirror M15 is positioned immediately after the fθ lens system FT to bend the optical axis AXf1 by 90 degrees, and the beam LB1 emerging from the fθ lens system FT is reflected by the mirror M15 by 90 degrees so that it becomes parallel to the Zt axis. A second cylindrical lens CYb is positioned between the mirror M15 and the sheet substrate P, and the beam LB1 emerging from the fθ lens system FT is focused as a spot light SP on the sheet substrate P. The spot light SP is then scanned one-dimensionally by the rotation of the rotating polygon mirror PM to form a drawing line (scanning line) SL1 parallel to the Yt axis (Y axis).

[0019] In the imaging unit MU1 shown in Figure 2, the lens system LGc and photoelectric sensor DT, located on the opposite side of the polarizing beam splitter PBS from the mirror M12, receive the reflected light from the sheet substrate P due to the projection of the light spot SP. By analyzing the waveform of the photoelectric signal from the photoelectric sensor DT, positional information of the pattern already formed on the sheet substrate P can be obtained. Also, in Figure 2, the surface OPa is set to be the rear focus of the lens LGa and the front focus of the lens LGb, and the beam LB1 is focused to a beam waist with a diameter of several tens of microns at the surface OPa. Therefore, the beam LB1 passing through the lens LGb becomes a parallel beam with a diameter of several millimeters or more. The first cylindrical lens CYa and the second cylindrical lens CYb, in cooperation with the fθ lens system FT, correct the positional fluctuation of the light spot SP (imaging line SL1) in the Xt direction due to differences in the tilt of each reflective surface of the rotating polygon mirror PM.

[0020] FIG. 3 is a perspective view showing the arrangement of drawing lines SL1 to SL6 on a sheet substrate P supported by a rotating drum DR and the arrangement of alignment systems ALGn (ALG1 to ALG5). In FIG. 3, scale disks RSD of an encoder measurement system are fixed to shafts Sft at both ends of the rotating drum DR, coaxially with the rotation center line AXo. The outer circumferential surface of the scale disk RSD is formed with graduations SD1 and SD2, on which grating lines are engraved at a constant pitch along the circumferential direction. The circumferential positional changes of the graduations SD1 and SD2 are measured with submicron resolution by encoder heads EH1, EH2, and EH3, respectively, located at three positions along the circumferential direction. Furthermore, a ring-shaped reference surface Rst is formed on the side end surface of the scale disk RSD, which is parallel to the XZ plane. Micro-displacements of the reference surface Rst in the Y direction are measured with submicron resolution by displacement sensors YS1, YS2, and YS3, respectively, located at three positions along the circumferential direction.

[0021] Of the drawing lines SL1 to SL6 formed on the sheet substrate P by the drawing units MU1 to MU6, the odd-numbered drawing lines SL1, SL3, and SL5 located upstream in the conveyance direction of the sheet substrate P are arranged parallel to the rotation center line AXo (Y-axis) and spaced at regular intervals (approximately the length of the drawing line) in the Y direction. Similarly, the even-numbered drawing lines SL2, SL4, and SL6 located downstream in the conveyance direction of the sheet substrate P are arranged parallel to the rotation center line AXo (Y-axis) and spaced at regular intervals (approximately the length of the drawing line) in the Y direction. The patterns drawn by the drawing lines SL1 to SL6 are exposed so as to be joined together as the sheet substrate P is conveyed.

[0022] Five alignment systems ALG1 to ALG5 are arranged at predetermined intervals in the Y direction as alignment system ALGn upstream of the odd-numbered drawing lines SL1, SL3, and SL5 with respect to the conveyance direction of the sheet substrate P. Alignment system ALG1 detects an alignment mark formed near the edge of the sheet substrate P on the -Y direction side via an objective lens system OBL and a leading edge mirror MR, and alignment system ALG5 detects an alignment mark formed near the edge of the sheet substrate P on the +Y direction side via a similar objective lens system OBL and a leading edge mirror MR. The detection areas (detection fields) of each of alignment systems ALG1 to ALG5 are arranged in a row in the Y direction, and the circumferential orientation of the detection areas as viewed from the rotation center line AXo is set to match the circumferential orientation of the reading position of encoder head EH3 as viewed from the rotation center line AXo.

[0023] Furthermore, the circumferential orientations of the odd-numbered imaging lines SL1, SL3, and SL5 as viewed from the rotation center line AXo are set to coincide with the circumferential orientation as viewed from the rotation center line AXo of the reading position of encoder head EH1, and the circumferential orientations of the even-numbered imaging lines SL2, SL4, and SL6 as viewed from the rotation center line AXo are set to coincide with the circumferential orientation as viewed from the rotation center line AXo of the reading position of encoder head EH2. Furthermore, as shown in FIG. 2 above, the optical axis AXf1 of the fθ lens system FT of imaging unit MU1 is bent by mirror M15 and set to be perpendicular to the tangent plane that contacts the surface of sheet substrate P at the position of imaging line SL1. Therefore, the extension of the optical axis AXf1 is directed toward the rotation center line AXo, and the intersection of the optical axis AXf1 with the sheet substrate P is the midpoint of the length of imaging line SL1 in the Y direction (main scanning direction).

[0024] Next, the schematic configuration of the beam switching unit BDU shown in FIG. 1 will be described with reference to FIGS. 4 and 5. In this embodiment, the beam switching unit BDU is provided with a function (mechanism) for measuring fluctuations in the beams LBa and LBb from the two laser light sources 10A and 10B. FIG. 4 is a top view of the configuration within the beam switching unit BDU in a plane parallel to the XY plane, and FIG. 5 is a perspective view showing the arrangement of optical components near the optical path from the laser light source 10B in FIG. 4 to the first acousto-optical element (AOM) AM6 used for switching. All optical components, including the laser light sources 10A and 10B in FIG. 4, are mounted on the optical base OBP shown in FIG. 1. The laser light sources 10A and 10B are, for example, fiber amplifier laser light sources such as those described in International Publication Nos. WO 2015 / 166910 and WO 2018 / 164087. Therefore, in this embodiment as well, the intensity modulation based on the drawing data of the beams LB1 to LB6 projected onto the sheet substrate P from each of the drawing units MU1 to MU6 is performed by high-speed switching of the seed light beams in the infrared wavelength range of each of the laser light sources 10A and 10B using an electro-optical element (EO element) or the like that responds to a clock signal of 100 MHz or more, for example, 400 MHz.

[0025] The optical paths of beams LBa and LBb from laser light sources 10A and 10B will be described below, but for convenience of explanation, the optical path of beam LBb from laser light source 10B will be described first. Beam LBb from laser light source 10B has a wavelength in the ultraviolet wavelength range (e.g., 400 nm or less) that exposes the photosensitive layer on sheet substrate P, and is emitted in the -X direction as a parallel beam with a diameter of approximately 1 mm. Beam LBb from laser light source 10B is incident on beam splitter 30B, which has high transmittance and low reflectance of several percent to 10%, and the transmitted beam LBb is guided to the even-numbered imaging units MU2, MU4, and MU6. Beam MBb reflected by beam splitter 30B is used to measure beam fluctuations via mirror 31B, lens GL1b, mirror 32B, and lens GL2b, as will be described in detail below.

[0026] After passing through beam splitter 30B, beam LBb passes through correction optical system 11B, which fine-tunes the tilt of beam LBb's traveling direction and slightly shifts it laterally in a plane perpendicular to the beam, before entering beam splitter 12B, which has a transmittance of less than a few percent. The beam that passes through beam splitter 12B is received by light intensity monitor 13B, which measures the intensity of beam LBb from laser light source 10B. Beam LBb is reflected in the -Y direction by beam splitter 12B, where it is converted into a parallel beam by reduction relay optical system 14B, with its beam diameter reduced to about half (0.5 mm). It is then redirected by mirror system 15B to a path traveling in the +X direction, where it enters acousto-optic modulator AM6, the first stage of switching, as a parallel beam.

[0027] When the acousto-optic modulator AM6 is in the off state (non-deflected state), the beam LBb passes through the acousto-optic modulator AM6 as is and passes through the condenser lens 16B, collimator lens 17B, and mirror 18B to be incident as a parallel beam on the second-stage switching acousto-optic modulator AM4. In the XY plane, an epi-mirror IM6 with its reflective surface tilted 45 degrees with respect to the XY plane is located at the rear focal position of the condenser lens 16B. The epi-mirror IM6 is positioned so that it reflects in the -Z direction only the first-order diffracted beam generated when the acousto-optic modulator AM6 is in the on state (deflected state), and does not irradiate the undiffracted zeroth-order beam (part of the beam LBb).

[0028] Here, the optical path from laser light source 10B to epi-illumination mirror IM6 will be described in more detail with reference to FIG. 5. As shown in FIG. 5, beam LBb emitted from laser light source 10B and passing through correction optical system 11B and reduction relay optical system 14B is reflected in the −Z direction by mirror 15B1, which constitutes mirror system 15B, and then reflected in the +X direction by mirror 15B2. When beam LBb reflected by mirror 15B2 enters acousto-optical modulator AM6, which is in the on state, beam LB6 is generated from acousto-optical modulator AM6 as a first-order diffracted beam deflected at a constant diffraction angle in the −Z direction. Acousto-optical modulator AM6 is positioned at the front focal position of condenser lens 16B so as to satisfy the conditions for Bragg diffraction, and is positioned so that beam LBb (or a zeroth-order beam) that passes directly through acousto-optical modulator AM6 passes simultaneously with the optical axis of condenser lens 16B. With this configuration, beam LB6, which is a first-order diffracted beam that has passed through focusing lens 16B, becomes parallel to the optical axis of focusing lens 16B, passes through a position that is eccentric from the optical axis in the -Z direction, reaches epi-illumination mirror IM6, and is reflected in the -Z direction.

[0029] Furthermore, before entering the condenser lens 16B, beam LBb (or the zeroth-order beam) and beam LB6 as the first-order diffracted beam are both parallel beams with a diameter of approximately 0.5 mm. However, at the rear focal point of the condenser lens 16B, they both form beam waists with a diameter of approximately 0.1 to 0.2 mm and are separated in the Z direction. Therefore, only beam LB6 can be reflected in the -Z direction by the epi-illumination mirror IM6. Furthermore, two lenses GL1b and GL2b, which transmit the measurement beam MBb reflected by the beam splitter 30B, form a 1x magnification relay imaging system. As shown in FIG. 4, a plane Psb optically conjugate with the exit of beam LBb from laser light source 10B is formed at the rear focal point of lens GL2b. Mirrors 31B and 32B and lenses GL1b and GL2b form a light guide system that guides the measurement beam MBb to a fluctuation detection optical unit composed of a triangular mirror 33 and a detection unit 34.

[0030] In this embodiment, the light guide system is made up of mirrors 31B and 32B and lenses GL1b and GL2b, but the fluctuation detection optical unit may be made up of detection unit 34, and the light guide system may also include beam splitter 30B, mirrors 31B and 32B, lenses GL1b and GL2b, and one of the reflecting surfaces of triangular mirror 33.

[0031] Returning to FIG. 4 , the explanation continues. The rear focal position of the condenser lens 16B is set to coincide with the front focal position of the collimator lens 17B at the subsequent stage. The optical axes of the condenser lens 16B and the collimator lens 17B are arranged coaxially. The beam LBb (or the zeroth-order beam) passing through the condenser lens 16B is converted again into a parallel beam with a diameter of approximately 0.5 mm by the collimator lens 17B, reflected by mirror 18B, and incident on the second-stage acousto-optic modulation element AM4, which is arranged under the Bragg diffraction condition. The beam LBb incident on the acousto-optic modulation element AM4 is reflected in the −X direction by mirror 19B, and then incident on the third-stage acousto-optic modulation element AM2, which is arranged under the Bragg diffraction condition, via the condenser lens 20B, which is arranged similarly to the condenser lens 16B, the collimator lens 21B, which is arranged similarly to the collimator lens 17B, and the mirror 22B. Here, too, the rear focal position of the condenser lens 20B and the front focal position of the collimator lens 21B are set to coincide with each other. Then, an epi-illumination mirror IM4 similar to the epi-illumination mirror IM6 is disposed at the rear focal position of the condenser lens 20B, and only the beam LB4, which is a first-order diffracted beam generated when the acousto-optic modulation element AM4 is in the on state, is reflected in the −Z direction by the epi-illumination mirror IM4.

[0032] Beam LBb that has passed through the third-stage acousto-optic modulation element AM2 is reflected in the +X direction by mirror 23B, passes through condenser lens 24B and collimator lens 25B, and enters beam splitter 26B. Beam splitter 26B is set to have high transmittance and low reflectance, and beam LBb (or the zeroth-order diffracted beam) that has passed through beam splitter 26B is absorbed by beam trap 27B. A portion of the beam reflected by beam splitter 26B is received by photodetector 28B, which measures the intensity and position of beam LBb (or the zeroth-order diffracted beam) that has passed through the three acousto-optic modulation elements AM6, AM4, and AM2. At the beam waist position of beam LBb between focusing lens 24B and collimator lens 25B (the position of the rear focal point of focusing lens 24B), a reflecting mirror IM6 and a reflecting mirror IM2 similar to reflecting mirror IM4 are arranged, and only beam LB2, which is a first-order diffracted beam generated when acousto-optical modulation element AM2 is in the on state, is reflected in the -Z direction by reflecting mirror IM2.

[0033] When viewed in the XY plane, each of the epi-mirrors IM6, IM4, and IM2 matches the arrangement in the XY plane of the mirror M10 (see Figure 2) of each of the even-numbered imaging units MU6, MU4, and MU2. Therefore, as shown in Figure 4, each of the epi-mirrors IM6, IM4, and IM2 is installed at a regular interval on a line Kb parallel to the Y axis in the XY plane, and is also arranged in the same position in the Z direction. Furthermore, laser light source 10A is the same as laser light source 10B, and further, the optical path arrangement (arrangement of each optical component) in the XY plane of beam LBa from laser light source 10A is rotated by 180 degrees in the XY plane from the optical path arrangement (arrangement of each optical component) of beam LBb from laser light source 10B.

[0034] Beam LBa (e.g., pulsed light with a wavelength of 400 nm or less) from laser light source 10A is emitted in the +X direction as a parallel beam with a diameter of approximately 1 mm. Beam LBa from laser light source 10A enters beam splitter 30A, which has high transmittance and low reflectance of several percent to 10%. The beam LBa that passes through beam splitter 30A is guided to odd-numbered imaging units MU1, MU3, and MU5. Beam MBa reflected by beam splitter 30A passes through mirror 31A, lens GL1a, mirror 32A, and lens GL2a and is used to measure beam fluctuations. After passing through beam splitter 30A, beam LBa passes through correction optical system 11A, which fine-tunes the tilt of beam LBa's traveling direction and slightly shifts it laterally in a plane perpendicular to the beam, before entering beam splitter 12A, which has a transmittance of several percent or less. The beam transmitted through the beam splitter 12A is received by a light intensity monitor 13A that measures the intensity of the beam LBa from the laser light source 10A.

[0035] Beam LBa reflected in the +Y direction by beam splitter 12A is converted by reduction optical system 14A into a parallel beam with a beam diameter reduced to 0.5 mm (approximately half the original diameter). Then, mirror system 15A (similar to mirror system 15B shown in FIG. 5, including mirrors 15A1 and 15A2) redirects the beam to travel in the -X direction, where it is incident as a parallel beam on first-stage acousto-optical modulator AM1, which is positioned under the Bragg diffraction condition. When acousto-optical modulator AM1 is in the off state (non-deflected state), beam LBa passes through acousto-optical modulator AM1 as is and passes through condenser lens 16A, collimator lens 17A, and mirror 18A to be incident as a parallel beam on second-stage acousto-optical modulator AM3, which is positioned under the Bragg diffraction condition. In the XY plane, at the rear focal position of condenser lens 16A, an epi-illumination mirror IM1 with its reflecting surface tilted 45 degrees relative to the XY plane is located. The incident-light mirror IM1 is positioned so that it reflects only the first-order diffracted beam generated when the acousto-optical modulator AM1 is in the on state (deflected state) in the -Z direction, and does not irradiate the undiffracted zeroth-order beam (part of beam LBb).

[0036] The rear focal position of the condenser lens 16A is set to coincide with the front focal position of the subsequent collimator lens 17A. The optical axes of the condenser lens 16A and the collimator lens 17A are arranged coaxially, and beam LBa (or the zeroth-order beam) passing through the condenser lens 16A is converted again into a parallel beam with a diameter of approximately 0.5 mm by the collimator lens 17A, reflected by mirror 18A, and incident on the second-stage acousto-optic modulation element AM3. After incident on the acousto-optic modulation element AM3, beam LBa is reflected in the +X direction by mirror 19A and then incident on the third-stage acousto-optic modulation element AM5, which is arranged under the conditions of Bragg diffraction, via condenser lens 20A, which is arranged similarly to the condenser lens 16A, collimator lens 21A, which is arranged similarly to the collimator lens 17A, and mirror 22A.

[0037] Here too, the rear focal position of the condenser lens 20A and the front focal position of the collimator lens 21A are set to coincide with each other. Then, an epi-mirror IM3 similar to the epi-mirror IM1 is disposed at the rear focal position of the condenser lens 20A, and only the beam LB3, which is a first-order diffracted beam generated when the acousto-optic modulation element AM3 is in the on state, is reflected in the −Z direction by the epi-mirror IM3.

[0038] Beam LBa that has passed through the third-stage acousto-optic modulation element AM5 is reflected in the -X direction by mirror 23A, passes through condenser lens 24A and collimator lens 25A, and enters beam splitter 26A. Beam splitter 26A is set to have high transmittance and low reflectance, and beam LBa (or the zeroth-order diffracted beam) that has passed through beam splitter 26A is absorbed by beam trap 27A. A portion of the beam reflected by beam splitter 26A is received by photodetector 28A, and the intensity and position of beam LBa (or the zeroth-order diffracted beam) that has passed through the three acousto-optic modulation elements AM1, AM3, and AM5 are measured.

[0039] A vertical mirror IM5 similar to the vertical mirrors IM1 and IM3 is disposed at the beam waist position of the beam LBa between the condenser lens 24A and the collimator lens 25A (the position of the rear focal point of the condenser lens 24A). When the acousto-optic modulator AM5 is in the on state, only the beam LB5, which is a first-order diffracted beam, is reflected in the -Z direction by the vertical mirror IM5. When viewed in the XY plane, the vertical mirrors IM1, IM3, and IM5 each match the arrangement in the XY plane of the mirror M10 (see Figure 2) of each of the odd-numbered imaging units MU1, MU3, and MU5. Therefore, as shown in Figure 4, the vertical mirrors IM1, IM3, and IM5 are disposed at regular intervals on a line Ka parallel to the Y axis in the XY plane, and are positioned at the same position in the Z direction.

[0040] The switching acousto-optic modulation elements AM1 to AM6 are all positioned to satisfy the conditions of Bragg diffraction. Two lenses GL1a and GL2a, which transmit the measurement beam MBa reflected by the beam splitter 30A, form a 1x1 relay imaging system, and as shown in FIG. 4, a plane Psa optically conjugate with the exit of the beam LBa from the laser light source 10A is formed at the rear focal point of the lens GL2a. The mirrors 31A and 32A and the lenses GL1a and GL2a form a light guide system that guides the measurement beam MBa to a fluctuation detection optical unit consisting of a triangular mirror 33 and a detection unit 34. In this embodiment, the optical path length and the optical path return position of the measurement beam MBa from the exit of the laser light source 10A to the triangular mirror 33 (and the detection unit 34) are set to be the same as the optical path length and the optical path return position of the measurement beam MBb from the exit of the laser light source 10B to the triangular mirror 33 (and the detection unit 34).

[0041] In the above configuration, the exit of laser light source 10A is optically conjugate with the crystal in first-stage acousto-optic modulation element AM1 via correction optical system 11A and reduction optical system 14A, and the exit of laser light source 10B is optically conjugate with the crystal in first-stage acousto-optic modulation element AM6 via correction optical system 11B and reduction relay optical system 14B. Furthermore, the position of the reflective surface of each of epi-illumination mirrors IM1 to IM6 in FIG. 4 is optically conjugate with a surface OPa (see FIG. 2) set inside each of imaging units MU1 to MU6. As a result, the position of the reflective surface of each of epi-illumination mirrors IM1 to IM6 (the position where each of beams LB1 to LB6 is focused as a beam waist) and the image plane of spot light SP focused on sheet substrate P are optically conjugate with each other.

[0042] The optical path arrangement in the XY plane from laser light source 10A to beam trap 27A and photodetector 28A and the optical path arrangement in the XY plane from laser light source 10B to beam trap 27B and photodetector 28B are rotated by 180 degrees relative to each other in the XY plane, and their imaginary center of rotation (center point of point symmetry) PG is set to the midpoint in the X direction between lines Ka and Kb in Figure 4, and the midpoint in the Y direction between epi-illumination mirror IM1 located furthest in the -Y direction and epi-illumination mirror IM6 located furthest in the +Y direction. Therefore, the imaginary line segment connecting epi-illumination mirror IM1 and epi-illumination mirror IM6, the imaginary line segment connecting epi-illumination mirror IM2 and epi-illumination mirror IM5, and the imaginary line segment connecting epi-illumination mirror IM3 and epi-illumination mirror IM4 are set to intersect at the center point PG of point symmetry in the XY plane. The midpoint in the X direction between the lines Ka and Kb coincides with the position of the central plane Cp shown in FIGS. 1 and 3, so the central point PG is located within the central plane Cp.

[0043] As shown in Fig. 5, the optical axis of beam LBb emitted from laser light source 10B and the optical axis of beam LBb passing through the first-stage acousto-optic modulation element AM6 are set to have a predetermined distance (height difference) in the Z direction by the two mirrors of mirror system 15B. Therefore, beam MBb reflected by beam splitter 30B located immediately after laser light source 10B travels in the -Y direction in the space above acousto-optic modulation elements AM1-AM6 in the +Z direction, is reflected by mirror 31B at 90 degrees in the +X direction, and is further reflected by mirror 32B at 90 degrees in the -Y direction. As shown in Fig. 4, the optical axis of beam LBa emitted from laser light source 10A and the optical axis of beam LBa passing through the first-stage acousto-optic modulation element AM1 are set to have a predetermined distance (height difference) in the Z direction by the two mirrors of mirror system 15A. Therefore, beam MBa reflected by beam splitter 30A, which is arranged immediately after laser light source 10A, travels in the +Y direction in the space above acousto-optic modulation elements AM1 to AM6 in the +Z direction, is reflected by mirror 31A at 90 degrees in the -X direction, and is further reflected by mirror 32A at 90 degrees in the +Y direction.

[0044] Furthermore, the center line of measurement beam MBb reflected by mirror 32B (the optical axis of lenses GL1b and GL2b) and the center line of measurement beam MBa reflected by mirror 32A (the optical axis of lenses GL1a and GL2a) are set to be parallel to the Y axis and coaxial with each other, and are set to intersect with the normal to the XY plane passing through center point PG. A triangular mirror 33 is disposed at the position of center point PG, reflecting both beam MBa and beam MBb in the +X direction. Beams MBa and MBb (both parallel luminous fluxes) reflected by triangular mirror 33 and traveling in the +X direction are incident on detection unit 34, which monitors beam fluctuations. The triangular mirror 33 and detection unit 34 form a fluctuation detection optical unit. The optical path length from the exit of laser light source 10A to beam splitter 30A is set to be the same as the optical path length from the exit of laser light source 10B to beam splitter 30B.

[0045] Figure 6 is a perspective view showing the specific positional relationship between the triangular mirror 33 and the detection unit 34 in Figure 4, and the Cartesian coordinate system XYZ in Figure 6 is set to be the same as the Cartesian coordinate system XYZ in Figure 4. In Figure 6, the triangular mirror 33 has a reflecting surface 33a that reflects the beam MBa traveling in the +Y direction at a right angle to the +X direction, and a reflecting surface 33b that reflects the beam MBb traveling in the -Y direction at a right angle to the +X direction, and the reflecting surfaces 33a and 33b are set to form a right angle (90 degrees) in the XY plane. A normal line parallel to the Z axis passing through the center point PG is set to be perpendicular to the extension of the center lines of the beams MBa and MBb before they reach the triangular mirror 33.

[0046] The detection unit 34 includes a telecentric reduction relay optical system (detection lens system, imaging system) composed of a pair of lenses 34A and 34B arranged along an optical axis AXu, a two-dimensional image sensor (CCD sensor or CMOS sensor) 34C, a beam splitter (half mirror) 34E, and a second image sensor (CCD sensor or CMOS sensor) 34G. The optical axis AXu is set parallel to the X axis, and its extension is set perpendicular to a normal line parallel to the Z axis passing through the center point PG. The pair of lenses 34A and 34B (detection lens system, imaging system) reduce the spacing and each beam diameter in the YZ plane of two beams MBa and MBb incident on the lens 34A parallel to the optical axis AXu by a predetermined reduction ratio, and project the beams onto the imaging surface of the first image sensor 34C. Here, the position of the front focal point of the lens 34A is set to coincide with the planes Psa and Psb shown in FIG. 4. Therefore, the imaging surface of the imaging element 34C is set so as to be in a conjugate relationship (imaging relationship) with each of the emission ports of the laser light source 10A and the laser light source 10B.

[0047] Between the pair of lenses 34A and 34B, the two beams MBa and MBb (parallel beams) incident on the lens 34A converge as beam waists, and a condensing plane Ph where the beams intersect is set. The condensing plane Ph is set at the rear focal position of the lens 34A and the front focal position of the lens 34B (the position of the pupil plane of the imaging system formed by the lenses 34A and 34B). The reflecting surface of a beam splitter (half mirror) 34E, which is disposed between the lens 34A and the condensing plane Ph, is set at an angle of 45° with respect to the XY plane, and reflects portions of the beams MBa and MBb that have passed through the lens 34A in the -Z direction. The portions of the beams MBa and MBb reflected by the beam splitter (half mirror) 34E are condensed as spot light at approximately the same position on the imaging surface of a second image sensor 34G, which is disposed at the rear focal position of the lens 34A (i.e., the position corresponding to the condensing plane Ph). Although not shown in Figure 6, if the illuminance of the measurement beams MBa and MBb is high compared to the imaging sensitivity of the imaging elements 34C and 34G, a neutral density filter (ND filter) may be placed in the optical path between the triangular mirror 33 and the beam splitter (half mirror) 34E.

[0048] In the configuration of the detection unit 34 shown in FIG. 6, for example, if the beam MBa projected onto the reflecting surface 33a of the triangular mirror 33 is shifted in parallel by ΔXa in the +X direction from a specified position (design position), the beam MBa incident on the lens 34A is shifted in parallel by ΔYa in the +Y direction, the same amount as ΔXa. In this case, the position of the beam waist of the beam MBa formed on the focusing surface Ph does not move from the position of the optical axis AXu within the focusing surface Ph. Therefore, the position of the spot light of the beam MBa focused on the imaging surface of the second imaging element 34G does not change either. Similarly, if the beam MBa incident on the lens 34A is shifted in parallel by ΔYa in the +Y direction, the center line of the beam MBa passing through the center of the focusing surface Ph (the position where the optical axis AXu passes) is tilted from the specified state (design state) within the XY plane. Therefore, if the reduction magnification of the reduction relay optical system formed by lenses 34A and 34B is β, the position of beam MBa imaged on the imaging surface of first imaging element 34C shifts by β·ΔYa (=β·ΔXa) in the −Y direction from the specified position (design position).

[0049] Furthermore, if the beam MBa projected onto the reflecting surface 33a of the triangular mirror 33 is tilted by Δθa from a specified state (design state) in the XY plane, for example, the beam MBa incident on the lens 34A will also be tilted by Δθa from a specified state (parallel to the optical axis AXu) in the XY plane. This tilt Δθa corresponds to the tilt of the beam LBa at the position of the exit of the laser light source 10A. In this case, the position of the beam waist of the beam MBa formed on the focusing plane Ph varies by ΔYθa in the Y direction from the position of the optical axis AXu within the focusing plane Ph, and the position of the spot light of the beam MBa imaged on the imaging plane of the second imaging element 34G shifts from the specified position (the position through which the optical axis AXu passes) in the Y direction by an amount corresponding to the magnitude of the tilt ΔYθa.

[0050] On the other hand, the reduction relay optical system formed by lenses 34A and 34B results in an imaging relationship between a plane Psa (see FIG. 4) conjugate with the exit port of beam LBa of laser light source 10A and the imaging surface of first image sensor 34C, so that if only the tilt of beam LBa changes at the position of the exit port of laser light source 10A, the position of beam MBa imaged on the imaging surface of first image sensor 34C does not change. As described above, first image sensor 34C can detect the parallel position change component of the fluctuations of measurement beams MBa and MBb (i.e., beams LBa and LBb), and second image sensor 34G can detect the tilt change component of the fluctuations of measurement beams MBa and MBb (i.e., beams LBa and LBb).

[0051] FIG. 7 is a diagram schematically illustrating the state of beams MBa and MBb projected onto the imaging surface of the first imaging element 34C, and FIG. 8 is a diagram schematically illustrating the state of spot light of beams MBa and MBb projected onto the imaging surface of the second imaging element 34G. In FIG. 7, the Y- and Z-axes set on the imaging surface correspond to the Y- and Z-axes of the Cartesian coordinate system XYZ set in each of FIGS. 4 to 6, and correspond to the Y- and Z-direction positional displacement directions of beams LBa and LBb on the plane of the emission ports of laser light sources 10A and 10B. Reference point CFa set on the imaging surface represents the position onto which measurement beam MBa is projected when beam LBa from laser light source 10A is emitted without parallel shift. Similarly, reference point CFb set on the imaging surface represents the position onto which measurement beam MBb is projected when beam LBb from laser light source 10B is emitted without parallel shift.

[0052] As shown in FIG. 7, when beam MBa is shifted in the −Y and +Z directions relative to reference point CFa, beam LBa emitted from the exit of laser light source 10A is shifted parallel in the −Y and +Z directions. Similarly, the shift of the projection position of beam MBb relative to reference point CFb represents the parallel shift of beam LBb emitted from the exit of laser light source 10B in the Y or Z direction. A control unit (not shown) analyzes and processes image information from image sensor 34C to determine the parallel shift error amounts of beams LBa and LBb (the shift amount of beam MBa from reference point CFa and the shift amount of beam MBb from reference point CFb). Based on the determined error amounts, the parallel shift error of beam LBa is corrected by correction optical system 11A shown in FIG. 4, and the parallel shift error of beam LBb is corrected by correction optical system 11B shown in FIG. 4.

[0053] 8, the θy axis set on the imaging surface of the second imaging element 34G represents the amount of tilt in the Y direction within the XY plane of each of the beams LBa and LBb at the exit ports of the laser light sources 10A and 10B, and the θz axis represents the amount of tilt in the Z direction within the XZ plane of each of the beams LBa and LBb at the exit ports of the laser light sources 10A and 10B. Furthermore, the reference point CFg on the imaging surface represents the position where each spot light of the measurement beams MBa and MBb is projected when the beams LBa and LBb are emitted from the laser light sources 10A and 10B without tilt. Note that, if the laser light sources 10A and 10B are fiber amplifier laser light sources, the θz direction also serves as the walk-off orientation of the wavelength conversion elements (harmonic generating crystals for second harmonic wave, third harmonic wave, etc.) provided inside the laser light sources.

[0054] In the example of FIG. 8, the spot light of beam MBa is approximately aligned with reference point CFg, and the spot light of beam MBb is projected displaced from reference point CFg in the -θz direction. This indicates that beam LBb from laser light source 10B is emitted from the exit at an angle in the -θz direction. A control unit (not shown) analyzes and processes image information from image sensor 34G to determine the tilt error amounts of beams LBa and LBb (the amount of deviation of the spot light of beam MBa from reference point CFg and the amount of deviation of the spot light of beam MBb from reference point CFg). Based on the determined tilt error amounts, the tilt error of beam LBa is corrected by correction optical system 11A shown in FIG. 4, and the tilt error of beam LBb is corrected by correction optical system 11B shown in FIG. 4.

[0055] Note that, since the spot lights of beams MBa and MBb projected onto the imaging surface of second imaging element 34G are both set to be positioned at reference point CFg in design, even if they are imaged as shifted spot lights, as shown in Figure 8, it is not possible to determine whether the spot lights are measurement beam MBa derived from beam LBa or measurement beam MBb derived from beam LBb. Therefore, when drawing a pattern on sheet substrate P, the image information from imaging element 34G can be sampled at the timing when either beam MBa or MBb is projected onto the imaging surface of imaging element 34G, taking advantage of the fact that there are periods when only laser light source 10A emits beam LBa and periods when only laser light source 10B emits beam LBb.

[0056] Alternatively, a shutter (movable light-shielding plate) may be provided in either or both of the optical path of measurement beam MBa from beam splitter 30A to triangular mirror 33 in Figure 4 and the optical path of measurement beam MBb from beam splitter 30B to triangular mirror 33, so that at least one of beams MBa and MBb is not projected onto the imaging surface of imaging element 34G.

[0057] FIG. 9 is a perspective view showing an example of a specific optical configuration of correction optical system 11B shown in FIGS. 4 and 5 (correction optical system 11A in FIG. 4 is also the same). Cartesian coordinate system XYZ in FIG. 9 is set to be the same as the Cartesian coordinate systems XYZ set in each of FIGS. 4 to 6. Beam LBb (parallel light beam) from beam splitter 30B (see FIG. 5) is incident on a beam shifter consisting of a quartz parallel plate HV1 that can be tilted about center line SF1 that is perpendicular to optical axis AXb and parallel to the Y axis, and a quartz parallel plate HV2 that can be tilted about center line SF2 that is perpendicular to optical axis AXb and parallel to the Z axis. Beam LBb is shifted parallel to the Z direction by the tilt of parallel plate HV1, and shifted parallel to the Y direction by the tilt of parallel plate HV2.

[0058] After passing through the parallel plate HV2, the beam LBb passes through a quartz prism plate RD1 that can rotate about the optical axis AXb, and then passes through a quartz prism plate RD2 that can also rotate about the optical axis AXb. Each of the prism plates RD1 and RD2 is formed in a wedge shape with a first surface that is perpendicular to the optical axis AXb and a second surface that is tilted relative to the first surface, facing each other. By adjusting the angle of each of the two prism plates RD1 and RD2 about the optical axis AXb, it is possible to fine-tune the tilt of the traveling direction of the beam LBb emerging from the prism plate RD2.

[0059] The tilt adjustment of each of parallel plates HV1 and HV2 and the adjustment of the rotation angle of each of prism plates RD1 and RD2 may be performed by an actuator controlled by a command from a control unit (not shown) based on the parallel shift error amount and tilt error amount measured by each of image sensors 34C and 34G shown in Fig. 6. Correction optical system 11A shown in Fig. 4 is configured similarly to correction optical system 11B shown in Fig. 9.

[0060] As described above, in this embodiment, a portion of each of the beams LBa and LBb emitted from the two spatially separated laser light sources 10A and 10B is split by the beam splitters 30A and 30B (one of 30A and 30B corresponds to the first beam splitter, and the other corresponds to the second beam splitter) to generate measurement beams MBa and MBb. The optical paths (mirror arrangements, etc.) on the beam MBa side and the beam MBb side to the detection unit 34, which measures beam fluctuations, are set to be identical in length. Furthermore, the optical path length to the triangular mirror 33, which combines the two measurement beams MBa and MBb so that they are parallel and close to each other, can be set long. Therefore, even slight fluctuations (parallel shift errors or tilt errors) in the beams LBa and LBb emitted from the laser light sources 10A and 10B can be detected as a relatively large positional deviation on the imaging surfaces of the image sensors 34C and 34G.

[0061] 1 to 5, if the relative positional relationship or relative tilt between the beam LBa emitted from the laser light source 10A and the beam LBb emitted from the laser light source 10B changes, an error may occur in the relative positional relationship between the odd-numbered drawing lines SL1, SL3, and SL5 and the even-numbered drawing lines SL2, SL4, and SL6, reducing the accuracy of joining the patterns drawn by each of the drawing lines SL1 to SL6. Therefore, using FIGS. 10 to 14, we will explain how the patterns drawn by each of the drawing lines SL1 to SL6 are misaligned due to changes in the emission states of the beams LBa and LBb from the laser light sources 10A and 10B (parallel shift error and tilt error).

[0062] 10 is a perspective view illustrating the state of shift of beam LBb from laser light source 10B when the beam LBb is shifted in parallel along the optical path from laser light source 10B to the first-stage acousto-optic modulation element AM6 shown in FIG. 5. In FIG. 10, when beam LBb emitted from the exit of laser light source 10B shifts (translates) in the −Y direction from a predetermined optical axis AXb (see FIG. 9) as indicated by arrow Ay1, beam LBb immediately after passing through beam splitter 30B and correction optical system 11B also shifts in the −Y direction as indicated by arrow Ay2. Therefore, beam LBb reflected by beam splitter 12B and traveling in the −Y direction shifts in the −X direction immediately before entering reduction relay optical system 14B. Because reduction relay optical system 14B is also an imaging system that forms an inverted image, beam LBb emitted from reduction relay optical system 14B shifts in the +X direction as indicated by arrow Ay3. Furthermore, the beam LBb bent in the +X direction by the mirror 15B2 is incident on the acousto-optic modulation element AM6 in a state where it is shifted in parallel in the -Z direction from the predetermined optical axis AXb as indicated by the arrow Ay4.

[0063] Furthermore, when beam LBb emitted from the exit of laser light source 10B shifts (translates) in the +Z direction from a predetermined optical axis AXb (see FIG. 9 ) as indicated by arrow Az1, beam LBb passes through beam splitter 30B, correction optical system 11B, beam splitter 12B, and reduction relay optical system 14B and is reflected in the −Z direction by mirror 15B1, shifting in the −Y direction as indicated by arrow Az3. Therefore, beam LBb bent in the +X direction by mirror 15B2 is incident on acousto-optic modulation element AM6 while being parallel-shifted in the −Y direction from the predetermined optical axis AXb as indicated by arrow Az4. When beam LBb emitted from the exit of laser light source 10B is parallel-shifted as indicated by arrow Ay1, measurement beam MBb reflected by beam splitter 30B is shifted in the −X direction as indicated by arrow Ay5. When beam LBb emitted from laser light source 10B is parallel-shifted as indicated by arrow Az1, measurement beam MBb is shifted in the +Z direction as indicated by arrow Az5.

[0064] Fig. 11 is an exaggerated view showing the states of the drawing beams LB2, LB4, and LB6 directed toward the even-numbered drawing units MU2, MU4, and MU6, respectively, when the beam LBb from the laser light source 10B is shifted in parallel in the -Y direction as indicated by the arrow Ay1 in Fig. 10. For ease of understanding, it is assumed in Fig. 11 that the position and inclination of the beam LBa are not adjusted by the correction optical system 11B shown in Fig. 9. 4, the beam LBb from the laser light source 10B is passed in series through the even-numbered acousto-optical modulation elements AM6, AM4, and AM2, and therefore in Fig. 11, the optical path from the first-stage acousto-optical modulation element AM6 to the condenser lens 16B, epi-mirror IM6, and collimator lens 17B is shown in the upper row, the optical path from the second-stage acousto-optical modulation element AM4 to the condenser lens 20B, epi-mirror IM4, and collimator lens 21B is shown in the middle row, and the optical path from the third-stage acousto-optical modulation element AM2 to the condenser lens 24B, epi-mirror IM2, and collimator lens 25B is shown in the lower row. Note that the Cartesian coordinate system XYZ in Fig. 10 is set to be the same as the Cartesian coordinate systems XYZ in Figs. 4 and 5.

[0065] 11, when beam LBb incident on the first-stage acousto-optical modulation element AM6 is shifted in parallel in the -Z direction as indicated by arrow Ay4 in Fig. 10 with respect to a predetermined optical axis AXb, the zeroth-order diffracted beam that travels straight without being deflected by acousto-optical modulation element AM6 in the on state intersects with optical axis AXb at the rear focal position of condenser lens 16B, i.e., the position of epi-mirror IM6, and then is collimated again by collimator lens 17B into a parallel beam parallel to the optical axis AXb, and is incident on second-stage acousto-optical modulation element AM4 in a state shifted in parallel in the +Z direction. The first-order diffracted beam deflected at a predetermined diffraction angle by acousto-optical modulation element AM6 in the on state converges to form a beam waist at the position of epi-mirror IM6 as drawing beam LB6.

[0066] Because the acousto-optic modulator AM6 is located at the front focal position of the condenser lens 16B and the epi-mirror IM6 is located at the rear focal position of the condenser lens 16B, the beam LB6 traveling from the condenser lens 16B to the epi-mirror IM6 is tilted in the XZ plane rather than parallel to the optical axis AXb. However, the position of the beam waist of the beam LB6 converging at the position of the epi-mirror IM6 does not change in the YZ plane even if the beam LBb incident on the acousto-optic modulator AM6 is shifted in parallel as indicated by arrow Ay4. However, the beam LB6 reflected by the epi-mirror IM6 is tilted in the -X direction in the XZ plane with respect to the optical axis AX6, which is an extension of the optical axis of the beam expander formed by lenses LGa and LGb in the imaging unit MU6 (see Figure 2).

[0067] When the first-stage acousto-optical modulator AM6 is in the off state and the beam LBb incident on the acousto-optical modulator AM6 is shifted in the -Z direction as indicated by arrow Ay4, the beam LBb passes through the acousto-optical modulator AM6, the condenser lens 16B, and the collimator lens 17B and is incident on the second-stage acousto-optical modulator AM4 while being shifted in the +Z direction in the XZ plane. Again, the acousto-optical modulator AM4 is located at the front focal position of the condenser lens 20B, and the epi-mirror IM4 is located at the rear focal position of the condenser lens 20B. When the acousto-optical modulator AM4 is in the on state, the zeroth-order diffracted beam traveling straight without being deflected by the acousto-optical modulator AM4 intersects with the optical axis AXb at the rear focal position of the condenser lens 20B, i.e., the position of the epi-mirror IM4, and is then collimated again by the collimator lens 21B so as to be parallel to the optical axis AXb. The beam LBb then enters the third-stage acousto-optical modulator AM2 while being shifted in the -Z direction.

[0068] Furthermore, the first-order diffracted beam deflected at a predetermined diffraction angle by the acousto-optic modulator AM4 in the on state converges as the drawing beam LB4 to form a beam waist at the position of the epi-mirror IM4. The beam LB4 traveling from the condenser lens 20B toward the epi-mirror IM4 is tilted rather than parallel to the optical axis AXb in the XZ plane. However, the position of the beam waist of the beam LB4 converging at the position of the epi-mirror IM4 does not change in the YZ plane even if the beam LBb incident on the acousto-optic modulator AM4 is shifted in the +Z direction. However, the beam LB4 reflected by the epi-mirror IM4 is tilted in the -X direction in the XZ plane with respect to the optical axis AX4, which is an extension of the optical axis of the beam expander formed by the lenses LGa and LGb in the drawing unit MU4 (see Figure 2).

[0069] When the first-stage acousto-optic modulation element AM6 and the second-stage acousto-optic modulation element AM4 are both in the Off state and the beam LBb incident on the acousto-optic modulation element AM6 is shifted in parallel in the -Z direction as indicated by arrow Ay4, the beam LBb passes through the acousto-optic modulation element AM6, the condenser lens 16B, the collimator lens 17B, the acousto-optic modulation element AM4, the condenser lens 20B, and the collimator lens 21B, and then incident on the third-stage acousto-optic modulation element AM2 while being shifted in parallel in the -Z direction in the XZ plane. In this case as well, the acousto-optic modulation element AM2 is placed at the front focal position of the condenser lens 24B, and the epi-illumination mirror IM2 is placed at the rear focal position of the condenser lens 24B. When the acousto-optical modulation element AM2 is in the on state, the zeroth-order diffracted beam that travels straight without being deflected by the acousto-optical modulation element AM2 intersects with the optical axis AXb at the rear focal position of the focusing lens 24B, i.e., the position of the epi-illumination mirror IM4, and is then collimated again by the collimator lens 25B into a parallel beam of light that is parallel to the optical axis AXb.

[0070] Furthermore, the first-order diffracted beam deflected at a predetermined diffraction angle by the acousto-optic modulator AM2 in the on state converges as the drawing beam LB2 to form a beam waist at the position of the epi-mirror IM2. The beam LB2 traveling from the condenser lens 24B toward the epi-mirror IM2 is tilted rather than parallel to the optical axis AXb in the XZ plane. However, the position of the beam waist of the beam LB2 converging at the position of the epi-mirror IM2 does not change in the YZ plane even if the beam LBb incident on the acousto-optic modulator AM2 is shifted in the -Z direction. However, the beam LB2 reflected by the epi-mirror IM2 is tilted in the -X direction in the XZ plane with respect to the optical axis AX2, which is an extension of the optical axis of the beam expander formed by the lenses LGa and LGb in the drawing unit MU2 (see Figure 2).

[0071] As described above, when the beam LBb emitted from the laser light source 10B is shifted so as to shift parallel to the Y direction as indicated by the arrow Ay1 in Fig. 10, the positions of the beam waists of the beams LB6, LB4, and LB2 formed at the respective positions of the epi-mirrors IM6, IM4, and IM2 do not change. These beam waists (focusing points) are conjugate (imaging) with the spot light SP on the sheet substrate P of the beams LB6, LB4, and LB2 that are ultimately projected onto the sheet substrate P from each of the imaging units MU6, MU4, and MU2. Therefore, even if the beam LBb emitted from the laser light source 10B is shifted parallel to the Y direction as indicated by the arrow Ay1 in Fig. 10, the positions of the even-numbered imaging lines SL6, SL4, and SL2 do not change.

[0072] The same applies to the odd-numbered imaging units MU1, MU3, and MU5 to which the beam LBa is supplied from the laser light source 10A, and the positions of the odd-numbered imaging lines SL1, SL3, and SL5 do not change even if the beam LBa emitted from the laser light source 10A is shifted in parallel in the Y direction in Fig. 4. However, as shown in Fig. 11, the beams LB1 to LB6 directed from the respective epi-illumination mirrors IM1 to IM6 to the respective imaging units MU1 to MU6 are inclined in the XZ plane.

[0073] These tilts are called telecentricity errors, and are the tilts of the center lines of the beams LB1 to LB6 projected onto the surface of the sheet substrate P relative to the normal to the surface of the sheet substrate P. The effects of the telecentricity errors (positional deviation of the spot light SP in the sub-scanning direction) occur during defocusing, so it is advisable to always set the focal plane where each of the beams LB1 to LB6 is focused as the spot light SP and the surface of the sheet substrate P within the depth of focus. When the effects of the telecentricity errors cannot be ignored, the parallel plate HV2 shown in FIG. 9 can be used to adjust the beams LBa and LBb from the laser light sources 10A and 10B so that they are translated in the Y direction.

[0074] Next, a case where the beam LBb emitted from the exit port of the laser light source 10B is shifted in parallel in the +Z direction as indicated by the arrow Az1 in FIG. 10 will be described with reference to FIG. 12. FIG. 12 is an exaggerated view showing the states of the drawing beams LB2, LB4, and LB6 directed toward the even-numbered drawing units MU2, MU4, and MU6, respectively, when the beam LBb from the laser light source 10B is shifted in parallel in the +Z direction as indicated by the arrow Az1 in FIG. 10. The Cartesian coordinate system XYZ in FIG. 12 is the same as the Cartesian coordinate system XYZ in FIG. 4. As described with reference to FIG. 10, when the beam LBb from the laser light source 10B is shifted in parallel in the +Z direction as indicated by the arrow Az1, the beam LBb incident on the first-stage acousto-optic modulation element AM6 is shifted in parallel in the −Y direction with respect to the predetermined optical axis AXb as indicated by the arrow Az4.

[0075] 12, the diffraction direction of each of the acousto-optical modulation elements AM6, AM4, and AM2 in the On state is the -Z direction in a plane parallel to the XZ plane, so that beam LB6 as a first-order diffracted beam emitted from the acousto-optical modulation element AM6 in the On state travels parallel to the optical axis AXb in the XY plane and enters the condenser lens 16B. After passing through the condenser lens 16B, beam LB6 forms a beam waist at the center in the Y direction of the reflecting surface of the epi-illumination mirror IM6 (a position offset from the optical axis AXb in the -Z direction) and is reflected in the -Z direction. Meanwhile, the zeroth-order diffracted beam from the acousto-optical modulation element AM6 intersects with the optical axis AXb in the space above the epi-illumination mirror IM6, passes through the collimator lens 17B, and becomes a collimated beam parallel to the optical axis AXb, which enters the second-stage acousto-optical modulation element AM4. Therefore, even if the beam LBb incident on the acousto-optical modulator AM6 is shifted in parallel in the Y direction as indicated by the arrow Az4, the position of the beam waist of the beam LB6 focused at the position of the epi-illumination mirror IM6 does not change in the XY plane.

[0076] When beam LBb incident on acousto-optical modulator AM6 is shifted in the Y direction as indicated by arrow Az4 and the first-stage acousto-optical modulator AM6 is in the off state, beam LBb shifted in the -Y direction with respect to optical axis AXb is incident on acousto-optical modulator AM4. When acousto-optical modulator AM4 is in the on state, beam LB4 as a first-order diffracted beam emerges from acousto-optical modulator AM4, travels parallel to optical axis AXb in the XY plane, and is incident on condenser lens 20B. After passing through condenser lens 20B, beam LB4 forms a beam waist at the center in the Y direction of the reflecting surface of epi-illumination mirror IM4 (a position offset in the -Z direction from optical axis AXb) and is reflected in the -Z direction.

[0077] On the other hand, the zeroth-order diffracted beam from acousto-optical modulation element AM4 intersects with the optical axis AXb in the space above the epi-mirror IM4, passes through collimator lens 21B, becomes a parallel beam parallel to the optical axis AXb, and enters the third-stage acousto-optical modulation element AM2. Therefore, even if the beam LBb entering the first-stage acousto-optical modulation element AM6 is shifted in parallel in the Y direction as indicated by arrow Az4, the position of the beam waist of beam LB4 focused at the position of the epi-mirror IM4 does not change in the XY plane.

[0078] Similarly, when beam LBb incident on acousto-optical modulator AM6 is shifted in the Y direction as indicated by arrow Az4, and both the first-stage acousto-optical modulator AM6 and the second-stage acousto-optical modulator AM4 are off, beam LBb shifted in the -Y direction with respect to optical axis AXb is incident on acousto-optical modulator AM2. When acousto-optical modulator AM2 is turned on, beam LB2 emerges from acousto-optical modulator AM2 as a first-order diffracted beam, traveling parallel to optical axis AXb in the XY plane and incident on condenser lens 24B. After passing through condenser lens 24B, beam LB2 forms a beam waist at the center of the Y direction of the reflecting surface of epi-illumination mirror IM2 (a position offset from optical axis AXb in the -Z direction) and is reflected in the -Z direction. Furthermore, the zeroth-order diffracted beam from acousto-optical modulator AM2 intersects with optical axis AXb in the space above epi-illumination mirror IM2, passes through collimator lens 25B, and travels as a collimated beam parallel to optical axis AXb. Therefore, even if the beam LBb incident on the first-stage acousto-optic modulation element AM6 is shifted in parallel in the Y direction as indicated by the arrow Az4, the position of the beam waist of the beam LB4 focused at the position of the epi-illumination mirror IM4 does not change in the XY plane.

[0079] As described above, even when the beam LBb emitted from the laser light source 10B is shifted so as to be parallel shifted in the Z direction as indicated by the arrow Az1 in Fig. 10, the positions of the beam waists of the beams LB6, LB4, LB2 formed at the respective positions of the epi-mirrors IM6, IM4, IM2 do not change. Therefore, even if the beam LBb emitted from the laser light source 10B is shifted parallel to the Z direction as indicated by the arrow Az1 in Fig. 10, the positions of the even-numbered drawing lines SL6, SL4, SL2 do not change.

[0080] The same applies to the odd-numbered imaging units MU1, MU3, and MU5 to which the beam LBa from the laser light source 10A is supplied. Even if the beam LBa emitted from the laser light source 10A is shifted in parallel in the Z direction in Fig. 4, the positions of the odd-numbered imaging lines SL1, SL3, and SL5 do not change. However, as shown in Fig. 12, the beams LB1 to LB6 directed from the respective epi-mirrors IM1 to IM6 to the respective imaging units MU1 to MU6 have a telecentricity error inclined with respect to a plane parallel to the XZ plane. Therefore, when the effect of the telecentricity error cannot be ignored, the beams LBa and LBb from the respective laser light sources 10A and 10B can be translated in the Z direction using the parallel plate HV1 shown in Fig. 9 to adjust the state of incidence of the beams LBa and LBb on the initial-stage acousto-optic modulation elements AM6 and AM1 (coaxiality with the optical axis AXb).

[0081] Next, a case where the beam LBb is emitted at an inclination from the exit of the laser light source 10B will be described. The exit of the laser light source 10B is set in a conjugate relationship with the crystal in the first-stage acousto-optic modulation element AM6 by the correction optical system 11B and the reduction relay optical system 14B shown in each of Figures 4, 5, and 10. If the reduction ratio of the reduction relay optical system 14B is 1 / Mb (Mb > 1), the tilt angle of the beam LBb incident on the acousto-optic modulation element AM6 with respect to the optical axis AXb is larger by a ratio corresponding to the reciprocal of the reduction ratio 1 / Mb than the tilt angle of the beam LBb with respect to the optical axis AXb at the exit of the laser light source 10B.

[0082] 13 is an exaggerated view showing the states of the imaging beams LB2, LB4, and LB6 directed toward the even-numbered imaging units MU2, MU4, and MU6, respectively, when the beam LBb incident on the first-stage acousto-optic modulation element AM6 is tilted with respect to the optical axis AXb in a plane parallel to the XZ plane. When the beam LBb from the exit port of the laser light source 10B is tilted in the Y direction in a plane parallel to the XY plane without the tilt correction of the beam LBb by the two prism plates RD1 and RD2 in the correction optical system 11B shown in FIG. 9, the beam LBb incident on the acousto-optic modulation element AM6 is tilted with respect to the optical axis AXb in a plane parallel to the XZ plane. The Cartesian coordinate system XYZ in FIG. 13 is set to the same as the Cartesian coordinate systems XYZ in FIGS. 4 and 12.

[0083] 13, if beam LBb incident on the first-stage acousto-optic modulation element AM6 in the On state is tilted slightly counterclockwise with respect to optical axis AXb in a plane parallel to the XZ plane, the zeroth-order diffracted beam not diffracted by acousto-optic modulation element AM6 is incident on condenser lens 16B at a tilt with respect to optical axis AXb. The zeroth-order diffracted beam travels along the optical path from condenser lens 16B to collimator lens 17B, parallel to optical axis AXb but slightly decentered in the +Z direction from the optical axis AXb, and forms a beam waist in the space above intermediate epi-illumination mirror IM6. The zeroth-order diffracted beam emerging from collimator lens 17B travels at a slight tilt clockwise with respect to optical axis AXb in a plane parallel to the XZ plane.

[0084] On the other hand, beam LB6, a first-order diffracted beam diffracted by acousto-optical modulator AM6, is deflected at a predetermined diffraction angle relative to the zeroth-order diffracted beam and enters condenser lens 16B. However, beam LB6 emerging from condenser lens 16B travels parallel to optical axis AXb along an optical path spaced apart from optical axis AXb in the -Z direction and is reflected by epi-illumination mirror IM6 in the -Z direction so as to be parallel to optical axis AX6. However, beam LB6 reflected by epi-illumination mirror IM6 is decentered in the -X direction with respect to optical axis AX6. Therefore, the focal point where beam LB6 forms the beam waist is shifted in the -X direction from the original position on optical axis AX6 in the XY plane, and the spot light SP of beam LB6 projected from imaging unit MU6 onto sheet substrate P is also displaced in the sub-scanning direction (Xt direction in FIG. 2) corresponding to the -X direction.

[0085] Furthermore, when beam LBb incident on acousto-optical modulator AM6 is tilted slightly counterclockwise with respect to optical axis AXb in a plane parallel to the XZ plane and the first acousto-optical modulator AM6 is in the Off state, beam LBb is incident on acousto-optical modulator AM4 in a plane parallel to the XZ plane and tilted slightly clockwise with respect to optical axis AXb. When acousto-optical modulator AM4 is in the On state, the zeroth-order diffracted beam not diffracted by acousto-optical modulator AM4 is incident on condenser lens 20B in a tilted state with respect to optical axis AXb. This zeroth-order diffracted beam travels along the optical path from condenser lens 20B to collimator lens 21B, parallel to optical axis AXb but slightly decentered in the -Z direction, and forms a beam waist in the space above intermediate epi-illumination mirror IM4. The zeroth-order diffracted beam emerging from collimator lens 21B travels in a plane parallel to the XZ plane and tilted slightly counterclockwise with respect to optical axis AXb.

[0086] On the other hand, beam LB4, a first-order diffracted beam diffracted by acousto-optical modulator AM4, is deflected at a predetermined diffraction angle relative to the zeroth-order diffracted beam and enters condenser lens 20B. However, beam LB4 emerging from condenser lens 20B travels parallel to optical axis AXb along an optical path spaced apart from optical axis AXb in the -Z direction, and is reflected by epi-illumination mirror IM4 in the -Z direction so as to be parallel to optical axis AX4. However, beam LB4 reflected by epi-illumination mirror IM4 is decentered in the -X direction with respect to optical axis AX4. Therefore, the focal point where beam LB4 forms the beam waist is shifted in the -X direction from the original position on optical axis AX4 in the XY plane, and the spot light SP of beam LB4 projected from imaging unit MU4 onto sheet substrate P is also displaced in the sub-scanning direction (Xt direction in FIG. 2 ) corresponding to the -X direction.

[0087] Furthermore, when beam LBb incident on acousto-optical modulator AM6 is tilted slightly counterclockwise with respect to optical axis AXb in a plane parallel to the XZ plane and acousto-optical modulators AM6 and AM4 are both in the Off state, beam LBb is incident on acousto-optical modulator AM2 in a plane parallel to the XZ plane and tilted slightly counterclockwise with respect to optical axis AXb. When acousto-optical modulator AM2 is in the On state, the zeroth-order diffracted beam not diffracted by acousto-optical modulator AM2 is incident on condenser lens 24B in a tilted state with respect to optical axis AXb. This zeroth-order diffracted beam travels along the optical path from condenser lens 24B to collimator lens 25B, parallel to optical axis AXb but slightly decentered in the +Z direction from the optical axis AXb, and forms a beam waist in the space above intermediate epi-illumination mirror IM2. The zeroth-order diffracted beam emerging from collimator lens 25B travels in a plane parallel to the XZ plane and tilted slightly clockwise with respect to optical axis AXb.

[0088] On the other hand, beam LB2, a first-order diffracted beam diffracted by acousto-optical modulator AM2, is deflected at a predetermined diffraction angle relative to the zeroth-order diffracted beam and enters condenser lens 24B. Beam LB2 emerging from condenser lens 24B travels parallel to optical axis AXb along an optical path spaced apart from optical axis AXb in the -Z direction and is reflected by epi-illumination mirror IM2 in the -Z direction so as to be parallel to optical axis AX2. However, beam LB2 reflected by epi-illumination mirror IM2 is decentered in the -X direction with respect to optical axis AX4. Therefore, the focal point where beam LB2 forms the beam waist is shifted in the -X direction within the XY plane from the original position on optical axis AX2, and the spot light SP of beam LB2 projected from imaging unit MU2 onto sheet substrate P is also displaced in the sub-scanning direction (Xt direction in FIG. 2) corresponding to the -X direction.

[0089] As described above, when the beam LBb emitted from the laser light source 10B is tilted with respect to the predetermined optical axis AXb in a plane parallel to the XY plane in FIG. 5 or 9, the positions of the beam waists of the beams LB6, LB4, and LB2 formed at the respective positions of the epi-mirrors IM6, IM4, and IM2 are displaced in the +X or −X direction. This causes the positions of the even-numbered imaging lines SL6, SL4, and SL2 to fluctuate in the sub-scanning direction (the Xt direction in FIG. 2). The same applies to the odd-numbered imaging units MU1, MU3, and MU5 to which the beam LBa from the laser light source 10A is supplied. When the beam LBa emitted from the laser light source 10A is tilted with respect to the predetermined optical axis (the design optical axis) in a plane parallel to the XY plane in FIG. 4, the positions of the beam waists of the beams LB1, LB3, and LB5 formed at the respective positions of the epi-mirrors IM1, IM3, and IM5 are displaced in the +X or −X direction. As a result, the positions of the odd-numbered rendering lines SL1, SL3, and SL5 will fluctuate in the sub-scanning direction (the Xt direction in FIG. 2).

[0090] 14 is an exaggerated view showing the states of the imaging beams LB2, LB4, and LB6 directed toward the even-numbered imaging units MU2, MU4, and MU6, respectively, when the beam LBb from the exit port of the laser light source 10B is tilted with respect to the optical axis AXb in a plane parallel to the XZ plane, resulting in the beam LBb incident on the first-stage acousto-optic modulation element AM6 being tilted with respect to the optical axis AXb in a plane parallel to the XY plane. Without the tilt correction of the beam LBb by the two prism plates RD1 and RD2 in the correction optical system 11B shown in FIG. 9, when the beam LBb from the exit port of the laser light source 10B is tilted in the Z direction in a plane parallel to the XZ plane, the beam LBb incident on the acousto-optic modulation element AM6 is tilted with respect to the optical axis AXb in a plane parallel to the XY plane. The Cartesian coordinate system XYZ in FIG. 14 is set to the same as the Cartesian coordinate systems XYZ in FIGS. 4 and 12.

[0091] As shown in Figure 14, when beam LBb incident on the first-stage acousto-optic modulation element AM6 in the On state is tilted slightly clockwise with respect to optical axis AXb in a plane parallel to the XY plane, the zeroth-order diffracted beam not diffracted by acousto-optic modulation element AM6 travels straight in the same direction as the incident beam LBb in the XY plane and enters condenser lens 16B. The zeroth-order diffracted beam travels parallel to the optical axis AXb along the optical path from condenser lens 16B to collimator lens 17B, slightly decentered from the optical axis AXb in the -Y direction, and forms a beam waist in the space above intermediate epi-illumination mirror IM6. The zeroth-order diffracted beam emerging from collimator lens 17B travels at a slight tilt counterclockwise with respect to optical axis AXb in a plane parallel to the XY plane.

[0092] On the other hand, beam LB6, a first-order diffracted beam diffracted by acousto-optic modulator AM6, travels the same optical path as the zeroth-order diffracted beam in the XY plane, but is deflected in the -Z direction at a predetermined diffraction angle and enters condenser lens 16B. Beam LB6 emerging from condenser lens 16B travels parallel to optical axis AXb along an optical path spaced apart from optical axis AXb in the -Y direction, and is reflected by epi-illumination mirror IM6 in the -Z direction so as to be parallel to optical axis AX6. However, beam LB6 reflected by epi-illumination mirror IM6 is decentered in the -Y direction relative to optical axes AXb and AX6. Therefore, the focal point of beam LB6, which forms the beam waist, is shifted in the -Y direction from the original position of optical axis AX6 in the XY plane, and the spot light SP of beam LB6 projected from imaging unit MU6 onto sheet substrate P is also displaced in the main scanning direction (Yt direction in FIG. 2) corresponding to the -Y direction. That is, the entire image line SL6 formed by the scanning of the spot light of the beam LB6 is shifted in the main scanning direction from its designed position.

[0093] Furthermore, when beam LBb incident on acousto-optical modulator AM6 is tilted slightly clockwise with respect to optical axis AXb in a plane parallel to the XY plane as shown in Figure 14 and acousto-optical modulator AM6 is in the Off state, beam LBb is incident on acousto-optical modulator AM4 in a plane parallel to the XY plane at a slight counterclockwise tilt with respect to optical axis AXb. When acousto-optical modulator AM4 is in the On state, the zeroth-order diffracted beam not diffracted by acousto-optical modulator AM4 travels straight in the same direction as incident beam LBb in the XY plane and is incident on condenser lens 20B in a tilted state. The zeroth-order diffracted beam travels along the optical path from condenser lens 20B to collimator lens 21B, parallel to optical axis AXb and slightly decentered in the -Y direction from the optical axis AXb, and forms a beam waist in the space above intermediate epi-illumination mirror IM4. The zeroth-order diffracted beam emerging from the collimator lens 21B travels in a plane parallel to the XY plane, tilted slightly counterclockwise with respect to the optical axis AXb.

[0094] On the other hand, beam LB4, a first-order diffracted beam diffracted by acousto-optic modulator AM4, travels the same optical path as the zeroth-order diffracted beam in the XY plane, but is deflected in the −Z direction relative to the zeroth-order diffracted beam at a predetermined diffraction angle and enters condenser lens 20B. Beam LB4 emerging from condenser lens 20B travels parallel to optical axis AXb along an optical path spaced apart from optical axis AXb in the −Y direction and is reflected in the −Z direction by epi-illumination mirror IM4 so as to be parallel to optical axis AX4. However, beam LB4 reflected by epi-illumination mirror IM2 is decentered in the −Y direction relative to optical axes AXb and AX4. Therefore, the focal point where beam LB4 forms the beam waist is shifted in the −Y direction from the original position on optical axis AX4 in the XY plane, and the spot light SP of beam LB4 projected from imaging unit MU4 onto sheet substrate P is also displaced in the sub-scanning direction (Xt direction in FIG. 2 ) corresponding to the −Y direction. That is, the entire image forming line SL4 formed by the scanning of the spot light of beam LB4 is shifted in the main scanning direction from the designed position. The shift direction of image forming line SL4 on the sheet substrate P is the same as the shift direction of image forming line SL6.

[0095] Furthermore, when beam LBb incident on acousto-optical modulator AM6 is tilted slightly clockwise with respect to optical axis AXb in a plane parallel to the XY plane, as shown in Figure 14, and acousto-optical modulators AM6 and AM4 are both in the Off state, beam LBb is incident on acousto-optical modulator AM2 in a plane parallel to the XY plane, tilted slightly clockwise with respect to optical axis AXb. When acousto-optical modulator AM2 is in the On state, the zeroth-order diffracted beam not diffracted by acousto-optical modulator AM2 is incident on condenser lens 24B in a tilted state with respect to optical axis AXb in the XY plane. This zeroth-order diffracted beam travels parallel to optical axis AXb along the optical path from condenser lens 24B to collimator lens 25B, slightly decentered from optical axis AXb in the -Y direction, and forms a beam waist in the space above intermediate epi-illumination mirror IM2. The zeroth-order diffracted beam emerging from collimator lens 25B travels in a plane parallel to the XY plane, tilted slightly counterclockwise with respect to optical axis AXb.

[0096] On the other hand, beam LB2, a first-order diffracted beam diffracted by acousto-optic modulator AM2, travels the same optical path as the zeroth-order diffracted beam in the XY plane, but is deflected in the −Z direction relative to the zeroth-order diffracted beam at a predetermined diffraction angle and enters condenser lens 24B. Beam LB2 emerging from condenser lens 24B travels parallel to optical axis AXb along an optical path spaced apart from optical axis AXb in the −Y direction and is reflected by epi-illumination mirror IM2 in the −Z direction so as to be parallel to optical axis AX2. However, beam LB2 reflected by epi-illumination mirror IM2 is decentered in the −Y direction relative to optical axes AXb and AX2. Therefore, the focal point of beam LB2, where the beam waist is located, is shifted in the −Y direction from the original position on optical axis AX2 in the XY plane, and the spot light SP of beam LB2 projected from imaging unit MU2 onto sheet substrate P is also displaced in the sub-scanning direction (Xt direction in FIG. 2 ) corresponding to the −Y direction. That is, the entire image forming line SL2 formed by the scanning of the spot light of the beam LB2 is shifted in the main scanning direction from its designed position. The shift direction of the image forming line SL2 is the same as the shift directions of the image forming lines SL6 and SL4.

[0097] From the above, when the beam LBb from the exit port of the laser light source 10B is tilted in the XZ plane with respect to the optical axis AXb, the even-numbered drawing lines SL2, SL4, and SL6 formed on the sheet substrate P are simultaneously shifted by the same amount in the main scanning direction (the Yt direction in FIG. 2). This situation also occurs with the odd-numbered beams LB1, LB3, and LB5 generated by the beam LBa from the laser light source 10A, and the odd-numbered drawing lines SL1, SL3, and SL5.

[0098] However, the set of odd-numbered imaging units MU1, MU3, and MU5 and the set of even-numbered imaging units MU2, MU4, and MU6 are arranged rotated 180° around the normal line passing through the center point PG shown in Fig. 4. Therefore, when the beam LBa emitted from the laser light source 10A and the beam LBb emitted from the laser light source 10B are both tilted in the +Z direction or the -Z direction within the XZ plane with respect to their respective predetermined optical axes, the odd-numbered imaging lines SL1, SL3, and SL5 and the even-numbered imaging lines SL2, SL4, and SL6 are shifted in opposite directions to each other in the Y (Yt) direction on the sheet substrate P. Conversely, if the beam LBa emitted from the laser light source 10A is inclined in the +Z direction by an angle ΔθLa in the XZ plane with respect to the predetermined optical axis, and the beam LBb emitted from the laser light source 10B is inclined in the −Z direction by an angle ΔθLb in the XZ plane with respect to the predetermined optical axis, and the angles ΔθLa and ΔθLb are equal, then the odd-numbered drawing lines SL1, SL3, and SL5 and the even-numbered drawing lines SL2, SL4, and SL6 will shift by the same amount in the same direction in the Y direction.

[0099] 11 and 12, if the beams LBa and LBb from the respective output ports of the laser light sources 10A and 10B are shifted in parallel with respect to the designed optical axis, the positions of the drawing lines SL1 to SL6 on the sheet substrate P will not fluctuate significantly, but a telecentric error will occur. On the other hand, as explained in Figures 13 and 14, if the beams LBa and LBb from the respective output ports of the laser light sources 10A and 10B are shifted in such a way that they are tilted with respect to the designed optical axis, the positions of the drawing lines SL1 to SL6 on the sheet substrate P will fluctuate in the X direction (sub-scanning direction) or Y direction (main scanning direction) depending on the direction and amount of tilt.

[0100] In this embodiment, the detection unit 34 shown in FIG. 6 can separately measure the relative parallel shift variation and the relative tilt variation between the beam LBa from the exit of the laser light source 10A and the beam LBb from the exit of the laser light source 10B. Therefore, the correction optical systems 11A and 11B including the parallel plates HV1 and HV2 and the prism plates RD1 and RD2 shown in FIG. 9 can be adjusted so as to reduce stitching errors due to positional variations of the drawing lines SL1 to SL6.

[0101] [Variation 1] The magnification of each of the relay imaging system formed by two lenses GL1a and GL2a shown in FIG. 4 and the relay imaging system formed by two lenses GL1b and GL2b can be set to a magnification or reduction other than unity. For example, the magnification of each relay imaging system can be set to a magnification α, so that the images of the exit ports of laser light sources 10A and 10B formed on surfaces Psa and Psb in FIG. 4 are magnified α times. As an example, if the magnification α is set to 4 times (α = 4) and the reduction magnification β of the reduction relay optical system formed by lenses 34A and 34B shown in FIG. 6 is set to 1 / 2 (β = 0.5), the positional displacement of beams MBa and MBb on the imaging surface of image sensor 34C will be twice (= α × β) the error in the parallel shift of beams LBa and LBb at the exit ports of laser light sources 10A and 10B, thereby improving measurement sensitivity.

[0102] Alternatively, the magnification of each of the relay imaging system formed by two lenses GL1a and GL2a and the relay imaging system formed by two lenses GL1b and GL2b may be a reduction magnification γ, so that the images of the exit ports of laser light sources 10A and 10B formed on surfaces Psa and Psb are reduced by γ. As an example, if the reduction magnification γ is set to 1 / 2 (γ=0.5) and the reduction magnification β of the reduction relay optical system formed by lenses 34A and 34B shown in FIG. 6 is also set to 1 / 2 (β=0.5), the proportionality constant between the amount of change in the tilt of beams LBa and LBb at the exit ports of laser light sources 10A and 10B and the amount of displacement of the positions of the spot light beams MBa and MBb on the imaging surface of imaging element 34G becomes larger than when the relay imaging system formed by lenses GL1a and GL2a and the relay imaging system formed by lenses GL1b and GL2b are set to the same magnification, thereby improving measurement sensitivity.

[0103] In view of the above, the reduction relay optical system formed by lenses 34A and 34B in detection unit 34 may be configured to have a magnification of 1x (reduction magnification β=1), and the magnification of each of the relay imaging system formed by lenses GL1a and GL2a and the relay imaging system formed by lenses GL1b and GL2b may be configured to be switchable so that it is an enlargement magnification when measuring the parallel shift error of beams LBa and LBb emitted from each of laser light sources 10A and 10B with image sensor 34C, and is a reduction magnification when measuring the inclination error of beams LBa and LBb with image sensor 34G.

[0104] Second Embodiment By using the beam switching unit BDU with the configuration shown in Figures 4 to 10 almost as it is and slightly modifying the internal configuration of each of the drawing units MU1 to MU6 shown in Figure 2, it becomes possible to achieve multi-spotting, in which two or three spot lights are simultaneously projected from each of the drawing units MU1 to MU6 onto the sheet substrate P while scanning along each of the drawing lines SL1 to SL6.

[0105] 15A to 15C are diagrams illustrating the incidence state and diffraction efficiency of beam LBb from laser light source 10B incident on acousto-optical modulator AM6, the first stage of the beam switching unit BDU. FIG. 15A shows acousto-optical modulator AM6 viewed in the XZ plane of the Cartesian coordinate system XYZ. Beam LBb from laser light source 10B typically enters acousto-optical modulator AM6 coaxially with optical axis AXb. Because acousto-optical modulator AM6 is positioned to satisfy the Bragg diffraction conditions for incident beam LBb (parallel light beam), beam LB6 as a first-order diffracted beam is deflected at a predetermined diffraction angle in the −Z direction with respect to optical axis AXb. If beam LBb incident on acousto-optical modulator AM6 is tilted by an incident angle θz from its coaxial state with optical axis AXb in a plane parallel to the XZ plane, beam LB6 as a first-order diffracted beam will also be tilted in a plane parallel to the XZ plane in proportion to the incident angle θz.

[0106] 15B is a view of the acousto-optical modulator AM6 viewed in the XY plane of the Cartesian coordinate system XYZ, in which the beam LBb from the laser light source 10B is normally incident on the acousto-optical modulator AM6 coaxially with the optical axis AXb. The acousto-optical modulator AM6 is positioned to satisfy the Bragg diffraction conditions for the incident beam LBb (parallel light beam), so the beam LB6 as a first-order diffracted beam travels parallel to the optical axis AXb when viewed in the XY plane. If the beam LBb incident on the acousto-optical modulator AM6 is tilted by an angle of incidence θy from a state coaxial with the optical axis AXb in a plane parallel to the XY plane (in the non-diffraction direction), the beam LB6 as a first-order diffracted beam maintains the angle of incidence θy in the XY plane, while being deflected at a predetermined diffraction angle in the −Z direction in the XZ plane.

[0107] Comparing the case of Figure 15A with the case of Figure 15B, the intensity of beam LB6 (first-order diffracted beam) from acousto-optical modulation element AM6 was measured as shown in Figure 15C. Figure 15C is a graph schematically showing the change in the intensity of beam LB6 (first-order diffracted beam) with respect to the incident angle θz in the diffraction direction of beam LBb incident on acousto-optical modulation element AM6 and the incident angle θy in the non-diffracted direction. The horizontal axis of Figure 15C represents the incident angle θz and the incident angle θy, the origin (0) represents the state in which beam LBb is incident on acousto-optical modulation element AM6 coaxially with the optical axis AXb, and the vertical axis of Figure 15C represents the diffraction efficiency (%), which is the ratio of the intensity of beam LB6 (first-order diffracted beam) to the intensity of incident beam LBb.

[0108] The characteristic CCz in the graph shown in Figure 15C represents the change in diffraction efficiency in the case of Figure 15A, and the characteristic CCy in the graph shown in Figure 15C represents the change in diffraction efficiency in the case of Figure 15B. As is clear from these characteristics CCz and CCy, when the incident beam LBb is tilted at an angle θz from the normal state toward the diffraction direction of the acousto-optic modulation element AM6 as shown in Figure 15A, the condition for Bragg diffraction is not met, and the diffraction efficiency drops sharply with changes in the incident angle θz. In contrast, when the incident beam LBb is tilted at an angle θy from the normal state toward a direction (non-diffraction direction) perpendicular to the diffraction direction of the acousto-optic modulation element AM6 as shown in Figure 15B, the drop in diffraction efficiency with changes in the incident angle θy is gradual.

[0109] Therefore, in this embodiment, when viewed in the XY plane, two beams (parallel light beams) are made to intersect and enter the first-stage acousto-optic modulation element AM6 at an angle of ±θya with respect to the optical axis AXb as shown in Fig. 15C. In this case, the two beams entering the first-stage acousto-optic modulation element AM6 are supplied from different laser light sources (e.g., 10B1 and 10B2) having the same configuration.

[0110] 16 is a perspective view showing the state of two beams on the optical path from the first-stage acousto-optic modulation element AM6 to the epi-illumination mirror IM6 of the beam switching unit BDU according to the second embodiment. The Cartesian coordinate system XYZ in FIG. 16 is set to be the same as the Cartesian coordinate system XYZ in FIG. 4, and the two beams (both parallel beams) incident on the first-stage acousto-optic modulation element AM6 are beams LSa and LSb supplied from different laser light sources (e.g., 10B1 and 10B2) of the same configuration. As described in FIG. 15C, the angle of incidence of beam LSa on acousto-optic modulation element AM6 is set to an angle −θya from the optical axis AXb in the XY plane, and the angle of incidence of beam LSb on acousto-optic modulation element AM6 is set to an angle +θya from the optical axis AXb in the XY plane. Each of the two beams LSa and LSb becomes a parallel beam with a diameter of approximately 1 mm to 0.5 mm, intersects and travels within the crystal of the acousto-optic modulation element AM6, and then travels straight as zeroth-order diffracted beams LSa0 and LSb0 (dotted lines), which enter the focusing lens 16B.

[0111] When acousto-optical modulator AM6 is in the On state, acousto-optical modulator AM6 generates a first-order diffracted beam LSa1 (solid line) deflected at a predetermined diffraction angle in the −Z direction with respect to the zeroth-order diffracted beam LSa0, and a first-order diffracted beam LSb1 (solid line) deflected at a predetermined diffraction angle in the −Z direction with respect to the zeroth-order diffracted beam LSb0, and these beams are incident on condenser lens 16B. The zeroth-order diffracted beams LSa0 and LSb0 emerging from condenser lens 16B pass through the space above epi-illumination mirror IM6 and are incident on the subsequent collimator lens 17B, shifted in parallel by the same distance in the +Y and −Y directions from the optical axis AXb in a plane parallel to the XY plane.

[0112] On the other hand, when viewed in the XZ plane, the first-order diffracted beam LSa1 emerging from the condenser lens 16B is shifted in the -Z direction from the zeroth-order diffracted beam LSa0, travels parallel to the optical axis AXb, and is reflected in the -Z direction by the downward-facing, 45-degree reflecting surface of the epi-illumination mirror IM6. Similarly, when viewed in the XZ plane, the first-order diffracted beam LSb1 emerging from the condenser lens 16B is shifted in the -Z direction from the zeroth-order diffracted beam LSb0, travels parallel to the optical axis AXb, and is reflected in the -Z direction by the downward-facing, 45-degree reflecting surface of the epi-illumination mirror IM6. Here, the first-order diffracted beam LSa1 reflected in the -Z direction by the reflecting surface of the epi-illumination mirror IM6 is referred to as beam LB6a, and the first-order diffracted beam LSb1 is referred to as beam LB6b.

[0113] As explained above with reference to Figures 11 to 14, if the optical axis AX6 is an axis that passes through the center of the reflecting surface of the epi-mirror IM6 in the Y direction and is perpendicular to the optical axis AXb, then beam LB6a advances while shifted in parallel from optical axis AX6 by a predetermined distance ΔYL in the +Y direction, and beam LB6b advances while shifted in parallel from optical axis AX6 by a predetermined distance ΔYL in the -Y direction. Because the reflecting surface of the epi-mirror IM6 is set at the rear focal point of the condenser lens 16B, each of beams LB6a and LB6b forms a beam waist at the reflecting surface of the epi-mirror IM6 and then advances as a diverging beam. The diameter of the beam waist at the reflecting surface of the epi-mirror IM6 is approximately several tens of microns.

[0114] 4, the second-stage acousto-optic modulation element AM4 is set optically conjugate with the first-stage acousto-optic modulation element AM6 by a 1x relay system consisting of a condenser lens 16B and a collimator lens 17B, and the third-stage acousto-optic modulation element AM2 is set optically conjugate with the second-stage acousto-optic modulation element AM4 by a 1x relay system consisting of a condenser lens 20B and a collimator lens 21B. For this reason, when the acousto-optic modulation element AM6 shown in FIG. 16 is in the off state, first-order diffracted beams LSa1 and LSb1 are not generated, and the beams LSa and LSb that are incident on the acousto-optic modulation element AM6 pass directly through the condenser lens 16B and the collimator lens 17B along the optical paths of the zeroth-order diffracted beams LSa0 and LSb0 and are incident on the second-stage acousto-optic modulation element AM4. In this case, the incident angles of the two beams LSa and LSb on the acousto-optic modulation element AM4 (tilt angles with respect to the optical axis AXb in the XY plane) are the same as the incident angles of the beams LSa and LSb incident on the acousto-optic modulation element AM6. Similarly, when the first-stage acousto-optic modulation element AM6 and the second-stage acousto-optic modulation element AM4 are both in the Off state, the incident angles of the two beams LSa and LSb on the third-stage acousto-optic modulation element AM2 (tilt angles with respect to the optical axis AXb in the XY plane) are the same as the incident angles of the beams LSa and LSb incident on the acousto-optic modulation element AM6.

[0115] The two beams LB6a and LB6b reflected in the -Z direction by the incident mirror IM6 in Figure 16 are each divergent light beams, but their chief rays (central rays) are parallel to the optical axis AX6. If the two beams LB6a and LB6b are spaced apart in the Y direction at the position of the incident mirror IM6, the two beams LB6a and LB6b will be incident on the mirror M10 of the imaging unit MU1 (MU2 to MU6 are also the same) shown in Figure 2, and the two spot lights ultimately projected onto the sheet substrate P will be shifted from each other by a certain distance in the Y (Yt) direction, i.e., the main scanning direction (a distance obtained by reducing the distance 2ΔYL in Figure 16 by a predetermined ratio). This is consistent with the state described above in Figure 14.

[0116] Therefore, in this embodiment, the optical configuration of each of the light path adjustment units BV1 to BV6 shown in FIG. 1 is slightly modified. In the first embodiment, each of the light path adjustment units BV1 to BV6 was composed of multiple reflecting mirrors, a relay optical system with multiple lenses, a tiltable quartz parallel plate, and the like. In this embodiment, a rotator mechanism is provided within the light path adjustment unit BV6 (as are BV1 to BV5) that rotates the two beams LB6a and LB6b incident on the first mirror M10 of the imaging unit MU6 (as are MU1 to MU5) by 90 degrees around the optical axis. In the following description, any of the imaging units MU1 to MU6 will be referred to as imaging unit MUn (n = 1 to 6), and the two beams incident on each imaging unit MUn will also be referred to as beams LBna and LBnb (n = 1 to 6).

[0117] Figure 17 is an exaggerated view of the state of two beams LB6a and LB6b passing through the optical path from the epi-illumination mirror IM6 shown in Figure 16 through the optical path adjustment unit BV6 (see Figure 1) to the lens LGa (see Figure 2) in the imaging unit MU6. The Cartesian coordinate system XYZ in Figure 17 and the Cartesian coordinate system XtYtZt in the imaging unit MU6 are set the same as in Figures 1 to 6 and 16, respectively. Figure 17 is a light path diagram seen in the XZ plane, and the extension line of the optical axis AX6 passing through the center of the epi-illumination mirror IM6 in the Y direction is set coaxially with the rotation axis LE6 (corresponding to LE1 in Figure 2), which serves as the rotation center when the entire imaging unit MU6 is slightly rotated.

[0118] The two beams LB6a and LB6b (divergent light) reflected in the -Z direction by the epi-illumination mirror IM6 appear to overlap in the direction perpendicular to the paper surface (Y direction) in Figure 17, but they are positioned symmetrically in the Y direction across the optical axis AX6 and enter the optical path adjustment unit BV6. The optical path adjustment unit BV6 is composed of mirrors M30, M31, and M32 tilted at 45 degrees in the XZ plane, mirror M33 positioned at an angle of 45° + θu / 2 (θu is shown in Figure 1) with respect to the YZ plane, lenses Gv1, Gv2, and Gv3, and an image rotator (hereinafter simply referred to as the rotator) IRD. As disclosed in, for example, JP-A-8-334698 and WO 2018 / 164087, the rotator IRD is composed of two reflecting surfaces that intersect with the optical axis AX6 and are arranged in a mountain shape in the direction of the optical axis, and a third reflecting surface that is arranged away from the mountain-shaped ridge of the two reflecting surfaces and parallel to the optical axis AX6.

[0119] In Figure 17, two beams LB6a and LB6b from the epi-mirror IM6 are reflected by mirror M30 at right angles in the +X direction and enter lens Gv1. The front focal point of lens Gv1 is set to be the reflecting surface of epi-mirror IM6, i.e., the position of the beam waists of beams LB6a and LB6b. Therefore, beams LB6a and LB6b passing through lens Gv1 are both converted into parallel beams, but when viewed in the XY plane, they intersect at plane Pva, the rear focal point of lens Gv1. Plane Pva is set optically conjugate with acousto-optic modulator AM6 by a relay system consisting of condenser lens 16B and lens Gv1 shown in Figure 16. The two beams LB6a and LB6b intersecting at plane Pva are reflected by mirror M31 at right angles in the -Z direction and enter lens Gv2, whose front focal point is set at plane Pva.

[0120] After passing through lens Gv2, the two beams LB6a and LB6b are each converted into a converging beam and again travel along optical paths parallel to the optical axis AX6, sandwiching the optical axis AX6, before being reflected by mirror M32 at a right angle in the -X direction. Each of the two beams LB6a and LB6b converges to form a beam waist at plane Pvb, the rear focal point of lens Gv2, and then diverges as it enters rotator IRD. Plane Pvb is conjugate with the reflecting surface of mirror IM6 (or its immediate vicinity) via the relay system formed by lenses Gv1 and Gv2. Therefore, on plane Pvb, which is parallel to the YZ plane, the beam spots (beam waist positions) of beams LB6a and LB6b are positioned symmetrically in the Y direction across optical axis AX6.

[0121] The rotator IRD is rotated around the optical axis AX6 so that the third reflecting surface, which is parallel to the optical axis AX6, is tilted 45° with respect to both the XY and XZ planes. As a result, the two beams LB6a and LB6b incident on the rotator IRD exit the rotator IRD and enter the lens Gv3 in a state where they are rotated 90° around the optical axis AX6 as a whole. The two beams LB6a and LB6b emerging from the rotator IRD are both diverging beams, but their chief rays (central rays) are parallel to the optical axis AX6. Furthermore, the front focus of the lens Gv3, including the optical path length of the rotator IRD, is set at the position of the plane Pvb. Therefore, the beams LB6a and LB6b that pass through the lens Gv3 are converted into parallel beams and tilted so that they intersect with each other in the XZ plane.

[0122] Beams LB6a and LB6b are reflected in the -Z direction by mirror M33, which is tilted at an angle of (45° + θu / 2) with respect to the YZ plane, intersect at a plane Pvc that is tilted at an angle θu with respect to the XY plane, and then enter mirror M10 in imaging unit MU6. Beams LB6a and LB6b (both parallel beams) reflected in the -Xt direction by mirror M10 each enter the first-stage lens LGa that constitutes the beam expander shown in Figure 2, tilted in the XtZt plane with respect to the optical axis (optical axis AX6) of lens LGa. Because the front focal point of lens LGa is set at the plane Pvc, beam spots (beam waists) SP6a and SP6b of beams LB6a and LB6b are formed on plane OPa, the rear focal point of lens LGa, at symmetrical positions in the Zt direction across the optical axis.

[0123] The plane OPa is conjugate with the imaging plane (surface of the sheet substrate P) that is ultimately set by the fθ lens system FT and second cylindrical lens CYb in the imaging unit MU6. Therefore, the spot light of each of the two beams LB6a and LB6b projected from the imaging unit MU6 onto the sheet substrate P is condensed at a predetermined interval in the Xt direction (sub-scanning direction). For each of the other imaging units MU1 to MU5, by providing optical path adjustment units BV1 to BV5 including a rotator IRD, as in FIG. 17, it is possible to condense the spot light of each of the two beams LBna and LBnb at a predetermined interval in the Xt direction (sub-scanning direction). Therefore, in this embodiment, a total of four laser light sources are installed: two laser light sources 10B1 and 10B2 for the two beams LBna and LBnb supplied to the even-numbered drawing units MU6, MU4, and MU2, respectively, and two laser light sources 10A1 and 10A2 for the two beams LBna and LBnb supplied to the odd-numbered drawing units MU1, MU3, and MU5, respectively.

[0124] In this embodiment, as shown in FIGS. 4 and 6, laser light source 10A1 and laser light source 10B1 can be arranged point-symmetrically with respect to center point PG, and laser light source 10A2 and laser light source 10B2 can be arranged point-symmetrically with respect to center point PG. Furthermore, triangular mirror 33 and detection unit 34 shown in FIG. 6 may be divided into two sets: one set that receives the beams from laser light source 10A1 and laser light source 10B1, and the other set that receives the beams from laser light source 10A2 and laser light source 10B2. Furthermore, when the spot light SPa represents the spot light on sheet substrate P produced by beams LBna (n = 1 to 6) projected onto sheet substrate P from each of imaging units MUn (n = 1 to 6), and the spot light SPb represents the spot light SPb on sheet substrate P, it is desirable to be able to accurately monitor (measure) the positional fluctuations of the two spot light SPa and SPb on sheet substrate P.

[0125] FIG. 18 is a diagram showing an example of optical paths for guiding beams from four laser light sources 10A1, 10A2, 10B1, and 10B2 applied to this embodiment to first-stage acousto-optic modulation elements AM6 and AM1. The Cartesian coordinate system XYZ in FIG. 18 is set the same as in FIG. 4, and the same components and arrangement as in FIG. 4 are assigned the same reference numerals. Laser light sources 10A1 and 10A2, which are arranged side by side in the Y direction, emit beams LSA1 and LSA2 (parallel beams), respectively, in the +X direction. Beam LSA1 is reflected obliquely in the +Y direction by mirror M40a, and beam LSA2 is reflected obliquely in the -Y direction by mirror M40b. Beam LSA1 reflected by mirror M40a and beam LSA2 reflected by mirror M40b are reflected by each of the two reflecting surfaces of V-shaped mirror M40c so as to form a predetermined intersecting angle in the XY plane.

[0126] As explained above in FIG. 15C, each of the two beams LSA1 and LSA2 reflected by the V-shaped mirror M40c enters the prism block VP1, which adjusts the incident angle ±θya to a value suitable for the first-stage acousto-optical modulation element AM1. The two beams LSA1 and LSA2 (parallel light beams) emerging from the prism block VP1 travel at a predetermined inclination with respect to the optical axis AXa in the XY plane and intersect within the crystal of the first-stage acousto-optical modulation element AM1. Furthermore, in the optical path from the prism block VP1 to the first-stage acousto-optical modulation element AM1, there are provided a half-wave plate WP1 that can rotate about the optical axis AXa, and a polarizing beam splitter PBS1 that reflects a portion of each of the two beams LSA1 and LSA2 in the +Y direction as a measurement beam MBa'. The ratio between the transmission intensity of the two beams LSA1 and LSA2 at the polarizing beam splitter PBS1 and the reflection intensity split off as the measurement beam MBa' can be adjusted by the rotation angle of the half-wave plate WP1.

[0127] Measurement beam MBa' (including a partial intensity of each of the two beams LSA1 and LSA2) is received by triangular mirror 33 and detection unit 34 described above in FIG. 6, and the relative fluctuation of beams LSA1 and LSA2 is measured. To correct fluctuations in the position and tilt of beams LSA1 and LSA2 from the respective emission ports of laser light sources 10A1 and 10A2, correction optical systems such as those shown in FIG. 9 are provided in the optical path between laser light source 10A1 and mirror M40a and in the optical path between laser light source 10A2 and mirror M40b. Although not shown in FIG. 18, a relay optical system including lenses GL1a, GL2a, etc., shown in FIG. 4, may be provided in the optical path from polarizing beam splitter PBS1 to detection unit 34, if necessary.

[0128] Two beams LSB1 and LSB2 supplied to each of the even-numbered imaging units MU6, MU4, and MU2 are emitted from laser light sources 10B1 and 10B2, respectively. Beam LSB1 from laser light source 10B1 travels through mirror M42a, which is similar to mirror M40a, V-shaped mirror M42c, which is similar to V-shaped mirror M40c, and prism block VP2, which is similar to prism block VP1, and is incident on the first-stage acousto-optic modulation element AM6, tilted at a predetermined angle with respect to optical axis AXb in the XY plane. Furthermore, beam LSB2 from laser light source 10B2 travels through mirror M42b, which is similar to mirror M40b, V-shaped mirror M42c, and prism block VP2, and is incident on the first-stage acousto-optic modulation element AM6, tilted at a predetermined angle with respect to optical axis AXb in the XY plane.

[0129] A half-wave plate WP2 and a polarizing beam splitter PBS2 are disposed in the optical path between the prism block VP2 and the acousto-optic modulation element AM6, and a measurement beam MBb' formed by splitting a portion of each of the two beams LSB1 and LSB2 is received by a detection unit 34 via a triangular mirror 33. In this embodiment as well, the entire optical arrangement of the laser light sources 10A1 and 10A2, mirrors M40a and M40b, V-shaped mirror M40c, prism block VP1, half-wave plate WP1, and polarizing beam splitter PBS1 and the entire optical arrangement of the laser light sources 10B1 and 10B2, mirrors M42a and M42b, V-shaped mirror M42c, prism block VP2, half-wave plate WP2, and polarizing beam splitter PBS2 are in a point-symmetric relationship with respect to the center point PG in the XY plane.

[0130] In the above Figure 18, the optical system consisting of mirrors M40a, M40b, V-shaped mirror M40c, and prism block VP1 (or the optical system consisting of mirrors M42a, M42b, V-shaped mirror M42c, and prism block VP2) functions as a synthesis optical system that synthesizes two beams LSA1, LSA2 (or LSB1, LSB2) within the first-stage acousto-optical modulation element AM1 (or AM6) so that they intersect at a predetermined crossing angle (for example, 0°<θy≦1°) with respect to the non-diffraction direction (Y direction).

[0131] 4, the optical path arrangement of the measurement beam MBa' (MBa) from the laser light source 10A to the triangular mirror 33 (or the detection unit 34) and the optical path arrangement of the measurement beam MBb' (MBb) from the laser light source 10B to the triangular mirror 33 (or the detection unit 34) do not necessarily have to be in a point-symmetric relationship rotated 180° about the center point PG, but may be in a line-symmetric relationship in the XY plane. Specifically, they may be in line-symmetric arrangement with respect to a center line parallel to the X axis orthogonal to the normal to the center point PG, or may be in line-symmetric arrangement with respect to a center line parallel to the Y axis orthogonal to the normal to the center point PG.

[0132] FIG. 19 is a diagram schematically illustrating the scanning behavior of two spot beams SPa and SPb projected onto a sheet substrate P in the second embodiment of the configuration shown in FIGS. 16 to 18. Here, the diagram representatively illustrates the main scanning behavior of the spot beams SPa and SPb by the two beams LB6a and LB6b projected from the imaging unit MU6 shown in FIG. 17. When the two beams LB6a and LB6b enter the imaging unit MU6 as shown in FIG. 17, the two spot beams SPa and SPb are positioned on the sheet substrate P separated by a center-to-center distance ΔXS in the Xt direction (sub-scanning direction) as shown in FIG. 19. Here, the effective diameter (1 / e of the peak intensity value) of each of the spot beams SPa and SPb is 2 If the diameter at which the intensity is 1 / 2 of the normalized intensity (or the diameter at which the intensity is 1 / 2) is φs (μm), the center-to-center spacing ΔXS is set to satisfy the relationship ΔXS ≥ 1.5 φs, for example. However, to minimize the effects of optical aberrations, the center-to-center spacing ΔXS can be reduced to ΔXS = 0.5 φs (where the spots SPa and SPb overlap at half the diameter φs). Conversely, if the center-to-center spacing ΔXS is 10 times or more the diameter φs, the effects of aberrations will cause distortion in the shapes of the spots SPa and SPb, and the telecentricity error will increase. Therefore, when α is an integer greater than or equal to 1, the center-to-center spacing ΔXS can be set to the general formula ΔXS ≥ 0.5 α φs (α = 1, 2, 3, etc.).

[0133] When the laser light sources 10B1 and 10B2 are fiber amplifier laser light sources with an oscillation frequency of 400 MHz, the spot light SPa and SPb are each pulsed in the Yt direction (Y direction), which is the main scanning direction, in response to a clock signal CLK with a period of 2.5 nS. Therefore, the spot light SPa and SPb are set to overlap with each other at 1 / 2 of the diameter φs in the Yt direction. That is, the rotation speed of the polygon mirror PM is set so that the scanning speed Vss of the spot light SPa and SPb in the Yt direction is Vss = (φs / 2 μm) / 2.5 nS. Similarly, the movement speed of the sheet substrate P in the Xt direction is also set so that the spot light SPa or SPb overlaps with each other at 1 / 2 of the diameter φs in the Xt direction. Therefore, when only a single spot light SP is projected from the drawing unit MU6, the movement speed of the sheet substrate P is set so that the spot light SP scans the sheet substrate P to form drawing lines SL6a, SL6a', SL6b, SL6b', ..., SL6f, SL6f', ... arranged at a pitch of ΔXT (=φs / 2) in the Xt direction, as shown in Figure 19.

[0134] On the other hand, when two spot lights SPa and SPb are arranged side by side in the Xt direction as shown in Figure 19, the spot lights SPa and SPb are scanned simultaneously in the main scanning direction as the polygon mirror PM rotates. Therefore, as shown on the right side of Figure 19, in order to make the spot lights ultimately struck on the sheet substrate P overlap at an interval of φ / 2 in the Xt direction, the sheet substrate P can be moved so that a single drawing line formed by the simultaneous scanning of the two spot lights SPa and SPb becomes SL6a, SL6b, SL6c, .... Therefore, whether there is a single spot light SP or two spot lights SPa and SPb, if the rotation speed of the polygon mirror PM is not changed (the scanning speed Vss is kept the same), when two spot lights SPa and SPb are used as shown in Figure 19, the movement speed of the sheet substrate P in the Xt direction can be doubled, that is, the exposure processing time for the sheet substrate P can be halved.

[0135] In this modification, two beams LSA1 and LSA2 (or LSB1 and LSB2) from two laser light sources 10A1 and 10A2 (or 10B1 and 10B2) are passed through the acousto-optical modulator AMn in the beam switching unit BDU at a predetermined intersecting angle. However, three laser light sources may be provided so that three beams intersect at the acousto-optical modulator AMn. The third beam is set coaxially with the optical axis AXa or AXb passing through the first-stage acousto-optical modulator AM1 or AM6 in the beam switching unit BDU. In this case, the third spot light projected onto the sheet substrate P from each of the imaging units MUn is set between the two spot lights SPa and SPb shown in FIG. 19. It is preferable that the three spot lights are set so as not to overlap one another in the Xt direction in FIG. 19, but the center-to-center spacing ΔXS of each of the three spot lights in the Xt direction may be set to ΔXS = 0.5 φs (a state in which the spot lights aligned in the Xt direction overlap each other by 1 / 2 of the diameter φs).

[0136] [Variation 2] FIG. 20 is a perspective view showing a modified example of the imaging unit MUn (MU1 to MU6) shown in FIG. 2. The configuration of the imaging unit MUn in FIG. 20 is disclosed in, for example, International Publication No. 2019 / 082850, and will be briefly described below. Components in FIG. 20 that have the same functions as components in FIG. 2 are designated by the same reference numerals. The Cartesian coordinate system XtYtZt is also set in the same way as in FIG. 2. The main differences between this modified example and FIG. 2 are that an imaging system using lenses LGd and LGe is provided in the optical path between the first cylindrical lens CYa and the polygon mirror PM, and three mirrors M14a, M14b, and M14c that fold the optical path are provided in the optical path between the lens LGe and the polygon mirror PM. The imaging system using lenses LGd and LGe establishes an imaging relationship between the rear focal position of the first cylindrical lens CYa and the reflecting surface Rp1 of the polygon mirror PM.

[0137] 16 to 18, in the imaging unit MUn of FIG. 20, two beams LBna and LBnb are supplied from each of the two laser light sources 10A1 and 10A2 (or 10B1 and 10B2) via the incident mirror IMn (n = 1 to 6) of the beam switching unit BDU and the optical path adjustment unit BVn (n = 1 to 6). However, the rotator IRD in the optical path adjustment unit BVn (n = 1 to 6) shown in FIG. 17 is omitted due to the imaging system using lenses LGd and LGe in the imaging unit MUn and the optical path bending using three mirrors M14a, M14b, and M14c. Therefore, in FIG. 17, each of the two beams LB6a and LB6b (parallel light beams) reflected by mirror M10 of the imaging unit MU6 and incident on the lens LGa is tilted symmetrically with respect to the optical axis in a plane that includes the optical axis of the lens LGa and is parallel to the XtYt plane. Therefore, each of the spots (beam waists) SP6a and SP6b formed on the plane OPa shown in FIG. 17 is located on a line that intersects with the optical axis and extends in the Yt direction (Y direction).

[0138] 19, when the imaging unit MUn (n=1 to 6) according to the above-described modification 2 is used, the two spot lights SPa and SPb can be arranged at a constant center-to-center distance ΔXS in the Xt direction, so that the exposure processing time for the sheet substrate P can be reduced to half the exposure processing time using a single spot light SP. Note that the photoelectric sensor DT in FIG. 20 is installed at a position optically conjugate with the two spot lights SPa and SPb projected onto the sheet substrate P, and therefore is configured with a two-part photoelectric element that separately receives the reflected light from the sheet substrate P due to the projection of the spot light SPa and the reflected light from the sheet substrate P due to the projection of the spot light SPb.

[0139] [Variation 3] In the first and second embodiments described above, a spot-scanning imaging unit MUn (n = 1 to 6) is used, which uses a polygon mirror PM and an fθ lens system to one-dimensionally scan spot light SP (or SPa, SPb) projected onto the sheet substrate P, which is the surface to be illuminated, while modulating the intensity in response to imaging data. However, the imaging unit MUn may also be configured using a maskless exposure method in which a variable light intensity distribution generated by reflected light from a digital mirror device (DMD) or a spatial light modulator (SLM) is projected onto the sheet substrate P by a projection imaging system.

[0140] In this case, one imaging unit is made up of one DMD (or SLM) and one projection imaging system, and a plurality of such imaging units are arranged in the width direction (Y direction) of the sheet substrate P. When a plurality of laser light sources are used as light source devices that supply exposure beams (illumination light beams for the DMD or SLM) to each of the plurality of imaging units, the fluctuations in the beams emitted from the respective laser light sources can be accurately monitored (measured) using a detection unit 34 as shown in Fig. 6 above.

[0141] [Variation 4] FIG. 21 is a perspective view showing a modified version of the configuration of a portion of the optical path adjusting unit BV6 shown in FIG. 17, with the Cartesian coordinate system XYZ set the same as in FIGS. 17 and 16. As described in FIG. 16, two beams LSa and LSb (parallel light beams) are incident on the acousto-optic modulation element AM6 at a certain crossing angle on either side of the optical axis AXb in the XY plane. When the acousto-optic modulation element AM6 is in the on state, a first-order diffracted beam LSa1 of the beam LSa and a first-order diffracted beam LSb1 of the beam LSb are generated and condensed by the condenser lens 16B to form beam waists at the reflecting surface of the epi-illumination mirror IM6 (inclined 45 degrees from a plane parallel to the XY plane). The two first-order diffracted beams LSa1 and LSb1 reflected in the −Z direction by the reflecting surface of the epi-illumination mirror IM6 are incident on the mirror M30a as beams LB6a and LB6b, respectively.

[0142] The reflecting surface of mirror M30a is tilted 45 degrees from a plane parallel to the XY plane, and beams LB6a and LB6b are reflected in the -X direction. Beams LB6a and LB6b reflected by mirror M30a are reflected in the -Y direction by mirror M30b, which has a reflecting surface tilted 45 degrees from a plane parallel to the XZ plane, and then reflected in the -Z direction by mirror M30c, whose reflecting surface is tilted 45 degrees from a plane parallel to the XY plane. The two beams LB6a and LB6b reflected by mirror M30c are incident on lens Gv1 shown in Figure 17. The chief ray (central ray) of each of beams LB6a and LB6b is parallel to the optical axis AX6 of lens Gv1 on the optical path from epi-illumination mirror IM6 to lens Gv1 via mirrors M30a, M30b, and M30c, and is located symmetrically across the optical axis AX6. The front focal point of the lens Gv1 is set at the position of the reflecting surface of the epi-illumination mirror IM6 via an optical path formed by mirrors M30a, M30b, and M30c.

[0143] As shown in Fig. 21, by providing mirrors M30a, M30b, and M30c after the epi-illumination mirror IM6, the positions at which the central rays of the two beams LB6a and LB6b (divergent light beams) enter the lens Gv1 are separated in the Y direction by ΔYL across the position of the optical axis AX6. The mirrors M30a, M30b, and M30c in Fig. 21 have the same function as the image rotator IRD in the optical path adjustment unit BV6 shown in Fig. 17. Similarly, the two beams LBna and LBnb (n = 1 to 5) reflected in the -Z direction by each of the other epi-illumination mirrors IM1 to IM5 are incident on the lens Gv1 included in each of the optical path adjustment units BVn (n = 1 to 5) via the mirrors M30a, M30b, and M30c in Fig. 21.

[0144] When this modification is used, the optical path adjustment unit BVn (n = 1 to 6) with the rotator IRD removed as shown in Fig. 17 and the imaging unit MUn (n = 1 to 6) as shown in Fig. 20 can perform main scanning along the imaging line SLn (n = 1 to 6) with two spot lights SPa and SPb as shown in Fig. 19 arranged with a center-to-center spacing ΔXS in the Xt direction. According to this modification, the optical path adjustment unit BVn (n = 1 to 6) is configured by the lenses Gv1 to Gv3 and mirrors M31 to M33 shown in Fig. 17, and the mirrors M30a, M30b, and M30c in Fig. 21.

[0145] [Variation 5] As shown in FIGS. 4 and 5, the beam splitters 30A and 30B that split a portion of the light intensity (energy) of the beams LBa and LBb from the laser light sources 10A and 10B, respectively, as the measurement beams MBa and MBb may be polarized beam splitters (corresponding to PBS1 and PBS2 described in FIG. 18). In this case, rotatable half-wave plates (corresponding to WP1 and WP2 described in FIG. 18) are provided between the laser light source 10A and the beam splitter 30A and between the laser light source 10B and the beam splitter 30B shown in FIG. 4 (or FIG. 5). The light intensity ratio between the exposure beam LBa (LBb) transmitted through the beam splitter 30A (30B) and the measurement beam MBa (MBb) reflected by the beam splitter 30A (30B) can be adjusted by changing the rotation angle position of the half-wave plates. Therefore, by individually adjusting the rotation angles of the half-wave plate on the laser light source 10A side and the half-wave plate on the laser light source 10B side, it is possible to adjust the intensity of the spot light SP projected onto the sheet substrate P from each of the odd-numbered drawing units MU1, MU3, and MU5 and the intensity of the spot light SP projected onto the sheet substrate P from each of the even-numbered drawing units MU2, MU4, and MU6 to be the same.

[0146] [Variation 6] 1 to 6, laser light sources (first light source devices) 10A that supply exposure beams LBa are provided for the three odd-numbered imaging units MU1, MU3, and MU5, and laser light sources (second light source devices) 10B that supply exposure beams LBb are provided for the three even-numbered imaging units MU2, MU4, and MU6. However, a similar detection unit 34 can also be provided in a pattern imaging apparatus (exposure apparatus) that performs splice exposure on patterns drawn by two imaging units, and in which one laser light source (light source device) is provided for each of the two imaging units. Furthermore, when four sets of three imaging units to which a beam from one laser light source (light source device) is supplied are provided and splice exposure is performed by a total of 12 imaging units MU1 to MU12, four laser light sources are provided. In this case, the optical paths of measurement beams MBa, MBb, MBc, and MBd generated by splitting beams LBa, LBb, LBc, and LBd from the four laser light sources are set, for example, as shown in FIG.

[0147] FIG. 22 is a schematic diagram showing the optical paths of measurement beams MBa, MBb, MBc, and MBd generated by splitting beams LBa, LBb, LBc, and LBd from four laser light sources 10A, 10B, 10C, and 10D, respectively. In FIG. 22, the X direction of the Cartesian coordinate system XYZ is the sub-scanning direction in which the sheet substrate P moves, and the Y direction is the main-scanning direction of the spot light projected from each of the 12 imaging units MU1 to MU12. In this modification, beam LBa emitted in the +X direction from the exit of laser light source 10A is directed via beam splitters 30A and 12A to pass through odd-numbered acousto-optic modulation elements AM1, AM3, and AM5 in series as shown in FIG. 4, and is then supplied to odd-numbered imaging units MU1, MU3, and MU5. In this modification, laser light source 10B is disposed back-to-back with laser light source 10A in the X direction. The beam LBb emitted in the −X direction from the exit of the laser light source 10B is directed via beam splitters 30B, 12B, etc. to pass in series through the even-numbered acousto-optic modulation elements AM2, AM4, and AM6 as shown in FIG. 4, and is supplied to the even-numbered drawing units MU2, MU4, and MU6.

[0148] 22 represents the center point of point symmetry of the arrangement of the 12 drawing units MU1 to MU12 in the XY plane, similar to the center point PG in Fig. 4. The remaining two laser light sources 10C and 10D are arranged in a point-symmetric relationship with respect to the arrangement of the two laser light sources 10A and 10B rotated 180° around the center point PG. Note that the two laser light sources 10A and 10B and the two laser light sources 10C and 10D are also arranged symmetrically in the XY plane with respect to a center line that passes through the center point PG and is set parallel to the X axis.

[0149] Beam LBc emitted in the +X direction from the emission port of laser light source 10C is directed via beam splitters 30C, 12C, etc. to pass through odd-numbered acousto-optic modulation elements AM11, AM9, AM7 in series, and is supplied to odd-numbered imaging units MU11, MU9, MU7. Beam LBd emitted in the -X direction from the emission port of laser light source 10D, which is arranged back-to-back with laser light source 10C in the X direction, is directed via beam splitters 30D, 12D, etc. to pass through even-numbered acousto-optic modulation elements AM12, AM10, AM8 in series, and is supplied to even-numbered imaging units MU12, MU10, MU8.

[0150] The measurement beam MBa, which is split by beam splitter 30A from beam LBa from laser light source 10A, is directed toward triangular mirror 33' located at center point PG via mirror 31A and a relay optical system (lenses GL1a and GL2a in FIG. 4), not shown. The measurement beam MBb, which is split by beam splitter 30B from beam LBb from laser light source 10B, is directed toward triangular mirror 33' located at center point PG via mirror 31B and a relay optical system (not shown). The measurement beam MBc, which is split by beam splitter 30C from beam LBc from laser light source 10C, is directed toward triangular mirror 33' located at center point PG via mirror 31C and a relay optical system (not shown). Similarly, the measurement beam MBd, which is split by beam splitter 30D from beam LBd from laser light source 10D, is directed toward triangular mirror 33' located at center point PG via mirror 31D and a relay optical system (not shown).

[0151] Figure 23 is a perspective view showing the relative positions of the triangular mirror 33' and the detection unit 34 that constitute the variable optical detection system, and the Cartesian coordinate system XYZ is set as in Figure 22. As shown in Figure 22, the four measurement beams MBa, MBb, MBc, and MBd heading toward the triangular mirror 33' are set so that their optical paths are parallel to the Y axis. The triangular mirror 33' has two reflecting surfaces 33a' and 33b' inclined at 45° from the XY plane so that their ridgelines are parallel to the X axis. The two measurement beams MBa and MBb traveling in the +Y direction are each reflected in the +Z direction by the reflecting surface 33a' of the triangular mirror 33' and enter the lens 34A of the detection unit 34, which is configured in the same way as in Figure 6, parallel to the optical axis AXu. Similarly, the two measurement beams MBc and MBd traveling in the -Y direction are each reflected in the +Z direction by reflecting surface 33b' of triangular mirror 33' and enter lens 34A of detection unit 34 parallel to optical axis AXu.

[0152] 6, the detection unit 34 in FIG. 23 also includes a lens 34B, a beam splitter (half mirror) 34E, a first image sensor 34C, and a second image sensor 34G. When the imaging surface of image sensor 34C is divided into four quadrants, measurement beams MBa, MBb, MBc, and MBd are projected into each quadrant. Furthermore, the four measurement beams MBa, MBb, MBc, and MBd each have a focal point formed at approximately the center of the imaging surface of image sensor 34G. Therefore, when measuring fluctuations using image sensor 34G, image information captured at the timing when one of the four measurement beams MBa, MBb, MBc, and MBd is supplied, i.e., when one of the four laser light sources 10A, 10B, 10C, and 10D is emitting a beam, can be acquired.

[0153] Even in a pattern exposure apparatus using four laser light sources (light source devices) 10A, 10B, 10C, and 10D as in this modification, the arrangement and optical path length of the optical members (mirrors and lenses) forming the optical paths of the measurement beams MBa, MBb, MBc, and MBd from each laser light source to the triangular mirror 33′ (fluctuation detection optical unit) can be set to be the same. Furthermore, as shown in FIG. 22, when viewed in the XY plane, the optical paths of the measurement beams MBa, MBb, MBc, and MBd can be set to be point-symmetric with respect to the center point PG or to be line-symmetric with respect to a line passing through the center point PG and parallel to the Y-axis or X-axis. Therefore, the measurement sensitivity and measurement accuracy when measuring fluctuations of the beams LBa, LBb, LBc, and LBd emitted from each of the four laser light sources 10A, 10B, 10C, and 10D can be set to be the same, so that relative shift fluctuations and tilt fluctuations of the four beams LBa, LBb, LBc, and LBd can be accurately captured.

[0154] Although Figure 22 illustrates a configuration in which one laser light source supplies beams to three drawing units, the configuration is not limited to this, and the pattern exposure device may be configured to have multiple (two or more) laser light sources (light source devices), and each laser light source distributes and supplies the emitted beam to two or more drawing units.

Claims

1. A pattern exposure apparatus, a first light source device that emits a first beam; a second light source device that emits a second beam; a plurality of acousto-optic modulation elements through which the first beam and the second beam pass in series; a plurality of drawing units that use diffracted beams of the first beam and the second beam generated from the plurality of acousto-optic modulation elements as spot beams and draw a pattern on a substrate by one-dimensionally scanning the spot beams; a combining optical system that combines the first beam from the first light source device and the second beam from the second light source device so that the first beam and the second beam intersect and pass through at a predetermined crossing angle within a first-stage acousto-optical modulation element among the plurality of acousto-optical modulation elements; Equipped with a first spot light formed by the first beam and a second spot light formed by the second beam, the first spot light being projected from each of the plurality of drawing units onto different positions on the substrate;

2. 2. The pattern exposure apparatus according to claim 1, a pattern exposure apparatus in which each of the plurality of drawing units has a rotating polygon mirror that deflects the diffracted beam, and the diffracted beam deflected by the rotating polygon mirror is used as the spot light to perform the one-dimensional scanning;

3. 3. The pattern exposure apparatus according to claim 1, a pattern exposure apparatus comprising a relay optical system disposed in an optical path between the plurality of acousto-optical modulation elements so as to make each of the plurality of acousto-optical modulation elements optically conjugate with each other;

4. 3. The pattern exposure apparatus according to claim 1, the combining optical system combines the first beam from the first light source device and the second beam from the second light source device within a first-stage acousto-optical modulation element among the plurality of acousto-optical modulation elements so that the first beam and the second beam intersect and pass at the predetermined crossing angle with respect to a non-diffraction direction orthogonal to a diffraction direction in which the diffracted beam is generated.

5. 3. The pattern exposure apparatus according to claim 1, When the one-dimensional scanning is a main scanning, the spot light is sub-scanned in a direction perpendicular to the main scanning direction, a crossing angle of the first beam and the second beam by the combining optical system is set so that the first spot light by the first beam and the second spot light by the second beam, which are projected onto the substrate from each of the plurality of drawing units, are positioned at a center interval ΔXS in the sub-scanning direction.

6. 6. The pattern exposure apparatus according to claim 5, wherein when an effective diameter of each of the first spot light and the second spot light is φs, the center-to-center distance ΔXS is set to satisfy a relationship of ΔXS≧0.5·α·φs (α is an integer of 1 or greater).

7. 6. The pattern exposure apparatus according to claim 5, The sub-scanning is performed by moving the substrate relative to the plurality of drawing units.

8. 8. The pattern exposure apparatus according to claim 7, A pattern exposure apparatus comprising a rotating drum that moves the substrate in the sub-scanning direction.

Images